<%BANNER%>

Improving Nitrogen Management in Potatoes through Crop Rotation and Enhanced Uptake


PAGE 1

IMPROVING NITROGEN MANAGEMENT IN POTATOES THROUGH CROP ROTATION AND ENHANCED UPTAKE By FERNANDO MUNOZ A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Fernando Munoz

PAGE 3

This work is dedicated to Luz Marina, Jos Fernando, and Isabela

PAGE 4

ACKNOWLEDGMENTS I would like to express my gratitude to Dr. Rao S. Mylavarapu, my advisor, for his support, time and confidence, and for allowing me develop my ideas about root systems and their relationship with nutrient uptake. My appreciation is also extended to my committee members, Drs. Chad Hutchinson, Kenneth Portier, Thomas Obreza, and Yuncong Li, for their support, useful comments, suggestions, and patience. I thank the staff at the Hastings Research and Education Centers Yelvington Farm in Hastings, Florida especially Pam Solano and Douglas Gergela; it would have been too hard to do this work without their opportune aid. I am so grateful to Dr. Kenneth Portier for all the time devoted to support me with the statistical analysis of my research data. His kindness and patience were so valuable to get this study done. Special thanks go to Dr. Jim Chambers from Louisiana State University and Dr. Nick Comerford at University of Florida for authorizing me to use GSRoot, the automated root length measurement software used here to estimate the potato root parameters. Their kindness facilitated my work. My appreciation and gratitude go to Joseph Nguyen and Martin Anderson; their excellent assistance helped me get this study successfully done. Thanks go to my friends, Jaime Sanchez, Daniel Herrera, and Leighton Walker, for their unconditional friendship throughout these three years in UF. We shared funny talks that softened rough times. iv

PAGE 5

I acknowledge the faculty, students, and staff in the Soil and Water Science Department for allowing me to be here. For sure these three years sharing classrooms, seminars, and informal talks will be significant in my professional life. I would like thank my parents and brothers for their support from a distance. I also thank my aunt Nubia for her unselfish and opportune support. Finally, I want to say thanks to Luz Marina, my wife, and to Jose Fernando and Isabela, my children, for their unconditional support and love. v

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES.............................................................................................................x ABSTRACT......................................................................................................................xii CHAPTER 1 INTRODUCTION........................................................................................................1 Nitrates as a Non-Point Source of Water Pollution......................................................1 The Nitrate Problem in Northeast Florida....................................................................2 Organization of this Dissertation..................................................................................4 2 NITRATE-N LOSSES FROM POTATOES GROWN UNDER DIFFERENT CROP ROTATIONS AND NITROGEN RATES.......................................................6 Introduction...................................................................................................................6 Materials and Methods.................................................................................................9 Site Description.....................................................................................................9 Soils.......................................................................................................................9 Statistical Design and Experimental Layout.......................................................10 Crop Production Practices...................................................................................10 Plant materials..............................................................................................10 Pest management..........................................................................................12 Irrigation management.................................................................................12 Nutrient management...................................................................................12 Water and Soil Sampling Schedule.....................................................................13 Soil sampling................................................................................................13 Water sampling............................................................................................14 Tissue Sampling Schedule...................................................................................16 Potato tissue..................................................................................................16 Sorghum and cowpea tissue.........................................................................17 Potato yield...................................................................................................17 Statistical Analysis..............................................................................................18 Concentrations of NO3-N in the water table................................................18 vi

PAGE 7

Concentrations of NO3-N and NH4-N in the soil.........................................18 Nitrogen uptake by the crops.......................................................................19 Potato yield...................................................................................................20 Results and Discussion...............................................................................................20 Concentrations of NO3-N in the Water Table under Potato................................20 Concentrations of NO3-N in the Soil Solution....................................................30 Concentrations of NO3-N in the Water Table under Cowpea and Sorghum.......33 Concentrations of NO3-N and NH4-N in the Soil................................................34 Nitrogen Uptake by the Crops.............................................................................39 Sorghum.......................................................................................................39 Cowpea.........................................................................................................40 Potato............................................................................................................41 Potato Yield.........................................................................................................42 Conclusions.................................................................................................................47 3 POTATO ROOT DISTRIBUTION STUDY.............................................................49 Introduction.................................................................................................................49 Materials and Methods...............................................................................................51 Site Description...................................................................................................51 Statistical Design.................................................................................................51 Root Sampling Methodology..............................................................................51 Root Washing......................................................................................................53 Root Scanning and Image Analysis.....................................................................53 Root Variables.....................................................................................................53 Soil Strength, Bulk Density, and NO3-N concentration by Depth......................54 Statistical Analysis..............................................................................................55 Root sampler................................................................................................55 Soil strength, bulk density, and NO3-N concentrations by depth.................55 Roots.............................................................................................................56 Results and Discussion...............................................................................................56 Root Sampler Device...........................................................................................56 Soil Strength and Bulk density............................................................................57 Rooting Depth, Root Length, and Root Surface Area.........................................58 Root Length Density............................................................................................60 Root Surface Area Density..................................................................................61 Specific Root Length...........................................................................................63 Specific Root Surface Area.................................................................................64 Conclusions.................................................................................................................66 4 SUMMARY, OVERALL CONCLUSIONS AND FUTURE RESEARCH..............67 Summary.....................................................................................................................68 Crop Rotation and N rates...................................................................................68 Concentrations of NO3-N in the water table under potato...........................68 Concentrations of NO3-N in the soil solution..............................................69 Concentrations of NO3-N in the water table under cowpea and sorghum...69 vii

PAGE 8

Concentrations of NO3-N and NH4-N in the soil.........................................69 Nitrogen recovery by the summer cover crops............................................70 Potato N uptake at full flowering.................................................................70 Potato yield...................................................................................................71 Potato Root Distribution Study...........................................................................71 Root sampler................................................................................................71 Soil strength, bulk density, and NO3-N by depth........................................72 Rooting depth, root length, and root surface area........................................72 Root length density and specific root length................................................72 Root surface area density and specific root surface area..............................73 Overall Conclusions....................................................................................................73 Future Research..........................................................................................................76 APPENDIX A REGRESSION PARAMETERS................................................................................77 B ANOVA TABLES FOR POTATO YIELD PARAMETERS....................................82 C GREEN BEAN YIELD..............................................................................................86 LIST OF REFERENCES...................................................................................................87 BIOGRAPHICAL SKETCH.............................................................................................93 viii

PAGE 9

LIST OF TABLES Table page 2-1 Nitrogen rate treatments and nutrients applied to potato crop each year.................13 2-2 Precipitation during crops and intercrops periods during the study period..............29 2-3 Effect of crop rotation on soil NO3-N concentration at potato planting..................38 2-4 Nitrogen uptake of cowpea by year at incorporation time.......................................41 2-5 ANOVA of the annual potato N uptake over the 3 year period...............................41 2-6 Marketable and Total yield of potato under rotation by N rate................................45 2-7 Tuber exported N, % TKN, and specific gravity of potato under year by N rate interaction.................................................................................................................47 3-1 Soil strength, Bulk density, and Soil NO3-N by depth.............................................57 3-2 Root length and root surface area by diameter class and soil depth. ......................60 B-1 2001 ANOVA table for potato yield parameters...................................................83 B-2 2002 ANOVA table for potato yield parameters....................................................84 B-3 2003 ANOVA table for potato yield parameters....................................................85 ix

PAGE 10

LIST OF FIGURES Figure page 1-1 Localization map of the tri-county agricultural area (TCAA) in northeast Florida...3 2-1 Field plot plan, crop rotations and N rates...............................................................11 2-2 Crop cycles, sampling schedules, and precipitation pattern over the three-year period of study..........................................................................................................15 2-3a Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the first year...............................................................................21 2-3b Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the first year...............................................................................22 2-4a Predicted NO3-N concentration in the water table under four crop rotations and five N rates during the second year..........................................................................24 2-4b Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the second year..........................................................................25 2-5a Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the third year.............................................................................27 2-5b Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the third year.............................................................................28 2-6a Predicted NO3-N concentration in the soil solution under two crop rotations and five N rates during the second year..........................................................................31 2-6b Predicted NO3-N concentration in the soil solution under two crop rotations and five N rates during the second year..........................................................................32 2-7 NO3-N concentrations in the water table under sorghum and cowpea....................33 2-8 Predicted NO3-N concentration in the soil by effect of five N rates.......................35 2-9 Predicted NH4-N concentration in the soil by effect of five N rates.......................36 2-10 Nitrogen uptake of sorghum by residual effect of the fertilizer applied to potato...40 x

PAGE 11

2-11 Potato N uptake at full flowering during the study period.......................................42 2-12 Predicted N rate for maximum marketable potato yield..........................................43 3-1 Procedure to take the root samples...........................................................................52 3-2 Distribution and sample size into the slice...............................................................57 3-3 Detail of potato roots thickening..............................................................................59 3-4 Root length density distribution...............................................................................61 3-5 Root surface area density distribution......................................................................62 3-6 Specific root length distribution ...............................................................................64 3-7 Specific root surface area distribution......................................................................65 xi

PAGE 12

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy IMPROVING NITROGEN MANAGEMENT IN POTATOES THROUGH CROP ROTATION AND ENHANCED UPTAKE By Fernando Munoz August 2004 Chair: Rao S. Mylavarapu Major Department: Soil and Water Science Impact on water quality due to nutrient loading of our finite water resources is a growing concern. Nitrate and phosphates originating from agricultural operations have been identified as the primary contributors to such non-source point pollution of water. The Tri-County Agricultural Area (TCAA) in north-east Florida is predominantly dedicated to intensive potato production system with producers often resorting to over-fertilization, particularly with N sources. Best management practices (BMPs) including crop rotation and enhanced N uptake have been proposed as options to minimize the risk of nitrate contamination of the surface and groundwater in the St. Johns River watershed. To evaluate the effect of NO3-N concentration on shallow water table and to determine optimum N requirement for an economically successful potato crop (var. Atlantic), a 3-year field study with four potato-cover crop rotations and five N rates was conducted in the TCAA. Increased NO3-N concentrations in the shallow water table were recorded immediately after N application at planting but not after the side dressing. At the end of xii

PAGE 13

potato season an increase in soil NO3-N resulting from residual effect of N side-dress was found. While nitrate concentrations produced linear and quadratic increase of sorghum N uptake across N rates, cowpea did not exhibit any response. Rotation with green bean in the fall increased soil NO3-N concentration at potato planting in the next season, and a high rainfall between during the intervening period increased the potential for nitrate leaching and run-off. Potato in rotation with sorghum produced higher yields than in rotation with cowpea. Sorghum was found to be a more efficient catch crop for residual NO3-N. A study of potato root distribution under the effect of three N rates was done in the third year. There was no effect of N rate on root length density, specific root length, root surface area density, and specific root surface. Observations suggested that potato root system was constrained due to soil compaction, and increased soil bulk density recorded at the study site. A soil region where the potato root system showed highest activity was identified which could guide precise fertilizer placement. xiii

