Material Information

Alternate Title:
Proceedings of the Soil and Crop Science Society of Florida
Alternate Title:
Soil and Crop Science Society of Florida proceedings
Abbreviated Title:
Proc.- Soil Crop Sci. Soc. Fla.
Soil and Crop Science Society of Florida -- Meeting
Place of Publication:
Hollywood, Fla
Soil and Crop Science Society of Florida
Publication Date:
Volume 51, 1992
Physical Description:
48 volumes : illustrations, portraits ; 23-28 cm


Subjects / Keywords:
Soil science -- Congresses ( lcsh )
Crop science -- Congresses ( lcsh )
Crop science ( fast )
Soil science ( fast )
Conference papers and proceedings. ( fast )
serial ( sobekcm )
conference publication ( marcgt )


Chemical abstracts
Dates or Sequential Designation:
Vol. 16 (1956)-v. 64 (2005).
General Note:
Some issues accompanied by CD-ROM.
Statement of Responsibility:
the Soil and Crop Science Society of Florida.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright, Soil and Crop Science Society of Florida. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
01489546 ( OCLC )
43013238 ( LCCN )
0096-4522 ( ISSN )
020317058 ( Aleph )
S590 .S65 ( lcc )
631.4062759 ( ddc )

Related Items

Preceded by:
Succeeded by:


This item is only available as the following downloads:

Full Text




Logo of the Soil and Crop Science Society of Florida. Al the 40th Annual Business Meeting at 1120 h, 8 Oct. 1980 at the Holiday Inn, Longboat Key, Sarasota, FL, a report was presented by the ad hoc Logo Committee composed of J. J. Street, Chair, R. S. Kalmhacher, and K. H. Quesenberry. All Society members were eligible lo participate in the contest which would result in the selection of a logo design for the SCSSF. The presentation of the winning entry would be made at the 1981 Annual Meeting. David H. Hubbell was announced as the winner of the contest and was awarded a free IO-yr membership in the Society. The design, shown above, depicted the chief interests of the Society: Florida Soils, and Crops. Soils and Crops were given equal weight with Florida. Soils was depicted as the soil texture triangle which shows the percentage of sand, silt, and clay in each of the textural classes, but simplified so as to permit clarity in reduction when printed. Crops was shm,n as a stylized broadleaf plant, including the roots. Florida was shown, minus its keys, with only one physical feature in its interior, Lake Okeechobee. Enclosed within nine rays such as might be envisioned as being made by the sun emitting light (sunburst) behind the symbols are two concentric circles containing the words "SOIL AND CROP SCIENCE SOCIETY" in the top semicircle and "FLORIDA" at the bottom of the lower semicircle. A schematic likeness of the State of Florida occupies 0.5 of the area within the smaller circle, while the symbol for Soils and the symbol for Crops occupy 0.25 of the area each, witl1 the sum of the three parts totalling unity. The first printing of Society stationery following the award on 28 Oct. l 981, and the printed program of the Society since the 1982 meetings, have featured the logo. V. E. Green,Jr., Editor, Volume 45


1991 OFFICERS President: .............. R. S. Kalmbacher Agric. Res. and Ed. Center Route 1, Box 62, Ona, FL 33865 President-Elect: .............. G. C. Smart Entomology-N ematology Dep. Univ. of Florida Gainesville, FL 32611-0630 Secretary-Treasurer: ......................... ............................. C. G. Chambliss Agronomy Dep. Univ. of Florida Gainesville, FL 32611 Directors: N. R. Usherwood (1991) Potash and Phosphate Inst. 2801 Buford Hwy., NE Atlanta, GA 30329 F. M. Rhoads (1992) North Florida Res. and Ed. Center Route 3, Box 4370, Quincy, FL 32351-9529 J. M. Bennett (1993) Agronomy Dep. Univ. of Florida Gainesville, FL 32611-0500 EDITORIAL BOARD Editor: ........................... P. L. Pfahler Agronomy Dep. Univ. of Florida Gainesville, FL 32611-0500 Associate Editors: F. M. Rhoads (Soils) North Fla. Res.and Educ. Center Route 3, Box 4370 Quincy, FL 32351-9529 B. M. McNeal (Soils) Soil and Water Science Dep. Univ. of Florida Gainesville, FL 32611-0510 M. J. Williams (Crops) Subtropical Agric. and Res. Station, USDA P.O. Box 46, Brooksville, FL 32605-0046 Published annually by the Soil and Crop Science Society of Florida. Membership dues including subscription to annual Pro ceedings are $20.00 per year for domestic members and $22.00 per year for foreign members. At least one author of a paper submitted for publication in the Proceedings must be an active or honorary member of the Society except for invitational pa pers. Ordinarily, contributions shall have been presented at annual meetings; excep tions must have approval of the Executive Committee and the Editorial Board. Contributions may be (1) papers on original research or (2) invitational papers of a philosophical or review nature presented before general assemblies or in symposia. A charge of $30.00 per printed page m the Proceedings will be billed to the agency the author represents to help defray printing costs. Members are limited to senior authorship of one volunteer paper per volume; there is no limit for Junior authorships. SOIL AND CROP SCIENCE SOCIETY OF FLORIDA VOLUME 51 CONTENTS 1992 Dedication Honorary Life Members SOILS SECTION Effect of Nursery Fertitilization on Fruiting Response of Strawberry -E. E. Albregts, C. M. Howard and C. K. Chandler Mineral Content of Soils and Forage from Horse Farms in Marion County, Florida. I. Lead V vii and Cadmium E. A. Ott, S. Sund/off, and M. Tooker 3 Mineral Content of Soils and Forage from Horse Farms in Marion County, Florida. II. Minerals Required by Horses E. A. Ott, S. Sund/off, and M. Tooker 7 Polymer Conditioners for Florida Soils D. Z. Haman and M.E.B.Joyner 14 Copper Toxicity and Phosphorous Concentration in 'Florida 502' Oats -F. M. Rhoads, R. D. Barnett, and S. M. Olson 18 Mycorrhizal Amelioration of the Detrimental Effect of Biodune on Plant Growth -T. Aziz and D. M. Sylvia 20 Calibration Parameters for Radar Reinfall Estimation S. F. Shih 23 SPOT Satellite Data and GIS for Well Permitting and Management -Y. R. Tan and S. F. Shih 29 Using Remote Sensing and Geographical Information System in Water-Quality Assessment -B. E. Myhre, S. F. Shih, and D. A. Still 34 Phosphorous and Zinc Influence on Bermudagrass Growth J. B. Sartain 39 Managing Plant-Parasitic Nematodes in Crop Sequences R. N. McSorley and R. N. Gallaher 42 Landsat and SPOT Imagery Classification for Land Use Change Analysis in Lee County, Florida J. D. Jordan and S. F. Shih 45 Three Potential Amendments for Better Fertilization Utilization in Sandy Soils -T. L. Yuan 49 Determination of Nitrate Levels in Water Samples Using a Microplate Reader -C. D. Stanley and]. B.Jones 56 A Comparison of Water Quality Information Obtained from Depth-Integrated Versus Depth-Specific Groundwater Monitoring Devices W. D. Graham and D. Downey 58 Controlled-Release Fertilizer Use on Young 'Hamlin' Orange Trees -T. A. Obreza and R. E. Rouse 64 CROP SECTION Lateral Root Distribution Patterns in Stylosanthes guianensis Seedlings -John B. Brolmann and Peter J. Stoffel/a 69 Evaluation of Macroptilium atropurpureum (DC) Urb. Germplasm for Reaction to Foliar Diseases -Ronald A. Sonoda, A. E. Kretschmer,Jr., and T. C. Wilson 71 Evaluation of a Photovoltaic System for Supplying Water to Beef Cattle -]. J. Mullahey, Y. J. Tasi, and D. J. Pitts 75 Energycane Response to Harvest Management -P. Mislevy, M. G. Adjei, G. M. Prine, and F. G. Martin 79 Effect of Dolomite Rate and Placement on Leucaena Forage and Seed Yield -R. S. Kalmbacher, J. E. Rechcigl, and F. G. Martin 85 Reliability of Methods for Assessing Leaf Blotch Diseases in Wheat -]. Cybulska-Augustyniak, F. M. Shokes, R. D. Barnett and A. R. Soffes 90 Mycoflora of Seed from Advanced Breeding Lines in North Florida -Z. Weber, D. A. Berger, F. M. Shokes, R. D. Barnett, and D. L. Wright 96 Hessian Fly Control in Florida Wheat with Systemic Insecticides -J. B. Hartman, R. D. Barnett, A. R. Soffes, ahd R. K. Sprenkel 99 Occurrence and Control of Target Spot in Tobacco in Florida caused by Rhizoctonia solani Kuhn -T. A. Kucharek, R. Tervola, and]. Washington 103 Dry Matter Production and Forage Quality of Line 8400 Sty lo, Alyceclover, and Hairy Indigo -M. J. Williams, C. G. Chambliss, J.B. Brolmann, and S. L. Sumner 106 Assessment of Productivity and Persistence of Selected Tropical Forage Legumes -A. E. Kretschmer, Jr., B. J. Bowman, T. C. Wilson and G. H. Snyder 109 Evaluation of "Tropical" Corn as a Feedstuff for Swine -R. 0. Myer, D. L. Wright, L. E. Anderson, J. H. Brendemuhl, G. E. Combs, and D. W. Gorbet 116 Leucaena: A Forage and Energy Crop for the Lower South, USA -T. V. Cunilio and G. M. Prine 120 Effect of Water Deficits on Growth and Nitrogen Fixation of Hairy Indigo -U. Winzer, S. L. Albrecht, and J. M. Bennett 125 Some Factors Affecting Response of 'Florida 77' Alfalfa to Acid-Soil Amendments -B. W. Matthews, R. E. Joost, and L. E. Sollenberger 130 SOCIETY AFFAIRS Minutes Business Meeting 25 September 1991 Board of Directors Meeting 25 September 1991 R. S. Kalmbacher 136 R. S. Kalmbacher 137


Board of Directors Meeting 17 January l 991 Awards Luncheon Graduate Student Paper Contest Necrology and Resolultion of Sympathy Financial Report Committees 1992 Editorial Report Membership Lists Regular Members 1991 Emeritus Members 1991 Honorary Life Members 1991 Sustaining Members 1991 Suscribing Members 1991 Past and Present Officers Life Members Proceedings Dedication Recipients Soil and Crop Science Society of Florida Invitational Reviewers R. S. Kalmbacher 138 139 139 140 141 142 142 142 142 148 149 149 149 152 153 154 155


DEDICATION OF THE FIFTIETH-FIRST PROCEEDINGS SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Dr. William G. Blue DR. WILLIAM G. BLUE William G. Blue was born at Poplar Bluff, Missouri, and attended Butler County and Poplar Bluff public schools. He attended the University of Missouri from 1941 through 1950 with a 3-year interruption for military service during World War II. He served as reconnaissance sergeant in the 78th Infantry Division in Europe, and was awarded the Purple Heart. He received his B.S. in agriculture in 1947, his M.A. in soil science in 1948, and his Ph.D. in soil science in 1950. Dr. Blue began his employment in research with the Soil Science Department, University of Florida, in 1950 with emphasis on soil chemistry and soil fer tility problems related to forage production. He continued to emphasize this area of research until his retirement. He became involved in graduate student supervision in the early 1960s and served as committee chair for 17 Ph.D. and 15 M.S. students. He as a supervisory committee member for an additional 80 students, primarily in agronomy and animal husbandry. In 1968, he began teaching the graduate soil fertility course, instructing approximately 400 students in this course through his retirement in 1989. Dr. Blue also served as assistant chair of the Soil Sci-ence Department for several years, and as acting chair in 1982 and 1983. A major accomplishment of Dr. Blue's research was the characterization of the action of anhydrous ammonia in coarse-textured soils. He and Charles F. Eno did many studies with anhydrous ammonia related to soil chemistry and microbiology, and plant responses. They determined that anhydrous ammonia is a partial sterilant in soils, and they were the first to treat hay with this N source to increase effec tive protein concentration for cattle and to reduce spoilage in partially cured hay. This technique has become widely used. Another important long-term study showed that factors normally thought to reduce N volatilization and leaching had little effect on apparent N uptake by forage plants, particularly Paspalum notatum Flugge, but that continuity of fertilization management had major effects. Studies with poorly drained, coarse-textured soils showed that apparent N uptake was 40 to 50% during the first 4 or 5 years, but subsequently increased to the 70 to 90% range. This result occurred with no depletion of soil N. Dr. Blue also made important contributions to knowledge of


the action of K and P in coarse-textured and highly weathered soils. These efforts have resulted in approximately 200 publications. During his career, Dr. Blue participated in a number of overseas development projects involving the University of Florida. He lived in Costa Rica with his family from I 958 through mid-1960, working as a pasture agronomist. In 1985 Bill and his wife, Bernice, took a 2-year assignment in Cameroon, where he worked as a research and extension advisor on a university-development project. Dr. Blue also has had several other short-term overseas assignments during the course of his career. An active member of the American Society of Agronomy (ASA), Soil Science Society of America (SSSA), and Soil and Crop Science Society of Florida for many years, Dr. Blue published extensively in these societies' journals. He has served as Associate Editor of the SSSA Journal, and is a Fellow of both ASA and SSSA. Dr. Blue is also a member of Sigma Xi, Alpha Zeta, Gamma Sigma delta, and Alpha Gamma Rho. He has been named Professor Emeritus at the University of Florida. Dr. Blue was President of the Soil and Crop Science Society of Florida in 1961, and has served the Society in other capacities as well, including five years as editor of the Soil and Crop Science Society of Florida Proceedings. In recognition of his contributions to science, to Florida's agricultural industry, to world agriculture, and to the Soil and Crop Science Society of Florida, the members of the Society take great pleasure in dedicating the Proceedings of this Fifty-first Annual Meeting to Dr. William G. Blue.


HONORARY LIFE MEMBER Dr. Victor W. Carlisle DR. VICTOR W. CARLISLE Dr. Victor W. Carlisle received his B.S. from the University of Florida in 194 7 after a three-year interruption for military service during World War II. Dr. Carlisle began employment with the Soil Science Department in 1948 as a soil surveyor, but he returned to the university classroom and received his M.S. in soil science in 1954. Working full-time as a teaching assistant and later as an instructor, he completed his Ph.D. in soil genesis in 1960. During his career Dr. Carlisle has taught the introductory course in soil science, soil fertility and soil genesis and classification. Dr. Carlisle's research was primarily in soil characterization, classification, and survey. Dr. Carlisle has been author or coauthor of more than 120 scientific publications, including five book chapters. Dr. Carlisle has served and leadership capacities for numerous professional organizations, including the Soil and Water Conservation Society -Florida Chapter, the Soil and Crop Science Society of Florida, and the Florida Association of Professional Soil Class ifiers. He was honored with the USDA Certificate of Appreciation for dedicated service and outstanding leadership in the Cooperative Soil Survey of Florida; named Distinguished Soil Classifier by the Florida As sociation of Professional Soil Classifiers; and received the Professional Achievement Award from the Soil and Water Conservation Society. Dr. Carlisle is a member of ASA, SSSA, International Society of Soil Science, Soil and Crop Science Society of Florida, Florida Academy of Sciences, Florida Association of Professional Soil Classifiers, Florida Defender of the Environment, Soil and Water Conservation Society, Gamma Sigma Delta, and Sigma Xi. In recognition of his outstanding contributions to his profession and of his past service to the Society, the members of the Soil and Crop Science Society of Florida take great pleasure in electing Dr. Victor W. Carlisle to Honorary Lifetime Membership.


HONORARY LIFE MEMBER Dr. Earl S. Horner DR. EARL S. HORNER Dr. Earl S. Horner was born and spent his early life on a farm in Colorado. He graduated with a B.S. degree in agronomy from Washington State University in 1940 and a M.S. degree in agronomy from Michigan State University in 1942. He served in the U.S. Army during World War II (1942-46). He then attended graduate school at Cornell University, where he studied genetics and plant breeding and received his Ph.D. in June 1950. Following graduation, he was employed by the Agronomy Department at the University of Florida as assistant professor. He was promoted to associate professor in 1956, professor in 1964, and in recent years served as assistant chairman of the Agronomy Department. Dr. Horner's research involved both corn and legume breeding. He developed and released 'Florida 66' and 'Florida 77' alfalfa cultivars. He did much of the early selection and development on the white clover cultivar, 'Osceola'. Dr. Horner is a member of Alpha Zeta, Sigma Xi, Phi Kappa Phi, and Gamma Sigma Delta in 1973-74. He is a long-time member of the Soil and Crop Sci-ence Society of Florida (which he served as President in 1968), American Society of Agronomy, and Crop Science Society of America, and is a Fellow in ASA and CSSA. He was an associate editor of Agronomy Journal (1968-73) and Crop Science (1974-76), a technical editor of Crop Science (1982-84), and Editor of Crop Science and Editor-in-Chief of CSSA (1984-86). He was editor of the Soil and Crop Science Society of Florida Proceedings from 1973-83 and again in 1987. Dr. Horner chaired the committees of 18 graduate students and served on the committees of many other students. He wrote numerous research publications on plant breeding methods and taught a course of advanced plant breeding. In recognition of his outstanding contributions to agricultural research and education and of his past service to the Society, the members of the Soil and Crop Science Society of Florida take pleasure in electing Early S. Horner to honorary lifetime membership.


PROCEEDINGS, VOLUME 51, 1992 SOILS SECTION Effect of Nursery Fertilization on Fruiting Response of Strawberry E. E. Albregts*, C. M. Howard, and C. K. Chandler ABSTRACT Strawberry (Fragaria ananassa, Duch.) was grown with four NPK fertility levels in the nursery. Transplants from this nursery were evaluated in the fruit production field using the annual hill cultural system. The nursery fertility levels in mmho cm conductivity were 0.35, 0.70, 1.05, and 1.40 the first season and 0.70, 1.05, 1.40, and 1.75 the second season. These levels of fertility were determined immediately after fertilization which occurred every 7 to 14 days. Early marketable yields averaged over 2 clones the first season and 4 the second were highest with plants which received the lower nursery fertility levels while the average fruit weight increased with increasing nursery fertility level the first season. The percent marketable fruit and the total mar ketable fruit yields were not affected by nursery fertility treatments either season. Strawberry (Fragaria ananassa, Duch.) has acclimated to most areas of the world because of its genetic diversity (Darrow, 1966). The fruiting response of strawberry not only depends on the genetic background of the particular strawberry clone, but also on the local environment and how the plant is handled prior to and after it is transplanted into the fruiting field (Darrow, 1966). Daylength, temperature, time of planting, soil fertility, plant size, and plant storage prior to transplanting are some of the vari ables that affect the fruiting response of strawberry (Albregts and Howard, 1984; Darrow, 1966). Previous work has suggested that fertility level in the nursery may have an effect on the fruiting response of strawberry (Albregts and Howard, 1985). Normally, nursery fertility is kept high to enhance plant production. However, anthracnose infection is intensified by moderate to high levels of N (Smith, 1991 ). With some cultivars, such as 'Dover', soil fertility may be kept moderate to low during the latter part of the s~ason to reduce plant crowding and increase plant Size. MATERIALS AND METHODS Strawberry was grown during the 1989-90 and 1990-91 winter seasons at AREC-Dover on a Seffner fine sand (sandy siliceous, hyperthermic, Quartzipsammentic Haplumbrept) using the annual hill cultural system. Strawberry plants for this study were grown under four fertility regimes. These regimes were 0.35, 0. 70, 1.05, and 1 .40 mmho cm1 the first E. E. Albregts, C. M.Howard, and C. K. Chandler, Agric. Res. and Education Ctr., IFAS, Univ. of Florida, 13138 Lewis Gallagher Road, Dover, FL 33527. Florida Agric. Exp. Stn. journal Series no. N-00449. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51: 1-2 ( 1992) season and 0.70, 1.05, 1.40, and 1.75 mmho cm1 conductivity the second season of the saturated paste ex tracts taken immediately after each fertilization in the nursery during plant production. The saturated extract procedure was used to determine the fertility status of the nursery for two reasons. The procedure is simple, inexpensive, and a quick test that the grower can perform. In addition, on the soils that strawberry is grown the soluble salts measured are mostly those of N and K. These are the two nutrients which can readily leach from the unmulched beds. The nursery fertilizer was a 10.0-4.9-9.3 NPK mixture derived from NH4N03 triple super phosphate, and KCI. Fertilizer rate at each fertilization was 40, 80, 120, 160, and 200 kg ha1 of the 10.0-4.9-9.3 fer tilizer for the 0.35, 0. 70, 1.05, 1.40, and 1. 75 treatments. Plants in the nursery were fertilized from early June to plant harvest in mid-October each season on a 7 to 14 d schedule depending on rainfall. Soil soluble salts varied considerably between fertili zations in the nursery because of rainfall and irrigation. The clones cv. 'Dover' and breeding line Florida 79-1126 were evaluated the first season and Florida 83-37 and Florida 84-1932 were added the second season. Transplants were set in the fruiting field on 18 Oct 1989 and 23 Oct 1990 with 18 and 14 transplants replicate1 the first and second seasons, respec tively. The statistical design was a randomized complete block, having treatments in a factorial arrangement with four and five replicates treatment1 the first and second seasons, respectively. Fruit production beds were fumigated at 392 kg ha1 of the bed area with a mixture of 98% methyl bromide and 2% chloropicrin and mulched with black polyethylene. Fertilizer rates in kg ha1 were 224-28-206-33 (N-P-K-Mg) plus 29 kg ha1 of a micronutrient mix containing 3% Cu and B, 7% Mn and Zn, and 9% Fe. One fourth of the fer tilizer was mixed in the bed, and the remainder was banded in the bed center two inches deep. Overhead sprinkler irrigation was provided for soil moisture, plant establishment, and freeze protection. Soil moisture was maintained between -5 and -20 chars. Fruit was harvested twice weekly from January through April, graded, counted, and weighed. Marketable fruit included those free of rot, not misshapen, and weighing 10 g or more. Plants were visually evaluated for size and color several times each season. RESULTS AND DISCUSSION Marketable fruit yields, average of two clones, for January, February, and March, gave quadratic re-


2 Son, AND CROP SCIENCE SOCIETY OF FLORIDA Table 1. Effect of nursery fertility levels on fruiting response of two strawberry clones during 1989-90 season. Marketable fruit yield Avg. fruit wtt Treatments January February March Total Januarl Seasonal mmho/cm' ----------------------------------------------Mg ha-1-------------------------------------------------------------g fruit I --------------0.35 0.70 1.05 1.40 Linear Quadratic 6.1 7.0 6.3 5.2 NS 10.9 7.6 9.4 8.2 12.0 7.0 11.6 6.3 NS NS ** 25.1 21.3 15.7 25.0 23.4 15.7 25.8 22.0 15.6 23.4 20.2 15.3 NS NS NS NS ** NS tData are average of two clones except for the January average fruit weight, which is only for the fruit of Florida 79-1126. NS,*,**Nonsignificant or significant at P = 0.05 or 0.01, respectively. sponses to nursery fertility levels during the first season (Table I). The nursery fertilizer levels giving the highest yield response were the 0. 70 mmho cm I con ductivity level in January and March and the 1.05 mmho cm1 conductivity level in February. These fruiting differences were a response to time of fruiting and not differences in yield since seasonal fruit yields were similar. During the second season, the January marketable yield of the four clones decreased as the nursery fertilizer rate increased (Table 2). Only the average fruit weight of the Florida 791 126 clone for January the first season was different because of nursery fertility treatments. During the second season, the average fruit weight of the four clones increased with higher nursery fertility levels. In an earlier study (Albregts and Howard, 1985 ), the average fruit weight was higher with lower nursery fertility. The percent marketable fruit in this study were similar for all months and for the season as related to nursery treatments. The visual, subjective plant size for Dover the first season from plant establishment through early January varied with nursery fertility. For the 0.35, 0. 70, 1.05, and 1.40 mmho cm1 conductivity treatments the relative plant size for 16 Oct 1989 were 6.38, 6.13, 7.13, and 7.12; for 15 Dec 1989 8.25, 7.88, 9.38, and 9.88; and for 9 Jan 1990 9.00, 8.63, 9.15, and 9.88, respectively. The largest plants were rated 10.0 at each date. Plant size varied linearly because of treatment for all evaluation dates at the 5% level and variation on the last date was also quadratic. During the second season plant size differences were not significant once the plants became established in the fruiting field. Plant color differences were noted during plant harvest from the nursery each season. The plants in 0.35 and 0.70 mmho cm1 conductivity treatments were lighter green, especially the 0.35 treatment. Nevertheless, plant color became similar among all nursery treatments within 2 to 3 wk after setting in fruiting field. SUMMARY The 1.40 and the 1.75 mmt:;:) cm1 conducti,ity treatments gave reduced early fruit yields, but had Table 2. Effect of nursery fertility levels on fruiting response of four strawberry clones during 1990-91 season. Marketable fruit Mkt Avg. fruit yieldt fruitt wt.t Treatments January Total seasonal seasonal mmhocm-1 ---------Mg ha '---------% g fruit I 0.70 9.6 25.4 70.3 15.40 1.05 9.6 23.6 68.4 15.44 1.40 8.9 24.3 70.3 15.38 1.75 8.3 24.2 69.2 15.77 Linear NS NS Quadratic NS NS NS NS tData presented are average of four clones. NS, *Nonsignificant or significant at P = 0.05, respectively. no effect on total fruit yields. The early yield differences were probably related to later fruiting of these two treatments. Strawberry has two to four fruiting cycles a season, and these two treatments were in a later cycle than the other t.reatments. The early fruiting of some of the treatments can result in higher prices for the fruit during most years. Since early fruit yield differences were small, the price advantage would also be small. In addition, total fruit yields were similar both seasons, and thus nursery fertility levels probably would not significantly affect economic returns from the fruit production field. In the summer nursery, plant production was economically reduced with nursery fertilization treatments below 0.70 mmho cm 1. REFERENCES Albregts, E. E. and C. M. Howard. 1984. Strawberry production in Florida. Fla. Agric. Exp. Stn. Res. Bui. 841. Albregts, E. E. and C. M. Howard. I 985. Short-term cold storage and soil fertility during plant and fruit production on growth and fruiting response of strawberry. HortScience 20:411-413. Darrow, G. W. 1966. The strawberrv. Holt, Rinehart and Winston. NY., NY 447 p. Smith, B. J. 1991. Strawberrv anthracnose. Proc. North Am. Strawberry Growers Assn. Inc. Annual Meeting 79-86.


PROCEEDINGS, VOLUME 51, 1992 3 Mineral Content of Soils and Forage from Horse Farms in Marion County, Florida I. Lead and Cadmium E. A. Ott,* S. Sundlof, and M. Tooker ABSTRACT Most of the documented cases of Pb and Cd poisoning of horses are due to environmental pollution. Concern for the possibility of industrial pollution of the horse farms of Marion County, Florida has made it appropriate to determine the Pb and Cd concentrations in soil and forage in that area. This study was designed to provide a measure of past pollution and provide a base line for measuring future pollution. Soil (0 to 7.6 cm) and bahiagrass ( Paspalum notatum Flugge) samples were collected from two pastures on each of 3 7 farms in Marion County, Florida. Farms were selected to represent fixed distances of l.0 (Group A), 2.0 (Group B), 4.0 (Group C) and 8.0 (Group D) km from an industrial site 0.6 km east of Lowell, FL. Group E farms were selected along State Road 329 between Lowell and a dump site 1.6 km NW of Fairfield, Florida. Group F farms were selected at random within Marion County but 8.0 km or more from the Lowell site. Preselected farms that were found to be adjacent to Interstate 75 were also sampled along the highway right-of-way (Group G). Samples were analyzed for Pb and Cd. Mean Pb concentration in bahiagrass was 1.8 .2 mg kg-1 dry matter (DM) with a range of 0.31 to 5.91 mg kg-1 DM. Lead concentrations in groups A, B, C, D were not different ( P > .05). Group E was higher in Pb than group F ( P < .05). Mean Cd concentration in bahiagrass was 0.17 0.02 mg kg-1 DM with a range of 0.0 to 0.52 mg kg-1 DM. Group E was higher in Cd than groups D and F ( P < .05). Group B was higher in Cd than group F ( P < .05). Mean Pb concentration of soil was 12.5 0.6 mg kg-' with a range of 5.8 to 24.1 mg kg-' and no differences were detected between the groups ( P > .05). Correlation between soil and forage Pb ( r2 = -0.022) and Cd ( r2 = -0.006) was very low. Results of this study indicate that forage and soil Pb concentrations are below concentrations documented to cause intoxication in horses. There was no relationship between the proximity to the industrial site and the Pb and Cd concentrations in soil and forage. A modest increase in forage Pb was detected in pastures along the highway between the industrial site and a dumping site used to dispose of plant waste. Lead is a natural, but mmor, element found in soil, water and air and is grouped under the term mineral in the title of this paper, due to its importance to the well being of animals (NRC, 1980). Natural Pb concentrations in soil vary from 2 to 200 mg kg-' (average 16 mg kg-1), exclusive of areas near Pb mines or man-made concentrations such as smelters, industrial facilities and roadways (NRC, 1972). Natural concentrations in lakes and rivers are between I and IO ,g L-1 Lead can be present in the air as both Pb dust and aerosols. Both of these forms can be removed from the air by precipitation. Soils remote from man-made Pb sources receive, on the average, 1 g Pb cm-2 year' from precipitation and 0.2 ,g Pb cm-2 yea1-1 from dust fall, which adds about 0.04 mg E. A. Ott and '.1. Tooker, Animal Science lkp .. Univ. of Florida, Gainesville, FL 32611-0910 and S. Sundlof, Physiological Sciences, College of Veterinary Medicine, Univ. of Florida, Caines ville, FL ,l2611-0691. Florida Agric. Exp. Stn. Journal Series no. R-02101. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Prot. 51 :37 ( 1992) kg-' to the soil surface (0-15 cm) annually (TerHarr et al., 1967). Lead smelters and other industrial emissions have resulted in local areas with Pb concentrations in forages as high as 325 to 450 mg kg-' (NRC, 1972). Lead also accumulates along highways due to the burning of Pb containing fuels. Numerous studies of Pb deposition along highways have been conducted. In one such study, Pb content decreased in soil and forage with distance from the highway. Lead content of soil also decreased with depth in the soil pedon. At a typ ical location, Pb concentrations in grass were 68.2, 47.5 and 26.3 mg kg-' at 8, 16, and 32 m from the highway respectively. At these same locations, soil Pb concentrations, extracted by the HCL method, were 522, 378 and 164 mg kg-' at a depth of 5 cm, 460, 260 and 108 mg kg-' at a depth of 5-10 cm and 416, I 04 and 69 mg kg-' at a depth of 10-15 cm from the surface (Lagerwerff and Specht, 1970). Lead concentration in grass near major highways could be high enough to cause intoxication of horses when consumed for extended periods of time. Most horses exposed to these sources will consume other feeds and pasture grasses which are remote from the highway and which contain lower concentrations of Pb so that the total Pb intake is expected to be within tolerable levels. Lead content of plants can be influenced by soil Pb and Pb present in rainfall, ground water and air. Plant Pb content is not very closely related to Pb content of soil because apparently much of the Pb in soil is not available to the plant. Soluble Pb is probably a better indicator of Pb availability to plants. Soils with low organic matter and low pH are likely to have a higher portion of soluble Pb than soils with high organic matter and high pH (NRC, 1972). Lead avail ability to the plant appears to decrease with time suggesting that Pb binds to other soil constituents making it less soluble. About half of the Pb present in samples of grass grown along highways can be removed by washing, indicating that Pb on the surface of the leaves may be of as much importance to grazing animals as the Pb in the plant itself. Forage grown under experimental conditions on soils with Pb concentrations of 700 to 3000 mg kg-' dry matter (DM), only accumulated 15 mg Pb kg-1 (Mueller and Stanley, 1970). It can be concluded that forage plants containing higher than 15 mg Pb kg-1 accumulated the additional Pb from airborne or waterborne sources. Lead toxicirv in horses occurs most frequently in areas immediately surrounding Pb and zinc smelters and in animals consuming feed produced in those areas or contaminated with other sources of Pb (Schmitt et al., 1971). The horse has an ability to ex(:rete Pb in the urine and at low Pb intakes will excrete


4 Son, Al'\D CROP SCIENCE SOCIETY OF FLORIDA Pb as rapidly as it is taken in, If Pb intake exceeds the animal's excretion capacity, it will accumulate in the tissues. \\'hen Pb concentrations exceed the animal's tolerance, Pb poisoning develops. Aronson (1972) suggested that horses may be more susceptible to poisoning bv Pb than cattle since horses have been poisoned in areas where cattle were unaffected. Toxicity in horses was documented in animals consuming forage containing 80 mg Pb kg-1 (Hammond and Aronson, 1964). This would provide a daily intake of 1.7 mg Pb kg-I body weight. Feeding young growing horses diets containing 30 mg Pb kg-1 for 105 days resulted in no adverse effect on growth of the animals and no observable signs of Pb toxicity but tissue Pb concentrations did increase (Willoughby ct al., 1972). However, tissue Pb concentrations in these animals were much lower than those reported for animals exhibiting signs of Pb toxicosis. Cadmium is a cumulative toxin that interferes with Zn and Cu metabolism via competition for metalloenzymes. Natural concentrations of Cd in soil, water and air are quite low except in areas immediately surrounding industrial facilities such as Zn smelters and plating operations. Blood, liver and kidney Cd concentrations increase with Cd intake. Domestic animals consuming diets with less than 1.0 mg kg-1 will generally have liver and kidney concentrations less than 1 mg kg-I fresh tissue (NRC, 1980). Susceptibility to Cd toxicity is apparently related to the Cu, Zn and Fe intakes of the animal. Deficiencies of these minerals make the animal more susceptible to Cd poisoning. Natural Cd concentrations in forages are generally less than 0.5 mg kg-I DM but are higher in forages fertilized by urban sewage sludge. Cadmium concentrations in the diet of 20 mg kg-' and higher are detrimental to livestock. Maximum tolerable dietary concentration of Cd for horses is estimated to be 0.5 mg kg-1 DM (NRC, 1980). MATERIALS AND METHODS Concentric circles with radii of 1.0 (Group A), 2.0 (Group B), 4.0 (Group C) and 8.0 (Group D) km were inscribed on a Marion County, Florida map with their centers at an industrial site 0.6 km east of Lowell, Florida. The center of the circles is on or near MidFlorida Mining's kitty litter manufacturing and soil decontamination plant. Horse farms were selected on or near each circle to provide a uniform distribution of sampling sites around each circle. A fifth group of farms (Group E) was selected along State Road 329 (SR329) between Lowell and a dumping site used to dispose of plant wastes 1.6 km NW of Fairfield, Florida. Pastures adjacent to the highway were sampled. A sixth group of farms (Group F) was selected at random within Marion County but outside the 8 km circle to represent farms 8.0 to 25.0 km from Lowell. Each selected farm was notified by mail of the interest in sampling their soil and forage and was provided a form granting approval and requesting information on pasture mandgemp;~t and fertiliza tion. Thirty seven farms were sampled in this study. Soil and forage samples were collected from two pastures on each farm. Sampling sites in each pasture were selected by walking a pattern over the entire field and stopping every 10 to 15 paces to collect a sample. Soil samples were collected with a stainless steel, curved trowel. At each sampling site, thatch, if present, was scraped aside, a uniform sample of soil removed to a depth of 7.6 cm and placed in a plastic bucket. At least 20 samples were collected from each pasture. The composite sample, 1.4 to 2.3 kg of soil, was bagged for transportation to the laboratory. Bahiagrass (Paspalum notatum Flugge) samples, which represented the forage being consumed by the ani mals, were collected by clipping the grass at a height of 2.5 cm above the soil using the same selection pattern noted above. The samples were bagged and transported to the laboratory. A third soil and forage sample was collected from the three farms adjacent to Interstate 75 (175) (Group G). These samples were collected 6.0 to 7 .5 rn inside the pasture along the right-of-way fence parallel to the highway to determine the impact of automobile emissions on soil and forage mineral concentrations. All samples were collected between 25 Apr and 31 May 1990. Each pasture sampled in this study was located on a soil map (U.S. Dep. of Agric., 1979) and the predominant soil map units recorded. Soil samples were allowed to air dry in their bag and were then sieved through a 2 mm screen to remove stones and large pieces of organic matter. Soil samples were mixed and 2 g subsamples extracted according to Environmental Protection Agency method 3050 (EPA, 1987). This procedure digests the organic component with concentrated HNO3 and H202 followed by extraction with concentrated HCl. The extract was analyzed for Pb and Cd by inductively coupled argon plasma spectroscopy (ICAP). Forage samples were dried at G0C for 72 h, ground in a Wiley mill through a 2.0 mm stainless steel screen and stored in Nalgene vials. Forage samples were ashed at 500C for 20 h, digested in cone. HNO3 and analyzed for Pb and Cd by ICAP. Duplicate analyses were run on all samples and blank, NIST standards, soil standards, spiked samples and sensitivity analyses were run to insure accuracy of the analyses. The data were analyzed according to the procedure for General Linear Models and Duncan's Multiple Range Test (SAS Institute Inc., 1985). Due to differences in the number of replications of soil types the modification proposed by Kramer ( 1956) was used to test the effect of soil type on Pb and Cd content of soil. RESULTS The analytical procedure employed in this study resulted in Pb recovery of 113% in the soil standard and 105% in the forage standard. Ninety three percent of the Pb added to soil samples was recovered. These values indicate the inherent variability of this type of analyses.


PROCEEDINGS, VOLUME 51, 1992 5 Table l. Lead and Cd concentrations in soil and forages from selected Marion County, Florida farms. Farm No. No. Soil' Fora e group farms pastures Pb Cd Pb Cd --------------------------------mg kg-' --------------------------------A 4 8 mean 13.6 1.3 a* 0.90 0.13 a 2.56 0.43 ab 0.22 0.06 abc range 6.5 to 23.5 0.31 to 1.79 0.51 to S.73 0 to 0.52 B 6 12 mean 14.1 0.7 a l.25.17a 2.27 0.4 7 ah 0.23 Cl.05 ah range 9.5 to 19.0 0.62 to 3.04 0.46 to 5.58 0.05 to 0.50 C 8 l fi mean 12.8 0.4a l.14.16a 2.32 0.39 ab 0.19 0.04 abc range 8.5 to 16.4 0 to 3.33 0.52 to 5.88 0 to 0.50 D 7 14 n1ean 13.1 I. I a 0.95 0.13 a 1.32 0.31 ab 0.12 0.04 be range 5.8 to 24.1 0to2.51 0.31 to 5.33 0 to 0.41 E 3 6 mean 10.2 0.5 a 1.08 0.14 a 3.08 0.32 a 0.30 0.03 a range 7.8 to 12.3 0.50 to 1.80 1.4 7 to 4.40 0.20 to 0.41 F 9 l(j mean 11.0 0.5a 0.94.!Sa 0.82 0.26 b 0.07 0.0] C range 5.9 to 17.0 0to4.12 0.35 to 1.62 0to0.14 Total 37 74 mean 12.5 0.6 1.0 0.1 1.8 0.20 0.17 0.02 range 5.8 to 24.1 0 to 4.1 0.31 to 5.91 0 to 0.52 *Means in the same column with the same letter are not significantly different (Duncan's Multiple Range Test, P = 0.05). 'Soil samples collected from surface to a depth of 7.6 cm. Lead None of the forage samples collected in this study contained concentrations of Pb near the 80 mg kg-1 documented to cause toxicity in horses (Hammond and Aronson, 1965) or the 30 mg kg-1 shown to result in Pb accumulation in growing horses (Willoughby et al., 1972). Mean Pb concentration in forage was 1.8 0.2 mg kg-' D1\1 with a range of 0.31 to 5.91 mg kg-1 DM. Groups A, B, C and D were not different (Table 1). Values from farms along SR329 (Group E) were higher than those for group F (P < .05) suggesting that fuel emissions or dust loss from vehicles may be causing modest elevation of forage Pb along the highway between the industrial site and a dumping site used to dispose of plant wastes. Mean soil Pb concentration for the 37 farms sampled was 12.5 .6 mg kg-1 in the surface 7.6 cm with a range of 5.8 to 24.1 mg kg-'. The mean is below the 16 mg kg-' reported as the mean soil Pb concentration for the United States. Considerable variation was found within each group and no differences were detected among groups (P > 0.05). Correlation between soil and forage Pb concentrations was very low (r2 = -0.022). Ten of the 11 farms that had at least one pasture containing 15 or more mg Pb kg-' soil were within 8 km of Lowell; eight were north and west of Lowell and two were southwest of Lowell. This distribution would not be compatible with prevailing wind patterns for the area suggesting that emissions from the industrial site at Lowell is not the primary source of the soil Pb on farms sampled in this study. The three soil and forage samples collected immediately adjacent to 175 were compared to the other samples collected on the same farms to minimize other sources of variations (Table 2). Although no statistically significant differences were detected between the three sample groups (P > 0.05), the samples collected next to the highway had numerically higher soil Pb concentrations than the other samples. No effect on forage Pb was detected. Results suggest that highway emissions prior to the reduction in the use of Pb containing fuels may have caused increased soil Pb as described previously (NRC, 1972), but current emissions and existing soil Pb are apparently not causing elevated forage Pb. Cadmium Cadmium concentrations of both soil and forage were near the lower detection limit of the analytical procedure. The Cd concentrations in both soil and forage were too low to present any health problems for animals. Mean soil Cd concentration was 1.0 0.1 mg kg-1 with a range of 0 to 4.1 mg kg-' (Table Table 2. Lead and Cd concentrations mean in soil and forage from farms along 175. Pasture not next to highway Pasture next to highway Sample parallel to right-of-wav' 11 3 3 3 Soil' Fora re Lead Cadmium Lead Cadmium ---------------mg kg-' ------------15.6.2 l.01.29 0.7!i.l9 0.10.06 16.8.2 1.12.22 0.60.07 0.08.02 25.2 5.2 1.40 .25 0.69 .12 0.10 .03 'Soil samples collected from surface to a depth of 7.6 cm. 'Sample Group G.


6 SOIL AND CROP SCI F'.\JCE SOCIETY OF FLORIDA 1). No differences were detected for soil Cd between any of the groups (Table I). Mean forage Cd was 0.17 0.02 mg kg-I DM with a range of Oto 0.52 mg kg-I DM. Group E had higher forage Cd than D and F (P < 0.05) and group B was higher than group F (P < 0.05). Proximity to 175 had no effect on soil and forage Cd concentrations (Table 2). There was no correlation between soil Cd and forage Cd in this study (r" = -0.006). Soil Type Thirteen soil types were identified on the 37 farms (74 pastures). Distribution of the soil types are shown in Table 3. Kendrick loamy sand (KE) predominated in 35% of the pastures sampled and in over half of the pastures in groups A, B, and C. Arrendondo sand (AR) predominated in 24% of the pastures. Other soil types each represented less than 10% of the total. Four types Apopka sand (AP), Gainesville loamy sand (GA), Lochloosa fine sand (LO) and Micanopy fine sand (MC) were represented by only one sample. Soil Pb was influenced by soil type (P < 0.0002). Soil type GA (24 mg kg-I) was higher in Pb than the other soil types (Table 3) except Zuber loamy sand (ZU) (16 mg kg I), Hague sand (HA) (JG mg kg-I), Flemington loamy sand (FM) ( 15. l mg kg-I) and LO ( 14.6 mg kg-I). Soil types ZU, HA, FM and LO all had higher soil Pb concentration than CA (7.6 mg kg-I). No other differences were detected. Soil Cd was higher in FM (1.9 mg kg-I) than CA (0.1 mg kg:-'). No other differences were detected. Interpretat10n of the relationship between soil type and mineral concentrations must consider the limited number of samples available for many of the soil types. DISCUSSION Lead and Cd concentrations in soil and forage are the result of natural concentrations of these elements in the soil, deposition from non-local airborne sources, local emission from industrial and highway sources and perhaps fertilization programs (NRC, 1972, 1980). This study was designed to determine whether local emission from an industrial site in Marion County, Florida was impacting on Pb and Cd concentrations in soil and forage on horse farms in the vicinity of the plant and whether the concentration of these elements was high enough to cause detrimental effects on animals grazing those pastures. Although horses are less susceptible to Pb toxicity clue to Pb acetate ingestion than cattle (Dollahite et al., 1978), horses appear to be more susceptible to smelter emissions (Aronson, 1972). This may be due to the difference in grazing habits between the two species. Cattle generally graze tall grass while horses prefer short grass and are more likely to pull grass and consume roots and soil. Lead concentrations in both soil and bahiagrass analyzed in this study are considerably below the 80 mg Pb kg-1 reported to be toxic in grazing horses (Hammond and Aronson, 1964) and the 30 mg Pb kg-I that caused elevated tissue Pb in growing horses without causing signs of Pb toxicosis (Willoughby et al., 1972). The results suggest that, at the present time, Pb and Cd concentrations in soil and forage on horse farms around the industrial site near Lowell, Florida are not a health hazard to horses grazing those pastures and recent emissions do not appear to have had a major impact on the Pb and Cd present in the environment of farms in the immediate vicinity of the plant. Since no baseline values exist for the farms sampled in this study prior to initiation of plant operations, these data will provide a reference for future studies. CONCLUSIONS Lead and Cd concentrations in forage samples collected from 37 farms in Marion County, Florida were well below those considered to be toxic to graz-Table 3. Number of pastures in each group and soil type and soil Pb and Cd concentrations.'' Soil type A AP Apopka sand AR Arredondo sand BC Blichton sand CA Candler sand FM Flemington loamy sand GA Gainesville loamy sand HA Hague sand KA Kanapaha fine sand KE Kendrick loamy sand 5 LO Lochloosa fine sand MC Micanopy fine sand SP Sparr fine sand I zc Zuber loamy sand I Total no. 8 'Soil Survey of Marion County Area, Florida, 1979. 'Mean SE. Grou B C D E 5 7 2 2 I 2 3 7 9 2 I 4 12 16 14 6 All F farms Pb Cd --------mg kg-' --------I 1 8.1 abc* 0.5 ab 4 18 10.6 .6bc 0.8 .2 ab 2 4 I 1.0 .5 be 0.9 .3ab 3 3 7.6 l.0c 0. .I b 4 6 15.1 l.8ab l.9.6a l 24.6 a 1.5 ab 2 16.0 3.2 ab 0.9 .6 a 3 9.5 1.4 be 1.0 .2ab 2 26 13.1 0.7 be 1.1 .2 ab I 1 14.6 abc l.0ab I 9.0 be 0.9 ab 3 13.0 l.7bc 0.9 I ab 5 16.1 !..~ab t.5 .4 ab 18 74 *Means in the same column followed by the same letters arc not statistically different (Duncan's Multiple Range Test, P = (l.05).


PROCEEDINGS, VOLUME 51, 1992 7 ing horses. Lead concentrations in the top 7.6 cm of soil from these farms was variable (5.8 to 24.1 mg kg-1 ) but even the highest values detected would not likely cause grazing animals to ingest toxic amounts of Pb. There was no linear effect of distance from an industrial site on soil or forage Pb. Soil Pb was numerically higher on farms within 8.0 km of Lowell than farms throughout the remainder of the county but this difference appears to be due primarily to differences in soil type. Correlation between soil and forage Pb and Cd were low indicating that a major portion of the soil Pb and Cd was insoluble and not available to the plant. ACKNOWLEDGEMENTS This research was supported by a grant from the Florida Thoroughbred Breeders Association. REFERENCES Aronson, A. L. 1972. Lead poisoning in cattle and horses following long-term exposure to kad. Am. J. Vet. Res. 33:627-629. Dollahite,]. W., R. L. Younger, H. R. Crookshand. L. P.Jones and H. D. Petersen. 1978. Chronic lead poisoning in horses. Am .J. Vet. Res. 39:961-964. Dollahite,]. W., L. D. Rowe and]. C. Reagor. 1975. Experimental lead poisoning in horses and spanish goats. SW Vet. 28:40-45. Hammond P. B. and A. L. Aronson. 1964. Lead poisoning in cattle and horses in the vicinity of a smelter. Annals New York Acad. Sci. II I: 595-611. Kramer, C. Y. 1956. Extension of multiple range tests 10 group means with unequal number of replicates. Biometrics 12:307310. Lagerwerff, J. V. and A. W. Specht. 1970. Contamination of roadside soil and vegetation with cadmium, nickel, lead and zinc. Env. Sci. Tech. 4:583-586. Mueller, P. K. and R. L. Stanley. 1970. Origin of lead in surface vegetation. AIHL Report No. 87. State of California Dept. of Public llcalth, Air and Industrial Hygiene Lab. Berkeley, CA p 1-15. NRC (National Research Council). I 972. Lead: Airborne lead in perspective. National Academy of Sciences, Washington, DC. NRC (National Research Council). 1980. Mineral tolerance ot domestic animals. National Academy of Sciences, Washington. DC. SAS Institute Inc. 1985. SAS Users guide: SAS Institute Inc., Cary, NC. Schmitt, N., G. Brown, E. L. Devlin, A. A. Larsen, E. D. McCausland and J.M. Sa\'ille. 1971. Lead poisoning in horses. ,\rch. Environ. Health 23:185-195. Soil Survey of Marion Co. Area Florida, 1979 U.S. Dep. of Agric. and Univ. of Florida, IF AS, Soil Scicrnc Dcp. Test methods for evaluating solid waste: phvsical/chernical methods 3rd ed. 1987. Vol. 1-A : Laboratorv manual physical/ chemical methods. Washington, DC: U.S. En\'iron. Protection Agency, Office of Solid Waste and Emergency Response p. 3050. TerHaar, G. L., R. B. Holtzman and ll. F. Lucas Jr. 1967. Lead and lead-210 in rainwater. Nature 216:'l5'l-354. Willoughby, R. A .. T. Thirapatsakun and B. J. McSherry. l'.172. Influence of' rations low in calcium and phosphorus on blood and tissue lead concentrations in horses. Arn. J. Vet. Res. 33: 1165-1173.


PROCEEDINGS, VOLUME 51, 1992 7 Mineral Content of Soils and Forage from Horse Farms in Marion County, Florida II. Minerals Required by Horses E. A. Ott,* S. Sundlof, and M. Tooker ABSTRACT Soil (0 to 7.6 cm) and bahiagrass ( Paspalum notatum Flugge) samples were collected from two pastures on each of 37 farms in Marion County, Florida in order to determine whether in an industrial facility burning clay and contaminated soil was impacting on horse farms in the vicinity of the facility. Farms were selected to represent fixed distances of 1.0 (Group A), 2.0 (Group B), 4.0 (Group C) and 8.0 (Group D) km from the industrial site 0.6 km east of Lowell, Florida. Group E farms were selected along State Road 329 between Lowell and a dump site 1.6 km NW of Fairfield, Florida. Group F farms were selected at random within Marion County but 8.0 km or more from the Lowell site. Preselected farms that were found to be adjacent to Interstate 75 'E. A. Ott a?d Mel Tooker; Animal Science Dep., Univ. of Fl;>nda, Gamesv11le, FL 326ll-09IO: and S. Sundlof, Physiological Sciences, College of Veterinary Medicine, Univ. of Florida, Gaines ville, FL 32611-0691. Florida Agric. Exp. Stn. Journal Series no. R-02100. *Corresponding Author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51:7-14 (1992) were also sampled along the highway right-of-way (Group G) to determine if highway emmissions were affecting forage and soil composition. Soil samples were analyzed for elements and related chemical characteristics. Forages were analyzed for Ca, P, Mg, K, Fe, Mn, Zn, Cu and Se. Group differences ( P < .05) were detected for soil P, Fe, Mn, Zn, Al, Na, NH4 N03 pH and electrical conductivity (EC). Except for N03 no linear relationships were detected due to distance from Lowell. A portion of the group effects may have been due to variations in soil type which wer~ shown to infl1;1~nce soil K, Fe, Mn, Cu, pH and EC. Average bah1agrass composltmn (mean SE on dry matter basis) was Ca, 3.5 .07 g kg-'; P, 1.9 .03 g kg-'; Mg, 0.16 .03 g kg-'; K, 13.6 .3 g kg-'; Fe, 60 2 mg kg-'; Mn, 75 5 mg kg-; Zn, 21 2 ~g kg-'; Cu, 4.7 .1 mg kg-'; and Se, 0.025 .008 mg kg-1 Fee_dmg ~rogr.ams for horses _receiving substantial portions of their nutrient mtake from bah1agrass pasture will need to be balanced to provide appropriate amounts of Ca, P, Fe, Mn, Zn, Cu and Se. provide a substantial portion of the daily nutnent mtake of most horses. When that forave is pasture, as it is during a major portion of the ye;;-on


8 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA the horse breeding farms in Marion County. Florida, a knowledge of the nutrient content of that pasture is essential for the proper planning of feeding programs. Since most of the horse pastures in Marion County are bahiagrass (Paspalurn nutatum Flugge), an understanding of the nutrient content of bahiagrass is essential to planning feeding programs and formulating appropriate supplements. The elements included in this study arc referred to collectively as minerals by nutritionists and that term is included in the title to identilv the focus of the studv. Nutrient cont~nt of forages is a 'function of species, maturity, soil nutrient availability, moisture availability and season. The relationship between soil nutrient availability and the nutrient content of the plant has been examined by many researchers (Kelling and Matocha, 1990). It can be generally concluded that if other nutrients arc not limiting, increasing availability of a required nutrient will result in increased plant growth and increased tissue concentrations of that nutrient. When excesses of the element are available, tissue mineral concentrations increase but yield decreases (Brown, 1970). For some forage species such as alfalfa this relationship and the interrelationship between nutrients has been explored in great detail (Kelling and Matocha, 1990). For the warm season perennial grasses such as bahiagrass, the data are much less complete. Snyder and Kretschmer (1988) reported on the mineral content of bahiagrass leaf samples collected from eight Florida counties and used the Diagnosis and Recommendation Integrated System (DRIS) to determine fertilization needs based on nutrient ratios in plant tissues. Other workers (Mislevy et al., 1990) reported on the composition of bahiagrass grown on phosphatic clay land reclamation sites. Both of these studies reveal considerable variability in the element content of bahiagrass due to growing conditions. This study was conducted to provide more accurate data on the element content of bahiagrass pasture grown in Marion County, Florida and to identify relationships between soil and forage element content. MATERIALS AND METHODS Bahiagrass and soil samples were collected from 37 farms in Marion County, Florida in order to determine whether an industrial facility burning clay and conLaminated soil was impacting on horse farms in the vicinity of the facility (Ott et al., 1992). Farms were selected to represent fixed distances of 1.0 (Group A), 2.0 (Group B), 4.0 (Group C) and 8.0 (Group D) km from an industrial site 0.7 km east of Lowell, Florida. Group E farms were selected along State Road 329 between Lowell and a dump site 1.6 km NW of Fairfield. Group F farms were selected at random within Marion County 8.0 to 25 km from Lowell. Soil and forage samples were collected from two pastures on each farm. Sampling sites in each pasture were selected by walking a pattern over the entire field and stopping every 10 to 15 paces to collect a sample. Soil samples were collected with a curved, stainless steel trowel. At each sampling site, thatch, if present, was scraped aside, a uniform sam ple of soil removed to a depth of 7 .6 cm and placed in a plastic bucket. At least 20 samples were collected from each pasture. The composite sample, 1.4 to 2.3 kg of soil, was bagged for transportation to the laboratory. Bahiagrass samples which represented the forage being consumed by animals grazing in the pasture were collected by clipping the grass at a height of 2.5 cm above the soil while walking the pattern noted above. The samples were bagged and transported to the laboratory. A third soil and forage sample was collected from the three farms adjacent to Interstate 75 (175) (Group G). These samples were collected 6.0 to 7.5 m inside the right-of-way fence parallel to the highway to identify the impact of automobile emissions on soil and forage mineral concentrations. All samples were collected between 25 Apr and 31 May, 1990. Soil samples were air dried, seived through a 2 mm screen, subsampled and submitted to the Florida State Soils Laboratory for analyses for soil pH, Ca, Mg, P, K, Zn, Cu, Mn, Al, Fe, Na. NH.1 NO3 organic matter (OM) and electrical conductivity (EC). The .Florida State Soils Laboratory uses the Mehlich I extraction procedure. Sample sites were identified on a soil map (Soil Survey of Marion County Area Florida, 1979) and primary soil map units recorded for correlation with soil analyses. Forage samples were dried, ground, sub-sampled, ashed at 500C for 20 h, solubilized in concentrated HN03 and analyzed by inductively coupled argon plasma spectroscopy (ICAP) for Ca, Mg, P, K, Zn, Cu, Mn, Al, and Fe. Selenium was analyzed by the flurometric procedure of Whetter and Ullrey (1978). All analyses were nm in duplicate. Data were analyzed according to the procedure for General Linear Models (SAS, 1985). Group differences were tested by Duncan's Multiple Range Test (SAS, 1985 ). Due to differences in the number of replications of soil types the modification of the Duncan's Multiple Range Test proposed by Kramer ( 1956) was used to test the effect of soil type on soil mineral content. RESULTS AND DISCUSSION Soil Mean Ca concentration of soil was 1154 mg kg-1 with a range of 137 to 3328 mg kg-1 No differences were detected between groups (Table 1). Soil Ca concentration recommended for bahiagrass is 600 mg kg 1 Qones, 1975). Mean soil P was 137 mg kg-1 with a range of 32 to 621 mg kg-'. Groups C, E and F had lower soil P than group G (P < 0.05). No other differences were detected. All of the samples collected were in the high to very high range. Mean soil Mg was 122 mg kg-' with a range of 25 to 319 mg kg-'. No differences were detected between groups. All values were in the medium to high range. Mean soil K was I 06


Table 1. Minerals and other physical characteristics of Marion County soils {fl to 7.6 cm from surface). No. No. Group farms past. Ca p Mg K Fe Mn Zn Cu Al Na NH4 NO3 OM pH EC ------------------------------------------mgkg'---------------------------------------------% '"d :ic A 4 8 Mean 1601 a' 198ab 145a 122a 15b 9.5b 7.0a 0.21 a 226b 9.3bc l.58ab 5.63a 3.6a 6.40a 0.16a 0 SE 350 64 16 19 2 0.8 1.0 0.04 12 1.0 0.28 2.29 0.4 0.13 0.03 n t"1 B 6 12 Mean 1306a 190ab 133a 128a 2lab 9.9b 5.3b 0.70a 238b l l.6ab 2.18a 3.75ab 3.9a 5.93ab 0. 14ab t"1 ti SE 226 40 17 14 2 1.3 0.5 0.31 33 0.8 0.47 0.81 0.2 0.11 0.01 ... z C 8 16 Mean 1143a 105b 119a 99a !6ab 9.4b 4.2bc 0.30a 217b 8.4c 2.05a 2.69ab 3.1 a 6.39a 0.13ab Cl SE 198 14 II 10 2 0.7 0.3 0.03 20 0.5 0.36 0.47 0.2 0.13 0.01 .CJJ D 7 14 Mean 894a 137ab 116a 83a 20ab 7.8b 3.4c 0.38a 227b 9.9abc 0.72b 1.36b 3.0a 5.96ab 0.!Ob < SE 170 37 21 10 3 1.1 0.3 0.11 19 0.8 0.16 0.20 0.3 0.12 0.00 0 r' E 3 6 Mean 892a 97b 107a 107a 27a l l.4ab 5.0bc 0.28a 202b 8.6e 1.93a 3.83ab 3.0a 6.02ab 0.14ab C: SE 192 26 13 18 5 1.5 0.4 0.05 23 0.8 0.24 0.86 0.1 0.18 0.01 a:: t"1 F 9 18 Mean 1163a I00b 116a 103a 23ab 9.6b 3.9bc 0.35a 200b 10.Sabc l.34ab 2.11 b 3.4a 6.09a 0.l3ab Ul SE 110 II 6 10 3 0.9 0.3 0.04 9 0.6 0.19 0.38 0.1 0.13 -0.01 G 3 3 Mean 1085a 230a 120a 128a 24ab 14.7a 3.2c 0.41a 315a 12.6a 0.50b 1.50b 3.6a 5.53b -0.!Ob '-.0 SE 209 45 41 18 6 7.0 0.5 0.10 20 1.6 0.50 1.04 0.5 0.29 0.00 '-.0 N) Totals 37 77 Mean I 154 137 122 106 20 9.6 4.5 0.38 222 10.0 1.54 2.83 3.3 6.11 0.13 Range 137-3328 32-621 25-319 27-234 6.0-51 3.2-28.3 0.09-0.3 0.09-4.0 112-463 5.05-9 0-5.55 0-20 1.1-5.7 5.0-7.4 0.08-.34 P< 0.350 0.072 0.791 0.160 0.158 0.238 0.034 0.275 0.244 0.010 0.007 0.016 0.10 0.019 0.034 'Means in the same column followed by a common letter are not different (P < 0.05) by Duncan Multiple Range Test.


10 Soll, AND CROP SCIENCE SOCIETY OF FLORIDA mg kg I with a range of 27 to 234 mg kg-I. No differences were detected between groups. Some of the values were in the low range (10-35 mg kg ') and some in the very high range (> 125 mg kg-I). Mean soil Fe content was 20 mg kg-' with a range of6 to 51 mg kg'. Group A had lower Fe than group E (15 vs 27 mg kg'. P < 0.05). No other differences were detected. Mean soil Mn was 9.6 mg kg-' with a range of 3.2 to 28.3 mg kg-'. Group G had higher soil Mn than groups A, B, C, D and F (P < 0.05). Mean soil Zn was 4.5 mg kg-' with a range of 0.9 to 10.3 mg kg-I. Group A had the highest soil zinc (7.0 mg kg '), followed by group B (5.3 mg kg-I) and group E (5.0 mg kg-I). Mean soil Cu was 0.38 mg kg---1 with a range of 0.09 to 4.0 mg kg-'. No differences were detected between farm groups. Mean soil Al was 222 mg kg---1 with a range of 112 to 463 mg kg-'. Group G had higher soil Al (~) 15 mg kg-') than the other groups (P < .05). The Na mean was 10.0 mg kg-1 with a range of 5.0 to 15.9 mg kg-'. Group G Na (12.6 mg kg-') was higher than groups C Na (8.4 mg kg-I) and E Na (8.6 mg kg---1 ) (P < .05). No other differences were detected. Soil NH4 concentrations were variable and although group differences were detected, soil NH4 concentrations are probably more closely related to fertilization programs than location. Soil NO3 concentrations were highest for group A (5.63 mg kg ') and lowest for groups D (1.36 mg kg-'), G ( 1.50 mg kg-') and F (2.11 mg kg-') (P < 0.05). There appeared to be a linear decrease in soil NO3 with distance from Lowell. Soil organic matter was not influenced by location. Soil pH was higher for groups A (6.40), D (6.39) and F (6.09) than for group G (5.53) (P < 0.05). Other groups were intermediate. Soil EC was very low but some group differences were detected. Group A had higher EC (O.Hi) than groups D (0.10) and(~ (0.10) (P < 0.05). No other differences were detected. Since soil samples collected in this study were taken only to a depth at of 7 .6 cm, their direct comparison with values collected to a more standard depth of 15 cm must consider the dilution effect due to soil depth. Values reported in this study may therefore be higher than values detected when more standard soil sample procedures are used. Soil type did not influence soil Ca, P, Mg, Zn, Al, Na, NH1 NO3 and OM concentrations (Table 2). Differences identified between groups (location) would therefore be considered of more importance for these factors. All of the other soil parameters varied with soil type. Potassium was higher in Lochloosa fine sand (LO) (187 mg kg-') than in Gainesville loamy sand (GA) (42 mg kg-1), Candler sand (CA) (64 mg kg---1 ) and Arredondo sand (AR) (80 mg kg-1 ) (P < .05). Other soil types were intermediate in Kand not different. Soil Cu was higher in GA (I. 76 mg kg-') than all other soil types except Zuber loamy sand (ZU) (1.18 mg kg') and Cu was higher in ZU than Flemington loamy sand (FM) (0.20 mg kg-') (P < 0.05). Other soil types were not different. Soil Mn was higher in LO (17.6 mg kg-') than GA (3.8 mg kg-'), CA (6.1 mg kg-') and AR (7.2 mg kg-') (P < .05). Soil types LO, FM, Blichton sand (BC), ZU, GA, Sparr fine sand (SP), Kanapaha fine sand (KA) and Micanopy fine sand (!\IC) were higher in Fe than were Kendrick loamy sand (KE), AR, Hague sand (HA) and CA (P < .05). Soil types BC, ZU, GA, SP, KA, MC, KE, AR, and HA were higher in Fe than was CA (P < .05). No other differences were detected. Soil EC was higher in AP than all the soil types except LO, SP and MC. Soil pH was higher in CA (6.6) than GA (5.2) and AP (5.3) (P < .05). The pH was higher in KE (6.3) than in GA (P < 0.05). No other differences were detected. It is of interest that the single sample of GA had the lowest Ca, Mg, K, Zn, Mn, NO3 OM, EC and pl I and the highest Cu and Al concentrations of the 77 samples collected. Forage Bahiagrass pasture samples collected from Marion County, Florida horse farms were analyzed for minerals required by horses. The results (Table 3) are expressed on a dry matter basis. Two of the forage samples collected were not bahiagrass and were eliminated from the report. Calcium concentration of bahiagrass was 3.6 0.007 g kg-' (mean SE) with a range of 2.3 to 6.0 g kg---1 There were no differences between any of the groups (P > 0.05). Phosphorus content of the forage was 1.9 .03 g kg-' with a range of 1.3 to '.U g kg-1 Groups A and E had higher P than group G (P < .05). No other differences were detected. Magnesium concentration was 1.6 0.03 g kg-' with a range of 1. 1 to 2.3 g kg '. Groups C and E were higher than group G (P < 0.05). No other differences were detected. Potassium averaged 1.36 .3 g kg-' with a range of 8.2 to 20.2 g kg-'. No differences between groups were detected (P > (l.05). All of the macromineral concentrations except K were below the mean values reported for bahiagrass composition in samples collected from eight Florida counties which did not include Marion County (Snyder and Kretschmer, 1988). Trace mineral concentrations in the forages were more variable than the macrominerals. Iron averaged 60 2 mg kg' with a range of 26 to 115 mg kg-'. Group E had higher concentrations than groups D and G (P < 0.05). Groups A, C and F had higher values than group G (P < 0.05). Forage Mn averaged 75 5 mg kg---1 with a range of 14 to 238 mg kg-'. Groups B, E and G had higher concentrations than A (P < 0.05). No other differences were detected. Zinc content of bahiagrass was 21 2 mg kg-' with a range of 10 to 31 mg kg-'. No differences were detected between groups (P > (l.05). Mean forage Cu was 4.7 0.1 mg kg-1 with a range of 2.8 to 7.0 mg kg-1 No differences between groups were detected. Selenium content of bahiagrass was 0.025 0.008 mg kg' with a range of O to 0.17 mg kg-'. Mean Fe, Mn, Zn, and Cu content of the bahiagrass were lower


Table 2. Minerals and other chemical and physical characteristics of soils in various soil map units in Marion County, Florida (0-7.6 cm from surface). Soil tvpe N Ca p Mg K Zn Cu :Vin Al Fe Na NH, '.\!0:1 OM EC pH '"O -mg kg-1 -7c ----------------------------------0 AP l 514a' 92a IOOa 97ab 5.8a 0.23bc 13.0ab 221a 50a 8.7a 2.52a 4.00a 3.7a 0.28a 5.3bc n M AR 18 843a 94a !Ola 80b 3.7a 0.36bc 7.2b 201a 14cd 9.3a 0.84a 1.75a 2.7a 0.1 !b 6.2abc M BC 4 1069a 117a 116a 127ab 5.6a 0.48bc 12.3ab 223a 30bc 11.4a 1.76a 4.00a 3.Sa 0.12b 5.6abc 8 CA 3 1078a 87a 121a 64b 4.3a 0.27bc 6.1 b 186a 10d I I. I a l.OOa 1.83a 2.6a 0.1% 6.6a z FM 6 1590a 190a 162a 106ab 3.2a 0.20c 8.9ab 200a 32c 10.9a .67a 1.08a 4.0a O.!Ob 6.1 abc () [/} GA I 360a 130a 53a 42b .9a 1.76a 3.8b 330a 24bcd IO.Sa 0.50a 1.00a 2.2a O.!Ob 5.2c < HA 2 696a 84a I !Sa 108ab 3.2a 0.36bc 9.7ab 320a 14cd 12.0a 0.25a l.OOa 4.3a O.!Ob 6.0abc 0 KA 3 760a 56a 112a 95ab 4.8a 0.25bc 8.9ab 163a 21 bed 7.3a 1.68a 4.50a 2.9a 0.13b 6.labc t"' KE 26 1466a 156a 134a I 13ab 5.2a 0.25bc 10.6ab 226a 17cd 9.5a 2.13a 3.65a 3.6a 0.14b 6.3ab c::: LO I ll!Oa 209a 122a 187a 2.9a 0.39bc 17.6a 236a 39ab 12.2a 3.0 a 4.0 a 3.2a 0.20ab 5.9abc MC I 530a 67a 94a 98ab 4.4a 0.25bc 9.7ab 127a 19bcd 7.5a 3.0 a 4.0 a 3.5a 0.16ab 5.6abc "' SP 3 1377a 153a !!Sa 153ab 6.Sa 0.44bc l l .4ab 219a 2lbcd 9.8a 2.8 a 5.8 a 3.7a 0.17ab 6.0abc -zu 5 976a 185a 126a 136ab 5.0a l.18ab 9.0b 257a 26bc 11.4a 1.9 a 2.5 a 3.6a 0.1 lb 5.7ab, -P< 0.17 0.70 ().55 0.03 0.03 0.005 0.01 0.29 (l.0001 00.55 0.002 0.34 0.008 0.003 0.02 <.O <.O N:) 'Soil types are identified as follows: Apopka sand (AP), Arredondo sand (AR), Blichton sand (BC), Candler sand (CA), Flemington loamy sand (FM), Gainesville loamy sand (GA), Hague sand (HA), Kanapaha fine sand (KA), Kendrick loamy sand (KI), Lochloosa fine sand (LO), MicanO(J\ fine sand (MC), Sparr fine sand (SP), Zuber loamy sand (ZU). 1Means in the same column with the same letter are not different (P < .05). -


Table 3. Bahiagrass mineral concentration (dry matter basis). [/'; ,.., V ... No. No. r Group farms pastures Ca p Mg K Fe Mn Zn Cu Se 7'. .. '-' g kg-' -mg kg-1 -n A 4 8 mean 3.7 a' 2. la 1 .5ab 12.8 a 59bcd 42hcd 19a 4.7a 0.020ah SE 0.2 0.2 1 0.09 5 6 2 .4 0.008 'V -,;: B 6 12 mean 3.5 a 2.0ab l.5ab 14.8 a 67ab 88a 18a 4.7a 0.006b /Jl SE 0.1 0.0 0.0 0.5 5 9 1 .2 0.002 n ""' C 8 14 mean 0.34a 1.Sab 1.6a 1.34a 64abc 73ab 15a 4.8a 0.028ab t"'1 SE 0.2 0.1 0.1 0.7 5 9 I .3 0.006 z '"' D 7 14 mean 4.0 a l.9ab l.6ab 1.32a 49cd 78 16a 4.5a 0.0 l 9ab SE 0.2 0.0 0.1 0.9 2 11 1 .2 0.007 /Jl F 3 6 mean 3.6 a 2.1 a 1.7a 1.43a 78a 89a 18a 5.5a 0.049a 0 n SE 0.4 0.1 0.2 01.3 13 13 I .3 0.023 ""' F 9 18 mean 3.4 l.9ab l.6ab 1.37a 57bcd 7lab 18a 4.5a 0.035ab a .., SE 0.1 0.1 0.1 0.7 4 10 1 .2 0.010 -< G 3 3 mean 3.5 a 1.7b 1.4b 12.0 a 42d 110a 18a 4.7a 0.016ab 0 SE 0.4 0.0 0.1 0.5 3 64 2 .2 0.009 "Tl Total 37 75 mean 3.6 1.9 1.6 13.G GO 75 17 4.7 0.025 'Tj r SE .07 0.3 0.3 0.3 2.1 5 2 .1 0.008 0 range 2.3-6.0 1.3-3.1 1.1-2.3 8.2-20.2 26-115 14-238 10-31 2.8-7.0 0-0.17 ti "Means in the same column with common letters are not different (P < 0.05) by Duncan Multiple Range test.


PROCEEDINGS, VOLUME 51, 1992 13 Table 4. Means ( SE) and correlations between soil and forage mineral elements. Mineral Means SE element Soil Forage Correlation P< -mg kg-' -Ca 1154 3547 0.218 0.057 p 137 13 1913 0.237 0.038 Mg 122 1574 0.317 0.005 K 106 13467 0.073 0.531 Mn 9.6.5 75 0.517 0.0001 Fe 20.0 1.2 60 -0.029 0.800 Zn 4.5.2 21 0.230 0.044 Cu 0.38.05 4.7.1 -0.115 0.317 than mean values reported by Snyder and Kretschmer ( 1988). The correlation between soil and forage minerals (Table 4) was significant for P (r2 = 0.238, P < 0.038), Mg (r2 = 0.317, P < 0.005), Mn (r2 = 0.517, P < 0.0001), and Zn (r2 = 0.230, P < 0.044). No other relationships were detected. Mineral content of bahiagrass pastures sampled in this study were compared with NRC (1989) composition data (Table 5). Mean values were below values previously reported (NRC, 1989) except for Mn. Bahiagrass mineral content was also compared with the nutrient requirement of mature horses at maintenance to identify supplementation requirements. Mean concentrations of forage macrominerals, Ca, P, Mg and K were all above the NRC ( 1989) requirements for maintenance of mature horses (Table 5). Forage on some farms had Ca and P concentrations below the animal's requirement, however, and grazing horses would need supplementation if the pasture was the only source of these minerals available to the animals. Magnesium and K content of the forage from all of the farms was adequate to meet the NRC (1989) recommendations. Gestating mares, lactating mares, growing foals and working horses have higher requirements for one or more of these minerals. Since these groups of animals would seldom rely on bahiagrass pasture as their only source of nutrients, requirements for supplemental minerals will depend on the combination of feeds offered to the animals. Eight of the 75 bahiagrass samples contained Fe concentrations below the horse's maintenance requirement of 40 mg kg-I (NRC, 1989) (Table 5). Sixteen of the samples contained Mn concentrations below the maintenance requirement of 40 mg kg-I (NRC, 1989). Supplementation of both of these minerals is recommended for all farms to minimize the likelihood of a deficiency. All of the bahiagrass samples contained Zn and Cu concentrations below the maintenance requirement of mature horses and most of the Se concentrations were also below the requirement. Supplementation of these three minerals is also essential to the well-being of the grazing animal. Mineral supplementation of forages can be provided to horses via additions to the concentrate, if one is being fed, and/or the offering of free-choice minerals. If forage Ca or P concentrations are below requirements (NRC, 1989) or are present in improper ratios, a complete mineral should be offered providing Ca, P, salt (NaCl), Fe, Mn, Zn, Cu and Se. If Ca and P concentrations in the forage are adequate to meet the requirements, trace mineralized salt can be offered. Although Na and Cl concentrations in bahiagrass were not determined in this study, these minerals are generally present in very low concentrations in most forages (NRC, 1989) making salt additions to the diet essential. The addition of macro-and microelements to the soil will increase forage mineral content and production (Brown, 1970). The low correlation between soil and bahiagrass mineral concentrations found in this study, however, indicate that other factors including soil type and other soil components may also influence the availability of soil minerals to bahiagrass. CONCLUSIONS Marion County, Florida soils are variable in mineral content but most pastures sampled in this study contained adequate mineral concentrations for forage production. Correlation between soil type and critical mineral elements required for forage growth was low. Soil NO3 was the only soil component that decreased in concentration with distance from Lowell Table 5. Comparison of bahiagrass mineral concentrations and mineral requirements of horses.' Mineral element Ca p Mg K Fe Mn Zn Cu Se 'Dry matter hasis. 1NRC, 1989. MeanSE' 3.6.01 1.9 0.00 1.6 0.00 13.6 0.03 60 2 75 21 2 4.7 .I 0.025 0.008 Bahia rass This study Range 2.3 to 6.0 l.3to3.l 1.1 to3.3 8.2 to 20.2 26to 115 14 to 238 10 to 31 2.8 to 7.0 0to0.17 'Units: g kg-' for Ca, P, Mg, P and mg kg-' for Fe, Mn, Zn, Cu, Se. Requirements for NRC' Maintenance of Mean horses' 4.4 2.4 3.0 1.7 2.7 0.9 15.3 3.0 85 40 75 40 29 40 7.2 IO 0.06 0.10


14 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA indicating that emissions from the industrial facility at Lowell have had little influence on soil and forage on the farm in the vicinity of the facility. Bahiagrass mineral concentrations of Zn, Cu and Se from all pastures sampled in this study were below :'-JRC ( 1989) requirements for mature horses indicating that supplementation programs for these minerals are essen tial for horses grazing bahiagrass pastures. Bahiagrass Fe and Mn concentrations were also below requirements for mature horses grazing some of the pastures sampled. ACKNOWLEDGEMENTS This research was supported by a grant from the Florida Thoroughbred Breeders Association. REFERENCES Brown, 1970 Plant Analysis. Missouri Agric. Exp. Sta. Bull. SB 881. Columbia, .\10. Jones, D. W. 1975. Pasture and forage crops for horses. Agronomy Facts No. 53. Florida Coop Ext. Ser. Gainesville, FL. Kelling, K. A. and J. E. l\lat.ocha. I \190. Plant analysis as an aid in fertilizing forage crops. In: R. L. Westerman, Ed. Soil Testing and Plant Analysis 3'" Ed. Soil Sci. Soc. Am. Madison, WI. Kramer, C. Y. 1956. Extt'nsion of multiple range tests to group means with unequal numbers of replicates. Biometrics 12:307-310. Marteus, D. C. and W. L. Lindsay. 1990. Testing soils for copper, iron, manganese and zinc. In: Soil Testing and Plant Analysis 3rd Ed. R. L. Westerman, Ed. Soil Sci. Soc. Am. Inc. Madison, WI. .\1islevy, P., W. G. Blue and C. E. Roessler, 1990. Productivity of clay tailings from phosphate mining: II forage crops. J. Environ. Qual. 19:694-700. NRC. 1989. Nutrient requirements of horses, 5th Rev. Ed. National Research Council, National Academy of Sciences, Washington, D.C. Ott, E. A., S. Sundlof and M. Tooker. 1992. Mineral content of soil and forage from horse farms in Marion County, Florida. I. Lead and cadmium. Soil Crop Sci. Soc. Florida Proc. 51:3-7. Snyder, G. H. and A. E. Kretschmer, .Jr. 1988. A DRIS analysis for bahiagrass pastures. Soil Crop Sci. Soc. Florida Proc. 47:56-59. Whetter, P. A. and D. E. llllrcy. I \)78. Improved fluorometri, method for determining selenium. J. Assoc. Off. Anal. Chem. 61:927.


14 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Polymer Conditioners for Florida Soils D. z. Hanman* and M. E. B. Joyner ABSTRACT This paper reviews the state of the art of soil polymer conditioners. Organic polymers can have a profound effect on the soils to which they are applied. They can improve soil waterholding capacity, aggregation, infiltration, and soil stability. There are two basic types of organic polymers: linear polymers and crosslinked superabsorbing polymers. The possible applications of these polymers under Florida's conditions are discussed in this paper. It is concluded that landscaping and turf improvement are two areas where the advantages of polymer application may outweigh the cost. ----------Florid a receives rainfall of 1200-1500 mm/y-1 (50-60 in/yr1), by most standards an abundant sup_ply of fresh water. Yet Florida requires supplemental irrigation for successful crop production and landscape maintenance. This contradiction is due to the very low water-holding capacity of Florida's soils and the uneven distribution of rainfall through the year. Recently, there has been increasing competi~ion for water between Florida's agriculture and rapidly D. z. Haman, Dep. of Agricultural Engineering, Univ. of Florida, Gainesville, FL 32611-0570 and M.E.B. Joyner, Graduate Student, Agronomy Dep., Univ. of Florida. Florida Agric. Exp. Stn. Journal Series no. R-02085. *Corresponding author. Mention of trade names or proprietary products is for the convenience of t~e reader only, and do~s not constitute endorsement or preferential treatment by the Uni versity of Florida. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51: 14-18 (1992) growing population.To ensure a suf~cient supply ~f good-quality water for every~me in the iuture,_ 1t 1s necessary to increase the effioency of water use m all sectors including agriculture, industry, landscape management, and households. At the present time, most of the ramfall m Flonda is not effective. The water-holding capacity of the soil is so low that a majority of rainwater is lost from the root zone. Consequently, an increase in soil waterholding capacity would result in significant decrease of the irrigation requirement. Two approaches can be used to retain water in sandy soil. One is to prevent the water from evaporating from the surface by an evapo:ation b~n.ier, the other is to add substances to the soil that will mcrease the soil water-holding capacity by capturing and retaining a larger fraction of the water passing through. The objective of this paper is to summarize a recent report to the South Florida Water Man~geme~t District regarding organic polymers and their appli cation as soil amendments (Joyner and Haman, 1991). The report reviews the state of the art of soil amendments and conditioners and provides over two hundred references. The report discusses mulches, inorganic and organic undecomposed soil additives, compost, and organic polymers. The major emphasis of the re view is on various organic polymers and their applica tion under Florida conditions.


PROCEEDINGS, VOLUME 51, 1992 15 ORGANIC POLYMERS The effects of synthetic organic polymers on the soil are very similar to those of the naturally-occurring humates and are often compared to them in the literature (Schnitzer and Poapst, 1967; Senn and Kingman, 1975). The organic polymers can be subdivided into two categories, water soluble and water insoluble. The water soluble, linear polymers are very useful for improving soil structure, while water insol-, uble crosslinked polymers can hold vast amount of water, so their addition can increase water-holding capacity of the soil. Both types of conditioners have the potential to improve the water-holding capacity of Florida's sandy soils. Linear Soil-Aggregating Polymers and Related Products In the history of polymer chemistry, linear polymers were the first to be synthesized. They were found to affect and alter soil structure and soil chemistry and alleviate soil crusting, thus promoting aggregation, curbing erosion, and enhancing soil moisture status and aeration (Quastel, 1954). As the technology of polymer synthesis evolved, it was discovered that the linear polymers could be modified to form crosslinkages, giving rise to a type of polymer with very different properties and superior water retention. Both types are now used for soil conditioning. Linear polymers were first called synthetic poly electrolytes, a name indicating that these products carried electrical charges similar to soil constituents like clay and organic matter (Hedrick and Mowry, 1952). In 1952, researchers at Monsanto reported that certain of these water-soluble, high-molecularweight polymers at very low concentrations (0.1-2.0 % by weight) (Hedrick and Mowry, 1952) were effec tive as soil aggregators. One of the most active was a polyacrylonitrile named CR-189. CR-189 and subsequent modifications, many marketed under the umbrella name Krilium, were the forerunners of such products as HPAN, H-SPAN, VAMA, PAM, PVA, PV Ac, and others (Table 1). An additional compound used as a soil conditioner is guar, a polysaccharide linear polymer derived from guar beans, which has many of the properties of the synthetics (Wallace, 1986). Table 1. Summary of the most widely used linear polymerst. From Azzaro (1980); Johnson and Veltkamp (1985); Quastel (1954). Guar HPA:--; H-SPAN PAM PVA PVAc VAMA natural polysaccharicle polymer from guar bean hydrolyzed polyacrvlonitrile hydrolyzed starch polyacrylonitrile polyacrylamide polyvinyl alcohol polyvinyl acetate modified vinylacetate-maleic acid compound 'Most linear polymers can be chemically altered to make crosslinked types bearing the same name or acronym. All linear organic polymers including guar are composed of single or branched polymer chains. Their basic chemical composition makes them water soluble and predisposes them to have naturally nega tive charges. This charge can be enhanced (made more negative), lessened (made less negative, approaching neutral), or changed to positive during synthesis or afterwards by substituting carboxyls, amines, alcohols, or other groups. Because similar charges repel one another, one might conclude that the negatively charged anionic linear polymers are not adsorbed as readily on clay particles. Just the opposite is true (A. Wallace, El Segundo, Calif., pers. comm.) because the anionic polymers have carboxyl (-COOH) groups that bind with divalent and trivalent cations of minerals such as calcium, magnesium, and aluminum, which, in turn, bind with the clay by the mechanisms of hydrogen bonding (De Boodt and Verdonck, 1972), charge neutralization bonding (Harris et al., 1966), protonation, ion dipole coordination, or polyvalent metal cation bridging (Edwards and Bremner, 1967). The binding causes the clay particles to flocculate (A. Wallace, pers. comm.). Cationic polymers bind directly to the clay particles, often with amide bonds (-NHx)-In either case, the result is flocculation with consequent aggregation of the clay particles. It should be noted that the efficacy of linear polymers in treating soil problems is usually tied to the amount and type of clay in that soil and the sorts of problems to be treated. Linear polymers are particularly effective in soils with substantial clay fractions because they are electrically active and interact with the electrical charges of clays. The degree of change in soil properties largely depends on the specific type of clay in the clay fraction, even in Florida's sandy soils with a very small percentages of clay (Joyner and Haman, 1991). Crosslinked Superabsorbing Polymers The insoluble, crosslinked polymers are known as super-adsorbents, hydrogels, or super slurpers. Because they have remarkably high water-absorbing properties, they are able to hold in excess of 120-500 times their own weight in water (Johnson, 1984; Kullmann et al., 1986). Depending on their original form, granular or pellet, when superabsorbents are hydrated they form an extensive gelatinous mass or discrete gel particles (Johnson, 1984). They are capable of cyclic absorption and desorption over relatively long periods, providing plants with available water from a slow-release, renewable soil reservoir (Johnson, 1984). The crosslinked polyacrylamides (PAMs) that show superior performance release 10 percent of their considerable water reserves (more than 200 grams per gram of polymer) at tensions as low as 15 kPa (0.15 bars) (Johnson and Veltkamp, 1985). Generally, about 90 percent of water stored by the polymer is available to plants. Superabsorbents are marketed under various, in consistent names. The main categories are polyacrylam-


16 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA ides (PAM), polyvinylalcohols (PV A), and polyacryloni trile starch copolymers (usually marketed as H-SPAN) (Hemyari and Nofziger, 1981; Johnson, 1984; Miller, 1979). In electronmicroscopic examination (Johnson and Veltkamp, 1985) high performance crosslinked PAM products were found to have chains that fold back on themselves, with sufficient hydrogen bond crosslinks to keep their structure. This creates a cellular form, not unlike a honeycomb, with 80 to 85 percent of the available water in but not bound to the folds of the polymer, readily available for slow release. The remaining 15 to 20 percent is somewhat more tightly bound in the polymer structure but most or all still available for plant use. The products that performed less well had too few or too many crosslinks for optimum water retention. Superabsorbents are used primarily in nurseries and allied businesses as soil and media conditioners to aid germination, to extend intervals between waterings, to reduce transplant shock, and as a buffer to reduce wilting (Greenwood et al., 1978; Rogers and Anderson, 1981; Whitmore, 1982). According to Ingram and Yeager (1987), plant performance remains the most meaningful method of evaluating water relations in container media. This implies that the water-holding capacity of a medium per se is secondary; the ultimate criterion is the condition of the plants. Consequently, research on gel amendments to nursery mixes have emphasized water relations of the plants. Table 2 summarizes the results of trials emphasizing plant water status. Most research on superabsorbents has been for nursery application because of severe economic constraints to large-scale uses. Although relatively less work has been done that relates to landscape installa tions, fruit and vegetable growing, or field crops, most of that work was done on sandy soils or soils with a high sand content. THE PERSISTENCE OF ORGANIC POLYMERS Most of the organic polymers are relatively stable molecules. Because research rarely continues more than several growing seasons, there is little information on the absolute persistence of polymers in the soil. According to Arthur Wallace (El Segundo, Calif., pers. comm.), most polymers break down at a rate of about 10 percent a year. This degradation is mostly due to mechanical damage to the molecules as a result of tillage and cultivation and from foot and vehicular traffic. Guar, a natural polymeric product, is subject not only to mechanical damage but also to microbial attack. Consequently, its breakdown, similar to other nonlignin organic materials in the soil, occurs within about two years. THE ECONOMICS OF USING POLYMERS Improvements in polymer synthesis and the consequent increases in the desirable qualities of the products applicable to soil conditioning have reduced Table 2. Effects of superabsorbentst on plant water status. Product Plant Effect on water status Viterra* chrysanthemum improved hours to wilt, reduced waterings, improved shelf life Viterra cmerana no improvement Liqua-gel, ligustrum reduced waterings H-Span type# tomato w/sand transplants improved hours to wilt w/finer texture no improvement Viterra 2t t marantha increased medium water content, improved hours to wilt, increased shelf life; pilea no significant increase of medium water content, improved hours to wilt, increased shelf life Veterra2H marigold improved hours to wilt zinnia improved hours to wilt Viterra 2 tomato improved qua!. of transplants Super sluper,, tomato improved qua!. of transplants Agrosoke,~ transplants spider (no reduction in watering) plant 2X mfg. recommended amt. = 50% increase in plant size Boston fern (no reduction in watering) 'ferra-sorb## no influence on plant size tomato pure gel + fert. = improved seedlings hours to wilt Alcosorb## tomato pure gel + fert. = improved seedlings hours to wilt t. Viterra = polyethylene oxide; Viterra 2 = potassium propenoate propenamide copolymer; Liqua-gel, Super Slurper, I-ISPAN = hydrolyzed starch polyacrylnitrile; Moisturite = starchgrafted polyacrylate polymer; Terra-sorb = polyacrylamidepotassium acrylate; Alcosorb = polyacrylamide; Agrosoke = polyacrylamide. *. Still (1976); Tu et al. (1985). ,_ Taylor et al. (1986); #_ Henderson et al. (1986); t t. Conover et al. (I 979); H_ Gehring et al. (1980); Willingham et al. ( 1981); ,,_ Wang et al. (1987) ##_ Pill, (1988). the quantities required to obtain the desired effects anywhere from a factor of 10 to a factor of 100. Where it might have taken 2 percent by weight when research first began, today it may take 0.2 percent, 0.02 percent, or even less to obtain comparable re sults. At this time, the linear PA.Ms in particular require much lower application rates than were true even ten years ago (Wallace and Wallace, 1986; Wal lace, 1990). The larger crosslinked PAM molecules, too, hold and release more water in the soil than did the earlier prototypes. Despite the extensive body of literature relating to both linear and crosslinked polymers (Joyner and Haman, 1991) it is difficult to obtain information on costs of bulk purchases of polymers. Most of the polymers available for retail sales are crosslinked PA.Ms (polyacrylamides) in small quantities repackaged by distributors and intended for use in houseplant media and other small-scale applications. The


PROCEEDINGS, VOLUME 51, 1992 17 majority of this crosslinked PAM is manufactured by American Cyanamid under the name "Aquastore." American Cyanamid then distributes the product to wholesale or retail repackagers or to quantity consumers. Aquastore is available in bulk at the prices presented in Table 3. Linear PAM also is available in bulk from Complete Green, El Segundo, California, for US$9.46/kg1 (US$4.29/lb-1 ) in 50 pound bags. Air Chem of Allentown, Pennsylvania, manufactures polyvinyl alcohol (PV A) and polyvinyl acetate (PV Ac) in bulk. The prices are: PVA: US$4.96/kg-1 (US$2.25/lb') US$3.35/kg' (US$ l .52/lb-1 ) in pallet (2500 lbs), FOB Allentown PVAc: (US$1.665/lb') US$3.67 /kg-1 US$3.67/kg' freight free. (US$ 1.665/lb-1 in pallet (1500 lbs), There was no current source found for polyacrylonitrile (PAN) in commercial quantities, although it can be purchased in what are known as research quantities 25 g to 25 kg (.05 IL to 11.3 lb) from scien tific suppliers for approximately US$5 for 25 grams (about 0.05 lb or 0.8 oz.). When contemplating the application of organic polymers, two factors must be considered: the scale of the application and the cost in terms of the expected return. The superabsorbing, crosslinked polymers are already well accepted and widely used by nurseries and plant propagators. Field, turf, and landscaping applications in Florida, however, are rel atively unexplored areas. Field application of linear or crosslinked polymers to improve water-holding capacity and structure must be justified by the expected economic return. Few if any crops currently grown would warrant the expense entailed when other, less expensive, alternatives already exist and have proven successful. Landscaping and turf improvement are two areas where the advantages of polymer application may far outweigh the cost. Landscaping, quite simply, carries the implicit expectation that the installed grass, plants, and trees will survive-that the newly-finished appearance will persist or improve with time. Florida Table 3. Wholesale prices of crosslinked PAM Aquastore: Form/Quantity t Price kg lb US$ kg-' US$ lb-' Granular I to 875 I to 1,929 7 .56 3.43 900 to 4,525 I ,984 to 9,973 6.79 3.08 4,550 to 13,600 I 0,028 to 29,974 5.80 2.63 13,625 or more 30,030 or more 4.80 2.18 Powder I to 875 1 to 1,929 7.76 3.52 900 to 4,525 1,984 to 9,973 7.05 3.20 4,550 to 13,600 I0,028 to 29,974 6.15 2.79 13,625 or more 30,030 or more 5.25 2.38 tPackaged in 25 kg (55.1 lb) bags soils with low water-holding capacity require very frequent water applications, each in relatively small amounts. This type of watering regime frequently re sults in poor establishment of landscaping materials or excessive water use. One potential treatment to ameliorate, if not solve, these problems is the addition of organic polymers to the soil. The superabsorbers can retain applied water and release it to the plants. The linear polymers, in situations such as turf establishment for lawns or golf courses, could be used to improve the soil structure and improve the soil water-holding ca pacity. NEED FOR RESEARCH IN FLORIDA Most of the research on organic polymers has been conducted in climates or soils different from those that predominate in Florida. Consequently, generalizations are possible but specific applications must be speculative Qoyner and Haman, 1991). To determine how well the superabsorbents would perform in Florida, research should be performed using typical soils of distinct physical and chemical characteristics. Based on existing research, there should be a selection of potentially effective products which would be applied to the soils in various amounts using representative water sources and frequently-recommended fertilizer applications. The length of time that various products persist in the soil should also be determined for this climate and these varying experimental conditions. From the existing research results, the use of the linear polymers is even more dependent on the existing local conditions. Research similar to that outlined above for crosslinked polymers should be conducted incorporating the additional component of clay content. It is necessary to determine whether the intrinsic clay fraction, as small as it is, has the ability to retain and utilize the polymers. It is even more important to determine whether the addition of small amounts of clay, readily available from phosphate mining, might be the key to utilizing the soil-improving properties of the linear polymers. REFERENCES Azzam, R. A. I. 1980. Agricultural polymers: Polyacrylamide preparation, application. and prospects in soil conditioning. Comm. Soil Sci. Plant Anal. 11:767-834. Conover, C. A., and R. T. Poole. 1979. Influence of pH on activity of Viterra 2 and effects on growth and shelf life of Maranta and Pilea. Proc. Florida State Hort. Soc. 92:332-333. De Boodt, M., and 0. Verdonck. 1972. The physical properties of the substrates in horticulture. Acta Hort. 26:37-44. Edwards, A. P., andJ. M. Bremner. 1967. Microaggregates in soils. J. Soil Sci. 18:64-73. Gehring, J. M., and A. J. Lewis, III. 1980. Effect of hydrogel on wilting and moisture stress of bedding plants. J. Am. Soc. Hort. Sci. 105:511-513. Greenwood, G. E., G.D. Coorts, and R. D. Maleike. 1978. Research conducted to determine value of various soil amendments. Am. Nurseryman 148: 12-13. Harris, R. F., G. Chesters, and 0. N. Allen. 1966. Dynamics of soil aggregation. In Advances in agronomy, Vol. 18. A.G. Norman, ed. New York: Academic Press. 18:107-169.


18 Sou. AND CROP SCIENCE SOCIETY OF FLORIDA Hedrick, R. M., and D. T. Mowry. 1952. Effect of synthetic polyelectrolytes on aggregation, aeration, and water relation ships of soil. Soil Sci. i3:427-441. Hemyari, P., and D. L. Nofziger. 1981. Super simper effects on crust strength, water retention, and water infiltration of soils. Soil Sci. Soc. Am. J. 45: 799-801. Henderson, J. C:., and D. L. Hensley. 1986. Efficacy of a hydrophilic gel as a transplant aid. HortScience 21 :991-992. Ingram, D. L., and T. H. Yeager. 1987. Effects of irrigation frequency and a water-absorbing polymer amendment on ligustrum growth and moisture retention by a container medium. J. Environ. Hort. 5:19-21. Johnson, M. S. 1984. Effect of soluble salts on water absorption by gel-forming soil conditioners. J. Sci. Food Agric. 35: 1063-1066. Johnson, M. S., and C.J. Veltkamp. 1985. Structure and functioning of water-storing agricultural polyacrylamides. J. Sci. Food Agric. 36:789-793. Joyner, M. E. B., and D. Z. Haman. 1991. Soil amendments and conditioners for Florida-Literature and research review: The state of the art. South Florida Water Management District. Contract No. 88-152-0506-A3. West Palm Beach, Florida I06pp. Kullmann, A., J. Lehfeldt, and H. Benkenstein. 1986. Homoktalaj fizikai es fizikai-kemiai tulajdonsagainak megvaltozasa szerves gel hatasara. fThe effect of an organic gel on the physical and physicochcmical properties of a sandy soil.] Agrokemia cs Talajtan 35:39-48. Miller, D. E. 1979. Effect of H-SPAN on water retained by soils after irrigation. Soil Sci. Soc. Am. J. 43:628-629. Pill, W. G. 1988. Granular gels as growth media for tomato seed lings. HortScience 23:998-1000. Quastel,J. H. 1954. Soil conditioners. Ann. Rev. Pl. Phys. 5:75-92. Rogers, C. S., and R. C. Anderson. 1981. Prairie grass response to strip mine spoil amended with sewage sludge. Bull. Ecol. Soc. Am. 62:143. Schnitzer, M ., and P. A. Poapst. 1967. Effects of a soil humic compound on root initiation. Nature 213:598-599. Senn, T. L., and A. R. Kingsman. 197 5. A report on hum ate re search. S.C. Agr. Expt. Stn. Res. Ser. No. 165. Clemson, S.C.:Clemson University. G4pp. Still, S. M. 1976. Growth of 'Sunny Mandalay' chrysanthemums in hardwood-bark-amended media as affected by insolubiliLed poly(ethylene oxide). HortScience 11 :483-484. Taylor, K. C., and R. G. Halfacre. 1986. The effect of hydrophilic polymer on media water retention and nutrient availabilitv to Ligustrum lucidum. HortScience 21: 1159-1161. Tu, Z. P., A. M. Armitage, and H. M. Vines. 1985. Influence of an antitranspirant and a hydrogel on net photosynthesis and water loss of cineraria during water stress. HortScience 20:386388. Wallace, A. 1986. A polysaccharide (guar) as a soil conditioner. Soil Sci. 141 :371-373. Wallace, A. 1990. The decade of the 1980s for iron nutrition and interactions in plants. HortScience 25:838. Wallace, A., and G. A. Wallace. 1986. Effects of very low rates of synthetic soil conditioners on soils. Soil Sci. 141 :324-327. Wang, Y-T., and C. A. Boogher. 1987. Effect of a medium-incorporated hydrogel on plant growth and water use of two foliage species. J. Environ. Hort. [,: l 2:,-12i. Whitmore, T. E. 1982. Transplant survival improved. Christmas Trees. 10: 10-11. Willingham, J. E., Jr., and D. L. Coffey. 1981. Influence of hydrophilic polymer amended soil on growth of tomato transplants (abstract). HortScience I 6(3). ACKNOWLEDGMENTS The authors express their appreciation to the South Florida Water Management District for support of this project.


18 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Copper Toxicity and Phosphorus Concentration in 'Florida 502' Oats F. M. Rhoads*, R. D. Barnett, and S. M. Olson ABSTRACT Severe foliage chlorosis has been observed for oats (Avena sativa L.), grown in rotation with tomatoes (Lycopersicon esculentum Mill.) sprayed with copper hydroxide [Cu (OH2)] to control bacterial diseases where soil Cu level was near 100 mg kg-1 It has been reported that Cu toxicity reduced P uptake by crop plants. This research was conducted to determine P concentra tion of oat plants as a function of soil-Cu at three lime rates. Oat plants were grown 7 wk in pots containing 2 kg of soil each from the A horizon of Orangeburg loamy fine sand (fine-loamy, siliceous thermic Typic Kandiudult). Copper rates of O, 100, 200, and 400 mg kg-' from Cu(OH)2 were mixed with the soil along with 0, 3, and 6 g kg-1 of lime to give a factorial arrangement of 12 treatments, with all treatments receiving 100 mg kg-1 P. A treatment containing 200 mg kg-1 P, 400 mg kg-1 Cu and 6 g kg-' lime was also included. Dry-matter yield of oat shoots and roots; tissue content of Cu, P, and Ca; soil-test Cu, P and Ca; and soil pH were determined. Copper concentration in oat plants was positively correlated (P<0.001) with applied Cu, while P concentration, P F. M. Rhoads, R. D. Barnett, and S. M. Olson, North Fla. Res. and Educ. Ctr., Rt. 3, Box 4370, Quincy, FL 32351-9500. Fla. Agric. Exp. Stn. Journal Series no. N-00516. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51: 18-20 ( 1992) uptake, and plant dry weight were negatively correlated (P<0.001) with applied Cu. Yield of oat plants was almost doubled by increasing P from 100 to 200 mg kg-1 with 400 mg kg-1 Cu and 6 g kg-1 lime. The data suggest that Cu toxicity of oats is due to Cu-induced P deficiency. Copper-containing bactericides are used in North Florida tomato (Lycopersicon esculentum Mill.) fields to control bacterial diseases that reduce yield and quality of the staked tomato crop. As much as 30 lb acre-1 Cu may be applied seasonally in cases of severe disease outbreak (Kucharek, 1990). Research results show that Cu moves only slightly in soil and tends to accumulate in the plow layer (Miller et al., 1987). High levels of soil Cu can cause phytotoxicity to many crop plants (Reuther and Labanauskas, 1965). Crops such as oats (Avena saliva L.), wheat (Triticum aestivum L. em Thell.), and corn (Zea mays L.), are grown in rotation with tomatoes in order to minimize loss of tomato yield from soil-borne pests. Oats growing during the fall of 1989 in a field that had previously produced tomatoes, exhibited se-


PROCEEDINGS, VOLUME 51, 1992 19 vere foliage chlorosis where soil Cu levels were near 100 mg kg-1 Previous research showed that soil Cu levels above 100 mg kg-1 were toxic to tomatoes at soil pH levels below 5.5, and that lime reduced toxic ef fects of Cu in soils with pH values below 6.5 (Rhoads et al., 1989). Further studies showed that lime and P reduced toxic symptoms for oats at soil pH levels below 5.5 (Rhoads et al., 1991). An earlier report indicated that high soil Cu levels reduced P uptake by crop plants (Reuther, 1957). The objective of this research was to determine P content of oat plants as a function of soil Cu at three lime rates. MATERIALS AND METHODS Soil used in this experiment was collected from the A horizon in a field of Orangeburg loamy fine sand (fine-loamy, siliceous thermic Typic Kandiudult). Mehlich ]-extractable Cu was 2 mg kg-', extractable P was 90 mg kg-1 extractable Ca was 30 mg kg' and soil pH was 6.5. Copper hydroxide, triple superphosphate, and calcite were added to air-dry soil and mixed thoroughly. Rates of Cu were 0, 100, 200, and 400 mg kg' and rates of calcite were 0, 3, and 6 g kg-1 producing a four by three factorial arrangement containing 12 treatments that received 100 mg kg' Peach. A treatment that received 200 mg kg-1 P, 400 mg kg' Cu and 6 g kg' lime was also included. The soil and nutrient mixes were allowed to incubate in aerated plastic bags for 2 wk after adding 0.15 L water kg' soil. Oat seedlings (cv.'Florida 502') were collected from the field 9 January 1991 and transplanted into pots (4 seedlings pot-1 ) each containing 2 kg (air-dry basis) soil. Each pot received 0.10 L of a solution containing 3.0 g of NH1NO3 L-1 and 8.0 g of KNO3 L'. Plant shoots (above ground portions) were harvested after 7 wk of growth. Roots were separated from the soil by screening and then washed in tap water. Shoots and roots were dried to constant weight at 70C and ground to pass a 20-mesh sieve. Samples of plant tissue were ashed and dissolved in a 0.lM HCl solution. Soil samples were air-dried, and extracted with Mehlich 1 (0.05 M HCl in 0.0125M HS04 ) extractant. Determination of Cu by atomic absorption, Ca by flame emission, and P by the molybdenum blue method was carried out for both the plant and soil extracts. Soil pH was determined in a l: 1 (vol/vol) soil/water suspension. The experimental design was a randomized complete block containing 13 treatments and four repli cates. Analysis of variance procedures, regression analyses, and single degree-of-freedom comparisons were used to determine if differences between treatments were significant (Steel and Torrie, 1960). RESULTS AND DISCUSSION Lime reduced oat dry-matter yield at O and 100 mg kg-1 applied Cu, had no effect at 200 mg kg' Cu, and increased dry matter at 400 mg kg' Cu (Table Table I. Yield and Cu concentration of oat plants and soil pH from pots of soil treated with different levels of Cu, lime and P. Soil treatment Yield Cu Concentration Soil Cu Lime p Tops Roots Tops Roots pH mg kd-1 g kg-1 mg kg-1 -----g pot'-----0 100 200 400 0 0 0 0 100 ll.43at 3.66 100 9.25 C 3.79 100 6.27 e 2.33 100 0.78 g 0.27 -----mg kg-'-----13 35 50 63 33 671 1745 1860 6.5 6.5 6.2 5.4 ------------------------------------------------------------------------------------------0 3 100 8.73 b 3.22 12 38 6.7 100 3 100 6.57 d 3.00 33 482 6.9 200 3 100 6.38e 2.20 49 1190 6.9 400 3 JOO 2.52 f 1.34 64 2827 6.4 0 6 100 7.60b 2.69 15 47 7.1 100 6 100 6.11 cl 2.68 37 476 7. I 200 6 100 4.61 e 1.79 53 1412 7.0 400 6 100 2.28 f 1.02 61 3300 6.7 -------------------------------------------------------------------400 6 200 4.371 I.75 51 3012 6.7 +Effect of lime on top wt was compared within each Cu level. Lime vs no lime was compared as follows: letters a and b were used at the zero Cu level, c and cl at 100 mg kg-1 Cu, eat 200 mg kg-' Cu, and f and g at 400 mg kg-1 Cu. Values within each Cu level followed by different letters are significantly different (P<0.01). "P in creased yield of tops at 400 mg kg-1 Cu and 6 g kg' lime (P< 0.001). 1). Copper concentration in oat tissue was not influenced by lime application. However, soil pH was 6.5 with no lime or Cu added. Addition of Cu at 400 mg kg' reduced soil pH to 5.4 with no lime, whereas lime application maintained soil pH at 6.4 or above for all applied Cu levels. In a previous similar experiment, addition of lime increased oat dry-matter yield and decreased Cu concentration with soil pH below 5.5 and applied Cu at 200 or 400 mg kg-1 (Rhoads et al., 1991). Data from this experiment agree with previous experiments which suggest that lime does not reduce Cu toxicity if unlimed soil pH is above 6.5 (Rhoads et al., 1989, 1991). Increasing applied P from 100 to 200 mg kg' almost doubled the dry-matter yield of oat plants with 400 mg kg' Cu and 6 g kg' lime (Table 1). Copper concentration in oat plants was positively correlated with applied Cu (Tables 1 and 2), while P concentration, P uptake, and plant dry-weight were negatively correlated with applied Cu (Table 2). In each case, the correlation was stronger with soil Cu than with tissue Cu. Since Ca concentration in oats increased with increased applied Cu, the toxicity could not have been due to Cu-induced Ca deficiency (Table 2). However, the fact that plant dry weight, P concentration, and P uptake were all negatively correlated with soil-and tissue-Cu level suggests that Cu toxicity in oats is due to a Cu-induced P deficiency as reported by Reuther (1957). Oat dry-matter yield decreased as applied rates of Cu and lime increased. Increased P in combination with high lime and Cu caused an increase in dry-matter yield of oat plants.


20 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table 2. Regression analyses showing P, Cu, and Ca concentration in oat plants, P uptake by oat plants (mg pot'), and dry weight of oat plants (g pot') as functions of soil Cu (SC) and tissue Cu (TC). The correlation coefficient (r) and the probability (P) of a larger r value are also shown. Function % P = 0.38-0.000405SC % P = 0.43-0.00354TC ppm Cu = 21.57+0.084SC % Ca = 0.20 + 0.002SC P uptake = 32.22-0.057SC P uptake= 40.63-0.526TC Plant wt. = 8.97-0.013SC Plant wt. = 10.83-0.12TC r -0.852 -0.759 +0.826 +0.819 -0.899 -0.851 -0.866 -0.804 p <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 REFERENCES 1. Kucharek, T. I 990. Plant protection pointers. Extension Plant Pathology Report No. 6, 34 p. Univ. of Florida Coop. Ext. Serv. 2. Miller, W. P., D. C. Martens, and L. W. Zelazny. 1987. Shortterm transformation of copper in copper-amended soils. J. En viron. Qua!. 16:176-181. 3. Reuther, Walter. 1957. Copper and soil fertility. p. 128-185. In Alfred Stefferud (ed.) Soil: Yearbook of Agric. U.S. Dept. of Agric., Washington, D.C. 4. Reuther, W., and C. K. Labanauskas. 1965. Copper. p. 157179. In H. D. Chapman (ed.) Diagnostic criteria for plants and soils. Univ. of California, Riverside. 5. Rhoads, F. M., S. M. Olson, and A. Manning. 1989. Copper toxicity in tomato plants. J. Environ. Qua!. 18:195-197. 6. Rhoads, F. M., S. M. Olson, and R. D. Barnett. 1991. Soil contamination from copper pesticides. p. 20-24. In A. B. Bottcher (ed.) Environmentally sound agriculture. Proc. Con ference, Orlando, FL. 16-18 April 1991. Florida Coop. Ext. Serv. IFAS, Univ. of Florida. 7. Steel, R. G. D., and J. H. Torrie. 1960. Principles and procedures of statistics. McGraw-Hill, NY.


20 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA M ycorrhizal Amelioration of the Detrimental Effect of Biodune on Plant Growth T. Aziz and D. M. Sylvia* ABSTRACT The instability of sand dunes and beaches due to erosion is a severe problem in many coastal areas. Pfizer Inc. has developed a sand-stabilizing gel called Biodune which helps to protect sand dunes from erosion; however, establishment of certain plant species in treated sand has been unsuccessful. A greenhouse experiment was conducted to evaluate the effect of Biodune on the symbiotic interaction of two coastal plant species (Panicum amarum Ell. and Uniola paniculata L.) with a vesicular-arbuscular mycorrhizal (V AM) fungus. Biodune was mixed with pasteurized sand at three rates, inoculum of a Glomus sp. was placed in one-half of the pots, and seeds of test plants were sown. Plants were fertilized with 1/4-strength Hoagland's solution every other day for IO wk and harvested 2 wk thereafter. Plants grown in Biodune-treated sand were smaller and had less shoot-P uptake than plants grown without Biodune. Percentage mycorrhizal col onization of U. paniculata roots was greater in Biodune-treated sand than nontreated sand while mycorrhizal colonization of P. amarurn was not affected. Plants with V AM colonization had greater total dry weight and shoot-P uptake compared to non mycorrhizal plants at all the levels of Biodune tested. We conclude that V AM inoculation offsets some of the deleterious ef fects of Biodune on plant establishment in dune sand. Coastal plants have an important role in protecting sand dunes and beaches from erosion. Establishment of plants in coastal regions, however, may be difficult because of the severe environment and dune instability. Use of beneficial microorganisms, such as T. Aziz and D. M. Sylvia, Soil and Water Science Dep., 2169 McCarty Hall, Univ. of Florida, Gainesville, FL 32611-0290. Florida Agric. Exp. Stn. Journal Series no. R-02147. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 :20-23 (1992) V AM fungi, has been suggested for rapid establishment of plants in coastal sands because they reduce environmental stresses (Koske and Polson, 1984; Syl via and Williams, 1992). Sylvia ( 1989) reported that when the pioneer dune plant, U. paniculata, was colonized with a V AM fungus in a nursery and then transplanted to a beach, it grew more rapidly than when not inoculated. In contrast to biological amendments, Pfizer Inc. (Pfizer Central Research, Eastern Point Road, Groton, CT 06340) has developed a gel-like substance called Biodune for the stabilization of coastal dunes. Biodune is a biodegradable, aqueous polymer gel which consists of a gelant (Biodune IOOPS) and a crosslinker (Biodune IOOXS). This material is mixed with beach sand and water to form a sand/gel composite. In field trials, this composite has provided protection to coastal sand dunes from winds and waves; however, it has been difficult to establish certain plant species on dunes treated with Biodune (Auerbach et al., I 990). Vesicular-arbuscular mycorrhizal fungi may help establish plants in dunes treated with this chemical. The o~jective of this study was to investigate the influence of Biodune on the symbiotic asso ciation between V AM fungi and two coastal plant species, P. amarum and U. paniculata. MATERIALS AND METHODS Biodune was mixed with pasteurized (8SC for 6 h, treated twice) sand (collected from Anastasia State Recreation Area, St. Augustine Beach, FL) and water


PROCEEDINGS, VOLUME 51, 1992 21 at 0.25, 0.5, and 1 times the recommended rate. The recommended rate was; 284 g of Biodune 100 PS, 168 g of Biodune 100 XS, and 2500 mL of cool water mixed with 7.5 kg of dry sand. The materials were mixed manually to a crumbling,jelly-like texture. Approximately 380 mL of Biodune-treated or nontreated sand was placed in each DeePot (600 mL ca pacity, J. M. McConkey & Co. Inc., 1615 Puyallup Street, Sumner, WA 98390) that contained sterile gravel at the bottom to prevent soil loss through the drainage holes. Prior to mixing sand with Biodune, inoculum of a VAM fungus (Glomus sp., INV AM isolate 925) was mixed with the sand for the mycorrhizal treatment at the rate of 5 g per 100 g of sand. The inoculum consisted of spores, fungal hyphae, and colonized root pieces from a 6-mo-old pot culture of the fungus, stored for about I yr at 4C. Seeds of P. amarum (bitter panicum/beach panic grass, Accession No. Atlantic 421136) were obtained from the USDA Plant Materials Center, 14119 Broad St., Brooksville, FL 34601. Seeds of U. paniculata (sea oats) were obtained from Horticultural Systems, Inc., Golf Course Road, Parish, FL 34219. Seeds were surface disinfested by immersion into a NaOCl solution (2.5 mL L-1 water) for 15 min, followed by six rinses with water. Approximately IO seeds were placed in each DeePot at a depth of 3 cm below the surface. Seeding was completed within 15 min of mixing Biodune with sand and water. The experiment was arranged in a completely randomized design with two V AM inoculation treatments ( + or-) and the following Biodune treatments: none, recommended rate (X), and 0.5X and 0.25X for P. amarum; and none and the recommended rate for U. paniculata. There were 8 replicates per treatment. Plants were grown in a greenhouse with mean max. and min. temperatures of 32.5 and 23.3C, respec tively, and mean max. PAR of 1485 mol m2 s-1 All plants were watered at the same time as needed and fertilized with 1/4-strength Hoagland's solution (Hoagland and Arnon, 1950) every other day starting 2 wk after seeding through 10 wk. Plants were harvested 12 wk after planting. Shoot and root weights were determined after drying sam ples at 70 C for 48 h. Percentage V AM colonization was determined by the gridline-intersect method (Giovannetti and Musse, 1980) after clearing and staining roots according to Phillips and Hayman (1970). Shoot-P content of P. amarum was determined colorimetrically (Murphy and Riley, 1962) after dryashing samples in a muffle furnace at 500 C. Data were analyzed by ANOV A (SAS Institute, Inc., 1985) and standard errors were calculated for the means. RESULTS AND DISCUSSION Biodune had no significant effect on percentage mycorrhizal colonization of P. amarum (Fig. la), but did increase percentage colonization of U. paniculata (Table 1). However, application of Biodune at the recommended rate reduced total root masses colonized by V AM fungi from 61 to 15 mg plant in P. amarum, and from 13 to 7 mg plant in U. paniculata (data derived from Fig. la, 2c and Table 1). No colonization was observed in the noninoculated plants. Shoot-P concentration was less in noninoculated P. amarum than inoculated plants when Biodune was used at the recommended rate (Fig. lb). In inoculated P. amarum, shoot-P concentration decreased with the first application, but increased at the higher rates of Biodune. There was a significant interaction between Biodune treatment and mycorrhizal inoculation on shoot-P concentration. Shoot-P uptake decreased with increasing concentration of Biodune (Fig. le). The decrease, however, was most pronounced with the first addition of Biodune. In the presence of Biodune, shoot-P uptake of the mycor rhizal P. amarum was higher than the nonmycorrhizal plants (P=0.0215). Shoot height of P. amarum decreased with increasing concentration of Biodune (Fig. 2a), and at higher concentrations, shoot height was greater for mycor rhizal plants than for nonmycorrhizal plants. Similar results were observed for shoot and root dry weights (Fig. 2b,c). The analyses of variance showed highly significant (P-s0.01) effects of both treatments (Biodune and mycorrhizae) on the above mentioned growth variables. The root to shoot ratio of P. amarum increased with increasing concentration of Biodune (Fig. 3). Mycorrhizal plants had a lower root to shoot ratio than nonmycorrhizal plants, especially at the higher Biodune concentrations. Similarly, Biodune depressed shoot and root dry weights of U. paniculata (Table 1). Mycorrhizal U. paniculata had greater total dry weight than nonmycorrhizal plants when Biodune was added. Mycorrhizal inoculation also re-Table I. Effect of Biodune and mycorrhizal inoculation Uniola paniculata growth in the greenhouse for 12 wk. Biodune Myc. Root Shoot dry Root dry rate moc. colonization Height weight weight R:S ratio o/r_ cm mg plant' mg plant' 0 + 10 3+ 32.0 6.1 140 84 86 54 0.61 0.10 0 31.8 7.4 161 140 85 78 0.53 0.06 IX + 27 4 33.4 11 72 10 26 7 0.36 0.06 0 0 23.7 1.2 41 7 23 4 0.57 0.04 + values are the mean of 8 replicates SE


22 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA ,.,.----__ 50 cR z 40 0 I-a 1 <( 30 -N z 0 20 -_J u z 0 0 u I-0 0 O:'. 10 0 4 u~ 3 o_ 0, f-E 2 0 '--" w :x::: <( l-0 I (_/) o_ C =i 0 o_ Q_ 0 1.0 0.8 0.6 1-o, 0.4 o E 0 I (/] 0.2 O.C 0 Inoculated Noninoculoted \ \ -, --C --. ----'-------'-----~---L___ 0 oc 0.25 0.50 0.75 1.00 BIOOUNE CONCENTRATION Fig. I. Effect of Biodune on (a) mycorrhizal colonization of root, (b) shoot-P concentration, and (c) shoot-P uptake of VAM inoculated and noninoculated Panicum amarum. Values represent the mean of 8 replicates SE. duced the root to shoot ratio of U. paniculata in the presence of Biodune. Lack of growth response of non-Biodune treated plants to V AM inoculation was due to the high level of P fertility maintained in pots. Phosphorus was not limiting the growth of these plants. The increasing shoot-P concentration in the inoculated P. amarum plants with increasing levels of Biodune is characteris tic of a mycorrhiza-mediated response. The reduction of shoot-P concentration with Biodune addition in the noninoculated plants reflects the severe inhibitory effect of Biodune on P uptake in the absence of mycorrhizae. These results indicate that, despite an inhibitory effect of Biodune on mycorrhizal activity, V AM fungi are able to aid plants in the uptake of P E u ----I-I C) w I I-0 0 I (/) C 0 a. '--CJ'I ----...: >-Cl:'. 0 I-0 0 I (/) ,...._ +-' C 0 a. '--CJ'I ----...: >-a:: 0 I-0 0 a: RJ ~------------------, fi (] 40 2:J () 0 jQ 0.25 0.20 0 15 0 0 C_ CS (] (l CJ 0.25 0.20 -0.' 5 0.10 -0.05 0.00 o Inoculated Noninoculated \ \ \ ----...... t t~ a b C %=-............. --I 0 00 0.25 0.50 0 7'j 1.00 BIODUNE CONCENTRATION Fig. 2. Effect of Biodune on (a) shoot height, (b) shoot, and (c) root dry weights of VAM inoculated and noninoculated Panicum amarum. Values represent the mean of 8 replicates SE. from Biodune-treated sand leading to enhanced plant growth. Biodune likely also has other deleterious effects on plants. Both plant species treated with the chemical had brown leaf tips. This "burning" was observed in only a few plants at the lowest concentration of Biodune whereas at the highest concentration most plants had the symptom. Although the exact reason for this "burning" effect is not known, we speculate that the material holds water tightly precluding plants from obtaining the water and causing effects similar to saline/drought conditions. This could be the reason


PROCEEDINGS, VOLUME 51, 1992 23 2.0 0 f---~-f <( 1. 6 Ct::' f-1.2 ~-, 0 ;>::::~ 0 -4;'-+' ::r:: 0.8 Ul f-0 I -1 c > c~ !,_,1 i ::ed 0 0.4 0 r i ,; ; r :1 cu Int c d Ct::' 0.0 0.00 0.25 0.50 0. 75 1.00 BIODUNE CONCENTRATION Fig. 3. Effect of Biodune on the root:shoot ratio of V AM inoculated and noninoculated Panicum amarum. Values represent the mean of 8 replicates SE. why, despite significant increases in dry matter production due to mycorrhizal inoculation, only partial recovery from Biodune-induced growth depression could be achieved. In conclusion, we found that Biodune depressed the growth of P. amarum and U. paniculata; however, growth depression was decreased by V AM inocula tion. We recommend that VAM fungi be tested further for possible use with Biodune to encourage plant establishment in sand dunes treated with this chemical. REFERENCES Auerbach, M. H., L. M. Ehrhart., 0. !vi. Bundy, and B. L. Edge. 1990. Environmental compatibilitv of a sand/gel composite system for coastal dune stabilization. p. I 69-184. In L.S. Tait (ed.) Beaches: Lessons of Hurricane Hugo. Proceedings of the Third Annual National Conference on Beach Preservation Technology, Florida Shore and Beach Preservation Association, Tallahassee, FL. Giovanetti, M., and B. Mosse. 1980. An evaluation of techniques for measuring arbuscular mycorrhizal infection in roots. '-Jew Phytol. 84:489-500. Hoagland, D. R., and D. I. Amon. 1950. The water-culture method for growing plants without soil. Cir. Calif. Agric. Exp Stn., No. 347. Koske, R. E., and W. R. Polson. 1984. Are VA mycorrhizae required for sand dune stabilization? BioScience. 34:420-424. Murphy, j., and J. P. Riley. 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta. 27:31-36. Phillips, J. M., and D. S. Hayman. 1970. Improved procedures for clearing roots and staining parasitic and vesicular-arbuscular mycorrhizal fungi for rapid assessment of infection. Trans. Br. Mycol. Soc. 55: 158-161. SAS Institute, Inc. 1985. SAS User's guide: Statistics. SAS Inst. Inc., Cary, NC. Sylvia, D. M. 1989. Nursery inoculation of sea oats with vesiculararbuscular mvcorrhizal fungi and outplanting performance of Florida beaches. J. Coastal Res. 5:747-754. Sylvia, D. M., and S. E. Williams. 1992. VA mycorrhizae and environmental stresses. In R.G. Linderman and G. Bethlenfalvay (ed.). VA mycorrhizae and sustainable agriculture. ASA, CSSA, SSSA Spec. Pub!., Madison, WI. (in Press).


PROCEEDINGS, VOLUME 51, 1992 23 Calibration Parameters for Radar Rainfall Estimation S. F. Shih ABSTRACT The large spatial and temporal variability of rainfall distribution is difficult to estimate accurately. Although a dense raingage network can be used to improve the accuracy of rainfall estimation, a dense network is costly to maintain and operate. Alternatives that can be used to estimate rainfall in a relatively large area, such as radar systems, should be investigated. However, the process for determination of the calibration parameters ( a and b), used in the radar echo (Z) -rainfall rate (R) relation are still in the research and development stages. In this paper, the avail able literature concerning these a and b values has been reviewed and analyzed according to eight rainfall types (i.e., average, various, thunderstorm, convective, continuous/steady, stratiform, orographic, and drizzle). The results showed that the coefficient of variation for a is much greater than that for b. This implies that further research should focus more on the determination of the a value than on the b value. Both a and b S. F. Shih, Agricultural Engineering Dep .. Univ. of Florida, Gainesville, FL :12611-0570. Florida Agric. Exp. Stn. Journal Series no. R-01792. Contribution published in Soil Crop Sci. Sor. Florida Proc. 5 I :2'.l-29 ( I 992) are significantly different among most of the rainfall types. It appears that several values of a and bare required to accomplish the rainfall estimation for a given area. Rainfall is important to several disciplines, including meteorology, climatology, hydrology, and agriculture. It has practical applications in hydrology and agriculture as a quantifier, predictor, and basis for modeling. Accurate knowledge of the spatial and temporal distribution of rainfall is important to planning and management of water resources. Water resources management is important to Florida because: (1) agriculture is a key industry and a major water user; (2) population growth rate predicts an increase in municipal/residential water consumption; and (3) salt-water intrusion poses a threat to water supply in coastal areas (Wilson, 1982). The large spatial and temporal variability of rain fall distribution requires a dense raingage network if acceptable accuracv of rainfall estimation is to be


24 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA achieved. However, Shih ( 1982) indicated that there is a nonlinear relationship between the increase in number of gages and the improvement in accuracy, and the improvement is limited to the reduction of temporal variation. A dense network is also costly to maintain and operate. Additionally, Creutin and Obled (1982), and Tabios and Salas (1985) compared various methods for point rainfall estimation using raingage data. They concluded that none of the methods examined were able to fully account for the statistical properties of the observed rainfall fields. Moses and Barrett ( 1986) mentioned that among the quantities needed as inputs to water-budgeting models, rainfall is the most variable, and the most difficult to obtain with sufficient accuracy because of its high variability both spatially and temporally. The National Weather Serv ice Network reports that there are approximately 110 rainfall stations for the 58,568 mi2 (1.5 X 105 km2 ) of the state of Florida. This means that the typical raingage network in Florida is only one raingage per 530 mi2 (1.37 x 10' km2). This raingage density will lead to a ratio of areas of about 0.35 ft2 (i.e., standard 8 -in diameter of raingage) to 530 x 640 x 43560 ft2 or approximately 1: 4 x 10w. This ratio illustrates the accuracy problem involved in raingage data. Furthermore, measurement of rainfall by gages is affected in particular by the interrelated factors of topography, site, wind and gage design (Barrett and Martin, 1981). The gage catch may be representative of a small or large area depending upon slope, aspect, elevation, and location in relation to hills and ridges. Due to these inherent limitations of the raingage design for rainfall measurements, point data could contain ambiguities producing a decrease in accuracy of regional estimates. Thus, an alternative method such as a radar system should be investigated further. One important procedure involved in radar rainfall estimation is to properly choose the parameters used in the radar reflectivity factor (Z) -rainfall rate (R) relation which is called the Z-R relation. This relation involves two calibration parameters called a and b. Battan (1973) provided an extensive summary of the a and b values reported in international literature up to 1970. However, Battan did not provide information after 1970, and Wojtiw (1987) indicated that although the equations listed in Battan ( 1973) differed markedly at small and large values of R, with the exception of curves for orographic rains, most of them do not differ greatly for rainfall intensities between 1 and 8 in. h-1 The errors and problems involved in radar rainfall estimation has been discussed by several researchers (Damant et al., 1983, Dalezio, 1988). In general, the applicability of radar systems to rainfall estimation is still in the research and development stages. Particularly, the calibration parameters must be pre-determined before the radar system can be implemented in an area for rainfall estimation. Therefore, the purposes of this study were: (1) to review the literature available after 1970 concerning the radar Z-R relation; (2) to analyze the calibration parameters as a function of rainfall type; and (3) to discuss the potential application of these findings to the Florida situation. MATERIALS AND METHODS Radar Method Gilman ( 1964) indicated that the ability to determine the areal distribution of precipitation intensity depends on the type of radar employed, and that the best all-purpose weather-search radar would have the following properties: (a) wavelength such that the ef fect of precipitation attenuation is minimized, (b) power and pulse length selected to ensure that the lowest significant amounts of precipitation are detected at maximum range, (c) correction for range, (d) beam width as narrow as possible, (e) antenna large enough to receive the weakest possible reflected energy. Miller and Thompson (1975) mentioned that the amount of backscatter from raindrops depends strongly on their size and the wavelength of the radio waves. For example, with a radio wavelength of close to 1 cm, raindrops and snow crystals having a diameter of 1 mm or more can be easily detected. The much smaller cloud droplets and ice crystals (diameter < 0.2 mm) (1 in = 25.4 mm) scatter so little energy that only extremely powerful and sensitive radar systems can "see" them except at very short wavelengths (< 1 mm). But a sufficiently powerful radar system that generates and receives such extremely short wavelengths is very difficult to build. Several methods of using radar to estimate rainfall have been summarized by Doviak ( 1983). The reflec tivity method was used in this study. Reflectivity Method Unfortunately, there is no analytical relationship connecting the estimated radar reflectivity (Z, mm6 / m3 ) (1 in = 25.4 mm) to the rainfall rate (R, mm/ hour). However, an empirical formula called the Z-R relation, which is commonly used to estimate rainfall intensity from the radar echo, is described by the geometric function (1) Where a and b are calibration parameters to be determined. Battan ( 1973) provided various relations reported in international literature up to 1970 (Table 1). In Canada, Wojtiw (1987) compared maximum rainfall amounts derived from radar data to those recorded by surface raingages using three different Z-R relations developed by Marshall and Palmer (1948) with Z = 200 R160, Hood (1950) with Z = 295R161 and Wojtiw (1984) with Z = 168R172 Wojtiw (1987) suggested that the Wojtiw Z-R relations gave overall estimates of maximum rainfall amounts that were closer to ground measurements than either of the other relations. The Marshall and Palmer (1948) Z-R relationship was used by Damant et al. (1983) for studying the errors involved in the Thiessen technique for estimat-


Data Source No. 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 PROCEEDINGS, VOLUME 51, 1992 25 Table 1. Empirical relations between reflective factor Z(mm/m'), and rainfall intensity R(mm/hr) (from Battan 1973). Z value Location 320R,.44 Washington, D.C. 214RJ.5 H Washington, D.C. 224R1 4 Ynyslas, Great Britain 630R,.45 Shoeburyness, England 208R,.53 Hawaii 190R,." Various locations 220R,.60 Various locations 295R1.r.12 Canada l80R1 55 Cambridge, Mass. l 27R'"' Australia 16.6R,.55 Hawaii 3 IR,.'' Hawaii 290R,.4 Hawaii 396R,.35 Central Illinois 486R1.37 Central Illinois 380R,.24 Central lllinois 313R,."' Central Illinois 150R,.54 Mount El'brus, USSR 257R,."' Mount El'brus, USSR 398R,.47 Mount El'brus, USSR 162R116 Lexington, Mass. 215R,.34 Lexington, Mass. 350R,.42 Lexington, Mass. 31 0R,.34 Lexington, Mass. 220R,.54 Leningrad, USSR 303R" Near Moscow, USSR 405R1.9 Near Moscow, USSR 289R,.'" Near Moscow, USSR 109R164 Kandia, India 342R142 Delhi, India 700R,.6 Tokyo,Japan 300R,.6 Tokyo,Japan 200R,., Tokvo,Japan 200R,., Tokyo.Japan 219R,.4 Poona, India 67.6R1.94 Poona, India 66.5R,.92 Poona, India 204R,.'0 Kiev, USSR 205R!.4 H Mostly Miami, Fla. 300R,."' Mostly Miami, Fla. 450R,.16 Mostly Miami, Fla. 184R,.28 Various locations 278R,.'0 Entebbe, Uganda 240R,.'0 Lwire, Congo I 76R118 Palma 151R,."' Barza, Italy l 79R,.2 Karlsruhe, Germany 227R1 '11 Karlsruhe, Germany l 78R,." Karlsruhe, Germany 150R,.23 Karlsruhe, Germany 137R,.36 Axel Heiberg Land 330R,.11 Chernozem, USSR 298R,.46 Vashnevo, l1SSR 520R,.81 Tucson, Arizona 730R155 France 255R,.4 France 426R,.'0 France 286R,.4 Miami. Florida 22 IR,.32 Majuro, Marshall Islands 30IR1.4 Corvallis, Oregon 311R,.41 Bogar, Indonesia Remarks 8 rain intensities, each a mean of about IO storms of same intensity 98 Storms-original data 5 rainstorms 4 rainstorms 50 storms, orographic rain Various types of rain Various types of rain 270 samples, 7 rainstorms; light rain l-3rnm/hr; heavy thunderstorms 50mm/hr 63 rain samples, widespread rain both uniform and variable; showers and thunderstorms Showers, 8 months of observation Orographic rain within cloud Orographic rain at cloud base Nonorograhic rain-thunderstorms 1.270 I-minute observations-all rains 560 I-minute observations-thunderstorms 330 I-minute observations-rain showers 380 I-minute observations-continuous rain Rain (melted granular snow and strongly granulated particles), 344 spectra, 6 rains Rain (melted snow of average granulation), 367 spectra, 7 rains Rain (melted non-granulated snow), 140 spectra, 4 rains Stratiform rains, 16 April 1954 Stratiforrn rains, 23 April 1954 Stratiform rains, 27 April 1954 Strati form rains, 28 April 1954 Showers and steady rain Various types ofrain, R <7 mm/hr Various types ofrain, 760 mm/hr Orographic, Monsoon rains Nonorographic, Monsoon rains One day, probably warm rain One dav, continuous rain Air ma;s showers Prewarm frontal rain Thunderstorms S tea

26 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table I. (cont.) Empirical relations between reflective factor Z(mm/m'), and rainfall intensity R(mm/hr) (from Battan 1973). Data Source No. Z value Location Remarks No. G2 63 64 6!> 66 67 68 G9 267R' ,, 230RL40 372Rl47 593R''' 256R' 41 140Ru, Woodv Island, Alaska Franklin. North Carolina Champaign, Illinois Flagstaff, Arizona 250R'' 500R11 Island Beach, l\.j. Locarno-Monti, Swit. Locarno-Monti, Swit. Locarno-Monti, Swit. ing areal rainfall over the Yamaska Basin in Canada. They found that the errors for 13 analyzed storms had values between 3% and 69% for the entire basin. For individual subcatchrnents, which varied in size between 6 mi" and 500 mi", errors were as high as 380%. However, errors when radar rainfall was greater than 1.18 inches in a subcatchment reached up to 57%. The Richards and Crozier (1983) Z-R relationship, Z = 205R' '", has been recently used by Dalezio ( 1988) in the Grand River Basin of southern Canada for evaluating the accuracv of radar rainfall measurements in operational hydrology. Several rainfall analyses (bivariate statistical analysis, radar univariate analysis, the reciprocal-distance raingage interpolation model, the Brandes 1975 field adjustment procedure, and the raingage analysis) were compared to evaluate radar hydrology. He concluded that most of the techniques in the comparison performed simi larly, with the exception of the radar univariate analysis which was considered unsatisfactory. He also concluded that, in general, radar tends to underestimate heavy rainfall and to overestimate light rainfall. This implies that a set of a and b values may be unsuitable for covering the entire range from light to heavy rainfalls. This is in accordance with previous findings and suggests that the radar systematic bias still deserves special attention in order to improve rainfall Drizzle Widespread rain Thunderstorm rain estimates. In fact, none of the compared methods can fully account for the radar systematic bias. Calibration Parameter Analysis Since large variations exist in the calibration parameters as shown in Table I, it is extrernely difficult to choose the proper a and b values to be used in Z-R relation. However, based on the existing Z-R re lations, there may have been some patterns of a and b values which are related to different types of rainfall. Thus, as Table 1 shows, eight rainfall types, which are categorized as average, various, thunderstorm, convective, continuous/steady, stratiform, orographic, and drizzle, were analyzed in this study. RESULTS AND DISCUSSION Calibration Parameters a and b Sixteen past-1970 Z-R relations are summarized in Table 2. As Tables 1 and 2 show, several observations can be made. First, the a values varied from 16.6 to 730.0 with an average of 278.3, and the b values varied from 1.16 to 2.87 with an average of 1.492. This large variation in both a and b implies that it is very difficult to choose their proper values. Second, both a and b varied not only with rainfall Table 2. Z-R relations for radar rainfall estimation since 1970. Data Source No. Z value Reference Remarks 70 300R1 ',ll Joss and Waldvogel (1970) Canada 71 300Rl40 Woodley et al. (1975) Florida, Convective rain 72 l55R1"" Cain and Smith (1976) 73 200R'" Barrett and Curtis (1976) Orographic rain 74 486R'" Barrett and Curtis (1976) Thunderstorms 75 32lRL44 Atlas and Ulbrich (1977) Julian elate l 74, Switzerland 7li 366R' sq Atlas and Ulbrich (l 977) Julian date l 95, Switzerland 77 456R' 47 Atlas and Ulbrich ( 1977) Julian date 220, Switzerland 78 377RL42 Atlas and Ulbrich ( l ~177) All three clays, Switzerland 79 295Rl43 Richard and Cozier ( l 983) Southern Ontario, Canada 80 l68R '" Wojtiw (I 981) Alberta, Canada 81 230RL65 Boucher and Wieler ( 1985) Massachusetts 82 230R1 '11 Austin ( 1987) All types of rain, :\/ ew England 83 400R'"0 Austin (1987) Thunderstorms, New England 84 230R' '" A us tin (1987) Convective rain, New England 85 JOOR'"' Austin (1987) Drizzle rain, New England


PROCEEDINGS, VOLUME 51, 1992 27 Table 3. Mean, standard deviation and coefficient of variation of calibration parameters used in Z-R relations for different types of rainfall. a Rainfall Data' type source no. x SD General 1, ... 85 278.3 136.4 Various rain 2,6, 7,9,25,38 233.6 47.3 Thunderstorm 8, 13, 15, 30, 35, 41,54,69, 74,83 398.6 106.0 Convective 16, 30, 40, 58, 71, 84 306.3 Continous/steady 17,32,36,39,68 227.1 Stratiforrn 21,22,23,24 259.3 Orographic 5,29, 73 172.3 Drizzle 67, 85 120.0 'Data source numbers are given in Tables 1 and 2. types but also with season, location, topography, and land use/cover condition. Third, since both a and b are influenced by so many variables, the type of rainfall appears to be a most important variable for helping to pre-determine which values of a and b should be used. Unfortunately, the result of radar rainfall estimation can be evaluated only by its relative agreement with other sensors of rainfall measurements. At present, there is no single sensor which can claim high accuracy in rainfall estimation. Although the raingage data could contain ambiguities in regional rainfall estimation, its data continues to be used to provide point information at ground level. Thus, if there is a significant discrepancy between the radar rainfall measurement based on rainfall type selection and the raingage data, both a and b values could be re-adjusted according to other parameters (i.e., season, location, topography, and land use/cover condition). In other words, several values of a and b may b~ required to accomplish the rainfall estimation in a given area. Rain{ all Types Influencing Calibration Parameters The mean, standard deviation, and coefficient of variation for the a and b values of eight rainfall types are given in Table 3. Several observations can be 51.0 98.9 86.1 55.0 28.3 Calibration eararneters b CV x SD CV 49.0 1.492 0.221 14.8 20.2 1.546 0.109 7.1 26.2 1.406 0.153 10.9 16.7 1.343 0.109 8.1 43.5 1.554 0.251 16.2 33.2 1.315 0.110 8.4 31.9 1.627 0.091 5.6 23.6 1.450 0.071 4.9 made. First, the mean values of a varied from 120 for drizzle to 398.6 for thunderstorm. The difference is about 230%. The b value for stratiform rain was 1.315, which was only 24% difference from 1.627 for orographic rain. This implies that it is more difficult to choose the a value than the b value. Second, the coefficients of variation for the a value of the eight rainfall types are all greater than 16%, while those for the b value of only three of the rainfall types are greater than 10%. This implies that the error as sociated with the selection of the b value is much smaller than that for the a value. In other words, future research into calibration parameter determination in the Z-R relation should focus more on the a value than the b value. Comparison of the Calibration Parameters Difference Among Rainfall Types The mean value differences among the different types of rainfall were examined using a t-test. The results of the t-test are listed in Table 4 and 5. Several observations can be made as follows: For calibration parameter a value (Table 4): 1. The a value was significantly lower for average rainfall than for thunderstorm, but higher Table 4. Mean-value differences of the calibration parameter, a, among different types of rainfall using t-test. Rainfall Various Thunder-type ram storms Average NS Various rain Thunderstorm Convective Continuous and steady Stratiform Orographic *significant difference at 0.05 level. **significant difference at 0.0 I level. NS non-significant difference. ** ** Convective NS KS Continuous and steady Stratiform Orographic Drizzle NS NS NS ** NS NS NS ** ** ** ** NS NS ** ** NS NS NS NS NS NS


28 SOIi, A'.\:D CROP SCIE:'\:CE SOCIETY OF FLORIDA Table 5. Mean-value differences of the calibration parameter, b, among different types of rainfall using t-test. Rainfall Various Thunder-type rain storms Average NS Various rain Thunderstorm Convective Continuous and stead\ Stratiform Orographic *significant difference at 0.05 level. **significant difference at 0.01 level. NS non-significant difference. NS NS Convecti\'e NS ** NS than for drizzle; and there was no difference among the other five types of rainfall. 2. Various types of rainfall had a significantly lower a-value than did thunderstorm and convective rainfall, but higher a-value than drizzle; and there was no difference among the other three types of rainfall. 3. Thunderstorm a-values were significantly higher than those for continuous/steady rain, stratiform rain, orographic rain, and drizzle. 4. Convective rain had a significantly higher avalue than did orographic rain and drizzle. 5. All other tests showed no difference. For calibration parameter b value (Table 5): 1. Various types of rainfall had a significantly higher b-value than did convective rain. 2. Orographic rain had a significantly higher bvalue than did convective rain, stratiform rain, and drizzle. 3. All other tests showed no difference. Practical Aj,f1lication The results of this study can be used to improve the application of the radar rainfall estimation technique in Florida in three ways: First, as discussed earlier, the calibration parameters a and b must be predetermined before the system is implemented. According to Miller and Thompson (1975), the average annual number of days with thunderstorms is about 80 clays for some areas of Florida. In other words, both a and b under the thunderstorm category should be the values first considered for use in Florida, followed by those under the convective rain category. Second, the vegetation condition in Florida undergoes seasonal changes which may influence the a and bvalues. Third, since the area of Florida covers about 58,568 square miles, this large area may affect the regional applicability of a and b values. Thus, the final values chosen for a and b in Florida can reflect not only the rainfall types but also seasonal vegetation condition and the site of interest. REFERENCES Atlas, D., and C. W. Ulbrich. 1977. P"!1-ami arpas-intcgratcd rainfall measurement bv microwave attenuation in t~e 1-3 cm Band. J. Applied Meteorology I 6: 1322-1331. Continuous and steady Strati form Orographic Drizzle NS NS NS NS NS NS NS NS NS NS NS NS NS NS ** NS NS NS NS ** NS ** Austin, P .\1. 1987. Relation between radar reflectivity and surface rainfall. Monthly Weather Review 115: l053-1070. Barrett, E. C., and L. F. Curtis. 1976. Introduction to Environmental Remote Sensing. John Wiley & Sons, Inc. New York, N.Y. Barrett, E. C., and D. W. Martin. 1981. The Use of Satellite Data in Rainfall Monitoring. Academic Press, New York 340 pp. Battan, L. J. 1973. Radar Observation of the Atmosphere. Univer sity of Chicago Press 324 pp. Boucher, R. J., and J. G. Wieler. 1985. Radar determination of snowfall rate and accumulation. J. Climate Applied Meteorol ogy 24:6873. Cain, D. E., and P. L. Smith. 1976. Operational adjustment of radar estimated rainfall with raingage data: a statistical evalua tion. Preprints 17th Conf. Radar Meteorology, Seattle, Am. Meteorology Soc. 533-538. Creutin, J. D, and C. Obied. 1982. Objective analysis and mapping techniques for rainfall field: an objective comparison. Water Resources Res. 18(2):413-431. Dalezios, N. R. 1988. Objective rainfall evaluation in radar-hydrology. J. Water Resources Planning and Management. Am. Soc. Civil Engineers 114(5):531-546. Damant, C., G. L. Austin, A. Bellon, and R. S. Broughton. 1983. Errors in the Thiessen technique for estimating areal rain amounts using weather radar data. J. Hydro!. 62:81-94. Doviak, R. J. 1983. A survey of radar rain measurement techniques. J. Climate, Applied Meteorology 22:832-849. Gilman, C. S. 1964. Rainfall. In Handbook of Applied Hydrology, ed. by V. T. Chow, McGraw-Hill Book Co., pp. (9-1)-(9-68). Hood, A. D. 1950. Quantative measurements at 3 and 10 cm of radar echo intensity from precipitation. Report No. 2155, Na tional Research Council of Canada, Toronto, Canada. Joss, .J-, and A. Waldvogel. 1970. A method to improve the accu racy of radar measured amounts of precipitation. Preprints 14th ConL Radar Meteorology, Tucson, Am. Meteorology Soc. 237-238. Marshall, J-S. and W. M. K. Palmer. 1948. The distribution of raindrops with size. J. Meteorology 5:165-166. Miller, A., and J.C. Thompson. 1975. Elements of Meteorology. Charles E. Merrill Publishing Company, Columbus, Ohio, 362 pp. Moses, J. F., and E. C. Barrett. 1986. Interactive procedures for estimating precipitation from satellite imagery. In Hydrologic Application of Space Technology, ed. by A. I. Johnson, Int. Assoc. Hydro!. Sciences, Publication No. 160, 25-40. Richard, W. G., and C. L. Crozier. 1983. Precipitation measurement with a C-band weather radar in Southern Ontario. Atmosphere -Ocean 2(12):125-137. Shih, S. F. 1982. Rainfall variation analysis and optimization of gaging systems. Water Resources Res. 18(4): 1269-1277. Tabios, III, G. Q., and J. D. Salas. 1985. A comparative analysis of techniques for spatial interpolation of precipitation. Water Resources Bull. 21(3):365-380. Wilson, W. E. I 982. Estimated effects of projected groundwater withdrawels on movement of the saltwater front in the Florida aquifer, 1976-2000, West-Central, Florida. USGS Water Supply Paper 2189, 24 pp.


PROCEEDINGS, VOLUME 51, 1992 29 Wojtiw, L. 1984. Use of radar to obtain intensity-duration estimates for rainstorms in central Alberta. Atmospheric Sci. Dep., Alberta Research Council, 53 pp. Wojtiw, L. 1987. Use of radar to derive a storm intensity-duration curve. Water Resources Bull. 23(5):849-855. Woodley, W. L., A. R. Olsen, A. Herndon, and V. Wiggert. 1975. Comparison of gage and radar methods of convective rain measurement. J. Applied Meteorology 14:909-928.


PROCEEDINGS, VOLUME 51, 1992 29 SPOT Satellite Data and GIS for Well Permitting and Management Y. R. Tan and S. F. Shih* ABSTRACT Land use classification from the Systeme Probatoire d'Observation de la Terre (SPOT) satellite data was demonstrated for well permitting and management. A geographic information system (GIS) technique was used to integrate the satellite-based land use classification with the well permit data. GIS processing (over lay and analysis) of the two data sets quickly provided the information about the current usage condition of permitted wells. The combination of remote sensing and GIS techniques proved to be a very useful tool for well permitting and management. Thousands of wells have been drilled in Florida for agricultural water usage. In 1978, the United States Geological Survey (USGS) estimated there were about 15,000 wells in Florida (Healy, 1978). Saline contamination (Healy, 1978; Burns, 1983) and increasing water demand (Khanal, 1980) have made managing and regulating those wells the focus of water management agencies. Consequently, a well permitting program was enacted within the South Florida Water Management District (SFWMD). The major functions of the well permitting program are not only to issue new well permits, but also to monitor the usage and condition of the permitted wells (MWU, 1985). Unfortunately, in managing and controlling wells several problems may be encountered. First, due to the large number of wells drilled throughout the region, current usage conditions of wells are very difficult to update by conventional methods. Second, the change of land use after well construction may affect agricultural permitting. Third, the lack of current land use data prevents issuing a new well permit and scouting an unused well. Fourth, land use data and well permit databases are not compatible. Satellite remote sensing data contribute a practical and efficient technique for land use classification (Welch, 1985; Shih, 1988; Coleman et al., 1990). The advantages of satellite data are the synoptic coverage over large areas and the periodic availability of data Y. R. Tan and S. F. Shih, Agricultural Engineering Dcp., Univ. of Florida, Gainesville, FL 32611-0570. Florida Agric. Exp. Stn. Journal Series no. R-02096. *Corresponding author. Trade names have been used to pro\'ide specific information. Their mention does ot constitute endorsement by the authors. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 : 29-34 ( 1992) for updating land use information. Rogers and Shih (1987) found that Landsat Multispcctral Scanner (MSS) data could provide maps of permitted agricultural areas for identifying potential agricultural water use. Recently, the French Systemc Probatoire d'Observation de la Terre (SPOT) satellite data was used for many land use mapping studies (Welch, 1985; Jordan and Shih, 1991 ), but its utility in well permitting and management has not been investigated. The geographic information system (GIS) techniques provide a convenient, yet powerful tool for manipulating and analyzing the geo-referenced data (Lo and Lineback, 1987; Jadkowski and Ehlers, 1989; Tan and Shih, 1990). The integration of SPOT satellite data and GIS techniques for well permitting and management has not been emphasized by water resources workers. The purposes of this study were: to determine the feasibility of obtaining land use information from SPOT satellite data for well management operations and to demonstrate the integration of satellite remote sensing data and GIS techniques for water well permitting and its associated management. MATERIALS AND METHODS Study Area and Source of Data The study was conducted in St. Lucie county, south Florida (Fig. 1), which covers about 639 mi2 (1,654 km2). A SPOT satellite image (SPOT scene #622-295) covering the county was acquired on Oct. 3, 1987 and used to classify land use. The SPOT image has a 20 m spatial resolution and three spectral bands which are green: 0.50-0.59 m, red: 0.61-0.69 m, and near infrared: 0.79-0.89 m. A total of 1,982 permitted wells with longitude and latitude information was provided for the county by the SFWMD. Land Use Classification The image processing system used for classifying the SPOT satellite image was the Earth Resources Laboratory Application Software (ELAS) which was developed by the National Aeronautic and Space Administration (Graham ct al., 1986). An unsupervised classification was performed using the Gaussian maximum likelihood classifier available in ELAS. The


30 Sou. AND CROP SCIENCE SOCIETY OF FLORIDA Study Area (639 sq mi) Fig. 1. Location of study area. classified image was geo-referenced to the Universal Transverse Mercator (UTM) coordinates. The ground-truth information was gathered from: 1) ae rial color infrared (ACIR) photographs which were obtained on Feb. 5, 1987 at a flying height of 12,000 ft (3,657 m) and a 1 :24,000 scale; and 2) field visits which were conducted in August 1987. The groundtruth information was used to determine the proper type of land use classified by the SPOT imagery. Storing Land Use and Well Permit Data into GIS A GIS which includes an IBM* AT compatible computer system and the Arc/Info* software (pc version 3.4D) were used. Because Arc/Info (pc version) allows a maximum of 5,000 arcs per polygon, the classified and geo-referenced image was divided into several subscenes. Each subscene has an area similar to a USGS 7 .5 minute series map. Then, the land use data were converted to Arc/Info format and stored as a data layer in a GIS database through the GRIDPOLY command available in Arc/Info. The well permit data file provided by the SFWMD was entered into the GIS database and stored as a separate data layer. The coordinates (latitudes and longitudes) of this well permit data layer were transformed into a UTM coordinate system to be compatible with the satellite-based land use layer in the GIS database. There are two reasons for converting this permit data file into a GIS data layer. First, this GIS data layer will quickly provide a synoptic coverage of county-wide well locations. Second, the GIS permitted well layer can be overlain and analyzed with land use data layer. Within the GIS environment, the two data layers (land use information and well permit data) were overlain using the IDENTITY command in Arc/Info to identify the usage condition for each of the permitted wells. As mentioned earlier, the main usage of the permitted wells was for agricultural production. The originally permitted agricultural usage is changed if a well is located outside the agricultural lands and wells located outside agricultural lands require more regulatory and management attention. By using the GIS graphical output capability, locations of the wells outside agricultural lands were plotted onto the land use map. Fig. 2 illustrates the entire procedure. Pilot Study Area and Data Overlay A pilot study area was selected for demonstration purposes. The land use in the pilot study area was representative of the entire county land use condition and the pilot area covered 81 permitted wells ( 4 .1 % of the permitted wells in the county). Thus, the area outlined in the upper-central part of Fig. 3 was chosen as the pilot study area which has an area of 14 mi2 (36 km2 ) or about a quarter of a USGS 7 .5 minute series quadrangle map and represents 2.4% of the entire county area. RESULTS AND DISCUSSIONS Classification of Land Use The unsupervised classification of the SPOT image produced 29 spectral response classes for the entire county. After ground-truthing, these 29 classes were aggregated into eight land use types (Table 1). Well Data File SPOT Satellite Image Image Classification Image Georeferencing Ground-truthing Spectral Classes Aggregation Arc/Info GIS System and Database UTM Conversion of Well data Spatial Data Overlay/Analysis Report Generation of Land use/wells Map generation of Land use/wells Fig. 2. Procedures for image processing and GIS data analysis.


PROCEEDINGS, VOLUME 51, 1992 31 I ""';":: .. ,~ ..... Fig. 3. Locations of county-wide permitted wells and pilot study area. These land use types mainly included water, marshes, urban lands/clearings, pine forests, pastures, citrus, hard woods, and a small class of mining areas. The land use types indexed as F, G, and H (Table 1) are considered as agricultural lands. These three land use types together represent 63.88% of the county area. The most common land use types in the county were pasture lands and citrus groves. The second most common land use was pine forests. The separation among citrus, pasture lands, and hardwoods was not emphasized in this study since the hardwood forest land is only a minor portion of land use in the county and this study was cumerned with the agricultural usage (e.g. citrus, pastures, row crops, etc) of the permitted wells. Locations of County-wide Permitted Wells Based on the GIS graphical output, locations of the 1,982 permitted wells in the county are depicted in Fig. 3 along with township and section lines and major highways (Florida turnpike and Interstate-95). A slight concentration of wells is noted in the central parts of the county (Fig. 3). New wells should not be permitted in a high well >ncentration area except where the unused wells li.,ve been identified and plugged. More attention ~hould be devoted to the high well concentration areas because a change of land use would significantly affect the well usage. In general, the synoptic view of county-wide well loca tions can provide information for the well permitting program in two ways: 1) to evaluate new permit applications more comprehensively; and 2) to improve management of the county-wide permitted wells. Overlay of Land Use with Well Permit Data A summary of the various land use types in the pilot study area is presented in Table 2. The agricultural lands indexed as F, G, and H (Table 2) occupied 87.9% of the pilot area and are depicted in Fig. 4. The land use data in the pilot study area are overlain with the well permit data. Results from the GIS overlay of satellite-based land use with well permit data indicated that most of the permitted wells in the pilot study area are still under agricultural usage. Therefore, the continued use of those wells is permissible. However, 11 permitted wells in the pilot study area were identified as being used in non-agricultural lands such as pine trees, marshy areas, and urban lands (Table 2). At.ten-Table 1. Land use/covers from 1987 SPOT data for St. Lucie County*, Florida. Index A B C D E F G H Total Spectral response class 2, 7, 9, 12, 16, 17, 23,24,29 18,20 10, 13,22 14, 19, 21, 25, 26 28,3,27 I, I 1 5,6,8 4 Land use type Water Mining operation Marshes Urban/build-up lands, clearings Pine forests Unimproved pastures, sapling citrus trees Improved pastures, young citrus, lawns Citrus trees, hardwoods *Total area in the entire county image is 639 mi2 (l.6 X 10' km2 ) including some coastal water areas. **I ha= 2.471 ac. Area Percent ha** % 16,331 9.8 39 0.0 9,072 5.5 16,023 9.7 I 8,318 11.1 42,878 25.9 45,964 27.8 16,878 10.2 165,504 100.00


32 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table 2. Permitted wells and land use/covers for pilot study area. Number Index of wells Land use type Area ha* A 0 Water 16 B 0 Mining operation 0 C 2 Marshes 98 D 7 Urban/build-up lands, clearings 160 E 2 Pine forests 201 F 30 Unimproved pastures, sapling 1,731 citrus trees G 36 Improved pastures, young citrus, 1,274 lawns H 4 Hardwoods, citrus 160 Total 81 3,640 *I ha= 2.471 ac. tion should be given to those 11 wells, particularly the seven wells which are located in urban land use types (Table 2). Those seven wells should be plugged since their usage has been changed. The other four wells which are located in marshes and pine forests should also be inspected to determine whether their usage is beneficial. Theoretically, the unused wells or the wells which are not used beneficially should be plugged. If a change of permitted agricultural usage of a well is detected, regulatory actions should be taken promptly as follows: 1) review the original permit of the well; 2) inform the well owner about the problem, 3) make a field visit to the well site; 4) revise and update the permit of the well if it is being used beneficially or to plug it to save water resources. The 11 wells represent 14% of the total permitted wells in the pilot study area. If this result is extrapolated to the entire county which has 1,982 permitted wells, there could be about 200 to 300 permitted wells that have permit problems. Based on a USGS statewide estimate of 15,000 wells (Healy, 1978) and on the plugging efforts as well as the newly permitted wells since 1978, it is evident that several thousand wells with permit problems exist in Florida. This indicates that the combination of satellite data and GIS techniques are feasible for improving well permitting and management. In south Florida, land use can undergo significant changes within a five-year period Gordan and Shih, 1991). Therefore, the usage condition of wells should be periodically monitored, perferably once a year. With the GIS capabilities, locations of the wells with a permit problem can be plotted onto the land LEGENDS Agr iclJI tural londs. \~ Urbon I onda ond clearing. Mor$hs. Water. SCALE: Fig. 4. Land uses of pilot study area.


PROCEEDINGS, VOLUME 51, 1992 33 "' i lands and clearing .;:,d f I ood-\le I I s t I ) ".: i l h pi,rn. i l pr--ob i em. r;~1 th l L j We t I :: a ) .,., i OU l _perm i l prob I em ; I I >SCALE : 1---... i Fig. 5. Plot of wells with permit problem on land uses in pilot study area. use map using color shades and markers. This plot will provide helpful information for the crews conducting field visits to those well sites. Fig. 5 provides an example of those GIS capabilities. Difficulties and Recommendations A methodology needs to be developed to refine the land use classification from SPOT satellite data for south Florida. Such a refinement will allow the well permitting and management personnel to better monitor the well usage in the region. For example, the separation among citrus, pasture lands, and hardwood forests will enable water resources managers to identify the usage of wells according to the three land use types. Therefore, a change of well usage (e.g. from pasture lands to citrus groves) within the agricultural lands can be identified. Such a usage change will affect the original permitted withdrawal allocation. Based on the Feb. 5, 1987 ACIR photography used as ground-truth in this study, citrus and pastures were registered in the ACIR transparencies with distinct colors. This provides some indications that those land use types might be separable. Early in the spring season, citrus trees are at a growing and blossoming stage while most pastures have not reco\'ered from winter defoliation. Therefore, selection or SPOT images from early spring months (e.g. Feb. and March) could improve land use classifications. Input of additional hydrological data (e.g. canals, rivers, water bodies, highways, hydrological boundaries, aquifer depth, etc.) to the current GIS database will be beneficial to the well permitting and management program. In the process of permitting a new well, the GIS database will allow one to quickly obtain the information about the nearbv land use, the effect on the withdrawal of neighborin'g wells, the effect of new well discharge on nearby water bodies, etc. SUMMARY This study demonstrates the feasibility of using SPOT satellite data and GIS techniques for improving the well permitting and management program. SPOT satellite imagery is a practical and efficient approach for obtaining timely land use information for monitoring well usage conditions. The advantages from using satellite data include not only the satellite synoptic coverage over large areas, but also the periodically available data for updating land use information. This allows the monitoring of the usage of permitted wells in the region. By entering the well permit data file into a GIS database and converting it into a spatial data layer, a


34 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA synoptic view ot county-wide well locations can be obtained quickly. This GIS capability will provide useful information for the well permitting and mamagcrnent agencies to better evaluate new permits and focus their management efforts. By using satellite remote sensing data and GIS overlay techniques, wells with changed usage condition can be identified timely and efficiently. In the meanwhile, the GIS graphical output capabilities can plot those wells with permit problems onto the current land use map, rendering very useful information before conducting field visits to those well sites. ACKNOWLEDGEMENT This project was partially supported by the South Florida Water Management District (SFWMD). The authors wish to thank: M. Piper of the SFWMD for the permit data; and 0. Lanni of the Remote Sensing Applications Laboratory of the Agricultural Engineering Department at the University of Florida for assistance in image processing. REFERENCES Coleman, T. L., L. (;udapati, and J. Derrington. 1990. Monitoring forest plantations using Landsat Thematic Mapper data. Remote Sens. Environ. 33:211-221. Burns, Wm. Scott. 1983. Well plugging applications to the interaquifer migration of saline ground water in Lee county, Florida. Tech. Pub. 83-8. South Florida Water Management District, West Palm Beach. FL. Graham, M. H., B. G.Junkin, M. T. Kalcic, R. W. Pearson, and B. R. Seyfarth. 1986. EI .AS-earth resources laboratory application software. Vols. I and II, NASA. !\at!. Tech. Lab., Earth Resources Lab., MS. Healy, H. G. 1978. Appraisal of uncontrolled flowing artesian wells in Florida. CSGS Water-Resources Investigations 78-95, Tallahassee, FL. Jadkowski, M.A. and M. Ehlers. 1989. GIS analysis of SPOT image data. Tech. Papers of ASPRS/ACSM Annual Convention, April 2-7 1989, Baltimore, MD. Jordan, J. D. and S. F. Shih. 1991. Landsat and SPOT imagery classification for Lind use change analvsis in Lee county, Florida. Soil Crop Sci. Soc. Florida Proc. 51: (in press). Khanal, N. N. 1980. Advanced water supplv alternatives for the upper east coast planning area. Parts I and II, Tech. Pub. 80-6, South Florida Water Management District, West Palm Beach, FL. Lo, T. H. C. and N. G. Lineback. 1987. Geographic information system for natural resources management in Alabama. Tech. Papers of ASPRS/ACSM Annual Convention, March 29 April 3 1987, Ballimore, MD. Management of water use (MWU), Permit information manual, Vol. III, 1985. South Florida Water Management District, West Palm Beach, FL. Rogers, J. W. and S. F. Shih. 1987. Using Landsat data for land use classification in agricultural land permitting program. ASAE Paper No. 87-2562, 1987 Intl. Winter Meeting of the Am. Soc. of Agric. Engineers, Dec. 15-18 1987, Chicago. IL. Shih, S. F. 1988. Satellite data and geographic information svstem for land use classification. J. Irrig. Drainage Engineering, Am. Soc. Civil Engineers, 114(3):505-519. Tan, Y. R. and S. F. Shih. 1990. GIS in monitoring agricultural land use changes and well assessment. Trans. ASAE, 33(4): 1147-1152. Welch, R. 1985. Cartographic potential of SPOT image data. Photogrammetric Engineering and Remote Sensing, 51(8):1085-IOlJI.


34 SOIL AND CROP SCIENCE SOCIEIY OF FLORIDA Using Remote Sensing and Geographical Information System in Water-Quality Assessment B. E. Myhre, S. F. Shih,* and D. A. Still ABSTRACT The need for management of water quantity and quality has become obvious in recent years. Over this same time period, two technologies have evolved: remote sensing and computers. Additionally, geographic information systems (GIS) were developed for main frame and personal computers. Two Systeme Probatoire de l'Observation de la Terre (SPOT) satellite scenes were classified by land use. This paper also demonstrates the use of GIS and satellite derived land use in water quality assessment. A pro gram was developed to allow a user to examine agricultural and silvicultural areas in relation to water quality parameters. Land use is an important parameter in hydrologic studies. These studies include water quantity and B. E. Myhre and S. F. Shih, Agricultural Engineering Dep., Univ. of Florida, Cainesville, FL 32611-0570. D. A. Still, St. Johns River Water Mai,ag,rnent District, Palatka, FL 32178-1429. Florida Agric. Exp. Stn. Journal Series no. R-02075. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florid11 Proc. 51 :34-38 (1992) quality management, which have recently become significant (USEPA, 1973, 1984). Substantial research has gone into the modeling and quantifying of nonpoint source pollution (Haan et al., 1982; Beasley, 1977; and Marani and Delluer, 1986). Land use and land use activities can be partially correlated to nonpoint source pollution (Novotny and Chesters, 1981). Thus, there is a need to identify the amount and spatial extent of different land uses for hydrologic analysis. Many existing hydrologic/water quality models (CREAMS, AGNPS, and others) use the Soil Conservation Service runoff equations (SCS, 1972). Additionally, the Universal Soil Loss Equation (USLE) is very popular for erosion modeling. Both of these equations are dependent on land use. Therefore, to successfully predict water quantity or quality land use must be determined. Accurate land use data, which are often difficult to acquire, have traditionally been obtained through aerial photography, and recently,


PROCEEDINGS, VOLUME 51, 1992 35 satellite imagery (Still and Shih, 1985; Shih, 1988; and Tan, 1988). Remote sensing data provide spectral and spatial information for the classification of land use and land cover which can be used by the SCS runoff equation and USLE-based models. Landsat was the primary source of remotely sensed data having a resolution of 79 m for the multi-spectral sensors (MSS) and 30 m for the thematic mapper (TM) sensors in comparison to 20 m resolution of the Systeme Probatoire de !'Observation de la Terre (SPOT) high resolution visible (HRV) system. In addition to models, geographic information systems (GIS) with or without remote sensing have been used to estimate land use to determine nonpoint pollution and its sources (Pelletier, 1985; Potter et al., 1986; Hession and Shanholtz, 1988; Oslin et al., 1988; and Rudra et al., 1991; for example). These previous works did not allow a user to interactively evaluate land use and water quality. Furthermore, water resource managers often need spatial information on water and land use and water quality without the complexities of modeling. This paper addressed the work accomplished within a GIS framework in relation to water quality data for agricultural and silvicultural areas. The source of the land use data is classified from the SPOT digital images. Also, all data were integrated into a GIS; Arc/Info specifically, which allows a user to query the data of interest. The general objective of this research was to use satellite data and GIS to establish a water-quality assessment program for the Lower St. Johns River Basin (LSJRB) and Lake George Basin (LGB). Specific objectives include: to 1) develop a conceptual model for integrating satellite data, water-quality data, hydrological/political layers with GIS; 2) convert existing water-quality data into GIS format; 3) convert the land use data classified from the SPOT images into the GIS format; and 4) demonstrate a practical application which will be to develop an interactive program within the GIS to view the LSJRB or LGB, hydrologic areas (rivers, creeks, and lakes), agricultural and silvicultural areas, and water-quality station locations and data. MATERIALS AND METHODS Study Area The two study areas in northeast Florida were the LSJRB and the LGB. The primary ecosystem of the basins is southern Florida flatwoods (hyperthermic zone). The soils in the area are mainly Spodosols, which are nearly level, somewhat poorly drained sandy soils. Smaller areas of Histosols (level, poorly drained organic soils underlain by marl and/or limestone) and Entisols (level to sloping, excessively drained thick sands) also exist in the two basins (Fernald and Patton, 1984). The land use in the basins historically included forest (conifers and hardwoods), agriculture (citrus, potatoes, cabbage, corn, onion, and improved pasture), wetlands, rangeland, urban, barren, and water. In recent years, fern operations and improved pasture areas have increased in extent. Satellite Data and Image Classifier Land use was determined by the use of SPOT imagery. Two scenes (SPOT scene #619-290 and #619-291) from 29 May 1988 were analyzed using Earth Resources Laboratory Application Software (ELAS) at the Remote Sensing Applications Laboratory (RSAL) of the Agricultural Engineering Department in the University of Florida. Each scene covers an area of approximately 5400 km2 over the LSJRB and the LGB, respectively (Fig. 1). The SPOT satellite HRV system has three spectral bands: green (0.50-0.59 m); red (0.61-0.69 m); and near infrared (0. 79-0.89 m), and has a resolution of 20 m. ELAS spectrally classifies satellite imagery into land use categories using a maximum likelihood classifier which utilizes a Bayesian statistical procedure (ELAS, 1989). LO~ER ST.JOHNS RIVER BASIN I I L I I I= ;-LAKE GEORGE BASIN 7 I = S~T SCENES Fig. l. Location of SPOT images.