PAGE 14

CHAPTER 1 INTRODUCTION Nitrates as a Non-Point Source of Water Pollution Worldwide, agriculture has been identified as a major contributor of nitrate-N to surface water (Randall and Mulla, 2001). A recent study showed that the mean annual input of N to the Gulf of Mexico from the Mississippi-Atchafalaya River Basin, the third largest river basin in the world, was estimated to be 1,568,000 t yr-1 in the period 1980-1996; the mean annual flux of N was about 61% nitrate-N, 37% organic-N, and 2% ammonium-N (Goolsby et al. 2001). The high N load, mainly as nitrate, has been related to increased eutrophication and hypoxia in the Mississippi River Delta (Rabalais et al. 2002). A US nationwide study reported nine times greater total N concentration downstream from agricultural lands than downstream from forested areas (Omernik, 1977). Along with several major horticultural crops, potato (Solanum tuberosum L.) cropping regions have been suspected of adding excess nitrate to surface and underground waters in Germany (Honisch et al. 2002) and Canada (Milburn et al. 1990). The nitrate concentration in rural well water in New Brunswick (Richards et al. 1990) and the nitrate concentration in streams flowing through the cropped fields in Ontario (Hill and McCague, 1974) have been associated with the proportion of the land cropped to potatoes. Therefore, emphasis on how intensive management practices impact the environment has been the primary focus of current day research (Ruser et al. 1998; Honish et al. 2002). 1

PAGE 15

2 Application of higher rates of N than are required for optimum potato production coupled with the susceptibility of NO3leaching by rainfall and irrigation have been the primary focus of current day research (Milburn et al. 1990). Furthermore, studies characterizing N uptake by potato roots have shown that N fertilizers are generally applied in excess of the optimal rate for maximum yield (Tyler et al. 1983; Joern and Vitosh, 1995a; Prunty and Greenland, 1997). Over-fertilization increases the risk of environmental pollution because N not taken up by the plant can move out of the root zone. Synchronization between N availability in the soil and N uptake by the potato root has been proposed as a core strategy to maximize yield and minimize N loss (Ruser et al. 1998; Waddel et al. 2000; Mortvedt et al. 2001). Furthermore, the efficiency of N uptake can be improved by placing fertilizer in bands and/or applying fertilizer in multiple, relatively small applications (Ruser et al. 1998; Waddel et al. 1999). Measures such as suppression of tile-drainage systems, construction of wetland restoration areas before discharge into ditches and rivers, application of correct rate of N at the optimum time, development of improved soil N testing methods to predict availability of mineralized N and carryover from previous crops, improved management of green and animal manures (Randall and Mulla, 2001), and increasing N uptake, and use efficiencies (Errebhi et al. 1998, 1999) are accepted strategies to reduce nitrate losses to surface and groundwater. The Nitrate Problem in Northeast Florida The TCAA (Tri-County Agricultural Area), located in Flagler, Putnam and St. Johns counties in northeast Florida (Fig. 1-1), contains approximately 15,000 ha of irrigated cropland, predominantly potato and cabbage farms. For more than a century, agricultural production and services have contributed significantly to the economic

PAGE 16

3 viability of the basin. Non-point source pollution resulting from storm water runoff from agricultural systems is considered to be one of the major sources of pollution in the low Figure 1-1. Localization map of the tri-county agricultural area (TCAA) in northeast Florida. Approximate location of the study site. St. Johns River Basin (LSJRB). It has been estimated that nonpoint pollutant sources induced by human activities may account for as much as 36% of the amount of pollutants entering the basin today (SJRWMD, 2004). Past work has identified the TCAA as a

PAGE 17

4 possible source of non-point pollution from enriched nutrient runoff. Enhanced algal growth in the St. Johns River coinciding with peak runoff associated with the TCAA potato season have been reported (SJRWMD, 1996). Best management practices (BMPs) to control water runoff and to improve fertilizer management by reduced rates and improved timing lowered the phosphorus runoff into the St. Johns River watershed 22.5 % the first year and about 55 % in 15 years (SJRWMD, 2004). A study comparing the traditional water furrow irrigation-drainage system with a system using subsurface irrigation and drainage showed that surface runoff nitrate concentrations were often higher than concentrations in the subsurface drains (Campbell et al. 1985). Best management practices to minimize potential runoff and leaching of nitrates from potato-cropped lands including nutrient management, irrigation and sediment management, and crop rotation with low N requiring crops have been implemented in the TCAA (Hutchinson et al. 2002). This three year study was initiated to document the impact the impact of legume rotation crops on NO3-N concentrations in the water table under potato beds, N uptake by cover crops, potato tuber production, and possible reduction of the N rate of soluble fertilizers currently used in the TCAA for potato production. Along with this study, a study of the potato root system addressed to describe its spatial distribution under a sufficient N fertilization regime. The knowledge of the spatial distribution of potato would allow in the future optimize fertilizer placement in order to increase potato N uptake. Organization of this Dissertation This work is organized in four chapters. First chapter is a general introduction addressing the leaching/runoff problem from potato fields with emphasis in northeast

PAGE 18

5 Florida potato production. The second chapter reports results of an 3-year experiment with four crop rotations including legume crops as supplemental N source to five N rates applied to potato. Effect of rotation and N rate on NO3-N concentration, N uptake by the crops and potato yield are discussed. The third chapter describe and report results of a study of the potato root distribution under effect of three sufficient N rates. The fourth chapter is addressed to summarize, state general conclusions, and suggest future research topics.

PAGE 19

CHAPTER 2 NITRATE-N LOSSES FROM POTATOES GROWN UNDER DIFFERENT CROP ROTATIONS AND NITROGEN RATES Introduction Potato has been reported as a crop with high potential to generate groundwater contamination due to its high N demand to produce optimum yields. Farmers usually apply N fertilizer in excess of the recommended rate as an insurance against N losses due to N deficiency. In northeast Florida, potato is grown as a late winter or early spring crop using seepage irrigation. Seepage irrigation is a subsurface irrigation system where the depth to or the level of water in the perched water table is controlled taking advantage of the naturally existing hardpan averaging at a depth of 1.5 meters. Seepage irrigation is very unique to this region of the country in contrast with the overhead irrigation used in the rest of the potato producing states in the US. The sandy nature of Florida soils, excess nitrate in the potato beds, and seasons with high rainfall can create conditions resulting in N leaching and/or runoff with the subsequent eutrophication of water bodies. In northeast Florida, most of 9000 ha under potato production are located in the Tri-County Agricultural Area (TCAA) a major agricultural region under the St. Johns River Water Management District (SJRWMD) boundaries. The SJRWMD being responsible for both the water quality and quantity is particularly concerned about the potential nutrient losses from the TCAA to the St. Johns River (Livingston-Way, 2000). The agricultural practice of alternating crops on a particular during a defined period of time (crop rotation) has several benefits, including interrupting plant disease and insect 6

PAGE 20

7 cycles, increasing mineral or organic content, and improving soil structure. This practice can be used to store N in a catch crop during fallow periods, decreasing potential nitrate leaching (Nielsen and Jensen, 1985; Schroder et al. 1996). Cover crops fit the definition of a rotational crop and are used extensively in production systems that utilize organic fertilizers. Cover crops also provide soil coverage during the periods when no commercial crops are grown, serve as a N sink after the commercial crop is harvested, and act as a green manure for subsequently planted cash crops (Goulding, 2000). Using a cover crop as a green manure lessens the impact of nitrate leaching by maintaining the N in a slow-release form (Thorup-Kristensen and Nielsen, 1998). However, the mineralization of green manures under specific environmental conditions must be known to insure that mineralization of residues is synchronized with crop need (Griffin and Hesterman, 1991; Porter et al. 1999; Schmidt et al. 1999). In the US, crop rotation with cover crops is used in areas such as the Great Plains, Florida, the northeast and along the west coast where well drained soils and intensive farming of row crops like potato cause high nitrate concentrations in groundwater (Mueller and Dennis, 1996). It has been further reported that plant residues of arable crops such as potato increased N leaching over at least two winters in free drained soils (Mitchell et al. 2001). The great challenge in establishment of a successful crop rotation with catch crops is synchronizing the availability of nutrients from mineralized residues with nutrient demand of the target crop in the rotation. Generally, most commercial crops are fertilized with close to optimum rates and the primary objective of catch crops would be to replace only part of the soluble fertilizer N (Thorup-Kristensen, 1994).

PAGE 21

8 Factors and processes such as content of nitrate in the soil, N mineralization and biomass production of the catch crop (Jensen, 1992), rooting depth of the catch crop, climate, soil type, and rooting depth of the succeeding crop (Thorup-Kristensen and Nielsen, 1998) must be taken in account to select the most appropriate catch crop species under specific environmental conditions. Nitrogen residues from fertilizers in the soil after potato harvest are generally high and depending of the irrigation applied, could be deep in the soil profile (Webb et al, 2000). Therefore, the catch crop used in rotation with potatoes would require an efficient root system in order to recover the N not taken up by the potato (Thorup-Kristensen and Nielsen, 1998; Singh and Sekhon, 1977). C/N ratio and incorporation time of catch crop must also be considered and adapted to the specific environmental condition (Thorup-Kristensen and Nielsen, 1998). Legume cover-crops such as lupin (Sanderson and MacLeod, 1994), green bean (Hutchinson et al. 2002) and red clover (Porter and Sisson, 1991), and cereal crops wheat and maize (Singh and Sekhon, 1976a, 1976b, 1977), sorghum (Hutchinson et al. 2002), winter rye, wheat, and barley (Shepherd and Lord, 1996; Vos and van der Putten, 2000) have been used as green manures and/or additional cash crops in rotation with potato. Modification of potato production practices including rotation with cash crops and green manure crops adapted to specific environmental situations could help producers increase net revenues by reduction of purchased inputs and increase tuber quality and/or yield (Westra et al. 1995). This research studied the effect of replacing sorghum sudangrass, the traditional summer cover crop used in the TCAA, with cowpea and/or the introduction of green bean as a fall cash crop.

PAGE 22

9 The objectives of this study were i) to identify and to quantify elevations in nitrate concentrations in the perched water table under potato beds due to leaching and ii) to study the effect of legume crops as possible rotation with potato on nitrate leaching and potato yield. Materials and Methods Site Description The study included three potato seasons in the spring of the years 2001, 2002, and 2003 at the UF/IFAS Plant Science and Education Unit, Yelvington Farm in Hastings, FL. The mission of the research farm is to conduct applied and adaptive production research in vegetables, agronomic, and other crops produced in north Florida and to disseminate new information to the academic and agricultural communities. The facility contains 20 ha of land for university research in plant and soil science. Weather data is continuously recorded at the research site through the Florida Automated Weather Network (FAWN) Soils The soil at the experimental site was classified as sandy, siliceous, hyperthermic Arenic Ochraqualf belonging to the Ellzey series (USDA, 1981). This soil has a sandy loam layer about 1 m deep with a clayey restricting layer below. As a result the profile is very poorly drained profile even though the saturated hydraulic conductivity in the top 1 m is about 10 cmh-1. This soil is about 94% sand, 2.5% silt, and 3.5% clay (Campbell et al. 1978). In its natural state, the water table is within 25 cm of the surface for 1 to 6 months in most years and slopes are less than 2% (USDA, 1981).