36 Sou, AND CROP SCIENCE SOCIETY OF FLORIDA Geographic Information System Water-Quality Assessment Program Development The RSAL uses the Arc/Info GIS package developed by Environmental System Research Institute (ESRI, 1990).' Arc/Info is a vector based system that can perform overlays, analysis, and graphic output of spatial databases. The Arc/Info system is run on a Compaq 386, 20 Mhz, 120 MegaByte PC, which has an EGA monitor (resolution 640 X 300 pixels). A digitizing tablet (CalComp 9100) is used for entering map related data and a Hewlett Packard Proplot plotter is used for output. Data Layers The different layers (called coverages in Arc/Info) were hydrologic boundaries, roads, water bodies, district boundary, county boundaries, creeks, and U.S.G.S. quadrangle topographic map boundaries, agricultural and silvicultural areas, and water-quality stations. The St. Johns River Water Management District (SJRWMD) provided the hydrologic boundaries, roads, district boundary, water bodies, creeks, county boundary, and quadrangle boundary Arc/Info cover ages. After the SPOT images were classified they were converted from raster (i.e., rectangular grid) to vector (i.e., polygon) format data within Arc/Info. The im ages will then be trimmed to include only agricultural and silvicultural areas of over 16.2 ha (40 acres). The 16.2 ha size cutoff is due to the number of farming operations greater than 16.2 ha, the quantity of data available, and other factors. Water-Quality Data The water-quality station data are from STORET data which were gathered by the Environmental Protection Agency (EPA). These data are in an ASCII text format and were converted to the Arc/Info database manager, Dbase III Plus. The water-quality data included chemical parameters (alkalinity, BOD, Chlorophyll A, conductance, diversity index, DO, fluoride, 13 nitrogen val ues, three phosphorus values, Ph, fecal coliform, and organic carbon) and physical parameters (color, transparency, temperature, suspended solids, and turbidity). The important water-quality parameters for this study were nitrogen, phosphorus, turbidity, and total suspended solids. Nitrogen and phosphorus were studied because they are in a form which is readily available for plant uptake and, when transported to a water body via runoff, can cause eutrophication of that water body. Sediment and suspended solids were studied because they are often the transport mechanisms for other pollutants. Agriculture and silviculture are among the major sources of these water pollutants. 'Trade names have been used to provide specific information. Their mention does not constitute endorsement by the authcw< A water-quality assessment program (W AP) was developed to allow the user to determine agricultural and silvicultural areas (total and percentage of watershed basin) and distance from a selected area to the nearest water body. This ARC/INFO compatible program is interactive, and will allow the user to select a particular agricultural or silvicultural operation within the project area for analysis. Attribute data, including the STORET water-quality measurements and operation information from the selected site can then be accessed. The operation information can be either Consumptive Use Permit (CUP) or Soil Conservation Service (SCS) formats. RESULTS AND DISCUSSION Integration of Different Data Sources With GIS The merging of the satellite data and other data layers was accomplished through the GIS (Fig. 2). The GIS was used to convert the raster, tabular, and vector data into a mutually compatible format. The GIS was used to access and/or convert the different database layers. A more detailed discussion of each area follow. Establishment of Water-Quality Database The water-quality station data from the STORET database were used to create another coverage within Arc/Info. A simple BASIC program was developed to separate the data into two files. One file contained the geographic locations of the stations, while the other contained historical water quality data. The sta tions locations were transformed from latitude/longitude to universal transverse mercator (UTM) coordinates. The locations of the water quality station within the LSJRB are shown in Fig. 3. The water-quality data were imported into Dbase III, which is directly accessible with Arc/Info. The Water Quality Data CUP, BMP, etc. Raw Satellite Image ELAS Classified Image Ground-Truthing Final Classified Image GIS Farm Data r---~ USER Hydrologic & Political Layers Fig. 2. Conceptual model of merging remote sensing data, water-quality data, and hydrological/political layers with GIS.


PROCEEDINGS, VOLUME 51, 1992 37 HYDRO~OGIC BOUNDARY EB ffi WATER DUALITY STATIONS :fl ff \ ( \ \ Fig. 3. Water-quality station locations for Lower St. Johns River Basin. two data files were related to each other by station narne. Merging of SPOT Data into GIS Both SPOT scenes were divided into smaller areas and converted from raster to vector data using the GRIDPOLY command in Arc/Info. The division of the scenes was due to software limitations of Arc/Info (maximum of 5,000 arcs per polygon). Once the areas were in Arc/Info, the agricultural and silvicultural areas greater than 4.1 ha\10 acres) were reselected to create new coverages. These cover ages were used to determine contiguous areas of greater than 16.2 ha. Functions of Water-Quality Assessment Program The outline of major WAP menu function is depicted in Fig. 4. The four major functions are: 1. The reviewing of CUP permits (land owner identification, locations of farms, allocations, and other pertinent farm data). 2. The reviewing of water-quality (WQ) parameters for the WQ monitoring stations within a watershed and do some statistical analysis of WQ parameters. 3. The interactive measurement of the distance between any two points (e.g., from a stream to a farm), the location of points of interest, and the calculation of area enclosed by a polygon during a review process. 4. The quick generation of thematic maps for selected areas of interests with regard to CUP or WQ management. The program operates on a geographic information The WAP progu.,:'.1 Version 1. O J/19'J2 REWIEWS UPDATE SF.TUI' Qll!T r~~-j for W,~.l ,_codes I' for cLit-a Dir for data ---------=--------------Quit ,,,-;: Creatr> CUP owner list I! Create-l"!Q 1_:_,.t I 'l..'pdatc-cr_F r.,.1p I' 't:pdate=\;c station lo,::1t __ _i_ons Fig. 4. Outline of major menu functions for water-quality assessment program. system (GlS) database which contains topographical and hydrological features. If the proper procedures described are followed to enter the necessary data in the proper formats and to specify the setup configurations, this program can be extended to other watersheds for the same review purposes. Application of Water-Quality Assessment Program A program was developed to allow a user to access the different database layers. The program control diagram is shown in Fig. 5. The main menu has five (5) options: LAYERS; MEASURE; QUALITY; DATA; FARM DATA; AND ZOOM PLOT. The LAYERS option allows the user to display one or more of the 10 hydrologic, political, or land use !avers. The MEASURE option has the sub-options: WHERE, AREA, and LENGTH. WHERE gives the x and y UTM coordinates for a desired location, while the AREA and LENGTH sub-options supply area and length measurements respectively, for a given location. The QUALITY-DATA options allows the user to select a station and a water quality parameter and view raw data. Additionally, mean values for a period of record can be determined. The FARM-DATA option enables the user to select a farm or silvicultural area (greater than 16.2 ha). The relevant consumptive use permits (CUP), irrigation and drainage systems, and best management practices (B\1P) activities for the selected farm can then be viewed. The ZOOM-PLOT option permit the user to zoom in on an area of the map or optionally review


38 Son, AND CROP SCIENCE SOCIETY OF FLORIDA LAYERS MEASURE QUALITY_DATA FARM_DATA ZOOM_PLOT QUIT I I SELECT FARM WHERE VIEW DATA AREA LENGTH RETURN RETURN SELECT STATION VIEW DATA SELECT ZOOM AREA STATISTICS ZOOM ALL RETURN RETURN HYDROLOGIC BOUNDARY AGRICULTURE AREAS SILVICULTURE AREAS WATER BODIES CREEKS DISTRICT BOUNDARY COUNTY BOUNDARY QUAD BOUNDARY QUALITY STATIONS RETURN Fig. 5. Flow diagram for water-quality assessment program. the entire hydrologic basin after zooming in on an area. CONCLUSIONS This paper demonstrated the conceptual development of using remote sensing data in a GIS for water quality assessment. Two SPOT satellite images were classified and converted from raster to vector by the GIS. Additionally, water quality data and different hydrological and political layers were converted and imported into database layers. These layers were used in an interactive program which allowed a user to examine agricultural and silvicultural areas in relation to water quality parameters. Future work should include the coupling of this program with water quality models which could include topographic effects. ACKNOWLEDGEMENTS This project is a cooperative effort between the SJRWMD and the RSAL. The authors would like to thank Messrs. 0. Lanni, J. D. Jordan, and Y. R. Tan for their assistance in this study. REFERENCES Beasley, D. B. 1977. ANSWERS: A mathematical model for simulating the effects of land use and management on water quality. Ph.D. Thesis, Purdue Univ. West Lafayette, IN. ELAS. 1989. Earth-resources laboratory applications software, user reference. Eds. A. M. Beverley and P. G. Penton. National Aeronautics and Space Administration, John C. Stennis Space Center, Sci. and Technology Laboratory. Environmental Systems Research Institute, Inc. 1990. Users manual: PC Arc/Info Starter Kit and PC Arc/Info Arcplot. Red lands, CA. Fernald, E. A., and D. J. Patton. 1984. Water resources atlas of Florida. Institute of Sci. and Public Affairs. Haan, C. T., H. P. Johnson, and D. L. Brakensiek. 1982. Hydrologic modeling of small watersheds. Am. Soc. Agric. Engineers, St. Joseph, MI. Hession, W. C., and V. 0. Shanholtz. 1988. A geographic information system for targeting non-point source agriculture pollu tion. J. Soil Water Conserv. 43(3):264-266. Marani, A., and J. W. Delluer. 1986. Agricultural non-point source pollution: model selection and application. Developments in Environmental Modeling, Elsevier, New York. Novotny, V., and G. Chesters. 1981. Handbook of non-point pollu tion sources and management. Van Nostrand Reinhold Com pany, New York. Oslin, A. J ., R. A. Westsmith, and D. S. Morgan. 1988. STREAMS: A basin and soil erosion model using CADD, remote sensing and GIS to facilitate watershed management. Proc. ASAE Int. Symp. Modeling Agri., Forest, and Rangeland Hydrology, Chicago, IL, December 12-13, 1988, pp. 470-477. Pelletier, R. E. 1985. Evaluating nonpoint pollution using remotely sensed data in soil erosion models .J. Soil Water Conserv. 40(4):332-335. Potter, W. B., M. W. Gilliland, and M. D. Long. 1986. A geographic information system for prediction of runoff and non-point source pollution potential. Hydrologic Applications of Space Technology, Int. Assoc. Hydrologic Sci. Publ. No. 160, pp. 437446. Rudra, R. P., W. T. Dickinson, and P. Donaghy. 1991. Watershed management using NPS model and GIS. Proc. Int. Conf. Computer Applications Water Resources. Tamsui, Taiwan,July 3-6, 1991, pp. 1201-1209. Shih, S. F. 1988. Satellite data and geographic information system for land use classification. J. Irrigation Drainage Engineering, Am. Soc. Civil Engineers 114: 505-519. Still, D. A., and S. F. Shih. 1985. Using LANDSAT data to classify land use for assessing the basinwide runoff index. Water Re sources Bull. 21:931-940. Tan, Y. R. 1988. Remote sensing and geographic information for studying agricultural landuse changes and abandoned well identification in St. Lucie county, Florida. Master thesis, Univ. of Florida, Gainesville, FL. U.S. Environmental Protection Agency. 1973. Methods for identifying and evaluating the nature and extent of non-point sources of pollutants. Office of Air and Water Programs, Washington, D.C. U.S. Environmental Protection Agency. 1984. Report to congress: non-point source pollution in the U.S, water planning division, Washington, D.C. U.S. Soil Conservation Service. 1972. SCS national engineering handbook. Sec. 4, hydrology. USDA-SCS, Washington, D.C.


PROCEEDINGS, VOLUME 51, 1992 39 Phosphorus and Zinc Influence on Bermudagrass Growth J. B. Sartain ABSTRACT Intensively managed turfgrass grown on golf greens and fairways often receives relatively high rates of fertilizer P and Zn. Some grasses have been found to exhibit P toxicity symptoms and/or Zn deficiency when grown under high P low Zn condi tions. A 3 yr study was conducted to evaluate the influence of elevated levels of soil P and Zn on the growth and quality of Tifway bermudagrass ( Cynodon dactylon (L.) Pers. x Cynodon transvaalensis Burtt Davy). Experimental plots were established on an Arredondo f.s. (Grossarenic Paleudult, loamy, siliceous, hypertbermic) which contained varying levels of Mehlich-1 extractable P. To study the effects of residual P and applied P, half of the plots received 0 P and the other half received 20 g P m-2 60 d-1 The plots were further divided and half received 2.5 and 12.5 g Zn m-2 the first year and 5 and 25 g Zn m-2 the second and third year. Residual soil P levels did not influence bermudagrass growth, but application of 20 g P m-2 60 d-1 reduced growth b~t not quality. Phosphorus uptake rate was not affected by P appli cation. Application of Zn did not influence growth nor Zn uptake rate. Extractable soil Zn levels over 252 mg kg-1 did not affect bermudagrass growth nor quality. In summary, bermudagrass can tolerate high levels of soil extractable P (478 mg P kg-1 ) and Zn without exhibiting toxicity effects. The negative influence of applied P on bermudagrass growth may be an indirect antagonis tic effect of soluble anionic P on NO'-N uptake and subsequent growth. Phosphorus application to soils containing high levels of extractable P has been shown to differentially influence the growth of ryegrass and bermudagrass. Addition of P to a phosphatic Arredondo fine sand (Grossarenic Paleudults, loamy, siliceous, hyperthermic) enhanced the growth rate of Medalist II ryegrass (Lolium perenne L.) and reduced the growth rate of bermudagrass ( Cynodon dactylon (L.) Pers. X Cynodon transvaalensis Burtt Davy). (Sartain and Dudeck, 1982). High levels of P in subterranean clover (Trifolium subterraneum L.) have been shown to produce P toxicity symptoms of necrosis in old leaves and depressed growth in new leaves (Loneragan et al., 1979). Italian ryegrass (Lolium multiflorum Lam. Tifton #1) has been shown to require 750 mg P kg-1 for satisfactory growth (Hylton et al., 1965). Increasing the P content in excess of 1000 mg P kg-1 did not induce additional dry matter accumulation. Kentucky bluegrass (Poa pratensis L.) dry matter yields were not reduced by application of 1700 kg P ha-1 (Juska et al., 1965). Phosphorus has been shown to interact with the uptake and leaf concentration of Mg, Ca and K in cool-season grasses (Triticum aestivum L.) (Reinbott and Blevins, 1991). Phosphorus and Zn nutrition in plants have often been linked. Loneragan et al. ( 1979) observed that application of P fertilizer induced or enhanced symptoms resembling Zn deficiency. In plants where P limited growth, additional P depressed Zn concentrations and induced Zn deficiency by J.B. Sartain, Soil and Water Sci. Dep. 106 Newell Hall, Univ. of Florida, Gainesville, FL 32611. Florida Agric. Exp. Stn. Journal Series no. N-00598. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51:39-42 (1992) promoting growth and diluting available Zn supplies. In plants where P did not limit growth, additional P increased P concentrations to toxic levels producing necrotic symptoms in old leaves and depressing new growth. Where plant tops had high Zn concentrations, P toxicity produced necrotic symptoms in the absence of any characteristic symptoms of Zn definency. Many of the soils used in turf grass production are high in P due to their mineral origin and continue to increase in available P in response to annual P fer tilizer applications. Due to the high Ca content of most of the irrigation water used on turfgrasses the mean pH of turfgrass producing soils tend to in crease. This increasing pH limits the solubility of Zn on these high P containing soils which establishes conditions ideal for Zn deficiencies. The interaction ef fects of P and Zn on bermudagrass have not been studied on Florida soils testing high in extractable P. Over the years through the application of mineral and organic sources containing Zn, high levels of Zn have accumulated in some turfgrass soils. Little is known relative to the effects of high soil Zn levels on the growth and quality of bermudagrass. This 3 yr study was established with multiple objectives. In one area, five levels of extractable P had been established in a previous experiment. This area was used to evaluate the influence of residual and applied P on the growth and P uptake of bermudagrass. In an adjacent area, also containing a range of soil P levels, Zn was applied to evaluate the influence of Zn and the interaction of P and Zn on the growth and P and Zn uptake of bermudagrass. MATERIALS AND METHODS Experimental units were composed of 2 X 3 m areas planted to Tifway bermudagrass (Cynodon dacty lon (L.) Pers.). Three replications of five rates of P (0, 10, 20, 40 and 80 g m-2 ) were established in a randomized complete block design in a previous experiment. These plots were divided. One of the set of 5 P level plots received P at the rate of 0 and 20 g P m-2 60 d-1 This set of treatments was designed to study the effects of the residual and applied P on the growth and P accumulation of bermudagrass. On other plots, Zn was applied at 2.5 and 12.5 g m-2 60 d-1 the first year and at 5 and 25 g m-2 60 d-1 the second and third years. Potassium as potassium sulfate (K2SO4 ) was applied every 90 d at 10 g K m-2 Nitrogen was applied every 30 d at 10 g N m-2 as ammonium nitrate (NH4NO3). Visual ratings were taken biweekly and clippings for dry matter production and nutrient uptake analysis were taken every 60 d. Soil samples were taken at the end of each warm season growth period. Plant tissue and soil samples were analyzed for Ca, Mg, K, P and Zn. Soils were extracted using Mehlich I (0.025 M HCl and 0.0125


40 SOIL AND CROP SCIF.1"CE SOCIETY OF FLORmA Table 1. Soil chemical properties prior to treatment application. Previous Mehlich-1 extractable+ P applied pH p Ca e11g K Zn g Pm-2 -----mg kg' -------() 6.2 151 12% 79 29 32 IO 6.0 192 120'.l 70 28 18 20 5.9 218 1251 67 28 25 40 6.0 313 1211 69 27 19 80 5.8 429 1271 G9 28 33 'Mchlich-I extractable (0.025 M HCI + 0.0125 M H2S01). JV! H2S01). Plant tissue was ashed at 500C and ash dissolved in 0.01 M HCI. All analyses were performed on an Atomic Absorption Spectrophotometer. Statistical techniques as outlined in SAS (1987) for regression and analysis of Yariance were used. Single degree of freedom contrasts were employed for identification of mean differences. RESULTS AND DISCUSSION Initial Conditions: This experimental site had been used in a pre\'ious P study in which five levels of P had been applied to establish a range in extractable P levels. These P treatments were applied 2 yr prior to the beginning of this study. The Mehlich-I extractable levels of selected nutrient elements are presented in Table 1. The unamended soils contain high levels of Mehlich-I extractable P (151 mg P kg-1 ) confirming the phosphatic nature of the Arredondo soil. Approximately a three fold increase in Mehlich-I extractable P is noted in response to the previously applied treatments. An excessively high P level (429 mg P kg-1 ) was extracted from plots receiving 80 g P m-2 Mehlich-I extractable leYels of Ca and Zn were sufficient for optimum turfgrass growth while extractable levels of Mg and K were marginal. Effect of Soil P: Previously applied P treatments and their corresponding extractable soil P levels had no influence on the growth rate of the bermudagrass during either of the 3 yr of this study (Table 2). I 11 the previous experiment, Mehlich-I extractable P levels declined over time (Sartain, 1979) and became Table 2. Effect of residual soil P level on bermudagrass growth rate. Previous Growth rate l' applied Isl Yr 2nd Yr :lrd Yr g p m-2 --------kg ha d -0 19.6 31.'l 24.7 10 23.7 2.~.6 24.1 20 18.7 20.9 21.6 40 20.9 24.9 19.6 80 19.3 23.'.l 18.4 ---P>F-----Contrasts 0 vs. 10 l'\S NS NS 10 vs. 20 NS NS '.\IS 20 vs. 40 NS NS NS 40,s. 80 NS NS NS Table 3. Effect of applied P on the uptake of Zn by bermudagrass by year. Previous applied P gm' () 10 20 40 80 Contrasts O vs. others Tissue Zn 1st Yr 2nd Yr 3rd Yr --------mg kg-' ---------135 '.lO0 323 131 304 321 117 325 '126 138 290 301 135 313 294 --P>F---------NS NS NS progressively less plant available. Although exces sively high levels of extractable P were observed no influence on growth was noted. There are two possible explanations for the observed response. One being that the Mehlich-1 extractant is extracting P which is not available to the turf grass and overestimating the P supply, or secondly bermudagrass has a very wide range of tolerance to accumulated P. Mehlich-I extractable P levels, as established by previous P application, did not influence bermudagrass tissue concentration of Zn (Table 3). Mean tissue Zn levels increased each year of the study indicating that some of the applied Zn was being taken up by the bermudagrass. Effect of Applied P: Bermudagrass growth rate was reduced in each of the 3 yr by the application of additional P (Table 4). These findings are in agreement with a previous study which showed that application of P reduced bermudagrass growth and enhanced ryegrass growth (Sartain and Dudeck, 1982). In the previous study a reduction in N uptake was observed in conjunction with the reduction in growth rate. It was suggested that the competitive effects of the N03-N and H2POrP anions may have been responsible for the observed reduction in N uptake and subsequent growth rate. Clipping yield and rooting of Merion Kentucky Bluegrass (Poa frratensis L.) were enhanced by the application of additional P (Watschke et al., 1977). Ammoniacal nitrogen is the predominant form of N in cool soils. It is possible that additional P enhanced N uptake by the cool season Kentucky Bluegrass which resulted in increased growth. Bluegrass tissue P levels were more than sufficient for optimum growth (0.32% P) even on plots were no Table 4. Effect of applied P on the growth rate of bermudagrass by year. Applied P gm-" () 20 Contrasts 0 vs. 20 Growth rate 1st Yr 2nd Yr 3rd Yr ------kg ha-1 cl' --------24.2 33.7 25.4 19.3 1 7.4 20.:"i ----------P> F----------0.01 0.0 I 0.(J I


PROCEEDINGS, VOLUME 51, 1992 41 50-----------------45 40 Jg 35 Ol m 30 Y = L13 + 233x P Flux, f-= 0.96,CV = 62%__ -/ ~j 0 ----ro I i 25 (3 ::r .~ .... ] 10 ,-----------,-----,--~ 4 6 8 10 12 14 16 18 Phosphorus Flux (mg ha-1d Fig. 1. Relationship between 'Tifway' bermudagrass growth rate and phosphorus flux as determined by the rate of P accumulation by the tissue. P was applied which indicates that the_ posi~ive response to applied P was not r c:lated phys10logically to P. A linear relationship (r2 = 0.96) was observed between bermudagrass growth rate and tissue P flux (Fig. 1), which suggests that the growth rate continued to increase as the tissue P concentration increased and that no toxic effects of elevated tissue P levels were observed. This finding suggests that the observed growth rate reduction in response to applied P is more related to the influence of P on N uptake than to the actual toxic effect of P since t_he levels of P in the tissue would not normally be considered as being excessive for optimum growth and the flux of tissue P was not changed over the entire growth range. Tissue Zn concentrations were reduced by the application of additional P (Table 5). This relationship between P and Zn has been noted by a number of researchers (Singh et al., 1986; Wagar et al., 1986 and Bowman et al., 1978). Numerous explanations have been proposed for this phenomenon. In some cases, a portion of the decrease in Zn levels was due to dilution and to the formation of an insoluble P-Zn complex in the roots. Other studi_es h_ave r~lat~d the Zn reduction to reduced mycorrhizal mfection m the roots of plants grown in treatments of high P applica tion rates (Singh et al., 1986). Table 5. Effect of applied P on Zn uptake by bermudagrass by year. Applied P gm-' 0 20 Contrasts 0 vs. 20 Tissue Zn !st Yr 2nd Yr 3rd Yr ---------mg kg-' --------151 316 355 124 258 310 ----------P>F----------0.01 0.01 0.01 Table 6. Effect of applied Zn on bermudagrass growth rate by year Applied Zn gm-2 2.5/5.0 12.5/25 Contrasts Growth rate !st Yr 2nd Yr 3rd Yr ------kg ha-' d --------23.2 26.0 23.0 20.4 25.2 21.7 ----------P>F----------2.5/5.0 vs. 12.5/25 NS i\JS NS Effect of Applied Zn: Bermudagrass growth rate was not influenced by the application of Zn (Table 6). Due to lack of influence of applied Zn on the growth rate of the bermudagrass during the first year of the study, Zn application rate was increased five fold to induce a growth reduction. It is apparent that bern:iudagrass can withstand relatively high levels of applied Zn without exhibiting toxicity symptoms. This observation is supported by a linear relationship between the growth rate and tissue Zn flux [Growth rate (kg ha-1 d-1 ) = 5.01 + 0.002 Zn flux (kg Zn ha-' d-1 ) (r" = 0.73)]. Soil levels of Mehlich-I extractable Zn increased dramatically in the Oto 5 cm zone in response to applied Zn (140 and 252 mg Zn kg-')_ fro~ the mean extractable Zn level of 25 mg kg-1 which existed prior to Zn applicati:)n (Table 7)_. Most o_f the Zn remained in the top five cm of soil and did not leach similar to the P. A level of 252 mg Zn kg-1 would be considered as high, but no detrimental effects on growth rate were observed. Bermudagrass visual quality ratmgs_ were ~ot n~fluenced by any of the treatment vanables m _this study. The m~an visual quality ratings for the f2rst, second and third vear of the study were 7.65, 1.44 and 7.87, rcspecti,;cly. The ratings were taken based on a 1 to 9 scale with 9 representing excellent turfgrass quality. All ratings were over 7.0 which is indicative of a high quality turfgrass. Based on the range in extractable soil P and Zn and in tissue levels of P and Zn cited above, it appears that bermudagrass can maintain growth and quality over a ":~Y wide range of soil and tissue P and Zn levels. Addiuonally, no relationship was found to exist between the P and Zn ratio in bermudagrass tissue and growth rate ?f turfgrass (data not shown). A critical ratio of P/Zn m Table 7. Effect of applied Zn and Pon the :'t:e~lich-1 extractabl_e levels of Zn and P in soil by depth at termmat10n of the experiment. Soil A!.>!.> I ied Z 11 (g m-2 ) A pl_llied P (gm-') Depth 2.5/12.5 5/25 F () 20 F cm ---mg Zn kg-1 ------mg Pkg-' ---Oto 5 140 252 ** 200 478 ** 5 to 10 2.4 3.0 NS 225 347 ** 10 to 20 1.9 2.2 NS 190 27-1 ** **indicates significance at the 0.0 I 'Ir level; NS signifies no significant differences.


42 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA the tissue of some plant species has been suggested by some researchers (Brown and Tiffin, 1962) but this does not appear to be true for bermudagrass. CONCLUSIONS Applied P induces much greater response in bermudagrass growth rate and P uptake than residual levels of P. Growth rate reductions in response to applied P appear not to be related directly to the accumulation of excessive levels of P in the tissue but to other factors, possibly to an antagonistic effect on N uptake. Application of high rates of Zn and accumulation of excessive levels of extractable soil Zn did not influence bermudagrass growth rate. The lack of a reduction in the rate of uptake of Zn and the lack of a correlation between growth rate and P/Zn ratio suggests that under the conditions of this experiment Zn was not deficient nor toxic. Bermudagrass can maintain optimum growth and quality over a very wide range of soil and tissue P and Zn. REFERENCES Bowman, R. A., S. R. Olsen, and F. S. Watanabe. 1978. Greenhouse evaluation of residual phosphate by four phosphorus methods in neutral and calcareous soils. Soil Sci. Soc. Am.J. 42:451-454. Brown, J. C. and L. 0. Tiffin. 1962. Zinc deficiency and iron chlorosis dependent on the plant species and nutrient-element balance in Tulare clay. Agron. J. 54:356-358. Hylton, L. 0., Jr., A. Ulrich, D.R. Cornelius, and K. Okhi. 1965. Phosphorus nutrition of Italian rvegrass relative to growth, moisture content and mineral constituents. Agron. J. 57:50'>508. juska, F. V., A. A. Hanson. and C .J. Erickson. 1965. Effects of phosphorus and other treatments on the development of red fescue, merion, and common Kentucky bluegrass. Agron. J 57:75-78. Loneragan, J. F., T. S. Grove. A. D. Robson, and K. Snowball. 1979. Phosphorus toxicity as a factor in zinc-phosphorus interactions in plants. Soil Sci. Soc. Am. J. 43:966-972. Reinbolt, T. M. and D. G. Blevins. 1991. Phosphate interaction with uptake and leaf concentration of magnesium, calcium and potassium in winter wheat seedlings. AgTon. J. 83: 1043-1046. Sartain, J. B. 1979. Mobility and extractability of phosphorus applied to the surface of tifway bermudagrass turf. Soil Crop Sci. Soc. Florida Proc. 39:2-4. Sartain, J. B. and A. E. Dudeck. I 982. Yield and nutrient accumulation of tifway berrnudagrass and overseeded ryegrass as influenced by applied nutrients. Agron. J. 74:488-491. SAS Institute, Inc. 1987. SAS user's guide: statistics. 1987 ed. SAS Institute, Inc., Cary. NC. Singh, J.P., R. E. Kararnaros, and .J. W. B. Stewart. 1986. Phosphorus-induced zinc deficiency in wheat on residual phosphorus plots. Agron .J. 78:668-675. Wagar, B. I., J. W. B. Stewart, and J. L. Henry. 1986. Comparison of single large broadcast and small annual seed-placed phosphorus on yield, and phosphorus and zinc content of wheat on chernozemic soils. Can. J. Soil Sci. 66:209-215. Watschke, T. L., D. V. Waddington, D.J. Wehner, and C. L. Forth. 1977. Effect of P, K, and lime on growth, composition, and 321' absorption by merion Kentucky bluegrass. Agron. J. 69:825828.


42 Sou, AND CROP SCIENCE SOCIETY OF FLORIDA Managing Plant-Parasitic Nematodes in Crop Sequences R. McSorley* and R. N. Gallaher ABSTRACT Plant-parasitic nematodes, particularly root-knot nematodes (Meloidogyne spp,), are a major limitation to crop production in tropical and warm temperature regions of the world. In these areas, multiple-cropping is the usual practice, and the selection of rotation crops or sequence of crops in multiple cropping can determine the severity of nematode problems in subsequent crops. Several examples of effects of selected crops on population buildup of Meloidogyne incognita (Kofoid and White) Chitwood in north-central Florida are reviewed. Damage to soybean (Glycine max [L.] Merr.) was more severe following a winter cover crop of crimson clover (Trifolium incan:ratum L.) than following rye (Secale cereale L.). In another experiment, populations of M. incognita increased greatly on summer crops of corn (Zea mays L.) or soybean, but remained low following sorghum (Sorghum bicolor [L.] Moench). Finally, the advantage of sorghum in maintaining low populations of M. incognita was confirmed in three additional field tests. Similar performance was observed with sorghum and sorghum-sudangrass (S. sudanense [Piper] Stapt). It is evident that poor hosts such as sorghum or R. McSorley, Entomology and Nematology Dep., University of Florida, Gainesville, FL 32611-0620; and R. N. Gallaher, Agronomy Dep. University of Florida, Gainesville, FL 32611-0541. Fla. Agric. Exp. Stn. Journal Series no. N-00463. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 :42-45 ( 1992) rye can be useful in crop sequences for keeping populations of root-knot nematodes low, but more research is needed to discover additional crops and cultivars useful for the management of rootknot and other nematode pests. The advantages of crop rotation in managing plant-parasitic nematodes have been demonstrated by numerous examples from the southeastern United States Qohnson, 1982). In Florida, double or multiple cropping is often practiced. In some instances, high value cash crops are alternated with crop covers. In these cases, it is important that the cover crops selected do not aggravate existing problems in subsequent cash crops. Populations of soil borne pests, such as paint-parasitic nematodes, will rise or fall in a site depending on the sequence of crops planted. Therefore, the choice of crops included in a site is critical in minimizing nematode problems, but much research is needed to determine the effects of candidate crops and cultivars on the populations of a range of nematode species. Although a polyspecific assemblage of plantparasitic nematodes exists in many Florida fields (McSorley and Dickson, 1989), choice of crop se-


PROCEEDINGS, VOLUME 51, 1992 43 quence can be simplified by emphasizing management of key nematode pests, such as root-knot nematodes (Meloidogyne spp.) or the sting nematode (Belonolaimus longfraudatus Rau). Various species and genera of plant-parasitic nematodes will respond differently to the same crop sequence (Gallaher et al., 1991). Often it is practical to target the root-knot nematodes, which tend to be the most damaging group in many systems. This paper reviews briefly the management of M. incognita by crop selection in north-central Florida. Additional information and data on other nematode species have been published elsewhere (Gallaher et al., 1991; McSorley and Gallaher, 1991 a, b). Three examples are illustrated: 1) effect of a winter cover crop on root-knot damage to soybean in the following summer; 2) comparison of nematode population buildup on summer crops of sorghum, corn, and soybean; 3) comparison of nematode population buildup among corn and sorghum cultivars. MATERIALS Ar<. D METHODS Effect of Winter Cover Crop. On 2 Apr 1990, a soybean blend (Hiebsch, 1990) was planted in Arredondo fine sand (loamy, siliceous, hyperthermic Grossarenic Paleudults) at the University of Florida Green Acres Agronomy Research Farm in Alachua County. During the previous winter, half of the plots were planted to 'Wrens Abruzzi' rye and half to 'Dixie' crimson clover. On 31 July, serious damage from M. incognita was noted in the soybeans following crimson clover, and so plant dry weights and nematode samples were collected from both rye and crimson clover sites (McSorley and Gallaher, 199 la). Nematodes were extracted from soil samples using sieving and centrifugation (Jenkins, I 964), and root damage from nematode galling was rated on a O to 10 scale (Zeck, 1971). Data from rye and crimson clover sites were compared using analysis of variance (Freed et al., 1987). Effects of Summer Crop. Nematode population increase was compared on corn, sorghum, and soybean in an Arredondo fine sand (94% sand, 3.5% silt, 2.5% clay; pH 6.7%; 2.0% organic matter) at the Green Acres Agronomy Research Farm. Following a winter cover crop of rye, the site was planted on 20 May 1990 with Pioneer X304C corn for silage, Pioneer X304C corn for grain, DeKalb FS25E forage sorghum, DeKalb BR64 grain sorghum, and 'Centennial' soybean. The five treatments (crops) were arranged in a randomized complete block design with five replicates. Plot maintenance and design are described in detail elsewhere (Gallaher et al., 1991). Plots were sampled for initial (Pi) and final (Pf) nematode densities on 6 June and 12 Sept, respec tively. Nematodes were extracted from 0.2 pt (100 cm3 ) soil subsamples using a modified sieving and centrifugation procedure (Jenkins, 1964). Nematode count data were log IO-transformed, and single degree of freedom orthogonal contrasts (Freed et al., 1987) were determined for corn vs. soybean, sorghum vs. soybean, and corn vs. sorghum. Effect of Corn and Sorghum Cultivars. Nematode population increase was compared on corn and sorghum cultivars in two adjacent fields at the University of Florida Pine Acres Farm in Marion County. The soil was an Arredondo sand-Gainesville loamy sand association (92% sand, 3% silt, 5% clay; pH 5.6; 2.8% organic matter). In both fields, the experimental de sign was a randomized complete block design with six treatments and eight replications. Treatments were four corn cultivars (two temperate hybrids, Pioneer 3320 and Northrup King 508; two tropical hybrids, Pioneer X304C and an experimental open-pollinated synthetic, Florida SYN-I), the sorghum hybrid De Kalb FS25E, and the sorghum-sudangrass hybrid De Kalb SX-17. Individual plots consisted of four rows 75 cm apart and 9 m long. Plots in the east field were planted 20 May 1990, and harvested for forage in late August. Plots in the west field were planted 2 Apr 1990, harvested for forage in early July, and then replanted on 20 July for double cropping. Sorghum and sorghum-sudangrass were permitted to ratoon for the second crop. The second crop of corn and sorghum was harvested in early November. Nematode samples were collected for the west field on 17 Apr, 18 July, and 22 Oct; the east field was sampled on 5 June and 30 August. Plot management, harvest procedures, and nematode extraction are described in detail elsewhere (McSorley and Gallaher, 1991b). Nematode densities among the various cul tivars were compared using analysis of variance of log I 0-transformed data, followed by mean separations using Duncan's multiple range test. Single degree of freedom contrasts were determined for corn vs. sorghum. Table l. Nematode densities and yields of soybean following winter cover crops of rye or crimson clover. Quality evaluated Nematodes: Criconemella spp. Meloidogyne incognita Paratrichodorus minor Pratylenchus spp. Root damage ratingt Soybean plant dry matter: Leaf Pod Root Stalk Total Winter Crop Rye Clover -----nematodes I 00 cm' soil----28 97 82 638 ** 14 8 64 76 1.00 3.75*** -------dry matter, g m'------449 216 *** 145 42 ** 167 87 661 230*** 1422 574 *** *, **, *** Significant differences between rye and crimson clover plots at P :s 0.05, P :s 0.01, and P :s 0.001, respectively. tRoots rated for galling on a scale from 0 (no galling) to 10 (all roots completely covered with galls) (Zeck, 1971).


44 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table 2. Initial (Pi) and final (Pf) nematode densities on corn, sorghum, and soybean crops. Criconemella spp. Meloidogyne incognita Paratrichodorus minor Pratylenchus spp. Crop Pi Pf Pi Pf Pi Pf Pi Pf -----------------------------------------------------Nematodes 100 cm' soil ----------------------------------------------------Corn (silage) 0 14 4 189 4 14 8 337 Corn (grain) 2 77 I 114 7 5 6 172 Sorghum (forage) I 141 I 5 17 15 10 132 Sorghum (grain) I 314 I 2 9 3 10 112 Soybean 15 67 0 69 6 6 4 108 Orthogonal contrasts: Corn vs. soybean *** nst ns ns ns ns ns ** Sorghum vs. soybean *** ns ns ns ns ns ns Corn vs. sorghum ns ns *** ns ns ns ** *, **,***Orthogonal contrasts significant at P :5 0.05, P:5 0.01, and P :5 0.001, respectively. tNS = contrast not significant at P :5 0.05. RESULTS AND DISCUSSION Effect of Winter Cover Crop. The plant-parasitic nematodes Criconemella spp., M. incognita, Parat richodorus minor (Colbran) Siddiqi, and Pratylenchus spp. were associated with the soybean crop grown in this site, but only M. incognita showed differences (P :5 0.05) in densities related to winter crop (Table 1). Densities of M. incognita juveniles in soil and gall indi ces of the soybean root systems were much greater following crimson clover than following rye. Dry matter yield of all portions of the soybean plant (leaf, pod, root, stalk, total) was increased (P :5 0.05) following rye compared to crimson clover (Table 1). Effect of Summer Crop. The population buildup of four nematodes on summer crops of corn, sorghum, or soybean is summarized in Table 2. The most important result is that relatively low final populations of M. incognita built up on sorghum compared to corn (P :5 0.001) or soybean (P :5 0.05). This confirms results of a previous study by Gallaher et al. ( 1988), suggesting that sorghum may be an effective crop for keeping populations of this pest at low levels. Criconemella spp. built up on sorghum more than on corn (P :5 0.05), but this is not of particular concern, since Criconemella spp. are not very damaging to most crops. Effect of Corn and Sorghum Cultivars. Results of these three tests also demonstrated that much lower populations of M. incognita resulted after sorghum than after corn (Table 3). In all three tests, high final populations of root-knot nematodes resulted on the four corn cultivars tested, but low populations occurred following both sorghum cultivars. Thus, re gardless of cultivar, the effect of crop on M. incognita population was similar to the previous test (Table 2) with sorghums suppressing root-knot buildup compared to corn. There were some differences (P :5 0.05) among cultivars in final populations of Criconemella spp., P. minor, and Pratylenchus scribneri Steiner (Table 3). However, such differences were in consistent. Population increases of these nematode species on corn and sorghum are discussed in more detail elsewhere (McSorley and Gallaher, 1991 b ). As mentioned previously, management of rootknot nematodes will be an important concern in many situations, because these pests are so damaging to a wide range of crops. The other nematode species examined here are either not very damaging to crops (e.g. Criconemella spp.), or few differences resulted Table 3. Final nematode densities on corn and sorghum cultivars in three plantings in Marion County, 1990. Criconemella spp. Meloidogyne incognita Paratrichodorus minor Pratylenchus scribneri Crop Cultivar Wt W2 E WI W2 E WI W2 E WI ---------------------------Nematodes 100 cm' soil --Corn Pioneer 3320 608c:j: 1182 a 508a 437 a 191 a 375 a 11 b 18 a 22 a 28c Corn Northrup King 508 1047 be 1155 a 682 a 409a 162 a 147 a 43a 16 a 22 a 74ab Corn Pioneer X304C 1002 be 637 a 518 a 762 a 234a 437 a 18 b 11 a 16 a 104ab Corn Florida SYN-I 1363 a 1292 a 586a 654a 249a 306a 16b 19a 22 a 62b Sorghum DeKalb XS-17 830bc 599 a 647 a 6b Ob 4b 59a 6a 36a 166a Sorghum DeKalb FS25E 1799 a 1480 a 894a 10 b 5b 13 b 44a Sa 22 a 190a Corn vs. sorghum: ns ns ns *** *** *** *** ns ns *** ***Contrast significant at P :5 0.001. tWI = west field, first crop; W2 = west field, second crop; E = east field. :j:Means in columns followed by the same letter do not differ (P :5 0.05) according to Duncan's multiple-range test. ns = not significant at P :5 0.05. W2 E 255a 952a 332a 907 a 411 a 1102 a 600 s 998a 400a 120b 270a 204 b ns ***


PROCEEDINGS, VOLUME 51, 1992 45 from the crops grown (with e.g. P. minor). Sorghum, sorghum-sudangrass, and rye were effective alternatives to corn, soybean, or crimson clover in keeping populations of M. incognita low. Corn, although apparently tolerant of some M. incognita infection (McSorley and Gallaher, 1991a), will increase rather than decrease populations when used as a cover crop or rotation crop. Selection of appropriate cover crops, rotation crops, and crop sequences may be an effective method for reducing or managing some plant-parasitic nematodes. Preliminary research has been successful in identifying some candidate crops useful for nematode management, but it is clear that information on the response of many nematode species to a wide range of crops and cultivars is needed. As this information accumulates, producers will be able to make better informed choices of crop sequences which will minimize nematode numbers and damage. REFERENCES Freed, R., S. P. Eisensmith, S. Goetz, D. Reicosky, V. W. Smail, and P. Walberg. 1987. User's guide to MSTAT (Version 4.0). Michigan State Univ., East Lansing. Gallaher, R. N., D. W. Dickson, J. F. Corella, and T. E. Hewlett. 1988. Tillage and multiple cropping systems and population dynamics of plwtoparasitic nematodes. Suppl. J. Nematol. 20:90-94. Gallaher, R. N., R. McSorley, and D. W. Dickson. 1991. Nematode densities associated with corn and sorghum cropping systems in Florida. Suppl. J. Nematol. 23:668-672. Hiebsch, C. 1990. 1989 Florida soybean variety trials. Agro. Res. Rep. AY-90-01. Agronomy Dep., Univ. Florida, Gainesville. Jenkins, W. R. 1964. A rapid centrifugal-flotation technique for separating nematodes from soil. Plant Dis. Reptr. 48:692. Johnson, A. W. 1982. Managing nematode populations in crop production. p. 193-203 In R. D. Riggs (ed.) Nematology in the Southern Region of the United States. Southern Coop. Ser. Bull. 276, Arkansas Agric. Exp. Stn., Fayetteville. McSorley, R., and D. W. Dickson 1989. Effects and dynamics of a nematode community on maize. J. Nematol. 21:462-471. McSorley, R., and R. N. Gallaher. 1991a. Cropping systems for management of plant-parasitic nematodes. p. 38-45. In A. B. Bottcher, K. L. Campbell, and W. D. Graham (eds.) Proc. Environmentally Sound Agric. Conf. Orlando, 16-18 Apr 1991. Florida Coop. Ext. Serv., Gainesville. McSorley, R., and R. N. Gallaher. 1991b. Nematode population changes and forage yields of six corn and sorghum cultivars. Suppl. J. Nematol. 23:673-677. Zeck, W. M. 1971. A rating scheme for field evaluation of root-knot nematode infestations. Pfl.-Nach. Bayer 24:141-144.


PROCEEDINGS, VOLUME 51, 1992 45 Landsat and SPOT Imagery Classification for Land Use Change Analysis in Lee County, Florida J. D. Jordan and S. F. Shih* ABSTRACT The use of Landsat Thematic Mapper (TM) and Systeme Probatoire de l'Observation de la Terre (SPOT) images in monitoring land use change was demonstrated in Lee County for the period of 1983 to 1987. An overlay technique was used to create a land use change map from the processed images. Results of a four-category (Agricultural/Irrigated, Urban/Clearings, Forest/Wetlands, and Water) analysis indicate a near reversal of the land use change trend found in the same area in a 1972-1982 study. The 1972-1982 trend consisted of agricultural and urban expansion into forests and wetlands, while the reforestation of Agricultural/ Irrigated land was the primary trend from 1983 to 1987. Regional-scale assessment of shifts in such land use categories as agriculture, forest, wetland, and urban areas is of interest to water management and planning agencies. Obtaining land use information for large areas by conventional aerial-photography based techniques is a tedious process (Tan and Shih, 1990). Satellite-based land use assessment techniques offer a convenient alternative. Archived imagery suit-J. D. Jordan and S. F. Shih, Agricultural Engineering Dep., Univ. of Florida, Gainesville, FL 32611. Florida Agric. Exp. Stn. Journal Series no. R-02028. *Corresponding author. Mention of trade names is for convenience. and does not imply endorsement by the authors. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 :45-49 (1992) able for detailed land use change analysis is available from several satellite systems. However, the various satellite imagery types are characterized by different spatial resolutions and spectral bands (see Table I), which present a challenge to the land use change analyst. Established land-use classification and image georeferencing techniques arc available, but the practical application of these methods to the geographic information system (GIS) projects of many resourcemanagement agencies is still in the developmental stage. A demonstration of the use of Landsat Multispect ral Scanner (MSS) and Thematic Mapper (TM) im ages in monitoring land use change in Lee County, Florida was conducted by Jordan and Shih (1991). That project covered a study area consisting of 72% of Lee County for the period from 1972 to 1982. The trend observed was a net expansion of agriculture and urban land use, and a corresponding reduction in forest and wetlands. In order to further demonstrate the use of different satellite image sources in land use change analysis, a new project was undertaken in the same Lee County study area. The objective of this study was to monitor the land use change from 1983 to 1987 using TM and Systeme Probatoire de !'Observation de la Terre


46 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table 1. Satellite data for land use change analysis. Earliest Spatial Number of System vear resolution spectral bands m Landsat MSS 1972 80 4 Landsat TM 1982 30 7 SPOT 1986 20 3 Note: Landsat TM band 6 is thermal-infrared at 120 m resolution. SPOT has an additional panchromatic band at IO m resolution. (SPOT) images. This study period contains the era of land use changes brought about by the hard freezes of the mid-l 980s. MATERIALS AND METHODS TM and SPO' f images of the 208,354 ha Lee County study area (see Figure 1) were obtained for the dates of 13 July 1983 and 8 October 1987, respectively. This selection was made on the basis of cloudfree image availability. The image months were chosen from the south Florida warm wet-season, so that both the evergreen and deciduous tree species would comprise forested land-use. In addition, the rowcropped fields within the study area are typically planted only in the winter months, so that these fields are under a fallow condition in both July and October. Land use change maps were derived from the two images by a four-stage process of classification, georeferencing, class identification, and land-use overlay analysis. The Earth Resources Laboratory Application Software (ELAS) image-processing package was used to obtain spectral classes by the unsupervised maximum-likelihood technique. The details and advantages of this Bayesian statistical method have been described by Lillesand and Kiefer ( 1979) and Graham et al. (1985). Bands 2 (green), 3 (red), and 5 (near-infrared) of the TM image were used to obtain 48 spect ral classes for land use classification; while bands 1 (green), 2 (red), and 3 (near-infrared) of the SPOT image produced 59 spectral classes. ELAS georeferencing of the classified images was accomplished using road intersections as ground-control points (GCPs). The geographic coordinates of the GCPs were transferred from USGS 7.5 Minute quadrangle sheets to the images using a digitizing tablet. Resampling for registration (geographic correction) produced a pixel size of 20 m for the SPOT image, and 40 m for the TM image. This registration was performed using a first-order global polynomial surface model and the nearest-neighbor resampling technique described by Lillesand and Kiefer ( 1979) and Beverley and Penton (1989). Thus, the pixels of the registered SPOT image were the same size (20 m) as those of the original SPOT image, while the pixels of the registered TM image were somewhat larger (40 m) than those of the original TM image (30 m), in order to allow an exact overlay of the SPOT and TM images for change analysis. The root-meansquared (r.m.s.) error was kept to 0.5 pixel or less. Identification of the land use type corresponding to each spectral class was performed for both images using 1 :24,000 scale aerial color-infrared (ACIR) photographs of a 61,800 ha portion of the study area, in addition to visits to the ground-truth sites denoted as fields Bl and B2 in Figure 1. The identified spect ral classes were then aggregated into four principal categories: I) Agricultural/Irrigated land-consisting of row-crop fields, pastures, large lawns, golf courses, and grassed roadsides; 2) Urban/Clearings-including buildings, pavement, construction sites, and clear ings (for new citrus groves, etc.); 3) Forest/Wetlandconsisting of pine flatwoods, palm hammocks, mature citrus groves, swamps, exotic pest-tree stands, and open marshes; and 4) Water-including lakes, rivers, and seawater. Pine flatwoods consisted of natural and plantation Slash Pine (Pinus elliotii Engelm.) with an understory of Saw-palmetto (Serenoa repens Bartr.) mixed with various evergreen shrubs (Ceratiola ericoides Michx., /lex spp., Lyonia spp., etc.) and wild grasses. Palm hammocks consisted of pure stands of Cabbage Palm (Sabal palmetto Walt.). Swamps primarily contained Bald-cypress (Taxodium distichum L.), various Oaks (Quercus hemisphaerica Bartr., Q. nigra L., and Q. vir gi,niana Mill.), Pop Ash (Fraxinus caroliniana Mill.), Wax-myrtle (Myrica cerifera L.), and Willow (Salix caroliniana Michx.); and near the coast contained Red Mangrove (Rhizophora mangle L.), Black Mangrove (Avicennia germinans L.), Strangler-Fig (Ficus aurea Nutt.), and Sea-grape (Coccoloba uvifera Jacq.). Exotic pest-tree stands consisted of Punk-tree (Melaleuca leucadendron L.), Brazilian Pepper (Schinus terebin thifolius Raddi), and Australian Pine (Casuarina equisetif olia Forster). Overlaying the registered 1983 and 1987 class ified land-use images was accomplished using ELAS in combination with a program that recognized the 16 possible combinations of the four land use categories. Figure 2 shows the 1983-1987 land use change map for the study area. An enlarged portion of the land use change map corresponding to the ground-truthed field B 1 is shown in Figure 3. RESULTS AND DISCUSSION Overlaying the 1983 TM and 1987 SPOT images produced 12 land-use change categories, of which 6 were large enough to be notable(~ 5% of areal extent of 1983 land use category). These results are given in Table 2. The largest percentage land use changes from 1983 to l 987 occurred for the Agricultural/Irrigated category (64.66% changed to Urban/Clearings and Forest/Wetland) and the Urban/Clearings category (41.74% changed to Agricultural/Irrigated and Forest/Wetland). About 11.79% of Forest/Wetland underwent change to Agricultural/Irrigated and Urban/Clearings land use. The other land use


PROCEEDINGS, VOLUME 51, 1992 47 N t o--FORT MYERS FIELD 82 I I / I HENDRY COUNTY I I I I I I I I I----, I I I I I L ____ .1 I 1--ACIR I I I I / STUDY AREA I I I I I I I I ---.1 LANDSAT TM AND SPOT STUDY AREA rc----,COl..l,E~ ICOUNTY I I I I J I I I I I I I Fig. 1. Lee County study area. changes, which involved water-bodies, were relatively minor in extent. The net results (see Table 3) for the 1983-1987 period were a substantial decrease in Agricultural/Irrigated land, a substantial increase in Forest/Wetland (due to reforestation), and a small increase in Urban/ Clearings (due to urban growth and development of new citrus groves). The very minor increase in Water may be attributable to a combination of tidal differences, lake-level differences, and the construction of retention ponds (some temporary, some permanent) in both agricultural and urban areas. However, it is so small that, for the purposes of this study, it may be considered inseparable from the edge-effect inherent in overlaying raster imagery. Visits to the study area in 1985 and 1987 indicated that the reforestatio~ of pastures and row-crop fields was taking three forms: 1) the spread of exotic pest trees onto abandoned land Qordan, 1987); 2) conversion to pine plantations; and 3) maturing of new citrus groves. The exotic pest trees consist of fast-growing and readily-spreading species (particularly Brazi-


48 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Fig. 2. Lee County study area 1983-1987 land use change map. Land use categories are denoted A for Agricultural/ Irrigated, U for Urban/Clearings, F for Forest/Wetland, and W for Water. Fig. 3. Enlarged portion of 1983-1987 land use change map. This area corresponds to field Bl in Figure I. Pixel size is 20 m. Land use categories are denoted A for Agricultural/Irrigated, U for Urban/Clearings, F for Forest/Wetland, and W for Water. lian Pepper), which under the local conditions are capable of reforesting a disturbed site in only 4-5 years. Factors influencing the reforestation of Agricultural/Irrigated areas could include the abandonment of many row-crop operations in response to both changing winter-vegetable profitability and state-mandated plugging of artesian wells Qordan and Shih, 1988), as well as a southward shift in Table 2. Landuse change trends in Lee County study area. Percent changed to: Percent changed Agric./ Urban/ Forest/ Landuse type (1983 1987) Irrig. Clearing Wetland Water Agric./Irrig. 65.02 32.05 32.61 0.36 (55,498 ha) Urban/Clearings 42.88 22.36 19.38 1.14 (37,445 ha) Forest/Wetland 14.05 6.17 5.62 2.26 (78,776 ha) Water 3.38 0.09 1.31 1.98 (36,635 ha) Table 3. Overall land use changes in Lee County study area. Land use type Agricultural/Irrigated 1983 TM 1987 SPOT Urban/Clearings 1983 TM 1987 SPOT Forest/Wetland 1983 TM 1987 SPOT Water 1983 TM 1987 SPOT Percent of total area 26.66 15.69 17.96 21.15 37.80 45.02 17.57 18.13 Change in percent of total area -10.97 +3.19 +7.22 +0.56 Florida citrus production m response to repeated hard freezes. CONCLUSIONS The results of this 1983-1987 study indicate a near reversal of the land use change trend found in the same Lee County study area in the 1972-1982 study. The conclusion of the earlier study was that forests and wetlands had given way to agricultural and urban expansion. Reforestation of agricultural land was the primary trend observed in this 19831987 project. Change patterns observed in both studies will be of interest to further investigations of links between land use and surface-hydrologic parameters such as runoff factors and temperature patterns. This study demonstrates that with proper attention to ground-truthed land-use classification and image registration, images from different years and satellites can be overlain for practical land-use change assessment.


PROCEEDINGS, VOLUME 51, 1992 49 ACKNOWLEDGEMENTS This research was partially funded by the St. Johns River Water Management District (SJRWMD) and the South Florida Water Management District (SFWMD). REFERENCES Beverley, A. M., and P. G. Penton, eds. I 989. ELAS Science and Technology Laboratory Software, Vols. I and II, May 1989 re vision. National Aeronautics and Space Administration, John C. Stennis Space Center, Science and Technology Laboratory, MS. Graham, M. H., B. G. Junkin, M. T. Kalcic, R. W. Pearson, and B. R. Seyfarth. 1980. ELAS---Earth Resources Laboratory applications software, Vols. I and II, January 1985 revision. National Space Technology Laboratories, Earth Res. Lab., MS. Jordan, J. D. 1987. Application of remote sensing techniques to abandoned well assessment in Lee County, Florida. Masters thesis, University of Florida, Gainesville, FL. Jordan, J. D., ands'. F. Shih. 1988. Use of remote sensing in abandoned well assessment. Trans. Am. Soc. Agric. Engineers, 31(5): 1416-1422. Jordan, J. D., and S. F. Shih. 1991. Satellite and aerial photographic techniques for use in artesian well assessment. Intl. Conference on Computer Applications in Water Resources (ICCWR) Proceedings, Vol. 2., Tamkang University, Taiwan, R.O.C. Lillesand, T. M., and R. W. Kiefer. 1979. Remote sensing and image interpretation. John Wiley & Sons, New York, NY. Tan, Y. R., and S. F. Shih. 1990. GIS in monitoring agricultural land use changes and well assessment. Trans. Am. Soc. Agric. Engineers, 33(4): 1147-1152.


PROCEEDINGS, VOLUME 51, 1992 49 Three Potential Amendments for Better Fertilizer Utilization in Sandy Soils T. L. Yuan ABSTRACT Conservation of soluble nutrients to increase the fertilizer efficiency in sandy soils may be achieved by using lime, gypsum, and zeolite as amendment materials. Lime not only increases soil pH and supplies Ca and Mg if dolomite is used, but also interacts with soluble P to form slow-release Ca-phosphates. Gypsum and the byproduct of the Florida phosphate industry, phosphogypsum, when used in combination with lime would achieve the same purpose and avoid over-liming. Zeolites having a high cation exchange capacity may effectively retain soluble cations from fertilizers, especially K and NH4-N, against leaching loss in sandy soils. The K-and NH4-saturated zeolites may be considered as slow-release nutrient sources. Minerals that contain P and K may render these nutrients more available when mixed with the zeolites. Potential benefits of the zeolites for Florida sandy soils need to be substantiated. More than four thousand years ago, farmers in China learned to use farm manure, night soil, crop residues, wood ash, composts, etc., to increase crop yields. They found that applications of green manure and oilseed meals were equally helpful. Such amendments were also used in Europe in early Greek and Roman periods. As our knowledge of plant nutrition increases, other amendment materials have been added to the list, including guano, bone meal, lake brines, rock phosphate, basic slag, and most important of all, the commercial fertilizers which are manufactured either by synthesis or by processing of raw minerals to supply one or more plant nutrients. Although the supply of plant nutrients is the prime purpose for adding soil amendments, they have also been used for the improvement of soil phys-T. L. Yuan, Soil and Water Science Dep., Univ. of Florida, Gainesville, FL 32611-0313. Fla. Agric. Exp. Stn. Journal Series no. N-00529. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 :49-55 (1992) ical, chemical, and/or microbiological conditions. Ap plication of organic manures is not only to supply plant nutrients but also to improve the soil structure by forming aggregates and to conserve water. The use of lime in acidic soils may remove possible toxic effects of certain elements and stimulate microbial activities. In the late l 940's and early 1950's, synthetic materials called soil conditioners were developed to increase the water stability of aggregates and discussed in the June, 1952 issue of Soil Science. A status report on these materials also appeared in the January, 1953 issue of Soil Science Society of America Pro ceedings. Since the energy cost is getting higher in recent years and heavy fertilization in sandy soils may cause environmental problems, attention is to be given to a new group of amendments which may preserve the nutrient elements against leaching and thus, increase the fertilizer efficiency. In this paper, the possible use of lime, gypsum, and zeolites as potential amendments for this purpose is discussed. SOME CHARACTERISTICS OF SANDY SOILS To discuss effectively the amendments for Florida sandy soils, an understanding of soil characteristics is necessary. Florida is a part of the coastal plain. Its soils range in elevation from sea level to slightly over 100 meters. Limestone underlies the entire state but is exposed only over limited areas. The limestone deposits are commonly buried beneath more recent deposits of sands, clays, marls, and organic materials. Under the influence of various soil-forming proces ses, widely different soils have developed. With the exception of Vertisols, Aridisols, and Oxisols, all other soil orders are represented in Florida. Among


50 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA them, Spodosols, Entisols, and Ultisols are the most prominent (Fig. 1). There are also about 2.4 million acres (about 10,000 km2 ) of Histosols in the State (Smith et al., 1967). A large acreage is in the Everglades, south of Lake Okeechobee. Spodosols are distributed most widely in the state. They are derived from marine deposits of non-calcareous sands, occurring in nearly level to level or depressional areas. Their water table is high and drainage is usually poor. Incomplete decomposition of organic matter in the surface horizon produces considerable amount of organic acids which dissolve minerals and during the leaching process, carry the dissolution products with them to the lower depth of the profile. These organo-mineral complexes may be deposited within two or three feet of the surface to form a dark to black, compact or locally cemented layer called spodic horizon, also referred to as organic hardpan. Between the surface and spodic horizons is the coarse-textured, light-colored E horizon. The Spodosols are acidic and contain very little clay, although many Spodosols may have an argillic horizon below the spodic horizon with a high clay content of about 150 to 350 g kg-1 For the surface soils, a clay content less than 10 g kg-' is common and the fertility is low. These soils have little retention capacity for either cationic or anionic nutrients in the surface. Entisols are most common on the ridge sections of the State. They are usually well to excessively drained soils and developed from marine deposits of non-calcareous sands and clays. They may also be influenced by limestone and calcareous sands and clays. They have a deep and slightly acidic profile and a texture ranging from sand to loamy sand. Their fer tility status is low but generally higher than that of Spodosols. The leaching of nutrients is also consider able. Ultisols occur on coastal plains and are derived either from marine deposits of non-calcareous sands SOIL ORDERS 1. ALFISOLS 2. ARIDISOLS .. 3. ENTISOLS 4. HISTOSOLS 5. INCEPTISOLS .. 6. MOLLISOLS .. 7. OXISOLS ... 8. SPODOSOLS 9. ULTISOLS 10. VERTISOLS ... *Widely interspersed areas *Minor occurrences ***None recognized in Florida Note: Small areas of contrasting soils would be shown at a larger mapping scale. Fig. 1. Soils of Florida. and clays or from limestone and calcareous sands and clays. They may also have developed from alluvial materials. The texture of their surface horizon ranges from sand to sandy loam, but clay content generally increases with depth. These soils are highly weathered and have a deep profile with a slightly to moderately acidic reaction. Their cation exchange capacity (CEC) may still be low, but the fertility level is generally higher than members of the other two soil orders. During the soil-forming process, many bases have been leached. Left behind in the clay fraction are minerals including hydrous Al and Fe oxides. Therefore, P retention in Ultisols is considerably higher than that in the other soils. Although profile characteristics and properties are very different among soils of these three orders, they each have a sandy epipedon or surface soil. In general, the CEC of the surface horizon or the capac ity to retain basic nutrients is low (Table 1), either due to their low clay contents or the nature of their clay fraction. The amount of organic matter is very important in these soils. Most of their CEC is from this fraction. The retention of non-metallic nutrients, such as P, is similar. Although phosphate is relatively immobile in Ultisols, its movement in other soils can be considerable (Neller, 1947). Therefore, heavy fer tilization on Florida cash crops may induce a large leaching loss of nutrients and thus, reduce the fer tilizer efficiency. When a compact and impermeable layer is present in the lower horizon, such as an argil lic or a spodic horizon, a very large portion of dissol ved nutrients in the leachates may be transported to streams, rivers, and lakes causing environmental problems. These problems occur particularly in the Spodosol areas. The use of suitable amendments may alleviate at least some of these problems. POTENTIAL AMENDMENTS Liming Materials Liming of acidic soils has been recognized for over a century as an important practice to reduce soil acid ity. It raises the soil pH to a proper level to give a more favorable condition for better crop growth. For acidic soils in Florida, liming does more than correct soil acidity. Inasmuch as Florida soils are generally low in nutrients, application of calcite supplies Ca as a nutrient and that of dolomite supplies both Ca and Mg. The inclusion of liming materials as amendments Table 1. Selected characteristics of surface horizons of virgin sandy soils. Entisols ( 56) Spodosols (37) Ultisols (91) (Yuan, unpublished data) pH O.M. Clay ----g kg' ----5.5 5.0 5.4 17 17 IO 22 11 66 t Number of samples tested in parentheses. CEC mmol( +) kg' 33 34 32


PROCEEDINGS, VOLUME 51, 1992 51 in this discussion is aimed at their reactions with other nutrients. As mentioned earlier, Florida's acidic sandy soils are of very low fertility. Highly profitable cash crops such as vegetables, ornamentals, and fruits, particularly citrus, have been fertilized very heavily. It is interesting to note that cultivated soils generally contain considerable amounts of nutrients, particularly P. Rhue et al. ( 1979) summarized their soil test results of more than one thousand samples each of Entisols, Spodosols, and Ultisols, and found that 63, 48, and 55% of the samples had high P values for the three respective groups (Table 2). Potassium values were rated medium and high for 31, 42, and 42% of the samples, respectively. The median Ca contents ranged from 1955 kg ha-1 (1744 lbs ac1 ) in Spodosols to 780 kg ha 1 (696 lbs ac1 ) in Ultisols with ~ntisols being intermediate between the two. Magnesmm accumulation was similar, averaged 133 kg ha1 (l 18 lbs ac1 ) in Spodosols and 73 kg ha-1 (65 lbs ac1 ) in Ul tisols. The high level of P accumulation ~n the Spodosols is particularly striking,. b~caus~ soil ~onstituents which are capable f retammg this nutrient are in very minute quantities in the surface horizon of these soils. The only explanation is the possibility of an interaction between the applied lime and P. Table 3 shows the total and Mehlich I-extractable nutrient contents of an Ellzey sand. The samples were taken from four plots of a field experiment. Plots L and P represent lime and P treatments, respectively, and subscripts 0 and 2 were rates applied three years earlier. This soil, which had been under cultivation for over 70 years, contained 18 and 15 g kg-1 of clay and organic matter, respectively, with a CEC of 28 mmol( +) kg-1 of soil. The high Ca and Mg contents may be attributed to the dolomite applications in the past management program. The a~ditional lime application (2.8 tons ac1 or 6.27 metr~c tons ha-1 ) ~hree year earlier further mcreased their contents m L2 plots. Both P and K contents were very high. The former may be assumed largely 'available' as indicated by the values obtained from the extraction of Mehlich I reagent. The extremely low extractable K values were a result of K being in insoluble mineral forms. Although the plot variation of P contents w~s co?-siderable, data showed a greater accumulat10n m Table 2. Phosphate and potassium levels of cultivated sandy soils in surface horizon. (Rhue and Sartain, 1979) Nutrient Phosphate Potassium Entisols (1150) Spodosols (1004) Ultisols (1096) Entisols ( 117 4) Spodosols ( 1039) Ultisols ( 1061) Levels of P or K Low Medium High -----------%of samples---------25 42 26 69 58 58 13 10 19 17 19 24 63 48 55 14 23 18 tNumber of samples tested in parentheses. Table 3. Total and Mehlich I-extractable nutrients from four soil treatments. (Yuan et al., 1985) Content Treatment"" p K Ca Mg Zn Mn ----------------------mg kg-1---------------------Total LoPo 540 690 810 230 130 26 LoP2 521 620 690 210 151 24 L2Po 614 815 1155 315 144 28 L2P2 697 695 1310 290 151 33 Mehlich I-extractable LoPo 270 68 520 40 10 5 LoP2 169 44 353 20 7 4 L2Po 290 60 700 80 8 5 L2P2 336 60 740 76 5 5 tLo and Po: no lime and phosphate applied. L2 and P2 in: 6278 kg ha-1 of dolomite and 48 kg ha of elemental phosphorus diammonium phosphate resp., applied three years earlier to an Ellzey sand (Typic Humaquept). limed plots (Table 3). The effect oflime on P accumulation in the soil profile, as expressed in Mehlich I-extractable form, is shown in Fig. 2. The P accumulation curves were parallel to those of Ca. The relationship is statistically significant. No such relations were shown with Al and Fe. Blue ( I 970) studied effects of lime on retention of applied P in an acidic Leon fine sand and found that the retention was linearly related with the amount of lime applied. The solubility of the retained nutrients is illustrated in Fig. 3. The amounts of water-soluble Ca and Mg were larger from the additional lime plot than from the plot without lime. More K and P, on the other hand, were found to be water-soluble from the plot without additional lime. The results indicate that lime reduced the leaching of both nutrients. The amount of release each time from the soils was small but continuous, suggesting that it had become a slow-release fo~m. The advantage of lime application in the reduction of P movement has been recognized. In some dairy farms in south Florida, application of lime has been 0 800 Double -Acid Extractable, mo kg Fig. 2. Distribution of double-acid-extractable P, Ca, Al, and Fe in soil profiles (Yuan, et al., 1985).


52 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA 60 }40 r : 20 a: 1: 0 -! p 'S z I 40 120 (.) Number of Leochlna, 50ml Fig. 3. Cumulative nutrient release by water leaching (Yuan, et al., 1985). made in the form of sea shells or lime rocks in intensive animal loading areas. Calcitic and dolomitic limestones are not the only liming materials. The others include burnt or quick lime, hydrated lime, marl, sea shells, as well as al kaline industrial by-products. Their effects on retention of nutrients may well be different, at least in their reaction rates. However, Blue (1970) did not find any difference in his study of the effect of two lime sources, CaC03 and Ca(OHh. Gypsum and Phosphogypsum It is known that over-liming has an adverse effect on plant growth. The immobilization of some essen tial micronutrients is one of the disadvantages. To reduce the P movement in soils without over-liming them, a neutral Ca source may be considered. Gypsum is a salt of a strong acid and a weak base by nature. Its reaction is slightly acidic. Addition of gypsum as a soil amendment increases the amount of soluble Ca which may interact with P to form certain compounds having lower solubility. However, the amount of retention was rather small with gypsum alone (Table 4). When applied in combination with lime, the efficiency of gypsum to retain P was greatly increased. It is probably due to the fact that gypsum has a relatively high solubility and its reaction with P is not as efficient as when it is hydrolyzed in the pres-Table 4. Phosphate retention by gypsum and phosphogypsum with and without calcium carbonate addition. (Yuan and Lucas, 1989) Calcium carbonate (g) 0 0.2 0.5 Amount Materialt used pH p pH p pH p g mg mg mg Gypsum 0 5.21 0 8.08 1.64 8.00 0.98 0.5 5.35 0.05 7.44 1.96 7.59 1.97 1.0 5.22 0.18 7.26 1.98 7.44 1.96 2.0 5.33 0.26 7.28 1.97 7.50 1.97 Phosphogypsum 0 5.21 0 7.93 1.66 7.92 1.12 0.5 5.12 -0.04 7.09 1.77 7.26 1.85 1.0 4.93 -0.11 7.10 1.86 7.28 1.89 2.0 4.95 -0.10 6.90 1.67 7.07 1.76 t All combinations were treated with a 25 mL solution contain ing 2.0 mg P. ence of lime to form a separate solid phase. Increase of P retention by increasing amounts of either lime or gypsum in the mixture was not shown in Table 4 because only 2 mg P was used in the experiment. This amount of added P was almost completely retained by the mixture of 0.5 g of gypsum and 0.2 g of calcium carbonate. Two dairy-farm soils, which were high in pH and Pas a result of sea shell incorpo ration, were incubated with gypsum (Yuan et al., 1989). The water-extractable P of the Myakka soil was greatly reduced (Table 5). Although the soluble P content of the Immokalee soil was low, the gypsum effect was still considerable. Due to increasing uses of triple superphosphate fertilizer, soil content of Sas a nutrient becomes limiting for plant growth (Blue et al., 1981). Gypsum is a natural S source. In central Florida, there are stockpiles of phosphogypsum available as a by-product of the P industry. The total amount is estimated to be several hundred thousand tons. More will be generated each year. Only a very small fraction has been used agriculturally or elsewhere. This material may be used not only as a possible economic source of Ca and S but also of P. For this reason, it was included in the gypsum study (Yuan et al., 1989) and the results are also shown in Table 4. The particular phosphogypsum sample contained 3.2 g kg-1 elemental P. Because of its low pH (4.5), there was a release Table 5. Waterand Mehlich I-soluble P of gypsumand phosphogypsum-treated soils with a high pH and high P (Yuan and Lucas, 1989). Myakka Immokalee Amount Materialt applied pH Water Mehlich pH Water Mehlilch g mgkg-1 mgkg-1 None 0 8.22 264 950 7.02 16 198 Gypsym 1 7.29 65 889 6.53 11 184 2 7.27 55 860 6.59 11 195 5 7.26 43 873 6.58 12 182 Phosphogypsum 2 7.22 73 933 6.53 24 212 5 7.18 76 947 6.49 29 212 tThe materials were added to 100 g soil samples and incubated for one week at field-moisture capacity.


PROCEEDINGS, VOLUME 51, 1992 53 of P even in the presence of added P. When lime was present, more P was retained by the lime-phosphogypsum mixture than by lime alone at a high lime level. The reason for larger differences at the high lime than at low lime level was discussed in the cited report. Like gypsum, phosphogypsum may be used to reduce the solubility of Pin water, but only in soils with high P accumulation (Table 5). In low soluble P soils, some P may be released from phosphogypsum itself. However, the presence of radioactive isotopes in phosphogypsum has caused some concern for using this material in agriculture because of the pos sibility of these isotopes accumulating in the soil, moving into the groundwater, and being absorbed by plants. In their review on gypsum use in soils, Shainberg et al. ( 1989) using 226Ra and Cd data from Mays and Mortvedt claimed that phosphogypsum should not present environmental problems when used at appropriate rates (0-10 Mg ha1 ) for agricultural purposes. Zeolite Zeolites are hydrated alumino-silicate minerals. They were first discovered more than 200 years ago. Aluminum and Si oxides in these minerals form an infinitely extended framework which encloses interconnected channels or cavities occupied selectively by relatively large exchangeable cations and water molecules. Zeolites have high CEC. Because of these properties, they have been used in industries as molecular sieves, ion exchangers, and catalysts. Other uses have also been made commercially and experimentally including NH4 ion removal for aquaculture and uranium mine wastewater; odor control for chicken farming and cat litter; and removal of heavy metal ions from nuclear, mine, and industrial wastewaters (Clifton, 1987). Because large zeolite de posits were not found until in the late l 950's, many of the zeolites used in industries have been synthetic products. At the present time, more than 100 zeolites have been synthesized and about 50 natural zeolites identified (Clifton, 1987). Since late l 950's, large deposits have been discovered in many countries (Haw kins, 1984). Their distribution and property holders in the United States were reported by Sheppard (1971) and Hawkins (1984), respectively. One zeolite occurrence was recognized in Florida near Caryville of Washington County. The total natural zeolite re sources in the United States, as cited by Clifton ( 1987) were estimated to be 10 trillion tons. Only 13000 tons were mined. Attempts to use zeolites in agriculture have been made only recently. Excellent reviews on their potential uses have been compiled by Pond et al. (1984). Ming et al. ( 1989) listed them as soil amendments, slow-release fertilizers, dietary supplemer-tts in animal nutrition, carriers of pesticides, and deodorizers and moisture-control agents for animal manures. The re sults from available research seem to suggest that they will be excellent amendment materials for sandy soils. They have high CEC, ranging from 1 to 3 mmol( +) g1 material. The cations on their exchange sites are alkali and/or alkaline-earth metals, i.e., Ca, Mg, K, and Na, depending on the zeolite. Those which contain the first three would serve as nutrient sources. They may act as a reservoir for these nutrients from fertilizers through ion exchange and reduced leach ing. In addition, they are generally stable at an acidic pH. Among the identified zeolites, only eight are suf ficiently abundant to be mined. They are analcime, chabazite, clinoptilolite, erionite, heulandite, laumonite, mordenite, and phillipsite. For agricultural purposes, chabazite, clinoptilolite, erionite, mordenite, and phillipsite are more useful with clinoptilolite ranking first (Hawkins, 1984). Table 6 shows average compositions of these five zeolites as given by Sheppard et al. (1982). Barbarick et al. (1984) reviewed the work on zeolites used in agronomy and horticulture. They cited literature on the reactions between NH4 ions and zeolites. The high affinity and selectiv ity of clinoptilolite, phillipsite, and erionite for NH4 ions would reduce the loss of N from NH4 fertilizers through volatilization and nitrification. When secondary sewage effluents and other industrial was tewaters are treated with clinoptilolite, NH4-saturated byproducts may eventually be used as N fertiliz ers. MacKown et al. (1985) conducted a column study on a loamy sand, an Entisol. The soil was treated with two different zeolites, one rich in clinoptilolite and the other rich in erionite. The water-saturated soil zeolite mixtures were added with NH4-N. The columns were then leached with deionized water containing a nitrification inhibitor. They found that zeol ites effectively enhanced the retention of NH4 ions Table 6. Composition of the more agriculturally-useful zeolitic tuffs (Sheppard and Gude, 1982). Dominant mineral in the tuffl" Chabazite (5) Clinoptilolite ( 16) Erionite (7) Mordenite (5) Phillipsite (5) tNumber of samples tested in parentheses. Average Composition MgO CaO ----------------------------------------------------------g kg I ----------------------------------------------------------561 142 19 16 35 25 17 659 119 17 9 17 27 26 587 133 17 II 27 30 40 646 111 22 20 26 25 34 563 142 25 11 8 51 50


54 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA K 180 Z 160 ... 140 120 tu :3 100 u.. tf:j BO tu > 60 .J 40 :::, ::;; :::, 20 u 0 zEouTE RATE r, ,,-,, ZEOLITE E -.-00 -0-12.5 25,0 -o-50.0 180 160 140 120 100 80 60 40 20 ZEOLITE K 0 lidCJ...-L---'--L--_._-----..1...J 0 0,2 0,5 1,0 1,5 2.0 2.5 3.0 0 0,2 0,5 1,0 1,5 2,0 2.5 3.0 EFFLUENT VOLUME (liters) Fig. 4. Cumulative effluent NH4-N leached from Rositas soil receiving erionite (E) and clinoptilolite (K) natural zeolite amend ments (MacKown and Tucker, 1985), against leaching (Fig. 4). In addition to N, Barbarick et al. (1984) cited the work by Hershey et al. and stated that a large proportion of the exchange sites of some zeolites was filled with K and these zeolites could serve as slow-release K fertilizers. Lai et al. (1986) studied the P release from a Florida phosphate rock (mainly carbonate apatite) by mixing it with a natural clinoptilolite-rich zeolite tuff as well as the same material saturated with NH4 Na or H ions. They found that the water-extractable P increased in the following order: phosphate rock alone < natural zeolite < NH4-zeolite < Na-zeolite < H-zeolite (Table 7). They also mixed soils from four different soil orders with the phosphate rock and treated the mixtures with NH4-saturated clinoptilolite. Increase in water-extractable P was also noted with the treatment. The amount of increase declined as the number of extractions increased. Addition of fresh NH4-saturated clinoptilolite or replenishment of NH4 ions in the form of NH4Cl to replace Ca in the zeolite further increased the P release by water (Fig. 5). Barbarick, et al. (1990) used an NH4-exchanged clinoptilolite which was mixed in soils with a North Carolina rock phosphate to evaluate its effect on P release in a pot experiment. The ratios of zeolite to the rock phosphate varied from O to 7 .5 at 1.5 increments. They found that the dry matter yields of sorghum-sudangrass from a Weld loam (Aridic Paleustoll) increased linearly with increasing zeolite:rock phosphate ratios. Plant P concentrations and uptake were also generally increased. The plant response was not that definite from another soil, a Read Feather ls (Lithic Table 7. Effect of clinoptilolite-rich tuff on the dissolution of phosphate rock. (Lai and Eberl, 1986) Form oftuff Natural NH4-saturatcd Na-saturated H-saturated Phosphate rock only mgL-1 P released 2.8 8.3 I0.4 67.6 0.6 Final pH 7.35 8.07 7.6'.1 3.75 7.20 6 A B 5 en 4 E "O Ql 3 en 0 Q) Ql 0::: 2 a.. -ep~t__. ..... ..-....... __.. 0 L---'-~-~---'-~ L_...J__L___J__L.,_...,___,L.___,_...._,_-'---" 0 2 4 6 8 100 4 8 12 16 20 Number of Extractions Fig. 5. Renewal of P release from zeolite-phosphate rock-soil systems by additional NH4-clinoptilolite (A) and NH4Cl (B) ap plications after 7th and 14th water extractions, respectively (Re drawn from Figs. 2 and 3 of Lai and Eberl, 1986. Several data in Bare new and slightly different from the original). Four soils used, a Haplustoll, a Paleudult, an Eutrustox, and a Dystrandept, are represented by -oil, -ult, -ox, and -ept, respec tively. Cryoboralf), which was deficient in K, but application of zeolite did increase the total P uptake. Therefore, P may be applied with zeolite to sandy soils in rock phosphate form rather than in soluble P form as in fertilizers to avoid P leaching. Amendment of sandy soils with zeolite to increase the growth and quality of plants has also been demonstrated. Barbarick, et al. (1991) applied a zeolite from South Dakota to a Keith soil with a mill shale rock from Idaho containing 9.9% P. They found that the application of zeolite had a beneficial effect on the dry matter production, nutrient concentration, and nutrient uptake of sorghum-sudangrass. Ferguson et al. ( 1986) in Arizona found that amendment of sand with 5 or 10% zeolite (clinoptilolite) significantly increased germination and establishment of creeping bentgrass. The quality of the turf grass was also increased. SUMMARY Amendment materials discussed above may prevent leaching loss of nutrient elements through adsorption, precipitation, or exchange reaction and thereby result in better fertilizer utilization. The materials themselves may become nutrient sources as a result of their interactions with fertilizers. Among the three discussed, lime has long been put into practice, though not for the stated reasons. Application of gypsum with lime will probably increase the effectiveness in reducing the P movement in sandy soils and preventing overliming. For cationic nutrients, particularly K and NH4-J\;, the use of zeolites has great potential benefits. These nutrients can be held on the exchange sites of the zeolites against extensive leaching loss when soluble fertilizers are applied. Zeolites


PROCEEDINGS, VOLUME 51, 1992 55 which contain considerable amount of K as the predominant exchangeable cation can be used directly as a K source. The NH4-N held by zeolites may not be readily lost through nitrification or gaseous forms. Their high CEC will certainly help improve the chemical properties of sandy soils. Parham ( 1989) of the Congressional Office of Technology Assessment, US Congress, has assessed the potential benefits of zeol ites in agriculture of developing countries and discussed how the application of zeolites would improve chemical properties of different soils, including high rainfall/temperature tropical soils, acidic sandy soils, and arid/semi-arid soils. He also cited Mumpton's work that addition of zeolite to sandy soils would be extremely effective at removing moisture from the air even where the humidity is low. This moisture, he concluded, could benefit crops during extended periods of little rainfall. Since zeolites are relatively new in agriculture as a soil amendment, more studies are needed both in the laboratory and in the field. However, they will certainly be a promising amendment material and should receive the attention of Florida agricultural scientists .. In addition to the three materials discussed, there are other amendment materials, such as colloidal phosphate from settled slime ponds, sludges and wastewaters from municipal wastes, fly ash from coals, calcium silicate slags, etc., which contain considerable amounts of nutrients. They can be used as nutrient sources but do not have the properties of conserving the soluble fertilizer nutrients, except perhaps colloidal phosphate. Therefore, they are not discussed in this paper. REFERENCES Barbarick, K. A., and H.J. Pirela. 1984. Agronomic and horticultural uses of zcolites: A review. 93-103. In Pond, W. G. and F. A. Mumpton (eds.): Zea-Agriculture. Westview Press, Boulder, co. Barbarick, K. A., D. D. Eberl, and T. M. Lai. 1991. Pine Ridge zeolite and Fort Hall mill shale P effects on sorghum-sudangrass. Colorado Ag. Exp. Stn. Tech. Bui. TB91-2. Barbarick, K. A., T. M. Lai, and D. D. Eberl. 1990. Exchange fertilizer (phosphate rock plus ammonium-zeolite) effects on sorghum-sudangrass. Soil Sci. Soc. Am. J. -~4:911-916. Blue, W. G. 1970. The effect of lime on retention of fertilizer phosphorus in Leon fine sand. Soil Crop Sci. Soc. Florida Proc. 30:141-150. Blue, W. G., E. Jacome, J. Afre, D. Perez, S. Brown, and D. W. Jones. 1981. Sulfur and manganese deficiencies as causes of poor plant growth on Florida sandy soils. Soil Crop Sci. Soc. Florida Proc. 40:95-101. Clifton, R. A. 1987. Natural and svnthetic zeolites. U.S. Bur. Mines Info. Circ. IC 9140. Ferguson, G. A., I. L. Pepper, and W.R. Kneebone. 1986. Growth of creeping bentgrass on a new medium for turfgrass growth: Clinoptilolite zeolite-amended sand. Agron. J. 78: I 095-1098. Hawkins, D. B. 1984. Occurrence and availability of natural zeol ites. 69-78. In Pond, W. G. and F. A. Mumpton (eds.): Zeo-Agriculture. Westview Press, Boulder, CO. Lai, T. M., and D. D. Eberl. 1986. Controlled and renewable release of phosphorus in soils from mixtures of phosphate rock and ammonium-exchanged clinoptilolite. Zeolite 6:129-132. MacKown, C. T., and T. C. Tucker. 1985. Ammonium nitrogen movement in a coarse-textured soil amended with zeolite. Soil Sci. Soc. Am. J-49:235-238. Ming, D. W., and F. A. Mumpton. 1989. Zeolites in soils. 873-911. In Dixon, J. B. and S. B. Weed (eds.): Mineral in Soil Environments. 2nd ed., SSSA Book Ser. 1, Soil Sci. Soc. Amer. Neller, J-R. 1947. Mobility of phosphate in sandy soils. Soil Sci. Soc. Am. Proc. 11 :227-230. Parham, W. 1989. Some potential agricultural applications for developing countries. l\iatural Zeolites. 107-115. Natural Resources Forum. Butterworth & Co. Ltd., Stoneham, MA. Pond, W. G., and F. A. Mumpton. 1984. Zeo-Agriculture. Westview Press, Boulder, CO. Rhue, R. D., and J-B. Sartain. 1979. A survey of the fertility status of Florida soils as indicated by selected soil test results. Soil Crop Sci. Soc. Florida Proc.38:112-116. Shainberg, I., M. E. Sumner, W. P. Miller, M. P. W. Farina, M.A. Pavan, and M. V. Fey. 1989. Use of gypsum on soils: A review. Adv. Soil Sci. 9:1-111. Sheppard, R. A. 1971. Zeolites in sedimentary depo~its of the United States-A review. Chap. 22:279-310. In Flamgan, E.M. and L.B. Sand (eds.): Molecular sieve Zeolite-1. Adv. in Chem. Ser. 101, Amer. Chem. Soc. Sheppard, R. A., and A. J. Gude, 3d. 1982. Mineralogy, chemistry, gas adsorption, and NHrexchange capacity for selected zeolitic tuffs from the Western United States. U.S. Geo!. Survey. OpenFile Rept. 82-969. Smith, F. B., R. G. Leighty, R. E. Caldwell, V. W. Carlisle, L. G. Thompson, Jr., and T. C. Mathews. 1967. Principal soil areas of Florida. Fla. Agr. Exp. Stn. Bui. 717. Yuan, T. L., D.R. Hensel, R. D. Rhue, and W. K. Robertson. 1985. Nutrient status of a sandy H umaquept under long-term potato cultivation. Soil Crop Sci. Soc. Florida Proc. 44:93-97. Yuan, T. L., and D. E. Lucas. 1989. Laboratory test of potential amendment materials for phosphate retention in sandy soils. Soil Crop Sci. Soc. Florida Proc. 48: 127-131.


56 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Determination of Nitrate Levels in Water Samples Using a Microplate Reader C. D. Stanley* and J.B. Jones ABSTRACT The feasibility of using a spectrophotometric microplate reader system for the determination of N03 levels in water samples with the intent of reducing hazardous waste normally generated with common laboratory procedures was investigated. The brucine colorimetric assay was used. Variables that were studied included the effect of sample volume on results, the effect of reagent choice [chloroform (CHC13 ) or sulfuric acid (H2SO4 ) as solvent], and whether mixing samples in sample wells prior to analysis affected final results. Precise sample volume is critical for accurate results, since absorbance is measured vertically through the sample well. The effect of mixing method on results varied greatly with the solvent used. When using H2SO4 for mixing and reacting samples in the wells of the microplate, results were quite comparable to those for samples mixed and reacted in test tubes prior to absorbance determinations. When samples were mixed in-well, CHC13 as a solvent for brucine was determined to be unacceptable due to its volatile nature. A major ad vantage of mixing samples in wells is the substantial reduction in reagent waste generated as opposed to samples conventionally processed in test tubes. A maximum sample volume of 200L was determined optimum to allow adequate mixing of the samples. Microplate reader systems were initially developed for use in the serodiagnostic application for enzyme-linked immunosorbent assays (ELISA). ELISA involves using enzymes linked to antibodies to detect serologically, homologous antigenic determinants (2). ELISA also is useful for detection and quantification of bacterial and viral pathogens in plant tissue or in other environments (3,4,5). The system uti lizes a microplate (commonly measuring 10 cm x 15 cm) with 96 wells (each with 300 L capacity) into which samples are placed. Depending on the enzyme system used, a colorimetric analysis at an optimum wavelength is performed on each individual well as the microplate is scanned in a spectrophotometric reader. The analytical process is rapid, with individual absorbance determinations for all 96 wells performed in a matter of a few seconds. A study was conducted to evaluate the feasibility of using a microplate reader to perform routine determinations of N03 levels in water samples. The rationale behind this investigation was that the microplate reader system has the potential to generate sub stantially lower amounts of hazardous waste products compared to common laboratory procedures, and that it is fast with respect to analysis and data handling. Although the microplate system itself is costly and other instruments such as rapid continuous-flow automated analyzers designed for water analysis are currently available, it was determined that the effort would be of value to establish the reliability of the methodology with a microplate reader. Factors that were investigated included: 1) the effect of sample C. D. Stanley and J B. Jones, Gulf Coast Research and Education Center, Univ. of Florida, IF AS, 5007 60th Street East, Bradenton, FL 34203. Florida Agric. Exp. Stn. Journal Series no. R-02068. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 :56-58 ( 1992) volume on absorbance readings; 2) the effect of reagents used; and 3) the effect of method used for mixing and reacting the samples prior to absorbance determinations. MATERIALS AND METHODS The instrument used in this study was an SL T Labinstruments Easy Reader Model 400 AT with an absorbance wavelength range of 400-700 nm, precision of 0.5%, absorbance resolution of 10, and a measurement time of 5 secs for spectrophotometric absorbance determinations in 96 wells simultaneously and individually. The instrument utilizes a self-blanking dual-wavelength procedure. For N03 analysis, measurement and reference wavelengths were 405 nm and 690 nm, respectively. The microplates used were polystyrene with flat-bottom individual wells. The instrument measures absorbance vertically through each sample well. Consequently, we felt that determination of the effect of sample volume on measured absorbance levels was needed. This was accomplished by determination of absorbance on 100, 150, 200, 250, and 300 L of a prepared 5 mg L-1 N03 standard. The brucine method described by Barker (1) was chosen as the N03 determination procedure for this study because of simplicity and speed, requirement for few reagents, low interference potential, and a desirable N03 detection range (0.2 to 10 mg mL1). In addition, the method can use more than one solvent for the brucine. We chose to evaluate the feasi bility of using either chloroform or sulfuric acid as solvents in this study. The effects of preparation and mixing methods on final results were investigated using conventional methods when sample solutions were prepared in test tubes prior to transfer to the microplates. These re sults were compared to a method where all sample and reagent additions, reactions, and mixings occurred in the individual wells on the microplates. The mixing was accomplished using a Titertek mechanical shaker/mixer (Flow Industries, Inc.) specifically designed for microplates. High precision autopipettes (capable of I L increments) were used to add required reagents to samples, with the smallest amount added at any time being 18L. Standard N03 solu tions prepared from KN03 (0, 2, 5, and 10 mg L1 ) were used to evaluate the effectiveness of the microplate reader and the methods used. The effect of each variable was determined using 4 replications. The data were subjected to statistical analysis to compare the effects of preparation and mixing. RESULTS Figure I illustrates the importance of sample volume control on absorbance, showing the critical nature of consistently using a fixed sample volume for


PROCEEDINGS, VOLUME 51, 1992 57 1.2 ~-------------------~~ 8 0.8 r:: Y 0.00441 0.00369X R2 0.99 ti! -e 0.6 0 OJ .c <( 0.4 0.2 0C'O---~--~--~--~--~---~-~ 0 50 100 150 200 250 300 350 Sample Volume (microliters) Fig. I. Effect of sample volume (100, 150, 200, 250, and 300 L) on the absorbance (using a 5 mg L1 N03 standard) measured in individual wells (Each point within each volume level repre sents a replication at that level). analysis; care must be exercised to ensure precision in the delivery of such small volumes. The maximum sample volumes in wells was recommended to be 200 .L when solution mixing was required, since volumes greater than that could be spilled during the mixing process. The effectiveness of using CHC13 as the brucine solvent with respect to mixing method is shown in Figure 2. It is quite evident from the deviations of the plots that there were some substantial problems with mixing the sample and reagents in the microplate wells. The final results were not a surprise since, while the microplates were being prepared, the CHC13 tended to volatilize quite rapidly. This caused variability in the final sample volume in the wells. There also appeared to be some chemical reaction between the microplate and the CHC13 further compounding the problem. The conventionally-mixed treatment showed promising results but, because 1.2 ,---------------------~ 0.8 G> C IV -f 0.6 0 ., .a 1 C IV -f 0.8 0 ., .a <( 0.6 0.4 0.2 o~--~--o 2 __ y Premixed 4 6 8 10 12 Nitrate Concentration (mg/L) Fig. 3. The effect of mixing method on measured absorbance levels for different nitrate concentrations using sulfuric acid as the solvent for brucine (Each point within each nitrate concentration level represents an average of 4 replications at that level).


58 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA volume are added individually to each well, the margin for error is small and good laboratory technique is required. With good micro-volume technique, the same percentage effects of precision of measurement apply as with macro-volumes. It is likely that additional spectrophotometric analyses for other water quality determinations may be accomplished using this instrument, but it will require additional evaluation of each method to determine the practicality and feasibility of using a microplate reader system for such pruposes. REFERENCES 1. Barker, A. V. 1974. Nitrate determinations in soil, water and plants. Mass. Agric. Exp. Stn. Bull. No. 611. 2. Clark, M. F. and A. N. Adams. 1977. Characteristics of the microplate method of enzyme-linked immunosorbent assay for the detection of plant viruses. J. Gen. Virol. 34:475-483. 3. De Kam, M. 1982. Detection of soluble antigens of Erwinia salicis in leaves of Salix alba by enzyme-linked immunosorbent assay. Eur. J. For. Path. 12: 1-6. 4. Lommel, S. A., A. H. McCain, and T. J. Morris. 1982. Evalua tion of indirect enzyme-linked immunosorbent assay for detec tion of plant viruses. Phytopathology 72:1018-1022. 5. Van Vuurde,J. W. L. and N.J. M. Roozen. 1990. Comparison of immunofluoresence colony-staining in media, selective isola tion of pectate medium, ELISA and immunofluoresence cell staining for detection of Erwinia carotovora subp. atroseptica and E. chrysanthemi in cattle manure slurry. Neth. J. Pl. Path. 96:7589.


58 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA A Comparison of Water Quality Information Obtained From Depth-Integrated Versus Depth-Specific Groundwater Monitoring Devices W. D. Graham* and D. Downey ABSTRACT Methods for the design, construction and installation of an inexpensive device for the multi-level sampling of groundwater quality in shallow, sandy, cohesionless aquifers are described. The multi-level sampler takes samples from several small, discrete zones in the aquifer and therefore is capable of providing accurate vertical contaminant concentration profiles. Analysis of depth trends is an important aspect of groundwater quality monitoring, since contaminant concentrations usually vary markedly in the vertical direction. Water quality information ob tained from a network of multi-level samplers is compared to similar information obtained from conventional groundwater monitoring wells at an agricultural field site. Results confirm that water quality samples from the conventional monitoring wells provide only an averaged contaminant concentration for the aquifer over the depth of the well screen, while the multilevel samplers provide information on the distribution of contaminant with depth. In general, with the effort it takes to drill one hole, more information may be obtained by installing a multilevel sampler than a conventional monitoring well. In recent years, many new devices have become available for groundwater-quality monitoring. These devices provide various means of access to the groundwater for acquisition of samples prior to chemical analysis. The selection of an appropriate device will depend on 1) hydrogeological conditions at the monitoring site; 2) availability and terrain accessi bility of the drilling equipment; and 3) construction materials needed to prevent the adsorption of conta-W. D. Graham, Agricultural Engineering Dept., Univ. of Florida, Gainesville, FL 32611-0570; D Downey, Agricultural Engineering Dep., Univ. of Florida, Gainesville, FL 3261 I-0.S70. Florida Agric. Exp. Stn. Journal Series no. R-02079. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 :58-63 (1992) minants onto or the release of contaminants from the monitoring devices. Cherry et. al. (1983) and Litaor ( 1988) presented comprehensive reviews of alternative groundwater sampling and soil solution sampling devices, respec tively; as well as guidelines for the appropriate selection of each device. Neilson (1991) also reviewed recommended practices and procedures for conducting groundwater quality investigations. The purpose of this paper is to describe the construction and installation of an inexpensive device for depth specific, multi-level sampling of groundwater quality in sandy, cohesionless aquifers, and to compare water quality information obtained from depth-integrated versus depth-specific groundwater monitoring devices at an agricultural site in Florida. DEPTH-INTEGRATED VERSUS DEPTH SPECIFIC SAMPLING Early in the selection process for groundwaterquality monitoring, it is necessary to determine whether depth-integrated or depth-specific samples are desired. A depth-integrated sample is one that is obtained when water is pumped from a well that has a long screen, or from a well with an open borehole. As pumping occurs water can flow from various depth levels at various rates. A depth-specific sample, on the other hand, is obtained from an isolated depth interval in a well or borehole. Depth-integrated sampling can identify the presence of a contaminant in the groundwater system, but cannot determine the actual depth or in-situ concentration of the contaminated zone. The concentrations obtained from a depth-integrated sample are


PROCEEDINGS, VOLUME 51, 1992 59 normally dependent on the length of the screened interval, the depth of the pump intake and the rate or time-period of pumping. The depth-integrated approach may be appropriate in situations where one simply desires, with minimum drilling effort, to determine whether or not contamination exists at a particular monitoring site. It is often suitable for evalua tions of the quality of public water-supply wells. Depth trends are usually an important aspect of detailed groundwater quality investigations, since contaminant concentrations in bedded deposits can vary markedly in the vertical direction. In some situ ations, the entire zone of contamination may occupy only a small part of the total aquifer thickness. When depth-specific sampling is performed, the water sample is drawn from a narrow interval in the borehole in a manner that minimizes mixing of water from different depth zones. It is usually necessary to do depth-specific sampling at several depths for each sampling location, in order to determine overall trends in groundwater quality. APPROACHES FOR DEPTH-SPECIFIC SAMPLING There are three designs that can be used for permanent depth-specific sampling systems (See Fig. 1). These include (1) conventional multiple-borehole piezometer nests; (2) multiple-level, single borehole packer sampling; and (3) multiple-level, single borehole samplers. A multiple-borehole piezometer nest consists of two or more boreholes screened over different depths at approximately the same location. New methods have recently been developed primarily to diminish the drilling costs associated with the establishment of multiple-borehole monitoring networks. Multiple-level packer sampling refers to sampling from different levels in a long screened interval or in an open borehole. Each interval is temporarily isolated by means of inflatable packers and the sample is drawn from the zone between the packers. The packers are then deflated and the packer assembly is raised or lowered to a different level so that another sample can be taken. Between sampling times, however, the borehole provides an open pathway for movement of groundwater from one depth level to another. In this paper an inexpensive multi-level, single borehole sampler which prevents short-circuiting during periods of non-sampling is presented. DESIGN CONSIDERATIONS FOR THE MULTI LEVEL SAMPLER The multi-level sampling device described in this paper was developed by researchers at the University of Waterloo, Ontario, Canada (Cherry et al., 1983) and has subsequently been used for many research and commercial applications (Mackay et al., 1986; Garabedian, 1987; TVA, 1988; Hyman, 1990). A schematic of this device is shown in Fig. 2. In general, the device consists of an outer casing (usually PVC) with small port holes drilled into it. A bundle of smalldiameter collection tubes (either metal, polypropylene, or teflon) extend from the ground surface clown the inside of the casing. Each of the tubes is attached in turn to a sampling port which protrudes slightly from the casing pipe. Screening materials are used to cover each sampling port to prevent soil from entering the tubing. A summary of the criteria which should be considered in the design and construction of a multi-level sampler (MLS) is presented below. In addition to cost considerations, the chemical and physical properties of the soil, groundwater and contaminant(s) of interest must be carefully considered before an MLS is selected for any particular site. CONVENTIONAL PIEZOMETERS MULTIPLE PIEZOMETERS IN A SINGLE CASING PACKER SAMPLING IN A LONG WELL SCREEN ----------+ __. _. _. _. __. ----_STANDPI PE TER PIEZOME SCREEN VSLOTTED OR TIP CASING NARROW DIA. / / / / / / / / / / / / / / _,, / / / / / -PIEZOMETER 777777777777777 Fig. 1. Three approaches for depth-specific sampling. I I I I INFLATED PACKERS _. ;:::+-_. I ::14--I j I I : ~WELL I I SCREEN I I I I '---' 777777777777 777


60 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA CASING PIPE COLLECTION TUBING COUPLINGS Fig. 2. Schematic of the multi-level sampling device (from Cherry et al., 1983). Soil Collapsibilit_v Soil collapsibility is of primary concern when using the MLS. Where the soil is unable to collapse tightly around the MLS (i.e. in clayey deposits), groundwater may move upward or downward along the borehole zone, so that samples withdrawn at a particular depth may not be representative of what is actually occurring at that deptl, '.'1 the aquifer. Unconsolidated, cohesionless soils are optimal for accurate implementation of MLSs. Depth of Water Table Below Ground Surface The MLS device described in this paper requires a vacuum pump to draw groundwater through the subsurface sampling ports to the ground surface. Since the theoretical suction lift of a vacuum pump is approximately 10 m (1 atmosphere), this type of MLS cannot be used at sites where the water table is more than 8 to 9 m below the ground surface. Contaminant Considerations The materials selected for constructing the MLS will depend on the contaminants of interest. If or ganic constituents (such as pesticides) are being sampled the inner tubing, fittings and screening should be constructed of metal, glass or teflon. In addition, no glue, tape or caulk should be used for the construction procedure, since these substances may leach organic compounds into the subsequent sample. If inorganic constituents such as nitrate, phosphate or chlorides are being monitored, polyethylene tubing and fittings, and nylon screening material are appropriate. Glue, tape and/or caulk also may be used for construction, since compounds leached from these materials should not affect inorganic ion analyses. CONSTRUCTION METHODS FOR THE MULTI-LEVEL SAMPLER Fig. 3 shows a detailed schematic of the construction methods and materials required for a MLS designed to sample organic constituents (Hyman,1990). The sampling ports consist of brass elbows which are held to the PVC casing pipe by thin brass nuts. The outer threaded end of the elbows can be sanded, if necessary, to prevent the edges from damaging the STAINLESS STEEL HOSE CLAMP GLASS WOOL ALUMINUM SCREEN, 20 MESH STAINLESS STEEL SCREEN, 200 MESH STAINLESS STEEL HOSie CLAMP 4 76c,ALUMINUM TUBING '\_ BRASS ELBOW CONNECTOR Fig. 3. Sampling port detail for an MLS designed to sample organic constituents (from Hyman, 1990).


PROCEEDINGS, VOLUME 51, 1992 61 screening materials. The inner threaded ends are connected to aluminum tubing which extends to the surface inside the PVC casing pipe. The aluminum tubing is 60/61 grade, 0.4 76 cm diameter with a 0.89 mm wall, and is connected with brass unions where necessary. The PVC is threaded, Schedule 40, purchased in 3 m sections. Size 36 stainless steel hose clamps are used to fix the screening materials over the sampling port. All interior and exterior surfaces should first be rinsed with acetone and then distilled water to clean the metal of possible cutting oils. As each port is installed, its order along the sampler is labeled to keep track of its depth during construction. The materials required to construct a 6 m MLS with 10 sampling ports for organic-constituent sampling cost approximately U.S. $160 and a single device takes approximately 2 to 3 h to construct. Fig. 4 shows a detailed schematic of the construction methods and materials required for a MLS designed to sample inorganic constituents (Garabedian, 1987). The Schedule 40 PVC is purchased in 3 m sections, and connected where necessary with glued couplings and fittings. The 0.63 cm outside diameter polyethylene tubing is purchased in 30 m sections. Stainless steel wire is used to attach the nylon screening material (210-mesh nylon cloth or nylon stocking material) over the sampling port, and to attach the sampling port in turn to the PVC casing pipe. Holes are drilled into the PVC casing pipe just above the desired depth. The holes should be angled downward, and should be just large enough to accommodate the polyethylene tubing. Five cm of tubing should be left sticking out of the sampling port and then cut at an angle. A double thickness of nylon screening material is wrapped over the polyethylene tube and it is then bound in place with stainless steel STAINLESS STEEL WIRE WRAP NYLON SCREENING MATERIAL (DOUBLE THICKNESS) 0.635 cm POLYETHYLENE TUBING 5. l cm PV CASING PIPE STAINLESS STEEL WIRE WRAP Fig. 4. Sampling port detail for an MLS designed to sample inorganic constituents (from Garabedian, 1987). wire. The sampling port is in turn fastened to the PVC casing, also with stainless steel wire. As each port is installed, its order along the sampler is labeled to keep track of its depth. Color-coded tape, with a different color representing each depth, is recommended. The materials required to construct a 6 m MLS with 10 sampling ports for inorganic-constituent sampling cost approximately U.S. $40, and a single device takes approximately 1 to 2 h to construct. INSTALLATION METHODS FOR THE MULTI LEVEL SAMPLER Multi-level samplers (MLSs) which are to be set at depths greater than 3 m below the ground surface are most easily installed by means of a truck-or trailer-mounted drill rig equipped with continuousflight hollow-stem augers. At each sampling location, the string of augers is advanced to the desired depth and the MLS is lowered clown the hollow stem to the bottom of the hole. As the augers are gradually backed out of the hole, the aquifer sands should close in tightly around the MLS below the water table. This occurs because the sand has little cohesion, and because the pressure of the water in the aquifer is much higher than the atmospheric pressure in the empty augers. If the MLS is to be set at a depth of less than 3 m, it can be installed manually. Using this method, a 10.16 cm PVC casing is installed by hand-augering the borehole to approximately 2 m, then driving the casing to the desired depth using a hammer or other portable impact device. Once the desired depth has been attained, the MLS is lowered into the casing. The casing may then be removed using jacks or a tripod winch. Although this method is labor intensive, it can be used in areas inaccessible to drill rigs. Hand augering/hammering is preferable to jetting in the casing with water, because it introduces no foreign water into the hole. After the MLS has been installed and the aquifer material has collapsed so tightly around it that the MLS cannot be moved back and forth by hand, the open area around the MLS above the water table should be backfilled with soil brought up by the auger. (Note-it may take up to 24 hours for the aquifer material to collapse tightly around the MLS). Protective well covers with lockable lids may then be cemented into the ground over each sampler. SAMPLING METHODS FOR THE MULTI-LEVEL SAMPLER Samples may be obtained by applying suction to each sampling tube using a peristaltic vacuum pump. Initially, flow rates are often low because of clogging by fine particles. The flow rate can usually be increased by alternate episodes of back-flushing and pumping (except in the case where the sample point is situated in a layer of clay or silt, or where the port's screening material has been punctured during the installation process). After each slow-yielding sample-


62 Son. AND CROP SCIENCE SOCIETY OF FLORIDA point has been "developed" in this manner, the MLS should be allowed to set for one or two weeks so that the groundwater can redevelop equilibrium conditions around the MLS. After development, the entire ensemble of tubes in the MLS should be sampled simultaneously, if pos sible, so that a horizontal flow pattern is induced toward the MLS. Multi-head peristaltic pumps, such as those manufactured for laboratory or medical appli cations, are ideal for this purpose. Because of the small-diameter tubing, the MLS generally yields water at a few liters per minute or less. All standing water from the sampling tubes should be removed prior to sampling, so that the sample is free from the effects of chemical changes that may have occurred while the water resided in the tubing. Once the standing water has been removed, it is necessary to pump an additional volume sufficient to obtain a representative sampling of the formation water. It is desirable to apply generally similar pumping rates and duration on each sampling occa sion at a particular MIS If hydrogeologic conditions permit, consistency from MLS to MLS is also desirable. FIELD DEMONSTRATION An array of piezometers and multi-level samplers was installed at a mature citrus grove in Manatee County, Florida. In the eastern and central portion of Manatee County (where the field site is located), the shallow (surficial) aquifer consists mostly of medium to fine-grained, well-sorted, quartz sand. The aquifer ranges in thickness from about 3 m to 30 m, with sandy clavs of the Bone Valley and Hawthorne formations forming its base. Depth to the surficial aquifer ranges from zero meters in coastal and flat, poorly drained areas to about 3 m below the land surface in topographically high areas. Average depth to the water table at the field site is approximately 2 .5 m. Fig. 5 shows a site plan of the subsurface monitoring network installed at the field site. The purpose of the monitoring program was to 1) characterize local o Piezometer Multi-level Sampler o Rain Gage N G rove \\-.__ D Hydraulic /. 11 r SrJ ~/ -======'==': Pasture (non-grazed) Hydraulic o Gradient Grove \ \ \ \ \ \ i.; Fig. 5. Site Plan for the subsurface monitoring network. subsurface flow patterns and nitrate levels in the surficial aquifer, and 2) compare water quality information obtained from the fully screened piezometers versus the multi-level samplers. The piezometers were screened over the top 3 m of the surficial aquifer, while the MLSs were constructed with 5 ports spaced at 0.6 m intervals over the top 3 m of the aquifer. Figs. 6 and 7 show vertical nitrate profiles along transect A-A' for two typical sampling episodes conducted during the summer of 1991. These vertical profiles indicate that there is a zone of groundwater with nitrate concentrations above the EPA maximum contaminant level (MCL) of IO mg L-1 at the top of the surficia: aquifer. In general, however, nitrate le ,'els drop below 10 mg L within 2 meters of the top of the aquifer, and below 5 m L-1 within 3 m of the top of the aquifer. This indicates that, although agricultural activity appears to be impacting the surficial groundwater at this site, there are not likely to be adverse public health effects because of the limited extent of the impacted water. Fig. 8 compares the water quality information obtained from a neighboring piezometer and MLS (piezometer 2 and MLS 1) at the downstream encl of the field site for a series of sampling dates during the fall of 1991. The water quality information from the piezometer indicates surficial groundwater concentrations above the MCL for all three sampling dates. The information from the MLS, however, again shows that the nitrate levels are above the MCL only in the top few meters of the aquifer. These results confirm that the nitrate concentrations obtained from the piezometers do not reflect in-situ concentrations, but rather depth-integrated concentrations over the length of the piezometer screen. If detailed vertical profiles were not available at this site, the zone of contaminated groundwater could mistakenly be assumed to represent conditions over the entire aquifer depth. Thus, with the effort it takes to drill one hole, more information is obtained by installing a multilevel sampler than a conventional monitoring well. 50 Nitrate/Nitrite Concentrations 21 June 1991 100 '50 200 Distance Along Well Transect (m) 250 -7 I 300 Fig. 6. Vertical nitrate contours along transect A-A' on 21 June, 1991 (contour interval 2 mgL-1 dashed contours indicate nitrate concentration above 10 mgL-1 asterisks indicate sampling locations).


PROCEEDINGS, VOLUME 51, 1992 63 I 25.o Nitrate/Nitrite Concentrations 29 July 1991 Distance Along Well Transect (m) Fig. 7. Vertical nitrate contours along transect A-A' on 29 July, 1991 (contour interval 2 mgL-1, dashed contours indicate nitrate concentration above 10 mgL-1 asterisks indicate sampling locations). REFERENCES Cherry, J. A., R. W. Gillham, E.G. Anderson, and P. E. Johnson. 1983. Mitigation of contaminants in groundwater at a landfill: A case study, 2. Groundwater monitoring devices. J. Hydro!. 63:31-49. Garabedian, S. P. 1987. Large-scale dispersive transport in aquif ers: Field experiments and reactive transport theory. Ph. D. Dissertation, \fassachusetts Institute of Technology, Cambridge. \1A, 290 pp. Hyman, J. A. 1990. Grounclwater monitoring in three dimensions using the multi-level sampler. M.S. Thesis, Massachusetts Institute of Technology, Cambridge, MA, 121 pp. Litaor, M. I. 1988. Review of soil solution samplers. Water Resour. Res. 24(5):727-733. MacKay, D. M., D. L. Freyberg, and P. V. Roberts. 1986. A natural gradient experiment on solute transport in a sand aquifer. 1. Approach and overview of plume movement. Water Resour. Res. 22(13):2017-2029. Neilson, D. \1. (ed). 1991. Practical Handbook of Ground-Water Monitoring. Lewis Publishers, Chelsea, MI, 717 pp. Tennessee Valley Authority. 1988. Evaluation of tracer sampling devices for the macrodispersion experiment. Electric Power Research Institute Tech. Rep. No. EA-5816. Nitrate/Nitrite Concentrations Multilevel Sampler versus Piezometer E -Q) a, _j ro Q) Cl) C ro Q) Q) .0 <( C 0 > Q) w 16 15 14 13 12 0 / I I I I I II I I 27 August 1991 5 10 30 September 1991 28 October 1991 15 20 25 Concentration (mgC ) 30 Fig. 8. Comparison of water quality information obtained from piezometer 2 and multi-level sampler 1 (Vertical bars indi cate concentration of piezometer samples for date indicated).


64 SoJI. AND CROP SCIENCE Soc1nv OF FLORIDA Controlled-Release Fertilizer Use on Young 'Hamlin' Orange Trees T. A. Obreza* and R. E. Rouse ABSTRACT Groundwater quality concerns have caused emphasis to be placed on fertilization practices which in~~ease fertilizer. efficiency. The use of controlled-release N fert1hzer has potential to reduce leaching losses. The research objective was to evaluate controlled-release vs. conventional dry fertilizer in terms of citrus tree growth, frnit yield, and juice quality. Complete fertilizers containing all water-soluble N or water-soluble/controlled-release N blends were applied to newly-planted 'Hamlin' orange (Citrus sinensis L. Osbeck) trees on Carrizo citrange (C. sinensis x Poncirus trifoliata L. Raf.) rootstock for 3 yr at N rates of O, 13 25 50 and 100% of the Univ. of Florida, IFAS recommended ra:e. During the study period, water-soluble N was applied 15 times, while the controlled-release sources [IBDU, methylene urea (MU), Osmocote+MU, and IBDU briquets+MU] were applied from 5 to 8 times. The critical point for fruit yield est!mated by a linear plateau model occurred at the 50% rate. At this rate, no differences in trunk cross-sectional area, canopy volume, leaf N and K concentrations, yield, or juice quality occurred among fertilizer sources. The cost of using controlled-release ma terials was higher than for conventional, water-soluble sources at recommended application frequencies, due to higher material cost. Controlled-release fertilizers have a role in the management of young citrns trees if high-frequency application of water-soluble fertilizer is not feasible. New citrus tree plantings in Florida have been substantial since the freezes of the early 1980s. Over 48,000 ha of new groves have been established since 1988 (Florida Dept. of Agr. and Consumer Serv., 1990). Land used for citrus production in southwest Florida (Charlotte, Collier, Glades, Hendry, and Lee counties) has increased from 21,000 ha in 1980 to 51,000 ha in 1990. Expansion is expected to continue through the l 990s, much of which will occur in the relatively new citrus-growing region of southwest Florida. One concern associated with the expansion of citrus is the effect of fertilization practices on groundwater quality. Nitrogen fertilizer efficiency, defined as the percentage of applied nutrient taken up by the crop, is often low because of the mobility of N fer tilizer in sandy Florida soils. Nitrogen fertilizer effi ciency can be increased through the use of controlledrelease N sources, which potentially reduce N leaching and improve efficiency of plant recovery (Khalaf and Koo, 1983). Previous studies have shown that controlled-release N fertilizer can be applied to citrus trees with lower frequency than water-soluble N, ~ith no ~ecrease in tree growth. A number of experiments w1~h newly-planted citrus trees have compared three to six annual applications of water-soluble N v~. one_ to three applications of controlled-release _N [mcl1:1dmg uramite, sulfur-coated urea (SCU), 1sobutyhdene T. A. Obreza and R. E. Rouse, Southwest Florida Research and Education Center, L1 niv. of Florida, IF AS, P. 0. Drawer 5127, Immokalee, FL 33934. Florida Agric. Exp. Stn. Journal Series no. R-02084. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 :64-68 (1992) diurea (IBDU), and "coated" N]. At the end of 1 and 2 yr, no differences in trunk cross-sectional area or canopy volume were observed when the sou~:-ces were applied at the same rate (Rasmussen and Smith, 1961; Jackson and Davies, 1984; Marler et al., 1987; Yuda et al., 1987; Ferguson et al., 1988; Zekri and Koo, 1991.). Koo (1986) applied different N fertilizer sources to bearing citrus trees and observed increased fruit production with IBDU and SCU over a water soluble N source. This study was designed to evaluate the performance of N-P-K fertilizers containing a portion of the N or K in controlled-release form under southwest Florida flatwoods citrus conditions. The objective was to compare tree growth, initial fruit yield, and juice quality between trees fertilized frequently with a water-soluble N and K source and infrequently with controlled-release N and K sources. MATERIALS AND METHODS The experiment was conducted in a large newlydeveloped commercial citrus grove in southwest Florida. The site was similar to most new groves in the region in terms of soil types and previous land use. Land which had been in unimproved pasture for more than 25 yr was disked, laser-leveled, and formed into two-row beds during the summer and autumn of 1988. 'Hamlin' orange (Citrus sinensis L. Osbeck) trees on Carrizo citrange (C. sinensis x Pon cirus trifoliata L. Raf.) rootstock were planted in March 1989 at a spacing of 2.7 m (in-row) x 7.6 m (between-row). The irrigation method was seepage, where water was supplied to the trees through upward flux from a water table maintained approximately 0.60 to 0. 7 5 m below the soil surface. The experiment was conducted within four adjacent tree rows (two 2-row beds). The soils were classified as Pople fine sand (loamy, siliceous, hyperthermic Arenic Ochraqualf) and Holopaw sand (loamy, siliceous, hyperthermic Grossarenic Ochraqualf). The area was delineated into eight blocks, each four rows wide by 16 trees long. Each block contained 16 plots, consisting of four adjacent trees within a row. In each block, 16 fertilizer treatments were assigned to plots at random (Table 1). Three complete N-P-K fertilizers (treatments 1-12) were evaluated over the 3-yr duration of the experiment. Two additional materials (treatments 1315) were evaluated for the first year of planting only. The conventional material (CONV) contained onlv water-soluble N, while the isobutylidene diurea (IBDU) and methvlene urea (MUI) sources contained a portion of the N in controlled-release form. The Osmocote (OSM) and Woodace briquets (BRIQ) contained essentially all controlled-release N, P, and K. A second methylene urea source (MU2) was applied over the OSM and BRIQ treatments in years 2 and 3 of the experiment (Table 1).


PROCEEDINGS, VOLUME 51, 1992 65 Table I. N-P-K fertilizer treatments applied to young 'Hamlin' orange trees, 1989-1991. Fertilizer source, Relative No. of Complete fertilizer annual 3-yr total analvsis, and Treatment fertilizer rate t applications application rate application vear~ appliedt :s;o. (1989, 1990, 1991) (1989, 1990, 1991) (1989, I 990, 1991) I\ p K % kg tree1 ------------kg tree1 -------------CONY, 8-1.8-6.6 1 13, 13, 13 6,5,4 0.34, 0.68, 1.02 0.16 0.04 0.14 (1989-91) 2 25, 25. 25 6,5,4 0.68, 1.36, 2.04 0.33 0.07 0.27 '.l 50, 50, 50 6,5,1 1.,16. 2.72, 4.09 0.65 0.15 0.54 4 100,100,100 6,5,4 2.72, 5.45, 8.17 UI 0.29 l.08 IBDU, 8-1.8-6.6 5 13, 13, 13 3,3,2 0.34, 0.68, 1.02 0.16 0.04 0.14 ( 1989-91) 6 25, 25, 25 3,3,2 0.68, 1.36, 2.04 0.33 0.07 0.27 7 50, 50, 50 3,3,2 1.36, 2.72, 4.09 0.65 0.15 0.54 8 100, 100, 100 3,3,2 2. 72, 5.45, 8.17 1.31 0.29 1.08 MU I, 9-1 .8-6.6 9 13, 13, 13 2,2,2 0.,\0, 0.61, 0.91 0.16 0.04 0.12 (1989-91) 10 25, 25, 25 2,2,2 O.fil, 1.21, 1.82 0.33 0.08 0.24 11 50, 50, 50 2,2,2 1.21, 2.42, 3.63 0.65 0.16 (l.48 12 100, 100, 100 2,2.2 2.42, 4.85, 7.27 l.31 0.32 0.96 OSM, 17-2.6-7.5 13 13, 25, 25 1, 2, 2 0.16, 1.36, 2.04 0.30 0.06 0.24 (1989) 14 25, 50, 50 1, 2, 2 0.32, 2.72, 4.09 0.60 0.13 0.47 + MU2 8-1.8-6.6 ( 1990-91) BRIQ, 14-1.3-2.5 15 13, 100, 100 I, 2. 2 0.20, 5.45, 8.17 l. I 2 0.25 0.90 (1989) + MU2, 8-1.8-6.6 (1990-91) Control 16 0, 0, 0 0.00, (J.00, 0.00 0.00 0.00 0.00 tcoNV=water-soluble N source; IBDU=isobutylidene diurea; MUI and MU2=methylene urea; OSM=Osmocote; BRIQ=Woodace bri~uets Percentage of maximum N rate applied each year. The maximum rate of fertilizer applied each year (Treatments 4, 8, and 12) supplied 0.22, 0.44, and 0.65 kg tree1 of N during 1989, 1990, and 1991, re spectively. These rates are about 10% higher than the current average University of Florida, IF AS N recommendation for Florida citrus trees in each of the first three years after planting (Koo et al., 1984). Additional fertilizer treatments equaled 50, 25, or 13% of the maximum rate each year. For the balance of this paper, the fertilizer rates will be referred to in terms of the percentage of the maximum rate applied each year. In addition to the N-P-K content listed in Table 1, the conventional (CONV) fertilizer source contained all water-soluble N, 3.0% Mg, 0.3% Mn, and 0.05% Fe. The IBDU source contained half of its N in water-soluble form and half as controlled-release IBDU, with 2.4% Mg, 0.3% Mn, and 0.06% Fe. The MUI source contained 60% water-soluble N and 40% in controlled-release forms including methylene urea, sulfur-coated urea, and activated sludge, with 3.0% Mg, 0.5% Mn, and 0.03% Fe; the MUI source also contained half of its K as controlled-release sulfur-coated potassium sulfate (K2SO4). The OSM source contained all controlled-release N-P-K, with 1.5% Ca, 1.0% Mg, 4% S, 0.10% Mn, 0.40% Fe, 0.05% Cu, 0.05% Zn, (l.02% B, and 0.004% Mo. The BRIQ source contained primarily controlled-release N (from IBDU), P, and K, with 2.3% Ca, 1.2% Mg, 0.17% Mn, 1.12% Fe, 0.05% Cu, and 0.08% Zn. The MU2 source contained 40% water-soluble N and 60% in controlled-release forms including methylene urea and activated sludge, with 2.4% Mg, 0.30% Mn, and 0.02% Fe; the Kin the MU2 source was 100% water soluble. Following tree planting, soil samples were col lected from each block and initial trunk diameter measurements were made. The two middle trees in each 4-tree plot were measured 15 cm above the bud union in both the north-south and east-west direc tions. The cross-sectional area of the trunk was calculated from the mean radius. Initial fertilizer applications were made in April 1989. For the OSM and BRIQ treatments, an 8-cm deep, 1-m long trench was dug on two sides of the tree approximately 0.45 m from the trunk, the materials were applied evenly, and the trench was back-fil led. Fourteen briquets per tree were required to equal the rate in Table 1. The remaining treatments were applied by hand under the trees in a 1-m circle at first and then in a wider circle matching the outward expansion of the tree canopy as the trees grew. The CONV fertilizer was applied in April,June, July, September, October, and December of 1989; March, May, July, September, and November of 1990; and March, May, July, and November of 1991. The IBDU


66 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA fertilizer was applied in April, July, and November of 1989; March, June, and September of 1990; and March and August of 1991. The MU I fertilizer was applied in April and September of 1989; March and August of 1990; and March and August of 1991. The MU2 fertilizer was applied in March and August of both 1990 and 1991. Trunk cross-sectional area was measured at 7, 12, and 24 months after planting. Tree canopy width (measured in both the north-south and east-west di rections) and height were measured at 12 and 24 months after planting. Since the measurements indicated that the height was greater than the width, canopy volume was calculated based on the formula for a prolate spheroid: (4/3)('TT)(tree height/2)(mean canopy radius)2 Leaf tissue samples were taken from each plot in mid-March of 1990 and 1991. This time of year was chosen over the traditional August/September sampling period because it represented the end of one fer tilization year and the beginning of the next. Leaves were from the 5 to 6-month-old autumn growth, from non-fruiting terminals. They were ground to pass a 1-mm screen, and 1-g subsamples were ashed in a muffle furnace at 500 C. The residue was dissolved in HCl, and the resulting solution was analyzed for K, Ca, and Mg using atomic emisson/absorption spectrophotometry (Baker and Suhr, 1982) and for P using the ascorbic acid method (Council on Soil Testing and Plant Analysis, 1980). A 0.4-g subsample of the ground tissue was also digested in concentrated H2SO4 as part of a micro-Kjeldahl procedure. The resultant solution was diluted with water and analyzed for N using the indophenol blue method (Hanlon and DeVore, 1989). Soil samples were taken to a depth of 15 cm at the dripline of the trees in 1989 and 1990. Samples were taken at depth increments of Oto 15 cm and 15 to 30 cm in 1991. The soil was air-dried and passed through a I-mm screen. A 5-g subsample was extracted with 20 ml of Mehlich I solution (0.05 M HCl + 0.0125 M H2S04 ) and P, K, Ca, and Mg were measured as described above. The soil was also analyzed for pH (2: 1 water-to-soil ratio) and for Walkley-Black organic matter content (Hanlon and DeVore, 1989). Fruit yield determinations were made in October 1991 (31 months after planting). Fruit on the middle two trees in each plot were harvested and weighed. A random subsample from each plot was taken for fruit size estimation and juice quality analysis. The juice was extracted using a commercial juice extractor (FMC Corp., Agricultural Machinery Div.,Jonesboro, AR). Juice quality parameters measured included percentage juice fruit, 1 sugars as total soluble solids (TSS, in Brix), total titratable acid content as citric acid, and TSS/acid ratio. TSS per tree was calculated from yield, percent juice, and TSS in the juice. Analysis of variance and Duncan's multiple range test were used to determine fertilizer-source effects on tree growth and yield variables, and leaf nutrient concentrations, at each rate. A1, econc::1ic analysis of the different fertilizer programs was maue using the Table 2. Selected soil characteristics at the experimental site (data represent the means of eight samples taken at planting in I 989). Mehlich I-extractable Organic Depth Texture pH matter p K Ca Mg cm % ------------mg kg-1 ------------0-15 Sand 7.0 1.03 8 7 1160 20 15-30 Sand 5.3 1.42 5 5 570 14 following costs: For the formulated complete fer tilizer products, the cost of CONV was U.S. $ 0.13 kg, IBDU was U.S.$ 0.30 kg-I, MUI was U.S.$ 0.50 kg-I, MU2 was U.S.$ 0.34 kg-1, and OSM was U.S.$ 2.20 kg1 (Muraro and Holcomb, 1989). Fertilizer ap plication was assumed to be mechanical, at a cost of U.S.$ 0.05 tree1 RESULTS AND DISCUSSION The soil in which the trees were planted had not been previously fertilized and was very low in extract able P and Kat planting (Table 2). The organic matter content was higher in the 15-30 cm soil depth than in the top 15 cm due to burial of the original pasture surface layer under subsurface material which had been used to form the beds. The pH of the bed surface soil was considerably higher than that of the original surface due to carbonates in the subsurface material (Table 2). However, roots of the citrus trees were exposed to both soils during planting, since the planting holes were approximately 30 cm deep. The effect of fertilizer sources on tree growth at 12 months after planting is shown in Table 3. None of the tree-size parameters differed among sources at the 13, 50, or 100% rates. A significantly larger canopy volume was observed for the OSM source as compared to all other sources at the 25% fertilizer rate. This rate of OSM was equal to the amount recommended by the manufacturer for the first year of newly-planted trees. The canopy volume observed was the largest of any source-rate combination at 12 months. At the 100% rate, leaf N concentration was greater with IBDU than with MUI (Table 3). Leaf N was high across all treatments in 1990. Citrus nursery trees in Florida have been shown to have elevated levels of nutrients stored in their tissues, attributed to intensive fertilization practices in the nursery (Castle and Rouse, 1990). Tree-size measurements made at 7 months after planting indicated no growth differences between fertilized treatments and the unfertilized control (data not shown). The levels of leaf N in the 12-month-old trees may have been due to a carryover effect from high levels of N in the trees at planting. The effect of fertilizer sources on tree growth at 24 months after planting is shown in Table 4. No differences in measured growth parameters occurred among sources at any rate. Maximum tree size (trunk


PROCEEDINGS, VOLUME 51, 1992 67 Table 3. Effect of fertilizer sources on the growth response of 12-month-old 'Hamlin' orange trees. Fertilizer source CONV IBDU MUI OSM BRIQ CONV IBDU MUI OSM BRIQ CONV IBDU MUI OSM BRIQ Percentage of maximum fertilizer rate applied during yr 1 ( 1989) 13 25 50 100 Trunk cross-sectional area ( cm2 ) 6.49 8.02 at 8.44 7.20 6.32 6.68b 7.81 7.13 6.88 7.07 ab 7.20 6.48 6.95 8.01 a 6.10 Canopy volume (m') 0.91 l.16 b 1.42 1.20 1.00 l.17 b l.13 l.16 I.IO 1.03 b 1.30 1.09 l.14 1.51 a 0.82 Leaf N concentration(%) 3.06 3.08 3.13 3.36 ab 3.14 3.12 3.31 3.52a 3.13 3.11 3.07 3.26b 2.91 2.89 3.15 tMeans within columns followed by different letters are signif icantly different (P<0.05) according to Duncan's multiple range test. Means within columns containing no letters are not signifi cantly different from each other. Table 4. Effect of fertilizer sources on the growth response of 24-month-old 'Hamlin' orange trees. Fertilizer source CONV IBDU MUI OSM+MU2 BRJQ+MU2 CONV IBDU MUI OSM+MU2 BRIQ+MU2 CONV IBDU MUI OSM+MU2 BRIQ+MU2 Percentage of maximum fertilizer rate applied during yr 2 (1990) 13 25 50 100 Trunk cross-sectional area ( cm2 ) 31.57t 36.23 39.08 34.30 31.97 32.15 36.87 34.16 3l.l3 34.72 35.77 33.01 32.16 36.14 29.33 Canopy volume (m3 ) 7.99 8.53 10.42 9.04 8.06 8.15 9.07 8.74 7.47 8.51 10.00 8.81 8.13 9.33 8.15 Leaf N concentration(%) 2.52 2.59 2.81 2.70 2.67 2.70 2.82 3.04 2.65 2.93 3.07 2.93 2.46 2.84 2.91 tMeans within each column were not significantly different from each other (P<0.05) according to Duncan's multiple range test. cross-sectional area and canopy volume) was attained at the 50% rate for all sources. This rate was just above the "critical point" as estimated by a linear plateau model fitted to the tree growth-fertilizer rate data. The model represents a fertilizer response curve with two lines, one for the increasing portion and the other for the plateau portion of the curve. The point at which the lines intersect is the "critical point," where response to fertilizer ceases. Leaf N concentration across all treatments was about 0.4% lower than in 1990. All N levels were still in the optimum or high ranges for leaf tissue standards (Koo et al., 1984) regardless of fertilizer source or rate. Leaf K concentration data from 1990 and 1991 indicated that there was no difference in K uptake between the source containing 50% controlledrelease K (MU I) and the 100% water-soluble K sources (IBDU and MU2) (Table 5). The MUI and MU2 sources were each applied twice in 1990, but there was no increase in leaf K content where the controlled-release K source was used. The added expense of including controlled-release K in the MUI fertilizer blend was not justified under these experimental conditions. The effect of fertilizer sources on fruit yield and juice quality at 31 months after planting is shown in Table 6. Significantly lower fruit yields were seen where the first year fertilizer rate was 13% of maximum (BRIQ + MU2 and low rate of the OSM + MU2 treatments). Increasing the fertilizer rate subsequently to the 25% or 100% rates in yr 2 and 3 was not sufficient to produce a yield comparable to treatments which received these higher fer tilizer rates for 3 yr. However, where OSM was applied at the 25% rate in yr I and the rate then increased to the 50% level with MU2 in yr 2 and 3 (treatment 14), the yield was similar to that of treatments receiving the 50% rate for all 3 yr. Although maximum yield occurred at the I 00% rate, the 50% rate was estimated as the "critical point" by a linear plateau model fitted to the yield-fertilizer rate data. Table 5. Effect of fertilizer sources on the leaf K concentration of young 'Hamlin' orange trees. Fertilizer K Annual source sourcet applications 1989 IBDU ws 3 MUI CR 2 1990 IBDU ws 3 MU2 ws 2 MUI CR 2 Percentage of maximum fertilizer rate applied in yrs 1 and 2 13 25 50 100 LeafK concentration(%), 1990 l.93" 2.06 2.69 2.80 l.63 1.84 2.10 2.55 Leaf K concentration(%), 1991 1.57 2.38 2.78 2.80 2.20 2.65 2.87 1.47 1.63 2.46 2.44 tws= 100% water-soluble K source; CR=50% controlled-re lease K source "Means within each column were not significantly different from each other (P<0.05) according to Duncan's multiple range test.


68 Son, AND CROP SCIENCE SOCIETY OF FLORIDA Table 6. Effect of fertilizer sources on the yield and juice quality of 31-month-old 'Hamlin' orange trees. Fertilizer source CONV IBDU MUI OSM+MU2 BRIQ+MU2 CONV IBDU MUI OSM+MU2 BRIQ+MU2 Percentage of maximum fertilizer rate applied during yr 3 ( 1991) source 13 5.37 6.26 6.74 17.42 17.66 17.00 25 50 Fruit yield ( kg tree1 ) 11.0labt 14.36 a 12.66 ab 8.44 b 21.63 19.05 23.06 20.93 juice TSStlarid ratio 16.47 15.30 15.74 14.49 17.72 14.82 16.90 14.80 100 22.43 ab 22.17 ab 26.94 a 16.24 b 13.93 13.90 13.81 14.37 'Means within columns followed by different letters are signif icantly different (P<0.05) according to Duncan's multiple range test. Means within columns containing no letters are not signifi cantly different from each other. +TSS = Total soluble solids. Juice quality expressed as TSS/acid ratio did not differ with respect to fertilizer source at any rate. The ratio decreased as fertilizer rate increased, due to decreasing TSS. The juice acid level did not signficantly differ among fertilizer rates (data not shown). The TSS/acid ratio plays a large part in juice palatability: within certain limits, increasing ratio increases the palatability of the juice. Thus, juice quality was lower at higher fertilizer rates. However, TSS tree-1 is the most important factor which determines economic returns for processed fruit. Higher fertilizer rates produced lower TSS in juice, but TSS tree' was higher due to a higher yield of fruit tree'. The estimated cost of fertilization using different sources for the first 3 yr of 'Hamlin' orange tree growth at the 50% fertilizer rate is shown in Table 7. The application cost of the controlled-release sources is lower due to decreased application frequency. However, the higher cost of these materials is not nearly offset by the lower cost of application. The cost used in this calculation was based on a mechanical fertilizer application. If hand labor had been used, the cost difference between water-soluble and controlled-release sources would narrow. Nevertheless, the cost of using controlled-release sources for fertili zation of young citrus trees remains considerably higher than the use of a conventional source at recoirnnended application frequencies. Controlled-release fertilizers applied between 5 and 8 times over 3 yr performed as well as water-soluble fertilizer applied 15 times over the same period. N-P-K fertilizer applied at the 50% rate was sufficient to produce over 20 kg of fruit on a 2.5-yr-old tree, which is indicative of vigorous growth. Although increased cost is associated with ..: use of controlledrelease sources, they have a role in young-tree situa tions (resets or solid-set new plantings) if high :;e-Table 7. Estimated cost of fertilization for the first three years of 'Hamlin' orange tree growth at the 50% fertilizer rate Cost component Material Application Total Complete fertilizer product CONV IBDU MUI OSM+MU2 --------------------------U.S. $ tree'--------------------------1.06 0.75 1.81 2.45 0.40 2.85 3.63 0.30 3.93 3.02 0.25 3.27 quency application of conventional fertilizers is not feasible. Groundwater quality concerns also favor the use of controlled-release fertilizer due to increased fertilizer efficiency. REFERENCES Baker, D. E., and N. H. Suhr. 1982. Atomic absorption and flame emission spectrophotometry. In A. L. Page (ed.). Methods of soil analysis, part 2. Am. Soc. Agron., Inc., Madison, WI. Calvert, D. V. 1969. Effects of rate and frequency of fertilizer applications on growth, yield and quality factors of young 'Va lencia' orange trees. Proc. Florida State Hort. Soc. 82:1-7. Castle, W. S., and R. E. Rouse. 1990. Total mineral nutrient content of Florida citrus nursery plants. Proc. Florida State Hort. Soc. I 03:42-44. Council on Soil Testing and Plant Analysis. 1980. Handbook on reference methods for soil testing, revised ed. The Council on Soil Testing and Plant Analysis, Athens, GA. Ferguson,J.J., F. S. Davies, C.H. Matthews, and R. M Davis. 1988. Controlled-release fertilizers and growth of young 'Hamlin' orange trees. Proc. Florida State Hort. Soc. 101:17-20. Florida Dept. of Agriculture and Consumer Services. 1990. Com mercial citrus inventory. Florida Agr. Stat. Serv., Orlando, FL. Hanlon, E. A., andJ. M. DeVore. 1989. IFAS extension soil testing laboratory chemical procedures and training manual. Florida Coop. Ext. Serv. Cir. 812. Jackson, L. K., and F. S. Davies. 1984. Mulches and slow-release fertilizers in a citrus young tree care program. Proc. Florida State Hort. Soc. 97:37-39. Khalaf, H. A., and R. C. J. Koo. 1983. The use of controlled release nitrogen on container grown citrus seedlings. Citrus Veg. \lag. 46(9): 10,32. Koo, R. C. J. 1986. Controlled-release sources of nitrogen for bearing citrus. Proc. Florida State Hort. Soc. 99:46-48. Koo, R. C. J., C. A. Anderson, I. Stewart, D. P.H. Tucker, D. V. Calvert, and H. K. Wutscher. 1984. Recommended fertilizers and nutritional sprays for citrus. Florida Agr. Exp. Stn. Bui. 536D. Marler, T. E., J. J. Ferguson, and F. S. Davies. 1987. Growth of young 'Hamlin' orange trees using standard and controlled-re lease fertilizers. Proc. Florida State Hort. Soc. I 00:61-64. Muraro, R. P., and F.. D. Holcomb, Jr. 1989. Budgeting costs and returns for southwest Florida citrus production, 1988-89. Florida Agr. Exp. Stn. Econ. Info. Rpt. 261. Rasmussen, G. K., and P. F. Smith. 1961. Evaluation of fertilizer practices for young citrus trees. Proc. Florida State Hort. Soc. 74:90-95. Yuda, E., T. Nakayama, Y. Moriguchi, and S. Nakagawa. 1987. Effect of controlled release fertilizer on the growth of rnung satsuma mandarin trees. J. Plant Nutr. 10(9-16):1471-1478. Zekri, M., and R. C. J. Koo. 1991. Evaluation of controlled-release fertilizers for young citrus trees. J. Am. Soc. Hort. Soc. l 16(6):987-990.


PROCEEDINGS, VOLUME 51, 1992 69 CROPS SECTION Lateral Root Distribution Patterns in Stylosanthes guianensis Seedlings John B. Brolmann* and Peter J. Stoffella ABSTRACT Seedlings of eight genotypes of stylo [Stylosanthes guianensis (Aubl.) Sw.] (P numbers 1414, 7035, 7102, 7630, 8288, 8400, 8681, and 8682) were grown in pot culture in the greenhouse for 39 d. Shoot weights, stem heights, root weights, lateral root intensity (LRI) (number of lateral roots/cm-' of taproot), and patterns of l~teral root distribution were measured. Genotype 7102 had sigmficantly lo'"'.er root weight, shoot weight, stem height, and ~hoot:~oot ~at10 tha_n most of the other genotypes. Lateral root ~nt~ns1ty did not differ among genotypes. Chi-square analyses md1cated that genotypes 7102, 1414, and 8682 had an uniform lateral root distribution ( 1: 1: 1 : 1: 1 ratio for number of lateral roots originating at the 1-3, 3-5, 5-7, 7-9, and 9-11 cm length of taproot). Lateral root distribution patterns for genotypes 7035, 7630'. 8288, 8400, and 8681 was not uniform, and may be conform~ng_to ran?om or ?ther unequal non-random patterns. These data md1cate differential genotypic responses for lateral root distribution patterns in young stylo seedlings. ~ost Stylosanthes guianensis (stylo) plantings are est~bhshed by 'surface-sowing' (Gardener, 1978). Estab hshme_nt of stylo ~s influ_e?ced by seed viability, edaph1c and chmatIC cond1t10ns, cultural and management practices, and seedling vigor. Miles (1986) reported that diameter of the largest root below 7 cm ~as thicker for direct seeded than transplanted seedlm!5s or_ transplanted rooted-stem cuttings of S. guzanensis Perennial Stylosanthes spp. developed deeper roots and less branched root systems than annuals when each had similar size n;ots (Gardener, 1978). Pitman et al. ( 1986) reported stand establishment and plant persistence differences among and wi_thin StJlosanthes spp. Cook ( 1980) reported that mICroenvironmental factors affecting seed moisture co_ntei:it and soil moisture duration are related to germmat1on and emergence of tropical pasture species. Root morphological characteristics under field conditions have been described for black beans (Phaseolus vulgaris L.) (Stoffella et al., 1979), cowpea [Vigna unguiculata (L.) Walp.] (Kahn and Stoffella, 1987), tomato (Lycopersicon esculentum Mill.) (Stoffella, 1 ~83), and in laboratory conditions for bell peppers (Capsicum annuum L.) (Stoffella et al., 1988). Minimal information is available on stylo seedling root morphology and developmental characteristics subsequent to _ra?icle pi:otrusion: These root traits may have a role m 1mprov111g seedl111g emergence, growth, and development, particularly under stress microenvironments. The purpose of this investigation was to John B. Brolmann and Peter J Stoffella, Agric. Res. and Educ. Center, P. 0. Box 248, Fort Pierce, FL 34954-0248. Florida Agric. Exp. Stn. Journal Series no. R-01001. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 5 l :69-7 I ( 1992) describe seedling lateral root distribution patterns of several Stylosanthes guianensis genotypes. MATERIALS AND METHODS Seeds of ~ight Styl~sanOies guianensis genotypes were heat sca'.Ified a~ 70 C ior 3 h. to break dormancy and pregermmated 111 9 cm petri dishes filled with a I cm layer of silica sand and 19 ml of water for 4 cl. Single pregerminated seeds ( 1-2 mm radicle length) of each genotype were placed in 1.5 L pots containing screened (1 mm sieve) virgin Oldsmar fine sand (san;1y, siliceous, hyperthermic Alfie Haplaquods) pH -. ~-:J (2.1 kg/pot1). A 20-8.7-16.6 N-P-K liquid fer t1hzer (278 mg L') was applied 29 cl after seeding at 200 ml per pot'. Water was applied as needed throughout the experiment. Pots were placed in a greenhouse with mean day/night temperatures of 26 and l 9C, respectively, and mean relative humidity of 57%. S_eedlings were removed from pots 39 cl after seedmg by gently washing the sand from pots and adhering roots. Taproots were separated from stems a~d ~tern heights (base of hypocotyl to uppermost tnfoha~e leaf) recorded. Individual root systems were placed 111 polyethylene trays and submerged in water. Lateral roots were counted for five 2 cm sections of ~ap'.oot length under a binocular scope. Counts were m1_t1~ted_ at I cm below the top of each taproot. Roots ong111at111~ from the first cm of taproot were not c~mn~ed smce they could not be morphologically dis t111gmshed as lateral, basal, or adventitious roots (Zobel, 1986). Roots and shoots were oven dried for 3 days at 65C, weighed, and shoot:root ratio calculated for each plant. Lateral root intensity (LRI) was calculated by: total number of lateral roots in 10 cm taproot length divided by 10. The experimental design was a randomized complete block with genotypes replicated eight times, except for genotype 7102 which had six replications. Analyses of variance (ANOV A) was performed on all measured and calculated variables (SAS Institute, Inc., 1988). Genotype means were separated by a least significant difference (LSD) test, at the 0.05 probability level. Chi-square analyses were used to determine if lateral roots were uniformly distributed on the taproot by testing a 1: 1: 1: l: l ratio of number of lateral roots initiating from the top 1-3, 3-5, 5-7, 7-9, and 9-11 cm sections of taproot length. RESULTS AND DISCUSSION Shoot and root weights differed among genotypes (Table 1). Although genotype 7102 had the smallest


70 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table 1. Seedling shoot and root characteristics of Stylosanthes guianensis genotypes. Shoot GenoLype weight Root weighL -----------mg---------7035 11.08 1.69 7630 17.15 3.04 8288 15.86 2.60 8400 17.45 2.45 7102 5.23 1.33 1414 17.23 2.89 8682 15.76 2.70 8681 12.75 1.95 LSD (P = 0.05) 3.74 0.93 Shoot:root Stem ratio height cm 6.84 3.75 6.02 6.31 6.73 4.09 8.00 4.93 4.03 3.03 6.27 5.16 6.02 4.69 7.26 4.35 1.59 0.92 Lateral root intensity (LRI)1 3.48 4.35 3.93 3.65 4.20 3.93 3.98 3.71 NS:j: 1 LRI = Total number of lateral roots for the first 1 to 11 cm of taproot length divided bv 10. *NS = not significant. shoot weight and one of the smallest root weights and stem heights as compared with the other genotypes, shoot:root ratio (4.03) was also smaller for 7102 than for the other genotypes. This suggests that for genotype 7102, a greater portion of photosynthates were being utilized for root growth than shoots as compared with the other genotypes. Genotype 7102 belongs to the fine stem stylo (S. guianensis var. inter media) and is characterized as early flowering, cold tolerant, and lower yielding than other genotypes. Lateral root intensity (LRI), an estimate of lateral root productivity, did 'not differ among genotypes (Table 1). Stafford and McMichael (1990) reported seedling LRI differences among guar [Cyamopsis tet ragonoloba (L) Taub.] genotypes and various seed treatments. Chi-square analyses were used to test uniformity of lateral roots on the taproot (Table 2). Goodness of fit tests for the ratio 1: 1: 1 : 1: 1 indicated that lateral roots were randomly distributed along the first five 2-cm sections of taproot length of genotypes 7102, 1414, and 8682. Heterogeneity chi-squares were nonsignificant for the three genotypes, 7102, 1414, and 8682, suggesting that an adequate fit occurred. De viation chi-squares were significant for 7035, 7630, 8288, 8400, and 8681 genotypes, indicating a lack of fit to a 1: 1: 1: 1: 1 ratio. In these genotypes an unknown non-random distribution pattern of lateral roots may be occurring. These data indicate a differential genotypic distribution of lateral root in young stylo seedlings. O'Toole and Bland (1987) cited many investigations measuring root morphological differences among genotypes of monocotyledon and dicotyledon crop species. Genotypic differences in lateral root distribu tion may be important for nutrient and water absorb tion, stand establishment, seedling vigor, growth, and development, particularly in stressful environments. Zobel (1989) has cited many investigations indicating the effects of environmental factors on root growth, development, and distribution. Further research is required to determine the genetic, environmental, and physiological factors governing or influencing lateral root distribution. ACKNOWLEDGMENTS Mr. George Pfaffs technical assistance 1s gratefully acknowledged. REFERENCES Cook, S. J. 1980. Establishing pasture species in existing swards: a review. Trop. Grassl. 14:181-187. Gardener, C. J. 1978. Seedling growth characleristics of Stylosanthes. Aust. J. Agric. Res. 29:803-813. Kahn, B. A. and P. J. Stoffella. 1987. Rool morphological charac teristics of field-grown cowpeas. J. Am. Soc. Hort. Sci. 112:402406. Miles, J.W. 1986. Effect of establishment method on field performance of Stylosanthes guianensis lines. Aust. J. Exp. Agric. 26:325329. O'Toole, J. C. and W. L. Bland. 1987. Genotypic variation in crop plant root systems p. 91-145. In "I. C. Brady (ed.). Adv. Agron. Vol. 41. Academic Press, Inc., New York. N.Y. Pitman, W. D., .J. B. Brolmann, and A. E. Kretschmer, Jr. I 986. Persistence of selected Stylosanthes accessions in peninsular Florida, U.S.A. Trop. Grassl. 20:49-52. SAS Institute, Inc. 1988. SAS/STAT User's Guide. Release 6.03 ed. SAS Institute, Inc., Cary, N.C. Stafford, R. E. and B. L. Mc Michael. 1990. Primary root and lateral root development in guar seedlings. Env. Exp. Bot. 30:27-34. Stoffella, P. J 1983. Root morphological characteristics of fieldgrown tomdtocs. HortScience 18:70-72. Swffella, P.J., M. Lipucci Di Paola, A. Pardossi, and F. Tognoni. 1988. Root morphology and development of bell peppers. HortScience 23:1074-1077. Table 2. Goodness-of-fit tests for 1:1:1:1:1 ratio of lateral root distribution in Stylosanthes guianensis seedlings. Taproot sections t (cm) Deviation Heterogeneity Genotype 1-3 3 --::J 5-7 7-9 9-11 x Pvalue x Pvalue --------------------Lateral root no. --------------------7035 8.1 8.1 7.4 4.6 5.8 11.4 0.02-0.05 27.1 0.30-0.50 7630 11.0 10.0 9.6 6.5 6.4 16.6 <0.01 10.7 0.99 8288 9.9 9.6 8.0 6.6 5.1 16.5 <0.01 22.9 0.70-0.80 8400 9.5 6.6 8.6 6.5 5.3 13.0 0.01-0.02 18.6 0.90-0.95 7102 8.8 9.2 8.5 7.7 8.2 1.0 0.80-0.90 8.5 0.99 1414 9.3 9.0 8.1 6.4 6.5 7.4 0.10-0.20 14.7 0.98-0.99 8682 9.9 7.6 7.9 7.9 6.5 6.0 0.20-0.30 17.9 0.90-0.95 8681 10.4 7.6 6.3 6.6 6.3 13.1 0.01-0.02 16.3 0.95-0.98 tTaproot sections in 2 cm increments starting I cm below top of root.


PROCEEDINGS, VOLUME 51, 1992 71 Stoffella, P. J., R. F. Sandsted., R. W. Zobel, and W. L. Hymes. 1979. Root characteristics of black beans. II. morphological differences among genotypes. Crop Sci. 19:826-830. Zobel, R. W. 1986. Rhizogenetics (root genetics) of vegetable crops. HortScince 21 :956-959. Zobel, R. W. 1989. Steady-state control and investigation of root system morphologv p. 165-182. In J. G. Torrey and L. J. Winship (eds.) Application of continuous and steady-state methods to root biology. Kluwer Academic Publishers, Dordrecht, The Netherlands.


PROCEEDINGS, VOLUME 51, 1992 71 Evaluation of Macroptilium atropurpureum (DC) Urb. Germplasm for Reaction to Foliar Diseases Ronald M. Sonoda*, A. E. Kretschmer, Jr. and T. C. Wilson ABSTRACT Macroptilium atropurpureum (DC) Urb. accessions were evaluated for incidence of foliar fungal diseases in two replicated field experiments; one planted in 1982 containing 84 M. atropurpureum accessions and one planted in 1986 containing 177 M. atropurpureum accessions. Moderate to high incidences of rust, Uromyces appendiculatus (Pers.:Pers.) Unger var. crassitunicatus J. Irwin, were consistently observed during the cooler months of the year. Few or no pustules were observed on 22 M. atropurpureum accessions. Two of these accessions, IRFL 4655 and 4449, produced more (P= 0.05) forage than cultivar 'Siratro'. Foliar blight, Rhizoctonia solani Kuhn, was scattered throughout both plantings during the warmer, wetter months. Plants of three accessions in the 1982 planting were unaffected by foliar blight. However, at least one plant of all accessions in the 1986 planting was affected by foliar blight. There were no significant differences between incidence of foliar blight on Siratro and on other accessions. Accessions unaffected by powdery mildew, Oidium sp., were recorded in the 1982 planting when a moderate incidence of the disease was observed. Siratro was unaffected by powdery mildew. Seventy-nine of the 177 accessions of M. atropurpureum in the 1986 planting had significantly lower (P = 0.05) gray leaf spot, Phaeoisariopsis griseola (Sacc.) Ferraris, incidence than Siratro, during a severe epidemic. One accession unaffected by rust and gray leaf spot outyielded Siratro (P= 0.05). Macroptilium atropurpureum is a viny legume with potential as a forage plant for tropical pastures. The cultivar 'Siratro' has been grown to a limited extent in south and central Florida pastures since its introduction in the mid-1960's (4). Siratro was selected in Australia in the early 1960's (2) in a breeding program using plants collected from a relatively small geographical area. Since then, collections of M. at ropurpureum have been made from a wider geographical area (5, 6). The primary limitation of Siratro is its lack of persistence under grazing which is primarily due to its poor recovery after being heavily grazed (3). Another limitation of Siratro is its susceptibility to several dis eases (10). The most common and important diseases of Siratro in Florida are two foliar fungal diseases: R. M. Sonoda, A. E. Kretschmer, Jr. and T.C. Wilson, Univ. of Florida Agric. Res. and Educ. Center, P. 0. Box 248, Fort Pierce, FL 34954-0248. Florida Agric. Exp. Sta. Journal Series no. N00583. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51:71-74 (1992) foliar blight caused by R. solani (13) and rust (14), caused by U. appendiculatus var. crassitunicatus. Siratro is also often affected by grey leaf spot (15), caused by P. griseola, and powdery mildew (11) caused by Oidium sp. Beginning in the late l 970's, a program to identify M.atropurpureum genotypes with better agronomic qualities was initiated at the University of Florida, IF AS, AREC, Ft. Pierce. In the course of these studies, two replicated plantings consisting of a total of 261 M. atropurpureum accessions were made. Some of the results from these plantings have been published (7, 8, 14, 15). The following summarizes evalu ations of accessions in these plantings for incidence of the four foliar fungal diseases. MATERIALS AND METHODS Macroptilium atropurpureum accessions were obtained from various sources including the germ plasm bank at CIAT (Centro Internacional de Agricultura Tropical, Cali, Colombia), CSIRO, Cunningham (Brisbane, Australia) and Davies (Townsville, Au stralia) laboratories, and collections made by the junior author and his colleagues (7, 8). Two replicated plantings involving 84 and 177 genotypes, respectively, were established in 1982 and 1986, respec tively (8). The same area at the University of Florida, IF AS, AREC Ft. Pierce was used for the two plantmgs. Scarified and germinated seed dusted with 'cowpea type' inoculant were planted in 1: I horticultural vermiculite: sphagnum peat moss, in cells of styrofoam planting trays, in the greenhouse. Individual plants of accessions intended for the 1986 planting were planted in adjacent cells in planting trays in June 1986. The accessions intended for the 1982 planting were arranged in a randomized block design with 5 single-plant replicates in the styrofoam planting trays. These plants were inoculated with uredospores of U. appendiculatus var. crassitunicatus at the 3 to 5 true-leaf stage in May 1982, following a procedure described previously (14). These plants were transplanted to the field in July 1982. The 1986 plots were transplanted in July 1986. The planting arrangement in each case was single plant plots spaced


72 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA 2 m x 2 m apart, in a completely randomized block design with 4 replicates in the 1982 planting and 5 replicates in the 1986 planting. Shoots were periodically harvested by hand-harvesting to about 6 cm in height and about 10 to 15 cm in diameter. Shoots were dried at 70 to 75 C, and weight of dry matter (DM) determined for each replicate. Harvest dates for the 1982 planting were 30 Sep 1982, 5 Jan, 5 May, 5 Jul, 26 Aug, and 5 Dec 1983. The harvest dates for the 1986 planting were 2 Oct 1986, 8 Dec 1986, 1 Jun 1987, and 16 Jul 1987. Herbicides were applied to interplant areas immediately after harvests to control weed growth. Weeds were also controlled by mowing. Rust incidence in the field for the 1982 planting was assessed on 18 April 1983. Rust incidence was assessed three times in the 1986 planting, on 4 Nov 1986, 2 Mar 1988, and 11 April 1988. The rating scale used was: 1 = no lesions; 2 = a few small pus tules on a few mature leaves; 3 = light incidence of small to large pustules on scattered mature leaves; 4 = moderate incidence of lesions on most mature leaves; and 5 = heavy incidence of pustules on most mature leaves with occasional defoliation. Powdery mildew incidence was assessed in the 1982 planting on 18 April 1983 using the following rating scale: l = no disease to 12 = 100% of mature leaves with surface covered by mycelium. No assessment of powdery mildew incidence was made in the 1986 planting, as only a few plants were affected. Rhizoctonia solani foliar blight was assessed 13 Jul 1983 in the 1982 planting and on 4 Nov 1986 in the 1986 planting using the following rating scale: l = no lesions; 2 = lesions on fewer than 10% of the leaves of a plant; 3 = lesions on 11 to 30% of leaves of a plant; 4 = lesions on 31 to 80% of leaves of a plant and 5 = lesions on 81 to 100% of leaves of a plant; plants nearly or completely defoliated. Incidence of P. griseola grey leafspot in the 1986 planting was rated on 4 Nov l 986 using the following scale: l = no lesions 2 = a few lesions on a few mature leaves; 3 = light incidence of lesions on scattered mature leaves; 4 = moderate incidence of le sions on most mature leaves, some defoliation; 5 = heavy incidence of lesions on most mature leaves, with heavy defoliation occurring. No assessment of ~rey leafspot incidence was made in the 1982 plantmg. All disease incidence ratings were made by the senior author. Disease incidence ratings were made soon before a harvest, thus foliage coverage was usually at a maximum. Data was analyzed with SAS (12). Disease incidence data for each disease and each rating date were analyzed separately. Total yield over the 6 harvests for the 1982 planting was analyzed. Only the 16 Jul 1987 harvest data for the 1986 planting was analyzed. RESULTS Six of the 84 M. atropurpureurn accessions in the 1982 planting, IRFL 2937, 2938, 3531, 3440, 5755, and 5756 were unaffected when inoculated with U. appendiculatus var. crassitunicatus uredospores in the greenhouse. The same six accessions were unaffected by rust on 18 April 1983 after transplanting to the field and exposure to natural infection (Table 1). Rust symptoms were present each year in the 1982 planting. Although IRFL 5817 was a "rust-resistant" accession obtained from Australia, three of four plants inoculated in the greenhouse were infected. However, none of the plants of IRFL 5817 were infected (Table l) when rated in the field in April 1983. Siratro produced more DM (P = 0.05) than all but IRFL 3440 and 5817 among accessions unaffected by rust in the 1982 planting (Table 1). Two accessions affected by rust in the 1982 planting, IRFL 3146 and 1055, produced more DM than Siratro (P=0.05) (data not presented). Rust symptoms were present each year in the 1986 planting. There were 14 accessions unaffected or only lightly affected by rust, mean rust incidence ratings less than 2.0, in the 1986 planting (Table 2). Of these, only two accessions, IRFL 4655 and 4449 outyielded Sii'-atro (P=0.05) (Table 2). The other 12 accessions, unaffected or lightly affected by rust, yielded as much as Siratro. Rust incidence ratings made at different times (4 Nov 1986, 2 Mar 1988, and 11 April 1988) were consistent (Table 2). Powdery mildew symptoms were observed in the spring of 1983 on many accessions in the 1982 planting. All plants of 36 accessions, including Siratro, were free of powdery mildew symptoms on 18 April 1983. Mean powdery mildew incidence rating on the most severely affected accession, IRFL 5766, was 7. Some accessions were uniformly affected by powdery mildew. Within some accessions, however, some plants were lightly to moderately affected, powdery Table I. Reaction to rust, Uromyces appendiculatus var. crassitunicatus (UA), powdery mildew (PM), Oidium sp. and foliar blight, Rhizoctonia solani (RS) among rust resistant accessions of Macroptilium atropurpureum in 1982-84 test. Severity index t Accession" Yield lJA PM RS g plant' Siratro 1274 2.6 1.0 1.8 5817 1172 1.0 1.5 1.0 3440 1038 1.0 3.0 1.0 937 835 1.0 1.5 2.0 2938 503 1.0 2.3 2.5 5755 202 1.0 1.5 3.0 3531 515 1.0 1.0 1.0 5756 143 1.0 1.0 3.0 LSD(P=0.05) 389 1.2 NS 1.4 tseverity index for lJA, assessed 18 Apr. 1983 and RS, assessed 13 Jul. 1983: I = no lesions to 5 = plants with heavy incidence of disease and heavy defoliation. Severity Index for PM, assessed 18 April 1983: 1 = no disease to 12 = plant 100% covered with mildew symptoms. 'Accession University of Florida, IFAS, AREC Fort Pierce, germplasm collection accession number. Total dry matter yield, shoots harvested 30 Sep 1982, 5 Jan 1983, 5 May 1983, 5 Jul 1983, 26 Aug 1983 and 5 Dec 1983.


PROCEEDINGS, VOLUME 51, 1992 73 Table 2. Reaction to rust, Uromyces appendiculatus var. crassitunicatus, (UA), at different rating dates; grey leafspot, Phaeoisariopsis griseola (PG); and foliar blight, Rhizoctonia solani (RS) among the highest yielding Macroptilium atropurpureum resistant to rust and/or P. griseola in 1986-88 test. Severity indext UA PG RS Acc.t Yield November 1986 March 1988 April 1988 g plant' Rust and grey leafspot resistant 4655 292 1.0 1.2 1.0 1.0 2.4 4606 201 1.0 1.0 1.3 1.0 2.0 4414 189 1.0 1.2 1.3 1.0 1.8 4406 185 1.3 1.2 1.0 1.0 2.2 4605 156 1.0 1.4 1.6 1.0 2.2 4391 154 1.0 1.0 1.0 1.6 2.6 4399 147 1.0 1.2 1.3 1.2 2.0 4397 146 1.0 1.0 1.3 1.2 2.2 4408 125 1.3 1.2 1.3 1.0 2.0 4398 120 1.5 1. 7 1.5 1.8 1.8 4419 61 1.0 1.7 1.0 1.0 2.2 Rust resistant only 4449 296 1.0 1.0 1.0 2.6 1.6 4504 186 1.7 1. 7 1.7 2.6 1.4 4206 140 1.0 1.4 1.4 2.8 1.8 4668 108 1.0 1.0 1.2 2.2 2.6 4171 98 1.0 1.0 1.0 2.6 1.6 4626 85 1.5 1.9 1.5 2.8 1.6 Grey leafspot resistant only 4410 324 1.0 3.0 2.8 1.6 2.0 4680 222 2.5 2.4 2.5 1.0 1.8 4577 195 4.4 4.1 4.4 1.0 2.0 4268 168 2.6 2.2 2.6 1.0 2.2 4637 134 4.0 3.1 4.0 1.0 2.0 Siratro 114 4.0 3.3 4.0 2.8 1.4 LSD(0.05) 152 1.0 1.2 1.0 1.1 1.0 tseverity index for UA, assessed 4 Nov 1986, 2 Mar 1987 and 11 April 1987; RS, assessed 4 Nov 1986; and PG, assessed 4 l'liov 1986: = no lesions to 5 = plants with heavy incidence of disease and heavy defoliation. 'Accession University of Florida, IFAS, AREC Fort Pierce, germplasm collection accession number. Dry matter yield for 16 Jul. 1987 harvest. mildew incidence ratings of 2-5, while other plants of the accession were free of the disease. No significant difference in incidence of powdery mildew was found among accessions. No rating for powdery mildew incidence was made in the 1986 planting as only a few scattered plants were affected. Rhizoctonia solani foliar blight occurred each summer in the 1982 and 1986 plantings. There were significant (P = 0.05) differences in mean rating for foliar blight incidence among accessions in the 1982 plantings. Plants of accessions IRFL 3440, 3531 and 5817 in the 1982 planting were unaffected by foliar blight (Table 1 ). Although there were significant (P = 0.05) differences between accessions in ratings in the 1986 planting, at least one plant of each accession was affected by foliar blight. None of the accessions in either planting had mean foliar blight incidence less than that of Siratro. Differences in P. griseola grey leaf spot incidence, means ratings ranging from 1 to 4, were obvious during the severe P. griseola infestation soon after the 1986 planting was established. Seventy-nine of the 177 accessions had mean grey leaf spot incidence rat ings lower than that recorded on Siratro (P = 0.05 ). Only IRFL 4174, with a rating of 4.0, had a mean grey leafspot rating greater than that on Siratro, 1.8 (P = 0.05). Many accessions, especially IRFL 4174, suffered heavy leaf loss. One accession unaffected by rust and grey leafspot, IRFL 4655, outyielded Siratro (P = 0.05) (Table 2). DISCUSSION Consistency of the rust incidence ratings throughout the course of the 1986 field experiment (Table 2)


74 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA indicate that field screening at the AREC-Ft. Pierce research plots is an effective means for preliminary selection for rust-resistant M. atropurpureum. Field screening was also a satisfactory means of preliminary screening for reaction of M. atropurpureum to R. solani foliar blight. The ability to detect resistance to grey leafspot and powdery mildew under field conditions was not as consistent, as detectable levels of powdery mildew occurred only in the spring of 1983 in the 1982 planting and gray leaf spot only in the fall of 1986 in 1986 planting. When diseases were present, however, it proved easy to obtain preliminary disease susceptibility information for these accessions planted in the field for agronomic characteristic evaluation. Separate greenhouse screenings would require additional space as well as increased costs in terms of ma terial and time. Some of the M. atropurpureum accessions in the two plantings produced dense foliage even in the single plant, well-weeded, plots maintained for agronomic characterization. Other accessions, however, did not produce a dense cover. The foliar diseases affecting Siratro are usually more severe when plant canopies are dense (R. M. Sonoda, personal observa tion). Improvements in screening for disease incidence and severity, as well as comparisons of disease susceptibility among accessions, using agronomic field plots may be possible if individual plots consist of more closely planted plants of each accession and harvests for forage yield determination are scheduled to allow diseases to develop to their fullest. Since R. solani foliar blight can drastically reduce shoot yield of Siratro growing in dense canopies during warm, wet periods (13), comparisons under dense canopy conditions are warranted between Siratro and those accessions on which low incidences of foliar blight were recorded in the 1982 and 1986 plantings. The variable results for rust development on ac cession IRFL 5817, susceptible upon inoculation in the greenhouse, but resistant when planted in the field in 1982, were found to be due to a mixture of susceptible and resistant plants from the same seed lot. The mixed results with accession IRFL 4410 (Table 5), resistant in one reading in the field and susceptible in two others may have been due to a rating error or to the presence of more than one race of the pathogen. Recently the presence of different races of U. appendiculatus var. crassitunicatus in Mexico was determined (1). Two U. appendiculatus var. cras situnicatus isolates from Florida have behaved simi larly in a differential M. atropurpureum genotype screening test (1 and R. M. Sonoda, unpublished data), but further study is needed to determine if more than one race of the pathogen occurs in Florida. The results of these field evaluations indicate that there are high yielding M. atropurpureum lines resistant to rust and to P. griseola. One of these, accession IRFL 4655 is currently being prepared for release as a germplasm source (9). REFERENCES l. Bray, R. A., R. M. Sonoda, and A. E. Kretschmer, Jr. 1991. Pathotype variability of rust caused by Uromyces appendiculatus on Macroptilium atropurpureum. Plant Dis. 75:430. 2. Hutton, E. M. 1962. Siratro-A tropical legume bred from Phaseolus atropurpureus. Aust. J. Exp. Agr. Anim. Husb.2:117-125. 3. Jones, R. M., and G. A. Bunch. 1988. The effect of stocking rate on the population dynamics of Siratro in Siratro (Macrop tilium atropurpureum) -Setaria (Setaria spacelata) pastures in southeast Queensland. I. Survival of plants and stolons. Aust. J. Agric. Res. 39:209-219. 4. Kretschmer, A. E., Jr. 1972. Siratro (Phaseolus atropurpureus L.) a summer-growing perennial pasture legume for central and south Florida. Florida Agric. Exp. Stn. Circ. S-214. 5. Kretschmer, A. E., Jr. 1988. Accession list and passport data of tropical legumes collected or introduced into the germ plasm bank of the University of Florida's IF AS Agricultural Research and Education Center, P.O. Box 248, Ft. Pierce, Florida 34954 (alphabetical list). Ft. Pierce AREC Res. Rep. FTP-88-1. 6. Kretschmer, A. E., Jr., R. Reid, J. Gonzales R., and G. H. Snyder. 1987. Tropical forage legume collection trip in southern Mexico. Soil Crop Sci. Soc. Florida Proc. 46:80-83. 7. Kretschmer, A. E., Jr., R. M. Sonoda, R. C. Bullock, G. H. Snyder, T. C. Wilson, R. Reid, and J. B. Brolmann. 1985. Diversity in Macroptilium atropurpureum (DC) Urb. p. 155-157. In T. Okubo and M. Shiyomi (eds.) Proc. XV Int. Grassl. Congr. Kyoto, Japan. 24-31 Aug. 1985. The Science Council of Japan and The Japanese Society of Grassland Science c/o The National Grassland Research Institute, Tochigi-ken, Japan. 8. Kretschmer, A. E.,Jr, R. M. Sonoda, and T. C. Wilson. 1989. Yield, flowering pattern and leaf area diversity in Macrop tilium atropurpureum (DC.) Urb. p. 259-260. in R. Desroches (ed.) Proc. XVI Int. Grssld Congr. Nice, France. 4-11 Oct. 1989. Association Francaise pour la Production Fourragere, Versailles, France. 9. Kretschmer, A. E.,Jr., R. M. Sonoda, and T. C. Wilson. 1991. Characteristics of a high yielding, disease resistant Macrop tilium atropurpureum (DC) Urb. Soil Crop Sci. Soc. Florida Proc. 50:25-27. 10. Kretschmer, A. E., Jr., R. M. Sonoda, R. C. Bullock, and T. C. Wilson. 1992. Registration of IRFL 4655 Macroptilium at ropurpureum germplasm. Crop Sci. 32: (In Press). 1 l. Lenne, J. M., and R. M. Sonoda. 1985. Diseases of Macrop tilium atmpurpureum -a review. Trop. Grassl. 19:28-34. 12. SAS Institute Inc. 1985. SAS Users Guide; Statistics. Cary, NC. 13. Sonoda, R. M. 1980. Reduction of forage yield of Siratro by Rhizoctonia foliar blight. Plant Dis. 64:667. 14. Sonoda, R. M. and A. E. Kretschmer, Jr. 1982. Sources of resistance in Macroptilium atropurpureum to rust. Soil Crop Sci. Soc. Fla. Proc. 41:73-74. 15. Sonoda, R. M. and A. E. Kretschmer, Jr. 1989. Reaction of Macroptilium atropurpureum (DC.) Urb. accessions to angular leaf spot caused by Isariopsis griseola Sacc. p. 699700. in R. Desroches (ed.) Proc. XVI Int. Grssld Congr. Nice, France. 4-11 October 1989. Association Francaise pour la Production Fourragere, Versailles, France.