PAGE 23

10 Statistical Design and Experimental Layout The study was conducted at the same location all the three years of the study. The experiment was replicated four times, and arranged in a split-plot design where the main plot factor was crop rotation and the sub-plot factor was N fertilizer rate applied to potato (Table 2-1). Each block consisted of one bed 16.32 m wide (16 rows) by 160 m long (Fig. 2-1). The block was divided into two strips of 8.16 m (8 rows) by 160 m. Each of the strips (8.16 m x 160 m) was divided again in two parts resulting 4 areas (main plots) of 8.16 m wide (8 rows) by 80 m long in each block (Fig. 2-1). To each of these main plots a crop rotation was randomly assigned for summer and fall. In summer 2000 main plots planted with sorghum and cowpea as summer cover crops (SCC) were randomly assigned. Green bean and fallow main plots were randomly assigned in Fall 2000. In spring 2001, before potato planting each of the main plots was divided into five subplots (8 rows x 16m) and the five nitrogen rates (Table 2-1) were randomly assigned to each of the subplots. Randomization of both the main plot subplot treatments was maintained during the 3-year study period. A code consisting on three letters to identify the four rotations is detailed in Fig. 2-1. Crop Production Practices Plant materials Potatoes (Solanum tuberosum, var. Atlantic) were planted in early February each year. Seed pieces were cut to a target size of 57 g and planted at an in-row spacing of 20.3 cm with a 101.6 cm between-row spacing resulting in a crop density of 48,216 plantsha-1. The sorghum/sudan grass hybrid (Sorghum vulgare x Sorghum vulgare, var. Sudanese, var. SX17) and cowpea (Vigna unguiculata, var. Iron Clay) were seeded in a single row on a 101.6 cm between-row spacing and a within-row seed spacing of 2.5 cm

PAGE 24

Main Plot SpringSummer cover crop Fall crop 123 4PotatoPotatoPotato PotatoSorghum sudanSorghum sudanCowpeaCowpeaFallowGreen beansFallowGreen beans 0 kg/ha N 224 kg/ha N 112 kg/ha N 168 kg/ha N 280 kg/ha N 112 kg/ha N 224 kg/ha N 280 kg/ha N 168 kg/ha N 280 kg/ha N 0 kg/ha N 224 kg/ha N 280 kg/ha N 0 kg/ha N 224 kg/ha N 168 kg/ha N 168 kg/ha N 112 kg/ha N0 kg/ha N112 kg/ha N 280 kg/ha N 168 kg/ha N 224 kg/ha N 280 kg/ha N 0 kg/ha N 112 kg/ha N 0 kg/ha N 224 kg/ha N 0 kg/ha N 168 kg/ha N 280 kg/ha N 168 kg/ha N 0 kg/ha N 224 kg/ha N 112 kg/ha N 224 kg/ha N 112 kg/ha N112 kg/ha N280 kg/ha N168 kg/ha N32143241 0 kg/ha N 224 kg/ha N 280 kg/ha N 280 kg/ha N 112 kg/ha N 168 kg/ha N 0 kg/ha N 168 kg/ha N 112 kg/ha N 112 kg/ha N 168 kg/ha N 0 kg/ha N 0 kg/ha N 224 kg/ha N 280 kg/ha N 112 kg/ha N 280 kg/ha N 224 kg/ha N168 kg/ha N224 kg/ha N4312 0 kg/ha N 112 kg/ha N 224 kg/ha N 224 kg/ha N 112 kg/ha N 168 kg/ha N 280 kg/ha N 280 kg/ha N 168 kg/ha N0 kg/ha N31 0 kg/ha N 168 kg/ha N 224 kg/ha N 112 kg/ha N 280 kg/ha N 0 kg/ha N 224 kg/ha N 168 kg/ha N 112 kg/ha N 280 kg/ha N42 Rows 1-8Rows 9-16Block 1Block 2Block 3Block 4 ********** 11 Plots used for the root studyIrrigation furrow Code PSFPSGPCFPCG Figure 2-1. Field plot plan, crop rotations and N rates

PAGE 25

12 resulting in a density of 392,000 plantsha-1. Cowpea and sorghum were planted in June after potato harvest in each production year as summer cover-crops (SCC). Bush green bean (Phaseolus vulgare, var. Espada) was planted in September in a single drill per row. with in-row spacing of 2.5 cm thus resulting in a density of 392,000 plantsha-1. Cowpea and bush green bean were inoculated with specific Rhizobium strains. Plots not planted with bush green bean in fall were maintained as a weed-free fallow Pest management All ground was fumigated with 1, 3-dichloropropene (56 Lha-1) 4 weeks before planting potatoes for nematode control. Irrigation management The seepage irrigation system uses high row beds to plant potatoes and shallow water furrows to supply irrigation water and to remove drainage water. The natural high water table combined with the sandy and flat nature of the soils and landscape in this area make the irrigation and drainage by the subsurface seepage method possible. The rows were raised to 40 cm in height with a between-row spacing of 102 cm. Each bed consisted of 16 rows separated by irrigation-drainage furrows. The water furrow was about 20 cm deeper than the alley between rows in the bed. During dry periods water was added continuously to the head of the water furrow to maintain the water table 60 cm beneath the top of the bed. Nutrient management Each year at potato planting, 112 kgha-1 N as ammonium nitrate (NH4NO3), was banded and incorporated in the plots receiving N. However, the control plot was not fertilized with any source of N. Two weeks after emergence of potato plants, the remaining N rates of 56, 112, and 168 kgha-1 were side-dressed using the same NH4NO3

PAGE 26

13 at hilling to complete the final treatmental dosages of 0, 112, 168, 224, and 280 kgha-1 of N respectively (Table 2-1). Recommended N rate for potatoes in the region is 224 kg.ha-1 (Hochmuth and Hanlon, 2000). Based on pre-plant soil tests 34 and 168 kgha-1 of P and K, respectively, were incorporated I all the plots prior to potato planting in each production year (Table 2-1). No fertilizers or nutrients were applied at any time to either cowpea or sorghum crops. In each production year, around mid August both sorghum and cowpea (before the cowpea set seed) were chopped and incorporated into the soil as green manure. Bush green beans were fertilized pre-plant and a month after planting with 400 kgha-1 of 14-3-10 and 600 kgha-1 of 14-0-10 fertilizer blends respectively. Total N applied to bush green bean through the split application amounted to 140 kgha-1. Table 2-1. Nitrogen rate treatments and nutrients applied to potato crop each year. N (kg.ha-1) P (kg.ha-1) K (kg.ha-1) At planting Side-dress Total At planting 0 0 0 112 0 112 112 56 168 112 112 224 112 168 280 56 168 Water and Soil Sampling Schedule Soil sampling Soil samples were taken with an open side tubular auger (diam. 1.9 cm) to a depth of 20 cm. A composite soil sample from eight randomly selected points located in the two central rows of the plots was taken at each soil sampling time. A margin of at least 1 m was left at both ends of each sampled row in order to avoid border effects such as a rare sample contamination with soil or fertilizer movement between adjoining plots. Samples were collected in labeled plastic bags, air-dried, and sieved in preparation for subsequent

PAGE 27

14 laboratory analyses. Soil samples were tested for NO3-N, NH4-N, P, K, Ca, and Mg at the UF/IFAS Analytical Research Laboratory (ARL) in Gainesville per the procedures manual (Mylavarapu and Kennelley, 2002). Both NO3-N and NH4-N in solution were analyzed in an air-segmented continuous autoflow analyzer. Mehlich-1 extraction procedure also referred to as the "dilute double acid" extractant (0.05M HCl + 0.0125M H2SO4) was used to extract soil samples and determine P, K, Ca, and Mg by the ICP method EPA 200.7. Results from analyses of only soil NO3-N and NH4-N concentrations analyses were interpreted and reported in this chapter as other nutrients will be outside the scope of this manuscript. The soil sampling schedule throughout the 3-year period is described in Fig. 2-2. Water sampling A 10 cm diameter PVC pipe was buried in each potato plot to a depth of 70 cm from the top of each ridge. The pipe acted as a well casing providing easy access to sample the water table below each plot. Additionally, a lysimeter was buried at 30 cm depth on one of the central ridges to sample water from the soil solution at the root zone during the potato season. During the first year (2001) water was sampled from wells and lysimeters on weekly basis, and on a bi-weekly basis during the second (2002) and third (2003) year (Fig. 2-2). In summer 2003, sampling wells were installed to sample the water table under cowpea and sorghum in the plots where potato was fertilized with 0, 168, and 280 kg.ha-1 of N. Three bi-weekly samplings were taken following the same procedure for water sampling during the potato season. At each sampling time water samples from wells and lysimeters were collected in plastic vials and frozen immediately to measure NO3-N and NH4-N concentrations.

PAGE 28

0510152025 Rainfall (cm)2002 Potato Sorghum& Cowpea Bush green bean 0510152025 2001 Potato Sorghum & Cowpea Bush green bean 0510152025 FebMarAprMayJunJulAugSepOctNovDecJan2003 Water (well)Water (lysimeter)SoilTissue Side-dress N Sorghum &Cowpea Bush green bean Potato Root sampling 15 Figure 2-2. Crop cycles, sampling schedules, and precipitation pattern over the three-year period of study

PAGE 29

16 In the first year and second years wells were installed 3 and 8 days before N side-dressing respectively, and 2 days after side-dressing in the third year. Lysimeters were installed at the end of the potato season (early May), and 1 week before N side-dressing in the second and third years. Due to the late installation of the lysimeters in the first year (early May) and problems with recovery of the water samples from lysimeters during the third year, only samples from the second year were analyzed and included in this dissertation. The wells and lysimeters sampling schedule is illustrated in Fig. 2-2. Water samples from wells and lysimeters were tested at the UF/IFAS ARL in Gainesville. Water samples were tested in an air-segmented continuous autoflow analyzer for NO3-N and NH4-N (Mylavarapu and Kennelley, 2002). Only NO3-N concentrations are analyzed, reported and discussed in this dissertation. Tissue Sampling Schedule Tissue samples from the crops were tested for Total Kjeldahl Nitrogen (TKN), P, and K at the UF/IFAS ARL in Gainesville. Digestates from the Kjeldahl digestion process were tested using semi-automated colorimetric analysis (EPA Method 351.2) to determine nitrogen. The instrument (Alpkem autoanalyzer) was set up and calibrated. Calibration standards and quality control samples were digested in the same manner as the samples as per the procedures described in manual (Mylavarapu and Kennelley, 2002). Potato tissue Potato tissue was sampled at full flowering 60 to 65 days after planting (DAP) in each of the three-years (Fig. 2-2). Shoots of two plants from the two central rows in each plot were cut at ground level and divided in leaves and stems, oven-dried, weighed, ground, and individually labeled. At harvest time a sample of marketable potato tubers

PAGE 30

17 from each plot was taken. Tubers were diced, fresh weighed, oven-dried, dry weighed, ground, and individually labeled. Leaves, stems, and tuber samples were sent to the UF/IFAS ARL in Gainesville for analysis. Pooled TKN concentrations of leaves and stems were analyzed and reported as shoots TKN. Sorghum and cowpea tissue Sorghum and cowpea plants were chopped and incorporated into the soil in August (Fig. 2-2). Before incorporating sorghum and cowpea, a tissue sample of each to estimate biomass and N added to the soil was taken following the same procedure described for potato plant sampling. Ground tissue samples were sent to the UF/IFAS ARL in Gainesville. Potato yield In all potato plots, two-row segments 6.1 m long from the central rows were harvested with a mechanical harvester 100 DAP. A margin of at least 1 m was left at both ends of each sampled row in order to avoid the border effect of adjoining plots. Potato tubers were graded by size (tuber diameter) into five classes (B= <4.8, A1= 4.8-6.4, A2 = 6.4-8.3, A3 = 8.3-10.2, A4 = >10.2 cm). Marketable yield was comprised of classes A1 to A3 only. Total yield was calculated by the summation of all of the classes. Tubers were rated for general appearance, disease, and nematode damage. In this dissertation, total and marketable yield, specific gravity, and % TKN in the tubers were analyzed and reported. Agronomic or physiological interpretation of the yield results was not attempted.