PROCEEDINGS, VOLUME 51, 1992 75 Evaluation of a Photovoltaic System for Supplying Water to Beef Cattle J. J. Mullahey*, Y. J. Tsai, and D. J. Pitts ABSTRACT In Southwest Florida, cattle grazing rangeland and pastures may be remote from electric power sources making pumping of water difficult. A study was conducted in 1990-91 at the Southwest Florida Research and Education Center near lmmokalee, Florida to determine seasonal pumpage and system efficiency of a photovoltaic powered livestock watering system, and to evaluate this system in meeting the daily water requirements of beef cattle in Florida. The prototype system consisted of one 0.92 m photovoltaic array (108 photovoltaic cells), a 12-volt DC selfpriming diaphragm pump, a 1134 L stock water tank, and a total flowmeter. Total cost for the solar-powered watering system was $2850. To evaluate the system, water was drawn from a pond (2.5 m head) and piped to the stock tank. System performance (L d1 ) was described using a linear regression model that related daily pumpage to solar energy. Slope (L w- m2 ) of the regression line for winter (0.028) was significantly (P<0.10) higher than the fall (0.021) and summer (0.017) and similar to the spring (0.026) season. System efficiency was assessed using the concept of energy work and was reported using 3 ranges (5-10, 10-15, and 15-20%). System efficiency across all seasons was 13.1%, however, in the winter and spring seasons, a gre.,,;er frequency of the 16-20% system efficiency range occurred compared to the fall and summer seasons. The photovoltaic system studied could reliably supply water (5670 L) for 100 beef cows grazing native range in Florida with a yearly energy savings of 172 Kwh m yr' ($14.85). Florida is one of the major beef cattle producing states in the eastern L1 nited States, with cattle produced from a forage-based system consisting of rangeland and introduced pasture. Range is an important natural resource in Florida encompassing nearly 4 million ha, and is vital for water filtration (water quality), aquifer recharge, wildlife, recreation, and grazing livestock. Ranchers utilize introduced pasture for cattle grazing during the spring and summer after which cattle are rotated to rangeland for winter grazing. Cattle grazing rangeland may be remote from electric power sources, making pumping of water difficult. Windmills, historically used to pump water for livestock, are not dependable under all weather conditions. Using solar energy for pumping water is an alternative to power derived from fos sil fuels and windmills. One of the first recorded attempts at pumping water using solar energy was made by French engineer Salomon de Caux in the 1600's. He developed a solar heat engine as a power source (Garg, 1987). The first solid-state photovoltaic device was made in 1876 (Wolf, 1981). In recent years photovoltaic-powered water pumps have been put in operation in many parts of the world. One of the largest water pumps is in Nebraska (Garg, 1987). Photovoltaicpowered solar energy systems have demonstrated J. J. Mullahey, Y. J. Tsai, and D. J. Pitts, Southwest Florida Research and Education Center, Univ. of Florida, P. 0. Drawer 5127, Immokalee, FL 33934 Florida Agric. Exp. Sta.Journal Series no. R-02081. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51:75-79 (1992) higher reliability at lower cost than alternative methods of solar energy conversion (Helikson, 1989). Cost (installation, maintenance, and repair) comparisons showed a solar powered pumping system to be 60% of the cost of a windmill and 35% of a pumpjack run by a gasoline engine from a 9 m shallow well in Nevada (Balliette and Garrett, 1990). Increases in photovoltaic cell efficiency coupled with decreases in the cost of solar powered systems makes photovoltaic systems more comparable in cost to that of power derived from fossil fuels (Helikson et al., 1990). The main advantage of using direct solar energy as a power source for watering livestock is that it is an energy resource that is usually available both when and where it is needed. It is important voltage and amperage output from the photovoltaic array matches the connected motor's load demand over a spectrum of motor speeds (Helikson, 1989). Energy losses in photovoltaic-powered pumping systems occur because the pumping system load curve does not match the photovoltaic array optimum operation point line for various irradiance levels (Hori et al., 1985). Though it is impossible to completely match these curves, it is possible to determine the load impedence curve which maximizes the overall photovoltaic pumping system efficiency (Helikson, 1989). Designing photovoltaic-powered water systems can be approached using an energy storage device (backup battery) or by powering the pump directly from the photovoltaic system (Hsiao and Blevins, 1984). In Florida, information is lacking on system performance and efficiency of a photovoltaic-powered water system for supplying water to livestock. The objectives of this research were: 1) to determine seasonal pumpage and system efficiency of a photovoltaic-powered watering system; and 2) to evaluate this system in meeting the daily water requirements of beef cattle in Florida. MATERIALS AND METHODS A photovoltaic-powered watering system became operational in Aug. 1990 at the Southwest Florida Research and Education Center near Immokalee (26 27'N, 81 26'W). This solar-powered system consisted of the photovoltaic array [0.92 m2 ; 3 modules, 36 cells per module; provided 4.8 kW m2 d1 rated at 960 peak watts (Simpler Solar Systems Inc., Tallahassee, FL)], a 12-volt DC self-priming diaphragm pump, a 1134 L galvanized stock tank, and a flowmeter. Total cost (cost of drilling and casing a well was not included) for the solar-powered watering system was $2850 (photovoltaic array= $2000; pump, stock tank, plumbing, flowmeter, etc. $850). The photovol taic array was attached to a support structure which was manually adjusted monthly via an adjustable tilt angle mechanism (1.3 m above soil surface) to maximize interception of solar energy by the panel.


76 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA To evaluate this system, water was drawn from a pond (2.5 m head) and piped to the stock tank. Water overflow from the tank was prevented by means of a discharge or overflow pipe set 5 cm from the stock tank top which returned water to the pond. Daily total flow (liters) and total daily solar energy (W m2 ) were recorded from Aug. 1990 to Aug. 1991. Solar energy and pumpage data were used to calculate system performance, system efficiency, and an estimate of energy savings. To quantify system performance, a linear regression analysis was used: Y= a+ bX x = solar energy (W m2 d-1 ) Y = pumpage (L d-1 ) a= intercept (L d-1 ) b = regression slope (L W-1 m2 ) Data for all seasons were analyzed using the General Linear Models procedure (SAS, 1982) to determine differences among slopes of the regression lines. A significant (P<0.01) interaction (season x solar energy) resulted in use of the interaction coefficients to estimate the regression equations. Slopes of the estimated regression equations for all seasons were compared to the winter season with differences declared significant at P = 0.10. N 400 I E 3 >.. O') ,.........._ L Cl) Q) -0 C C W 0 Cl) L :::::, 0 0 0 (f) I-300 200 ...... ..... By definition of energy work, the system effi ciency was calculated as: efficiency(%) = (energy out/ energy in) *100 For the photovoltaic-powered watering system, energy out was equal to the work done by the pump to lift water of 2.5 m head. Energy in was the total solar energy available to the panel. System performance was analyzed by season [spring (Mar.-May), summer Qune-Aug.), fall (Sept. Nov.), winter (Dec.-Feb.)] of the year to determine optimal operational period(s) and related to when cattle would occupy rangeland. To evaluate system efficiency, a frequency analysis was used with three defined ranges (5-10, 10-15, and 15-20%). In addition, system performance was predicted based on the three ranges of system efficiency using regression analysis. Energy savings associated with the photovoltaic system were calculated based on average daily solar energy available for the year multiplied by average yearly system efficiency (kWh m-2 yr1)= [(solar energy (W m2 d-1) system efficiency(%)* 7 d wk-1 52 wk yr1)]/60000wL]. The number of cows that could be watered by this photovoltaic-powered watering system was calculated using seasonal daily pumpage and daily livestock water requirements. Daily livestock watering requirements were obtained from the USDA-SCS office in Gainesville, FL. (S. Brantley, 1991, personnel communication) were 45.3 to 56.7 L hd-1 d-1 ~. >.. 0 0 100 ................. .......... ........................................ 0 I AUG I I OCT Year 1990 I I I I DEC FEB I I I I I I APR JUN AUG Year 1991 Fig. 1. Daily solar energy collected from Aug. 1990 -Aug. 1991 at the Southwest Florida Research and Education Center, Immokalee, FL.


PROCEEDINGS, VOLUME 51, 1992 77 RESULTS AND DISCUSSION Total daily solar energy was greatest in the summer months and lowest in the winter (Fig. 1). In south Florida, the month of May usually marks the end of the dry season (Fernald and Patton, 1984). Total solar energy at the earth's surface for a 25N latitude aver ages about 31800, 502980, 520980, and 390000 W m-2 for January, April,July, and October, respectively (Doorenbos and Pruitt, 1977). These values are greater than the solar energy reported in this study (Fig. I). This is partially explained by the above-normal rainfall during the study period. Hence, increased cloud cover impeded solar energy reaching the earth's surface. Rainfall was highest in the summer rainy season and lowest during the dry season (Oct.-May) (Fig.2). High incidence of cloud cover during the rainy season inhibited the pumping ability of the photovoltaic system. This system was very sensitive to interruptions (cloud cover) in solar energy reaching the photovoltaic array. System performance by season of the year for 1990-91 was variable with the highest performance in the winter and poorest performance in the summer (Fig. 3). Slope (L W-1 m-2 ) of the regression line for winter (0.028) was significantly (P<0. l 0) higher than the fall (0.021) and summer (0.017) and similar to the spring (0.026) season. The four estimated regression equations used to compare slopes among the different seasons were: Fall Spring Summer Winter -li33.38 + 0.022(solar energy) -1717.97 + 0.026(solar energy) -993.58 + 0.016(solar energy) -1005.97 + 0.027(solar energy) Beef cattle graze rangeland in Florida from December-May which corresponds to the period of greatest system performance. During the winter and spring seasons, this photovoltaic-powered watering system reliably pumped 6688 and 5456 L d-1 respec tively, compared to 4110 and 6140 L d-1 for the sum-250 225 200 ,,...._ E 175 E ..._,, 150 C 125 -C: C 100 Cl'.'. 75 50 25 0 .... ~O-,-YR NORMAL AUG OCT Year 1990 DEC FEB .... APR JUN AUG Year 1991 Fig. 2. Monthly 1990-91 average rainfall collected from the Southwest Florida Research and Education Center, Immokalee, FL. 8 6 4 2 0 ,......_ (11 8 -0 C 6 0 (11 ::J 4 0 ..c f-2 ,,___, (11 L 0 QJ -+-' 8 -QJ 6 0, 0 4 o_ E ::J 2 Q_ 0 8 6 4 2 0 0 Y = -160 + r 2 = 0.44 Y = -1 040 + 0.028 X 00 & r 2 = 0. 6 3 ~oo o '~' 0 ~~.o o coo 0 Y = -1 720 + 0.026 X r2 = 0.76 ... ,,_ Y = -990 + 0.017 X .. Winter Spring Summer 100 200 300 400 Daily Solar EnergY., W m-2 (Thousands) Fig. 3. System performance during fall (Sept.-Nov.), winter (Dec.-Feb), spring (Mar.-May), and summer (June-Aug.) 1990-91 of a photovoltaic powered water system located at the Southwest Florida Research and Education Center, lmmokalee, FL. mer and fall, respectively (values based on 276000 and 300000 W m-2 d-1 of average daily solar energy in the winter/spring and summer/fall, respectively). If cattle need 56.7 L of water per day, then this system could supply enough water for 108 cows in the winter and spring, and 89 cows in the summer and fall. Watering cattle grazing remote rangeland using this photovoltaic-powered watering system would require additional hardware (stock tanks, float valve, and a rechargable battery). A rancher should have a 2-d supply of water to insure against temporary system failure (ie. pump failure) or periods of low pumpage associated with cloudy days. System efficiency averaged 13. l % with seasonal values of 11.5%, 13.7%, 16.2%, and 11 % for the fall, winter, spring, and summer, respectively. Distribution of the system efficiency within a season was compared in a histogram (Fig. 4). For all three ranges of system efficiency (5-10, 10-15, 15-20%), winter and spring had the greatest frequency of the highest effi ciency range. During the summer and fall, the high est frequency was for the 10-15% range.


78 Sou, AND CROP SCIENCE SOCIETY OF FLORIDA 100 r--------------------~ ..., c: 80 Q) u L Q) o._ 60 >, u C: 40 rr Q) L LJ.. 20 System Efficiency 5 -1 O :Ii [IT] IO :-1 5 % 15 :-20 % Fall Winter Spring Summer (Sep -Nov) (Dec -Feb) (Mar May) (Jun Aug) Fig. 4. Frequency analysis of system efficiency for a photovoltaic-powered water system during 1990-91 located at the Southwest Florida Research and Education Center, lmmokalee, FL. In this study, the large frequency (70%) of the 1_5-20% sys~em efficiency range in the spring was partially explamed by the available solar energy in April and May. In addition, motor impedence and pump choice influence system efficiency. Hori et al. (1985) demonstrated that system efficiency was especially dependant upon motor impedence. The effect of different motor characteristics on the flow pattern of a pump becomes apparent. Thomas (1987) suggested that centrifugal pumps are ideally suited for condi tions of moderate to high flow from wells, cisterns, ~r reservoirs. The pump choice for a specific application depends on the flow rate and head requirements for the system, choice of pump position, and driving electronics. Using a linear regression analysis, prediction of system performance based on the three ranges of system efficiency is shown in Fig. 5. For a given unit (W m2 d1 ) of solar energy, 0.011 L, 0.017 L, and 0.023 L of water was pumped for the 5-10, 10-15, and 1520% ranges, respectively. This points out the benefits in output (pumpage) as system efficiency was increased. Several means exist of increasing the effi ciency of a photovoltaic-powered watering system, thereby decreasing the overall cost of the system per daily hydraulic output. Efficiency can be increased through the use of maximum power point trackers, tracking mechanisms, and reflectors. However, for "low-cost" agricultural applications (i.e. ranching) these mechanisms are not normally cost effective. Energy savings from the photovoltaic-powered watering system, based on an estimated solar energy of 216000 W m-2 d I and an average system efficiency of 13.1 %, were calculated to be 172 kWh m-2 yr1 valued at $14.85. Realistically, ranchers would operate the system on rangeland for up to 180 cl., so savings would be less. However, the unit could be moved to a ground water well in a pasture for summer use. Implications A photovoltaic-powered system for pumping water to livestock in Florida is practical due to good 8 Y = 0.011 X 6 r2=0.95 4 2 ,---... en .. (A) -cJ C: 0 0 (/] 8 ::::J 0 ..c: Y = 0.017 X f---6 .. ...__, r 2 = 0.96 en L 0 ~ 0o ooo (!) 4 ~oa, 0 0 ~o o o +-' c,Oo(IO~ 'f O 80 001 oo o r o

PROCEEDINGS, VOLUME 51, 1992 79 ACKNOWLEDGMENTS The authors acknowledge the support of the South Florida Water Management District and the Governor's Energy Office. 'Use of trade names in this publication does not imply endorsement by the Institute of Food and Agricultural Sciences, University of Florida, of the products named. REFERENCES Balliette, J. F., and J. R. Garrett. 1990. Cost of pumping livestock water. p. 21-26. In Leslie J. Krvsl (ed.) Cattlemen's Update Symposium, Reno, NV., r-..-evada Coop. Ext. 13-20 Jan. 1990. Univ. of Nevada. Doorenbos, J., and W. D. Pruitt. 1977. Guidelines for predicting crop water requirements. Food and Agriculture Organization of the united Nations. Irrigation and Drainage. Paper# 24:144p. Fernald, E. A., and D. J. Pattern. 1984. \\'ater resource atlas ot Florida. Fla. State Univ. Press, Tallahassee, FL. Garg, H. P. 1987. Advances in solar energy technology, Vol. III: Heating. agricultural and photovoltaic applications of solar energy. D. Reidel Pub. Co., Dordrecht, Holland. Helikson, H.J. 1989. Evaluation of photovoltaic-powered pumping system for micro-irrigation. M.S. Thesis, University of Florida, Gainesville, FL. Helikson, II. J., D. Z. Haman, and C. D. Baird. 1990. Pumping water for irrigation using solar energy. EES-63. Univ. of Florida, Gainesville, FL. Hori, A., and T. Abe Kanematusu. 1985 .-\n optimum design of photovoltaic direct coupled water pumping system. Transactions of the IEEE, 160(8): 1626-163 I. Hsiao, Y. R., and B. A. Blevins. 1984. Direct coupling of photovoltaic power source to water pumping system. Solar Energy. 32:489-498. SAS. 1982. SAS User's Guide: Statistics. SAS Inst., Inc., Cary, N.C. Thomas, M. 1987. Water pumping: the solar alternative. SAND87-0804. UC-63. Sandia Laboratories, Albuquerque, NM. Wolf, M. 1981. Photovoltaic solar energv conversion svsterns. lnJ. E. Kreider and F. Kreith (ed.) Solar energy handbook. McGrawHill, NY.

PAGE 100

PROCEEDINGS, VOLUME 51, 1992 79 Energycane Response to Harvest Management P. Mislevy*, M. B. Adjei, G. M. Prine, and F. G. Martin ABSTRACT Biomass may have potential as a renewable energy source to meet Florida's future energy needs. Optimal production of methane from biomass necessitates the identification and management of high yielding cultivars. Energycane (Saccharum sp. L.), L 79-1002, was selected to monitor the effect of harvest treatments at plant heights of 4, 8, and 12 ft, mature in October (16 ft), and mature in December on agronomic characteristics from 1986 to 1989. All treatments received 50 lb acre-1yr1 P205 and 100 lb acre-1yr1 K20 in one application and 300 lb acre-1yr1 N in split applications applied prior to the growth of each harvest. An additional October treatment at the mature stage was imposed which received similar P205 and K20 but only 150 lb acre-1 N. The field-plot layout was a randomized complete block, with four replicates. Plants continually harvested at the 4-ft stage decreased 89% in biomass dry matter (DM) yield from year 1 to year 4. Plants harvested once annually (mature stage) in October, regardless of N rate, or December decreased in biomass DM 43%, yielding an average of 14.4 tons acre-1 after 4 yr. Tiller density and plant density generally showed a minimal reduction (&0.05) over 4 yr. when harvested at the mature stage. Nitrogen content and in vitro organic matter digestion of biomass tissue followed an inverse relationship with plant harvest treatment, decreasing 64 and 33%, respectively, between the 4 ft and mature December treatments. Total non-structural carbohydrates in the crown (3 in. stubble and 3 in. roots) tended to increase (P:s0.05) as harvest treatment was delayed. Data indicate harvesting this energycane entry at the mature stage could extend the longevity of the plant stand several years. P. Mislevy, Agric. Res. and Educ. Ctr., Univ. of Florida, Ona, FL 33865-9706; M. B. Adjei, Agric. Exp. Stn., Univ. of Virgin Islands, St. Croix, Virgin Islands 00850; G. M. Prine, Agronomy Dep., Univ. of Florida, Gainesville, FL 32611-0621; and F. G. Mar tin, Statistics Dep., Univ. of Florida, Gainesville, FL 32611-0621. Florida Agric. Exp. Stn. Journal Series no. N-00594. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida !'roe. 51:79-84 (1992) Florida imports about 85% of its energy. As world-wide fossil fuel availability decreases or becomes cost prohibitive, Florida will need to utilize alternative energy sources. Since the state has a climate for extended plant growth, biomass may be the most logical alternative energy source. It has been known for centuries that methane gas can be produced by anaerobic fermentation, but technology and chemical conversion required for commercial application of pipeline quality gas was not available (1) until recently. Energy from biomass can be obtained by combustion, fermentation, or thermal gasification. Energy from fermentation is the most practical because of the ability to handle wet or dry feed stocks, high quality and purity of gas produced, and flexibility in size of gasification facilities. Methane yield tends to vary among species and different parts of the same plant. Treatment of plants with various nutrients, especially N, during the growth period and the age of the plants at harvest time affect methane production (11). Shiralipour and Smith ( 11) indicated young tissue of napiergrass (Pen nisetum purpureum L.) produced more methane than the older tissue probably because younger tissue is less lignified. To have a successful methane production industry, there is a need to have a continuous, year-around supply of biomass. This feedstock can be obtained from stored material or from freshly cut, highly-productive biomass species. Obtaining optimal production of methane from biomass necessitates research on the identification and management of high yielding perennial cultivars. Cultivars must grow well under limited water and fer-

PAGE 101

80 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA tility and be tolerant to nematodes, insects, diseases, saturated soil conditions, etc. Energycane, L 79-1002, was one of the entries selected for further testing due to its high DM yield, rapid growth rate, and persistence under subtropical conditions (7). The purpose of this experiment was to study the influence of harvest treatments on herbage mass, quality, and other agronomic characteristics of L 791002 energycane. MATERIALS AND METHODS The experiment was conducted over a 4-yr period from 1986 to 1989 at the University of Florida's Ona Agricultural Research and Education Center. The soil was an Ona fine sand (sandy, siliceous, hyperthermic, Typic Haplaquod) which supported 'Pangola' digitgrass (Digitaria decumbens Stent.) for the previous 3 yr. The field plot layout was a randomized complete block with four replicates of six treatments, each measuring 15 by 20 ft with a 10-ft border (Table 1). The annual fertilization program and harvest number is outlined in Table 1. Soil pH averaged 5.0 during the 4-yr duration of the study. The energycane was planted in December 1984 from stem pieces about 0.3-m long at a distance of 2.5-ft apart within the row and 3.0 ft between rows. Each stem piece contained 2 to 3 nodes and was placed horizontally in the soil at about a 3in. depth using a tree-planting machine (R.A. Whitfield Mfg Co. Mableton, GA). 1 If a stem piece failed to produce a plant, it was replaced with a rooted plant resulting in a 100% plant stand. After four harvest years, plants were again counted to determine the influence of harvest treatments on plant stand. Harvest treatments were initiated about 1.5 yr after planting. A 10-by 3-ft sample area from the center row of each treatment was harvested to a 3-in. stubble to determine total biomass. Approximately 2 Table 1. Harvest treatments and cultural practices imposed on L79-1002 energycane, 1986-1989. Nitrogent Harvest treatment Harvest Application* Total number yr' lb acre-1 lb acre-1 yr-1 4 ft 4 75 300 8 ft 2 150 300 12 ft I 300 300 Mature (October) 1 300 300 Mature (October) 1 150 150 Mature (December) 1 300 300 t All treatments received one annual spring application of 50 lb acre-1 P205 100 lb acre-1 K20, 2.5 lb acre-1 Fe, 0.9 lb acre-1 Zn, 0.9 lb acre-1 Mn, 0.4 lb acre-1 Cu, and 0.4 lb acre-1 B. *Each N split was applied prior to the growth of each harvest. Indicates month treatment harvested. 1Mention of a proprietary product name is for identification purposes only and does not imply warranty or endorsement to the exclusion of other products. to 5 plants were selected as a subsample to represent average plant height of the treatment. The subsample was chopped into pieces 1 in. or less and dried at 140 F to determine percentage DM. A representative portion of plant material was selected from each subsample and analyzed for total N (4,5) and in vitro organic matter digestion (IVOMD) (8). Following each harvest, total live tillers (stems and leaves photosynthetically active) were counted in each 10-by 3-ft plot. In 1986 and 1987, an area of approximately 12 ft2 containing two plants was clipped to the soil surface as plants grew in 2-ft increments to obtain leaf area index (LAI) data. These data came from a portion of the plot which originally measured 20 ft long; 10 ft for yield and 10 ft to obtain one or two LAI treatments from each harvest treatment. The 2-and 4-ft LAI came from the 4 ft harvest treatment, 6 and 8 ft LAI from the 8 ft harvest treatment etc. All green leaves were separated from the stems at the collar region and passed through a Li-Cor portable area meter model LI 3000 (Li-Cor Inc, Lincoln, NE 68504). Core samples (4-in. diam) of plant crowns were removed from each treatment at each harvest (1986, 1987, and 1988) to monitor total non-structural carbohydrates (TNC) concentration (%). During the spring (March) of 1990 after the threat of frost was over, percent and yield (g) of TNC were again measured. Each core consisted of a 3-in. portion below the soil surface (roots) and 3-in. portion above the soil surface (stubble). Samples were immediately washed to remove soil and debris, heated to 212 F for 1 h and subsequently dried at 140 F in a forced-draft oven. Dried tissue was ground to pass a 1-mm mesh screen, sealed in plastic bags, and stored in a freezer with silica-gel desiccant. Extraction of TNC from ground tissue followed the enzymatic procedure modified by Smith (12). The extract was analyzed for reducing sugars using Nelson's (9) colorimetric adaptation of Somogyi's (13) copper-reduction method with glucose standards. For those response variables that were independent of years, a pooled-year analysis was performed. The three responses CP, IVOMD, and total tillers were found to have treatment effects which depended on years. A significant level of 0.05 was used for all F-tests. Within each year, the number of times a plant was harvested depended on the treatment. The number of harvests varied from 1 for treatments 3 to 6, to 4 for treatment 1. RESULTS AND DISCUSSION Biomass Yield, Tiller Density, Persistence, and LAI. Biomass DM yield of harvest treatments was independent of year. Harvesting well-established L 791002 energycane each time plants attained a height of 4 ft resulted in the lowest (P::c;0.05) biomass DM yield compared to all other harvest treatments (Table 2). Delaying harvest from the 4-ft plant height to the mature stage in October resulted in a 320% (P::c;0.05) increase in biomass DM yield. This averaged out to a

PAGE 102

PROCEEDINGS, VOLUME 51, 1992 81 Table 2. Influence of harvest treatments on total seasonal biomass yield averaged over 1986-1989. Harvest N Biomass treatment treatment yield lb acre' yr' tons acre' 4 ft 300 5.0d* 8 ft 300 10.6 C 12 ft 300 15.1 b Mature (October)t 300 21.0a Mature (October) 150 19.2 a Mature (December) 300 20.5 a *Means within the column followed by different letters are sig nificantly different (Duncan's Multiple Range Test, P:s0.05) tlndicates month treatment harvested. 100% (P::::;0.05) DM yield increase for each 4-ft increase in plant height. Alexander (2) indicated delaying harvest frequency from 2 to 12 months for sugarcane (Saccharum officinarum L.) and 2 to 6 months for napiergrass increased DM yield of biomass 700 and 140%, respectively. Once plants attained the physiological mature stage iL October, biomass DM yield remained relatively stable even if harvest treatments were delayed until December (P2=0.05). Decreasing the total N rate from 300 lb acre1 to 150 lb acre1 had no effect (P2=0.05) on total biomass yield (Table 2) when harvested in October. Nitrogen content of above-ground biomass was 0.6% and 0.3% on a DM basis for the 300 and 150 lb acre I N rates, respectively. Nitrogen uptake in the above-ground tissue was about 87% for the 300 lb N acre1 rate and 83% for the 150 lb N acre1 rate. Even though the high N rate did not increase dry biomass yield (P2:0.05), N content in the tissue was doubled which could increase methane production. Shiralipour and Smith (11) suggest that an increase in protein content of plant tissue, due to N fertilization may be a factor which favors methane production. They ( 11) increased N concentration from 1 to 10 mg L1 in the growth media consisting of Hoagland's solution minus N (6) and increased methane yield (std m3 kg-1 volatile solids added) from entire water hyacinth [Eichhornia crassipes (Mart.) Sohns] plants by 38%. Yield of biomass decreased (P:S0.05) linearly over years. Average biomass yield over all six treatments in 1986 was 21.3 tons acre1 (Table 3). These yields dropped a total of 52% over the 4-yr duration of the study. There was no harvest treatment X year interaction (P2=0.05) on biomass yield, despite the variation in treatment response over years (Table 3). Harvesting L 79-1002 energycane continuously at 4-ft intervals, resulted in a 89% yield decrease after 4 yr. De laying harvest until plants attained an 8-ft height (2 harvests year1 ) reduced the yield loss after 4 yr by 32 percentage units, when compared with the 4 ft harvest treatment (Table 3). Biomass yields at the mature stage, although approximately 50% of 1986 yields, still ranged from 13.8 to 15.5 tons acre1 in 1989, 4 yr after initiation of the study. These yields are quite respectable compared to the most productive annual species [forage or sweet sorghum (Sorghum bicolor L. Moench)] which yield about 10 tons acre1 year1 and are susceptible to Pythium and nematode problems when grown more than 2-yr consecutively (10). These data indicate harvesting plants at the mature stage, in October or December, with 150 or 300 lb N acre 1 would be acceptable in terms of yield and persistence, compared with the more frequent harvest treatments of 4 to 12 ft. Highest tiller density was found in the most frequent harvest treatment (4 ft) (Table 4). As harvest frequency was delayed from 4 ft to the mature stage, tiller density declined from 7.0 to 3.1 tillers ft2 and stabilize at this density. Tillers tend to develop throughout the warm season, even within the dense uncut biomass. Tillers may attain a height of 6 in to 2 ft and die, depending on light penetration in the plant canopy. Frequently harvested plots (4 ft) have less vegetative canopy and higher solar radiation reaching the soil surface, consequently develop and retain many tillers area2 However, as intensive harvest frequency (4 ft) continued over a 4-yr period, the plants weakened, and tiller density decreased about 68% (Table 4). Whereas, plants harvested at the mature stage in October continued to maintain about 3 tillers ft2 Table 3. Influence of harvest treatments X year on oven dry biomass yield and percentage biomass yield reduction over 4 yr. Harvest treatment 4 ft 8 ft I 2 ft Mature (October)t Mature (October) Mature (December) Avg. N treatment lb acre' 300 300 300 ,mo 150 300 Year 1986 1987 1988 1989 ------------------------------------tons acre' ---------------------------------10.1 5.4 3.2 1.1 16.7 10.5 8.1 7.1 23.8 15.0 12.2 9.3 27.1 20.9 20.5 15.5 21.0 23.7 18.1 13.9 28.8 22.1 17.4 13.8 21.3a* 16.3b 13.3c IO.Id *Means within a row followed by different letters are significantly different (Duncan's Multiple Range Test, P:s0.05). tlndicates month treatment harvested. Yield reduction after4 yr % 89 57 61 43 34 52

PAGE 103

82 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table 4. Influence of harvest treatments X year on tiller density and percentage tiller reduction at end of 4 harvest yr. Year Harvest N Tiller treatment treatment 1986 1987 1988 1989 reduction lb acre- yr-1 -----------tillers ft-1 ---------------------------------------% 4 ft 300 J0.6a*A 7.8 a B 6.2 aC 3.4aD 68 8 ft 300 7.2h A 4.4 be B 3.4 c C 2.9aC 60 12 ft. 300 3.7 C B 5.0b A 4.7bA 2.6a C 48 Mature (October/ 300 3.7 C A 3.2d A 3.2cA 3.0aA 19 Mature (October) 150 2.8 C A 3.5 cd A 3.0cA 2.7 aA 23 Mature (December) 300 3.5 C A 3.7 cdA 3.0cA UbB 65 *Means within columns (a,b,c) or rows (A,B,C) followed by different letters are significantly different (Duncan's Multiple Range Test, P50.05). tlndicates month treatment harvested. The 4-ft harvest frequency also had a profound effect on percent plant stand, which declined to 13% after four harvest seasons (Table 5) or an 87% loss. Harvesting plants at the mature stage had little effect on plant stand after 4 yr, averaging 97%. Leaf area index, determined at 2-ft increments, exhibited a quadratic response pattern, increasing until it peaked at 12.2 and 11.7 in 1986 and 1987, respectively (Fig. 1). Following the curvilinear pattern, LAI peaked at 8(1986) and 10-ft (1987) plant heights and then decreased as plant height increased to the 14-ft stage at which time plants began to flower. This curvilinear LAI pattern would be expected, since these tall biomass plants form a dense canopy. As plants attain heights of 10 to 14 ft, many of the bottom leaves become necrotic and senesce, due to lack of light. The average LAI of 6 at 14 ft represents the green leaves basically on the upper portion of the plant. Lack of N at this physiological stage ( 14 ft) may also have an influence on decreasing LAI. Nitrogen Concentration and IVOMD There was a treatment X year interaction (P:::;0.05) for N content and IVOMD. Generally, for most years the N content in biomass tissue was higher (P:::;0.05) at the 4-ft harvest treatment, than the more advanced harvest maturity stages (Table 6). Average N content over the 4-yr period was 1.4, 1.05, and 0.9% for the 4-, 8-, and 12-ft harvest stages, respec tively, and averaged 0.5% for all mature stages. Aver-Table 5. Influence of harvest treatment on plant stand after 4 yr. Harvest N Plant treatment treatment stand lb acre-1 yr' % 4 ft 300 13 d* 8 ft 300 85c 12 ft 300 89bc Mature (October/ 300 94ab Mature (October) 150 99a Mature (December) 300 97 a 'Means within the column followed by different letters are sig nificantly different (Duncan's Multiple Range Test, P50.05) tlndicates month treatment harvested. ;;;; age N content in plant tissue at the mature stage with 300 lb acre-1 N was 0.58% compared with 0.33% at the same stage of maturity with 150 lb acre-1 N. Shiralipour and Smith ( 11) indicated plant tissue with increased N content may be a factor which favors methane production. In vitro organic matter digestion of L 79-1002 biomass tissue harvested at the 4-ft stage averaged 53.3% (Table 7). This value is 9 to 12 percentage units higher than the biomass harvested at the 8to 12-ft harvest stage. Throughout this paper, it has been indicated harvesting at the mature stage produced highest biomass yield, uniform tiller production, and most persistent plant stand over the 4-yr period. However, this stage also yields the lowest IVOMD, 35% (Table 7). Presumably, such low IVOMD values (35%) would have less methane gas potential per pound of DM, but higher methane yields per acre due to higher biomass yields, than biomass tissue harvested at the 4-to 12-ft harvest stage (46% IVOMD). Bjorndal and Moore (3) trying to relate chemical characteristics to fermentability indicated great variation among and within species in plant chemical composition. Therefore, a complex relationship between plant chemistry and fermentabil-16 14 12 10 8 6 4 2 0 0 1986 LAI -~ ~--1 1987 LAI y = -6.3 + 4.4H 0.26H 2 y = 0.9 + 1.8H 0.09H 2 H = Plant Height (ft) -----------,:, ,$ I_] ':Ji rr 2 4 6 8 10 12 Plant Height (ft) 14 16 Fig. 1. Influence of plant height on leaf area index (LAI) of L79-1002 energycane during 1986 and 1987.

PAGE 104

PROCEEDINGS, VOLUME 51, 1992 83 Table 6. Influence of harvest treatments X year on N content of L79-1002 energycane biomass tissue. Harvest treatment N treatment 1986 1987 Year 1988 1989 4-yr average lb acre-1yr1 -----------------------------------------------------o/c, -------------------------------------------------4 ft 8 ft 12 ft Mature (October)t Mature (October) Mature (December) 300 300 300 300 150 300 1.3 a* C 1.1 b B 0.7 C C 0.6cdB 0.4d A 0.5cdA 1.3 a C 1.3 a A I.0bB 0.6c B 0.3 cA 0.6c A 1.4 a B 1.6 a A 1.4 0.9b C 0.9 b C 1.1 1.3 a A 0.7 be C 0.9 0.4cd B 0.9b A 0.6 0.2d A 0.4 C A 0.3 0.5 C A 0.5 C A 0.5 'Means within columns (a,b,c) or rows (A,B,C) followed by different letters are significantly different (Duncan's Multiple Range Test, Ps0.05) tlndicates month treatment harvested. Table 7. Influence of harvest treatments X year on in vitro organic matter digestion (IVOMD) of L 79-1002 energycane biomass tissue. Harvest treatment N treatment 1986 1987 Year 1988 1989 4-yr average lb acre' yr' ------------------------------------------------------% -----------------------------------------------------4 ft 8 ft 12 ft Mature (October/ Mature (October) Mature (December) 300 300 300 300 150 300 51.6 a* B 42.5 b B 38.5c C 32.5deA 30.2e B 33.7 cl B 52.3 a B 52.3 a B 57.1 aA 53.3 47.5 bA 42.7bB 43.3 b B 44.0 42.4 c A 43.0bA 40.0 b B 41.0 35.8d A 34.9c A 35.2 c A 34.6 35.5 d A 36.4 c A 34.6c A 34.1 37.2dA 34 .. c AB 36.6cAB 35.5 'Means within columns (a,b,c) or rows (A,B,C) followed bv different letters are significantly different (Duncan's Multiple Range Test, P:S0.05) tlndicates month treatment harvested. ity must be understood before the fermentability of a feedstock can be predicted from its chemical com position. Crown Total Non-structural Carbohydrates Harvest treatment X year interaction was not significant (P2:0.05) for crown total non-structural carbohydrates. Data pooled over years showed a general TNC increase as harvest frequency was delayed (Table 8). Harvesting L 79-1002 energycane at the mature stage regardless of harvest date or N rate, resulted in a high (P::s0.05) concentration of TNC (15.3%) in crown samples, compared with plants harvested at any younger (4-12 ft height) physiological stage (Table 8). Lowest (P::s0.05) TNC concentration (9.7%) was obtained when plants were harvested most frequently (4 ft) throughout the warm season. This TNC data helps explain why four harvests yr1, weakened the plants, and produced lowest DM yield (Tables 2 and 3) and poorest persistence (Table 5). Significant (P::s0.05) differences were observed in percentage TNC during the spring of 1990. Plants harvested the most frequently (4-and 8-ft harvest treatment) during the previous harvest year tended to be higher (P::s0.05) in TNC (5.3 to 6.6%) (Table 9). Whereas, plants harvested at the mature stage in October and December tended to contain a low to medium concentration of TNC (3.5 to 4.7%). The amount of TNC (g) was not different (P2:0.05) between harvest treatments, ranging from 1.9 to 4.3 g. This response variable takes into account% TNC X root mass. Therefore, even though % TNC was highest for the 4-to 12-ft treatments, root mass was low (due to frequent harvests) producing a non-significant (P2:0.05) difference between harvest treatments for TNC (%). These 1990 spring data do not reflect the results of biomass yield and % plant stand measured in this study. Even though plants were exposed to 27, 22, and 22 Fon 4, 24, and 25 Dec 1989, fol lowed by several frosts in January, 1990, one would Table 8. Influence of harvest treatments on average total nonstructural carbohydrates (TNC) over 3 yr. Harvest N Average crown treatment treatment TNC lb acre- yr' % 4 ft 300 9.7 c* 8 ft 300 12.2 b 12 ft 300 11.1 b Mature (October)t 300 15.6 a Mature (October) 150 16.1 a Mature (December) 300 14.2 a *Means within a column followed by different letters are signif icantly different (Duncan's Multiple Range Test, P:S0.05). t1ndicates month treatment harvested.

PAGE 105

84 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table 9. Influence of harvest treatments on total non-structural carbonydrates (TNC) in crown tissue during the spring of 1990 (90 d following the last harvest). 1990 Crown TNC Harvest N treatment treatment Percent Weight lb acre-' yr' g 4 ft 300 6.6 a* 3.5 a 8 ft 300 5.3 ab 4.3 a !2 ft 300 3.6 C 1.9 a Mature (October)t 300 3.5 C 2.3 a Mature (October) 150 4.6 be 3.2 a Mature (December) 300 4.7bc 3.5 a *Means within a column followed by different letters are signif icantly different (Duncan's Multiple Range Test, P:c;0.05). tlndicates month treatment was harvested. expect the % TNC in the 1990 spring sampling to have a similar relationship to data in Table 8, with plants harvested at the mature stage having the highest TNC concentration. CONCLUSIONS Harvesting L 79-1002 energycane at a 4-ft interval over a 4-yr period resulted in a 75% decrease in biomass yield compared with treatments harvested once annually, regardless of N rate. This same 4-ft harvest treatment also decreased 87% in plant stand, when compared with a single harvest treatment in October or December. As harvest treatments was delayed from the 4-ft multiple cut, to a single cut (mature harvest), N and IVOMD decreased 66% and 35%, respectively, at the same time crown TNC increased by 37%. REFERENCES 1. Abelson, Philip, H. 1987. Foreword. p. XI to XIV In W. H. Smith andJ.R. Frank (ed.) Methane from Biomass: A Systems Approach. Elsevier Appl. Sci., New York. 2. Alexander, A.G. 1982. Management of tropical grasses as a year-round alternative energy source p. 87-103 Symp. Energy from Biomass and Waste V. Lake Buena Vista, FL. 26-30 Jan. 1981. Inst. Gas Tech. Chicago, IL. 3. Bjorndal, K. A. and]. E. Moore. 1987. Chemical characteris tics and their relation to fermentability of potential biomass feedstocks. p. 355-365. In W. H. Smith and J. R. Frank (eel.) Methane from Biomass: A systems Approach. Elsevier Appl. Sci., New York. 4. Gallaher, R. N., C. 0. Weldon, and J. G. Futral. 1975. An aluminum block digestor for plant and soil analyses. Soil Sci. Soc. Am. Proc. 39:803-806. 5. Hambleton, L. G. 1977. Semiautomated method for simultaneous determination of phosphorus, calcium, and crude protein in animal feeds. J. Assoc. Off. Anal. Chem. 60:845854. 6. Hoagland, D. R. and D. I. Arnon. 1950. The water culture method for growing plants without soil. Berkeley, Calif. Agric. Exp. Sta. Circ. 34 7. 7. Mislevy, P., J. P. Gilreath, G. M. Prine, and L. S. Dunavin. I 987. Alternative Production Systems: Nonconventional Herbaceous Species. p. 261-276. In W. H. Smith and J. R. Frank (ed.) Methane from Biomass: A Svstems Approach. Elsevier Appl. Sci. New York. 8. Moore, J. E., and G. 0. Mott. 1974. Recovery of residual organic matter from in vitro digestion of forages. J. Dairy Sci. 57: !258-1259. 9. Nelson, N. 1944. A photometric adaptation of the Somogyi method for the determination of glucose. J. Biol. Chem. 153:375-380. JO. Prine, G. M., P. Mislevy, R. L. Stanley,Jr., L. S. Dunavin, and D. I. Bransby. 1991. Field production of energycane. elephantgrass, and sorghum in southeastern United States. Proc. Inst. Gas Tech. Energy from Biomass and Wastes XV. Washington DC. March 25-29, 1991. Inst. Gas Tech. Chicago, II. 11. Shiralipour, A., and P. H. Smith. 1984. Conversion of biomass into methane gas. Biomass 6:85-92. 12. Smith, D. 1981. Removing and Analyzing total non-structural carbohydrates from plant tissue. Wisconsin Agric. Exp. Stn. Res. Rep. l\o. R2107. Univ. of Wisconsin, Madison, WI. 13. Somogyi, M. 1945. A new reagent for the determination of sugars. J. Biol. Chem. 160:61-68.

PAGE 106

PROCEEDINGS, VOLUME 51, 1992 85 Effect of Dolomite Rate and Placement on Leucaena Forage and Seed Yield R. S. Kalmbacher*, J.E. Rechcigl, and F. G. Martin ABSTRACT Leucaena leucocephala (Lam.) de Wit is a high protein leguminous shrub that can be utilized for cattle feed. It can be grown in Florida, but soils must be limed. The purpose of this research was to determine the response of leucaena, K340, grown on a Spodosol to dolomite rate (0 to 8 Mg ha-1 ) and placement (1-m vs. 3-m treated strip) Soil pH and soil and leaf contents of Ca and Mg increased linearly (P<0.01) with increases in dolomite rate, and these contents were usually not affected by placement method. Yield of leaf, grazeable stem ( <8 mm) and seed were affected linearly (P<0.01) by dolomite rate, but increases always depended on placement, with higher yield coming from 3-m strip. Root activity, as determined by Sr uptake, was largely confined to dolomite-treated areas because of high (16 mg kg-') exchangeable Al content in untreated soil. Leucaena can be grown on a dolomite-strip treated area, but a 3-m wide strip is recommended with rates of 8 Mg ha- of dolomite. Leucaena is a tropical, leguminous shrub that has value as a high protein cattle forage (Gray, 1968; Oakes, 1968). There appears to be some potential for leucaena in Florida on well-drained soils where leaves and stems (<6 mm diameter) of 13 accessions (four harvests) annually averaged 6.2 Mg ha-1 in 2 yr (Othman et al., 1985). Soil reaction is important to the growth of leucaena. Ahmad and Ng ( 1981) in Malaysia considered performance of leucaena satisfactory at pH 4.8 to 5.0, and reported increases in foliar concentration of Al and Mn with decreasing pH. Hu and Cheng (1980) in China studied lime rates and found an increase in leucaena yield from pH 4.8 up to pH 7.0. Increases in yield and pH were accompanied by decreases in Al availability in soil. Koffa and Mori ( 1987) found small differences in leucaena seedling growth among four leucaena strains in the pH range 4.5 to 6.5, but Al concentration greatly affected seedling growth. Soils in Florida are predominantly acid, infertile Spodosols. Most ranches, at least in south Florida where most of the state's cattle are raised, are large and extensively managed. Liming is expensive and decisions concerning application of liming materials must be carefully considered. This study was conducted to determine the effect of dolomite on soil reaction and growth of leucaena. Because leucaena is usually established in rows, banding or placing dolomite in strips under the row was also studied in an effort to reduce the area treated. Our idea was to provide a practical way of amending soil pH so that leucaena could be established along fire guards or roads in native range, thereby providing cows with a protein bank in the late summer and fall when they utilize range. R. S. Kalmbacher and J. E_ Rechcigl, University of Florida, Agric. Res. and Ext. Center, Ona, Florida 33865; and F. G. Martin, Statistics Dep., Univ. Florida, Gainesville, Florida Agric. Exp. Sta. Journal Series no. R-02088. *Corresponding author. Contribution published in Soil CrojJ Sci. Soc. Florida Proc. 5 l :85-89 (l 992) MATERIALS AND METHODS Research was conducted at the Ona Agricultural Research and Education Center (AREC) on a Po~ona fine sand (sandy, siliceous hyperthermic UltIC Haplaquod). The plot area had never received lime or fertilizer and native vegetation had been removed by heavy disking. On 27 Apr 1988, the plot area was disked, and the following rates of dolomite were applied: 0, 1, 2, 4, 6 and 8 Mg ha-. Dolomite contained 730 g kg- CaCO3 and 230 g kg- MgCO3 with a CaCO3 equivalence of 100%. Dolomite was applied uniformly by hand to 3-by 6-m plots (3-m strip) or to a 1-m strip that extended the length of the 3by 6-m plot. The 6 ~g ha-1 rate of dolomite was not applied in the 1-m stnp method of placement. Between each plot was a 3-m untreated buffer. Plots were disked lightly after dolomite application. Seeds of leucaena K340, were germinated in March 1988 and then planted in peat pots. On 13 June 1988, leucaena plants were transplanted into the field. Each_ plot had a single row of five plants spaced 1 m apart m the row. Each plant was protected with a a 30-cm tall, wire-mesh cage. The entire experiment was hand fertilized in June 1988 with 25 kg ha-1 of P, 100 kg ha1 of K, 22.4 kg ha1of F-503 micronutrient mix (180, 70, 75, 30 and 30 g kg-1 of Fe, Zn, Mn, Cu and B, respectively) and 22.4 kg ha-1 S. Dolomite was not reapplied, but P and K were reapplied at 25 and 100 kg ha', respectively, in April 1989. On 5 Mar and 28 June 1990, 12 and 50 kg ha' P and K, respec tively, were applied at each date. Leucaena plants were harvested and plant height was determined on 28 Nov 1989, 27 June and 9 Dec 1990. At each harvest leaves and petioles were sampled first, followed by all stems <8 mm diameter. In June 1990, stems were not harvested. In December 1990, seed pods were harvested, and yield of seed and pods (after shelling seed) were determined. Leaves_ and stems were dried ( 40C for 72 hr) and weighed for dry matter yield determination. Leaf samples were ground and analyzed for N (Gallaher et al., 1975; and Hambleton, 1977). Tissue was ashed at 500C for 6 hr, and nutrients were extracted using 0.302 M HCl. Solutions were analyzed for Ca, Mg, K, Fe and Zn using an atomic absorption spectrophotometer and for P using a spectrophotometer and c~lorimetric method (Murphy and Riley, 1962). Soil was sampled (0 to 15 cm depth) from within the length of the row in April (pre-treatment), June and September 1988; May and October 1989; and February 1990. In May 1991, soil was sampled O to 7.5 cm depth. Samples were analyzed for pH, P, K, Ca, Mg, Fe, Mn, Zn, Cu at the University of Florida Soil Testing Laboratory (Hanlon and DeVore, 1989).

PAGE 107

86 SOIi. AND CROP SCIENCE SOCIETY OF FLORIDA Soil organic matter (OM) (Hanlon and De Vore, 1989) and exchangeable Al Uayman and Sivasubramanium, 1974) were determined on soil sampled April 1988 and May 1989, respectively. To determine the extent of leucaena root elongation laterally from the center of the plot (row), 50 g SrC13 was applied to plots that received 8 Mg ha1 dolomite in both 3-m and 1-m strip placement methods. On 23 Apr 1991, SrCl3 was applied in 100 ml of water (total per plot) in two shallow (2-to 3-cm deep) parallel trenches 1.5 m on both sides of the leucaena row. On 29 May, developing leaflets were sampled from plants in treated rows, dried, ground and analyzed for Sr using atomic absortion spectrophotometry. The eight, 8 Mg ha1 dolomite-treated plots were retreated on 29 May and resampled on 28 June then analyzed for Sr. Untreated SrC13 plants, which received 6 Mg ha1 dolomite, were sampled as a control for Sr content and the measured value subtracted from each treatment. The experimental design was a 6 by 2 incomplete factorial of dolomite rates and methods of placement in a randomized, complete block with four replicates. Data were analyzed using General Linear Models (SAS, 1985): R2 values were calculated to indicate the percentage of the variation explained by the regression equation after adjustment for all other sources of variation in the analysis. RESULTS AND DISCUSSION Soil and Tissue Responses Initial soil pH averaged 4.2, and OM content averaged 28 g kg1 On each of the subsequent dates when pH was monitored, soil pH was found to be a linear function of dolomite rate (Table 1). The effect of rate on pH did not change over time. The highest pH measured was 5.5 at 8 Mg ha1 rate. According to our data, each additional Mg ha1 of dolomite resulted in an increase of about one-tenth pH unit for the Oto 15-cm depth. In our study, 8 Mg ha1 resulted in only about a 1.3 unit increase in pH. Fiskell et al. (1964) applied 1. 79, 3.58 and 7. 17 Mg ha 1 limestone to a Leon fine sand with an initial pH of 4.2 and at 6 months after treatment found pH 4.4, 4.9 and 5.2 respectively. This report was similar to our findings. Henderson ( 1960) summarized results of soil testing in Florida and stated that 2.4 to 4.6 Mg ha1 of CaCO3 applied to a sand with an OM concentration of 20 to 40 g kg-1 on would result in a 1.0 pH unit increase in the upper 15 cm soil. Several other reports in the literature were similar to Henderson's findings. O'Donnell et al. ( 1991 ), working on Pomona fine sand at the Ona AREC, applied 0, 3.4, 6.8 and 10.2 Mg ha1 CaCO3 in a field study and found pH increased quadratically: 4.6, 5.4, 5.8 and 6.0, respectively. Breland and Locascio (1962) applied 1.1, 2.8 and 4.5 Mg ha1 of dolomite to an Ona fine sand and found pH increased linearly from 4.9 initially to 5.3, 5.8 and 6.2 at 7 mo post-treatment. Breland (1964) applied 1.7, Table I. Linear equations (P<0.05) from analyses of variance for response of soil pH, Ca, Mg and K to dolomite rate (0, 1, 2, 4, 6, 8 Mg ha) and placement methods (3-m vs. 1-m strip). Ona, FL. Element pH Ca Mg K June 1988 at 4.7 118 35 NS b 0.09 82.5 25.2 R2 0.64 0.54 0.53 September a 4.5 135 27 40 b 0.09 46.4 15.6 1.66 R' 0.65 0.62 0.69 0.25 May 1989 a 4.2 150 29 NS b 0.09 33.8 10.l R' 0.69 0.44 0.60 October a 4.7 182 11 NS b 0.09 53.3 19.3 R 0.73 0.67 0.69 February 1990 a 4.5 :j: 17 35 b 0.10 16.8 0.65 R' 0.70 0.46 0.68 0.08 ty = a + bX (dolomite rate, Mg ha 1). Elemental content is mg kg-'. *Dolomite rate and placement interacted for Ca content: 3-m strip = 116 + 81X and 1-m strip = 116 + 30X. Placement method significant: 3-m strip = 2.2 mg ha': 1-m strip = 1.2 mg ha 1 3.4 and 6. 7 Mg ha1 CaCO3 to a Leon fine sand with initial pH of 4.2, and pH after 240 d was 4.6, 5.0 and 5.5, respectively. Data in Table 1 are based on soil samples from Oto 15-cm depth. Since dolomite was surface applied and disked lightly, we suspected that much of the liming material was still near the surface. Soil pH in the upper 7.5 cm depended on dolomite rate: 4.5 + 0.1 ldolomite rate, Mg ha1 Predicted and observed pH (0-to 7.5-cm depth) at the 8 Mg ha1 rate were both 5.4, which was little different from that measured in samples collected Oto 15 cm. Soil pH was affected by method of dolomite placement (independently of rate) on May and October 1989 and February 1990 when soil pH in the 3-m strip exceeded that in the 1-m strip placement. The respective pH means were: 4.7 and 4.5; 5.0 and 4.8; 5.1 and 4.8. Method of placement was not significant (P>0.05) at June and September 1988. Soil Ca and Mg contents increased linearly with increases in dolomite rate (Table 1). Both Ca and Mg contents were not affected by method of placement except in February 1990 when each additional Mg ha-1 of dolomite resulted in an 81 mg kg1 increase in soil Ca content in :3-m strip placement compared to 30 mg kg1 for the 1-m strip placement. Soil P (avg 5.0 mg kg') and micronutrient contents (avg: Zn, 0.8; Cu, 0.05; Mn, 0.4; Fe, 8.4 mg kg1 ) were not affected (P>0.05) by dolomite rate or by placement methods. Soil K content was increased linearly by (P<0.05) dolomite rate on two of the five

PAGE 108

PROCEEDINGS, VOLUME 51, 1992 87 dates tested (Table 1). Soil K values were very low, and the increases brought about by increasing dolomite rate do not seem to be biologically important. Tissue Fe and Zn were the only minerals affected (P<0.05) by dolomite rate in 1989 (Table 2). Content of Fe decreased when dolomite rate increased to about 5 Mg ha 1 then Fe content increased. Content of Zn decreased with additions of dolomite up to 6 Mg ha-1 then increased in content at 8 Mg ha-1 We do not know why these responses occurred quadratically. Leucaena began growth in February and was about 10 months old at harvest in 1989. We felt advanced tissue age and delay after P and K fertilization could be responsible for lack of P and K response in 1989. For this reason fertilization and harvest were split in 1990 (March and June; June and December, respectively). In June 1990, content of Ca and Mg in leaves depended on dolomite rate and method of placement (Table 2). In December, Ca and Mg content were affected by rate only. Content of Kin leaves was affected by dolomite rate, but not by placement in both June and December. Neither rate nor placement affected tissue N, P or micronutrient contents in June, but both factors affected P content in December. These K concentrations for leaves are considerably lower (especially December) than those reported for leaves/stems of Florida-grown leucaena (Othman et al., 1985), but K fertilizer rates were 250 kg ha I yr-1 in that study. Our leaf Ca and Mg contents at the 8 Mg ha-1 dolomite rate are similar to those of Othman et al. (1985). Plant Yield Responses Inspite of periodic saturated soil during the rainy season (J unc to September), leucaena persisted through the study. There was no observed psyllid or leaping lice (Heteroj>yslla spp) damage in any year. Psyllids can be a major problem on leucaena in Florida (Othman et al., 1985). At the first harvest in November 1989, there was a dolomite rate by method of placement interaction for plant height. Leucaena plants were taller (P>0.01) m 3-m strip method of placement (cm=59+6.ldolomite rate, Mg ha1 ; R2=0.82) than 1-m strip placement (cm= 59 + 3.2dolomite rate, Mg ha-1 ; R2=0.82). The hard freeze (-5.5C) on 25 Dec 1989 killed all plant tops to ground level, but there was no plant mortality. In June 1990 (at the second sampling for yield) leucaena height was not affected by method of dolomite placement, but was affected by dolomite rate (cm=43+7.7dolomite rate, Mg ha-1 ; R2=0.74). Between June and December, 1990, which was the third sampling for yield, little additional growth in height occurred. Plants replaced leaves removed in June, and the height response to increasing dolomite rate was similar to that in June (data not shown). Both leaf and stem yield in November 1989 increased linearly to increases in dolomite rate, but there was a rate by placement interaction for leaf yield (Fig. 1). Stem yield was affected (P<0.02) by 60 Leaf __ [ Placement Stem -,,,,_ .. I Placement 3mstrip:z 10.2 gplarrt1 1mstrip= 6.9 gp1an11 50 40 -C: _!!! 30 Q. Cl 20 10 trip= 6.0 + 5.8 D 0 1m strip= 6.0 + 3.0 D I 0 0 3 4 5 6 Dolomite Mg ha-1 0 7 Fig. 1. Leaf and stem (<8 mm diameter) dry matter yield of leucaena, K340, grown on a Spodosol treated with five dolomite rates and two methods of placement. November 1989. Footnote: Independent variable (D) in equations is dolomite rate, Mg ha-1 Table 2. Equations for leucaena leaf tissue mineral contents (P<0.05) from analyses of variance where dolomite rate (0, 1, 2, 4, 6, 8 Mg ha) and placement were variables. Ona, FL. Date Dec89 June90 Dec90 Rcsponset Fe Zn Ca Mg K l' K Ca Mg Placement" Avg. Avg. 3m Im 3m Im Avg. 3m Im Avg. Avg. Avg. tFe,Zn in mg kg-1 of tissue dry matter. "Placement: 1-m or 3-m wide strip. Independent variable is dolomite in .'\1g ha 1 Intercept Linear* Quadratic R' 61 -4.7 0.50 0.12 29.0 -3.8 0.35 0.33 4 0.7 0.75 4 0.37 0.75 2.6 0.17 0.56 2.6 O.lfi 0.56 7.7 3.5 0.25 2.8 -1.26 0.43 2.8 -1.3 I 0.43 2.9 -1.'ll 0.13 9.5 1.31 0.52 3.4 0.34 0.51

PAGE 109

88 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA -C method of dolomite placement independently of rate. Greater yield of stems resulted from dolomite applied to the 3-m compared a 1-m strip. Each additional Mg ha of dolomite increased leaf yield 5.8 g plant1 in 3-m strip dolomite placement vs 3.0 g plant1 in 1-m strip placement. Leaf:stem ratio in 1989 was affected (P<0.02) by dolomite rate (ratio=2.18+0.16(dolomite rate, Mg ha1 ; R2 = 0.54) and dolomite placement (P<0.03) (1-m strip = 2.6; 3-m strip = 3.6). Total 1989 yield (leaves and stems from a single harvest) depended on both dolomite rate and method of placement. For 3-m strip placement, total yield was: g plant'= 10 + 7.3dolomite rate, Mg ha1 (R2 = 0.96) and for 1-m strip: 10 + 3.7dolomite rate, Mg ha1 (R2=0.96). Leaf yield in June and December 1990 increased linearly as a function of both dolomite rate and method of placement, which interacted (P<0.05) (Fig. 2). When comparing the slopes of equations for June vs. December leaf yield (Fig. 2), leaf responses were similar between June and December (within a placement method), but plants did not regrow as much, especially at the zero or low rates of dolomite. It seems that leucaena required 6 to 8 Mg ha1 of dolomite and 3 m-strip placement on this soil in order to produce December regrowth amounts similar to June yields. Leucaena stem yield in December 1990 (stems were not harvested in June) depended on the interaction of dolomite rate and method of placement. Stem yield (g plant1 ) in 3-m strip placement was: 2 + 2.3dolomite rate, Mg ha 1 (R2 = 0.80) and 2 + l. ldolomite rate, Mg ha1 (R2= 0.80) in 1-m strip placement plots. Grazeable stem yield was a relatively small percentage of total plant yield. Equations for total yield follow. Increasing dolomite rate, especially with 1-m strip placement, resulted in smaller stem yield in creases compared to leaf yield. 80 60 ~-~J.une ~--~ Placement 3m strip =011 + 7.5 D lf:=0.73 1m strip =011 + 5.6 0 fi!:=o.73 Placement ,3m strip =*-4 + 7.0 0 R2=0_91 ~t~=6-4 + 3.0 D R2= 0.91 .!!! 40 Q. Cl 20 3 4 5 6 Dolomite Mg ha -1 Fig. 2. Leaf dry matter yield of leucaena, K340, grown on a Spodosol treated with five dolomite rates and two methods of dolomite placement in June and December 1990. Footnote: Independent variable (D) in equations is dolomite rate, Mg ha. Leaf:stem ratio of June 1990 plus December 1990-harvested leaves:December-harvested stems was not significantly affected (P>0.05) by dolomite rate or placement (avg= 1.5). Leaf:stem ratio of December-harvested leaves:December-harvested stems was also not significantly affected by treatment (P>0.05) (avg. = 0.38). This later ratio was smaller because stems were not harvested in June and did not have to regrow as did leaves. Leucaena seed and pod yield were affected by the interaction of dolomite rate and method of placement (Fig. 3). Ecotype K340 seems to be a good seed producer at this location, and a considerable portion of total yield of the plant was attributable to seed and pods. Total 1990 dry matter yield (June leaf + December leaf, stem, seed and pod) depended on dolomite rate and method of placement. Dolomite increased total yield (g planr-1): -4 + 45. ldolomite rate, Mg ha 1 (R2 = 0.83) when dolomite was applied to a 3-m strip and -4 + 23.3dolomite rate, Mg ha-1 (R2 = 0.83) when dolomite was applied to a 1-m strip. In every case, yield responses for both methods of placement have been linear. The idea of placing liming materials in strips was to reduce cost and make it easier for ranchers to produce leucaena, but rates greater than 8 Mg ha1 should be investigated on Florida Spodosols to determine at what rate leucaena yield ceases to respond to increases in liming mate rials. Many of the leucaena responses to dolomite rate depended on method of dolomite placement. We suspected that roots were confined to the treated areas, and the larger the area, the greater the root volume possible for the plant. Strontium content of leaves averaged over the two sample dates was 225 mg kg1 when the 3-m strip was dolomite-treated compared to 75 mg kg1 when a 1-m strip was dolomite-treated, 140 Pod Placement m strip= 07 16.7 D 2 0-" I 120 1mslrip=-7+8.9D '100 C 80 Ill ii Cl 60 40 20 Seed ---------, Placement 3m strip= -7 + 11.7 D R2=o.65 1 m strip 'i7 + 4. 7 D P?-= 0 _65 .0 3 4 5 6 Dolomite Mg ha1 0 8 Fig. 3. Seed and seed-pod yield of leucaena, K340, grown on a Spodosol treated with five dolomite rates and two methods of dolomite placement. December 1990. Footnote: Independent variable (D) in equations is dolomite rate, Mg ha.

PAGE 110

PROCEEDINGS, VOLUME 51, 1992 89 both at 8 Mg ha-1 Some roots were leaving the 1-m dolomite-treated strip, but not many as evidenced by differences in plant Sr content. Exchangeable Al in the soil decreased line2rly as soil pH increased (mg kg-1 = 69.02 -13.16 pH, R~ = 0.55). Soil exchangeable Al content was a function of dolomite rate (16.8 -3.8dolomite rate + 0.32dolomite rate2, Mg ha-1 R2 = 0.58). Exchangeable Al in our untreated soil averaged 16.4 mg kg-1 Although this is a relatively low concentration, it represents a high percentage of the cation exchange capacity (CEC). On an Ona fine sand at Ona AREC, soil pH (0 to 15 cm depth) was found to be 4.5 with 27 mg kg-I Al, which represented 34% of CEC saturation (Rechcigl, unpublished data). We believe that at low pH levels (4.2) and Al concentrations (16 mg kg-I) of our soil, roots were restricted to the dolomite-treated areas. Expanding the area treated with dolomite expanded the area where Al was precipitated-out of the soil solution, which occurs at about pH 5.5. Zelazny and Fiskell (1971) concluded that liming Florida soils to pH 5 or 5.5 neutralizes exchangeable Al and reduce toxicity to crops. Treating a strip of soil with dolomite to raise pH and reduce Al toxicity may be a practical method to grow leucaena on native Spodosols in Florida, but a 3-m strip and rates of 8 Mg ha' of dolomite are suggested on soils like Pomona. When one hectare is extended in a 3-m strip, the strip is 3.3 km along a road or fire guard. Since leucaena is difficult to estab lish under the best conditions, one problem with this approach may be seedling survival due to grazing by wildlife, especially when leucaena is extended over a long narrow area. REFERENCES Ahmad, N., and F. S. P. Ng. 1981. Growth of Leucaena leucocephala in relation to soil pH nutrient levels and rhizobium concentration. Leucaena Res. Rep. 2:5-7. Breland, H. L. 1964. Effect of limestone and fertilizer application on certain soil properties. Soil Crop Sci. Soc. Florida Proc. 24:42-52. Breland, H. L., and S. L. Locascio. 1962. The effect of dolomite and phosphorus application on soil fertility measurements. Soil Crop Sci. Soc. Florida Proc. 22:60-68. Fiskell, J. G. A., S. L. Locascio, H. L. Breland, and T. L. Yuan. 1964. Effects of soil acidity and liming of Leon fine sand on the exchange properties and on watermelons as indicator plants. Soil Crop Sci. Soc. Florida Proc. 24:52-62. Gallaher, R. N., C. 0. Weldon, and J. G. Futral. 1975. An aluminum block digester for plant and soil analysis. Soil Sci. Soc. Am. Proc. 39:803-806. Gray, S. G. 1968. A review of research on Leucaena leucocephala. Trop. Grassl. 2:19-30. Hambleton, L. G. 1977. Semiautomated method for simultaneous determination of phosphorus, calcium and crude protein in animal feeds. J. Assoc. Agric. Chem. 60:845-852. Hanlon, E. A., and]. M. De Vore. 1989. IFAS extension soil testing laboratory procedures and training manual. Florida Agric. Exp. Stn. Circ. 812 (in press). Henderson, J. R. 1960. Results of research on soil testing and their use as guides to liming and fertilization for production of field crops. Soil Crop Sci. Soc. Florida Proc. 20:382-392. Hu, T. W., and W. E. Cheng. 1980. Effect of liming on the growth and nutrient levels of leucaena. Lcucaena News!. 1:31-32. Jayman, T. C. Z., and S. Sivasubramanium. 1974. The use of ascor bic acid to eliminate interference from iron in the aluminon method for determining aluminum in plant and soil extracts. Analyst 99:296-301. Koffa, S. N., and T. Mori. 1987. Effects of pH and aluminum toxicity on the growth of four strains of Leucaena leucocephala (Lam.) de Wit. Leucaena News!. 8:58-62. Murphy, J., and J. P. Riley. 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chem. Acta. 27:31-36. Oaks, A. J. 1968. Leuwena leucoaf1hala -description, culture, utili zation. Advancing Frontiers of Plant Sci. (India). 20: 1-114. O'Donnell, J. J., J. E. Rechcigl. W. D. Pitman, and D. M. Sylvia. 1991. Establishment and growth of Vigna parkeri on an acid Florida Spodosol in response to lime and phosphorus. p. 491500. In R. J. Wright et al. (eds.). Plant-soil interactions at low pH. Kluwer Academic Pub. Netherlands. Othman, A. B., M. A. Soto, G. M. Prine, and W.R. Ocumpaugh. 1985. Forage productivity of leucaena in the humid subtropics. Soil Crop Sci. Soc. Florida Proc. 44:118-122. SAS Institute. 1985. SAS/STAT guide for personal computers. Version 6 Edition. SAS Institute, Inc. Cary, N.C. Zelazny, L. W., and .J. G. A. Fiske!!. 1971. Acidic properties of some Florida soils II. Exchangable and titratable acidity. Soil Crop. Sci. Soc. Florida Proc. 31:149-154.

PAGE 111

90 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Reliability of Methods for Assessing Leaf Blotch Diseases of Wheat J. Cybulska-Augustyniak, F. M. Shokes*, R. D. Barnett, and A. R. Soffes ABSTRACT Inter-rater and test-retest reliabilities were evaluated for three methods of assessing leaf blotch on 'Coker 9766' wheat (Triticum aestivum L.). The methods included a single visual assessment of plots based on plant appearance, an individual leaf assessment on 12 plants in each plot, and a reflectance-based assessment using a multispectral radiometer. Assessments were made on forty plots with 10 different predetermined disease levels. Three raters were used so that inter-rater reliability could be determined. Raters were trained before each test assessment. Actual diseased leaves were used for disease recognition and a computer simulator was used to teach differentiation of disease severities. Two separate test-retests were performed to compare the reliabilities. Individual leaf assessment was the most reliable of the two visual methods in both inter-rater and test-retest reliabilities. Inter-rater reliability coefficients were 0.86-0.91 for flag leaf as sessments, 0.050-0.055 for whole plot assessments and 0.85-0.90 for reflectance-based assessments. Great improve.nent in test-retest reliabilities for the whole plot assessment occurred for the second test-retest and variability decreased. Coefficients of variation were low for the reflectance-based assessments (3. 7-5.2), intermediate for the individual leaf assessments (5.8-13.8), and highest for the whole plot assessments (17.9-30.4). Cost of the methods in time per plot was least for whole plot (8-9 s plot'), intermediate for reflectance-based assessments (42-48 s plot') and greatest for individual leaf assessments (42-91 s plot 1). The best visual method of assessing leaf blotch under the conditions of this study was the individual leaf assessment. Reliability of this method was comparable to the reflectance-based assessments but applicable over a wider range of conditions. Disease assessment is often necessary when comparing varietal performance, evaluating crosses, selecting breeding lines for further development, or just comparing disease control treatments. Accurate disease assessment is needed for calculation of the yield losses caused by disease Qames, 1971). Criteria for a good disease assessment method have been listed by Shokes et al. ( 1987) as: 1) an assessment method should be easy to use so that training can be done with rapid development of skill in its application; 2) it should allow a rapid estimation of disease intensity; 3) it should be applicable over a wide range of conditions; 4) it should provide an accurate measure of disease; 5) it should provide reproducible re sults. Often assessment methods developed by researchers on specific cultivars or germplasms in one environment are adopted by researchers in other areas of the world without any tests for accuracy or repeatability. It has been suggested that inter-rater and test-re test reliabilities should be determined to evaluate disease assessment methods (Shokes et al., 1987). Interrater reliability is a measure of the consistency between the assessors. Test-retest reliability is an estimate of the correlation between assessments by the J. Cybulska-Augustyniak, Dep. Genetics and Plant Breeding, Poznan Agricultural Univ., Poznan, Poland, 60-625; F. M. Shokes, R. D. Barnett, and A. R. Soffes, North Florida Res. and Educ. Center, Rt. 3 Box 4370, Quincy, Florida 32351. Florida Agricultural Experiment Stations Journal Series no. N00603. *Corresponding author. Contribution published in Soil Crofi Sci. Soc. Florida Proc. 51:90-95 (1992) same rater at different times. These two reliability measurements can be used to determine how well a method adheres to the criteria for a good disease assessment method. Many different assessment methods have been used with small grains and James ( 1971) has developed one of the most useful keys for these crops. Visual assessment is a fundamental aspect of many phytopathological research projects. The percentage of diseased area has been widely used and is actually a measurement of disease severity. Visual estimates of leaf blotch severity can be made in the field using keys to compare diseased plant leaves to standard diagrams Qames, 1971; Zadoks and Schein, 1979). Computer simulation programs that give training in dis ease assessment on small grains have also been developed. One program, DISTRAIN (Tomerlin, 1988), has been widely disseminated in the USA and has been used in our plant pathology research program as a training aid. Whole plot assessment is actually an evaluation of plant appearance in which disease incidence (the proportion of plants with disease symptoms) is a major component. Disease severity is also a component of whole plot assessment since leaves with high severity will greatly affect overall plot appearance. That effect will be much greater, however, if the incidence is high. Obviously other factors which affect plant appearance will also be a component of this assessment. Ahlrichs and Bauer ( 1982) and Nutter and Cun fer ( 1988) have shown that there is a strong linear relationship between green leaf area (GLA) and percent reflection of sunlight from plant canopies. Nutter ( 1989) used the measurement of percent reflectance with the multispectral radiometer as an o~jec tive method for estimating GLA to detect and quantify leaf disease in peanut canopies. Leaf blotch diseases commonly occur on wheat in north Florida. There are many ways to evaluate the severity of these diseases but the best methods have not been thoroughly investigated under north Florida conditions. 'fhe two major leaf blotch diseases in this area are spot blotch caused by Bipolaris sorokiniana (Sacc. in Sorok.) Shoem. (syns. Helminthos porium sativum P. K. & B., H. sorokinianum Sacc. ex Sorok) and septoria nodorum blotch caused by Leptos pharia nodorum E. Muller (anamorph Septoria nodorum Berk). A study was conducted in 1990-9 l to compare the inter-rater and test-retest reliabilities of three methods of assessing leaf blotch on wheat grown in north Florida. MATERIALS AND METHODS The wheat cultivar 'Coker 9766' was planted in plots 6.1-m long by 1.8-m wide, at the North Florida Research and Education Center, Quincy, Florida on 6 Dec 1990. Sixteen different fungicide treatments were applied in a randomized complete block design

PAGE 112

PROCEEDINGS, VOUJME 51, 1992 91 with four replications. Ten treatments were then chosen for use in the present study based on different disease levels according to an initial evaluation of leaf blotch disease severity on 26 Apr 1991. The initial evaluation was performed by an experienced plant pathologist using the visual assessment method of James (1971). The three different methods evaluated for assessing leaf blotch included two visual methods and a remote sensing method. One visual method was a whole plot assessment in which disease severity (percent disease) was assessed on the flag leaf and the F-1 leaf. The second visual method was an assessment of the percentage of leaf blotch on 12 individually marked flag leaves and 12 F-1 leaves randomly selected in each plot. All raters used the same leaves for evaluation. For reflectance-based assessments, the 800 nm band of the Cropscan Multispectral Radiometer (CROPSCAN Inc., Fargo, N.D., 58102)' was used to detect foliar disease gradients in the same three separate locations within each plot. Locations for radiometer assessments were uniformly spaced so that each measurement included four of the twelve plants assessed visually by the other methods. Although the original intent of this study was to evaluate septoria nodorum blotch, laboratory examination indicated that most of the leaf blotch present was caused by B. sorokiniana. Leaves and glumes were examined in the laboratory and incubated in moist chambers at room temperature to induce sporulation. Since symptoms of the two leaf blotches were similar and both organisms could sometimes be isolated from the same spot, the diseases were difficult to differentiate in the spring of 1991. The two diseases were considered as one for purposes of disease assessment in this study and were rated simply as leaf blotch. Three raters were used to test the reliability of the assessment systems. They were trained prior to each test to recognize leaf blotch disease using actual leaves removed from diseased plants. Raters were then trained to determine the percentage of leaf blotch using the computer program DISTRAIN (Tomerlin, 1988). This program permits raters to check their accuracy and work until they improve. During the training with the computer simulator each rater evaluated at least 30 leaves each, with low, medium, and high disease severities followed by 30 leaves with random severities. Raters were also trained by the resident plant pathologist to make whole plot assessments in the field by carefully, but quickly, scanning the plots while observing the percentage of leaf blotch on the leaves. Raters were trained in use of the radiometer in the field before the first test. No further training on the use of this device was considered necessary. Disease assessments were made four times in the field. Plots were rated independently by each rater for two tests and two retests for the measurement of 1Use of trade names in this publication does not imply endorsement by the Institute of Food and Agricultural Sciences, University of Florida, of the products named. test-retest reliabilities. The first test was assessed on 29 April with a retest on 30 April. The second test was conducted on the morning of 2 May and the re test was performed that afternoon. On the second test-retest the F-1 leaves were not assessed. The dis ease had advanced rapidly in plots with severe leaf blotch and many of the F-1 leaves had senesced. Inter-rater and test-retest reliabilities were estimated using the methods of Shakes et al. (1987). Re liability is determined to be the ratio of true variance to the total variance, therefore, if: rating = (mean for all plots) + (effect due to plots) + (effect due to rater) + (error), in which: 2 ai have a mean 0, variance pT 2 f3j have a mean 0, variance aj 2 Eij have a mean 0, variance aE then: 2 2 2 2 Inter-rater reliability is computed as p=
PAGE 113

92 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA that time of the year. Disease levels by the second test-retest were exacerbated by the wet conditions. High rainfall may have also been a factor in the predominance of spot blotch over septoria nodorum blotch. However, weather conditions were conducive to the development of both pathogens. Less than ideal weather and growing conditions may affect, but should not preclude an assessment method. One of the criteria for a good assessment system is that it must be useful over a wide range of conditions. Significant differences in estimation of disease severity were noted among both plots and raters for all three assessment methods (Table 1). Reliability coefficients estimated from the two-way ANOV A are reported in Table 2. Inter-rater reliabilities for the visual methods varied and the 12 individual leaf as sessments were always better than whole plot ratings. Inter-rater reliability coefficients close to 1.0 indicate that raters made similar estimates of diseases using the method on the same plots. This was the case with both the flag and F-1 leaf assessments using the individual leaf assessment. Inter-rater reliabilities considerably < 1. 0 indicates disagreement between raters on their estimate of disease with a method. This was consistently the case with the whole plot assessment in which inter-rater reliabilities were low, ranging from 0.50 to 0.56. The radiometric measurements Table 1. F value significance levels from the analyses of variance for three methods of assessing of leaf blotch on wheat (Triticum aestivum L.). Methodst Leaves of Whole plot 12 plants Flag Flag Source df leaf F-1 leaf F-1 Reflectance Plots 39 ** ** ** ** ** Test 1 Raters 2 ** ** ** ** Error 78 Plots 39 ** ** ** ** ** Retest 1 Rater 2 ** ** ** ** ** Error 78 Plots 39 ** ** ** Test2 Raters 2 ** ** ** Error 78 Plots 39 ** ** ** Retest 2 Rates 2 ** ** ** Error 78 .Significant at the 0.05 and 0.01 levels of probability, respec tively. tMethods are 1) assessment of whole plots based on leaf appearance, 2) percent disease severity on 12 individual leaves, and 3) reflectance-based assessments using the 800 nm band of a multispectral radiometer. *The first test was performed on 29 April with a retest on 30 April. The second test was on the morning of 2 May with the retest in the afternoon. F-1 leaves were not assessed on the second test-retest due to senescence of many leaves with the highest disease severities. Table 2. Inter-rater reliabilities for three methods of assessing leaf blotch, on wheat caused by Septoria nodorum Berk. and Bipolaris sorolciniana (Sacc. in Sorok.) Shoem., at the North Florida Research and Education Center, Quincy, Florida, in 1991. test Methodt Test 1 Retest 1 Test2 Retest2 -----------------r value ----------------Whole plot Flag leaf 0.50 0.54 0.56 0.55 F-1 0.52 0.50 12 ind. plants Flag leaf 0.87 0.86 0.90 0.91 F-1 0.95 0.91 Reflectance 0.90 0.89 0.85 0.89 tMethods are 1) assessment of whole plots based on leaf appearance, 2) percent disease severity on 12 individual leaves, and 3) reflectance-based assessments using the 800 nm band of a multispectral radiometer. *The first test was performed on 29 April with a retest on 30 April. The second test was on the morning of 2 May with the retest in the afternoon. F-1 leaves were not assessed on the second test-retest due to senescence of many leaves with the highest disease severities. and visual methods of estimating disease severity on 12 individual flag leaves had similar inter-rater relia bility coefficients. Slight improvement in inter-rater reliabilities was noted between the first test-retest and the second test-retest for whole plot evaluation of dis ease severity on flag leaves and assessment of the percentage of leaf blotch of 12 individual flag leaves (Table 2). There was little difference in inter-rater reliabilities between the flag and F-1 leaves for either the whole plot or the individual leaf assessments. In our study, the higher reliability coefficients obtained for the individual leaf assessments indicate that the precision of disease severity measurements was better when evaluating the percentage of diseased area of randomly selected individual leaves rather than the whole plot assessment (Table 2). James (1971) reported that the assessment of disease on individual leaves is an effective method and the error attached to the observation is small because the observer is assessing one disease on one leaf at one time. Our results would certainly support that conclu sion. Averaging several measurements also decreases the error variance and hence increases reliability (Shakes et al., 1987). Inter-rater reliabilities for whole plot assessments were poor. This may have been be cause variability within plots was great and the raters were unable to accurately assess the level of infection on a whole plot basis. Lack of time sometimes precludes assessing individual leaves, but for some dis eases individual plants must be examined closely. Our results showed that when visual assessments are required, leaf blotch in wheat should be assessed using individual leaves. The correlation coefficients between the tests and retests (test-retest reliabilities) are listed in Table 3. The whole plot assessment improved markedly for all three raters by the second test-retest. For the indi vidual leaf assessment, a fairly high level of consis-

PAGE 114

PROCEEDINGS, VOLUME 51, 1992 93 Table 3. Test-retest reliabilities of three trained raters assessing leaf blotch by three rating methods.t First test-retest Second test-retest Method RI, R2 R3 RI R2 R3 ----------------------r value --------------------Whole plot Flag leaf 0.78 0.66 0.64 0.90 0.87 0.87 F-1 0.64 0.61 0.56 12 ind. plants Flag leaf 0.88 0.94 0.90 0.97 0.98 0.99 F-1 0.94 0.95 0.90 Reflectance 0.92 0.92 0.89 0.92 0.93 0.98 tReliabilities listed are coefficients of correlation between the values obtained for each test compared to each retest. 'The first test was performed on 29 April with a retest on 30 April. The second test was on the morning of2 May with the retest in the afternoon. Methods arc I) assessment of whole plots based on leaf appearance, 2) percent disease severity on 12 individual leaves, 3) reflectance -based assessments using the 800 nm band of a multispcctral radiometer. ~The three raters are designated as Rl-R3. tency was noted for all raters. Correlation coefficients obtained from radiometric measurements were very consistent and similar to the test-retest reliabilities for the rating of 12 individual leaves. For whole plot assessment the data indicates that rater number 1 was more consistent than two other raters in the test-retest reliability. Rater number I was the only assessor with some prior experience rating foliar diseases of small grains. Experience can be a factor in accuracy and precision of disease assessment since improvement occurs with practice. Some raters have a greater ability to differentiate between disease severities than others. Similar conclusions are reported by Shokes et al. (1987) in the study ofleafspot in peanut. Our results are very consistent with the observations of these authors. Correlation of percent reflectance values and percentage of flag leaf blotch severity with the two visual methods were significant (P::s0.01) for all days of assessment (Table 4). Correlation coefficients were generally high and negative because of the inverse relationship between reflectance and foliar disease. These results show a good relationship between the visual assessments and the radiometric method (Table 4) and are consistent with the observations of Nutter and Cunfer (1988). They reported that dis ease gradients in barley can be determined by measuring the percentage of canopy reflectance. The low coefficients of variation for the reflectance measurements are a good indicator of its precision in making repeated measurements (Table 5). With disease assessment as with many types of field data any time a coefficient of variation is <20 it is generally considered acceptable. From the coefficients of variation (Table 5) it is evident that the raters also obtained a fairly high level of consistency using the individual leaf rating even on the first day. These coefficients of variation were less than half of that obtained with the whole plot assessment. This may have been an indicator of the effectiveness of the method and the prior training. Tomerlin (1988) reported that the major advantage of the program, DISTRAIN, is that it allows the user to determine how close the estimated severity is to the actual severity. A person can practice until an acceptable level of skill is obtained. It is difficult to train for making whole plot assessments. One possible method would be to assess individual leaves in a plot, then assess the whole plot with immediate comparison of results. This is a time-consuming training exercise and was not implemented for these tests. Such a training exercise also assumes that the individual leaf assessment is accurate; a reasonable assumption if proper care has been taken in performance of the assessment. Another factor that has to be considered with whole-plot assessment is that it may be biased by general plot appearance. That is, it is really a plot appearance score and plot appearance may be affected by factors other than dis ease. Any variations in soil conditions such as nutrient levels, hard pans, gradients in soil texture, etc. may Table 5. Coefficients of variation for each method of assessment. Method Test t' Retest 1 Test2 Retest 2 Whole plot 30.4 26.0 I 7.9 19.4 Individual leaf 13.8 9.2 5.8 6.8 Reflectance 3.7 4.1 5.2 4.7 tThe first test was performed on 29 April with a retest on 30 April. The second test was on the morning of 2 May with the retest in the afternoon. *Values given for reflectance, test I, arc based on the assessments of raters 1 and 2 only. due to a temporary problem with the radiometer for the third rater. Table 4. Correlations (P<0.01) between visual and reflectance-based assessments for three raters in two test-retest periods. Raters Rater I Rater 2 Rater 3+ Individual leaf assessments Whole plot assessments Test It Retest I Test 2 Retest 2 Test I Retest I Test 2 Retest 2 ------------------------------------------------------------------r value -------------------------------------------------------------------0.79 -0.58 -0.69 -0.74 -0.67 -0.53 -0.55 -0.55 -0.71 -0.fi3 -0.63 -0.65 -0.66 -0.74 -0.72 -0.68 -0.64 -0.64 -0.63 -0.48 -0.70 -0.80 -0.7'\ tThe first test was performed on 29 April with a retest on 30 April. The second test was on the morning of 2 May with the retest in the afternoon. *Data unavailable for reflectance-based assessment for third rater, first test, due to a temporary problem with the radiometer.

PAGE 115

94 SOIL AND CROP SCIEl\"CE SOCIETY OF FLORIDA have considerable effect on plant appearance. By the last assessment heavy rains had caused some lodging in a few plots which made whole plot assessment extremely difficult and may have affected the ratings. The percentage reflectance readings using the multispectral radiometer were made by three raters at different times during the day. There were differences of light reflected from wheat canopies in the measurements. The radiometer is designed to compensate for these differences. It has sensors 'looking up' to sense the incoming radiation for comparison to the incident radiation reflected from the plant canopy. The comparator circuitry adjusts the resulting output to compensate for changes in incoming radiation with time. However, readings should ideally be made between 1100 and 1400 h, the time when the sun is nearly directly overhead. The radiometer gives best results when light levels are consistent. It seems to work very well for example, when light is diffused by a complete cloud cover. Greater variation in measurements are noted when conditions are partly cloudy with frequent shadows cast by cloud movement. Inter-rater reliabilities were good for the radiometer (2:0.85) as were the test-retest reliabilities (2:0.89). Consistency is a major advantage of radiometric measurements. The correlations between the reflectance-based assessments and the visual assessments were lower than expected (Table 4). The visual assessments allow raters to compensate for undesirable factors such as lodging particularly with the individual leaf assessment. The radiometer was consistent in its measurements but was probably affected more than the other methods on the plots in which lodging had occurred. There is no way to compensate adequately for such unforeseen variables using remote sensing. Notes on crop conditions in relation to other variables must be kept in addition. The correlation of disease severity estimated using the three assessment methods for grain yield, test weight and TKW are given in Table 6. Yield was significantly correlated (P
PAGE 116

PROCEEDINGS, VOLUME 51, 1992 95 Septoria in 1991 with the extreme amounts of precipitation and mild temperatures. Our results show strong relationship between kernel weight and all the methods of assessment indicating that disease was certainly a major determinant of kernel weight. The amount of heaci blight due to infection of individual glumes and seed was probably more of a factor in determining yield in this study than was foliar dis ease. Time required for each plot assessment varied for all methods. The raters only used 8 to 9 s plor-1 to evaluate disease level in the whole plot assessment. Times required to assess disease severity on 12 individual leaves ranged from 69 to 91 s plot-1 for the first test-retest but decreased to a range of 42 to 48 s plot-1 at the second test-retest. Reflectance-based assessments required 36 to 48 s plot1 and had only a nominal decrease of< 10 s plot-1 by the second test-re test. However more time was required for movement from plot to plot with the radiometer than with the other methods since the instrument must be carried. It is awkward to carry since it is mounted on an aluminum pole 2.4 min length. CONCLUSIONS Results of this study indicate that raters should definitelv be trained prior to making assessments of leaf blotch diseases on wheat. Even experienced raters gain from retraining which serves to recalibrate the eye with a given rating scale. This study further supports the idea that some raters are better than others at detecting differences in disease severities and/or incidence, and that raters can improve with practice and experience. This is true for foliar dis eases of wheat as well as other crops. It was found that the individual leaf assessment and the radiometer measurements had similar inter-rater and test-retest reliabilities. The cost of these methods in time were considerably greater than that for the whole plot assessment but the greater precision gained would indicate that the cost was worthwhile. If very fast assessment of a great number of plots is needed and high accuracy and precision are not major considerations, the single whole plant estimate could be used. Overall the best method of estimating leaf blotch in our study was the individual leaf assessment. This method meets the criteria for good disease assessment. Although it costs slightlv more than the reflectance-based assessments in time, it is applicable over a wider range of conditions. ACKNOWLEDGEMENTS The authors would like to thank Dr. D. A. Zahn and the students in the Statistical Consulting Department at Florida State University for their helpful advice on the anal vsis of the data collected in this study. REFERENCES Ahlrichs, J. S., and \I. E. Bauer. 1982. Relation of agronomic and multispectral rt>flectance characteristics of spring wheat canopies. Tech. Report S R-12-04384, Laboratory for Applica tions of Remote Sensing, Purdue University, West Lafayette, II\. 26 pp. James, C. 1971. A manual of assessment keys for plant disease. Can. Dep. Agric. Pub!. 1458. Nutter, F. W., Jr. 1989. Detection and measurement of plant disease gradients in peanut with a multispectral radiometer. Phytopathology 79:958-963. Nutter, F. W., Jr., and B. M. Cunfer. 1988. Quantification of barley yield losses caused by Rhyndwsj1orium secalis using visual versus remote sensing assessment methods. Phytopathology 78: 1530. Shokes, F. M., R. D. Berger, D. H. Smith, and J.M. Rasp. 198i. Reliability of disease assessment procedures: A case study with late leafspot of peanut. Oleagineux 42:245-251. Tomerlin, J. R. 1988. Distrain: computer program for training people to estimate disease severity on cereal leaves. Plant Dis. 72:455-459. van den Bosch, R., J. C. Zadoks, and J. A. Metz. 1988. Focus expansion in plant disease: The constant rate of focus expansion. Phytopathology i8:54-58. Wagoner, P. E., and R. D. Berger. 1 \)87. Defoliation, disease, and growth. Phytopathology 77:393-398. Zadoks, J. C., and R. D. Schein. 1979. Epidemiology and plant disease management. Oxford University Press, New York. 247pp. Zillinsky, F. J. I 983. Common Diseases of Small Grain Cereals: A Guide to Identification. Centro International de Mejoramiento de Maiz y Trigo (CIMMYT). 141 pp.

PAGE 117

96 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Mycoflora of Seed from Advanced Wheat Breeding Lines in North Florida Z. Weber, D. A. Berger, F. M. Shokes*, R. D. Barnett, and D. L. Wright ABSTRACT Six co~~ercial c.ultivars and 54 advanced breeding lines of wheat (Tnt1cum aest1vum L.) were grown at two test sites in north Florida and examined for seed pathogens after harvest in 1991. The two test sites were the North Florida Research and Education Centers at Marianna, and Quincy, Florida. Rainfall was I 74 cm at the Quincy site and 108 cm at the Marianna site. Seed were surface disinfected and placed on a selective medium for septoria isolation. The medium allowed growth of a number of other fungi: A total of 30 seed from each genotype from each site were exam1_n~d. The pr~dominant fungi isolated were Bipolaris sorokm1ana (Sacc. m Sorok.) Shoem. and Alternaria alternata (Fr::Fr.) Keissi. Bipolaris was found in all 60 genotypes from Qumcy and 53 genotypes from the Marianna site. Alternaria was present in seed from all genotypes at both sites. Septoria nodorum Berk. was isolated with low frequency from seed of most of the lines tested, however, Septoria was not detected in seed of el~ven genotypes from both sites. Three Fusarium spp., F. equ1setJ. (Corda) Sacc., F. lateritium Nees:Fr., and F. tricinctum (Corda) Sacc., were isolated from wheat seed from both environments. This is the first time these species of Fusarium have been r~porte~ from wheat_ ii_i Florida. Two breeding lines were identified with no Septona mfection and only low levels of Fusarium and Bipolaris. There is potential for head resistance in some of the breeding lines tested since conditions were conducive to infection by the mycoflora at both sites. Over 100 different fungal species may be isolated from newly harvested wheat grain including Fusarium SPJ:>, Alternari~ spp., Helminthosporium spp., and Septorza spp. (Weise, 1987). Infection by these fungi can be a problem in humid field environments when rela tive humidity exceeds 90% and seed moisture is >20% (Weise, 1987). Infected seed may have various symptoms depending upon which pathogen(s) is present and grain may be blackened, discolored, s?runken, or ~hri~eled. Grain quality, seed germination, a~d gr~m yi~ld may be adversely affected by fungal mfectlon of seed heads. For example, yields may be decreased as much as 10 to 20% if high levels of infection by S. nodorum Berk. (teleomorph, Leptosp_hae1:a nodorum E. Muller), occur. Even if germinatioi:i is n~t aff:ected, the seed infection problem is still senous smce moculum may be carried over and contribute to a severe epidemic when that seed is used to produc~ a crop. Seed infection is a major contributing factor m severe outbreaks of Septoria nodorum blotch (Eyal et al., 1987). Septoria nodorum blotch (often called glume blotch) commonly occurs on small grains in Florida (Kucharek, 1988). Spot blotch cal:1sed by B. sorokiniana (syns. Helminthosporium satzvum P. K. & B., H. sorokinianum Sacc. ex Sorok.) is also commonly found on wheat in Florida. When it attacks the heads of wheat, this fungus causes shrivel-Z. Weber, Plant Pathology Dep., Poznan Agricultural Univ., Poznan, Poland; D. A. Berger, F. M. Shokes, R. D. Barnett, and D. L. Wright, North Florida Res. and Educ. Center, Rt. Box 4370, Quincy, Florida. 32351. Florida Agric. Exp. Stn .Journal Series no. N-00590. *Corresponding author Contribution published in Soil Crop Sci. Soc. Florida Proc. 5 I :96-99 ( 1992) in~ ~f the grain and may cause black point, a dark stammg of the embryo end of the seed (Zellinsky 1983). Nur:ierous soft r~d winter wheat genotypes are grown m north Flonda. Weather conditions in this area during the wowing seaso~ (November -June) are often conducive to head mfection by various fungi. Differences in number of seed of different w~eat breedi?g _lin~s infected by fungal pathogens m1g~t. ~e an m?ication of resistance. This is a good possibility, particularly when conditions are favorable for disease and the probability of disease escape is smal~. A ~tudy of 60 wheat genotypes, grown at two locations m north Florida, was undertaken in 1991 to determine the predominant fungal pathogens on seed. A second objective of this study was to ascertain whether ar~y of _the lines. mi_ght_ have significantly lower seed mfect1on as an mdicat1on of resistance. MATERIALS AND METHODS Wheat seed used in the experiment were from yield tests of six commercial cultivars and 54 advanced breeding lines grown in 1990-1991 at two lo cations, the North Florida Research and Education Centers at Quincy and Marianna. The commercial cultivars were 'Florida 301', 'Florida 302', 'Florida 303', 'Florida 301H', 'Traveler', and 'ATW270'. A complete list of all the breeding lines can be obtained from the third author. Wheat at the Quincy site was ~?'.)wn on a Dothan loamy, fine sand (fine, loamy, siliceous, thermic, Plinthic Kandiudults) and at the Marianna site on a Chipola loamy sand (loamy, siliceous, thermic, Arenic Hapludults). Wheat had not been planted in the field at the Quincy site in the past two years wheat had not been grown at the Mananna site m the past three years. Previous summer-winter rotations at the Quincy site were fallowwheat, soybean (Glycine max L.)-fallow, soybean wheat. Rotations at the Marianna site were corn (Zea mays L.)-fallow, corn-fallow, peanut (Arachis hypogaea L.)-wheat. Seed were planted at the Quincy site on 12 Dec 1990 and harvested on 3 June 1991. The Marianna test was planted on 28 Nov 1990 and harvested ~n 29 May 1991. ~oth sites were fertilized preplant with 560 kg ha' of 5-10-15 N-P-K. The Quincy site was top-dressed with 56 kg ha' 28-0-0-5 N-P-K plus sulfur on 4 Feb. 1991 and the same at the Marianna site on 12 Feb. 1991. Weeds were controlled with herbicides according to extension recommendations (Colvin and Brecke, 1991). Seeds were mechanically harvested with a small plot combine, dried to 12% moisture, stored in paper bags at l5C, and randomly sampled for the mycoflora study. Due to the large numbers of entries, the laboratory as_say of seed was conducted with three replicates over time. Seed were placed on agar plates in the laboratory on 24 July, 5 Aug., and 22 Aug., 1991. Ten seed from each of the genotype-location combi-

PAGE 118

PROCEEDINGS, VOLUME 51, 1992 97 nations were surface disinfested for 1.5 to 2 min in 2.6% sodium hypochlorite and plated on the selective medium for S. nodorum, SNAW (Manandhar and Cunfer, 1991), on sterile disposable petri dishes. Seed were incubated under near ultraviolet light with a 12 hr photoperiocl, at 20 to 24C. On the fourth day, wheat seed without microbial colonies or with small colonies were transferred to new petri dishes with SN AW medium. This was done to prevent the overgrowth of slow growing fungi by fast growing col onies (Shokes et al., 1992). Fast growing fungi from the seed were evaluated on the 8th through the 12th day of incubation, identified, and counted. Incubation of the slow growing fungi was continued until the colonies could be identified. A total of 30 seed of each genotype were examined from each site. Infection levels were determined as the percentage of seed of a given genotype infected by a particular pathogen. Analysis of variance was performed on the re sults obtained using a randomized complete block split-plot design. The agricultural statistics program MST A TC (Department of Statistics, Michigan State University, East Lansing, MI) was used for all analyses. RESULTS AND DISCUSSION Temperatures were very similar at both sites, but the Marianna site had slightly higher maximum and lower minimum temperatures than the Quincy site (Fig. 1). Monthly precipitation was generally higher at the Quincy site than at the Marianna site; seasonlong precipitation was 174 cm at the Quincy site, about double the seasonal average. Precipitation at the Marianna site was 108 cm, about 1.3 times the seasonal average. Weather at both sites was conducive for diseases of the foliage and heads throughout the growmg season. With the rigorous surface disinfestation given to seed before plating it is unlikely that any of the fungi isolated were surface contaminants. The fungi isolated from seed (Table 1) included A. alternata, B. Rainfall (cm) 100 Temperature (C) 90 -Q Rainfall M Rainfall Q Max. Temp. 80 40 30 20 10 0 Q Min. Temp. M Max. Temp.-,-M Min Temp. 11_ Nov. Dec. Jan. Feb. Mar. Apr. May Month 40 35 30 25 20 15 10 0 Fig. 1. Weather conditions during the wheat growing season in 1990-1991 at two sites, the North Florida Research and Educa tion Centers, Marianna (M), and Quincy (Q). Table I. Results of tests of 60 wheat genotypes for seed infection by various mycoflora. Fungal species Alternaria alternata Bipolaris sorokiniana Epicoccum sp. Fusarium spp. Septoria nodorum Other Fungi ( ;enotypes Mean seed Sites t Inf cctcd Not infected infected ----------No. --------------% ----Q 60 0 ,lO.O M 60 0 66.0 Q 60 0 56.0 M 53 7 11.0 Q 59 1 11.0 M 58 2 12.0 Q 5'.l 8 7.0 \! ti() 0 14.0 Q 39 21 05 M 36 24 4.0 Q 10 50 0.5 M 12 48 1.0 i'Genotypes were grown at two sites; the North Florida Research and Education Centers, Quincy, (Q(, ancl Marianna, (M), Florida, in 1990-1991. sorokiniana, Epicoccum spp. F. equiseti, F. lateritium, F. tricinctum, and S. nodorum. Alternaria and Epicoccum are considered to be saprophytic or just weakly parasitic and contribute to the sooty head molds often found on wheat (Weise, 1987). This could also be the case for the Fusarium spp. isolated in this study, however, Fusarium is a common cause of scab or head blight of wheat resulting in shrunken seed (Weise, 1987). These species of Fusarium had not been previ ously reported from wheat seed in Florida (Miller, 1991). The three Fusarium spp. were commonly isolated, sometimes from the same seed and for this study were grouped as Fusarium spp. for all analyses. Fifty-two of the genotypes at the Quincy site had low levels of infection by Fusarium spp. (:S7%), whereas all genotypes at the Marianna site had at least some infection by these fungi (:Sl4%). The frequency of isolation of B. sorokiniana was very high. It was found in seed of all genotypes at the Quincy site at ::; 10% level of infection, while it occurred at very high levels of infection (>50%) in 44 of the 60 genotypes tested at this site. At the Marianna site, infection levels were much lower with 43 genotypes having 13% or less of the seed infected with B. sorokiniana. Only one genotype at the Marianna site had >50% infection with this fungus and all except that one genotype had <33% of the seed infected with Bipolaris. Septoria nodorum was isolated infrequently from seed at both locations. At the Quincy location septoria was detected from 10% or less of the seed of 53 genotypes. Of the seven genotypes from the Quincy site with > 10% infection, the highest level recorded was 23%. Forty-nine genotypes at the Marianna loca tion had septoria detected in :S7% of the seed. Of the eleven genotypes with >7% infection, the highest

PAGE 119

98 SOIL AND CROP SCIENCt: SOCIETY OF FLORIDA Table 2. Infection levels (%) of seed of six commercial cultivars of wheat compared to 11 Septoriafree breeding lines from Marianna, and Quincy, Florida, in the 1990-1991 growing season.t S. nodorum B. sorokiniana Fusarium spp. Genotype Marianna Quincy Marianna Quincy Marianna Quincy ----------------------------------------------------------------% --------------------------------------------------------FL301 FL302 FL303 FL301H Traveler ATW270 FL8 I 72-Gl 16 FL8 l 72-G 160 FL8 l 72-G 162 FL8150-J9-Kl FL85238-G94 FL85363-G 187 FL85238-G65-Abl FL85377-G5-15 FL8338-G48 FL8156-Gl25-63 FL85238-G76 LSD (P:s0.05) 15 6 10 8 5 7 -:j: 11 13 3 12 10 10 13 07 67 13 17 10 67 10 17 27 23 50 7 13 0 17 13 13 10 15 43 13 0.7 83 20 1.0 63 12 0.3 37 03 0.3 40 13 0.0 47 17 0.7 77 33 2.0 57 23 1.3 87 40 1.0 87 03 0.3 63 13 0.3 90 13 0.3 60 7 0.7 63 17 0.3 23 3 0.0 37 17 0.3 70 7 0.3 25 18 1.2 tNumbers represent the mean for 10 seed. Ten seed of each genotype from each test site were plated three times (a total of 30 seeds for each genotype and site). *Seed of these genotypes had no Septoria detected when plated on a selective medium. level detected was only 17%. There were 11 breeding lines across both locations from which no Septoria was obtained. Infection of these breeding lines by Bipolaris and the Fusarium spp. are compared to that of the commercial cultivars in Table 2. Levels of infection of the seed of the cultivars by Septoria ranged from 5 to 15%. Breeding lines FL 8338-G48 and FL 8156-Gl25-63 had very low levels of Fusarium and Bipolaris in addition to no Septoria in both test environments. These lines need to be investigated thoroughly as potential sources of resistance to these seed pathogens. Frequency of isolation of other unidentified fungi was very low. Only 10 genotypes at the Quincy site and 12 at the Marianna site had other fungi detected and incidence of these fungi were low ranging from 0.5 to 1 %. Since these fungi were infrequently found, attempts at identification were not pursued beyond the initial microscopic examination. Differences (P::S0.001) in levels of seed infection were noted between sites (Table 3) for A. alternata, B. sorokiniana, and Fusarium spp. No differences (P>0.05) occurred between the two sites for S. nodorum, Epicoccum sp., or the other unidentified fungi from seed. Differences (P::S0.01) were noted between genotypes for Alternaria, Bipolaris, Fusarium, and Septoria infection of seed. A significant (P::S0.05) site by genotype interaction was observed for both Bipolaris and Septoria infection. Differences (P::S0.001) between genotypes for the number of seed infected with B. sorokiniana indicates that some lines may have resistance. The much higher incidence at the Quincy site was probably due to wetter growing conditions. The shorter rotation (2 vs. 3 yr) may also have been a contributing factor. Since almost all genotypes had Bipolaris detected in the seed (60 genotypes at Quincy and 53 at Marianna), the probability that any genotypes 'escaped' the disease is low. The significant site by genotype interaction (P::S0.001) for Bipolaris may indicate that resistance of some lines breaks down under very severe conditions. The data for Alternaria, Fusarium and Septoria was not as definitive as it was for Bipolaris but does indicate that some genotypes may have resistance. Weather and/or crop rotation may have an influence on the amount of infection by Table 3. Analysis of variance of levels of seed infection of 60 wheat genotypes at two locations in North Florida in 1990-1991. Alternaria Epicoccum Bipolaris Fusarium Source df alternata sp. sorokiniana spp.t Site* l *** n.s. *** *** Genotype 59 ** n.s. *** ** Site X Genotype 59 n.s. n.s. *** n.s. *,**,***Significant at 0.05, 0.0 I, and 0.00 I levels of probability, respectively. tFusarium lateritium, F. equiseti, and F. tricinctum were combined for all data analyses. *Locations of the tests were the North Florida Research and Education Centers, Quincy, and Marianna, Florida. Septoria Other nodorum fungi n.s. n.s. *** n.s. n.s.

PAGE 120

PROCEEDINGS, VOLUME 51, 1992 99 Alternaria and Fusarium spp. since the site effect was highly significant. This was not evident with Septoria in this study. CONCLUSIONS A number of different fungi may infect wheat seed in north Florida, several of which are common pathogens of wheat. The high incidence of B. sorokiniana in our study indicates that it may be the predominant pathogen of wheat seed in this area, particularly when very wet conditions occur in the spring. Since 11 genotypes of wheat had no Septoria infection detected, this warrants further investigation as it may be an indication of resistance. Two breeding lines, FL G8338-G48 and FL 8156-Gl25-63, had no Septoria and very low levels of Fusarium and Bipolaris in both test environments. REFERENCES Colvin, D. L. and B.J. Brecke. 1991. Weed control in small grains 1991. Weeds in the Sunshine. Florida Agric. Exper. Stn. Bull. SS-AGR-7. Eyal, Z., A. L. Scharen, J. M. Prescott, and M. Ginkel. 1987. The Septoria Diseases of Wheat: Concepts and Methods of Disease Management. International Maize and Wheat Improvement Center (CIMMYT), D.F., Mexico. Kucharek, T. A. 1988. Diseases of small grains in North and Central Florida. Plant Path. Fact Sheet PP-38. Florida Coop. Ext. Serv. 6 pp. Manandhar, J.B., and B. M. Cunfer. 1991. An improved selective medium for the assay of Septoria nodorum from wheat seed. Phytopathology 81:771-773. Miller, J. W. 1992. Other detections of special interest. Div. of Plant Industry, Florida Dep. Agric., Tri-Ology, 30:4-5. Shokes, F. M., Z. Weber, and R. D. Barnett. 1991. Fungi isolated from wheat seed on two Septoria nodorum (Berk.) Berk. selective media. Phytopathologia Polonica 13:31-34 Weise, M. V. 1987. Compendium of wheat diseases. 2nd ed. APS Press, St. Paul, MN. Zellinsky, F. J. 1983. Common Diseases of Small Grain Cereals: A Guide to Identification. International Maize and Wheat Improvement Center (CIMMYT), D.F., Mexico.

PAGE 121

PROCEEDINGS, VOLUME 51, 1992 99 Hessian Fly Control in Florida Wheat with Systemic Insecticides J. B. Hartman, R. D. Barnett*, A. R. Soffes, and R. K. Sprenkel ABSTRACT Hessian fly (Mayetiola destructor Say), the most damaging insect pest of wheat (Triticum aestivum L. em Thell) in the southeastern United States, is primarily controlled by the planting of resistant cultivars. However, the small number of adapted resistant cultivars limits management options for southeastern wheat growers. With susceptible cultivars, good fall control is usually obtained with fall application of phorate or disulfoton, but yield losses due to winter and spring generations of the fly may occur. This two year study evaluated the use of an after-planting appli cation of disulfoton for the control of winter and spring genera tions of the fly, and compared the use of insecticides as an alter native to the use of resistant wheat cultivars. Although afterplanting applications were found to reduce fly numbers, chemical treatments were not consistently linked to increases in grain yield. It was concluded that the use of systemic insecticides with Hessian fly susceptible cultivars did not compare favorably with the use of cultivars containing genetic sources of resistance to Hessian fly. Hessian fly causes the greatest damage to wheat of any insect pest in the southeastern United States. Hessian fly larvae feed on plant photosynthates, J. R. Hartman, R. D. Barnett, A. R. Soffcs, ad R. K. Sprenkel, North Florida Research and Education Center, Quincy, FL. 323t5 l. Florida Agric. Exp. Stn. Journal Series no. R-02137. *Corresponding author. Mention of a trademark or proprietary product does not constitute a guarantee or warranty by the U niver sity of Florida and docs not imply approval over other products that may be suitable. The or~! presentation of data associated with this manuscript was judged to have been best of the presentations made in the Graduate Student Paper Contest. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51 :99-102 (1992) stunting plant growth and reducing yield (Lidell and Schuster, 1990). Complete yield losses due to Hessian fly have been observed in Georgia (Buntin and Chapin, 1990). In the midwestern wheat growing region of the United States, Hessian fly has been controlled by using resistant cultivars and planting after the "fly free" date, the initial fall emergence of Hessian fly. In Florida, there is no "fly-free" date, and planting resistant cultivars is the most effective means of Hessian fly control (Buntin et al., 1990). However, very few resistant cultivars appear on the recommended variety list for Florida, and those few that do appear may have yield and quality deficiencies which make them unattractive to growers. Systemic insec ticides have been effectively used to control Hessian fly in wheat (Brown, 1960; Bigger et al., 1965; Buntin and Raymer, 1989; Chapin et al., 1989). In addition to lacking a "fly-free" date, Florida and South Georgia wheat growers may be confronted with six or more Hessian fly generations per wheat growing season Oohnson and Buntin, 1989). This high number of fly generations per wheat generation has been implicated in the relatively short life (as few as three seasons) of Hessian fly resistance genes in these areas (Taylor, 1989, personal communications). Buntin ( 1990) found that at planting in the row granular applications of phorate, disulfoton, ter bufos, and carbofuran were all effective in controlling seasonal infestations of the Hessian fly in Georgia. However, during two years of the three year study, the emergence of a November generation of Hessian

PAGE 122

100 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA fly, 35 days after planting, severely reduced effective ness of an at-planting insecticide treatment. His data also suggested that phorate and terbufos remained effective longer than disulfoton and carbofuran. This study was designed to evaluate the effective ness of an after-planting systemic insecticide application to reduce Hessian fly damage to winter wheat in North Florida. Additional objectives were to evaluate insecticide use with susceptible wheat cultivars as an alternative to the use of resistant cultivars, and to determine the effectiveness of using systemic insec ticides for the control of virulent biotypes of Hessian fly infesting resistant wheat. MATERIALS AND METHODS The study consisted of six wheat cultivars. 'Florida '.HH', 'Florida 302', and 'Florida 303', are all suscepti ble cultivars on the recommended list for Florida. 'Florida 30 I H' and 'Coker 9766' are also on the recommended list and contain the H7H8 duplicate dominant genes for Hessian fly resistance. Florida 301H is a backcross line of Florida 301 and contains the H6 gene in addition to H7H8. The pedigree of Florida 301H is Florida 301 *3/Knox 62. An advanced line, FL85363-Gl8-14, containing the H9 gene was also used. FL85363-G 18-14 is a backcross line of Florida 302, and its pedigree is Florida 302*3/Ella. Insecticide treatments used were: no insecticide was applied to the control (NT); a single application of phorate (0,0-diethyl S-ethylthiomethyl phosphorodithioate) (Thimet 20G, Cyanamid Co., Wayne, New Jersey) at 7.8 kg ha I was made at planting in the row (APT); and the APT treatment plus disulfoton (0,0diethyl S-2-ethylthioethyl phosphorodithioate) (Di Syston SEC, Mobay Corp. Kansas City, Missouri) at 1.2 L ha-1 timed to the nitrogen topdressing in Feeke's plant stage 4 (PS) (Large, 1954). The study was carried out for two years ( 1989-90 and 1990-91) at the Marianna Agricultural Research and Education Center near Greenwood, FL. on Chipola loamy sand (loamy, siliceous, thermic, Arenic Hapludult). A drill strip of Florida 302 was planted around the experimental plot area two weeks prior to planting the test to facilitate Hessian fly development. Wheat was drilled into 1.5-m x 3-m plots. Plots were fertilized with 560 kg per hectare 5-10-15 (N-Pr 05-K20) preplant. Plots were planted 15 Nov. 1989 and for the 1990-91 trials, plots were planted 16 Nov. 1990. A nitrogen topdressing with 56 kg per hectare 28-0-0-5 (N-P205-K20-S) and corresponding applica tion of disulfoton was made 6 Feb. 1990 and in the second year 12 Feb. 1991. Thirty centimeters of row were removed randomly from each plot immediately prior to harvest, on 1 May 1990 and 1991. From this sample, 10 culms were removed, thoroughly examined, and the number of larvae and pupae were determined. Percentage of tillers infested was calculated by dividing the number of culms with one or more Hessian fly larvae and/or pupae by the total number of culms dissected (i.e. 10 x 100 %). Grain yield (kg ha1 ) and test weight (kg hl-1 ) were measured. For analysis, counts of Hessian fly larvae and pupae per 10 culms were transformed to log + 1. Percentage of tillers infested were transformed by arcsin square root of the decimal fraction before analysis. Both transformations and analysis of variance were carried out using SAS procedures (SAS Inst., l 988). Data were analyzed as a split plot with 4 replica tions each year, with cultivars as whole plots and in secticide treatments as sub-plots. Locations of field plots differed and plots were randomized independently each year. Data were analyzed using the analysis of variance procedure, and appropriate means were separated using Duncan's Multiple Range Test (P < 0.05) (Duncan, 1955). RESULTS The main effects of cultivar and treatment affected all variables tested (Table 1). Year also affected grain yield, test weight, and larvae and pupae per 10 culms, but not the percentage of tillers infected. Table 1. Mean squares and significance levels from the variance analysis of grain yield (GY), test weight (TW), Hessian fly larvae and pupae per 10 culms (LP), and percentage of tillers infested (PTI). Source df GY TW LPt PTI* Years (Y) I 93348246 ** 27061 ** 1261 ( SA)* I (0.08) Blocks B(Y) 10 191088 88 ** 258 ( 1.6) 4 (0.07) Cultivars (C) 5 9804117 ** 118 ** 2773 (24.5) ** 43 (1.18) ** Y*C 5 1983939 ** 277 ** 148 ( 0.6) 2 (0.05) C B(Y) 50 262947 8 205 ( 1.0) 2 (0.05) Insecticide treatments (T) 2 I 913387 ** 28 ** 1427 ( 5.2) ** 14 (0.28) ** Y*T 2 352973 8 145 ( 0.3) I (0.0 I) C*T IO 660735 ** 21 ** 738 ( 2.7) ** 3 (0.07) Y*C*T 10 299243 ** 16 ** 82 ( 0.5) I (0.0'l) Error 120 105652 6 57 ( 0.9) 2 (0.05) *,**F values significant at the 0.05 and 0.01 probability levels, respectively. tresulls from analysis of log + I transformation of number of larvae and pupae per 10 culms. Mean squares from transformed values shown in parenthesis. +results from analysis of an arcsin square root transformation of the decimal fraction representing the number of tillers of a 10 culm sample infested with Hessian fly. Mean squares of Lransformerl values are shown in parenthesis and reported in radians.

PAGE 123

PROCEEDINGS, VOLUME 51, 1992 101 Overall, insecticide treatments did not increase grain yield and test weight of resistant cultivars (Ta bles 2 and 3). However, grain yield of susceptible cul tivars was increased by the APT treatment, and in 1989-90, grain yield of Florida 303 was higher in the PS than the APT treatment. Resistant cultivars Coker 9766 and FL85363-Gl8-14 had higher grain yield than the three susceptible cultivars when no insec ticide was applied. Florida 301H did not have signif icantly higher grain yield than its susceptible recurrent parent, Florida 301. In 1989-90, grain yield of Florida 301 and 303 was not different from that of Coker 9766 when given the PS treatment. No clear relationship between test weight and re sistance or susceptibility to Hessian fly was found. In 1989-90, Florida 303 at the PS level of insecticide treatment had a higher test weight than Florida 303 at the NT level. Also, in 1989-90, the highly suscepti ble cultivar Florida 302 had the lowest test weight at the NT and PS treatments but not at the APT treatment. Whereas, in 1990-91, Coker 9766 had lower test weight than FL85363-G 18-14 at the PS treatment, and again there were no differences between cultivars given the APT treatment. In 1989-90, the APT was sufficient to reduce the Hessian fly larvae and pupae per 10 culms on Florida 301 and 303 to the numbers not significantly different from those found on resistant cultivars. Florida Table 2. Grain yield, test weight and larvae and pupae counts per IO culms at three levels of insecticide treatment [no insecticide applied (NT), an at planting in the row treatment with phorate (APT), or an at planting treatment with phorate plus an application of disulfoton with the nitrogen topdressing (PS)], grown during the 1989-90 soft red winter wheat growing seasons at Marianna, Florida. Cultivar NT APT PS ------------Grain yield (kg ha1 ) --------------Coker9766 4170 at 4152 a 3951 a FL85363-Gl8-14 3506 b 3245 b 3252 b Florida 30 I H 3331 be 3037 b 3377b Florida 301 3037 cd 3435 b 3592 ab Florida 303 2718 d 3399b 3937 a Florida 302 1366 e 2402 C 2657 C ------------fest weight (kg hl1 ) -------------Coker9766 73.1 a 72.7 a 72.7 ab FL85363-G 18-14 72.0 ab 71.5 a 71.8 ab Florida 301 H 72.4 a 72.8a 72.8 ab Florida 301 72.9 a 72.5 a 73.1 a Florida 303 71.2 ab 73. la 73.7 a Florida 302 69.5 b 70.6 a 70.1 b Larvae and pupae counts (no. 10 culms) Coker9766 2(0.6) C 11(1.8) b 6(1.4) a FL85363-G 18-14 5(1.3)c 7(1.6) b 8( 1.7) a Florida 30 I H 1(0.4) C 2(0.9) b 0(0.0) b Florida 301 I 7(2.6) b 5( 1.2) b 5( 1.3) a Florida 303 27(2.6) b 6(1.5) b 4(1.2) a Florida 302 56(3.7) a 31(3.0) a 14(1.9) a tmeans within a column followed by the same letter are not significantly different (P < 0.05) according to Duncan's multiple range test. Table 3. Grain yield, test weight and larvae and pupae counts per 10 culms at three levels of insecticide treatment [no insecticide applied (NT), an at planting in the row treatment [no insecticide applied (NT), an at planting treatment with phorate plus an appli cation of disulfoton with the nitrogen topdressing (PS)], grown during the 1990-91 soft red winter wheat growing seasons at Marianna, Florida. Cultivar l\T APT PS ------------Grain yield (kg ha 1 ) --------------Coker 9766 2757 at 2524 a 2599 a FL85363-Gl8-14 2481 a 2667b 2524a Florida 30 I H 1470 b 1682 b 1656 b Florida 301 1380b 1861 b 1793 b Florida 303 1459b 1760b 1839 b Florida 302 1273 b 1667 b 1502 b -------------Test weight (kg hl1 ) -------------Coker9766 64.8 ab 64.2 a 63.9b FL85363-G 18-14 67.0a 67.0a 66.9a Florida 301H 63.8 b 64.7 a 64.8 ab Florida301 64.0 b 65.3 a 64.7 ab Florida 303 64.6 ab 64.3 a 65.0 ab Florida 302 65.2 ab 65.5 a 64.7 ab Larvae and pupae counts (no. IO culms) Coker 9766 2(0.7) C 2(0.8) be 3(0.7) be FL85363-G 18-14 2(0.8) C l(0.7)bc 3(0.8) be Florida 30 I H 0(0.0) C 0(0.0) C 0(0.l)c Florida 301 7(1.9)b 5(1.3) b 7( 1.5) ab Florida 303 :18(3A)a 14(2.6) a 6(1.3) ab Florida 302 11(2.5)ab 6(1.5) b 7(1.9) a tmeans within a column followed by the same letter are not significantly different (P < 0.05) according to Duncan's multiple range test. 302 had a higher count of larvae and pupae per 10 culms than all other cultivars in the study, while Florida 301H had the lowest and over all levels of insecticide treatment. Although Hessian fly populations were lower in 1990-91 [mean of larvae and pupae per 10 culms was 11.6 (89-90) vs. 6.3 (90-91)], the percentage of tillers infested were nearly identical [mean 17% (89-90) vs. 16%(90-91)]. Insecticide treatments had no effect on the levels of larvae or pupae per 10 culms or the percentage of tillers infested in the resistant cultivars (Table 4). An APT treatment reduced infested tiller counts in the susceptible cultivars. The PS treatment further reduced larvae and pupae counts in Florida 302 in 1989-90. DISCUSSION Chapin et al. ( 1991) found that post planting ap plications of disulfoton, even when timed to correspond with spring Hessian fly emergence, had no posi tive effects upon yield (although insect numbers were reduced). Similarly, in this study, reductions in insect numbers on the susceptible cultivars Florida 301 and 302 did not correspond with increases in grain yield. However, the PS treatment increased grain yield for Florida 303. This suggests that some cultivars may be more responsive to insecticide treatments. Florida

PAGE 124

102 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Table 4. Percentage of Hessian fly infested tillers at three levels of insecticide treatment [no insecticide applied (NT), an at planting in the row treatment with phorate (APT), or an at planting treatment with phorate plus an application of disulfoton with the nitrogen topdressing (PS)] of six cultivars grown during the 1989-90 and 1990-91 soft red winter wheat growing seasons at Marianna, Florida (growing seasons combined). Insecticide Cultivars treatment Coker 9766 FL85363-G 18-14 Florida 301 H Florida 301 Florida 303 Florida 302 NT 10(.22) at 12(.28) a 2(.05) a 30(.57) a 33(.59) a 43(. 72) a APT 11(.27) a 13(.3l)a 3(.11) a 18(.36) b 17(.37) b 34(.61) ab PS 9(.23) a 11(.29) a 1(.03) a 17(.34) b 13(.29) b 25(.50) b +means of percentage of tillers infested in a column followed by the same letter are not significantly different (P < 0.05) according to Duncan's multiple range test. Separations based on arcsin square root transformation of the decimal fraction representing percentage of tillers infested. Transformated values are shown in parenthesis and are reported in radians. 303 has higher levels of resistance of leaf rust (Puccinia recondita Rob. ex Desm. F. sp. tritici) and powdery mildew [Blumeria graminis (DC.) E. 0. Speer F. sp. tritici) than Florida 301 and 302. Thus the effects of disease are less limiting, and insecticide treatments may have a more positive effect on the yield of Florida 303. Heavy rains in April and May of 1991 led to severe outbreaks of leaf rust, powdery mildew, and septoria glume blotch (Leptosphaeria nodorum Muller (anamorph Septoria nodorum Berk.)). It is probable that this complex of diseases was more limiting to grain yield and test weight than Hessian fly for the 1990-91 season. Therefore the effect of insecticide treatments on susceptible cultivars was lower. Only in the case of Florida 303 with the PS treatment under the 1989-90 environment does a suscep tible cultivar match the best resistant cultivars. Florida 301 gave yields equal to Florida 301 H with or without insecticide. Both generally yielded lower than Coker 9766 or FL85363-Gl8-14. In 1989-90, Coker 9766 yielded over I t ha1 more in the NT treatment than Florida 303, without the added cost of the insecticide. This study suggests that, in the absence of pronounced differences in disease resistance, the use of Hessian fly susceptible cultivars in conjunction with a systemic insecticide is not a competitive alternative to the use of Hessian fly resistant cultivars. No measurable advantage to applying systemic in secticides to Hessian fly resistant cultivars was demonstrated by this study. REFERENCES Bigger, J. H., P. E. Johnson, and R. 0. Weibel. I 965. Controlling Hessian flv with phorate and disulfoton. J-Econ. Entomol. 58: I 083-1085. Brown, H. E. 1960. Insecticidal control of the Hessian fly. J. Econ. Entomol. 53:501-503. Buntin, G. D. 1990. Hessian flv (Diptera:Cecidomyiidae) management in winter wheat using systemic insecticides at planting. JAgric. Entomol. 7:321-331. Buntin, G. D., P. L. Bruckner, and J. W. Johnson. 1990. Management of Hessian fly (Diptera:Cecidomyiidae) in Georgia by delayed planting of winter wheat. J-Econ. Entomol. 83: 10251033. Buntin, G. D., and J-W. Chapin. 1990. Biology of Hessian fly (Diptera:Cecidomyiidae) in the southeastern United States: geographic variation and temperature dependant phenology. ]. Econ. Entomol. 83: 1015-1024. Buntin, G. D., and P. L. Raymer. 1989. Susceptibility of winter wheat and triticale to the Hessian fly. Georgia Agric. Exp. Stn. 389. Chapin, J. W., J. F. Grant, and M. J-Sullivan. 1989. Hessian fly (Diptera:Cecidomyiidae) infestation of wheat in South Carolina. J. Agric. Entomol. 6: 137-146. Chapin, J. W., M. J-Sullivan, and J. S. Thomas. 1991. Disulfoton application methods for control of Hessian ny (Diptera: Cecidomyiidae) on southeastern winter wheat. J-Agric. En tomol. 8:17-28. Duncan, D. B. 1955. Multiple range and multiple F tests. Biomet rics 11: 1-42. Johnson, J-W. and G. D. Buntin. I 989. Breeding for Hessian fly resistance in the southeastern United States. In Proc. 5th Int. Wheat Conf., Rabat, Morocco. Large, E. C. 1954. Growth stages in cereals. Plant Pathology 3: I 28129. Lidcll, M. C. and M. J<". Schuster. I 990. Distribution of the Hessian fly and its control in Texas. Southwestern Entomologist 15: 133145. SAS Inst. 1988. SAS language guide for personal computers. Re lease 6.03. Cary, NC.

PAGE 125

PROCEEDINGS, VOLUME 51, 1992 103 Occurrence and Control of Target Spot in Tobacco in Florida Caused by Rhizoctonia solani Kuhn T. A. Kucharek*, R. Tervola, and J. Washington ABSTRACT Target spot of tobacco (Nicotiana tobacum L.), caused by Rhizoctonia solani Kuhn, was first identified in Florida in trans plant beds and field plantings in 1989. Target spot continued to appear at increasing levels and at more sites in 1990 and 1991. No control measures were available for this serious disease. Ini tially, wounding was thought to be a predisposing factor for disease, but with studies in the greenhouse, it was determined that although wounding enhanced the severity of disease, wounding was not necessary for disease to occur. Because an effective control was needed, the fungicide iprodione 4F was tested and was effective for controlling target spot in the greenhouse, the trans plant bed, and on transplanted tobacco. lprodiione reduced lesion size by as much as 100% in greenhouse tests. Incidence of target spot was reduced by 50% (P=0.001) and 62% (P=0.05) in the transplant bed and in transplanted tobacco, respectively. Isolates from foliar lesions in tobacco were determined to be multinucleate and represented three anastomosis groups of R. solani. In 1989, a new leafspot in tobacco was identified in north Florida in plant beds and fields. The new disease became progressively more widespread in 1990 and again in 1991. Leafspots varied in size and appearance (3). Based upon microscopic examinations, agar culture isolations, inoculation of tobacco plants and reisolation of R. solani, it was determined that R. solani was the causal agent of the new leafs pot. Stem rot (sore shin) and root rot, caused by R. solani, have been present in Florida-produced tobacco for more than 20 yr and probably ever since tobacco has been produced in Florida. Within the past 5 to 6 yr, an increased number of inquiries about sore shin in tobacco has occurred. Leafspot caused by R. solani may have occurred at a low level in Florida prior to 1989 and thought to be another disease. Rhizoctonia leafspot, called target spot (5), has expressed symptoms in Florida that mimic brown spot, caused by Altemaria altemata (Fr. ex Fr.) Keissl.; blue mold, caused by Peronospora tabacina Adam; and frogeye leafspot, caused by Cercospora nicotianae Ellis & Everh. The first known exis tence of target spot in tobacco has been traced to a report in Brazil in 1948 (5). Reports about target spot in the United States (6,7), elsewhere (4), and our experience indicate that target spot has the potential to be a major yield-limiting disease of tobacco in Florida. Control of target spot by any means was lacking at the beginning of these investigations. The purpose of these studies included three primary investiga tions. First, studies in the greenhouse were initiated L A. Kucharek and J. Washington, Plant Pathology Dep., Univ. of Florida, Gainesville, FL 32611, R. Tervola, 1302 II St., S.W., Live Oak, FL 3206. Florida Agric. Exp. Stn. Journal Series no. R-00609. *Corresponding author. Use of trade names in this publication does not imply endorsement by the Institute of Food and Agricultural Sciences, University of Florida. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51:103-106 (1992) to determine if wounding was required for infection and if chemical control of target spot would be effec tive without phytotoxicity. Secondly, tests in the field were established to determine if chemical control could be achieved in the field. Thirdly, a cursory sur vey was done to determine the nuclear status and the anastomosis grouping (AG) for isolates that ong1-nated from target spots in tobacco in Florida. MATERIALS AND METHODS Greenhouse Tests Tobacco plants of the cultivar NC2326 with five to seven leaves were used for all tests in the greenhouse. The upper most, horizontally-oriented leaves were used for testing in the greenhouse. All tobacco plants were produced from seed planted in Metromix 200 (W.P. Grace & Co., Fogelsville, PA)' within clay dishes and transplanted into 10.2 cm plas tic pots with Metromix 200. All plants were fertilized to maintain a vigorous appearance. Leaf number per plant was uniform and the same leaf position was used for treatments within tests. Cultures of R. solani were grown on acidified potato dextrose agar. Inoculum consisted of 4 mm diam. mycelial disks that were cut aseptically with a cork borer from near the edge of an actively growing 3 to 4 d old colony of R. solani. For inoculated treatments, the mycelial plug was placed with the mycelium in contact with the upper leaf surface and centered between the midvein, side edge, apical tip, and basal edge of the leaf. Isolate 48, which originated from a foliar lesion in tobacco grown in Columbia County, Florida in 1989, was previously determined to be more aggressive than two other isolates (29 and 30) and was therefore used for inocula in the greenhouse tests. For the wounding studies, six and seven replicates were used for the first and second tests, respectively. In the first test, each replicate had three plants and one plant was used per replicate in the second test. Wounding was done by placing a small amount of carborundum on cheesecloth and rubbing the upper surface of the leaf gently prior to inoculation. After inoculation, the plants were placed within a chamber which received a constant mist from a humidifier for 5 cl after which disease assessments were made. Statis tical mean separations were done using Students t for unpaired comparisons. For the efficacy tests with iprodione 4F (3-[3,5,dichloro-phenyl)-N-[ l-methylethyl]-2 ,4-dioxo-1-imi dazolidinecarboxamide), six and five replicates were used for the first and second tests, respectively. Treatments with iprodione consisted of 1 ml of iprodione plus 0.5 ml of the adjuvant Triton CS-7 (Rohm and Haas Co., Philadelphia, PA) (blend of alky laryl polyethoxylate and sodium salt of alkylsulfonate

PAGE 126

104 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA dalkylate) in 100 ml of water. Fungicide treatments were applied with a paint brush to the upper leaf surface and allowed to dry before inoculation. After inoculation, the plants were placed within the misting chamber for 5 and 4 d at which time disease assessments were made for the first and second tests, re spectively. Field tests for efficacy of if1rodione Two tests in Suwannee County, Florida, one in a transplant bed and one in a transplanted field, were conducted to determine the efficacy of iprodione for the control of target spot. Naturally occurring inocula were relied upon for disease. A randomized complete block test design with four replicates was used for each test. The tobacco cultivar K326 was used in both tests. A rate of iprodione 4F of 2.34 L ha I ( 1 qt acre') was used in both field tests. Additionally, lower rates were evaluated in the transplanted field. When used, rate of the adjuvant, Triton CS-7, was 1.2 L ha I ( 1 pt acre-I). The final spray consisted of the fungicide, with or without the adjuvant, diluted in 234 L of water ha-1 (25 gallons acre I) and applied with a CO2 backpack sprayer at a spray pressure of 207 kPa (30 psi) through a single LE 6 (Delavan, Lexington, TN) nozzle per plot. The test in the transplant bed was conducted after the grower had completed transplanting and after the yellowed plants were fertilized. Target spot was present in leaves of some plants throughout the test site. Plots in each replicate consisted of an area 1.83m long by 0.31-m wide which was the swath width of the spray. This same zone was used for disease assessments. Sprayed and unsprayed plots were each bordered by an unsprayed zone 0.31-m wide. Disease assessments were made 8 d after the single spray ap plication. Disease assessments included the number of spots/5 leaves and the number of infected leaves. Mean separations were done with Students t for unpaired comparisons. The test in the transplanted commercial field was initiated 3 d after transplanting at which time the first fungicide applications were made. Each plot consisted of two rows 1.1-m apart and 9.2-m long. Target spot was present on some of the transplants which were just reviving from transplant shock (wilt). Four applications were made, 3, 11, 21 and 28 d after transplanting (DATP). Each spray was directed through a single nozzle centered along the row. A vigor rating of plants was made 11 DA TP and disease incidence was assessed at 33 DATP. Vigor ratings were made on a 1-4 scale with 4 being the best. Disease incidence was assessed as the number of plants with target spot. Laboratory Classification of Isolates Twelve isolates of Rhizoctonia spp. from tobacco were grown in potato dextrose broth. Mycelia from these cultures were stained in bisbenzamide trihydrochloride (Sigma Chemical Company, St. Louis, MO.) which was diluted in deionized water at a rate of 0.001g ml' which was the stock solution (Washington, unpublished PhD thesis). The final solution applied to a microscope slide for examination of mycelia was one-two drops of the stock solution and 10 ml-I deionized water. After a cover slip was placed on the specimen for at least 10 sec, the preparation was rinsed thoroughly by pipetting distilled water under one side of the cover slip while absorbing the rinsate on the other side of the cover slip with tissue paper. The stained specimens were examined with a Leitz Dialux 20 microscope fitted with a Dl3-572 Ploemopak 24 epifluoresent illuminator transmitting ultraviolet light (400 nm) through a Dl3-413 filter system to determine if cells of the cultures were bi-or multi-nucleate. Multinucleate isolates were tested for positive or negative anastomosis with known tester isolates of anastomosis groups (AG) 1, 2-1, 2-2, 3, and 4 obtained from D. Sumner at the University of Georgia. Pairings of known and unknown isolates were done on water agar-coated microscope slides. The presence or absence of anastomoses was determined micro scopically at 1 OOX. Visual observation of multiple anastomoses within pairings was required to classify an unknown culture to a specific anastomosis group. RESULTS AND DISCUSSION Greenhouse tests Wounding the leaves with carborundum increased the size of the lesions in two tests. In the first test, lesion diameter was 5.7 and 23.6 mm for the unwounded and wounded treatments, respectively (P= 0.05). In the second test, lesion diameter was 12.3 and 19.9 mm for the unwounded and wounded treatments, respectively (P = 0.05). Disease assessments for the two tests were made 7 and 5 d after inoculation for the first and second tests, respectively. These data suggest that damage to leaves can enhance disease but damage to leaves is not necessary for disease to occur. Similarly the sore shin phase is reported to be more severe after wounding (5). In the tobacco-growing region of north Florida, wind blown sand occurs in most if not all seasons and such an event may enhance target spot if favorable weather for infection follows. Iprodione reduced the size of lesions in both tests in the greenhouse. In the first test, leaves were wounded prior to application of iprodione and subsequent inoculation but were left unwounded in the second test. I prodione reduced the size of the lesions by 89 and 100% in the first and second tests, respec tively. Assessments were made 5 and 4 d after inoculation, respectively. Lesion size expanded to 16.2 and 17 .1 mm in the untreated controls for the two tests, respectively. No phytotoxicity occurred. The results from these two tests indicated that iprodione effectively controlled target spot that was ar tificially inoculated in the greenhouse. Target spot has been a serious problem in transplants of tobacco produced in the greenhouse in North Carolina (7).

PAGE 127

PROCEEDINGS, VOLUME 51, 1992 105 Most transplants of tobacco produced in north Florida are produced in ground-level beds that are covered with clear plastic covers until excess heat occurs. Although target spot has not been a major problem in Florida prior to the time plastic covers are removed, it is possible that this disease could become serious prior to removal of the plastic. The data presented herein indicate that iprodione can be used for effective control under similar conditions without phytotoxicity. Field tests for efficacy of iprodione Iprodione reduced the number of spots per five leaves by 66% (P = 0.05 ), the number of infected leaves by 50% (P = 0.001), and the overall disease index by 82% (P = 0.05) in a transplant bed 8 d after treatment (Table 1). Use of iprodione in the transplant bed is currently thought to be an important method for reducing disease in the transplant bed and the amount of disease that ultimately occurs after transplanting. Although all sources of inocula and their respective importance for target spot are not known at this time, one source of inoculum for transplanted tobacco is certainly that which occurs on the transplants. The use of multiple applications of iprodione on transplanted tobacco was not originally intended. Target spot did not progress until the plants became larger and thus, it was decided to maintain fungicide residues on the treated plants until the disease progressed (7). Applications of iprodione to transplanted tobacco reduced the incidence of target spot (Table 2). Iprodione reduced the incidence of disease by 62% (P = 0.05) when no adjuvant was added to the spray and further reduced the incidence of disease when the adjuvant, Triton CS-7, was added even though lower rates of iprodione were used. Efficacy of iprodione at lower rates is important because the cost of iprodione is high. Further studies on efficacy of iprodione at lower rates will be conducted. Collectively, the efficacy tests in the greenhouse, the transplant bed, and on the transplanted tobacco suggest that iprodione is a viable option for control of target spot of tobacco. An emergency label for use Table 1. Effect of one foliar spray of ipriodione 4F on the control of Rhizoctonia-induced target spot of tobacco in a transplant bed of the cultivar K326 in Suwannee County, Florida during 1991. Treatment Unsprayed I prodione treated No. spots/ % leaves Disease 5 leavest infected index* 14.0* 4.8 50*** 25 56.0* 10.3 *, ***Means are different at P = 0.05 and P = 0.00 l, respectively in unpaired Student"s t tests. tFive leaves selected/0.65 m' in each of four replications on 19 Apr. *Disease index = Average of no. spots/five leaves/plot X no. of leaves infected/plot. .34 L ha- of iprodio11e 4F + 1.2 L ha of the adjuvant (Triton CS-7) was applied in 234 L ha-1 of water at 207 kPa through one LE 6 nozzle in a 0.36m band on 11 April. Table 2. Effect of four foliar sprays of iprodione 4F on the early season control of Rhizoctonia-induced target spot of tobacco in a field planting of the cultivar K326 in Suwannee County, Florida during 1991. Treatment and rate ha-1 l' nsprayed lprodione 2.34 L' lprodione 1.8 L + adjuvant Iprodione 1.2 L + adjuvant I Apr 2.1 a* 2.6 a 2.6 a 3.0a No. plants' with spots 24Apr 19.Sa 7.5 b 6.3b 5.B b *Means within columns followed by the same letter do not differ (P = 0.05, Duncan's multiple range test). tVigor rating made 11 d after treatment is on a 1-4 scale with 4 being best. *No. of possible plants was 36 (I 8 in each of two rows). Assessments made 33 d after planting. Sprays applied in 234 L ha' of water through one LE6 nozzle/ row at 207 kPa. When an adjuvant (Triton CS-7) was used in the spray, the rate was 1.2 L ha'. Applications were made 21 March, 1 April, 11 April and I 9 April. Plants were transplanted on 18 March. of iprodione on tobacco was issued in 1991 based upon bioassays (in vitro; Kucharek, unpublished data). These data support the findings in the bioas says that iprodione is effective for control of target spot in tobacco. Laboratory classification of isolates Of the 12 isolates acquired from tobacco in Florida from 1989 to 1991, 11 isolates of Rhizoctonia spp. were determined to be multinucleate (Table 3). Except for isolate 4 which was isolated from stem tissue and determined to be R. zeae, the isolates were determined to be R. solani. Isolate 113 was the only one that was in AG 2-2. The majority of the isolates from foliar lesions were in AG 3. Isolate 48, which was used for the inoculation studies m the greenhouse, and isolate 47 were in AG 4. The occurrence of AG's 2-2 and 3 with foliar le sions in tobacco is not new. AG 2-2 and AG 3 were found to cause foliar lesions in tobacco in North Table 3. Classification of isolates of Rhizoctonia spp. from tobacco in Florida from 1989 to 1991. Plant R. solani part Nuclear anastomosis group No. origin County status or species 1 leaf Columbia multi AG3 3 leaf Hamilton multi AG3 4 stem Columbia multi R. zeae 19 leaf Lafayette multi AG3 29t leaf Columbia multi '.mt leaf Union multi AG3 46 stem Tavlor 47 stem M;dison multi AG4 4gt leaf Columbia multi AG4 ll2 leaf Lafayette multi AG3 113 leaf Lafayette multi AC 2-2 114 leaf Lafayette multi AG3 'Determined to be pathogenic based upon Koch's Postulates.

PAGE 128

106 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Carolina (6, 7) and South Africa (4), respectively, in recent years. The association of AG 4 with lesions in leaves of tobacco might be considered unusual. How ever, Chase (2) found AG 4 to be pathogenic to foliar tissues of several foliage ornamental species in Florida. Although Anderson ( 1) promotes the idea of each AG as an evolutionary unit, with specific AG's tending to be associated with specific diseases or crops, he also summarizes information about anas tomoses between isolates belonging to different AG's and relates to such occurrences as microevoluntionary events. At this time, it would not be wise to assume that most or all isolates of R. solani that cause foliar lesions in tobacco or any other crop should belong to a restricted group or groups. However, for future reference, a thorough survey is needed to determine the current status of AG types found in asso ciation with Florida-produced tobacco. SUMMARY The continual increase in severity of target spot, caused by R. solani, in tobacco in Florida and elsewhere in the southeastern United States, is of concern. Wounding of leaves enhanced the severity of disease but infection occurred without wounding. Ip-rodione reduced target spot of tobacco in the greenhouse, the transplant bed, and in a transplanted field. Because all of the isolates of R. solani associated with foliar lesions were determined to be multinucleate and were further classified into three AG groups, the potential for wide variations in this dis ease seems plausible. Cultural and additional chemical controls are needed. REFERENCES I. Anderson, :'\i. A. I 982. The genetics and pathology of Rhizoc tonia solani. Ann. Rev. Phytopath. 20:329-347. 2. Chase, A. R. 1991. Characterization of Rhizoctonia species isolated from ornamentals in Florida. Plant Dis. 75:234-238. 3. Kucharek, T.A. 1990. Rhizoctonia diseases in aboveground plant parts of agronomic and vegetable crops. Univ. of Florida Plant Pathology Fact Sheet PP 41. 4. Meyer,]. C., R .J. Van Wyk, and A.J. L. Phillips. 1990. Rhizoctonia leaf spot of tobacco in South Africa. Plant Path. 39:206207. 5. Shew, H. D. and G. B. Lucus. I 990. Compendium of tobacco diseases. APS Press, St. Paul, MN. 6. Shew, H. D. and C. E. Main. 1985. Rhizoctonia leafspot of flue-cured tobacco in North Carolina. Plant Dis. 69:901-903. 7. Shew, H. D. and C. E. Main. 1990. Infection and development of target spot of flue-cured tobacco caused by Thanatephorus cucumeris. Plant Dis. 74:1019-1013.

PAGE 129

106 So11, AND CROP SCIENCE SocIFTY OF FLORmA Dry Matter Production and Forage Quality of Line 8400 Stylo, Alyceclover, and Hairy Indigo M. J. Williams*, C. G. Chambliss, J.B. Brolmann, and S. L. Sumner ABSTRACT Preliminary legume evaluation trials showed that Line 8400 stylo [Stylosanthes guianensis (Aubl.) Sw.], although by growth habit a perennial, when grown as an annual remained vegetative later in the fall than many commonly used tropical annual legumes. This suggested that Line 8400 stylo might be useful as a stockpiled forage or late season hay crop in those areas where it failed to perennate due to climatic conditions. Planting date (April or June) and harvest date (August, September, October, or November) effects on seasonal dry matter (DM) distribution, total DM production, crude protein (CP), and in vitro organic matter digestibility (IVOMD) of Line 8400 stylo, common alyceclover, [Alysicaryus vaginalis (L.) DC.], and 'Flamingo' hairy indigo (Indigofera hirsuta L.) were determined in an irrigated trial in 1989. Dry matter yield of Line 8400 (l.6 -8.7 Mg ha') did not differ (P>.0.05) from that of comparable age alyceclover or hairy indigo at any harvest date. Seasonal DM distribution of June planted Line 8400 (>50% after August) was shifted later into the fall than either June-planted annual species (<20% DM after Au-M. J. Williams, USDA, ARS, Subtrop. Agric. Res. Stn., P. 0. Box 46, Brooksville, FL 34605; C. G. Chambliss, Agronomy Dep., Univ. of Florida, Gainesville, FL 32611; J. Brolmann, Florida Agric. Res. and Educ. Center, P. 0. Box 248, Ft. Pierce, FL 34954; S. L. Sumner, Univ. of Florida, Coop. Ext. Ser., 1702 Highway I 7-98 South, Bartow, FL 33830. Florida Agric. Exp. Stn. Journ~l Series no. R-00609. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51:106-I09 (1992) gust). In contrast, April-planted Line 8400 made <20% of total DM accumulation after August. In vitro organic matter digestibil ity and CP of Line 8400 across all harvest dates (467 -536 g kg-' IVOMD, 95 114 g kg-' CP) were similar (P>0.05) to comparable age alyceclover (433 -490 g kg-1 IVOMD, 93 107 g kg-1 CP) and generally superior (P<0.05) to comparable age hairy indigo (373 -470 g kg-1 IVOMD, 77 -89 g kg-', CP). This trial indicated that Line 8400 stylo when grown as an annual, regardless of planting date, would be as good if not superior in DM yield, CP, and IVOMD to common alyceclover and Flamingo hairy indigo. Forage quality of most warm-season grasses grown in Florida is generally insufficient during the fall to meet the nutritional requirements of replacement heifers. Annual legumes, such as hairy indigo, are of limited value during this period due to photoperiod-mediated flowering, which occurs around September, and subsequent rapid leaf loss after seed maturation. Line 8400 stylo, a recent development of the University of Florida, Institute of Food and Agricultural Sciences forage breeding program at the Agricultural Research and Education Center at Ft. Pierce (Brolmann, 1987), is a perennial legume that is a facultative annual in much of peninsular Florida due to frosts that occur during the winter season. In a preliminary legume evaluation trial at the USDA,

PAGE 130

PROCEEDINGS, VOLUME 51, 1992 107 ARS, Subtropical Agricultural Research Station (STARS) in Brooksville, FL, Line 8400 remained veg etative later in the year than most common tropical annual legumes grown in the state (Williams, 1988). This suggested that Line 8400 could be stockpiled for use in the fall without appreciable loss of yield or forage quality. Additionally, this period of extended vegetative growth could make Line 8400 a desirable alternative to alyceclover for hay production. Harvesting of Line 8400 could be scheduled to better coincide with more favorable hay curing weather conditions in the fall without the risk of significant yield and quality losses commonly associated with harvesting alyceclover at that time. As no direct comparison of seasonal DM distribution and forage quality of Line 8400 stylo and alyceclover or hairy indigo has been made, a series of studies was initiated to determine the feasibility of using Line 8400 as an annual forage crop for stockpiling or hay. This paper summarizes results of a preliminary trial evaluating the effects of planting date (April or June) and harvest date (August, September, October, or November) on st .:isonal DM distribution, total DM yield, and forage quality of these legumes. MATERIALS AND METHODS The test site, located on a private ranch near Haines City, FL, was irrigated with tertiary-treated sewage water. Although no record of rainfall or amount of irrigation water applied was kept, the waste-water-disposal schedule called for a minimum of 5 cm of waste water to be applied weekly throughout the test period. Plots were established on a Candler fine sand (hyperthermic, uncoated, Typic Quartizipsamment) in an area of the ranch recently converted from native rangeland (soil pH = 5.4). Immediately prior to the first planting date (10 April 1989), 336 kg ha-1 of 0-10-20 plus a micronutrient mix (0.15% Mn, 0.06% Cu, 0.36% Fe, 0.14% Zn, 0.06% B, and 0.001 % Mo) was applied to the plot area. Results from a soil test taken after the second planting date (23 June 1989) showed low levels of P (4 mg kg-1 ) and K (20 mg kg-1). An additional 196 kg ha-1 of 0-10-20 plus micronutrient mix was applied to all plots. Common alyceclover and Flamingo hairy indigo were planted at the rate of 22 kg ha 1 Line 8400 stylo seed was heat scarified 70 C for 4-6 hr and planted at the rate of 11 kg ha-1 All seed was inoculated with Bradyrhizobium spp. for alyceclover prior to planting. A split-split plot design with four replicates was used with planting date as the main plot, legume as the subplot, and harvest date (22 Aug, 19 Sep, 30 Oct, and 22 Nov 1989) as the sub-subplot. Legume subplots were planted with 15 cm interrow spacing using either a vegetable-type, single-row planter or small-plot cone planter. At each harvest date, the center 0.9 m X 4.6 m of each 1.8 m X 6 m sub-subplot was harvested with a sickle-bar mower set at a 10-cm stubble height to determine forage production. Seasonal DM distribution (%) was calculated for each harvest date based on (DMharvest date/DMmaximum yieid)*l00. An approximately 0.45-kg subsample from each subplot was dried in a forced-air oven 50 C for DM, CP, and IVOMD determination. Dried subsamples were ground to pass through a 1 mm screen. Crude protein (Gallaher et al., 1975) and IVOMD (Moore and Mott, 197 4) determinations were made at the Forage Evaluation Laboratory, University of Florida. Flowering time and presence of mature seed were also noted. All plots from the April planting date were harvested in August only. Only the stylo plots from this planting date were harvested at all subsequent harvest dates (see Results and Discussion). Data from the August harvest date (April-planted alyceclover, hairy indigo and stylo) were analyzed as a randomized complete block (RCB) model using the general linear models (GLM) procedure for PC-SAS (Statistical Analysis System, 1985). Data from subsequent harvests of April-planted stylo material was included as a fourth treatment in the analysis of June-planted material. These data (April-planted stylo and Juneplanted alyceclover, hairy indigo, and stylo) were first analyzed using the GLM procedure for a split-plot model (Statistical Analysis System, 1985) with legume as the main plot and harvest date as the subplot with the error terms of legume X replicate and residual error, respectively. If significant legume X harvest date interaction occurred, the data were further analyzed separately for each harvest date using a RCB model as previously described. Means were separated by Duncan's multiple range procedure (Statistical Analysis System, 1985 ). RESULTS AND DISCUSSION Only the April-planted material was harvestable at the first harvest date (August). There was no difference (P = 0.09) in DM yield among the three legumes at that date (4.7, 6.2, and 5.0 Mg ha-1 for alyceclover, hairy indigo, and Line 8400 stylo, respec tively). Neither April-planted alyceclover nor hairy indigo was harvested further due to flowering-induced senescence (see flowering discussion). Juneplanted alyceclover and hairy indigo exhibited similar flowering-induced senescence, although at a later date (see flowering date discussion) and were not harvested in November. Dry matter yield of Aprilplanted stylo and June-planted alyceclover, hairy indigo, and stylo are reported in Table 1. There was no legume X harvest date interaction (P = 0.20). As expected, because of the longer growing season, DM yield of April-planted Line 8400 was consistently higher (P<0.05) than that of any of the June-planted material. There was no difference (P>0.05) in average DM yield of June-planted legumes. June-planted material was not harvested in August due to insufficient growth ( < 10 cm harvest height). Although the growing period prior to the August harvest date was shorter after the June planting date (8 wk) than the April planting date (16 wk) and would be expected to affect DM yield, the non-

PAGE 131

108 Son, AND CROP SCIENCE SOCIETY OF FLORIDA Table 1. Seasonal dry matter (DM) production of Line 8400 stylo grown as an annual compared to common alyceclover and Flamingo hairy indigo in Central Florida in 1989. Legume Alyceclover Hairy Indigo Line 8400 Stylo Line 8400 Stylo Planting date June June June April Harvest Date September October November Mean ----------------------------------------Mg ha ---------------------------------------1.8 2.1 2.0 b* 2.7 2.5 2.G b 1.6 3.3 3.1 2.8 b 7.0 8.7 7.3 7.7 a 'Means in a column followed by the same letter do not differ (Duncan's multiple range test, P<0.05). harvestability of June-planted material was probably not solely due to planting date. In a companion study conducted the same year at STARS where only a June planting date was used (Williams et al., 1991 ), harvestable material of all three legumes was available in August (4.3 Mg ha 1 alyceclover, 3.0 Mg ha-1 hairy indigo, and Line 8400 0.6 Mg ha-1). Retarded growth of June-planted material in this study was probably a consequence of low soil fertility and irrigation practices at the site. Levels of P and Kat the June planting date were low, although fertilizer was applied in April. Irrigation water applied between the April fer tilizer application and June planting date probably leached away much of the soluble nutrients applied, particularly K. Symptoms of what was believed to be K deficiency were noted (C.G. Chambliss, personal observation) on newly-emerged June-planted material. These deficiency symptoms disappeared after the second fertilizer application. Low levels of K have been found to limit DM production of stylo lines (Brolmann and Sonoda, 1975). Similar to what was found by Williams et al. (1991), distribution of DM accumulation for Juneplanted Line 8400 in this study was shifted to later in the growing season than for June-planted alyceclover or hairy indigo. Over 80% of the total DM yield of June-planted alyceclover and hairy indigo was made by September, compared to <50% for Line 8400 stylo. In contrast to June-planted Line 8400 stylo, almost 80% of the total DM yield of April-planted Line 8400 stylo had accumulated by September. Early planting appeared to accelerate maturation of the true annuals, as measured by flowering date, but did not affect Line 8400 stylo. April-planted alyceclover and hairy indigo plots were in full bloom at the August harvest date, while June-planted alyceclover and hairy indigo did not flower until September. Line 8400 stylo, planted at either planting date, did not flower until October. Previous studies with perennial Stylosanthes spp. have shown that photoperiod and plant age interact to affect both length of vegetative growth period and date of flower induction. These combine to determine total forage yield and, importantly for a facultative annual, seed production (Ison an Humphreys, 1984). Further studies with Line 8400 stylo need to be conducted at differing locations in the state to determine the consequence of planting date on the interaction of seed production and forage yield. In August, IVOMD and CP of April-planted alyceclover (490 and 107 g kg, respectively) was similar to (P>(l.05) Line 8400 stylo (508 and 110 g kg-1 respectively) and both were higher (P<0.05) than IVOMD and CP of April-planted hairy indigo (433 and 89 g kg1, respectively). At subsequent harvests there was a legume X harvest date interaction (P= 0.02) for IVOMD due to the digestibility ofJuneplanted alyceclover and hairy indigo declining more rapidly across harvest dates than did the digestibility of Line 8400 stylo planted at either April or June (Table 2). There was no legume X harvest date interaction (P = 0.49) for CP content. Across harvest date CP content of either planting date of Line 8400 was higher (P<0.05) than June-planted hairy indigo and equivalent (P>0.05) to June-planted alyceclover (Table 3). Regardless of planting or harvest dates, the trend was for digestibility and CP of Line 8400 stylo to be equal to or superior to that of alyceclover or hairy indigo. It is thought that higher leaf to stem ratio of Line 8400, particularly after alyceclover and hairy indigo flower, accounts for the better forage quality of Line 8400 stylo in the fall. This trial showed, that regardless of planting date, when grown as an annual Line 8400 stylo was as good if not superior to either alyceclover or hairy indigo in terms of total DM yield and forage quality, particularly after October. Unfortunately preliminary hay production trials confirmed, that as with other Stylosanthes spp. (Haggar, 1969), curing Line 8400 stylo for hay may be a problem due to excessive leaf loss (M. J. Williams, personal observation). Further Table 2. In vitro organic matter digestibility (IVOMD) of Line 8400 stylo grown as an annual compared to common alyceclover and Flamingo hairy indigo at three harvest dates in Central Legume Alyceclover Hairy Indigo Line 8400 Stylo Line 8400 Stylo Florida in 1989. Harvest date Planting--------------date September October November ----------------------g kg-' ----------------------June 475 a 433 be June 470a 373 C June 536 a 571 a 477 a April 467 a 473 ab 459a 'Means in a column followed by the same letter do not differ (Duncan's multiple range test, P<0.05).

PAGE 132

PROCEEDINGS, VOLUME 51, 1992 109 Table 3. Crude protein (CP) content of Line 8400 stylo grown as an annual compared to common alyceclover and Flamingo hairy indigo in Central Florida in 1989. Legume Alyceclover Hairy Indigo Line 8400 Stylo Line 8400 Stylo Planting date June June June April Harvest date September October November Mean ---------------------------------------g kg-' ----------------------------------------93 77 96 96 98 79 114 104 95 95 95 a* 78 b 101 a 96a *Means in a column followed by the same letter do not differ (Duncan's multiple range test, P<0.05). efforts are needed to determine the feasibility of hay production. This suggests that where Line 8400 stylo must be grown as an annual, initially it may be best used as a stockpiled forage if cattle can be managed to consume seasonal DM accumulation prior to frost. REFERENCES Brolmann, J. B. 1987. Registration of FP-8400 pencilflower germplasm. Crop Sci. 27:153. Brolmann, J. B., and R. M. Sonoda. 1975. Differential response of three Stylosanthes guyanensis varieties to three levels of potassium. Trap. Agric. (Trinidad) 52: 139-142. Gallaher, R. N., C. 0. Weldon, and J. G. Futral. 1975. An aluminum block digester for plant and soil analysis. Soil Sci. Soc. Am. Proc. 39:803-806. Haggar, R. J. 1969. A guide to the management and use ol styio (Stylosanthes gracilis). Samaru Agric. News. 11:63-66. Ison, R. L.,and L. R. Humphreys. 1984. Reproductive Physiology of Stylosanthes. p. 257-278. In H. M. Stace and L.A. Edye (eds.). The biology and agronomy of Stylosanthes. Academic Press, Orlando, FL. Moore, J. E., and G. 0. Mott. 1974. Recovery of residual organic matter from in vitro digestion of forages. J. Dairy Sci. 68: 27322736. Statistical Analysis System (SAS). 1985. SAS Proceedures Guide for Personal Computers, Release 6 Edition. SAS Institute Inc, Cary, NC. Williams, M. J. 1988. Potential of some tropical forage legumes for Florida's sand ridge. Soil Crop Sci. Soc. Florida Proc. 47:184188. Williams, M.J., C. G. Chambliss,]. B. Brolmann, and S. L. Sumner. 1991. Dry matter production and quality of Line 8400 stylo in Central Florida. p. 61-64. In Forages: A versatile resource. Proc. Forage Grassld. Conf., Columbia, MO. 1-4 Apr 1991.

PAGE 133

PROCEEDINGS, VOLUME 51, 1992 109 Assessment of Productivity and Persistence of Selected Tropical Forage Legumes A. E. Kretschmer, Jr.*, B. J. Boman, T. C. Wilson, and G. H. Snyder ABSTRACT A small-plot replicated cutting experiment was conducted from 1979 to 1983 to select several tropical forage legumes for further evaluation under grazing conditions. The companion grass was 'Pangola' digitgrass (Digitaria decumbens Stent.). A rating system was devised to evaluate the 100 entries for legume vigor (L V) and legume plot coverage (LC). Genera included were Calopogonium, Centrosema, Desmanthus, Desmodium, Indigofera, Macroptilium, Neonotonia, Stylosanthes, Teramnus, Vigna, and Zornia, In June and September, after harvesting selected plots with a small-plot sicklebar mower, aerial forage was removed from all plots using a flail-type harvester. Yield and forage digestibility and crude protein data were obtained from selected, highest LC plots to determine productivity. Rating L V was more difficult than LC because of the large difference in growth habits. At the end of 1982, 45 entries were rated as having A. E. Kretschmer, Jr., B .J. Boman, and T. C. Wilson, Univer sity of Florida, IF AS, Agricultural Research and Education Center, P. 0. Box 248, Ft. Pierce, Florida 34954; and G. H. Snyder, Everglades Research and Education Center, P. 0. Box 8003, Belle Glade, Florida 33430. Florida Agric. Exp. Stn. Journal Series no. N-00582. *Corresponding author. Contribution published in Soil Crofi Sci. Soc. Florida l'ror. 51: I 09-116 (1992) 20% or less LC, and 25 with no plants, Legume vigor within a genus did not necessarily relate to LC (persistence). Legume coverage and L V ratings for climbing legumes were not related to legume yield or total yield, while for erect legumes LC, but not L V ratings, were correlated with legume yield. Persistence under grazing may not be related to persistence under clipping. Seventeen legumes were mostly selected from the experiment based on persistence under clipping and edaphic considerations. When placed under grazing evaluations, only seven were thought worthy of further research. Evaluation of tropical forage legumes at the University of Florida's IF AS Agricultural Research and Education Center, Ft. Pierce (ARECFP) began in the early 1960's (Kretschmer et al., 1992). Since that time, initial and secondary evaluations have been done with single-plant plots, plots of legumes alone, and with legume-grass mixtures under clipping and grazing. Selection criteria of the better legumes shifted from the higher yielding to the more persistent types. Information has been obtained on insect and disease problems and on fertility and liming requirements. During the l 970's manv of the newer

PAGE 134

110 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA legumes appeared to have the adaptability and persistence worthy of further evaluation. The object of this investigation was to further evaluate the better, more persistent previously-tested entries and newer accessions of legumes compared with several commercial cultivars. A visual rating system of legume coverage (LC) and legume vigor (L V) was used to estimate legume persistence, and selected plots were harvested for yield to complement and as sess the value of the ratings system. MATERIALS AND METHODS One-hundred thirty-six tropical legume acces sions were seeded (6 Aug. 1979) immediately after planting 'Pangola' digitgrass in an Oldsmar fine sand (sandy, siliceous hyperthermic family of Alfie Arenic Haplaquods) at the ARECFP. The "hardpan" layer (spodic horizon) was located about 1 m below the surface and contributed to temporary flooding during periods of high rainfall. Normally soil moisture was maintained above the spodic horizon during the dry periods because of a water table maintained in surrounding ditches. The cultivated area, soil pH of 6, had received adequate lime, P, and micronutrients in the past. Plots were fertilized annually in the spring with about 300 kg ha-1 of0-4.4-16.6 (N-P-K) mixture. Because 36 Aeschynomene accessions were included in this experiment and reported earlier (Kretschmer et al., 1986), only data on 100 legumes, including 28 species, are presented (Table 1). Four replications of each legume were seeded in plots 1.5 by 2.0 m in a randomized complete block design. Data on legume growth habit were separated into two broad groups, climbing (ie. 'Siratro') and erect (non-climbing, ie. Stylosanthes). Legume Coverage (LC) and Vigor (L V) Visual LC ratings were made eight, and L V six times during the 1979-83 experimental period. The accu racy of the ratings obtained in June 1983 was hindered by the encroachment of legumes from one plot to another and are not included in this study. A scale of 1 to 9 was used to subjectively note LC of the plot: 1 = below 10%, 2 = 20, 3 = 30, 4 = 40, 5 = 50, 6 = 60, 7 = 70, 8 = 80, and 9 = 90%. The scale for LV was I (very poor or no plants), 3 (poor), 5 (moderate), 7 (good), and 9 (excellent). Standard ANOV A and correlation procedures for LC and LV ratings and selective yields were used to compare accessions (SAS, 1985). Harvested Plots In June and September of 1980, 1981, and 1982, a sicklebar mower with an effective cutting width of 0.8 m and cutting height of 10 cm was used to harvest selected plots having the highest LC values (see Kretschmer et al., 1986 for more details). In many instances, not all replicates of an entry were harvested for yield and quality. Thus, statistical analysis was not used to determine differences. Legume and grass Table 1. Tropical forage legumes included in the field experi ment at the ARECFP. Legume (IRFL No.) Arachis sp. (2273) Legume (IRFL No.) Marroptilium atropurpureum (DC.) Crb. (483, 2117, 2118, 2206, 2207, 2208,2209,2211,2212,2213,2214, 2937, 2938) Calopogonium mucunoides Neonotonia wightii (Wight and Arn) Desv. (1545) (383, 1929) Centrosema pascuorum C. Mart. (I 940, 2143, 2281, 2950, 295l,2952,2953,2956) Centrosema pubescens Benth. (380,975, 1687, 1743, 2150, 2441) Centrosema schottii (Mill.) Schuman (2172, 2174, l"i'!O) Centrosema virginianum (I,.) Benth. (1526, 1935, 2029, 2948, 2949) Centrosemasp. (1939, 2126) Desmanthus virgatus ( L.) Willd. (474, 1857) Desmodium adscendens (Sw.) DC. (2058) Desmodium barbatum (L.) Benth. (907, 1208, 1737, 1923, 1979,2007,2038, 2042,2051,2077,2197, 2198,2199,2200,2201,2202, 2203, 2470) Stylosanthes guianensis (Ab.) Sw. (1805,2004,2024,2025, 7035, 7040, 7162,7919, 7935, 7937) Stylosanthes hamata (L.) Taub. (1233, 7303) Stylosanthes scabra J. Vogel (7938) Stylosanthes sympodialis Taub. (2271) St,lo.rnlhes ttubeerculata S. F. Blake (2'.l54)t Stylosanthes viscosa (L.) Sw.(I712, I 713) Teramus labia/is (L.F.) Spregle (141, 1842, 1843, 1845, 1846, 1847, 1848, 2130) Desmodium heterocarfNm (L.) Temmus uncinatus (L.) Sw. (1550) DC. (588, 1699, 1946) Desmodium intortum (Miller) Urb. (1022) Desmodium sp. (2276) Indigo/era sp. (725) tshort growing S. scabra type Vigna adenantha (G. Mey.) Marechal et al. ( 1806, 2138) Vigna luteola]acq. (2127) Zornia latifolia Smith (2045) components were dried in a forced-draft oven at 60C, and standard procedures were used to determine component yields, crude protein (CP), and in vitro organic matter digestibility (IVOMD) (Fethiere et al., 1992). Entries were rated before harvesting in June and September of 1980, 1981, and 1982, except after harvest in November 1982. After each harvest, the entire plot area was cut with a commercial 2.2 m wide flail-type harvester, and cut material was blown into a dump truck for removal. Flail stubble heights were 10 cm in June and 2.5 cm for September harvests. The very low fall cutting height was designed to place a severe biological stress on those entries having higher than cutting-height crown or stem-bud regenerative areas, or on late-flowering annuals not having produced mature seeds by the September harvest. RESULTS AND DISCUSSION LEGUME COVERAGE AND VIGOR Average LC ratings for all legumes increased from approximately 30% in 1979 to a maximum of about 40-50% in 1980 to June 1981, and then plateaued at about 30% in 1982 (Table 2). A better understanding of the decreasing plant coverage is shown by the number of entries with less than 20%

PAGE 135

PROCEEDINGS, VOLUME 51, 1992 111 Table 2. Mean legume plant coverage (LC) and vigor (L V) ratings from 1979 to 1983 and number of entries with populations rated below 2.0 or with no plants. Number of Entries* Date LCt LV LC rating below 2.0 no plants 1979 30 Nov 2.8 8 1980 13 June 4.6 5.9 18 Sept 4.7 4.6 5 0 1981 15 June 4.0 3.8 2 Sept 3.3 3.1 46 18 14 Dec 2.8 1982 !June 2.6 2.8 29Nov 3.1 3.2 45 25 tTotal of I 00 entries included. t I = less than I 0%, 5 = 50%, and 9 = 90% legume coverage of IJ!Ot. I = very poor or no plants, 5 = moderate, and 9 = excellent legume vigor. coverage and those without plants for the fall of 1980, 1981, and 1982. About half of the entries had LC ratings of less than 20% at the end of the experiment. On the other hand, coverage of several entries from several genera increased from planting and maintained adequate populations throughout the experimental period (see individual legume discussions). Legume vigor ratings may be of value when plants with similar growth habits are evaluated, but with the different growth forms included in this experiment, vigor was very difficult to assess. For example, a vigorous short-growing V. parkeri and a robust V. adena_ntha are difficult to assess when even a poor growmg V. adenantha appears more productive than a vigorous V. parkeri. Because of this, less emphasis was placed upon the L V ratings. Climbing Legumes Centrosema species. Centrosema pubescens, C. pas cuorum, and C. schottii entries performed poorly (Fig. 1 ). Five C. virginianum entries, especially IRFL 1935, were more persistent, if less vigorous, than commercial C. pubescens. IRFL 483 cv. Siratro maintained O_,_-+----+------l-----1-----t--------+---------<,--J Nov 79 Jun Sep 80 80 Jun Sep 81 81 Jun 82 Nov 82 Fig. 1. Centrosema spp. plot coverage ratings during 1979-82. Number of entries evaluated are in parentheses. c,, ~6 0 a:: ., E :::, -3 3 Geeeo a~ropurpureum IRFL 483 t3BBBEI Neonotonio (2) O -Calopogonium ( 1) Nov 79 Jun Sep 80 80 (U) Jun Sep 81 81 Jun 82 -----Nov 82 Fig. 2. Macroptilium atropurpureum, Calopogonium mucunoides, and Neonotonia wightii plot coverage ratings during 1979-82. Number of entries evaluated are in parentheses. higher plot coverage than did commercial centro IRFL 1687 or IRFL 1935 (Fig. 1,2). Calopogonium rnucu.noides. Coverage of the single accession increased dramatically in 1980 but fell rapidly to almost no plants in 1982 (Fig. 2). Macroptilium atropurpureum. Average plot coverage for the 13 en tries increased in 1980 to about 60 to 70% where they remained. Legume coverage ratings for alternative entries were no better than that for Siratro (Fig. 2). Neonotonia wightii. The two entries failed to maintain a significant population past 1980 (Fig. 2). Vigna species. Of the two V. adenantha entries, IRFL 1806 had higher LC ratings in December 1981 and June 1982 (P<0.05). Initial stand of IRFL 1806 was significantly less than Siratro, and Siratro was rated significantly higher (P<0.05) four out of the seven rating dates (Fig. 3). AV. luteola entry had a LC rating from about 50 to 70% in 1980-1981 but decreased rapidly to insignificant coverage by 1982 (data not shown). Vigna parkeri cv. Shaw a short, creeping legume, was present throughout the experiment, covering almost 50% of the plot by the fall of 1982 (Fig. 3). It was equal in coverage to V. adenantha for five of the rating dates, however, L V ratings were much less. Teramnus species. Teramnus uncinatus LC remained at about 40% until 1981 when plants disappeared. Teramnus labialis, a short creeping-twining plant, per-Nov 79 Jun Sep 80 80 Jun Sep 81 81 Fig. 3. Vigna adenantha, V. parkeri, and Siratro plot coverage ratings during 1979-82.