PAGE 31

18 Statistical Analysis Concentrations of NO3-N in the water table The assumptions of homogeneity of variance and normality of the population were checked for the water data sets (wells and lysimeters). Normality was checked by residual analysis using the Shapiro-Wilk test as implemented in the CAPABILITY procedure of SAS (SAS Institute, 2004). Concentration of NO3-N in water was log transformed and checked again for normality. An exploratory analysis consisting of a time series plot to see the general pattern of NO3-N concentration in water, and a general linear model analysis on the log transformed data set were performed. Using orthogonal polynomial contrasts the N rate factor on the subplot was partitioned into linear, quadratic, cubic, and quartic components to determine which regression best described NO3-N concentrations in water with-time. Subsequently, the RSREG procedure of SAS (SAS institute, 2004) was used to estimate the regression parameters and predicted values. The number of days after or before N side-dressing were used as the independent variables. A general linear model by individual sampling times was performed to check the effect of residual fertilization on the NO3-N concentration in the water table under sorghum and cowpea planted in summer 2003. The analysis was implemented as a split plot design where crops (cowpea and sorghum) were the main plot factor and residual effect from the five nitrogen rates applied to potato were the sub-plot factor. Means were separated by the Tukey test as implemented in the GLM procedure of SAS (SAS Institute, 2004) and confidence limits of 95% were calculated for each mean. Concentrations of NO3-N and NH4-N in the soil The soil sampling schedule was not the same for each year (Fig. 2-2) therefore a general linear model by individual soil sampling times was performed and therefore was

PAGE 32

19 not combined over years. The Tukey multiple comparison procedure was used to separate rotation and N treatment means when overall significant effects were observed. The N treatment effect was decomposed into linear, quadratic, cubic and quartic components using orthogonal polynomials in order to identify the response of soil NO3-N to the five N rates applied to potato. Soil NO3-N content at the time of the previous soil sampling was used as a covariate to check the effect of past NO3-N concentration on selected sampling times. Since N treatment had either linear or quadratic effect on soil NO3-N and NH4-N concentrations, regression analysis was used to test hypothesis and estimate regression parameters. A general linear model combining soil samplings at potato harvest was used to compare residual NO3-N before SCC planting. Nitrogen uptake by the crops Statistical analysis of N as TKN in the biomass of the summer cover crops at the time of incorporation into the soil was done to estimate N uptake and N provided by the soil incorporation of each of the crops. The assumptions of homogeneity of variance and normality of the population were checked. Normality was checked by residual analysis using the Shapiro-Wilk test as implemented in the CAPABILITY procedure of SAS (SAS Institute, 2004). A general linear model by individual years was used to determine the effect of crop rotation and N rate on N uptake by the SCC. Using orthogonal polynomial contrasts the N rate factor on the subplot was partitioned in linear, quadratic, cubic, and quartic components to determine the regression that best described total N uptake of the cover crops over-time. Cowpea did not respond to residual N from the potato fertilization therefore no regression was adjusted. Similarly, homogeneity of variance and normality of residues were checked prior to performing the statistical analysis for N uptake values for potatoes. A general linear

PAGE 33

20 model by individual years on potato N uptake at full flowering was performed. N rate factor was partitioned in its linear, quadratic, cubic, and quartic components by using orthogonal polynomial contrasts in order to determine the best regression to describe the effect of N rate on potato N uptake. Finally, the REG procedure as implemented in SAS (SAS Institute, 2004) was used to estimate regression parameters (Appendix A). Potato yield The assumptions of homogeneity of variance and normality of the population were individually checked for potato size classes. A correlation analysis including all of the variables was performed. Normality was checked by residual analysis using the Shapiro-Wilk test as implemented in the CAPABILITY procedure of SAS (SAS Institute, 2004). A general linear model by individual years was used to study yield response to crop rotation and N rate. The LSMEANS with Tukey adjustment as implemented in SAS (SAS institute, 2004) was used to separate individual factor means and/or interaction means when they resulted significant. Subsequently, a general linear model decomposing N rate in its linear, quadratic, cubic, and quartic components was used to determine the best regression explaining the response of marketable potato yield to N rate. The quadratic model explained the response of yield to fertilizer rates. The first derivative of the quadratic regression equation was solved for N rate to determine the N rate required to reach maximum marketable yield. Results and Discussion Concentrations of NO3-N in the Water Table under Potato The time of application of side-dressed ammonium nitrate, approximately 40 DAP (days after planting) was used as the reference point to describe the effect of the five rates of applied fertilizer on the NO3-N concentration in the water table (Figs. 2-3a to 2-5b).

PAGE 34

21 100203040506070-10 PCF 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.79 112 R2 =0.78 168 R2 =0.77 224 R2=0.73 280 R2=0.770203040506070-10Days since side-dress10 PCG 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.81 112 R2 =0.66 168 R2 =0.80 224 R2=0.77 280 R2=0.67 Figure 2-3a. Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the first year. PCF=Potato-Cowpea-Fallow, PCG=Potato-Cowpea-Green bean

PAGE 35

22 PSF 7002030405060-1010100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.76 112 R2 =0.74 168 R2 =0.81 224 R2=0.72 280 R2=0.78NO3-N (mg.L-1) PSG 7002030405060-10Days since side-dress10100.0510.10.0250.515050246810020 N rate (kg.ha-1) 0 R2 =0.64 112 R2 =0.87 168 R2 =0.79 224 R2=0.84 280 R2=0.81 Figure 2-3b. Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the first year. PSF=Potato-Sorghum-Fallow, PSG=Potato-Sorhum-Green bean

PAGE 36

23 The concentration of NO3-N in the water table was explained by linear and quadratic regressions (Appendix A). During the 3 year study period, NO3-N concentration in the water table under the four rotations exhibited decreasing trend from high values before N side-dressing to values close to zero two months after N side-dressing(Fig. 2-3 to 2-5). In the first year, at the time of the first water sampling (3 days before side-dressing) the predicted values for fertilized plots were in the range of 13.5 to 54.5 mgL-1 of NO3-N. The predicted values for the unfertilized plots were between 8.2 and 18.2 mgL-1 of NO3-N. In the PSF rotation the predicted value for the control plot was higher than the value for the 112 kgha-1 N rate suggesting a possible cross-contamination of the control plot from the adjoining plots. The declining trend was explained by a significant linear regression model with coefficients of determination ranging between 0.64 and 0.87 (Fig. 3-2). NO3-N in the water table under plots with the rotations PCF and PSG with 280 kgha-1 of N, and plots with the rotation PSF and 112 kgha-1 of N were described by a quadratic model. In spite of these exceptions, an overall similar pattern was observed for each one of the 4 rotations studied. No differences between N treatments including the control were observed (Fig. 2-3a, b). In the second year, coefficients of determination ranged from 0.57 to 0.32 (Fig 2-4a, b). The predicted values at the first sampling (8 days before N side-dressing) were in a range of 0.3 to 5.0 mgL-1 of NO3-N in the water table. No differences between fertilized and unfertilized plots were observed. The relative low concentrations of NO3-N in the water table in this year produced tendency for lower values of R2 and low adjustment of the models suggesting high variability when NO3-N concentration values

PAGE 37

24 PCF 0203040506070-1010 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.57 112 R2 =0.24 168 R2 =0.46 224 R2=0.44 280 R2=0.26 PCG 0203040506070-10Days since side-dress10 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.37 112 R2 =0.24 168 R2 =0.50 224 R2=0.30 280 R2=0.25 Figure 2-4a. Predicted NO3-N concentration in the water table under four crop rotations and five N rates during the second year. PCF=Potato-Cowpea-Fallow, Figure 2-4a. Predicted NO3-N concentration in the water table under four crop rotations and five N rates during the second year. PCF=Potato-Cowpea-Fallow, PCG=Potato-Cowpea-Green bean 24 PCF 0203040506070-1010 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.57 112 R2 =0.24 168 R2 =0.46 224 R2=0.44 280 R2=0.26 PCG 0203040506070-10Days since side-dress10 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.37 112 R2 =0.24 168 R2 =0.50 224 R2=0.30 280 R2=0.25

PAGE 38

25 PSF 7002030405060-1010 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.05 112 R2 =0.32 168 R2 =0.34 224 R2=0.46 280 R2=0.25 PSG 7002030405060-10Days since side-dress10 100.0510.10.0250.515050246810020 NO3-N (mg.L-1) N rate (kg.ha-1) 0 R2 =0.15 112 R2 =0.53 168 R2 =0.33 224 R2=0.10 280 R2=0.20 Figure 2-4b. Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the second year. PSF=Potato-Sorghum-Fallow, PSG=Potato-Sorghum-Green bean

PAGE 39

26 were close to the analytical method detection limit (0.5 mgL-1 of NO3-N). ). Even though, the low coefficients of determination a similar overall pattern was observed in the water table NO3-N concentration under the four crop rotations (Fig. 2-4a, b). A probable explanation for the low NO3-N concentration values was the low rainfall in the period between potato planting and N side-dressing in 2002 compared with the high rainfall in 2001 and 2003 (Table 2-2). In the third year, water samples from wells were taken only from the plots fertilized with 0, 168, and 280 kgha-1 of N. As in the first year, the NO3-N concentration in the fertilized plots followed a decreasing pattern. The predicted values for fertilized plots in the first sampling (2 days after side-dressing of N) were between 11.0 and 54.6 mgL-1 of NO3-N (Fig. 2-5). These values are similar to the predicted values at the first sampling during the first year. The only difference being sampling time which was 3 days before side-dressing in the first year (Fig. 2-3). Predicted NO3-N concentration values in the water table under fertilized plots treated with different fertilizer rates tended to be similar and unfertilized plots showed low NO3-N concentration values (0.8 to 2.0 mg.L-1) except in PCG rotation (Fig. 2-5a) where the effect of the two legumes in rotation with potato could explain the relatively high NO3-N concentration in the water table under the control plot for this rotation. Similarly to the second year, the relatively low NO3-N concentration in the water table under unfertilized plots produced low R2 and low adjustment to the regression model (Fig. 2-5a, b). The main difference between the first and third year is that the predicted values for the unfertilized plots in the first year were higher compared with the low values for the same plots in the third year. The difference

PAGE 40

27 PCF 0203040506070-1010 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.22 168 R2 =0.44 280 R2=0.60 PCG 7002030405060-10Days since side-dress10 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.82 168 R2 =0.54 280 R2=0.81 Figure 2-5a. Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the third year. PCF=Potato-Cowpea-Fallow, PCG=Potato-Cowpea-Green bean

PAGE 41

28 PSF 7002030405060-1010 100.0510.10.0250.515050246810020 NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.18 168 R2 =0.47 280 R2=0.55 PSG 7002030405060-10Days since side-dress10100.0510.10.0250.515050246810020 NO3-N (mg.L-1) N rate (kg.ha-1) 0 R2 =0.37 168 R2 =0.84 280 R2=0.83 Figure 2-5b. Predicted NO3-N concentration in the water table under two crop rotations and five N rates during the third year. PSF=Potato-Sorghum-Fallow, PSG=Potato-Sorghum-Green bean

PAGE 42

29 among unfertilized plots in the first and third year was due to high residual effect of N in the field from the past season. In each year, a declining pattern in NO3-N concentration starting with higher values prior to side-dressed application of N was a common observation under all rotations and N rates. This was evidently due to the application of 112 kg ha-1 of N through readily soluble ammonium nitrate to all the potato plots that received N treatments at planting when the root development is limited leading to inefficient fertilizer uptake and higher NO3-N concentration in the water table. This hypothesis is supported by the steady decline in NO3-N concentration after side-dress fertilization through the potato season in the fertilized plots. In the 3-year period NO3-N concentration in the water table under the four different crop rotations exhibited a similar pattern within season but different trends were observed during different seasons. In the second year, NO3-N concentrations in the water table between planting and side-dressing N fertilizations under all rotations were lower than 5 mgL-1 compared with the same period on the first and third years when the highest concentrations were close to 50 mgL-1 of NO3-N (Figs 2-3 and 2-5). High amounts of rainfall received during those years (Table 2-2) have resulted in higher leaching of NO3-N. During the first and third years, precipitation received during the month of planting was 47% and 51% higher respectively than the historic 25-yr average (NOAA, 2004) with occurrence of leaching rain Kidder et al. 1992) before the first Table 2-2. Precipitation during crops and intercrops periods during the study period. Potato Crops in rotation with potato Planting to N side-dress N side-dress to harvest Inter crop Sorghum/ Cowpea Inter crop Green bean Inter crop Year Rainfall (cm) 2001 15.98 5.86 16.32 27.37 5.97 54.24 5.87 2002 8.24 7.56 1.79 30.05 13.10 23.81 26.55 2003 22.30 11.18 12.19 38.75 14.48 22.46 6.32

PAGE 43

30 water sampling. The second year was not a typical year with a received precipitation 45% below the average. Therefore, less NO3could have leached from the potato beds in the second year compared to the first and third years. Concentrations of NO3-N in the Soil Solution The concentration of NO3-N in the soil solution under four crop rotations and five N rates during the second year (Fig. 2-6a. b), was explained by the quadratic regression model with coefficients of determination ranging between 0.47 and 0.89 (Fig. 2-6a,b) Eight days before N side-dressing, predicted values fluctuating between 156.0 and 63.4 mgl-1 of NO3-N were estimated for fertilized plots planted with PCG (Fig. 2-6a) and PSG (Fig. 2-6b) rotations, and predicted values for the non-fertilized plots planted under the same rotations were 40.4 and 25.0 mgl-1 of NO3-N respectively. In the plots with crop rotations including fall as a fallow period predicted values for fertilized plots were between 27.1 and 85.6 mgl-1 of NO3-N. Predicted values for non-fertilized plots with the same rotations were around 16.0 mgL-1 of NO3-N (Fig. 2-6a, b). Also, a comparison of predicted NO3-N concentrations at first sampling in PSF and PSG under fertilization showed that the concentration in plots with green bean in fall was twice as the concentration in fallow plots. A similar trend was observed when PCF and PCG were compared. Factors such as residual N from green bean fertilization, increase in the soil N due to the biological N fixation and/or mineralization of crop residues from green bean could have resulted in such a difference in the NO3-N concentration in the soil solution. An elevation in NO3-N concentration in the soil solution was detected between 40 to 60 days after N side-dress (Fig. 2-6a, b). A similar elevation in NO3-N concentration in water samples from the water table beginning 30 days after N side-dress (Fig. 2-6a, b) was detected. These elevations in NO3-N concentration in the soil solution and water

PAGE 44

31 1010.5150502468100200.2 3007002030405060-1010 PCFNO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.78 112 R2 =0.72 168 R2 =0.69 224 R2=0.47 280 R2=0.76 1010.5150502468100200.2 3007002030405060-10Days since side-dress10NO3-N (mg.L-1) PCG N rate (kg.ha-1) 0 R2 =0.82 112 R2 =0.72 168 R2 =0.64 224 R2=0.77 280 R2=0.59 Figure 2-6a. Predicted NO3-N concentration in the soil solution under two crop rotations and five N rates during the second year. PCF=Potato-Cowpea-Fallow, PCG=Potato-Cowpea-Green bean

PAGE 45

32 1010.5150502468100200.2 3007002030405060-1010 PSF NO3-N (mg.L-1)N rate (kg.ha-1) 0 R2 =0.89 112 R2 =0.68 168 R2 =0.87 224 R2=0.70 280 R2=0.81NO3-N (mg.L-1) 1010.5150502468100200.2 3007002030405060-10Days since side-dress10 PSG N rate (kg.ha-1) 0 R2 =0.79 112 R2 =0.63 168 R2 =0.58 224 R2=0.55 280 R2=0.49 Figure 2-6b. Predicted NO3-N concentration in the soil solution under two crop rotations and five N rates during the second year. PSF=Potato-Sorghum-Fallow, PSG=Potato-Sorghum-Green bean

PAGE 46

33 table close to potato harvest coincide with the suggestion of increased nitrification of the NH4+ applied at N side-dress fertilization (Fig. 2-8 and 2-9). The coincidence in the elevation of NO3-N concentration in the soil solution at 30 cm from the top of the ridge with the elevation in NO3-N concentration in the water table 40 cm below of the lysimeter cup suggest leaching of NO3-N. Concentrations of NO3-N in the Water Table under Cowpea and Sorghum In summer 2003, three bi-weekly samplings from the water table under plots planted with cowpea and green bean were taken in order to study the residual effect of the 0510152025303540CowpeaSorghumCowpeaSorghumCowpeaSorghumNO3-N (mg.L-1) 0 kg.ha-1 N 168 kg.ha-1 N 280 kg.ha-1 N First samplingSecond samplingThird sampling 0510152025303540CowpeaSorghumCowpeaSorghumCowpeaSorghumNO3-N (mg.L-1) 0 kg.ha-1 N 168 kg.ha-1 N 280 kg.ha-1 N First samplingSecond samplingThird sampling Figure 2-7. NO3-N concentrations in the water table under sorghum and cowpea. Vertical whiskers are confidence intervals at 0.95 level. fertilizer applied to potato in spring. At first sampling NO3-N concentrations in the water table under cowpea showed a significant increase when the N rate applied to potato in spring was increased from 168 to 280 kg.ha-1 and there was no difference between plots

PAGE 47

34 fertilized with 0 and 168 kg.ha-1 of N (Fig. 2-7). Also at the first sampling there was no difference in the NO3-N concentrations in the water table under plots planted with sorghum. The NO3-N concentrations in water table under sorghum were significantly lower than the NO3-N concentrations under cowpea in plots fertilized with 280 kg.ha-1 of N but there was no difference with cowpea planted in plots where potato was no-fertilized or fertilized with 168 kg.ha-1 of N (Fig. 2-7). These results showed that sorghum was a better NO3-N catch crop compared to cowpea. Including sorghum as a summer cover crop in the potato production rotation was found to have a significant potential in reducing applied N rates, improving N efficiency and minimizing leaching. Concentrations of NO3-N and NH4-N in the Soil In 2001 and 2003, 3 weeks after side-dress N was applied soil NO3-N concentration across N rates showed a positive relationship described by linear and quadratic regressions respectively (Fig 2-8a) Twelve weeks later, at planting of sorghum and cowpea a corresponding trend was detected (Fig 2-8b). The persistence of the N fertilizer in the soil indicated N over-fertilization to potato. Similarly to 2003, at seven weeks after N side-dressing soil NO3-N concentration increased linearly across N rates (Fig 2-8b) Ammonium-N concentration 3 weeks after N side-dressing exhibited a positive relationship described by a quadratic regression in both years (Fig. 2-9a). Prior to side-dressing, concentrations of NH4-N in unfertilized plots (0 N) and plots with fertilization only at planting (112 kg.ha-1 N rate) showed similar values possibly due to both high nitrification rate and/or leaching and runoff due to the high amount of rainfall after potato planting in 2001 and 2003 (Table 2-2). Predicted soil NH4-N concentrations at planting of sorghum and cowpea (12 weeks after N side-dressing) in 2001 and 2002, and 7 weeks after N side-dress in 2003 (Fig. 2-9b) decreased to similar concentration values observed

PAGE 48

35 0102030405060700112168224280Soil NO3-N (mg.kg-1l) 2001 R2=0.53** 2003 R2=0.57**A 0102030405060700112168224280N rate (kg.ha-1)Soil NO3-N (mg.kg-1) 2001 R2=0.51** 2002 R2=0.26** 2003 R2=0.17**B Figure 2-8. Predicted NO3-N concentration in the soil by effect of five N rates. AThree weeks after N side-dressing. B-At sorghum and cowpea planting in 2001 and 2002, and three weeks before potato harvest in 2003. Coefficient of determination significance: p <0.05 and **p<0.01

PAGE 49

36 051015202530350112168224280N rate (kg.ha-1)Soil NH4-N (mg.kg-1) 2001 R2=0.08* 2002 R2=0.25** 2003 R2=0.13**B 051015202530350112168224280Soil NH4-N (mg.kg-1)A 2001 R2=0.64** 2003 R2=0.55** Figure 2-9. Predicted NH4-N concentration in the soil by effect of five N rates. A-Three weeks after N side-dress. B-At sorghum and cowpea planting in 2001 and 2002, and 3 weeks before potato harvest in 2003 Coefficient of determination significance: p <0.05 and **p<0.01.