PAGE 136

112 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA formed poorly. This was due in part to the poor ini tial stand of plants for the eight entries. Twenty to 35% plant coverage was maintained until 1982, after which only a few plants remained (data not shown). Erect Legumes Arachis sp. This species, described by Kretschmer and Wilson ( 1988), survived well with coverage beginning at about 40%, rising to 60 to 70% in 1980-81, and falling to 40 to 50% during the remainder of the evaluation (Fig. 4). Desmodium species. The 18 D. barbatum entries were moderately effective in maintaining plant coverage, with IRFL 1923 being superior (Fig. 4). Because of its obvious adaptability, the species was more thoroughly evaluated later (Kretschmer et al., 1990). Of the three D. heterocarpon entries (two "ovalifolium" types), 'Florida' carpon desmodium IRFL 588 (Kretschmer et al., 1979) maintained superior coverage (Fig. 4). Desmodium intortum cv. Greenleaf maintained good coverage until the fall of 1981 when the decline of LC was dramatic (data not shown). Desmanthus virgatus. In spite of poor germination of the two entries, especially in one of the replica tions, initial 20% average plant coverage rose to about 30% where it stabilized (Fig. 4). Stylosanthes species. Average legume coverage rating of S. guianensis is shown in Table 5. Stylosanthes guianensis IRFL 2024 and 7040 had the highest LC of the 10 entries evaluated (data not shown). Plant coverage was similar for S. hamata accessions IRFL 7303 and cv. Verano IRFL 1233, ranging from about 30, 60, and 20% for 1979, 1980-81, and 1982, respec tively (Fig. 5). Stylosanthes scabra and S. sympodialis plant coverage was poor, especially in 1981-82. Stylosanthes tuberculata plant coverage was good throughout the experiment (Fig. 5). Survival of the two S. viscosa entries is known to be poor. Zornia latifolia. IRFL 2046 LC was very good, and was consistently higher than IRFL 996 from the ini tial establishment (data not shown). Coverage of IRFL 2046 was equal to or better than that of Stylosanthes and Desmanthus spp. throughout the rating period, although productivity was less. Q) E ::, :? J _J Nov 79 Jun Sep 80 80 ~Ar'Jd1i:o "'f-'-IRF. 2/..l~~ ..... ooroot,,m IRFL I 92J GBBBEl -Jun Sep 81 81 ck:~,r,licl .. .nn R~1_ '.Jl:38 lkl;-L 474 Jun 82 Nov 82 Fig. 4. Arachis, Desmodium barbatum, 'Florida' carpon desmodium, and Desmanthus virgatus plot coverage ratings during 1979-82. 9,--------------..------------------, "' '.35 a Ct'. Q) E ::, :? 3 _J Nov 79 Jun Sep 80 80 Jun Sep 81 81 Jun 82 Nov 82 Fig. 5. Stylosanthes guianensis, S. hamata, S. tuberculata, and Zornia latifolia plot coverage ratings during 1979-82. Number of entries evaluated are in parentheses. HARVESTED PLOTS Yield Aerial regrowth from September harvests was killed by frosts or freezes during each cool season, so that June yields represent regrowth from March in the first two seasons and from February in the 198182 season (Fig. 6). Even so, night temperatures were not consistently above 15 C until April or May at which time rapid legume growth commenced. There were 47, 46, and 35 days of 15 Corless minimum temperatures in March, April, and May for the 197980, 1980-81, and 1981-82 seasons, respectively. Data from selected, higher-yielding plots and those with several harvests are presented in Table 3. Best June yields were produced by S. guianensis IRFL 7035, but cutting at 10 cm damaged plants and slowed summer regrowth. Therefore they were not selected for harvesting in September as high-yielding legumes. Because there was incomplete harvest data of all legumes for the three years, it is difficult to make definitive statements concerning the yields. However, by eliminating entries IRFL 474, 1687, 7035, and 7303, and averaging legume and total yields of the remaining group, certain trends are evident. Average June legume yields for 1980, 1981, and 1982 were about 820, 940, and 720 kg ha-1 respectively; respective yields for September were 1860, 1260, and 722 kg ha-1 (note that many of the highly yielding entries were not harvested in 1982). Total yields (legume plus Pangola) in 1980, 1981, and 1982 u ,:_,25 1979/80 1980/81 1981/82 Fig. 6. Minimum daily winter and spring temperatures during 1979-82.

PAGE 137

PROCEEDINGS, VOLUME 51, 1992 113 Table 3. Legume and total legume plus Pangola oven dry weight yields. June Sept June Sept .June Sept lRFLt 1980 1980 1981 1981 1982 1982 No. LY* TY LY TY LY TY LY TY LY TY LY TY -------------------------------------------------------------------kg ha-1 -------------------------------------------------------------------------------------483 565 1530 827 1995 1116 2304 1644 5472 I I 78 2793 588 1287 2584 2252 2870 1781 3249 1378 4579 1055 2258 1806 1150 3468 4413 6750 784 2380 1278 4327 983 2836 474 180 1183 955 1682 2195 4033 1952 3905 1435 2930 2046 257 675 1539 2366 147 1069 741 1363 247 1677 1316 2556 1923 1496 2760 774 2427 931 2973 580 2522 2273 247 1154 903 1948 413 1320 831 3206 508 2812 594 1857 1687 1520 2380 2520 4285 2470 3757 7035 3914 5391 4446 6156 1150 3306 1935 1525 2717 2518 4'103 1283 2366 1159 3221 1311 2736 399 1819 7303 736 1321 537 1135 542 1221 2242 3724 t ARECFP accession number: 483 = Macroptilium atrop11rpureum cv. Siratro, 588 = Dr
PAGE 138

114 Sou, AND CROP SCIENCE SOCIETY OF FLORIDA Table 5. IVOMD percentages of tropical legumes and associated Pangola digitgrass. June Sept .June Sept June Sept IRFI! No. 1980 l 980 198 l l 981 1982 1982 x ---------------------------% ----------------------------483 Pangola 588 Pangola 1806 Pangola 474 Pangola 2046 Pangola 1923 Pangola 2273 Pangola 1687 Pangola 7035 Pangola 1935 Pangola 7303 Pangola x (legumes) x (Pangola) x (legumes}* 56.8 46.3 29.2 37.5 62.4 45.1 24.2 54.3 52.5 52.3 75.8 55.0 49.2 43.l 45.5 34.8 60.3 45.8 50.7 46.7 54.0 50.2 42.l 34.8 43.5 56.9 42.5 51.7 36.4 73.0 43.5 50.2 44.0 48.4 41.7 48.9 36.8 45.9 36.3 45.9 57.9 58.0 59.l 44.2 45.1 52.2 38.2 41.5 42.4 48. l 44.0 41.2 53.0 48.3 66.8 -~6.6 66.l 44.7 42.9 53.6 19.7 23.2 30.3 38.3 41.2 41.9 49.5 49.7 58.2 50.5 66.7 39.2 48.4 45.3 53.5 53.9 44.8 40.8 46.4 47.7 46.0 46.7 46.4 52.9 7 l. l 62.2 66.7 73.8 49.8 42.5 54.5 54.3 49.3 38.2 48.8 52.3 45.3 53.8 53.0 51.3 55.7 56.4 46.3 38.0 53.6 51.9 58.8 26.6 43.7 37.l 49.9 46.3 54.2 50.6 44.7 43.5 52.0 51.8 52.9 48.9 57.6 53.l 1ARECFP accession number (sec Table 3 for names). + Excluding 474, Desmanthus virgatus. 56.4 46.0 39.0 44.6 61.8 45.8 27.l 47.l 53.1 48.3 44.9 48.0 70.4 49.9 49.6 41.8 50.1 49.6 51.7 44.4 48.7 40.9 stem content of the samples. Excluding IRFL 474, mean legume IVOMD values were higher in June (mean for three June harvests = 54.8%) than that for three September harvests (49.3%). Correlations Correlations were made for LC, LV, legume yield, total yield, and growth habit (Table 6). Except for climbing legume yield and total yield, there were no other significant relationships between variables measured. Erect LC rating, however, was closely correlated to legume yield, and LV, as was erect legume yield and total yield (Table 6). This effect of erect legumes was sufficiently strong to provide significant correlations when all legumes are considered. Growth habit and LC were not significantly correlated, while L V was negatively correlated to growth habit. From these data it appears that LC or L V ratings of climbing legumes would have little predictive value for estimating total yield. From a practical standpoint, neither L V nor LC ratings may have much relation-Table 6. Correlation coefficients (r) of various legume coverage (LC) and vigor (L V) ratings, and yield. Correlation LC rating vs. legume yield L V rating vs. legume yield LC rating vs. total yieldt LV rating vs. total yield Legume yield vs. total yield LC vs. LV Growth habit vs. LC Growth habit vs. L V All Climbing Erect legumes n = I 9 n = 29 n = 48 ------------------r -------------------NS+ NS NS NS 0.84** NS 0.52** NS NS NS 0.87** o.s3 0.34** NS NS NS 0.86** 0.35* NS -0_39 ** = Significant at the 5% and I% levels of probabilities, respectively; NS = not significant. tLegume plus Pangola yield. ship to persistence under grazing as evidenced by the known poor persistence of M. atropurpureum under grazing and the excellent persistence of this species under cutting regimes (Jones, 1988). GENERAL DISCUSSION A field experiment with 100 tropical legume ac cessions in a conventional randomized block design with harvesting twice a year requires a large input of time and labor. This method relates to a green chop or haying system where yields and quality factors can be directly related to the growers' management program. Rating data obtained in this experiment is of less value when legume persistence under grazing is the most important element to measure. Cutting may not predict persistence of legumes with a large diversity of growth habits under grazing conditions. This subject has been discussed previously (Kretschmer, 1989; Hodgkinson and Williams, 1983; Jones and Walker, 1985; Grof, 1986). Siratro (M. atropurpureum) is an excellent example of persistence under clipping, yet poor persistence under grazing (Jones, 1988). Stylosanthes guianensis persists well under grazing but may not persist well when growth permits plant heights to reach 0.6 to 1.0 m or more prior to low cutting height (Kretschmer and Snyder, 1982). Carpon desmodium has persisted well in this and other experiments under clipping and under various grazing managements in mixture with various grasses (Kretschmer et al., 1979), even in mixtures with the various competitive bahiagrass (Paspalum notatum Fluegge) cultivars for more than ten years. A modified rating-clipping system was used in this experiment to identify and detect legumes for future testing under grazing. The selections were made more on a basis of plant coverage ratings than on yields. The constraint of very low cutting height in September may have eliminated from consideration some of the legumes. Criteria used for selecting legumes (Table 7) for additional evaluation under commercial grazing management in 1983 were based on good LC at the

PAGE 139

PROCEEDINGS, VOLUME 51, 1992 115 Table 7. List of legumes selected in 1983 for continued evaluation under grazing for various attributes and 1982 legume cover (LC) ratings. LC Establishment t l 'olerance t Nov. Legume and (IRFL No.) ease persist. Flood Drought 1982 Florida carpon desmodium (588) 5 5 3 3 7.8 Siratro (483) 5 1 2 5 8.0 Florida commercial aeschynomene ( 4 70) 5 3 5 2 1.0 Aeschynomene americana (I 725, I 726)" :1 3 4 4 6.5 Aeschynomene flumenensis (2854) 2 I 5 2 Aeschynomene histrix (2891)' 1 1 2 3 1.0 Aeschynomene villosa (2328, 2331, 2927, 4531)" 3 3 2 4 1.3 Arachis sp, (2273) 5 4 5 4 4.8 Alysicarpus vaginalis (3240)* 3 4 3 .S Centrosema virginianum (1935) 4 5 3 3 4.8 Desmanthus virgatus (4 74, 1857) I 2 2 5 '.l.8 Desmodium barbatum ( 1923) 4 ,\ 3 3 6.8 Lotononis bainesii (406) 1 4 2 5 Teramnus labialis (1841, 1848)* 2 I 5 1.3 Vigna adenantha ( 1806)" 5 2 3 2 6.0 Vigna parkeri (2977)" 3 2 3 2 4.8 Zornia latifolia (2046) 3 '.1 'l 5 6.0 t1 = poor, 3 = moderate, 5 = excellent "In spite of poor growth in this experiment, Teramnus labia/is was selected because of its persistence in mixture with stargrass in the wet-dry climate of Guanacaste, Cost Rica. Aeschynomene histrix is very leafy in the fall. Aeschynomene vagina/is was selected from other observations because of its perennial, prostrate stoloniferous growth habit. Vigna parkeri, and V. adenantha are being studied under grazing at the Ona AREC. Alysicarpus vagina/is, IRFL 3240 was not included in the present experiment. Aeschynomene spp. (IRFL 2929, IRFL 4531, and IRFL 2854) were not included in this discussion but were selected from previous published research (Kretschmer et al., 1986). end of the experiment, or based on persistence under grazing and edaphic adaptability from other research results. Since 1983, the list has been reduced because of inadequate establishment and/or lack of persistence under grazing. There has been an edaphic problem (causing a yellowing of foliage) with D. vir gatus and, except for two sites, establishment has been poor even after adequate seedling populations were attained. Arachis IRFL 2273 has proven to be moderately persistent (Kretschmer and Wilson, 1988) and has led to recent comparisons with 12 other nut-producing entries. A newly developed seed-harvesting method (unpublished) has led to increased interest in this Arachis. Centrosema virginianum IRFL 1935 has proven to have moderately good persistence in spite of seed scarcity for planting and low establishment plant populations, and only moderate productivity. Desmodium barbatum IRFL 1923 was promising enough in this experiment to evaluate a large number of entries (Kretschmer et al., 1990), however, Florida carpon desmodium IRFL 588 was more persistent and productive than IRFL 1923. Teramnus labialis has not established nor persisted well in spite of its persistence under grazing combined with grasses in a wetdry environment in Costa Rica. After testing in primarily strip-plot tests during the past eight years, the list of better legumes has been reduced to Arachis IRFL 2273, A. villosa entries from the previous experiment (Kretschmer, et al., 1986), a prostrate and perennial type of Alysicarpus vaginalis IRFL 3240, C. virginianum IRFL 1935, Vigna adenantha IRFL 1806, V. parkeri IRFL 2977, and Z. latifolia IRFL 2046. Siratro, commercial A. americana, and carpon desmodium (all with different growth habits) always were included as controls in the grazing tests. Methods to reduce time and seed quantities were discussed in some detail by Kretschmer ( 1989). To reduce time from acquisition to use for grazing, entries must be grazed as soon after the initial year's evaluation and seed enhancement as possible. Establishment year plants can be grazed the year after seed increase if provision for animals and proper planting sites are chosen in advance. REFERENCES Fethiere, R., L. E. Sollenberger, and J. E. Moore. 1992. Forage evaluation supports laboratory methods. Univ. Florida, Animal Nutrition Lab. Bldg. 477, Gainesville, FL 32611. Grof, B. 1986. Forage potential of some Centrosema species in Llanos Orientales of Colombia. Trop. Grassl. 20: 107-112. Hodgkinson, K. C., and 0. B. Williams. 1983. Adaptation to grazing in forage plants. p. 85-100. In J. G. Mel vor and R. A. Bray (ed.) Genetic resources of forage plants. CSIRO, East Melbourne, Australia. Jones, R. M. 1988. Inspection of old species evaluation trials after twenty years of farm grazing. Trop. Grassl. News 4:4-9. Jones, R. J., and B. Walker. 1983. Strategies for evaluating forage plants p. 185-201. In]. G. Mcivor and R. A. Bray (ed.) Genetic resources of forage plants. CSIRO, East Melbourne, Australia. Kretschmer, A. E.,Jr. 1989. Tropical forage legume development, diversity, and methodology for determining persistence. p. 117-l 18. In Marten, G. C., A. C. Matches, R. F Barnes, R. W. Brougham, R. J. Clements, and G. W. Sheath (ed.) Persistence of forage legumes. Proc. of a Trilateral Workshop, Honolulu, Hawaii. 18-22Juh 1988. Am. Soc. Agron., Madison, Wisconsin. Kretschmer, A. E., .Jr., J. 13. Brolmann, G. E. Snyder, and S. W. Coleman. 1979. 'Florida' carpon desmodium (Desmodium heterocarpon (L.) DC.) a perennial tropical forage legume for use in south Florida. Florida Agric. Exp. Stn. Circ. S-260.

PAGE 140

116 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Kretschmer, A. E., .Jr., J. B. Brolmann, R. M. Sonoda, G. H. Snyder, R. C. Bullock, and B.J. Boman. 1992. A tropical forage crops bibliography from lFAS, University of Florida, Agric. Res. and Educ. Center, Ft. Pierce, Florida (ARECFP). Ft. Pierce AREC Res. Rep. FTP-92-2. Kretschmer, A. E., Jr., R. C. Bullock, and T. C. Wilson. 1990. Evaluation of Desmodium barbatum (L.) Benth., a tropical forage legume. Soil Crop Sci. Soc. Florida Proc. 49:204-206. Kretschmer, A. E., Jr., and G. H. Snyder. 1982. Comparison of mixtures of seven tropical legumes and six tropical grasses in south Florida. Soil Crop Sci. Soc. Florida Proc. 41:67-72. Kretschmer, A. E., .Jr., G. H. Snvder, and T. C. Wilson. 1986. Productivity and persistence of selected Aeschynvmme spp. Soil Crop Sci. Soc. Florida Proc. 45:174-178. Kretschmer, A. E.,Jr., and T. C. Wilson. 1988. A new seed-producing Ararhis sp. with potential as forage in Florida. Soil Crop Sci. Soc. Florida Proc. 34:63-66. SAS Institute Inc. 1985. SAS Users Guide; Statistics, Cary, NC.

PAGE 141

116 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Evaluation of "Tropical" Corn as a Feedstuff for Swine R. O. Myer*, D. L. Wright, L. E. Anderson, J. H. Brendemuhl, G. E. Combs, and D. W. Gorbet ABSTRACT The feeding value of a tropically-adapted corn (Zea mays L.) hybrid (Pioneer X-304C) was evaluated as the primary energy feedstuff in typical swine diets. Six different trials, three involving growing-finishing swine (28 104 kg; n == 108 total), one involving finishing swine (56 106 kg; n == 96), and two involving young, starting swine (7 26 kg; n == 216 total), were conducted. Each trial involved direct comparisons of nutritionally adequate diets containing either tropical or conventional temperate hybrid corn. Corn crops of both types grown in three successive years (1988, 1989, 1990) at the same location were tested. Results of the feeding trials indicated that the feeding value (energy value) of the tropical corn used was estimated to be 90 to 95% that of the conventional corn used for young starting pigs and 95 to 100% for older growing-finishing pigs. The feed grain of choice for swine feeding in the southeastern USA is corn. Corn production in the southeastern USA is inadequate to carry the present swine and poultry populations, thus, these industries are dependent on more expensive imported corn. Corn hybrids that are potentially better suited to warm, humid subtropical and tropical environments are becoming available. The potential of and agronomic characteristics of "tropical corn" for the southeastern USA have been previously reported (Goldsworthy et al., 1974; Lilly et al., 1992; Wright et al., 1988; Lilly, 1991). Tropical corn is similar to conventional temperate yellow corn hybrids in color and appearance, however, tropical corn kernels tend to be smaller with the seed coat and endosperm being much harder than conventional corn (Wright et al., 1988). This combination of harder seed coat and endosperm may influence the feeding value of tropical corn grain for swine. Preliminary research has shown that diets formulated with tropical corn have a feed-R. 0. Mvcr and D. W. Gorbet. :'.\orth Florida, Res. and Educ. Center, Ma~ianna, FL 32446-7906; D. L. Wright, North Florida Res. and Educ. Ctr., Quincy, FL 32351-9529; L. E. Anderson, J. H. Brendemuhl, and G. E. Combs, Animal Science Dep. Univ. of Florida, Gainesville, FL 32611-0910. Florida Agric. Exp. Stn. Journal Series no. N-00587. *Corresponding author. *Contribution published in Soil Crop Sci. Soc. Florida Proc. 51: 116-119 (1992) ing value slightly less than diets formulated with conventional corn for swine (Williams et al., 1984). This study was conducted to evaluate the feeding value of a tropical corn hybrid, compared to conven tional temperate corn, when used as the primary energy feedstuff in typical swine diets. MATERIALS AND METHODS Six swine feeding trials were conducted, three with growing-finishing swine, one with finishing swine and two involving young, starting swine. Each trial involved a direct comparison of diets containing either conventional temperate or tropical corn. Three different year's crop (1988, 1989, 1990) of both types of corn were used for each of the three growingfinishing trials. The finishing trial used corn from the 1988 crop and the two starting trials used corn from the 1988 and 1989 crops. All trials, except for the finishing trial, were conducted at the NFREC in Marianna; the finishing trial was done at the University of Florida swine research unit in Gainesville. The temperate and tropical corns involved in these trials were grown under irrigation at the NFREC in Marianna. The hybrid of tropical corn grown each year was Pioneer Brand X304C. The conventional temperate corn for each year was a blend of four to six different hybrids which were grown in strips in the same field and were blended at the time of harvest. Each year, the temperate corn was planted in March and harvested in August and the tropical corn was planted in June and harvested in November. Representative samples of the tropical and blended temperate corns from each year's crop were analyzed for amino acid composition at a commercial laboratory (AAA Laboratory, Mercer Island, WA) with an amino acid analyzer (Model D-500, Durrum, Palo Alto, CA) according to manufacturer's recommended procedures. Crude protein, crude fat (ether extract), crude fiber, and ash contents were also determined using accepted procedures (AOAC, 1984). For each of the growing-finishing swine trials, 36 crossbred pigs with an average initial weight of 28 kg

PAGE 142

PROCEEDINGS, VOLUME 51, 1992 117 were used. Pigs were divided by sex, weight, and litter origin into pens of six pigs each. Each pen was assigned at random to one of the two dietary treatments (temperate or tropical corn) within each of three re plicates (blocks). Pigs were fed grower diets (Table 1) to an average pen weight of 56 kg per pig and then were switched to lower protein finisher diets to an average pen weight of 104 kg per pig. Pigs were housed in an open-sided building, in 2 x 4.5 m pens with solid concrete floors. During the second and third week of the growing phase and the third and fourth week of the finishing phase, chromic oxide was added to the diets at a level of 0.2%. After a 5 d adjustment period, "grab" fecal samples were obtained from at least five of the six pigs per pen for each of four consecutive days. Fecal samples were dried, ground, pooled by pen, and analyzed, along with the corresponding feed samples, for chromic oxide (Christian and Coup, 1954), dry matter (AOAC, 1984), and gross energy (adiabatic calorimeter, Isoperibol Bomb Calorimeter, Parr Instrument Co., Moline, IL). Dry matter and energy apparent digestibility coef icients were calculated (Maynard and Loosli, 1969). For the finishing trial, 96 crossbred pigs with an average initial weight of 56 kg were used. Pigs were divided by sex, weight and litter origin into pens of six pigs each. Each pen was assigned at random to one of the two treatments within each of eight repli cates (blocks). Pigs were fed finisher diets (Table l) until an average final body weight of 106 kg. Pigs were housed in an open-sided building, in 2 x 5 m pens with solid concrete floors. In each of the two starting pig trials, 108 crossbred pigs, with an average initial weight of 7 kg, were utilized. Pigs were assigned to pens of six pigs each by sex, initial weight, and litter origin. Within each of nine replicates (blocks), pens were assigned randomly to one of the two dietary treatments. Pigs were housed in an enclosed nursery in pens with expanded-metal floors that had a wood overlay across one end of each pen. Supplemental heat for the pigs was provided by infra-red lamps. Pigs were fed experimental diets for 35 d. Within each of the six trials, the two dietary treatments each contained one of the two corn types as the primary energy source. Composition of all experimental diets used is given in Table 1. Diets within Table 1. Composition of experimental diets. Ingredient Grouwr (28 to 56 kg) Growing-finishing trials (Marianna) Finisher (.59 to 104 kg) Finishing trial (Gainesville) Finisher (56 to 106 kg) Starting trials (Marianna) Starin (7 to 26 kg) --------------------------------------------% ------------------------------------------Ground graint Soybean meal (48%) Dicalcium phosphate Calcium carbonate Salt Vitamin premix Trace mineral premix Salenium premix Antibiotic premix Calculated composition:j::j::j:: Lysine Calcium Phosphorus tTemperate corn or tropical corn. 78.00 19.00 1.30 1.00 0.30 0.20:j: 0.05tt 0.15## 100.00 0.78 0.74 0.58 84.30 13.00 1.20 1.00 0.30 0.15 0.05tt 100.00 0.60 0.54 0.54 84.20 66.15 13.00 30.00 1.55 1.50 0.75 1.00 0.25 0.50 0.15,J 0.25# 0.15:j::j: 0.10 0.05,i,i --0.5Qitt 100.00 100.00 0.60 1.09 0.70 0.80 0.60 0.65 tProvided the following per kilogram of diet: vitamin A, 4400 IU; vitamin D3 660 IU; vitamin E, 18 IU; vitamin K activity, 2.2 mg; riboflavin, 3.5 mg; d-pantothenic acid, 13 mg; niacin, 18 mg; choline chloride, 440 mg; and vitamin B12, 18 ,g. Provided the following per kilogram of diet: vitamin A, 3300 IU; vitamin D3 530 IU; vitamin E, 13 IU; vitamin K activity, 2.0 mg; riboflavin, 2.2 mg; d-pantothenic acid, 9 mg; niacin, 13 mg; chlorine chloride, 330 mg; and vitamn B12, 13 ,g. ,JProvided the following per kilogram of diet: vitamin A, 5500 IU; vitamin D3 880 IU; vitamin E, 22 IU; riboflavin, 13 mg; d-pantothenic acid, 18 mg; niacin, 44 mg; choline chloride, 176 mg; and vitamin B12, 22 ,g. #Provided the following per kilogram of diet: vitamin A, .5500 IU; vitamin D3 880 IlJ; vitamin E, 22 IU; vitamin K activity, 3.3 mg; riboflavin, 4.4 mg; d-pantothenic acid, 18 mg; niacin, 22 mg; choline chloride, 550 mg; and vitamin B 12, 22 ,g. ttProvided the following per kilogram of diet: zinc, JOO mg; iron, 50 mg; manganese, 25 mg; cooper, 5 mg; iodine, 1.0 mg; and selenium, 0.1 mg. :j::j:Provided the following per kilogram of diet: zinc, 200 mg; iron, JOO mg; manganese, 55 mg; copper. 10 mg; iodine, 1.5 mg. Provided the following per kilogram of diet: zinc, 200 mg; iron, 100 mg; manganese, 55 mg; cooper, 10 mg; iodine, 1.5 mg; and selenium, 0.20 mg. ,J,JProvided 0.3 mg selenium per kilogram of diet. ##Provided 22 mg tylosin per kilogram of diet. tttProvided per kilogram of diet: chlortetracycline, 110 mg; sulfarnethazine, I JO mg; and penicillin, 55 mg. :j::j::j:Calculated using NRC (l 988) table values.

PAGE 143

118 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA each trial were formulated to NRC ( 1988) guidelines with grain (corn) level constant across treatments. The corns were ground in a hammer mill equipped with a 0.6 cm screen before incorporation into the experimental diets. Feed and water were available to the pigs at all times in all trials while pigs were on experiment. For each of the two starting swine trials and the finishing swine trial, average daily weight gains, feed to-gain ratios, and average daily feed intakes were computed on a pen basis. For the growing-finishing trials, average daily weight gains, feed-to-gain ratios, and average daily feed intakes were determined on a pen basis for the grower phase, the finisher phase, and the entire growing-finishing period. Apparent digestibilities also were computed on a pen basis. Data from all trials were analyzed by analysis of variance for a randomized complete block design (SAS, 1985). For consistency, the growing-finishing and starter trials were analyzed and reported separately as year by treatment interactions were noted for some parameters. RESULTS AND DISCUSSION Nutrient composition of both the conventional temperate and tropical corns used from the three crop years is presented in Table 2. In general, the tropical corn crops were slightly higher in crude protein and most of the amino acids than the temperate corns. Contents of other analyzed chemical components were similar between the two corn types within each year. Performance data from the growing-finishing trials are summarized in Table 3. Performance data presented are for the entire growing-finishing period because responses to treatment were of similar magnitude during the growing and finishing phases. Table 2. Nutrient composition of tropical and conventional temp erate corns usedt. Tropical corn Temperate corn Item Yrl:j: Yr2 Yr,3 Yr! Yr2 Yr3 ---------Nutrient composition, % ---------Crude protein 10.2 10.6 11.9 8.5 9.0 8.2 Ether extract (fat) 4.0 3.2 3.4 3.6 3.3 3.4 Crude fiber 2.1 2.8 2.4 2.6 2.8 2.9 Ash 1.4 1.5 1.4 1.3 1.4 1.2 Selected amino acids1: Lysine 0.29 0.29 0.31 0.28 0.28 0.28 Methionine# 0.17 0.17 0. Hi 0.17 0.16 0.16 Threonine 0.38 0.36 0.39 0.32 0.33 0.32 tValues are expressed on a 88% dry matter basis. :j:Year l -crop harvested in 1988; year 2 -harvested in 1989; and year 3 -harvested in 1990. Percent N x 6.25. ~Tryptophan was not determined. #Cvstine was not determined. Table 3. Performance of growing-finishing swine fed diets con taining tropical or conventional temperate corn (Marianna)t. Dietary corn type Item:j: Temperate Tropical ----------------------------------------Yr 1 ----------------------------------------Avg. daily gain 0.89 0.89 Avg. daily feed intake 2.73 2.72 Feed/unit gain 3.06 3.06 ---------------------------------------Yr 2 --------------------------------------Avg. daily gain Avg. daily feed intake Feed/unit gain 0.90 2.83 3.14 0.89 2.86 3.17 ----------------------------------------Yr 3 ---------------------------------------Avg. daily gain Avg. daily feed intake Feed/unit gain 0.95 3.01 3.18 0.95 3.16 3.33 tThree pens per treatment each year with six pigs per pen; on experiment from 28 to 104 kg. Units for gain and intake = kg. :j:Treatment means for each parameter do not differ (P>0. l 0). Corn crop year; yr I = 1988, yr 2 = 1989, and yr 3 = 1990. Within each trial, average daily weight gain, average daily feed intake, and feed-to-gain ratio were not affected (P>0.10) by dietary corn type. There was, however, a trend for slightly higher feed intake and slightly poorer feed-to-gain ratio for pigs in the third trial fed the tropical corn diets using the 1990 crop when compared to the corresponding temperate corn diets. Performance data from the finishing trial, which used corn from the 1988 crop only, are summarized in Table 4. Pigs fed the diet containing tropical corn had an average daily gain, average daily feed intake, and feed-to-gain ratio that were the same (P>O. l 0) as those of pigs fed the temperate corn diet. Performance data from the starting swine trials are presented in Table 5. Within each of the two trials, pigs fed diets containing either corn type had a similar (P>0.10) rate of daily weight gain. Pigs fed diets containing tropical corn, however, tended to consume more feed and have poorer feed-to-gain ratios than pigs fed temperate corn diets. The increase in average daily feed intake was significant Table 4. Performance of finishing swine fed diets containing tropical or conventional temperate corn (Gainesville; 1988 crop only)t. ltem:j: Avg. daily gain Avg. daily feed intake Feed/unit gain Dietary corn type Temperate 0.83 2.83 3.40 Tropical 0.84 2.86 3.39 tEight pens per treatment each year with six pigs per pen; on experiment from 56 to 106 kg. :j:Units for gain and intake = kg. Corn tvpe means for each parameter do not differ (P>0.10). Corn crop vear; yr I = 1988, yr 2 = 1989. and yr 3 = 1990.

PAGE 144

PROCEEDINGS, VOLUME 51, 1992 119 Table 5. Performance of young, starting swine fed diets containing tropical or conventional temperate corn (Marianna)t. Dietary corn type Itemt Temperate l'ropical -----------------------------------------Yr It---------------------------------------Avg. daily gain 0.52 0.52 Avg. daily feed intake 0. 99 1.0 I Feed/unit gain I. 90 I. 95 -----------------------------------------Yr 2 ----------------------------------------Avg. daily gain Avg. daily feed intake Feed/unit gain 0.51 0.90 1.76 0.53 0.98* 1.87 tNinc pens per treatment each year with six pigs per pen; on experiment from 7 to 26 kg. Units for gain and intake = kg. tCorn crop year 1988 and 1989, respectively, *Corn type means differ (P <0.05). Corn type means differ (P0. l 0) by corn type within the grower or finisher diet type within each trial. The exception was with energy di gestibility of the grower diets from the third trial using the 1990 crops. Energy digestibility of the tropical corn diet was found to be slightly but significantly (P<0.05) poorer than the corresponding temperate corn diet. Results of these trials indicated that both young, starting swine and the older growing and finishing swine readily consume diets containing tropical corn. However, with young, starting swine, there is evidence that the hard, flinty nature of tropical corn may interfere with its digestibility and thus its feed value. This negative effect was not apparent with the older finishing pigs as feed-to-gain ratios were not affected by dietary corn type and apparent diges tibilities of the finisher diets were also not affected. There was some evidence of reduced digestibility and subsequently feed value with the grower diets in one of the three growing-finishing trials. During the growing phase of a growing-finishing trial the pigs are also relatively young. Similar findings, in regards to efficiency of feed utilization of diets containing tropical corn for young, starting pigs, have also been observed by White and Lopez ( 1989). Thus, the rela tive feed value of the tropical corn used was estimated to be 90 to 95% that of conventional temperate corn for young, starting swine (7 25 kg) and 95 to 100% for growing-finishing swine (25 110 kg). Table 6. Apparent digestible dry matter and energy of diets containing either tropical or conventional temperate corn used in the growing-finishing trils (Marianna)t. Grower diets+ Finisher diets+ Dietary corn type Dietary corn type Item l'emperatc Tropical Temperate Tropical -------------------------------% -----------------------------------------------------Year I ---------------------------------------Dry matter 82 82 83 83 Energy 81 82 82 83 ----------------------------------------Year 2 --------------------------------------Dry matter 80 80 83 83 Energy 79 79 83 82 ---------------------------------------Y car 3 ---------------------------------------Dry matter Energy 81 80 80 78 82 80 81 80 tindicator method: 5-day adjustment, 4-day fecal collection. Three pens per treatment with six pigs per pen; collections from at least four pigs per pen each collection day; approximate pig weight 35 to 45 kg for grower diets and 70 to 80 kg for finisher diets. tComposition of diets given in Table I. *Treatment means for grower diets differ (P <0.05). ACKNOWLEDGEMENT The assistance of Mary Chambliss, Harvey Standland, John Crawford, Richard Rogers, Alvin Boning, Jr., Brian Kidd, and Dane Bernis is gratefully acknowledged. REFERENCES AOAC. 1984. Official Methods of Analysis (14th Ed.). Assoc. Offi cial Anal. Chern., Arlington, VA. Christian, K. R., and M. R. Coop. 1954. Measurement of feed intake by grazing cattle and sheep. 6. The determination of chromic oxide in faeces. N.Z. J. Sci. Technol. 36:328. Goldsworthy, P. R., A. F. E. Palmer, and D. W. Sperling. 1974. Growth and yield of lowland tropical maize in Mexico. J. Agric. Sci. 83:223-230. Lilly, D. P. 1991. Response of tropical corn to location, planting, and nitrogen rate. pp. 1-97. Univ. Florida M.S. Thesis, Univer sity of Florida, 97 p. ----, D. L. Wright, I. D. Teare, R. N. Gallaher, and R. L. Stanley. 1992. Tropical corn vs. planting date, and water. Proc. Soil Crop Sci. Soc. Florida 51 :(in press). Maynard, L.A., and]. K. Loosli. 1969. Animal Nutrition (6th Ed.). McGraw-Hill Book Co., New York. NRC. 1988. Nutrient Requirements of Domestic Animals, No. 2. Nutrient requirements of swine. 9th Rev. Ed. NAS-NRC, Washington, D.C. SAS. 1985. SAS User's Guide. Statistical Analysis System Institute, Inc., Cary, NC. White, C. E., and F. D. Lopez. 1989. Comparison of Pioneer X-304C tropical corn and Coker 77-B temperate varieties of corn in a swine feed trial. Univ. Florida An. Sci. Res. Rep. SW-1989-3. Williams, A. C., M. T. Coffey, and G. E. Combs. 1984. Comparison of the value of feeding tropical corn with commercial yellow corn in grower-finisher diets. Univ. Florida An. Sci. Res. Rep. AL-1984-4. Wright, D. L., I. D. Teare, D .J. Zimet, and B. T. Kidd. 1988. Tropical corn as an alternate crop in S. E. United States. Univ. Florida Res. Rep. NF-88-3, pp. 1-15.

PAGE 145

120 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Leucaena: A Forage and Energy Crop for the Lower South, USA T. V. Cunilio and G. M. Prine* ABSTRACT Leucaena [Leucaena spp., especially L. leucocephala (Lam.) de Wit] is a perennial woody legume which regrows each spring from below ground rootstock in colder subtropical and warm temperate climates where temperatures below -2C occur. The crop is useful for biomass (energy), browse for cattle, and sustainable agroforestry systems. Results of several biomass and forage experiments conducted during the 1980's on a two-replication, 373-accession leucaena nursery at Gainesville, Florida, which was planted on Lake sand (hyperthermic coated Typic Quartzipsamment) in the 1979 season, are reported. In December 1990, ratings were conducted to determine plant vigor and survivabil ity of selected accessions from the nursery of the then 12 yr-old stand. Of 12 accessions selected for their high biomass produc tion in 1982, only six [Pl No. 263695 (K8), 281608 (K28), 288005 (K67), 443482 (K49) 443647 and 443696], still had good stands and vigor. Accession Pl 443655 from a forage experiment had all plants living but relatively low vigor. Overall, 20 Ieucaena accessions were selected which still had good vigor and/or stands after 12 yr. These accessions should be considered for use in future research on leucaena in Florida and the Lower South USA where good persistence and vigor are required. The value of leucaena as fuel wood (biomass), a browse and green chop forage plant for livestock, and a mulch for soil is well known (FAO, 1989). This has prompted a number of diverse studies, mostly in the state of Florida, to determine the potential use of leucaena as a forage and/or energy crop and in sustainable agroforestry systems in the warm temperate and subtropical southern, USA. These studies also addressed the principal limitations of this multipurpose tree and shrub which include winterkill in subtropical and temperate regions, mimosinal toxicity to livestock, a susceptibility to psyllid (Heteropsylla cubana Sulc.) damage, and an intolerance to acid soil and poor drainage. In colder locations, top growth of leucaena is killed by prolonged periods of subfreezing temperatures in many winters, but plants regenerate during the following spring from underground rootstocks. It is not known how far north leucaena will survive in the southern USA, but it should survive very well in warmer parts of USDA Hardiness Zone 8a and in Zones 8b, 9, and higher (USDA, 1990). In Florida, we have noticed some difference in cold tolerance in mild winters among the 373 leucaena accessions planted in the leucaena nursery at Gainesville. However, top growth of all accessions is killed to the soil surface in severe winters. Winter survival of K636 in Texas also indicates the possibility of genetic differences within L. leucocephala Q. L. Brewbaker, 1990, personal comm.). Many investigajtions have been carried out to evaluate cold tolerance of Leucaena spp. in T. V. Cunilio, Center of Sustainable Agroforestry, Inc., Gaines ville, FL 32601; and G. M. Prine, Agronomy Dep., Univ. of Florida Gainesville, FL 32611-0500. Florida Agric. Exp. Stn.Journal Series no. R-0060 I. *Corresponding author. Contribution published in Soil Cro/J Sci. Soc. Florida !'roe. 51: 120-124 ( 1992) Texas, Florida, and Mexico (Glumac, 1986; Williams, 1987; Kendall et al., 1989). Mimosine toxicity can be controlled in ruminants by ruminal inoculation with microorganisms that degrade the toxic ruminal metabolite of mimosine (Hammond et al., 1989). Limiting leucaena to less than 50% of diet will also usually control mimosine toxicity. In the warmer subtropics and tropics, leucaena can grow from a leafy plant that is high in N to a small tree and be harvested over periods of varying lengths. Twelve selected leucaena accessions (out of 373 total) in Gainesville, Florida, during four growing seasons produced an average oven-dry biomass of 31.4 Mg ha1 yr1 with an oil energy equivalent of 86 barrels oil ha-1 (Green et al., 1988). Nutrient removal for the 1982 growing season did not differ among the 12 accessions with mean amounts of 210, 30, 180, 80, and 20 kg ha-1 ofN, P, K, Ca, and Mg, respectively (Othman, 1984). Tree age and density have shown little effect on caloric values but have major influence on moisture content and the specific gravity of the wood (Brewbaker et al., 1984). In an occasional mild winter such as occurred in 1990-1991 in Gainesville the top stems of many leucaena accessions are no~ killed by freezes and growth continued for another season. Normally though, above ground growth is killed_ by freezing temperatures in the cold subtropics. Dned stems of leucaena that have been killed can be allowed to remain standing in the field as regeneration takes place the following spring so that the biomass (living and dead) produced during a 2-yr cycle can be harvested in one operation (G. M. Prine, 1991, personal observation). Dead biomass loses some weight over the season but the energy value of the remaining biomass is the same as that of recently harvested stems (Ravenswaay, 1989). Commercial-quality wood chips from mediumhardwood species are acceptable to landscapers for ~omposting or for mixing with treated sewage sludge for potting material. Chips have also been used for coburning in small institutional incinerators (Green et al., 1988). Paper has also been made from Savadortype leucaena pulp (NAS, 1977). Although a much studied agroforestry component, Leucaena spp. have not yet been fully researched for their leaf manuring potential. A study in India using oats as the manured crop however, demonstrates that grasses can quickly (within one week) use leucaena leaf N and produce normal, dry forage yields (Pathak, 1985). In a 4-yr experiment on Haplaquent soils from phosphatic settling pond clays in central Florida, Mis levy et al. ( 1989) studied soil treatments as main plots and seven biomass crops including leucaena, as the subplots. He showed that L. leucocephala was easily established independent of soil treatment, contained the highest digestibility and crude protein concentration of above ground biomass material, and, of the

PAGE 146

PROCEEDINGS, VOLUME 51, 1992 121 seven biomass crops grown (five grasses and two legumes), had the second highest dry biomass yield of 58.5 Mg ha-1 yr'. One hundred of the 373 accessions of leucaena planted at Gainesville, Florida in 1979 were studied under four different defoliation systems in 1980 (Soto, 1982). It was found that as cutting frequency decreased from four to two cuttings per year at a 50-cm cutting height, grazeable material per plant increased by 48 to 100%. Using a mean of ten plants, Soto also found that the best accessions for producing grazeable forage were PI 263695 (K8) and PI 443663. Grazeable forage from Leucaena spp., was defined as stems (less than 6 mm in diameter), leaves, and petioles. Average yield of grazeable material from 14 selected leucaena accessions was 7 .2 Mg ha1 in 1981 and from 13 of the 14 accessions was 5.2 ~fg ha1 in 1983 (Othman et al., 1985). Four harvests were made in 1981 and three in 1983 (Othman et al., 1985). Psyl lid damage reduced forage yields in 1983 in the third cutting and completely destroyed the fourth cutting. Accession PI 263695 (K8) had the highest forage yield of 14.8 Mg ha-1 of grazeable dry matter in 1981. In 1983, nutrient removal in grazeable material for 13 accessions over two harvests totaled 160, 10, 100, 41, and 12 kg ha1 for N, P, K, Ca, and Mg, respec tively. Psyllid damage has prompted a selection program to develop leucaena for resistance to psyllid attack. Heavy psyllid damage in the summer of 1983 at Gainesville was followed by a severe freeze on 25 Dec 1983. In the 1984 growing season, no sign of psyllids was observed until just before winter. The 1984-85 winter also had a severe freeze and no insects were noticed until late the next growing season. From 1986 through 1990, there was little psyllid damage to the leucaena nursery at Gainesville, though psyllids were detected on several occasions. Predation and parasitic attack of the psyllids by various local insect predators have been noticed (Austin et al., 1990a,b). We now think future severe psyllid damage on leucaena in the humid cold subtropics will be confined to an occa sional short term outbreak, especially if psyllid-resistant leucaena accessions are planted. In pot and field tests of 98 and 18 selected leucaena accessions, respectively, from the Gainesville collection, Soffes (1984), showed almost total resistance to two root-knot nematodes (Meloidog'yru: incog nita and M. javanica). She found potential acid-soil tolerance in nine accessions of leucaena including ecotypes UF-1 and UF-2, PI 282405, PI 288011, Pl 286295, and CIAT lines 78-15/11, 22/9-7, 17-5/2-27, and 16/1-7. Also, two acid tolerant strains of Rhizohium spp. were found which would infect and fix nitrogen in Leuwena spp. In another study (Olivera y Sanchez, 1982), surface-soil liming allowed leucaena root systems to develop as deep as 90 cm in acid Ultisols and Oxisols. Total weight of K8 seed lings was found to increase in response to low concentrations of indoleacetic acid and gibberellic acid. Seed scarification by placing in 100C water, cooling and decanting is a practical method which resulted in an average of 94% germination for four leucaena acces sions (Olivera y Sanchez, 1982). In spite of these extensive studies, little has been written on the long term persistence and vigor of es tablished plantings. The object of this paper was to give final results of several earlier experiments and to determine persistence, 12 vr after establishment, of selected accessions from several Leucaena spp. that had been previously identified from the 373-acces sion leucaena nursery in Gainesville, Florida as potentially useful biomass and forage types. Accessions that did not stand out initially in early studies but which appeared vigorous and persistent in 1990 were also identified. MATERIALS AND METHODS The selected leucaena accessions for the present evaluation were part of a 373-accession nursery that was established in 1979 (Valencia, 1981). To establish this nursery, hand-scarified seed of each accession were planted in pots which were placed in a greenhouse until seedlings were 15 cm tall. They were then transplanted into one-row plots consisting of five seedlings spaced 100 cm apart. Rows were I m apart and each plot was replicated two times. Proper fertilization, irrigation, and weed and insect control allowed the leucaena accessions to become established by the end of 1979. Soil pH m 1979 ranged from 5.6 to 6.0. Othman ( 1984) applied 0.0, 24.5, and 93.0 kg ha1 of N, P, and K, respectively, in the springs of 1982 and 1983. Since 1983, no further fertilizer has been applied. Othman (1984) collected stem yield data on 62 and 53 visually selected high-biomass ecotypes from the 1982 and 1983 growth seasons, respectively. Of these, 12 accessions were selected for additional evaluation. The leucaena nursery was maintained by cutting top growth off several times a season with a rotary mower in 1984 and 1985. Biomass yields were then determined on the 12 selected accessions during the 1986 and 1987 growing seasons (Green et al., 1988). Plots of the selected accessions were cut in January and February at 4 to 5 cm above the ground. Fresh weights of whole-plot yields were determined at harvest and, by subsampling and oven drying to constant weight at 50C, the dry weight yields were calculated. In all four years, top growth was dead as a result of the winter freezes and most leaves had fallen so only stems were harvested. Fourteen and 13 leucaena accessions from the 373-accession nursery were selected for grazeable forage production in 1981 and 1983, respectively (Othman et al., 1985). Stems less than 0.6 mm in diameter, leaves and petioles made up the grazeable forage. Four forage harvests were made in 1981 and three in 1983. Psyllid damage reduced forage yields in 1983 in the third cutting and completely destroyed the fourth cutting We conducted new observations of the vigor and persistence on the entire leucaena collection in June and December 1990 and in January 1991, and fo-

PAGE 147

122 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA cused on the 12 high-biomass types selected by Othman et al., (1985). Due to the difficulty of distinguishing among accessions in June, observations presented are based on the condition of the material in December 1990 and January 1991 when aboveground plant parts were dead. The rating system used involved number of plants per accession and relative vigor based upon stem diameter. Population numbers could be no greater than 10 plants, i.e., 5 plants per plot totaled over 2 replications. Relative vigor numbers were 1 for small to 5 for large stem diameter. Selected mature plants were categorized as vegetative with no seed present (V), immature seed present (IS), and mature seed present (MS). An additional designation of low (L), medium (M) or high (H) for seed pod load was also used. Where mature seed (MS) was observed, the percentage of mature seed from the total seed load was estimated. RESULTS AND DISCUSSION The biomass (stem) yields of the 12 selected leucaena accessions are reported in Table 1, for the 1982, 1983, 1986, and 1987 growing seasons. The potential of leucaena for high biomass yields is shown by the overall mean yield of 31.4 Mg ha1 Individual accessions did much better. Because of the layout of the nursery where borders to single row plots were other leucaena accessions, it is possible that environmental enrichment occurred which made some of the biomass yield values higher than if the accessions had been bordered by plants of the same genotype. The high average yield of 31.4 Mg ha1 over the 12 acces sions reported here along with the high leucaena biomass yield reported for settling pond clays by Mis levy et al. (1989) make leucaena potentially a very productive valuable source of woody biomass. Table 1. The oven dry stem biomass yields of 12 selected leucaena accessions at Gainesville, Florida, during four growing seasons: 1982, 1983, 1986 and 1987. Data from Othman (1984) and Green et al. (1988). Oven dry biomass Acession PI number 1982 1983 1986 1987 Mean 263695 (KS) ---------------------------Mg ha-1 ----------------------------38.9at 26.5a 58.9a 44.9a 42.3 443696 35.2a 30.7 a 35.7 abc 35.3 ab 34.2 443541 31.8 a 20.6a 19.6 C 24.0 281607 (K28) 31.7 a 24.1 a 47.6abc 34.9 ab 34.6 286295(K62) 28.5 a 30.0a 31.2bc 26.1 b 29.0 28 I 608(K29) 28.5 a 28.6 a 45.8 abc 38.4 ab 35.3 288005 (K67) 28.1 a 27.9a 45.6abc 42.1 a 35.9 443674 27.6a 19.9a 31.5 be 31.4 ab 27.6 443482 (K419) 26.6a 21.0a 30.7 be 34.3 ab 28.2 443610 25.8a 21.6 a 27.6 be 28.7 b 25.9 370749 24.4 a 25.4 a 48.2 abc 38.2 ab 34.1 443483(K420) 23.4a 19.9a 31.8 be 25.1 b 25.3 Mean 29.3 24.7 40.2 33.4 31.4 tMeans followed bv the same letters in the same column are not different (P<0.05) according to Duncan's multiple range test (DMRT). Dry matter yield of the grazeable portion of selected leucaena accessions from the same leucaena nursery is shown in Table 2 for 1981 and 1983. Ac cessions on both sides of those selected were cut at the same time so a reasonably normal situation existed and there was less chance of environmental enrichment. The best accessions produced in excess of 7 Mg ha1 yr-1 of grazeable forage over two seasons even through psyllid damage 1983 reduced grazeable yield of the third cutting and completely destroying the fourth cutting. The average crude protein concentration over the 13 accessions studied in 1983 was 214 g kg-1 for the first cutting and 244 g kg-1 for the second cutting. The excellent grazeable yield and quality of leucaena forage make leucaena a good browse plant that could supplement grass pastures during summer and fall, when forage quality of grass pastures is often low. Of the 12 high-biomass accessions selected by Othman in 1982 (Table 1), six accessions, [Pl No. 443541, 286295(K62), 28 l 608(K29), 4436 I 0, 370749, and 443483(K420] either had poor persistence or low vigor by the end of the 1990 growing season (Table 3). The three lowest yielding accessions in Othman's study were not considered in the 1991 final persistence data. The poor persistence of Pl 286295 (K62) was disappointing, as it had also been found to have some psyllid resistance, possible acid soil tolerance, and high forage production capability. Despite the loss of half of Othman's biomass types, twenty leucaena accessions show excellent survival capability over an extended period of time in sub-Table 2. Grazeable forage yield of selected lucaena accessions established in 1979 over two growing seasons at Gainesville, Florida. Data from Othman et al., 1985. Accession Pl Number Oven dry yield of grazeable forage t 1981 1983 2-year mean 263695 (KS)t 443663 -------------------------Mg ha-I ------------------------14.8 a 11.3 ab 7.3 a 9.3 443660 I 1.1 ab 6.8ab 9.0 443711 8.2bc 4.7bc 6.5 443655 8.0bc 9.9a 8.9 443656 7.6bc 4.4 be 6.0 443664 7.9bc 6.1 b 7.0 443704 6.4 bed 7.4 a 6.9 443700 5.9cdc 5.9b 5.9 443594 6.7 bed 4.5 be 5.6 443710 4.6cd 4.4 be 4.5 443653 3.4 cd 2.2 cd 2.8 337088 2.3d 2.1 cd 2.2 288001 2.2 d l.0d 1.6 Mean over accessions 7.1 5.1 5.9 1'Total of 4 harvests in 1981 and only 3 in 1983 due to psyllid damage. *This accession was used in biomass study (Table l) where it was one of highest biomass yielding accessions in the 1983 season. Means followed by the same letters in the same column are not different (P<0.05) according to Duncan's multiple range test (DMRT).

PAGE 148

PROCEEDINGS, VOLUME 51, 1992 123 Table 3. Insect resistance, plant vitgor, and plant persistence of twelve selected leucaena accessions grown at Gainesville, Florida from 1979 to 1991. Vigor and persistence 1979* December 1990 Accession Biomass Insect number rank resistance + Vigor No. plants Vigor No. plants 263695 (KB) 2.0 4.7 IO 5.0 10 443696 2 1.0 2.8 IO 4.0 IO 4435411 3 >2.5 3.5 9 4.5 6 281607 K(28) 4 >2.5 2.8 8 5.0 9 286295 (K62)1 5 2.0 3.5 9 3.0 4 281608 (K29)1 6 >2.5 2.0 IO 0 288005 (K67) 7 >2.5 3.5 8 3.0 7 443674 8 >2.5 3.2 IO 5.0 8 443482 (K419) 9 >2.5 3.2 10 4.0 IO 4436101 10 2.0 3.0 8 1.5 5 3707491 l l 2.0 3.0 JO 2.5 5 443483 (K420)1 12 2.5 3.5 IO 0.5 I Overall mean 3.2 9.5 3.2 6.3 +Rating 0 = no insect damage to IO = all leaves completely covered with psyllids and no normal leaf growth (Othman, 1984). *Data of Valencia, 198 l. Vigor rating: I = low to 5 = high, based on stem size and number. ~Eliminated at end of 11 years dt to sparseness and/or low vigor in December 1990. tropical Florida (Table 4). In addition to the six highbiomass accessions which did persist well over the 12-year study period, there were 13 new accessions from the 373 accession nursery which were judged both persistent and vigorous. One additional accession, PI 443655, a L. leucocephala, which was borderline in vigor in 1990 but, because of the good early forage yield reported by Valencia ( 1981) during its second season of growth in 1980, still remains of interest. The order in which these entries are presented does not reflect a biomass or persistence/vigor rank per se; the first six accessions are the survivors of the original 12 selected by Othman (1984). Accession Pl 286248 (KIO) was found to be an exceptional forage type by Soto ( 1982) and to be fairly resistant (20% leafloss in August 1983) to shoot damage caused by psyllid (Othman, 1984). Only two of the 20 identified accessions in Table 4 are species other than L. leucocephala, and these two were found by Othman (1984) to exhibit very low leafloss due to psyllid damage. Accession PI 443740 is the only other persistent accession with psyllid resistance (15% leaf loss) as determined by Othman ( 1984). All 20 selected accessions survived the warm winter of 1990-91 with only leaves and small limbs killed by freezes and began normal growth in spring of 1991 on shoots developing from main trunks. CONCLUSION Leucaena showed good potential for furnishing rotational grazing for beef cattle particularly during late summer and fall when perennial grasses are of low quality. Research needs to be continued to determine the level of grazing (browsing) pressure that will allow leucaena to persist under subtropical conditions. Table 4. Vigor of stem regrowth, number of plants surviving, and seed condition of 20 selected leucaena accessions grown at Gainesville, Florida, from 1979 to winter of 1990-91. Seed condition Accession No. plants Maturity/yield/%mature numbers+ Vigor surviving Rep I; Rep 2 263695 (KS) 5 IO MS/M/10%; MS/USO% 281607 (K28) 5 9 MS/H/100%; IS 288005 (K67) 3 7 IS/H 443482 (K4 l 9) 4 9 IS/L 443674 5 8 IS; MS/H/5% 443696 4 IO V 443481 3 C) IS/L; V 443537 3 9 V; IS/L 443546 4 JO MS/M (L. esculenta) 286248 (KIO) 5 IO MS/L/80%; V 443586 3 9 MS/H/50%; MS/M/75% 435929 4 7 MS/UI0%; MS/L/25% (L. lanceolata) 435926 5 7 V (Leucaena sp) 443740 5 8 MS/M/50%; IS/L 443695 3 IO IS/L; IS/M 443714 3 9 IS; IS/H 443617 3 8 MS/M; IS/H 443468 4 8 MS/L/60%; MS/H/50% 443478 (K415) 5 8 IS/M 443555 1.5 IO V Overall mean 3.9 8.8 t Accessions are L. leuwcrphala except where marked differently. *Seed coindition:maturity values: V = vegetative only, IS = immature seed present, MS = mature seed; Seed pod load; L low, M = medium and H = high load (yield) of pods. The percentage shown is percentage of mature seed pods at time of rating. Because values sometimes varied between replications, results of both replications are shown when not the same values. Low vigor but strong upright stems with all 10 plants surviv ing.

PAGE 149

124 SOIL AND CROP ScrENCE SOCIETY OF FLORIDA Since the deliberate planting of woody perennials in production systems, a practice known as agroforestry, is relatively new in the more technology-based societies, there is a need for more research on the man~gement of crops and livestock in leucaena production systems that exploit its principles in order to p_romote sustainability (FAO, 1989). High annual yield of woody stems by leucaena makes it a potential energy crop where biomass is harvested each winter in cold subtropical and warm temperature climates. Any new research on leucaena where persistence and continued vigor are required should use one or more of the 20 superior accessions identified in this study. ACKNOWLEDGMENTS This research was partly supported in 1982 and 1983 through an agreement between the University of Florida, Institute of Food and Agricultural Sci ences and the Gas Research Institute of Chicago, Il linois and in 1986 and 1987 through a grant from the Governor's Energy Office. The Center of Sustainable Agroforestry, Inc. of Gainesville, Florida and a grant fro: Farm~orkers S:lf-Help, Inc., of Dade City, Flonda provided partial support for the research in 1990 and 1991. REFERENCES Austin, M. T., M. J. Williams and J. H. Frank. 1990a. Florida Leucaena psyllid trial (LPT). II. First year evaluation of psyllid damage. Leucaena Res. Rep 11 :6-9. A. C. Hammond,]. H. Frank, and C. G. Chambliss. 1990b. Florida Leucaena psyllid trial (LPT). I. First year establishment, yield, chemical composition, and survival. Leucaena Res. Rep. 11: 3-5. Brewbaker, .J. L., R. Van Den Beldt, and K. MacDicken. 1984. Fuel wood uses and properties of nitrogen-fixing trees. Pesquisa Agro.pecuaria Brasileira 19: 193-204. FAO. 1989. Technical Advisory Committee to the Consultative Group on International Agricultural Research p. 56-58. Sus tainable agricultural production: implications for international agricultural research. Rome. Glumac, E. L. 1986. Biomass production, survival, and cold tolerance of three species of Leucaena in South Texas. Leucaena Res. Rep. 7: 119-120. Green, A., G. Prine, D. Rockwood, C. Batich, J. Winefordner, D. Williams, B. Green .J. Wagner, J. Schwartz, H. Van Ravenswaay, D. Clauson, S. Mills, T. Yurchisin, R. Storer, and A. Feinberg. 1988. Co-combustion in community waste to energy systems. The Am. Soc. Mechanical Engineers. Vol. 4: 13-28. Hammond, A. C., M. J. Allison, M. J. Williams, G. M. Prine, and D. B. Bates. 1989. Prevention of Leucaena toxicosis of cattle in Florida by ruminal inoculation with 3-hydroxy-4-(lH)pyridone-degrading bacteria. Am. J. Veterinary Res. 50: 2 I 762180. Kendall, J., H. Margolis, and M. Capo. 1989. Relative cold hardiness of three populations of Leucaena leucocephala from Northeastern Mexico. Leucaena Res. Rep. 10: 19-21. Mislevy, P., W. G. Blue, and C. E. Roessler. 1989. Productivity of clay tailings from phosphatic mining: I. Biomass crops. J. Environm. Qua!. 18: 95-100. National Academy of Sciences (NAS) and the Philippine Council for Agriculture and Resources Research p. 24-25. 1984. Leucaena: promising forage and tree crop for the tropics. Pub lished by the National Academy of Sciences, Washington, D.C. Olivera y Sanchez, Eduardo. 1982. l,mcaena establishment germination, hormone applications on early development and nodulation, and methods to establish in acid soils. Ph.D. diss., Univ. of Florida, Gainesville, FL. (Diss. Abstr. 8382280). Othman, A. B. 1984. Evaluating Leucaena introductions for biomass and forage production. M.S. Thesis, Univ. of Florida, Gainesville, FL. Othman, A. B., and G. M. Prine. 1984. Leucaena accessions resistant to jumping plant lice. Leucaena Res. Rep. 5: 86-87. Othman, A. B., M. A. Soto, G. M. Prine, and W. R. Ocumpaugh. 1985. Forage productivity of Leucaena in the humid subtropics. Soil Crop Sci. Soc. Florida Proc. 44: 118-122. Pathak, P. S. 1985. l,eucaena leaf manuring and yield of oats. Leucaena Res. Rep. 6: 4 7-48. Ravenswaay, H. R .J. V. 1989. Co-burning biomass and nonhazardous waste in a modular incinerator. M.S. Thesis, Univ. of Florida, Gainesville, FL. Soffes, A. R. 1984. Meloidogyne, aluminum and rhizobium re lationships in Leucaena germplasm selections. Ph.D. diss., Univ. of Florida, Gainesville, FL. (Diss. Abstr. 8421074). Soto, M. A. 1982. Morphological components of yield in Leuwena. M.S. Thesis, Univ. of Florida, Gainesville, FL. USDA. 1990. Plant hardiness zone map. US Printing Office, Washington, DC. Misc. Publication 1475. Valencia, I. M. 1981. \/umerical classification of the genus Leucaena. M.S. Thesis, Univ. of Florida, Gainesville, FL. Williams, M. J. 1987. Establishment and winter survival of Leucaena spp. and Gliricidia sepurn in the cold subtropics. Leucaena Res. Rep. Vol. 8: 79-81.

PAGE 150

PROCEEDINGS, VOLUME 51, 1992 125 Effect of Water Deficits on Growth and Nitrogen Fixation of Hairy Indigo U. Winzer, S. L. Albrecht*, and J.M. Bennett ABSTRACT Hairy indigo (Indigofera hirsuta L.) is used as a forage, green manure and cover crop. In the subtropical environment of Florida, crops are frequently exposed to periods of severe drought during the growing season. The response of hairy indigo to water deficits was investigated in a greenhouse study conducted in Gainesville, FL. Plant water status, nitrogenase activity, and leaf diffusive resistance were measured during the course of an imposed water deficit. Hairy indigo maintained substantial shoot growth for 15 d after water was withheld. Root growth continued for as long as 21 d after imposing the water-deficit treatment, and thus the root:shoot ratio of water-deficient plants increased above those of the well-watered controls. Leaf water potentials of the stressed plants were similar to those of well-wa tered plants 9 dafter withholding water, and then declined to as low as -2.2 MPa by the end of the water-deficit period. Nitrogenase activity also decreased with declining leaf water potential, though leaf diffusive resistance was more sensitive to water deficit than any other measured parameter. Leaf diffusive resistance of plants in the water-deficit treatment increased rapidly as leaf water and turgor potentials declined below -2.0 and 0.2 MPa, respectively, suggesting that hairy indigo relies on the reduction of water loss from leaves as an effective way to reduce water loss during periods of severe drought. Hairy indigo (lndigofera hirsuta L.) is a tropical legume native to Africa and Asia that was introduced into the United States in 1908. Relatively high dry matter production and nitrogen content make it suitable for use as a forage, cover or green manure crop (Baltensperger et al., 1985). In addition, hairy indigo is a poor nematode host and can actually reduce soil populations of root-knot (Meloidogyne spp.), sting (Be lonolaimus longfr:audatus) and lesion (Pratylenrhus brachyurus) nematodes in fields when planted in rotation with nematode-susceptible crops (Reddy et al., 1986). Hairy indigo appears well-adapted to draughty, sandy soils; however,quantitative information on its response to water deficits is lacking and the mechanism for its presumed drought tolerance is not understood. In Florida, limited rainfall during the spring and fall months can drastically reduce soil moisture. Drought conditions may persist for short or extended periods and induce plant water deficits which are detrimental to crops at almost any stage of development. A decline in growth and productivity of forage legumes during these periods is common in a subtropical environment. Dinitrogen fixation by forage legumes is an effec tive and economic way to provide nitrogen for increasing the productivity of pastures in Florida. Hairy indigo should be capable, in symbiosis with Bradyrhizobium spp., of fixing substantial amounts of nitrogen. It has been shown that drought inhibits dinitrogen fixation by various grain legumes (Albrecht U. Winzer, S. L. Albrecht, and J. M. Bennett, Agronomy Uept., Univ. of Florida, Gainesville, FL 32611-0840. Florida Agric. Exp. Stn. Journal Series no. N-00596. *Corresponding author. Contribution published in Soil Crop Sci. Soc. Florida Proc. 51: 125-129 (1992) et al., 1984; Bennett and Albrecht, 1984; Pankhurst and Sprent, 1975; Sprent, 1972; Weisz et al., 1985) and forage legumes (Albrecht et al., 1980; Engin and Sprent, 1973). Bennett and Albrecht (1984) found that nitrogenase activity in soybean (Glycine max L.) was closely correlated with nodule water status, which appeared to be more sensitive to drought than either leaf water potential or stomata! resistance. Several theories have been formulated to explain the responses which occur in both the nodules and whole plant as a result of water deficits (Huang et al., 1975a, b; Pankhurst and Sprent, 1975; Pate et al., 1969; Sprent, 1976; Weisz et al., 1985). To effectivelv utilize the potential of hairy indigo in the nitrogen economy of Florida pastures, it is important to understand the response of dinitrogen fixation during periods of water deficit. The purpose of this study was to investigate the response of hairy indigo to water deficit and to relate the effects of soil water deficits on dinitrogen fixation to changes m leaf water status and leaf diffusive resistance. MATERIALS AND METHODS Cultural Practices and Greenhouse Environment A greenhouse experiment was conducted in Gainesville, FL during the spring of 1985. Hairy indigo (var. 'common') seeds, inoculated with a commercially available Bradyrhizobium inoculum (Nitrogen Co., Milwaukee, WI), were planted 5 March in plastic pots (18.0 cm dia. x 17.7 cm ht.) containing coarse, washed sand. Plants were thinned to one plant per pot and watered on alternate days with N-free Jensen's solution (Vincent, 1968). All pots were leached weekly with excess water, to prevent salt accumulation. Immediately before initiating the treatments, all pots were leached and then irrigated with 200 ml of N-free Jensen's solution. The temperature within the greenhouse ranged between 20C (night) and 35C (day), with relative humidity near 40% at midday. Photosynthetically-active radiation reached a maximum of 1500 E m-2 s-1 at midday on dear days. Treatments Pots were arranged in a completely randomized design with two water treatments (well-watered and water-deficit) replicated four times. The well-watered treatment was watered throughout the experimental period to prevent visible stress symptoms. For the water-deficit treatment, water was withheld from the plants from 66 to 87 cl after planting. Measurements Soil and plant characteristics were measured at 2 to 3 d intervals during the 21 cl drying cycle. Measurements were not begun until day 4 of the treatment period, as previous experiments (Albrecht and Ben-

PAGE 151

126 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA nett, unpublished results) have shown that plants grown in this environment do not exhibit water-stress symptoms until the treatment has been imposed for 4 to 5 d. Measurements of leaf water status, leaf dif fusive resistance and nitrogenase activity were made near midday, under full sun conditions. On each measurement date, soil samples were collected, using a 2-cm dia. aluminum tube, from the center of each pot and weighed immediately. Soil samples were then dried at 105C for 16 hand dry weights determined. On each sample date, weights of shoots, roots and nodules were determined after drying samples at 105C for 16 h. Nodules were removed from the root system and weighed before drying to determine their water content. Leaf Diffusive Resistance A steady-state diffusion porometer (Li-Cor, Inc., Model LI-1600, Lincoln, NE) was used to determine leaf diffusive resistance (Rs) on upper, fully-expanded leaves. Resistances of the abaxial (Rab) and adaxial (Rad) leaf surfaces were determined and total leaf Rs was calculated as: Rs= 1/Rad + 1/Rab, assuming that the individual surface resistances act in parallel. Leaf Water Status Leaf water potentials, osmotic potentials and turgor potentials were measured on one entire leaflet from an upper, fully-expanded leaf from each repli cate. The leaflet was placed in a Spanner-type (Spanner, 1981) thermocouple psychrometer (J. R. D. Merrill Spec. Equipment Co., Model 84-13, Logan, UT) and equilibrated for 4 h in a thermostatically-controlled water bath at 30C. The psychrometer output was recorded using a strip chart recorder and a dewpoint microvoltmeter (Wescor, Inc., Model HR 33T, Logan, UT) operating in the psychrometric mode. Output from each thermocouple psychrometer then was converted to water potential by comparison to calibration curves previously developed using NaCl solutions. Leaf osmotic potential was determined after freezing and thawing of the same leaf tissue following 4 h of equilibration at 30C. Turgor potential was calculated as the difference between leaf water ~otential and leaf osmotic potential. Nitrogenase Activity Nitrogenase act1v1ty was estimated by the acetylene (C2H2 ) reduction method (Hardy et al., 1968). Intact root systems were excised from the shoots and separated' from the soil. The root system and attached nodules then were quickly placed in a 75-ml serum vial, sealed with a rubber serum stopper, and 7.5 ml of air in the vial was exchanged with C2H2 to provide a 10%, C2H2 atmosphere. Samples were incubated at greenhouse temperatures. During the incubation period, 0.5 ml gas sa>::j'les were analyzed for ethylene (C2H4 ) at 30 min. intervals. Concentrations of C2H4 were determined with a gas chrom?tri-graph (Varian, Inc., Model 940, Sunnyvale, CA) fitted with a flame ionization detector. Nitrogenase ac tivity was computed from linear regression lines fitted to the time-sequence measurements. Specific nitrogenase activity was calculated by dividing total nitrogenase activity per plant by nodule dry weight. Analysis All data were analyzed with appropriate statistical methods using the Statistical Analysis System (SAS, 1985) at the Northeast Regional Data Center on the University of Florida Campus. RESULTS Gravimetric soil water content for the well-watered (control) plants was maintained between 110 and 150 g kg-1 except on clay 4 when it was below 80 g kg1 (Figure 1). No symptoms of water stress were observed for the control plants during the course of the experiment. Gravimetric soil water content decreased after water was withheld from the water-deficit treatment and was less than 1 % by 18 d after the treatment had been imposed. After rewatering on day 21, gravimetric soil water content in the water deficit treatment returned to almost 120 g kg-1 and was similar to that of the well-watered treatment. Vis ible water-stress symptoms (midday wilting) did not become apparent in the water-deficit treatment until 15 d after water was withheld. Shoot dry weight increased from 0.28 g to 1.44 g and from 0.27 to 1.18 g during the treatment period for the well-watered and the water-deficient plants, respectively (Figure 2a). For the first 18 dafter treatments were imposed, shoots of plants in both treatments accumulated dry matter at a similar rate. How ever, during the last few clays of the experiment, dry matter accumulation for the water-deficient plants appeared to be less than for the well-watered plants, although the differences were not statistically signifi cant. Dry weights of both root and nodule tissue in-160 ,---,---,-~--.--r---r--r----r--,----,--.----~ 140 !z~ 120 6 3: 100 Uc'.:' a::" 80 w c,, I-<( ;;:: 60 ....JI f;l c,, 40 I 0~ I 0----0-0-0 ........ 0 ~/ ""'/"" 20 0-0 Control Stress .'--------0 L__J__.L.______._ __,___,_.....J...._.1...----!...---===--'----' 0 2 4 6 B 1 0 12 14 16 1 B 20 22 24 DAYS AFTER IMPOSING TREATMENT Fig. 1. Gravimetric soil water contents of well-watered and water-deficit treatments. Each point represents a mean of four observations. The water-deficit treatment was rewatered on day 21. Each point is the mean of at least three determinations. The vertical bar represents the mean standard deviation.

PAGE 152

PROCEEDINGS, VOLUME 51, 1992 127 1.50 A 0 I /. !i: t!) i-o:::;:=i---w 1.00 3:: ~0 i:::::::i t3 0.50 0 i~ 0-0 CONTROL ::r: en e-e STRESSED 0.00 400 ,,..... B "' ___,..----- 5 JOO I !i: ---0 t!) -t-0-0_0 / w 200 3:: 0 I-100 ro-0 0 0 ct: i 0 100 ,,..... C "' I //l 5 BO I-::r: -~.__,,,, t!) 60 w 3:: 10----0 --0 0 40 o/~'~ :::, 20 0 0 z 0 '----'----'---'--~-~~-~-~-~-~~~ O 2 4 6 B 10 12 14 16 1B 20 22 DAYS AFTER lt.APOSING TREATt.AENT Fig. 2. Shoot dry weight (A), root dry weight (B) and nodule dry weight (C) of well-watered and water-deficient hairy indigo plants. Each point is the mean of at least four determinations. The vertical bar represents the mean standard deviation. creased in both treatments during the 21-d treatment period (Figure 2b and 2c). Differences in root dry matter accumulation for the two treatments were observed between 9 and 21 d of the treatment period. During the first 21 d after water was withheld, accumulation of dry matter by the roots of water-deficient plants was greater than for the well-watered plants. This ehhanced root growth for the water-deficient plants resulted in an increase in root (plus nodule):shoot ratio. For example, water deficit increased the root:shoot ratio by 113% and 76% on days 15 and 18, respectively. Nodule water content remained between 780 and 880 g kg-I for the well-watered plants throughout the experimental period. For the water-deficient plants, however, nodule water content decreased after day 9, from 822 to a low of 520 g kg-I on day 18 (Figure 3). Nodule water content was significantly lower for the stressed treatment from day 12 until day 21. After rewatering, the nodule water content of the nodules from the water deficit treatment increased once more to 870 g kg-I, which was similar to the control values. Leaf diffusive resistance of the well-watered plants ranged between 0.38 and 0.84 s cm-I, which is 900 ------------------~~ I 850 I _,.,..--0.__ I-~(\~-~ a z w 800 I-.s;: z (I) 0 (I) u .::: 750 ct: (I) w :i 700 !;;: 0 -~\~/ 3:: C: 650 w "' _J .:;: :::, .:. 600 e-e STRESSED 0 (I) 0 0 0-0 CONTROL z 550 500 0 2 4 6 8 10 12 14 16 18 20 22 24 DAYS AFTER IMPOSING TREATMENT Fig. 3. Water content of nodules excised from well-watered and water-deficient hairy indigo plants. Each point represents the mean of at least four observations. The water-deficit treatment was rewatered on day 21. The vertical bar represents the mean standard deviation. The "*" indicates that means differ statistically (P=0.10) on a given date. typical for nonstressed leaves (Figure 4). However, diffusive resistance for the water-deficient plants increased significantly on the ninth day after treatments were imposed, and continued to increase until plants were rewatered. Even 2 d after rewatering, leaf diffusive resistance of the previously water-deficient plants remained higher than for the well-watered controls, demonstrating residual effects from the water deficit. Leaf water potential (Figure 5a) and leaf turgor potential (Figure 5b) did not differ between treatments for at least 9 d after the water-deficit treatment was imposed. There was a rapid decline in leaf water potential, however, for the water-deficient plants from -0.61 to -1.99 MPa between days 9 and 12. Differences in leaf turgor potentials, on the other hand, were not observed until day 15 of the treatment period. Near the end of the drying cycle, leaf water potentials and leaf turgor potentials had declined to less than -2.0 MPa and near zero, respectively, indi-w u z (/] w ci::~ w' > E u (/] :::, (I) LJ...~ LJ... LJ... _J 6 5 4 3 2 I 0-0 CONTROL STRESSED 2 4 6 8 10 12 14 16 18 20 22 24 DAYS AFTER IMPOSING TREATMENT Fig. 4. Leaf diffusive resistance of well-watered and water-deficient hairy indigo plants. Each point represents the mean of four observations. Water-stressed plants were rewatered on day 21. The vertical bar represents the mean standard deviation. The "*" indicates that means differ statistically (P=0.10) on a given date.