PAGE 50

37 in unfertilized plots 3 weeks after side-dressing (Fig. 2-9a). In 2003, the decrease in soil NH4-N concentration between N side-dressing (Fig. 2-9a) and potato harvest (Fig. 2-9b), during a relatively low rainfall period (Fig. 2-2), along with increased NO3-N in the seventh week after N side-dressing in 2003 (Fig. 2-8b), showed that increase in NO3-N concentration at the end of the potato season was caused by nitrification of the NH4+ contained in the ammonium nitrate side-dressed to potato. The increase in NO3-N concentration in the soil solution at 30 cm from the top of the ridge between 40 and 60 days after N side-dressing in 2002 (Fig. 2-6a, b) supports the suggestion of increased nitrification before potato harvest. The accumulation of NO3-N in the potato beds at the end of the potato cycle could be explained by the relatively low rainfall between time of side-dress fertilization and harvest of potato compared with the high rainfall between potato planting and side-dress N application (Table 2-2). Furthermore, as seepage irrigation relies on upward capillary movement of water instead of the downward movement of the water under conventional overhead irrigation NO3accumulated in the relatively dry soil on the surface potato beds. Also aerobic conditions on top of the potato beds could result in an increased nitrification favoring build up of soil NO3-N. In 2003, the soil sampling at 3 weeks prior potato harvest revealed increased NO3-N concentration along with the increase in N rate (Fig. 2-8b). Three months later, at sorghum and cowpea incorporation, soil NO3-N concentration dropped close to 0.5 mgkg-1 and no statistical differences were detected among crop rotation or N treatments. Most of the residual NO3-N detected before potato harvest was evidently captured up by sorghum as was found in its increased N uptake values (Fig. 2-10) and/or leached to the water table from cowpea plots as depicted in Fig 2-7.

PAGE 51

38 In 2001 at green bean harvest, there was no difference in soil NO3-N concentration among crop rotations (Table 2-3). Two months later, at potato planting 2002, soil NO3-N concentration in plots planted with green bean in fall was significantly higher than plots left as fallow (Table 2-3). Soil NO3-N concentration in the rotation with cowpea in summer and fallow in fall (PCF) was significantly higher than soil NO3-N concentration Table 2-3. Effect of crop rotation on soil NO3-N concentration at potato planting G. bean harvest Potato planting G. bean harvest Potato planting 11-29-01 02-04-02 11-15-02 02-03-03 Crop Rotation mg.kg-1 PCG 3.90 a 10.58 a 5.11 a 1.94 a PCF 3.00 a 5.29 b 3.90 a 1.30 c PSG 2.96 a 9.91 a 2.86 a 1.56 b PSF 2.55 a 3.99 c 2.12 a 0.95 d Means in the same column followed by different letter are statistically different (p<0.01) in plots with sorghum in summer and left as fallow in fall (PSF) suggesting increased soil NO3-N concentration due to mineralization and nitrification of cowpea residues incorporated in summer into the soil (Table 2-3). Cowpea residues have an average C/N ratio of 15:1 (Havlin et al. 1999) which should enhance N mineralization. In summer 2001, soil incorporation of sorghum in PSF rotation could result in an initial immobilization of N due to its high C:N ratio (60:1) (Havlin et al. 1999). Six months later at potato planting in 2002, soil NO3-N concentration in PSF plots increased suggesting late mineralization of sorghum residues incorporated in summer 2001 (Table 2-3). In 2002 at green bean harvest, similarly to 2001, there were no statistical differences in soil NO3-N concentration (Table 2-3). However, in a couple of months at potato planting 2003, a decreased soil NO3-N concentration was observed (Table 2-3). Declining in soil NO3-N concentration at potato planting in 2003 was due to rainfall

PAGE 52

39 during the intercrop period at the end of 2002 which was five times higher than the rainfall during the similar intercrop period at the end of 2001 (Table 2-2). These results show that the beneficial effect of enhanced N availability by rotation with legumes easily can be eliminated in sandy soils by effect of rainy periods. Therefore, the increase in N available to potato grown in the following spring season from the rotation with green bean should be taken in account to calculate the fertilization rate for the following potato crop but the amount of rainfall should be considered also. Nitrogen Uptake by the Crops Sorghum Rotation did not significantly affect N uptake by sorghum. Significant linear response in the first year and quadratic in the second and third years of N uptake to residual soil NO3-N from the five N rates applied to potato 70 days before summer cover-crops (SCC) planting were observed (Fig 2-10). Overall decreasing sorghum N uptake during the 3-year period was observed (Fig. 2-10). An initial data exploratory study showed sorghum biomass highly correlated (0.91, p<0.0001) with sorghum N uptake, therefore, vegetative growth determined its N uptake. Decreasing N uptake by sorghum in spite of the trend to increased residual NO3-N across N rates through the study period suggests that growth of sorghum was limited by other factors, possibly in combination. Differences in sorghum N uptake among years could be explained in part by the combined effect of different length of the period between planting and soil incorporation and the amount of residual N from potato fertilization. In the first, second, and third years sorghum stayed in the field for 57, 77 and 48 days, respectively. These results reveal sorghum as an efficient NO3-N catch-crop and showed that sorghum grown for an extended period in the field leaving only a short time between potato harvest and

PAGE 53

40 0204060801001200112168224280N rate (kg.ha-1)N uptake (kg.ha-1) 2001 R2=0.36** 2002 R2=0.36** 2003 R2=0.42** Figure 2-10. Nitrogen uptake of sorghum by residual effect of the fertilizer applied to potato. Coefficient of determination significance: p <0.05 and **p<0.01. sorghum planting as feasible measures to reduce potential leaching of residual NO3-N from potato fertilization. The effect of a sorghum crop as an efficient N scavenger will be particularly apparent during the wet summer season (Fig. 2-2) as was showed by the significant less NO3-N leaching under sorghum plots compared with cowpea plots (Fig. 2-7). Cowpea Crop rotation and residual N from potato fertilization had no significant effect on N uptake of cowpea. N uptake of cowpea in control plots (0 N) was similar to N uptake in plots with residual N from fertilization to potato in spring (Table 2-4). This response was expected from a leguminous plant like cowpea, which can meet its N requirement through biological fixation with minimal dependence on soil N availability. Low demand for soil N by cowpea explains the occurrence of the second flux of NO3-N to the water

PAGE 54

41 Table 2-4. Nitrogen uptake of cowpea by year at incorporation time. (kgha-1) N rate (kgha-1) 2001 2002 2003 0 213 a 225 a 59 a 112 204 a 214 a 66 a 168 225 a 163 a 68 a 224 199 a 221 a 68 a 280 191 a 186 a 61 a Means followed by the same letter are not statistically different. table detected after potato harvest in 2003 (Fig. 2-7) when the summer rainfall peak occurred (Fig. 2-2). Nitrogen uptake in 2003 was significantly reduced compared to the corresponding values in 2001 and 2002. However, no difference among N rates persisted (Table 2-4) suggesting that there was another limiting factor at play other than N fixation. Similar to the reduced N uptake observed in sorghum (Fig. 2-10) and potato (Fig. 2-11) in 2003 compared with 2001 and 2002, the reduced N uptake of cowpea suggests the occurrence of a growth limiting factor affecting any crop. Potato Nitrogen rate affected significantly N uptake by potato shoots at full flowering all the three years (Table 2-5). Effect of crop rotation on N uptake of potato was not significant in any of the 3 years suggesting a short term effect of crop rotation (Table 2Table 2-5. ANOVA of the annual potato N uptake over the 3 year period Type III Mean Square 2 Source of variation DF 2001 DF 2002 DF 2003 Replication 1 2 22179.39* 2 3861.55 3 5004.21 Rotation 1 3 1610.55 3 6456.90 3 7750.47 Main Plot Error 6 3312.78 6 2250.81 6 3818.49 N rate 4 21837.80** 4 24594.45** 4 44526.97** Rotation*N rate 12 3509.94 12 1310.37 12 1723.61 Error 59 3878.20 59 1466.22 79 1311.33 1 Main plot effects tested using main plot error. 2 *=significant at type I error <.05. **=Significant at type I error <0.01

PAGE 55

42 5). In 2001, potato N uptake was described by a linear regression and was higher than N uptake in 2002 and 2003 (Fig. 2-11). In 2002 and 2003 potato crop N uptake was described for quadratic and linear regressions respectively, with similar overall N uptake (Fig. 2-11). The linear increase in N uptake across N rates 2001 and 2003 (Fig. 2-11) without response in marketable yield (Fig. 2-12) confirmed luxury consumption of N by Figure 2-11. Potato N uptake at full flowerin potato after it has reach its maximum yield. g during the study period. Coefficient of determination significance: p <0.05 and **p<0.01 Potato Yiel exploratory analysis showed that marketable yield in the N fertilized treatm3. there was a significant interaction between crop rotation and N rate (Appendix B). Yields 0501001502002500112168224280N rate (kg.ha-1)N uptake (kg.ha-1) 2001 R2=0.23** 2002 R2=0.45** 2003 R2=0.51** ld An initia ents were highly and significantly correlated with yield of potato sizes A2 and ATherefore, these two sizes were the main components of the potato yield. In 2001, N rate affected significantly most of the yield components (Appendix B). In 2002 and 2003

PAGE 56

43 Figure 2-12. Predicted N rate for maximum marketable potato yield. NYmax= N rate for maximum yield. Coefficient of determination significance: p <0.05 and **p<0.01. 05101520250112168224280 3035 NYmax=1642001 051015202530350112168224280Marketable Yield (Mg.ha-1) PCF R2=0.81** PCG R2=0.48** PSF R2=0.72** PSG R2=0.91** NYmax=196 NYmax=208 NYmax=210 NYmax=2152002 051015202530350112168224280 PCF R2=0.79** PCG R2=0.92** PSF R2=0.84** PSG R2=0.86** N rate (kg.ha-1) NYmax=234 NYmax=189 NYmax=205 NYmax=2192003R2=0.38**

PAGE 57

44 increased quadratically with increasing N rate. In 2001, there was no interaction between crop rotation and N rate (Appendix B) and The predicted N rate required for potato to reach maximum marketable yield (NYmax) was 164 kg.ha-1 of N for all four crop rotations(Fig. 2-12). In 2002, the predicted NYmax for the four crop rotations was between the range of 196 to 215 kg.ha-1 of N (Fig. 2-12). In 2003, the NYmax ranged between 189 and 234 kg.ha-1 of N (Fig. 2-12). The annual increase in the spread of the range containing the predicted NYmax indicated the effect of crop rotations over-time. A decline in yield with N rates higher than NYmax was observed in the study period (Fig. 2-12). These results were similar to reports of 168 to 196 kg.ha-1 for maximum yield and decreased yields above 280 kg.ha-1 of N in the same site (Hochmuth and Cordasco, 2000). In 2002 and 2003, rotations including sorghum in summer produced higher marketable yield than rotations with cowpea (Fig 2-12) suggesting late mineralization of sorghum residues from the past season as a possible source of N during the initial stages of the potato crop. In 2002, PSG rotation had the lowest NYmax among the rotations studied (Fig. 2-12). This event could be explained by a decreased overall C:N ratio in the PSG rotation caused by the inclusion of green bean which resulted in increased mineralization of sorghum residues and finally more N could be available at potato planting in 2003 reducing in this way the dependence of soluble fertilizers to reach the NYmax. Furthermore, in the intercrop period between green bean harvest-2001 and potato N side-dressing 2002 rainfall was three times less than the same period between 2002 and 2003 (Table 2-2), favoring the persistence in the soil of the min eralized N f rom sorghum and green bean residues (Table 2-3). In the same period in 2003, under a three