PAGE 153

128 SOIL AND CROP SCIFNCE SOCIETY OF FLORIDA _J <( i== z w 1---0,,...... CL o Q'. 0... w:::!' 1---~ LL _J _J <( f:= z w I---0 CL,,...... -1 -2 -3 0.6 0.5 0.4 0:: C 0 0.3 c., 0:: ::::, 0.2 I---LL 0.1 ....I 0.0 0 A I 0--0 CONTROL e-e STRESSED B I 2 4 6 8 10 12 14 1 6 18 20 22 24 DAYS AFTER IMPOSING TREATMENT Fig. 5. Leaf water potential (A) and leaf turgor potential (B) of well-watered and water-deficient hairy indigo plants. Each point represents the mean of four observations. Water-deficient plants were rewatered on day 21. The vertical bar represents the mean standard deviation. The "*" indicates that means differ statistically (P=O.lO) on a given date. eating severe plant water deficits. Visible stress symptoms also indicated that the plants were severely water-stressed before they were rewatered. Two days after rewatering, leaf water potentials and leaf turgor potentials still remained below those of the control plants, even though the soil water contents (Figure I) of the two treatments were similar. The severe stress apparently inhibited plants from a rapid recovery to full leaf water status after rewatering. Specific nitrogenase activity of well-watered plants was between 50 and 130 mo! ethylene (C2H4 ) formed h-1 g nodule dry wt' during the experiment, and seemed to increase slightly with time (Figure 6). Specific nitrogenase activity declined rapidly for the water-deficient plants after the ninth day of the stress period, decreasing from 79 to 33 mol C2H4 formed h-1 g nodule dry wt1 by day 12. Two days after rewatering, nitrogenase activity remained lower than for the well-watered plants. Total nitrogenase activity per plant was calculated by multiplying specific nitrogenase activity by the total dry weight of nodules on a given plant. This value increased in the well-watered plants from 1.19 to 7 .25 mol C2H4 formed h' plant' during the study, whereas for the water-deficit treatment it remained at 2.13 to 2.25 mo! C2H1 formed h plant for the first 9 d and then declined to near zero by day 18. Specific nitrogenase activity also increased with increasing nodule water content (Figure 7). Ni-I 120 ~3 I ::;;: 100 o----....... 0 ti 'C / O"-o/ <( w::, 80 VJ 'C <( 0 z C :><1\~ w o> 60 L'J o- n:: I I---.... ..c 40 ./ z :.i-u :r: 0-0 CONTROL f; UN ~-u 20 STRESSED w., CL~ -----VJ 0 E 0 -3 0 2 4 6 8 10 12 14 16 18 20 22 24 DAYS AFTER IMPOSING TREATMENT Fig. 6. Specific nitrogenase activity of well-watered and waterdeficient hairy indigo plants. Each point represents a mean of four observations. Water-deficient plants were rewatered on day 21. The vertical bar represents the mean standard deviation. The "*" indicates that means differ statistically (P=0.10) on a given date. trogenase activity was greatly reduced below 700 g H2O kg nodule fresh wt 'and undetectable below 550 g H2O kg nodule fresh wt 1 In general, above 7 50 g H2O kg nodule fresh wt1 the specific nitrogenase ac tivity increased rapidly, though erratically. There was little increase in specific nitrogenase activity above 850 g H2O kg nodule fresh wt'. DISCUSSION During the first 18 d after withholding water there were only slight differences in dry weight accumulation between the two treatments, even though soil water content had declined to low levels. This suggests th~t hairy indigo exhibits some tolerance to water deficits. Declining soil water content also enhanced root growth. After 9 d of the treatment period, root growth actually increased for the waterdeficient plants, resulting in a significantly higher root dry weight and an increase in root:shoot ratio. It has been suggested that an increase in the 1 100+---+---l---+--+---l--+--+--+---h-E -; so ti 'C 80 70 g 60 z w '?' 8 50 n:: I t: 1: 40 z :q. 30 u :r: UN 20 f;l ., CL
PAGE 154

PROCEEDINGS, VOLUME 51, 1992 129 root:shoot ratio of water-deficient plants can often be attributed to continued root growth coupled with lit tle or no shoot growth (Kramer, 1983). Increased root growth, allowing water extraction from a greater volume of soil, is one mechanism associated with improved drought resistance Qones et al., 1981). Changes in leaf diffusive resistance in response to the imposed water stress were observed earlier than changes in either leaf water or turgor potentials or the appearance of visible stress symptoms. Stomatal activity, as determined by leaf diffusive resistance, of hairy indigo appeared quite sensitive to decreasing soil water content. Regulation of stomatal closure is often considered to be an important feature of drought-resistant plants (Kaufmann, 1981). While closing of the stomata prevents water loss from a leaf, it also reduces carbon dioxide diffusion into the leaf and thereby decreases photosynthesis. Wong et al. ( 1985) contended that a close correlation between stomatal opening and photosynthetic activity exists in most plants. It is likely that the rapid increase in leaf diffusive resistance during the experiment also caused a decrease in photosynthetic activity for plants in the water-deficit treatment. If the water-deficit period had been extended, a considerable reduction in growth would probably have been observed. However, the ability to close stomata, even at the expense of photosynthesis and dry matter accumulation, should allow the plant to conserve water and thereby possibly survive until soil water is replenished. This rapid increase in diffusive resistance of water-deficient hairy indigo plants as leaf water and turgor potentials declined should provide an effective means for reducing water loss during periods of severe drought. Nitrogenase activity decreased only after the ini tial increase in leaf diffusive resistance had been observed, suggesting that nitrogenase activity and leaf diffusive resistance were not directly related. If leaf diffusive resistance and photosynthesis are closely as sociated for this legume, then the decline in nit rogenase, activity by plants in the water-deficit treatment was not immediately caused by limited photosynthetic capability, as some reports have suggested (Huang et al. 1975a, b). These results for hairy indigo are similar to those presented by Albrecht et al. ( 1984), who reported a decline in photosynthesis, measured as apparent canopy carbon exchange, in water-deficient soybean plants prior to any decline in nitrogenase activity. Under the conditions of this experiment, nitrogenase activity closely followed changes in leaf water potential and nodule water content. Up to day 9 of the water-deficit period, leaf water potential, nodule water content and nitrogenase activity were similar for the well-watered and water-deficit treatments. However, as soil water continued to decline, decreases occurred simultaneously in all three parameters. These results support the conclusions of Sprent and coworkers (Engin and Sprent, 1973; Sprent, 1971; Sprent, 1976) and Bennett and Al-brecht (1984), who reported that water status of the nodule was closely correlated with changes in nitrogen fixation. REFERENCES Albrecht, S. L., J. M. Bennett, and K. J. Boote. I 984. Relationship of nitrogenase activity to plant water stress in field-grown soy beans. Field Crops Res. 8:61-71. Albrecht, S. L., J. M. Bennett, and K. H. Quesenberry. 1980. Growth and nitrogen fixation of Aeschynomene under water stressed conditions. Plant and Soil 60:309-315. Baltensperger, D. D., E. C. French, G. M. Prine, 0. C. Ruelke, and K. H. Quesenberry. 1985. Hairy indigo, a summer legume for Florida. Univ. of Florida Agric. Expt. Sta. Circular S-318. Bennett, J. M., and S. L. Albrecht. 1984. Drought and flooding effects on N2 fixation, water relations and diffusive resistance of soybean. Agron. J. 76:735-740. Engin, M., and]. I. Sprent. 1973. Effects of water stress on growth and nitrogen-fixing activity of Trifolium repens. New Phytol. 72:117-126. Hardy, R. W. F., R. W. Holsten, and R. C. Burns. 1968. The acetylene-ethylene assay for nitrogen fixation: Laboratory and field evaluation. Plant Physiol. 43: 1185-1205. Huang, C. Y., J. S. Boyer, and L. N. Vanderhoeff. 1975a. Acetylene reduction (nitrogen fixation) and metabolic activities of soybean having various leaf and nodule water potentials. Plant Physiol. 56:222-227. Huang, C. Y.,J. S. Boyer, and L. N. Vanderhoeff. 1975b. Limitation of acetylene reduction (nitrogen fixation) by photosyn thesis in soybeans having low water potentials. Plant Physiol. 56:228-232. Jones, M. M., N. C. Turner, and C. B. Osmond. 1981. Mechanisms of drought resistance. In L. G. Paleg and D. Aspinall (ed.), The Physiology and Biochemistry of Drought Resistance in Plants. Academic Press, Sydney. pp. 15-53. Kaufmann, M. R. 1981. Water relations during drought. In L. G. Paleg and D. Aspinall (ed.), The Physiology and Biochemistry of Drought Resistance in Plants. Academic Press, Sydney. pp. 55-70. Kramer, P. J. 1983. Water Relations of Plants. Academic Press, New York, N. Y. 359 p. Pankhurst, C. E., and J. I. Sprent. 1975. Effects of water stress on the respiratory and nitrogen-fixing activity of soybean root nodules. J. Exp. Bot. 26:287-304. Pate, J. S., B. E. S. Gunning, and L. C. Briarty. 1969. Ultra structural functioning of the transport system of a leguminous root nodule. Planta 85: 11-34. Reddy, K. C., A. R. Soffes, G. M. Prine, and R. A. Dunn. 1986. Tropical legumes for green manure. II. Nematode populations and their effects on succeeding crop yields. Agron. J. 78:5-10. SAS Institute, Inc. 1985. SAS user's guide: Statistics. SAS Institute, Cary, NC. 956 pp. Spanner, D. C. 1951. The peltier effect and its use in the measure ment of suction pressure. J. Exp. Bot. 2: 145-168. Sprent, J. I. 1971. The effects of water stress on nitrogen-fixing root nodules. I. Effects on the physiology of detached soybean nodules. New Phytol. 70:9-17. Sprent, J. I. 1972. The effects of water stress on nitrogen-fixing root nodules. IV. Effects on whole plants of Vicia faba and Glycine max. New Phytol. 71:603-611. Sprent. J. I. 1976. Water deficits and nitrogen-fixing root nodules. In T. T. Kozlowski (ed.), Water Deficits and Plant Growth, Vol. IV. Academic Press, Sydney. pp. 291-315. Vincent, J. M. 1968. A manual for the practical study of rootnodule bacteria. Blackwell Scientific Publications, Oxford. 164 pp. Weisz, P.R., R. F. Denison, and T. R. Sinclair. 1985. Response to drought stress of nitrogen fixation (acetylene reduction) rates by field grown soybeans. Plant Physiol. 78:525-530. Wong, S. C., I. R. Cowan, and G.D. Farquhar. 1985. Leaf conductance in relation to rate of CO2 assimilation. III. Influence of water stress and photoinhibition. Plant Physiol. 78:830-834.

PAGE 155

130 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Some Factors Affecting Response of 'Florida 77' Alfalfa to Acid-Soil Amendments B. W. Mathews, R. E. Joost, and L. E. Sollenberger* ABSTRACT In many acid mineral soils, subsoil penetration by roots is severely limited by Al toxicity. A field study was initiated with an established first-year stand of 'Florida 77' alfalfa (Medicago sativa L.) growing on an Olivier silt loam soil (Aquic Fragiudalf, fine-silty, mixed, thermic) in southern Louisiana. The objective was to evaluate the effects of surface-applied phosphogypsum (PG), langbeinite (LB), and muriate of potash (potassium chloride, KCI) on subsoil exchangeable Al and on alfalfa growth under field conditions. Ten soil amendment treatments were imposed in four replications in a randomized complete block design. Treatments included KCI at rates of 1340, 2680 (single application), and 2680 (split application) kg ha-1 ; LB at rates of 3630, 7260 (single), and 7260 (split) kg ha-1 ; PG at rates of 4815, 9630 (single), and 9630 (split) kg ha-1 ; and a control. Both LB and PG moved considerable S into the subsoil (30-to 60-cm soil depth), but there was no corresponding decrease in subsoil exchangeable Al. Sulfur supplied as LB moved faster and deeper into the subsoil than PG-supplied S. Muriate of potash was not effective in displacing surface-soil Ca into the subsoil. None of the amendments produced yield increases over the control in the two years of the study, with examination of the soil profile at the end of the second year indicating that alfalfa roots had penetrated to a depth of 55 to 60 cm regardless of the treatment. A fragipan which began at a soil depth of 56 cm, rather than subsoil Al, appeared to be the most important factor limiting root development in this soil. Alfalfa does not produce high yields in most of the southeastern USA. Low yields and poor persis tence may be due in part to the detrimental effects of subsoil Al on root growth, thereby reducing plant uptake of water and nutrients (Sumner, 1990). Because of its limited solubility, surface application of lime generally has little effect on toxic subsoil exchangeable-Al levels. Deep incorporation of lime by physical means is both costly and often undesirable due to concurrent exposure of infertile subsoil (McCray and Sumner, 1990). Therefore, amelioration of high subsoil exchangeable-Al levels usually depends on leaching of surface-applied amendments. Surface-applied amendments are especially important in the maintenance of perennial forage crops, since tillage is not feasible following stand establishment. Volk (l 966) was among the first to mention the potential for using gypsum (CaSO4.2H2O) as an acid subsoil ameliorant. Since then, numerous studies have reported significant reductions of subsoil exchangeable Al following surface application of various gypsum materials (Alva and Sumner, 1990; McCray and Sumner, 1990; Sumner, 1990). This practice represents a sizeable opportunity to improve soils with moderate to high subsurface exchangeable B. W_ MaLhews and L. E_ Sollenberger, Agronomy Dep., Univ. of Florida, Gainesville, FL 32611-0900, and R. E. Joost, Southwest Center, Univ. of Missouri. Rt. 3, Mount Vernon, MO 65712. Florida Agric. Exp. Stn. Journal Series no. R-02077. *Corresponding aulhor. Contribution published in Soil Crof1 Sci. Soc. Florida Proc. 51: 130-135 ( I 992) Al levels in the Southeast, since by-product phosphogypsum (PG) resulting from the production of phosphoric acid has been accumulating in Florida, Louisiana, Mississippi, and North Carolina for over 30 yr. Though more effective than lime, amelioration of subsoil exchangeable-Al problems by leaching surface-applied PG is often a fairly slow process (Sumner, 1990). In a greenhouse leaching column study, Mathews and Joost (1990) found that mined langbeinite (LB, or K2SO4.2MgSO4), a more soluble SOi-material, was superior to PG in the reduction of subsoil exchangeable Al. It is well known that alfalfa has a high requirement for K (Lanyon and Griffith, 1988). Muriate of potash (potassium chloride, KCl) is the predominant K fertilizer used on alfalfa and most other crops. Ap plications of KCl also have been used to enhance Ca leaching and thereby the ameliorating effects of Ca on subsoil exchangeable-Al toxicity (Bouldin, 1979). The present research was initiated to: i) study the effects of surface-applied PG, LB, and KC! on soil chemical properties; and ii) study the effects of these amendments on yield and root development of Florida 77 alfalfa. MATERIALS AND METHODS Florida 77 alfalfa was established in October 1986 on an Olivier silt loam soil that had received 2240 kg ha-1 of dolomitic limestone (incorporated to a depth of 10 cm) 90 cl prior to planting. The site, located on the Perkins Road Farm, Baton Rouge, LA, previously had been cropped to cotton (Gossypium hirsutum L.) or corn (Zea mays L.) for over 25 yr. Ten soil amendment treatments were arranged in a randomized complete block design with four replications. Plot size was 3. 7 by 12.2 m. Granular grade KCl (500 g K kg-1 ) was applied at 1340 (low rate) and 2680 (high rate) kg ha-1 in single applications and at 2680 kg ha in two 1340 kg ha applications during 1987. Langbeinite (LB) (K2SO4.2MgSO4 ; 185 g K kg-1 110 g Mg kg-1 and 224 g S kg-1 ) rates, based on the K equivalent of the KCl rates, were 3630 (low rate), 7260 (high rate) and 7260 kg ha-1 in two 3630 kg ha-1 applications. Phosphogypsum (PG) (231 g Ca kg-1 169 g S kg', 3 g P kg-1 and 6 g F kg-') rates, based on the amounts of S equivalent to the LB rates, were 4815 (low rate), 9630 (high rate), and 9630 kg ha' in two 4815 kg ha-1 appli cations. Plots were initially topdressed on 23 May 1987 and the second half of the split applications was applied on 4 Oct. 1987. The PG amendment treatments and the control also received 336 kg K ha-1 as KCl to supply K at or slightly above levels typically recommended for alfalfa production in Louisiana. All plots received 100 kg P ha as triple superphos-

PAGE 156

PROCEEDINGS, V0LCME 51, 1992 131 phate. On 16 Mar. 1988, borax was applied to all plots at a rate of 3 kg B ha 1 Alfalfa was harvested at 25% bloom to a 10-cm stubble. Four harvests were made in 198 7 and five harvests in 1988. In July 1987 and 1988, the experiment was sprayed with sethoxydim to control gooseg rass (Eleusine indica L. Gaertn.) and dallisgrass (Pas palum dilatatum Poir.). The alfalfa also was sprayed as needed and, according to Louisiana State University guidelines, to control insects. Severe flooding prevented continuation of the study in 1989 but, even under more favorable conditions, alfalfa does not persist for more than 2 or 3 yr in southern Louisiana. Herbage yields were obtained from a 0.8-by I2-m swath cut through the center of each plot. Harvested material was weighed and a I-kg subsample collected. Subsamples were weighed and dried for 2 d in a forced-air drying room at 65C. Following drying, samples were reweighed and dry matter yields calculated. In September 1988, rooting depth was observed by the trench-face technique (Troughton, 1981 ). Six composited soil samples were collected in I5-cm increments (to a depth of 60 cm) from each plot 90 d (1987) and 1 yr ( 1988) after initial amendment application. Rainfall for the 90-d period was 610 mm, while rainfall for the year was 1830 mm; 950 mm occurred following the 4 Oct. 1987 split application. Soil samples were air-dried and ground to pass a 2-mm screen prior to analysis. Soil pH measurements were made in water following a 2-hr equilibration of a 1: I soil:liquid suspension. Exchangeable Al was measured after a 30-min extraction with IM KCl using a soil-to-solution ratio of 1: 10 (Bache and Sharp, 1976). Exchangeable Ca, Mg, and K were extracted by the neutral IM NH4OAc method outlined by Thomas (1982). Sulfur was extracted by the 0.5M NH4OAc + 0.25M HAc procedure of Bardsley and Lancaster (1960). All extractions were made on a reciprocating shaker at 180 strokes min-1 and elemental analyses of Al, Ca, Mg, K, and S were performed by 5.81 5.6 :~\ Al ~---5.4 5.2 0-15 15-30 30-45 Depth (cm) 45-60 1.5 1.2 ,.... Cl ...: 0.9 .-.. .:!:. 0 E o.s u_ <( 0 Fig. 1. Influence of soil depth on pH and exchangeable aluminum in an Olivier silt loam soil, 1988. inductively coupled plasma en11ss1on spectroscopy (ICPES). Soluble Cl was determined by standard AgNO3 titration of a I :8 soil:water extract (Jackson, 1958). Subsoil saturation extracts from composited replicates also were prepared and analyzed for Al by ICPES (Bache and Sharp, 1976). Duplicate 400-g samples of soil (oven-dry basis) were saturated with deionized water, equilibrated at laboratory temperature for 20 h, and transferred to a Buchner funnel lined with a 0.45-m filter paper for vacuum extraction of the soil solution. Data were analyzed using analysis of variance. Differences among treatment means were determined using Fisher's protected LSD test (Steel and Torrie, 1980). RESULTS AND DISCUSSION Soil Factors Soil pH and exchangeable Al were not influenced (P > 0.20) by any of the amendments in either 1987 or 1988 at soil depths ranging from O to 60 cm. Exchangeable Al increased with soil depth while pH decreased (Fig. I). Critical soil exchangeable-Al levels for alfalfa range from 0.2 to 0.9 cmol( +) kg 1 or from 4 to I 9% of the exchange complex (Moschler et al., 1960; Foy, 1964). Subsoil exchangeable-Al levels in this study were within or above the former range depending on soil depth (Fig. 1). Despite these moderate levels of subsoil exchangeable Al, subsoil solution Al levels were extremely low (Fig. 2). This phenomenon may be more common than previously believed in the southeastern USA, even in Ultisols with exchangeable-Al levels > 2 cmol( +) kg' (Rechcigl et al., 1988; McCray and Sumner, 1990; Mathews and Joost, 1990). The amendments produced no differences in subsoil exchangeable Ca in 1987 or 1988. Mean exchangeable-Ca levels in 1988 were 4.75 (SE = 0.53), 4.64 (SE = 0.37), and 4.09 (SE = 0.42) .-.. ~6 ~5 ::I ...... <( 4 C: .2 3 :i 0 en 2 I I Control [,2 KCI fl Langbeinlte w Phosphogypsum 30-45 45-60 Depth (cm) Fig. 2. Influence of the high rates of KCl (2680 kg ha'), langbeinite (7260 kg ha'), and phosphogypsum (9630 kg ha') applied in single applications on subsoil solution aluminum, 1988.

PAGE 157

132 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA cmol( +) kg-1 at the 15-to 30-cm, 30-to 45-cm, and 45to 60-cm soil depths, respectively. These relatively high amounts of subsoil exchangeable Ca in comparison to native levels ( < 2 cmol( +) kg-1 ; Miller et al., 1986) may be due to Ca leaching resulting from previous applications of ammonium nitrate and ammonium sulfate to the long-term cotton and corn crops, which had also received surface applications of limestone. As was expected, the PG amendments increased surface soil exchangeable Ca, though there were no differences among the PG treatments (Fig. 3). High heterogeneity was apparent for exchangeable Mg and reduced the ability to detect treatment differences. Exchangeable Mg was increased (P < 0.05) above that of the control within the 0-to 15-cm depth by LB application, though there were no differences among LB treatments (Table I). The high rate of LB applied in a single application also increased exchangeable Mg above the control within the 15-to 30-cm (P < 0.10), 30-to 45-cm (P < 0.05), and 45-to 60-cm (P < 0.10) soil depths. The high rates of KC! and LB increased (P < 0.05) exchangeable K over the control within the Oto 15-cm depth, though there were no differences between methods of application or K carrier (Table 2). The high rate of LB applied in single or split applications also increased (P < 0.05) exchangeable K over the control in the 15to 30-cm depth in 1988. Slightly greater mobility of K applied as LB than as KC) may have been due to competition of K with the Mg contained in the LB for cation exchange sites of the surface soil. Regardless of K source, these data indicate that most of the K remained in the top 30 cm of soil. Similar results were obtained by Rominger et al. (1977) while working with high rates of KCl and K2SO4 surface applied to a low fertility Alfisol in Wisconsm. The downward movement of Cl over time when KCl was applied at 2680 kg ha-1 in a single application 10 9 8 7 ...... 6 .:!:. 0 5 E u 4 ti 3 2 0 ----, I I Control -V,J Phosphogypsum (lo;-! [ i~J Phosphogypsum (sp~t-hlgh) Phosphogypsum (high) LSD(0.05) = 1.74, (0.10) = 1.44; SE= 0.58 v ff,?;-m --v/j~~ ____ J_/ / ---0-15 Depth (cm) Fig. 3. Influence of the low rate of phosphogypsum (4815 kg ha-1 ) applied in a single application, and the high rate of phosphogypsum (9630 kg ha') applied in split and single applications on surface-soil exchangeable Ca, 1988. Table I. Influence of KC!, langbeinite (LB), and phosphogypsum (PG) application on exchangeable Mg in an Olivier silt loam soil, 1988. Depth (cm) Amendment Rate Application 0-15 15-30 30-45 45-60 kgha-1 ----Mg, cmol( +) kg 1 ------Control 0.96 1.75 2.38 2.46 KC! !340 single 1.18 1.75 2.42 2.55 KC! 2680 split 0.81 1.21 1.99 2.37 KC! 2680 single 1.20 1.84 2.13 2.35 LB 3630 single 1.92 2.36 2.68 2.78 LB 7260 split 2.31 2.39 2.50 2.56 LB 7260 single 2.08 2.50 3.12 3.00 PG 4815 single 1.08 1.60 2.23 2.50 PG 9630 split 1.19 1.89 2.45 2.38 PG 9630 single 0.83 1.63 2.42 2.56 LSD(0.05) 0.63 t 0.52 SE 0.22 0.30 0.18 0.14 tNonsignificant at the 0.05 level using Fisher's protected LSD test, but LSD(0.10) = 0.73 for 15-30 cm depth and LSD(0.10) = 0.35 for 45-60 cm depth. is presented in Fig. 4. Loss of most of the Cl after 1 yr (the 1988 data) suggests little impedence to downward anion movement in this soil at depths of 0 to 60 cm. Chloride is known to be highly mobile in most soils and does not accumulate, except where rainfall is low (Rominger et al., 1977). The progressive downward movement of S with time at the high rates of LB and PG applied in a single application (7260 and 9630 kg ha 1 respec tively) is illustrated in Fig. 5. Soil from plots receiving the high rates of LB and PG had greater (P < 0.05) extractable S than did control plots in the 0-to 15-cm layer, but there were no significant differences between S carriers (Table 3). Extractable S also was greater in LB-and PG-amended plots than in control plots for both the 15-to 30-cm and 30to 45-cm soil layers. Table 2. Influence ofKCl, langbeinite (LB), and phosphogypsum (PG) application on exchangeable K in an Olivier silt loam soil, 1988. Depth (cm) Amendment Rate Application 0-15 15-30 30-45 45-60 kgha1 ------K, cmol( +) kg1 ------Control 0.30 0.18 0.20 0.14 KC! 1340 single 0.53 0.25 0.20 0.12 KCI 2680 split 0.69 0.26 0.16 0.16 KCI 2680 single 0.93 0.44 0.19 0.20 l.R 3G'.,O single 0.46 0.31 0.21 0.19 LB 7260 split 0.9i 0.54 0_25 0.18 LB 7260 single 0.i3 0.63 0.36 O.~O PG 4815 single 0.34 0.23 0.19 0.17 PG 9630 split 0.31 0.21 0.16 0.17 PG 9630 single 0.25 0.19 0.16 0.15 LSD(0.05) 0.38 0.27 t t SE 0.13 0.09 0.05 0.02 tNonsignificant at the 0.05 level using Fisher's protected LSD test.

PAGE 158

PROCEEDINGS, VOLUME 51, 1992 133 0--------------------~ 15 E I ; 301 a I GI C 45 60 0 1987 ) j 50 100 150 200 250 300 0,-----------------------15 e 30 t GI C 45 1988 60 -------------------+--~ o 50 100 150 200 Cl (mg kg1 ) 250 300 Fig. 4. Effects of time and soil depth on levels of soil chloride extracted from the control and the plots receiving the high rate of KCI (2680 kg ha1 ) in a single application, 1987 and 1988. In the 30-to 45-cm layer, the high rate of LB (7260 kg ha-1 ) applied in single or split applications increased (P < 0.05) extractable S above that of the other amendments. At the 45. to 60-cm layer, the high rate of LB (7260 kg ha1 ) applied in a single application also increased (P < 0.05) extractable S above that of the other amendments. The high rate of LB (7260 kg ha1 ) applied in a split application and the high rate of PG (9630 kg ha1 ) applied in a single application also increased (P < 0.05) extractable S above that of the control, though values were not different (P > 0.05) from each other or from those for the low rate of LB (3630 kg ha-1). Like the greenhouse study of Mathews and Joost (1990), this study demonstrated that LB-supplied S moved deeper and faster into the subsoil than PG supplied S. Differences between LB and PG can be attributed to LB being approximately 130 times more soluble than PG at 25C (Mathews and Joost, 1990). The fact that considerable S moved into the sub soil without reducing exchangeable Al levels suggests that little precipitation of Al-hydroxy-sulfate miner als, or indirect precipitation of hydroxy-Al in rc sponse to S042 for OH ligand exchange on hydrous oxide surfaces, occurred in this soil. This is in agree 0------------------~ Con,trol I I 15 e .._ 30 :[ GI Q 45 0 X 1987 La~. 8---I Phosphogypsum l) 60 ---+----t-----+---+---+---+-----+---0 50 100 150 200 250 300 350 400 450 500 0---------------------, 15 e .,2. 30 .:: C. GI Q 45 1988 0 60 ---+---+---+---+---+---+-~---+----~ 50 100 150 200 250 300 350 400 450 500 S (mg kg.1 ) Fig. 5. Effects of time and soil depth on levels of soil sulfur extracted from the control and the plots receiving the high rates of langbeinite (7260 kg ha1 ) and phosphogypsum (9630 kg ha) in single applications, 1987 and 1988. ment with results from a 3.yr field study (Gigger silt loam soil; Typic Fragiudalf, finesilty, mixed, ther mic) recently completed by Caldwell et al. (1990). It is possible that formation of Alhydroxy-sulfate min erals may only be important in situations in which Table 3. Influence of KCI, langbeinite (LB), and phosphogypsum (PG) application on extractable S in an Olivier silt loam soil, 1988. Amendment Control KC! KC! KC! LB LB LB PG PG PG LSD(0.05) SE Rate kgha' 1340 2680 2680 3630 7260 7260 4815 9630 9630 Depth (cm) Application 0.15 15-30 30 45 single split single single split single single split single S, mg kg1 --!ti 44 40 28 23 35 30 27 19 35 34 28 21 35 34 20 47 113 114 74 159 209 191 113 I 04 I 92 2 10 I 68 67 IOI 88 43 118 144 89 47 125 156 139 86 57 57 48 52 20 20 17 18

PAGE 159

134 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA subsoil pH is below 5.0 and solution Al high, either naturally or as a result of ammendment-induced hy drolysis of exchangeable Al (McCray and Sumner, 1990; Mathews and Joost, 1990). Precipitation of hydroxy-Al in response to SO4 2 for OH ligand exchange is probably highly dependent on soil levels of SO4 2-sorbing Al and Fe oxides (Alva and Sumner, 1990). Based on this, Sumner ( 1990) proposed that response to PG in the field may be predicted by the degree of salt sorption as re flected by electrical conductivity (EC) values. Responsive soils should give sorptionvalues 2 0. The Olivier subsoil gave sorption values of -217 and -146 mg kg1 for the 30-to 45-cm and 45-to 60-cm soil depths, respectively, correctly predicting the lack of soil response observed in the present short-term study. The pH of the Olivier subsoil (samples prepared in the same manner as for EC) was found to be 0.17 pH units higher in 0.005M CaSO4 than in 0.005M CaCl2 for both the 30-to 45-cm (4.83 vs. 4.66) and 45-to 60-cm (4.75 vs. 4.58) layers, suggesting that some replacement of OH by SO12 may take place when PG or LB is applied to the Olivier soil (Sumner, 1990). However, the data of Sumner ( 1990) also indicate that most PG-responsive soils give pH values that are 2 0.25 units greater in 0.005M CaSO4 than in 0.005M CaCl2 Since many of the Alfisols and Ultisols in southern Louisiana are of mixed or smectitic mineralogy (Amacher et al., 1989) and therefore gen erally lower in free Fe and Al oxides than predominantly kaolinitic soils, they may not be likely to respond to PG application through a SOi for OH-ligand exchange-mediated reduction in exchangeable Al (Alva and Sumner, 1990). The Olivier soil (35 to 40% kaolinite in the clay fraction at depths ranging from O to 60 cm) is known to contain less than half the amount of free Fe and Al oxides (Miller et al., 1986) typically found in PG-responsive soils (predominantly kaolini tic clay fractions; Alva and Sumner, 1990). As shown in Fig. 2, solution Al concentration was increased by PG and LB application. However, due to increased ionic strength and formation of the rela tively non-toxic AlSO4 + complexion, soil solution AlH activity may actually be reduced (McCray and Sumner, 1990). Since solution Al levels were low in the present study, speciation was not conducted. In any case, subsoil modification by leaching surface-applied amendments is a soil-specific and often slow process under field conditions (Alva and Sumner, 1990; McCray and Sumner, 1990; Sumner, 1990). More field experiments to evaluate long-term soil and crop responses are needed before conclu sions can be drawn on the relative effectiveness of PG and LB in reducing acid-soil infertility problems. Economically, PG has a substantial advantage over LB, since PG is a generally inexpensive material. Root Length Examination of the soil profilt> 1n September 1988 indicated that alfalfa roots penetratcci LO a depth of 55 to 60 cm, regardless of the amendment. Subsoil exchangeable-Al levels (Fig. 1) apparently did not provide sufficient toxic Al in solution to affect the rooting depth of Florida 77 alfalfa. This conclusion is supported by the fact that the concentration of total Al in the subsoil solutions was extremely low (Fig. 2). The high levels of subsoil exchangeable Ca (ratio of cmol( +) exchangeable Ca to cmol( +) exchangeable Al was 2 3) probably also helped prevent the appearance of Al-toxicity symptoms (McCray and Sumner, 1990). It appears that the fragipan (56-to 110-cm soil depth) may have been the most important factor limiting root development in the Olivier subsoil, for roots did not extend far into the pan. The role of fragipans in physically limiting root growth of alfalfa and other crops is well documented (Bradford and Blanchar, 1977). Recent work by Sumner et al. (1990) suggests that long-term PG leaching may result in direct physical improvement of hardpan conditions through improved flocculation and aggregation in subsoils where Ca levels are initially very low. This condition was not evident in the present study, but crop responses to gypsum on soils with hardpans (Caldwell et al., 1990) warrant further investigation. Crop Yields There were no differences (P > 0.20) in alfalfa yields as a result of the surface-applied amendments in 1987 (CV= HU%) or 1988 (CV= 17.6%). Mean yield for the field was greater (P < 0.05) in the second year (9.45 Mg ha1 ) than in the first year (6.89 Mg ha1). This can be attributed in part to the extra harvest in 1988. Although there were differences (P < 0.05) between the amendments with regard to plant K, Ca, Mg, and S concentrations, plant tissue concentrations for the macronutrients (N, P, K, Ca, Mg, and S) were in the sufficiency range for alfalfa (Lanyon and Griffith, 1988) regardless of amendment or year (Mathews and Joost, unpublished data, 1989). CONCLUSIONS Based on the results of this study, leaching of PG or LB appeared to be of little short-term benefit in reducing subsoil exchangeable-Al levels in this Alfisol. This may be because the subsoil pH was too high (2 5.0) and solution Al too low for the formation of Alhydroxy-sulfate minerals. In addition, KCl was not effective in displacing surface-soil Ca into the subsoil. Yield and root growth data from this study demonstrate that Florida 77 alfalfa can be successfully grown on some acid soils with the addition of surface lime and adequate P and K, provided that subsoil solution and exchangeable-Al levels are low enough that root growth is not inhibited. Rechcigl et al. (1988) came to similar conclusions for 'Arc' and 'WL 311' alfalfa. REFERENCES Alva, A. K., and M. E. Sumner. 1990. Amelioration of acid soil infertility by phosphogypsum. Plant Soil 128:127-134.

PAGE 160

PROCEEDINGS, VOLUME 51, 1992 135 Amacher, M. C., W. J. Day, B. A. Schumacher, P. M. Walthall, and B. J. Miller. 1989. A guide to the classification of soils of Louisiana. Louisiana Ag1ic. Exp. Stn. Bull. 803. Bache, B. W., and C. S. Sharp. 1976. Characterization of mobile aluminium in acid soils. Geoderma 15:91-101. Bardsley, C. E., and J. D. Lancaster. 1960. Determination of re serve sulfur and soluble sulfate in soil. Soil Sci. Soc. Am. Proc. 24:265-268. Bouldin, D. R. 1979. The influence of subsoil acidity on crop yield potential. Cornell Int. Agric. Bull. 34. Bradford, J. M., and R. W. Blanchar. 1977. Profile modification of a fragiudalf to increase crop production. Soil Sci. Soc. Am. J. 41:127-131. Caldwell, A.G., R. L. Hutchinson, C. W. Kennedy, and]. E .Jones. 1990. Effect of rates of lime and by-product gypsum on movement of calcium and sulfur into an acid soil. p. 264. In Agronomy abstracts. ASA, Madison, WI. Foy, C. D. 1964. Toxic factors in acid soils of the Southern United States as related to the response of alfalfa to lime. USDA Prod. Res. Rep. #80. Jackson, M. L. 1958. Soil chemical analysis. Prentice-Hall Inc. En gelwood Cliffs, NJ. Lanyon, L. E., and W. K. Griffith. 1988. Nutrition and fertilizer use. In A.A. Hanson (ed.) Alfalfa and alfalfa improvement. Agronomy 29:333-372. Mathews, B. W., and R. E. Joost. 1990. The effects of leaching surface-applied amendments on subsoil aluminum and alfalfa growth in a Louisiana Ultisol. Commun. Soil Sci. Plant Anal. 21 :567-581. McCrav, J.M., and M. E. Sumner. 1990. Assessing and modifying Ca and Al levels in acid subsoils. Adv. Soil Sci. 14:45-75. Miller, B. J., W. J. Day, and B. A. Schumacher. 1986. Loesses and loess-derived soils in the lower Mississippi Valley. Guidebook for soils-geomorphology tour. 1986 ASA Meetings (New Orleans). Louisiana Agric. Exp. Stn., LSU Agric. Center, Baton Rouge, LA. Moschler, W.W., G. D.Jones, and G. W. Thomas. 1960. Lime and soil acidity effects on alfalfa growth in a red-yellow podzolic soil. Soil Sci. Soc. Am. Proc. 24:507-509. Rechcigl, J E., R. B. Reneau Jr., and L. B. Zelazny. 1988. Soil solution Al as a measure of Al toxicity to alfalfa in acid soils. Commun. Soil Sci. Plant Anal. 19:989-100 I. Rominger, R. S., D. Smith, and L.A. Peterson. 1977. Influence of high rates of topdressed KC! and K2SO1 on recovery of K, Cl, and SO4-S by alfalfa and residual amount remaining in the soil. Commun. Soil Sci. Plant Anal. 8:489-507. Steel, R. C. D., and J. H. Torrie. 1980. Principles and procedures of statistics: A biometrical approach. 2nd. ed. McGraw-Hill Book Co., New York. Sumner, M. E. 1990. Gypsum as an ameliorant for the subsoil acidity syndrome. Fla. Inst. Phosphate Res. Pub!. No. 01-024090. Bartow, FL. Sumner, M. E., D. E. Radcliffe, M. McCrav, E. Carter, and R. L. Clark. 1990. Gypsum as an ameliorant for subsoil hardpans. Soil Tech. 3:253-258. Thomas, G. W. 1982. Exchangeable cations. In A. L. Page (ed.) Methods of soil analysis, Part 2. 2nd ed. Agronomy 9: 159-165. Troughton, A. 1981. Root mass and distribution. p. 159-177. In J. Hodgson et al. (eel.) Sward measurement handbook. Brit. Grassl. Soc., Hurlev, Berkshire, UK. Volk, G. M. 1966. Kn~w your fertilizers and lime. Florida Agric. Ext. Serv. Bull. 177B.

PAGE 161

136 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA SOCIETY AFFAIRS MINUTES Fifty-First Annual Business Meeting Soil & Crop Science of Slorida 25 September 1991 DR ROBERT S. K.ALMBACHER The meeting was called to order at 1645h by Pres. R. S. Kalmbacher. Old Business: Dr. Kalmbacher indicated that the minutes of the last business meeting were published in the proceedings. It was moved and seconded that the minutes be accepted as given. Motion passed. A financial report was presented by the Secretary/ Treasurer. It was moved, seconded and approved that this report be accepted as given and that it be published in the proceedings. Standing committee reports ;/ere presented: The audit committee report was given by Dr. Kalmbacher. It was moved, seconded and appro'.'ed that this report be accepted as given and printed in the proceedings. The site selection committee report was presented. A lengthy discussion followed relating to one facility versus another and to possibly changing the dates when the annual meeting would be held. There was a general consensus to hold the next annual meeting at the University Center Hotel during the last week of September, 1992. No formal action was taken. It was announced that the Lifetime Membership and Dedication Committee and the Necrology Committee reports would be given at the luncheon. The Editorial Committee report was presented. Associate editor for Crops, Dr. D. A. Knauft, had asked to step down. Dr. M. J. Williams had agreed to accept responsibilities as associate Crops editor. Dr. F. M. Rhoads had agreed to continue as associate editor for Soils. Dr. W. G. Blue, after serving for five years as editor, had asked to step down. No one had been found to serve as editor. Dr. Kalmbacher agreed to find someone to fill this important role. The nominating committee report was presented. Dr. N. R. U sherwood would be rotating off the board and Dr. D. A. Graetz had been nominated to fill that vacancy for director. Dr. Kalmbacher called for nominations from the floor. Dr. K. H. Quesenberry moved that nominations cease and that Dr. Graetz be elected by acclamation. Dr. J. M. Bennett seconded the mo tion. The motion passed. The committee nominated Dr. G. H. Snyder for president-elect. Dr. Kalmbacher called for nominations from the floor. Dr. Mislevy moved that nominations cease and that Dr. Snyder be elected by acclama tion. Dr. O'Connor seconded the motion. The motion passed. The Membership Committee, a new committee formed in 1991, was chaired by Dr. Brian McNeal. The committee reported that the different departments were well represented on the membership role, but more participation by members in the society meetings should be encouraged. The Sustaining Membership Committee report was presented by Dr. Usherwood. He indicated that membership on the committee should be expanded to more than one person. A letter was being drafted to send out to anyone who might be a candidate for sustaining membership. Also, a folder was being developed that will tell what the Society is about. Dr. Kalmbacher indicated that letters to sustaining members would go out in January.

PAGE 162

PROCEEDINGS, VOLUME 51, 1992 137 Dr. Kalmbacher introduced the idea of possibly publishing abstracts in the 1993 proceedings. There was a great deal of discussion concerning the pros and cons by the membership concerning this :dea. Dr. S. C. Schank suggested that the membership be surveyed concerning this question. Dr. O'Connor moved that the president, Dr. Kalmbacher, develop a survey document about this question as well as other questions that concern the society, distribute it to the membership and report the findings at the next annual business meeting. Dr. Sartain seconded the mo tion. Motion passed. Other Business: Dr. Sartain discussed the student paper contest and asked the membership to cooperate when asked to help with the judging of graduate student papers. Dr. Kalmbacher asked for any other business. Dr. R. A. Dunn moved for adjournment. About 20 members seconded the motion in unison. Motion passed. MINUTES Board of Directors Meeting Soil and Crop Science Society of Florida 25 September 1991 The meeting was called to order by President R. S. Kalmbacher at 1 000h in the Executive Board Room of the Ramada Hotel Resort, International Drive, Orlando, Florida. Other board members present were G. C. Smart, C. G. Chambliss, R. D. Barnett, N. R. Usherwood (1991), F. M. Rhoads (1992), and J. M. Bennett (1993). Old Business: The minutes of the 17 January 1991 board meeting were presented by Secretary/Treasurer, C. G. Chambliss. N. R. Usherwood moved that the minutes be approved. The motion was seconded by J. M. Bennett. Motion passed. Dr. Kalmbacher reviewed the need to obtain a tax exemption number and also a bulk mailing privilege. Dr. Kalmbacher agreed to work on obtaining the tax exemption number in the coming year. New Business: A financial report was given by C. G. Chambliss. The question was raised as to how the society's funds would be disposed of if the society were to be dissol ved. After some discussion, the consensus was that the money could be given to the University of Florida, IFAS, for a scholarship fund. No formal action was taken. The question arose as to how much money we should try to maintain in reserve. The general con sensus was that we should keep enough in reserve to cover one year's expenses. No formal action was taken. Dr. Kalmbacher mentioned that the board of directors and officers turn over fairly rapidly and due to the lack of continuity, policies, etc., established by the board may be forgotten. C. G. Chambliss moved that the incoming president (Grover Smart) appoint a committee to review the minutes of previous board and business meetings with the intent of developing an outline of important policies, rules and regulations that have been passed over the years, so that the present and future board(s) will have this outline to fol low. The motion was seconded by Dr. Usherwood. Motion passed. Dr. Kalmbacher asked for committee reports. The 1991 Program Planning Committee report was given by Dr. Smart. He said there were 62 presently registered, with 69 reservations for the luncheon and 74 for the cookout. Also, only a few changes were needed in the program. He indicated that the hotel management has been very helpful. Dr. Smart also discussed the cookout and program for Thursday night. Dr. Kalmbacher indicated that industry representatives and the Cooperative Extension Service had been very helpful in setting up this event. Dr. Smart said there would be two symposiums in the afternoon, one on Agronomic Nematology, and one on the new Environtron. The business meeting would be held at 4:00 pm and the reception at 6:00 pm. Dr. Smart discussed certain changes in the program. Thursday morning sessions would start at 8:30 am instead of 8:45, and the luncheon from 12:00 to 1 :45, with sessions starting again at 2:00 pm. Other changes were: Buses to load at 5:00 pm instead of 5:30 pm with the steak supper at 6:00 instead of 7:00. These changes were due to the shorter day length at this time of year. The steak supper would be sponsored in part by Dow, Vigoro, IMC and Douglass Fertilizer. Dr. Kalmbacher was congratulated for his ef forts in collecting funds and organizing the cookout. Dr. Kalmbacher presented the Audit Committee report. Records were examined and no discrepancies were found. Dr. Kalmbacher presented the Site Selection Committee report. It had previously been suggested that the committee consider Gainesville due to the present economic situation. After a discussion of different locations in Gainesville, Dr. Rhoads moved to hold the meeting at the University Center Hotel. The motion was seconded by Dr. Usherwood. Motion passed. The Lifetime Membership and Dedication of Proceedings Committee report was presented: Dr. E. S. Horner and Dr. V. W. Carlisle were nominated for lifetime membership. It was proposed that the proceedings be dedicated to Dr. W. G. Blue. A presentation would be given at the luncheon. Dr. Barnett moved that the report be accepted. Dr. Rhoads seconded. Motion passed. Editorial Committee Report: Dr. M. J. Williams had agreed to take over the responsibilities as Crops Editor since Dr. D. A. Knauft is stepping down. Dr. Rhoads had agreed to continue as Soils Editor. At this time, no replacement for Dr. Blue as Editor has been found.

PAGE 163

138 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Dr. Bennett suggested that Dr. P. L. Pfahler be contacted and asked to serve as Editor. Dr. Smart suggested that the Society might want to install a structured system where an individual would start as an associate editor, move up to editor and then out. Dr. Bennett suggested an editorial board of 5 or 6 people to spread the work load. Dr. Rhoads discussed the problems with outside reviews. Dr. Smart suggested that perhaps it would not be so difficult to get people to help if we spread out the work. The Nominating Committee report was presented: Dr. Usherwood rotates off the board and Dr. D. Graetz had been nominated to replace him. Dr. George Snyder had been nominated as president elect. Dr. U sherwood moved the report be accepted and presented to the membership at the business meet ing. Dr. Bennett seconded the motion and it passed. A report from the Membership Committee, chaired by Dr. B. L. McNeal, was presented: Dr. Usherwood suggested that the membership committee should be an on-going committee and that the society also needs a publicity committee and a longrange planning committee. There is a need to focus on the potential membership list put together by Dr. McNeal. The Sustaining Members Committee report was presented by Dr. Usherwood. He indicated that sustaining members (or potential members) have no concern over amount of membership dues. They view our meetings as very scientific. They see our programs as for their Ph.D.'s and would only send 1 or 2 people to get technical information and take it back. Our annual meetings are too technical for industry. They will not send sales people or field agronomists. They can't afford to pull them out of the field. The question is, where does industry fit in? What about a special afternoon or evening or special workshops? Can we be educational as a society? What is the return for industry's dues? Industry is looking for "how do we apply science", best management practices, etc. We are too technical for "practicing agronomists". A discussion followed concerning the possibility of a winter meeting in addition to and separate from our annual meeting to focus on a workshop, etc., to interact with industry. Focus on information useful to industry and focus on a specific topic. Dr. Bennett moved that the letter developed by Dr. Usherwood be modified by the board and sent to dealers of all agricultural products. The motion was seconded and approved. Dr. Smart suggested that sustaining membership be on a calendar year. We need to get communications straight with sustaining members. This may have been resolved, but if not, we need to adjust it. It was also agreed that the idea of publishing abstracts of all papers presented will be brought before the membership at the annual business meeting. The meeting was adjourned at 1200h. MINUTES OF BOARD OF DIRECTORS MEETING 17 January 1991, 310 Newell Hall University of Florida, Gainesville The board meeting was called to order by Dr. R. S. Kalmbacher. Dr. W. G. Blue gave an update on the progress being made in receiving new papers (for the proceedings). He suggested that Dr. D. A. Graetz process the water-quality papers. It was recommended that the keynote speaker's address by published in the section on sustainable agriculture. Dr. E. S. Horner agreed to write a page on the history of the Society by expanding on what Dr. V. E. Green, Jr. wrote for the luncheon program (to be published in Volume 50 of the proceedings). A treasurer's report was given: $10,000 in acertificate of deposit, and $1,346.77 in a checking account. There was discussion of sustaining membership. Dr. Kalmbacher suggested setting up a membership committee. Dr. J. M. Bennett said the young faculty are not participating ... and suggested that a letter and copy of the proceedings as well as a phone call (be sent to those who are not participating). It was suggested that an "elite" committee be set up. Dr. Kalmbacher (Dr. Smart) -We need a flyer that says who we are and what we are. Dr. N. R. Usherwood -moved that the president appoint a membership committee. The motion was second and passed. Omegda Overman was suggested (for the committee). The question was again posed -How to increase participation by young faculty members? (Sugges tions were made for other potential members.) Dr. Usherwood suggested contacting consultants. Dr. Bennett said there is a registry of consultants. Other possibilities are: seed industry reps., environmental engineering firms. Dr. Blue suggested a "symposium on seed". A publicity committee was suggested. Arti cles could be written for Agronomy News, Florida Grower and Rancher, etc. Dr. Blue suggested putting a note in Agronomy News about the 50th Anniversary. Dr. Smart presented comments about the perception of young faculty that must present a paper and publish. Dr. Blue suggested adding an abstract section (to the proceedings). Dr. Bennett gave an example of a student who gave a paper but needed a refereed publication. Dr. Smart moved, Dr. Bennett seconded -that the idea of abstracts be brought up at the next business meeting. Dr. Bennett Abstracts would be published without a full paper. A discussion followed concerning support from the administration. Also, abstracts and how they would be handled was discussed. Dr. Bennett suggested that abstracts could be sent in after the meeting. The 1991 meeting site was discussed: Dr. Smart stated that the proposal to meet with the Lime & Fer tilizer Conference was presented to them and they

PAGE 164

PROCEEDINGS, VOLUME 51, 1992 139 declined. The meeting dates for 1991 will be Wed., Thurs. & Fri., 25-27 September. The meeting site selected was the Ramada Hotel Resort, Orlando. Drs. Kalmbacher and Smart have met with the hotel administration. Meeting rooms will be complimentary. Each meeting room will be set up with tables, water and glasses. The rates for sleeping rooms are as fol lows: $69 single/double, three occupants $84, four occupants $99. A minimum of 25 sleeping rooms per night are needed. Two complementary sleeping rooms will be provided with a total of 90 room nights. An extended discussion followed concerning program planning details. It was stated that we don't have a publication telling what the organization is. Dr. U sherwood agreed to work on this publication. Dr. Blue said he had talked to Dr. Bob Barnes about printing the proceedings. The consensus was to stay with Painter Printing this year. Dr. Usherwood mentioned that we are giving a free registration to the sustaining members and ques tions as to whether or not this needs to be reconsidered. The consensus was to leave as is for now. Dr. Bennett seconded the motion. Respectfully submitted, Carol G. Chambliss Secretary-Treasurer AWARDS LUNCHEON An awards luncheon was held Thursday, 26 September 1991 at noon at the Ramada Hotel Resort with President Robert Kalmbacher presiding. The in vocation was given by Dr. Fred Rhoades and introduction of guests and dignitaries was made by Dr. Kalmbach er. Volume 51 of the proceedings was dedicated to Dr. William G. Blue by Dr. Charles Dean. Honorary lifetime memberships were awarded to Drs. Victor W. Carlisle and Earl S. Horner. Information about their career accomplishments was presented by Drs. Mary Collins and Thomas Obreza, respectively. Outgoing members of the Board of Directors were recognized for their service to the Society. Dr. Ronald Barnett served as Past President and Dr. Noble Usherwood was a Board Representative. Appreciation was also expressed to Drs. William G. Blue (editor), Fred M. Rhoades (soils) and David A. Knauft (crops) for their work on the Editorial Board. A necrology report was made by Dr. Luther Hammond. Dr. Grover C. Smart, President-elect introduced the guest speaker, who was Ms. AmedgaJ. Overman, Professor Emeritus, Gulf Coast Research and Education Center, Bradenton. Her address was "Development of Nematology within the Soil and Crop Science Society of Florida". The gavel was then passed to Dr. Grover Smart as the new President of the Society by retiring Presi dent Dr. Robert Kalmbacher. Dr. Smart presented a plaque to Dr. Kalmbacher for his contributions to the Society as President. Dr. Smart aqjourned the luncheon festivities. GRADUATE STUDENT PAPER CONTEST The fourth annual graduate student paper con test was held as part of the Soil and Crop Science Society of Florida annual meeting. Fourteen well organized papers were presented by graduate students representing the departments of Agricultural Engineering, Agronomy and Soil Science. Five Society members, having expertise in the subject matter of the papers, judged the presentations of the papers in the areas of organization, presentation, visual aids, and subject comprehension. Mr. John B. Hartman, a M.S. student in Agronomy, under the direction of Dr. Ronald D. Barnett, won first place and an award of $ I 00 for the paper entitled, "Hessian Fly Control in Florida Wheat with Systemic Insecticides". Mr. Hartman, a native of Nevada and graduate of the University of Nevada at Reno, received his M.S. from the University of Florida in May 1992 and is pursuing a Ph.D. degree in Genetics and Plant Breeding at the University of California at Davis. Mr. Bruce E. Myhre, a Ph.D student in Agricultural Engineering, under the direction of Dr. Sun-Fu Shih, won second place and an award of $50 for the paper entitled, "Using Remote Sensing and Geographic Information Systems in Water Quality As sessment". Mr. Myhre is currently working full-time at Environmental Services and Permitting, Inc. in Dr. Ronald D. Barnett, North Florida Research and Education Center (right) with his student, Mr. John B. Hartman, winner of first place in the graduate student paper contest.

PAGE 165

140 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA ,r~ I I -~ - --Dr. Sun-Fu Shih, Agricultural Engineering Dep. (right) with his student, Mr. Bruce E. Myhre, winner of second place in the graduate student paper con test. Dr. Donald A. Graetz, Soil and Water Science Dep. (left) with is student, Mr. Rodolfo R. Villapando, winner of third place in the graduate student paper contest. Gainesville, FL, while completing his graduate studies. Mr. Rodolfo R. Villapando, a Ph.D. student in Soil and Water Science, under the direction of Dr. Donald A. Graetz, won third place and an award of $25 for the paper entitled, "Water Table Effects on Nitrogen and Phosphorus Leaching from a Spodosol". Mr. Vil-lapando, a native of the Philippines and a graduate of the University of the Philippines at Los Banos, re ceived his M.S. from West Virginia University in Soil Fertility in May of 1989. He expects to obtain his Ph.D. in May 1993. His dissertation research is entitled "Reactivity and Mobility of Phosphorus in Spodosols. Presentations by graduate students were generally of high quality. We salute these graduate students for their presentations and encourage current graduate students to present their research findings at future meetings of our Society. We also encourage them to prepare manuscripts for publication NECROLOGY Dr. Gaylord M. Volk (82), Professor Emeritus of Soil Science, University of Florida, and Honorary Life Member of the Society, died 31 December 1990, in Tucson, Arizona. He was born 5 August 1908, in Oconto Falls, Wisconsin. Dr. Volk earned the B.S., M.S., and Ph.D. degrees in Soil Science at the Univer sity of Wisconsin ( 1928-32 and 1945-46). He retired in 1975 at age 67. Dr. Volk had a diverse and distinguished career: farmer, scouter for European corn borer, soil surveyor in Wisconsin and in Central America (United Fruit Company), investigator of physical properties of alkaline and gypsiferous soils for the U.S. Soil Conservation Service at Gallup and Albuquerque, New Mexico, Consultant in the establishment of a soils laboratory at the University of San Jose, Costa Rica, and a Researcher Soil Chemist (Professor) at the U niver sity of Florida (1939-1975). Dr. Yolk's early contributions in Florida included the application of soil physical principles to the building of soil surfaces for horse racing tracks and for natural-turf football fields. His major contributions were to Florida Agriculture in soil-plant-nitrogen studies. Management of the ammonia and nitrate nitrogen balance in the soil solved a leaf-roll problem in potato fields of Hastings. Successful management practices for the use of urea as a nitrogen fertilizer were established. He was active in early studies of leaching losses of fertilizer nutrients in sandy soils and the use of slow-release fertilizers to reduce those losses. His research on nitrogen losses due to volatili zation laid the foundation for later investigators. Dr. Volk was a charter member of the Soil and Crop Science Society of Florida and served as President in 1944. Volume 37 of the proceedings of this society was dedicated to him. Honors included membership in Sigma Xi, election as a Fellow of the American Society of Agronomy, and recipient of the 1959 Annual Research Award of the Florida Fruit and Vegetable Association. Dr. Volk is survived by two sons, Dr. Leonard Volk of Tulsa, Oklahoma, and Bill Volk of Salt Lake City, Utah, and three grandchildren.

PAGE 166

PROCEEDINGS, VOLUME 51, 1992 141 RESOLUTION OF SYMPATHY FINANCIAL REPORT Now, therefore, be it resolved that this expression of sorrow over this great loss and of sympathy to the immediate family of the deceased be spread upon the records of this society and a copy of the same be sent to the closest members of the family. Respectfully submitted Necrology Committee C. K. Hiebsch L. C. Hammond, Chair 1 July 1990 through 39 June 1991 ASSETS IN BANK (1 July 1990) Barnett Bank Checking account ..................................................... $10,0 l 7.58 Total in bank ............................................... $10,017.58 RECEIPTS Dues ........................................................................... Regular ...................................................... 2,890.00 Sustaining .. .. .... ... . .... .... .... .......................... .... -0Sale of Proceedings ................................................... Page charges, Volume 49 ......................................... Annual meeting ........................................................ Banquet ........................................................ 150.00 Registration ............................................... 2,585.00 Interest ...................................................................... Checking and savings account .................... 249.90 2,890.00 2,185.00 6,930.00 2,735.00 249.90 Total receipts ................................................................. $14,989.90 DISBURSEMENTS Publication of Volume 50 .......................................... 10,631.33 Postage ... .. .. .. .. .. .. .. . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .... ... 211. 7 4 Expenses of annual meeting ..................................... 2,569.05 Programs and supplies ................................ 849.03 Banquet and breaks .................................. 1,720.02 Office supplies .......................................................... Miscellaneous ............................................................ Barnett Bank CD ..................................................... .. 125.44 105.25 10,000.00 Total disbursements .................................................. $23,642.81 ASSETS IN BANK (30 June 1991) Barnett Bank Checking account ..................................................... 1,364.67 Certificate Deposit .... .... ... .............................. .... .... .... 10,000.00 Total in bank .............................................................. $11,364.67

PAGE 167

142 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA 1992 COMMITTEES (Date in parenthesis indicates year each member rotates off after the annual meeting.) Site Selection/Local Arrangements K. H. Quesenberry, Chair (1992) (1993) (1994) R. S. Mansell J. W. Noling C. G. Chambliss, Ex Officio Audit F. J. Coale, Chair ]. W. Mishoe R. D. Rhue (1992) (1993) (1994) Dedication of Proceedings and Honorary Lifetime Member Selection M. E. Collins, Chair (1992) (1993) (1994) Necrology C. K. Hiebsch, Chair D. L. Myhre E. E. Albregts Nominating E. A. Hanlon, Chair T. A. Kucharek S. Shih (1992) ( 1993) ( 1994) (1992) (1993) (1994) T. A. Obreza S. C. Schank Membership J. B. Sartain, Chair D. W. Dickson A. G. Hornsby A. E. Kretschmer E. B. Whitty P. S. Rao EDITORIAL REPORT The Editorial Committee for Volume 51 of the Proceedings consisted of P. L. Pfahler, Editor; F. M. Rhoades, Associate Editor for Soils; B. L. McNeal, Associate Editor for Soils; and M. J. Williams, As sociate Editor for Crops. The editor recognizes the invaluable contributions of Laura Barry, Nancy Byrd, Marge Lantrip, and Maria Pereira for providing as sistance in the publication of the Proceedings. Sixteen articles were related to soils and 15 articles were related to crops. MEMBERSHIP LIST (1992) (1992) (1993) (1993) (1994) (1994) Respectfully submitted, F. M. Rhoades (Associate Editor, Soils) B. L. McNeal (Associate Editor, Soils) M. J. Williams (Associate editor, Crops) P. L. Pfahler (Editor) In the interest of informing Society members and other readers of the proceedings about membership in our Society, lists of Regular, Subscribing, Honorary Life and Sustaining members follow. MARTIN B ADJEI RR 2 BOX I0,000 KINGSHILL, ST CROIX US VIRGIN ISLANDS 00850 STEPHAN L ALBRECHT BLDG 164 BOX 840 GAINESVILLE FL 32611-0840 EARLE ALBREGTS AREC 13138 LEWIS GALLAGHER RD DOVER FL 33527-9664 REGULAR MEMBERS 1991 L H ALLEN BLDG 164 BOX 840 GAINESVILLE FL 32611-0840 RA ALLEN 2059 CHAMPAGNE DR TALLAHASSEE FL 32308 ASHOK ALVA CITRUS REC 700 EXPERIMENT STN RD LAKE ALFRED FL 33850-2299 L M AMES ENTOMOLOGY/NEMATOLOGY BOX 650 GAINESVILLE FL 32611-0650 CLARENCE B AMMERMAN BLDG 477 BOX 900 GAINESVILLE FL 32611-0900 PAUL R ANAMOSA 5005 S 22 ST ARLINGTON VA 22206

PAGE 168


PAGE 169


PAGE 170


PAGE 171


PAGE 172


PAGE 173


PAGE 174

WILLIAM G BLUE 1501 t\W 31 ST GAINESVILLE FL 32605 FREDERICK T BOYD 5551 NW 4 PL GAINESVILLE FL 32607 VICTOR W CARLISLE 2169 MCCARTY HALL BOX 290 GAINESVILLE FL 32611-0290 RICHARD A CARRlGAN 2475 VIRGI,\/IA AVE NW #304 WASHIN(~TO'J DC 20037 PH EVERETT SW FL REC PO DRAWER 5127 IMMOKALEE FL 33934 NATHAN GAMMON JR 1403 NW 11 RD GAINESVILLE FL 32605 E O Painter Printing Co PO Box 877 DeLeon Springs, FL 32028 Green Cay Farms Attn: T W Winsbcrg Rt l Box 331B Boynton Beach, FL 33437 South Bay Growers Inc Attn: Stewart Swanson PO Drawer A South Bay, FL 33493 American Cyanamid Company Technical Information Service PO Box 400 Princeton NJ 08540 Arkansas State university D B Ellis Librarv Box 2040 State University AR 72467 Auburn University Draughon Library Serials Department Auburn AL 36849 PROCEEDINGS, VOLUME 51, 1992 HONORARY LIFETIME MEMBERS 1991 VICTOR E GREEN 3915 SW 3 AVE GAI:\'ESVILLE FL ,12607-2709 HENRY C HARRIS 4711 NW 39 TER GAINESVILLE FL 32607 J R HENDERSON 2240 NW 16 AVE GAINESVILLE FL 32605 EM HODGES 1510 HWY 64 WEST WAUCHULA FL 33873 EARL S HORNER 304 NEWELL HALL BOX 500 GAINESVILLE FL 32611-0500 DA YID W JONES 12300 NW 56 AVE GAINESVILLE FL 32606 GORDON B KILLINGER 3928 NW 25 CIRCLE GAINESVILLE FL 32606 SUSTAINING MEMBERS 1991 United States Sugar Corp Attn: J B Boy PO Drawer 1207 Clewiston, FL 33440 S Florida Water :Vfgt Dist Reference Center PO Box 24680 West Palm Beach, FL 33416 Desert Ranches of FL Paul Genho Star Route Box 1250 West Palm Beach, FL 33416 SUBSCRIBING MEMBERS 1991 Ballen Booksellers Int., Inc. 125 Ricefield Lane Hauppauge LI NY 11788 Biblioteca Escucla Agricola Panamericana Apartado 93 Tegucigalpa HO'.\DURAS Biblioteca-IVIC PO Box 958(3) Ann Arbor MI 48106-0961 LESTER T KURTZ TURNER HALL AGRONOMY I 102 S GOODWIN URBANA IL 61801 DE McCLOUD 304 NEWELL HALL IFAS 0311 GAINESVILLE FL 32611-0311 0 CHARLES RUELKE 2183 MCCARTY HALL BOX 300 GAINESVILLE FL 32611-0300 KNOWLES A RYERSON 4925 BATTERY LANE BETHESDA MD 20814-4978 ER:--JEST L SPENCER BOX 444 BROOKFIELD VT 05036 GEORGE D THORNTON PO BOX 833 VENICE FL 34284 Diamond R Fertilizer Co Ed Sullivan PO Box 938 Ft Pierce, FL 34954 A Duda and Sons Inc Attn: Dr Larry Beasley PO Box 257 Oviedo, FL 32765 Potash/Phosphate Inst Dr Noble Usherwood 2801 Buford Hwy NE Atlanta, GA 30329 Bibliotheek Landbouwuniv. 60020 Postbus 9100 6700 HA Wageningen THE NETHERLANDS British Library Doc. Sup. Ctr., Acc. Dept. Boston SPA Yorks LS23 7BQ Et\GLAND Buchhandling Behrendt PO Box 1206 D-5300 Bonn 1 WEST GERMANY 149

PAGE 175

150 So11, AI\'D CROP SCIENCE SocIETY OF FLORIDA Canada Department of Agriculture Library Biblioteque Sir John Carling Building Ottawa KI A0C5 CANADA CDM-KARDEX Purdue Univ Library West Lafayette IN 47907-1535 Central Soil Salinity Research Institute Karna! 132001 (Haryana) INDIA CIA T Library PO Box 526050 Miami, FL 33152 Citrus & Subtropical Fruit Research Institute Private Bag X 11208 Nelspruit 1200 SOUTH AFRICA Clemson Universitv The Library Clemson SC 29631 Colorado State University Library Ft Coliins CO 80521 Connecticut Agric. Exp. Sta. Valley Laboratory PO Box 248 Windsor CT 06095 Cornell University Albert R Mann Library Ithaca NY 14853 CSIRO -Librarian Black Mountain Librarv GPO Box 109 Canberr~ ACT 2601 AlJSTRALIA CSIRO -Librarian Davies Laboratory Private Mail Bag PO AITKENV ALE QLD 4814 AUSTRALIA CSlRO Division of Soils Private Bag No 2 Glen Osmond SA AUSTRALIA 5064 CSIRO -Librarian Trop Crops & Pastures Cunningham Lab Carmody Road/St Lucia QLD AUSTRALIA 4067 Curran Associates, Inc. Rt 23 PO Box 620 Claverack NY 12513 Div. Plant Industrv-Librarv Florida Dept. of Agricultur~ PO Box 147100 Gainesville FL 326I4-7100 DSIR -Land Resources Librarian, Private Bag Lower Hutt NEW ZEALAND Dupre Library, USL Serials Dept. 302 E St Mary Blvd Lafavette IN 70504 Florida State Universitv Robert Manning Strozeir Lib. Serials Division Ta!lahassee FL 32306 Hawaiian Sugar Planters' Association Library -Experiment Station PO Box 1057 Aiea HI 96701 Hoveldstreek-Biblioteek Highveld Region Library Private Bag X804 Potchefstroorn 2520 SOUTH AFRICA Indian Agricultural Research Institute (PUSA) Library New Delhi 110012 INDIA Institut fur Tropischen & Subtropischen Pflanzenbau Grisenbachstrabe 6 3400 Gottingen GERMANY International Paper Company Rt I Box 421 Bainbridge GA 31717 Kasetsart University Ser. Sec., Library ESS Bangkhen Bangkok 10900 THAILAND Koninklijk lnstituut voor de Tropen Bibliotheek-SSS \1auritskade 1092 AD Amsterdam HOLLA~D Louisiana State University Library Serials Dept Baton Rouge LA 70803-3342 Mississippi State University Mitchell Memorial Library Serials Tech Ser Mississippi State MS 39762 Mt. Albert Research Center DSIR, Librarian Private Bag, Auckland NEW ZEALAND Na vorsings Institut Vir DIE Direkteur Sitrus en Subtropiese Vrugte Privaat Sak Xll208 Nelspruit 1200 SOl!TH AFRICA Nissin Int. Transport USA, Inc. 12 Thornton Road Oakland NJ 07436-3116 North Carolina State Lniversity DH Hill Library/ACQC PO Box 7111 Raleigh NC 27695-7111 Oklahoma State University The Librarv Stillwater OK 74078 Oregon State University Library Serials Department Corvallis OR 97331 Roval V etcrinary and Ag. College Library 13, Bulowsvej Copenhagen V DENMARK State Library of Florida R A Gray Building Tallahassee FL 32301 Sveriges Lantbruksuniversitets Bibliotek Box 7071 S-7 50 07 lJ ppsala 7 SWEDEN Swets North America PO Box 517 Berwyn PA 19312 Texas A&M University University Library Serials Record College Station TX 77843 Toho University Noogakubu Library Yavoi Bunkyo-Ku Tokyo I 13 JAPAN TVA Technical Library PO Box 1010 Muscle Sholes AL '.~5660 U. Centrale de Versailles Agrement Douane I 115 Route de St. Cyr 78000 Versailles FRA'\JCE UFRGS-Bibliotew Fae. de Agronomia A. Bento Goncalves, 7712 91500 Porto Alegre -RS BRASIL Universidad de Sonora Biblioteca Escuela de Agric y Ganaderia APDO Postal 305 Hermosillo Sonora MEXICO Universidad de! Zulia American Book, Inc. 7824 NW 56 Street Miami FL 33166 University of California Bio-Agriculture Library Serials Department Riverside CA 92!i2 l Cniversity of California Serials Dept. Library Santa Barbara CA 93106

PAGE 176

University of California The C niversity Library Serials Records Section Davis CA 95616-5292 University of Florida CFREC. 2807 Binion Road Apopka FL 32703 University of Florida ERFC Library PO B<;X 8003 Belle Glade FL 33430-8003 University of Florida Southwest Florida AREC 1'0 Drawer 5127 lrnrnokalee FL 33934 L:niversity of Georgia Library/ Acq. Dept. Serials Section Athens GA 30602 L:niversity of Hawaii at 1\fanoa Thomas Hale Hamilton Library Serials Section, 2550 The Mall Honolulu HI 96822 University of Kentucky Agriculture Library Ag. Science Center North Lexington KY 40546 PROCEEDINGS, VOLUME 51, 1992 University of .\finnesota St. Paul Campus Library I 984 Buford Avenue St Paul MN 55I08 University of Nebraska The Library Lincoln NE 68588 University of Puerto Rico Agricultural Experiment Station PO Box 21360 Rio Piedras PR 00928 University of Puerto Rico Estacion Experimental Agricola Apartado 21380 Rio Piedras PR 00928 University of Puerto Rico Gen Library/ Acquisitions Dept. Mayaguez Campus Mayaquez PUERTO RICO University of Queensland St. Lucia Campus Serials Sec./Central Library St. Lucia, Queensland 4072 AUSTRALIA University of Tennessee Library Serials Department Knoxville TN 37996-1000 University of West Indies The Library St. Augustine, Trinidad WEST INDIES University of Wisconsin Steenbock Memorial Library 550 Babcock Drive Madison WI 53706 USDA-ARS Horticultural Research Lab 2120 Camden Road Orlando FL 32803 Virginia Polytechnic Institute University Libraries Serials Receiving PO Boxc 90001 lllacksburg VA 24062-9001 Waite Agricultural Research Institute Private Bag No. 1 Glen Osmond/South Australia 5064 AUSTRALIA Yankee Book Peddler, Inc. PO Box 29418 Maple Street C:ontoocook NH 0,3229 Yuksekogretim Kurulu Dokurnentasyon Merkezi 06530 Bilkent-Ankara TURKEY 151

PAGE 177

152 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Officers of the Society Year President President Elect Secretary-Treasurer Board Directors 1994 D.A. Graetz 1993 J.M. Bennett 1992 G.C. Smart, Jr. G.H. Snyder C.G. Chambliss F.M. Rhoads 1991 R.S. Kalmbacher G.C. Smart, Jr. C.G. Chambliss N.R. Usherwood 1990 R.D. Barnett R.S. Kalmbacher C.G. Chambliss G.C. Smart, Jr. 1989 J.B. Sartain R.D. Barnett C.G. Chambliss R.W. Johnson 1988 P. Mislevy J.B. Sartain D.D. Baltensperger/ J.W. Prevatt/ C.G. Chambliss J.B. Sartain/B.L. '.\1cNeal 1987 D.F. Rothwell P. Mislevy G. Kidder E.E. Albregts 1986 E.B. Whitty D.F. Rothwell G. Kidder D.R. Hensel 1985 J .G.A. Fiskell E.B. Whitty G. Kidder P. 'vf is levy 1984 G.M. Prine JG.A. Fiskcll G. Kidder D.V. Calvert 1983 F.M. Rhoads G.M. Prine G. Kidder G.M. Prine 1982 0.C. Ruelke F.M. Rhoads G. Kidder T.W. Winsberg 1981 A.J. Overman O.C. Ruelke J.B. Sartain F.M. Rhoads 1980 V.W. Carlisle A..J. Overman .J.B. Sartain P.H. Everett 1979 D.W.Jones V.W. Carlisle J.B. Sartain O.C. Ruelke 1978 W .L. Pritchett D.W. Jones J.B. Sartain A..J. Overman 1977 K. Hinson W.L. Pritchett D.W. Jones R.L. Smith 1976 H.L. Breland K. Hinson D.W. Jones A.L. Taylor 1975 A.E. Kretschmer, Jr. H.L. Breland D.W. Jones A.E. Kretschmer, Jr./ G.L. Gascho 1974 L.C. Hammond A.E. Kretschmer, Jr. D.W. Jones H.L. Breland 1973 F.T. Boyd L.C. Hammond D.W. Jones .J.T. Russell 1972 C.E. Hutton F.T. Boyd D.F. Rothwell 1971 E.M. Hodges C.E. Hutton D.F. Rothwell Board of Directors established 1970 W.K. Robertson E.M. Hodges J. NeSmith in 1972 with I, 2, and 3 year 1969 E.S. Horner W.K. Robertson J. NeSmith terms for the first three members. 1968 C.M. Geralclson E.S. Horner J. NeSmith 1967 G.B. Killinger C.M. Geraldson J. NeSmith 1966 C.F. Eno G.B. Killinger J. NeSmith 1965 V.E. Green, Jr. C.E. Eno R. V. Allison 1964 D.O. Spinks V.E. Green, Jr. R.V. Allison 1963 H.C. Harris D.O. Spinks R. V. Allison 1962 W.G. Blue H.C. Harris R.V. Allison 1961 W.H. Chapman W.G. Blue R. V. Allison 1960 J.R. Henderson W.H. Chapman R.V. Allison 1959 P.H. Senn J.R. Henderson R.V. Allison 1958 G.D. Thornton P.H. Senn R.V. Allison 1957 D.E. McCloud G.D. Thornton R. V. Allison 1956 R.W. Ruprecht G.D. McCloucl R.V. Allison 1955 F.H. Hull W. Reuther R.V. Allison 1954 KL. Spencer F.H. Hull R. V. Allison 1953 N. Gammon, Jr. E.L. Spencer R. V. Allison 1952 I.W. Wander N. Gammon, Jr. R.V. Allison 1951 R.A. Carrigan I.W. Wander R.V. Allison 1950 W.T. Forsee, Jr. R.A. Carrigan R.V. Allison 1949 W.T. Forsee, Jr. R.A. Carrigan R. V. Allison 1948 H.A. Bestor L.H. Rogers R.V. Allison 1947 II.A. Bestor L.H. Rogers R.V. Allison 1946 H. Gunter H.A. Bestor R.V. Allison 1945 W.E. Stokes H. Gunter R.V. Allison 1944 G.M. Volk W.E. Stokes R. V. Allison 1943 H. I. Mossbarger G.M. Volk R.V. Allison 1942 J.R. Neller H.I. Mossbarger R.V. Allison 1941 F.B. Smith JR. Neller R. V. Allison 1940 M. Peech F.B. Smith R.A. Carrigan 1939 R. V. Allison M. Peech R.A. Carrigan

PAGE 178

PROCEEDINGS, VOLUME 51, 1992 153 Honorary Life Members Editors* 1950 Selman A. Waksman Charles Franklin Kettering Sir Edward John Russell Merritt Finley Miller Frederick James Alway Sergei Nikolaevitch Winogradsky Walter Pearson Kelley Oswald Schreiner David Jacobus Hissink Charles Ernest Millar John Gordon DuPuis, M.D. 1954 LvmanJames Briggs I 956 Hard rad a Harold Hume Firmin Edward Bear Wilson Popcnoe 1960 Pettis Holmes Senn Knowles A. Ryerson James A. McMurtrey, Jr. Herbert Kendall Hayes Harold Cray Clayton Thomas Ray Stanton Gotthold Steiner Emil Truog John William Turrentine George Dewey Scarseth 1%1 .Joseph R. Neller Howard E. Middleton I %2 Frank L. Holland 1963 Herman Gunter Frank E. Boyd l
PAGE 179

154 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA Dedication of Proceedings Year Volume Person Year Volume Person 1939 I Robert M. Barnette 1966 26 Fred Harold Hull 1940 2 H.H. Bennett & 1967 27 Frederick Buren Smith Selman A. Waksman 1968 28 William Thomas Forsee, Jr. 1941 3 Harry R. Leach 1969 29 Henry Clayton Harris 1942 4-A Spessard L. Holland 1970 30 William G. Kirk 1942 4-B H. Harold Hume 1971-72 31 Alvin Thomas Wallace 1943 5-A Nathan Mayo 1973 32 Curtis E. Hutton 1943 5-B Wilmon Newell 1974 33 E. Travis York, Jr. 1944 6 Herman Gunter 1975 34 George Daniel Thornton 1945 7 Lewis Ralph Jones 1976 35 Marshall 0. Watkins 1946-47 8 Millard F. Caldweld 1977 36 Frederick T. Boyd 1948-49 9 Willis E. Teal, Col. USA 1978 37 Gaylord M. Volk 1950 10 The First 11 Honorary 1979 38 Gordon Beverly Killinger Lifetime Members 1980 39 Nathan Gammon, Jr. 1951 11 Charles R. Short 1981 40 Fred Clark 1952 12 Robert M. Salter 1982 41 John W. Sites 1953 13 J. Hillis Miller 1983 42 William K. Robertson 1954 14 Lyman James Briggs 1984 43 Charles F. Eno 1955 15 Lorenzo A. Richards 1985 44 Theodore (Ted) W. Winsberg 1956 16 T.L. Collins & Fla. Water 1986 45 Gerald 0. Mott Resource Study Commisson 1987 46 JG.A. Fiskell 1957 17 Firmin E. Bear 1988 47 Francis Aloysius Wood 1958 18 Harold Mowry 1989 48 Victor E. Green, Jr. 1959 19 Work & Workers in the Fla. 1990 49 Earl S. Horner Agr. Ext. Serv. 1939-59 1991 50 Amegda J. Overman 1960 20 Roger W. Bledsoe 1992 51 William Guard Blue 1961 21 Doyle Conner 1962 22 R.V. Allison 1963 23 J.G. Tigert 1964 24 J.R. Neller 1965 25 Active Charter Members of the SCSSF *In the earlier years, the Proceedings bear a date which coincides with the year of the annual meeting. During the years of World War II, publication was irregular and at least some of the volumes were published after the war. Also, the actual year that a publication is available becomes the "legal" date of a issue, not the year the meeting was held. Thus, volumes 32 and after bear a date one year after the annual meetings.

PAGE 180

PROCEEDINGS, VOLUME 51, 1992 SOIL AND CROP SCIENCE SOCIETY OF FLORIDA INVITATIONAL REVIEWERS PROCEEDINGS, VOLUME 51, 1992 C. H. Andrews Department of Agronomy P.O. Box 5267 \liss. State University \lis.sissippi State, MS :lll7(i2 P. Hack1nan Auburn University Dept. of Plant Pathology Auburn, AL 36849 K. R. Barker Plant Pathology Department ', .C. State University Ralcip;h, r-;c 27695 J. P. Beasley Jr. Rural Development Center P.O. Box 1209 Tifton, GA 31793 \I. K. lleutc Dept. of Plant Pathologv P.O. 7616 N .C. State L niversity Raleigh, NC 27695 C.R. Bogle S11pcrintendent L pper Coastal Plain St11. Route 2, Box 400 Rocky \fount, NC 27801 W. J. Bourgeois Citrus Research Station Rt. I, Box 628 Port Sulphur, LA 70081 W. D. Branch Agn)nomy Department Coastal Plain Exp. Station Tifton, GA 3 l 79'.l C.R.Camp USDA-ARS l'.O. Box 3039 Flotcnce, SC 29502 E. I'. Caswell Dept. of Nematology University of California Davis, CA 95616 S. H. Chien International Fertilizer Development Center P.O. Box 2040 l\1,1.scle Shoals, AL 35662 S. T. Chu S.D. State Lniversitv Agric. Eng. Department P.O. Box 2120 Brookings, SD 7007 M. D. Clegg I ll2 B Kiesselbach Res. Lib. Lniversity of Nebraska Linrnln. 1',E 68583 R. J. Clements CSIRO Div. Trop. Crops & Pastures, Cunningham Lab :JO{i Carmody Road St. Lucia, Queensland 4067 A.l-STRALIA B. G. Cook P.O. Box 395 Gympie, Queensland 4570 AUSTRALIA B. \I. Cunfer Dept. of Plant Pathology Georgia Agric. Exp. Stn. Experiment, GA 30212 B. E. Dahl Dept. of Range Sciences Texas Tech Universitv I .uhbock. TX 79409 J W. Demski Dept. of Plant Pathologv Georgia Agric. Exp. Stn. Experiment, GA 30212 R.J. Edling I,ouisiana State Universitv Agric. Eng. Department Baton Rouge, LA 70803 R. R. Gangler Dept. of Entomology Rutgers University P.O. Box 2cll ~ew Brunswick, NJ 08903 G. J. Cascho L'niwrsity of Georgia, CPES l'.O. Box 748 Tifton, GA 31793 R. N. Gates Dept. of Agronomy Coastal Plain Exp. Stn. P.O. Box 748 Tifton, GA 31793 T. J. Gerik Texas Agric. Exp. Stn. 808 E. Blackland Road Temple, TX 76502 G. V. Gooding Dept. of Plant Pathology N.C. State University Raleigh, NC 27650 P.H. Graham Soil Science Dept. Cnivcrsity of Minnesota St. Paul, MI 55108 J. B. Hacker CSIRO Div. Tropical Crops and Pastures 306 Carmody Road St. Lucia, Queensland 406 7 Al'STRALIA W. B. Hallmark Iberia Livestock Exp. Stn. P.O. Box 466 Jeanerette, LA 70','J,t C. D. Heatwole Virginia Tech Agric. Eng Dept. 30 I Seitz Hall Blacksburg, VA 24061 H.J Hill 1 719 W. Clinton Fresno, CA 93705 B. W. Hipp Texas A&\! University Res. and Ext. Center 17360 Coit Road Dallas, TX 75252 S. C. Hodges Coastal Plain Exp. Station P.O. Box 1209 Tifton, GA 31793 C. C. Holbrook Jr. USDA-ARS Coastal Plain Exp. Stn. P.O. Box 748 l"ifton, GA 31793 J.E. Houk Agronomy Department P.O. 748 Tifton, GA 31794 W. H. Hudnall Agronomy Department M. B. Stugis Hall Louisiana State Univ. Baton Rouge, LA 70803 R. N. II uettal USDA-ARS Beltsville Agric. Res. Center Beltsville, MD 2070.'i A. W. Johnson USDA-ARS P.O. Box 748 Tifton, GA 31793 R. M. Jones CSIRO Div. of Tropical Crops & Pastures 306 Carmody Road St. Lucia, Queensland 10G7 ACSTRALIA R. E. Joost Department of Agronomy Louisiana State University Baton Rouge, LA 70893 H.K. Kaja Department of Nematolog1 l ni,ersity of California Davis, CA 95616 155

PAGE 181

156 D. L. Karlen USDA-ARS National Soil Tilth Lab 2150 Pammel Drive Ames, IA 50011 D.R. Krieg Dept. of Plant & Soil Sciences Texas Tech. U niversitv Lubbock, TX 79409 S. A. Lewis Plant Pathology Department Clemson Gniversity Clemson, SC 296:\4 D.S. Loch P.O. Box 395 Gympie, Queensland 4570 AUSTRALIA R. H. Loeppert Soil and Crop Sciences Dep. Texas A&M University College Station, TX 77843 David Marshall Texas Agric. Exp. Stn. 17360 Coit Road Dallas, TX 75252 0. G. Merkle Agronomy Department Oklahoma State University Stillwater, OK 74078 M. W. \-lichaud Agric. Exp. Station R.R. 2, Box I 0,000 Kings Hill, St. Croix U.S. Virgin Islands 00850 F. P. Miller Department of Agronomy Ohio State University 2021 Coffey Road Columbus, OH 43210 J. A. Morgan USDA-ARS Crops Res. Laboratory I 70 I Center Avenue Ft. Collins. CO 80526 R. W. Mozingo Tidewater Res. and Cont. Education Center P.O. Box 7217, Holland Stn. Suffolk, VA 23437 Son. AND CROP SCIENCE SOCIETY OF FLORIDA .J.P. l\'oe Plant Pathology Dept. University of Georgia Athens, GA 30602 J. W. Odom Agronomy and Soils Dept. Auburn University Auburn, AL 36849 D. M. Oosterhuis Dept. of Agronomy Cniversity of Arkansas Fayetteville, AR 7270 I J. F. Pedersen Res. Geneticist-Plants USDA-ARS Agronomy Department University of I\' ebraska Lincoln, NE 68583 G. A. Pederson USDA-ARS P.O. Box 5367 Mississippi State, \-IS 39762 D. E. Pettry P.O. Box 5248 Mississippi State, MS 39762 T. 0. Powers Plant Pathology Dept. 406 Plant Science Hall University of N ehraska Lincoln, NE 68583 R. Richaud Dept. of Agronomy Louisiana State Univ. Baton Route, LA 70893 J. W. D. Robbins Irrigation-Mart Louisiana Tech Univ. Route 6, Box 1241 Ruston, LA 71270 R.E. Rouse Texas Agric. Res. & Ext. Ctr. 2415 E. Highway 83 Weslaco, TX 78596 P. L. Sims So. Plains Range Res. Stn. 2000 I 8 Street Woodward, OK 73801 D. H. Smith Agronomy Department C 106 Plant Sciences Colorado State University Fort Collins, CO 890523 E. L. Smith Agronomy Department Oklahoma State University Stillwater, OK 74078 G. R. Smith Texas A&M Uiversity Ag. Research Center Drawer E Overton, TX 77843 L. Stan Dept. Plant Pathology & l'\ematology Texas A&M University College Station, TX 77843 K. C. Stone USDA-ARS P.O. Box 3039 Florence, SC 29502 K. H. Tan Agronomy Departmellt 3111 Plant Science University of Georgia Athens, GA 30602 J. K. Teitzcl Research Station P.O. Box 20 South .Johnstone Queensland 4859 AUSTRALIA A.M.Thro Agronomy Department l,ouisiana State University Baton Rouge, LA 70803 D. C. Wolf Agronomy Department University of Arkansas Fayetteville, AR 7270 I A. G. H. Wollum Soil Science Department Box 7619 :--1. C. State University Raleigh, NC 27695 L. W. Zelazny Agronomy Department \'Pl & State University Hlacksburg, VA 24061