PAGE 58

45 times increased rainfall, the PSG rotation exhibited the highest requirement of N for maximum yield (Fig. 2-12) suggesting increased N leaching-runoff (Fig 2-12). Maximum marketable yield decreased over-time from 32 Mg.ha-1 in 2001 to Mg.ha-1 in 2003 (Fig. 2-12). The NYmax reported here (Fig. 2-12) are close to the current recommended N rate of 224 kg.ha-1 for maximum yield in the TCAA (HochmuTable 2-6. Marketable and Total yield of potato under rotation by N rate. 22 th Rotation N rate Marketable yield (Mgha -1 ) Total yield (Mgha -1 ) PCF 0 8.68 d 11.02 d PCF 112 21.31 bc 25.48 ab PCF 168 25.57 ab 29.73 a PCF 224 23.45 abc 27.22 ab PCF 280 22.63 abc 26.30 ab PCG 0 13.92 d 16.52 c PCG 112 23.25 abc 26.73 ab PCG 168 24.34 abc 28.00 ab PCG 224 24.49 abc 28.21 ab PCG 280 20.06 c 23.26 b PSF 0 8.46 e 11.06 d PSF 112 25.88 ab 29.09 a PSF 168 26.79 a 30.26 a PSF 224 24.83 abc 28.31 ab PSF 280 25.58 ab 28.79 a PSG 0 12.93 d 15.24 cd PSG 112 24.62 abc 28.08 ab PSG 168 26.18 ab 29.83 a PSG 224 26.71 a 30.52 a PSG 280 23.25 abc 26.65 ab Means followed by different letter are significantly different. Marketable yield and Cordasco, 2000). The small difference in yield observed among the plots with only fertilization at planting (112 kg.ha N) and plots with additional side-dressed fertilization plots (Fig 2-12) support previous reports that most of the N used for potato under subsurface irrigation in Hastings comes from the fertilization at planting (Elkashif and Locascio, 1983). That report suggests upward movement of soluble salts as water evaporated during dry periods reducing the need of side-dress fertilization. In this study, (p<0.0004). Total yield (p<0.0004). -1

PAGE 59

46 the increased residual effect of the side-dressed fertilization on soil NO3-N concentraclose to potato harvest (Fig. 2-8b) produced increase in sorghum N uptake in sum tion mer (Fig. d ble fertilization at ing anside-dresation have been reported (Hochmuth and Cordasco, 2000). obsern suggssibility of reach higto yields with only one side-ed fertion w timing placement and N rate. The inclusion of green bean in the crop rotatiosed significantly marketable tal yf unfets as compared to unfertilized plots planted with cowpea orghu summft as fallow in fall. Th no effect of N rate on etable total yato grown under the different crop rotations (Table 2-6). bservncrease by inclusion of green bean in fall and the lack of response ld totes cone potential of incorplegumes in fall as a possible ure toce then rates of soluble fertilizers. f potato tubers decreased significantly in unfertilized plots along the three-year period. There were however no differences in exported N from plots that received N rates during the three years (Table 2.7). Along with the decrease in exported N from unfertilized plots a general increase in the % TKN was observed throughout the period of study (Table 2-7). In a preliminary exploratory statistical analysis these two variables were negatively correlated (r=-0.28, p<0.0001) explaining why in spite of the reduction of yield over the three years the exported N in 2-10) supporting the suggestion that a sufficient side-dressed N fertilization wouldecrease the risk of NO3-N leaching-runoff without negatively affect potato marketayield. Higher yield with commercial N fertilization programs using only one side-dress application compared with experimental fertilization programs including plant d s fertiliz This vatio est the po her pota dress tiliza ith optimum ns increa and to ield o rtilized plo and s m in er and le ere was mark and ield of pot The o ed i in yield of yie N ra firmed th orating meas redu applicatio The export of N by total yield o 2

PAGE 60

47 tuber remained statistically the same. Specific gravity (Table 2-7) of fertilized plots was not affected by N rate through the 3-year period. Table 2-7. Tuber exported N, % TKN, and specific gravity of potato under year by N Year N rate N exported (kgha -1 ) % TKN Specific gravity rate interaction. 1 0 40.39 e 0.99 h 1.071 bc 1 100 79.65 cd 1.15 gh 1.078 ab 1 150 89.20 abc 1.23 fg 1.078 ab 1 200 82.15 bcd 1.22 fg 1.079 ab 1 250 85.71 bcd 1.42 ef 1.079 ab 2 0 27.27 e 1.20 fg 1.065 c 2 100 90.07 abc 1.63 de 1.074 abc 2 150 95.51 abc 1.71 de 1.075 abc 2 200 104.68 a 1.84 bcd 1.075 ab 2 250 99.08 ab 1.82 bcd 1.075 ab 3 0 16.93 f 1.75 cd 1.077 ab 3 100 69.01 d 1.79 cd 1.077 ab 3 150 90.85 abc 1.96 bc 1.078 ab 3 200 94.84 abc 2.05 ab 1.079 ab 3 250 99.62 a 2.25 a 1.080 a Means followed by different letter are significantly different. N exported (p<0.0001%TKN (p<0.0003), Specific gravity (p<0.0001). Conclusions Fertilization at potato planting generated increase in NO3-N concentration in the water table under potato beds. A similar pattern of the NO3-N concentration under N ), fertilized treatmthe rainfall peak in summer. ents was caused by the even N rate applied at potato planting. The amount of rainfall between potato planting and N side-dress determined the increase in NO3-N concentration in the water table. Rainy periods between potato planting and N side-dressing caused increased NO3-N concentration in the water table under fertilized potato beds. Nitrate and nitrification of ammonium from ammonium nitrate side-dress fertilization could generate a potential second flux of NO3-N leaching/runoff by effect of

PAGE 61

48 Summer cover crops with high demand of N could act as N catch crops diminishingthe possibility of NO3leaching in summer. On the contrary, legume crops due to their an g.h aBush green bean as fall crop increasO3-N at potato planting but the ount nfal harvest andlanting in termined how ch NIn 2001 and 2003, potato N uptake increased linearly even after potato reached its ximurket Potato yield ufour rotatnded quadratically N rate maketable yielde recommate for the region. tationludin in summer produced higher maield than rotations low dependence on soil N could not be efficient N catch crops especially when more th 168 k a-1 of N re applied to potato. ed soil N am of rai l between its potato p spring de mu O3-N persisted in the soil. ma m ma able yield. nder the ions respo to with ximum mar under th ended N r Ro s inc g sorghu m rketable y including cowpea.

PAGE 62

CHAPTER 3 POTATO ROOT DISTRIBUTION STUDY ns 26) and the effect of fertilizer placement and soil compuptake of nitrate and water than those nearer the soil surface (Asfary et al. 1982). In another study, two potato cultivars differing in their N acquisition rate from the soil demonstrated the importance of root length and surface area to estimate N uptake over root dry weight. This study suggested that the larger root system of one variety enabled it to absorb more nitrate when it was limited, which allowed it to produce its optimum yield at lower soil N availability compared with a second variety (Sattelmacher et al. 1990). However; these approaches may underestimate the influx rate because not all the root length and root surface area have equal nutrient uptake capacity. The region immediately behind the meristem has greater uptake activity than older segments, but these segments maintain some uptake capacity (Gao et al. 1998; Robinson et al. 1991). The improvements in technology have now enabled roots to be analyzed in detail. Root characterization studies have great potential to elucidate the role of root structure in N management. Introduction The potato root system began to be studied many years ago. Pictorial descriptioof the potato root system (Weaver, 19 action (De Roo and Waggoner, 1961) have been reported. Research reports on the relationship between the potato plant root characteristic water use and uptake of nutrients are scarce. Total root length of potato in the field has been used to calculate inflow rates of N, P, K and water into the plant. These studies demonstrate that roots below 30 cm were more active in 49

PAGE 63

50 Enhanced N use and uptake are likely results of proposed strategies to minimize nitrate leaching, including diminished fertilizer rates. In the short term, this could be accomplished by timely and precise placemil regions where it could be quickly and erm, the development of potato cultivars with high uptake anies by plant breeding would be part of they e elease ent of the fertilizer in so fficiently absorbed by the roots. In the long te d use efficienc solution to nitrate leaching from potato fields (Errebhi et al. 1998). In both cases a detailed knowledge of the root system is required. The precise placement of fertilizer in time and space is especially important for nutrients such as nitrate that have high mobilitin the soil. Root distribution studies are important to enhance the understanding of nutrient and water uptake in order to improve irrigation and nutrient management programs (Asfary et al. 1983). Nitrate leaching from potato crops has been considered one of the possible causes of water eutrophication in the St. Johns River watershed, therefore improved knowledgabout the potato root system under northeast Florida conditions is needed to enhance N uptake. In order to begin the collection of information about the potato root system, astudy of potato root distribution was initiated. The information obtained about root distribution could be used to develop fertilization regimes based on controlled rfertilizers in order to synchronize N release with plant demand. Knowledge of the root system distribution could be used as well to model the N cycle in the potato production system. Based in the above considerations the following objectives were stated: i) to design and test a new methodology to obtain a representative sample from the potato root systemin order to study vertical and horizontal root distribution. ii) to study the effect of

PAGE 64

51 suffic site with pling Methodology s) x 50.8 cm height (Fig 3-1a). Three wooden blocks keep the sheets in place with an even ient fertilization rates on the root distribution of the Atlantic potato variety used extensively by the farmers in the TCAA located in the St. Johns River watershed. iii)generate information that could be used as part of strategies to enhance fertilizer placement in order to maximize N uptake and minimize leaching and/or runoff of nitrate from the potato beds. Materials and Methods Site Description This study was conducted at the Hastings Research and Education Centers Yelvington Farm in Hastings, Florida. Soil and other details about the experimentalwere described in chapter 2 of this dissertation. Statistical Design The plots containing the potato-sorghum-fallow rotation under the three highest N rates (168, 224, and 280 kgha-1 in the crop rotation experiment (Fig 2-1) were used to study potato root distribution at full flowering (60 DAP) in the third year. At this time, Atlantic reaches its maximum vegetative growth and rooting depth (Stalham and Allen2001). The plots used in this study were set as a randomized complete block design three replications and the treatments were arranged under a split-split plot design wherenitrogen rate was the main-plot (Fig. 2-1), slice was the split-plot, and position into the slice was the split-split plot (Fig. 3-2). Root Sam A device to sample a representative portion of the potato root system was designed based on a sampler used by Vos and Groenwold (1986). The device, a slicer, was composed of two sharp-edged metal sheets of 101.6 cm long (distance between row

PAGE 65

52 separation of 10.2 cm (half the distance between plants) and three C clamps firmly hthe pieces together (Fig 3-1a). eld sub-samples (environments) arranged in three Each environment was coded by the first letter of the layer where it cam Figure 3-1. Procedure to take the root samples. (a) Root sampler; (b) leveled and buried into the ridge; (c) slice of ridge; (d) splitting sub-samples. The slicer was driven into the soil (Fig 3-1b), keeping it horizontal (Fig. 3-1c). Subsequently, part of the ridge was removed and the slicer containing the slice of soil was flipped on the ground and the upper sheet removed to access the ridge slice (Fig. 3-1d) Afterward, each slice was split in 14 A B D C layers (top, middle, and bottom) e from followed by a number to identify the position in the layer (Fig. 3-2). Soil samples containing roots where weighed individually. Two slices, one

PAGE 66

53 including a plant and another one between plants, were taken in each plot in order to get a BMP format as is required by G representative sample of the potato root system under each N rate. remimmeon transSRoot (Guddanti and Chambers, 1993) specialized root length and surface area. (<0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, 0.8-1.0, 1.0-1.2, >1.2 mm) to get root length and root and total root surface area in cm2 (TRSA) were calculated and reported as the total sum of Root Washing The roots were separated from the soil by washing them on a screen while oving organic matter debris. The soil samples containing roots were stored in a cooler at 7 C during the washing process. In the laboratory, roots were spread on white trays to enhance contrast and sorted from thin residues of organic matter. Afterward, roots were refrigerated in plastic bags containing a solution of sodium azide (0.01%) as preservative. Root Scanning and Image Analysis In order to enhance the image contrast and prior to scanning, roots were stained by rsion in a solution of methylene blue (1 g L-1) for at least 24 hr. Roots were spread lucent Plexiglas trays containing water and scanned on a HP Scanner at resolution of 300 dpi. Images acquired by this method were loaded and saved in Adobe PhotoDeluxe in format PDD. Root images stored as PDD were converted to grayscale software for root analysis. GSRoot was set up to classify roots by diameters, and estimate Root Variables After the scanning process, the root samples were oven dried at 105 C and weighed to estimate root dry weight (RDW) on an analytical scale Sartorius Model 1712 MP8 with readability of 0.00001 g. GSRoot was set up to measure seven diameter classes surface area for the individual root diameter classes. Total root length (TRL) in cm

PAGE 67

54 length and surface area of the individual root diameter classes and expressed as meters of root per square meter of soil (m root m-2 soil) and square centimet ers of root surface per oil (cm2 root surface m-2 soil) respectively. Using the calculated soil volumressed RL) expressed as cm root mg-1 root dry matter and sp SRL and nd 2). mpling, a study of the soil strength in the plots was done. A ter with a 30 cone-shaped tip was pushed into the soil 60 cm deep from the to where ity square meter of s e (SVOL) for each sample (see next section) root length density (RLD) expas cm root cm-3 soil and root surface area density (RSAD) expressed as mm2 root surfacecm-3 soil were calculated. RLD and RSAD intended to describe the exploratory capacity of the root system in terms of length and root surface by unit of soil volume, respectively. High values of RLD and RSAD mean high capacity of the root system to explore the soil. Specific root length (S ecific root surface area (SRSA) expressed as cm2 root surface mg-1 root dry matter were calculated in order to estimate the resources invested by the plant in terms oflength and root surface area by unit of root dry matter. Therefore, high values of SRSA indicate high efficiency of the root system in the assignment of resources. Total and marketable tuber yield of the plots where the root study was done were measured aanalyzed in conjunction with all of the plots in the crop rotation experiment (chapterSoil Strength, Bulk Density, and NO3-N concentration by Depth Two days before root sa Delmi penetrome p of the ridge in each sampled plot. The measure was taken in the same ridgethe root sampling was performed. The penetrometer provided a continuous plot (on a standard 7.6 by 12.7 cm card) of soil strength in megapascals (MPa). Using the method of undisturbed soil cores, soil samples were collected and used to estimate soil bulk densand soil moisture content in each one of the three layers. The soil volume (SVOL) of each one of the 14 sub-samples was calculated using their estimated soil dry weight and

PAGE 68

55 the estimated soil bulk density of the layer from which each one of the 14 sub-samples came from. The calculated values were plugged into the following relationship: V = W D where: V= Soil volume of each sub-sample, W= Soil Dry Weight of each sub-sample, and Bulk density of the layer (top, middle, or bottom) and applied to the data from each sub-sample. Along with the sampling to estimate bulk density, soil samples were taken to estimate NO3-N concentration in the three layers of each slice. Statistical Analysis Root sampler The soil wet-weight (SWW) of each sub-sample recorded at sampling time, the soil dry weight (SDW) estimated from the bulk density sampling, and the estimated sub-sample soil volume (SVOL) were used as response variables to evaluate the performanceof the designed root sampling device. Replication (bed) and slice were considered random effects and their contribution to the total variance of the dependent variables was D= 04). Soil s estimated using the VARCOMP procedure of SAS (SAS Institute, 20 trength, bulk density, and NO3-N concentrations by depth. Soil strength values obtained with the penetrometer and NO3-N concentration by depth were analyzed under a Split Plot Design using the GLM procedure of SAS (SAS Institute, 2004). Nitrogen level was the main plot factor and depth from 0 to 36 cm in increments of 12 cm from the top to bottom of the ridge was the subplot. Bulk density was analyzed as a RCB design for the same depths sampled with the penetrometer. The LSMEANS option with Tukey adjustment was used to separate soil strength and NO3-Nconcentration by depth. A regular Tukey test was used to separate bulk density means.

PAGE 69

56 Roots As exploratory analysis, scatter plots of the studied variables to search poss ible assocure of SAS/GRAPH (SAS Institute, our different models using the MIXED procedure and considering common evaluation process, RLD yzed using the model considering spatial correlations within and SRL was analyzed using the model considering heterogeneous variances, and Ss of relatively low compr iation were plotted using the GOPTION proced 2004). F pooled variance, heterogeneous variances for bed (replication) by N treatment, spatial correlations within slice for each slice, and spatial correlations within and between sliceswere evaluated. Furthermore, the same four models were re-evaluated using root dry weight (RDW) and soil volume (SVOL) as covariates. After the and RSAD were anal between slices, RA was analyzed with the model considering spatial correlations within and between slices. The LSMEANS option was used to perform means separation. MeanSRL and SRA were separated with the Bonferroni adjustment and SRL and SRA by the Fisher test with alpha=0.001. Results and Discussion Root Sampler Device The variance attributable to the sampling technique (slice) was ared with the variance of environments. This result could be caused by the irregulashape of the sub-samples coming from the external part of the ridge (Fig. 3-2). Soil volume was used to evaluate the root sampler performance. This effect was evaluated using the soil volume of each one of the subsamples taken in each slice. Average soil volume (cm3) and horizontal and vertical dimensions for each environment are included in Figure 3-2. Soil volume for each environment is the overall mean estimated through out all the soil slices sampled.

PAGE 70

57 1236 T21554T31635T4930T1740*2044180218672059204421902190217220752049 M1M2M3M4B1B2B3B4B5B6 0 24 201515151520cm 10cmFigure 3-2. Distribution and sample size into the slice. Soil volume (cm3), dimensions (cm), T1 to B6 = Environment code for identification. Soil Strength and Bulk density Soil strength differences among blocks (beds) were observed. Nitrogen treatment had no effect on soil strength. Highly significant differences were observed among all of the three layers studied (Table 3-1). The increase of soil resistance from the top to the bottom layer coincided with a siserved bulk density of the with the top and middle layers (Table 3-1). In some root samples coming from the 12 to 24 and 24 to 36 cm layers a pronounced thickness of the 3kg-1) gnificant increase in the ob bottom layer compared Table 3-1. Soil strength, Bulk density, and Soil NO-N by depth. Soil depth (cm) Soil strength (MPa) Bulk density (gcm-3) Soil NO3-N (mg 0-12 0.024 a 1.28 a 10.22 a 12-24 0.105 b 1.45 b 14.15 a 24-36 0.203 c 1.50 b 2.71 b Soil strength (p<0.0001); Bulk density (p<0.0001); Soil NO3-N (p<0.0082). roots were observed (Fig. 3-3). A similar pattern of thickness was reported for tomato roots impeded by a column of glass beads (Waisel, 2002). Inhibition of the elongation othe main axis and enhanced for f mation of lateral roots in young barley plants grown in

PAGE 71

58 fields with high bulk density also have been reported (Scott-Russell et al. 1974). Also, accumulation of ethylene in roots under mechanical stress and anaerobic conditions have been reported as the possible cause of aerenchyma formation in the root cortex of waterlogged maize roots (Drew et al. 1979). In the case of the potato roots, a combination of these two conditions could have been responsible for the increase in diameter of roots in the middle and bottom layer (Fig 3-3). Soil NO3-N concentration from 24 to 36 cm was lower than the concentration observed at the 0 to 12 and 12 to 24 cm depths. There was no difference between the two top layers (Table 3-3). The low concentration of NO3-N in the deepest layer could be explained by increased denitrification due to the top layers as have been reported in a study with lysimd cm e tested c was one of the havioons, it couldas evidence that Atlantic has a restricted root system growth t anaerobic conditions in the bottom layer and/or upward capillary flux of water causing concentration of nitrate in the two eters ( Patel, 2001). Rooting Depth, Root Length, and Root Surface Area The deepest roots were observed in the bottom layer (24-36 cm) and never excee36 cm. This is a shallow rooting depth compared to the maximum rooting depth of 60 reported for Atlantic on a deep sandy clay loam soil under overhead irrigation in the UK (Stalham and Allen, 2001). At that time 16 varieties of varying determinacy wer and some of them reached rooting depths close to 1 meter. Atlanti variet ies with a s llower ro ot system. E en thou gh that observat n was un der different conditi be taken in Hastings. The observed total root length (TRL) for Atlantic in Hastings was 0.763 km room2 soil (Table 3-2). Although in the literature there are no data on TRL of Atlantic it is a very low value as compared with reported values as high as 24 km root m2 for the

PAGE 72

59 cultivar Norin 1 (Iwama et al. 1999), 11.4 km root m2 for Russet Burbank (Lescyznki and Tanner, 1976), and even still lower than 1.4 to 4.6 km root m2 for White Rose (Bishop and Grimes, 1978). The vertical distribution of Atlantic TRL (Table. 3-2 )that 94% of the TRL was in the two upper layers (0 to 24 cm) but 63% of the TRL was concentrated in the middle layer (12 to 24 cm) and just 6% was in the bottom layer (24 to36 cm). The region from 12 to 24 cm could be the place where most of the NO3uptaketaking place under Hastings conditions. Seventy-six percent of the TRL was comprised by roots less than 0.4 mm in diameter suggesting this fraction was a important part of the Atlantic root system (Table 3-2). shows is Figure 3-3. Detail of potato roots thickening The observed total root surface area (TRSA) for Atlantic was 838.41 cm root m with 92 % of this area in the upper 24 cm of the ridge (Table 3-2) and 64.4 % in the 22middle layer (12 -24 cm). The fraction of roots with diameter less than 0.4 mm represented 76 % of the TRL but just 45 % ofthe inconveniences when TRL is used to estimate the apparent inflow rates of nutrients. the TRSA. This observation shows one of Therefore, approaches using TRL could underestimate the true values of nutrient uptake

PAGE 73

60 (Robinson et al. 1991). Determination of the root diameter class active in nitrate uptake will be a key factor to adjust fertilizer placement in order to improve its uptake. Table 3-2. Root length and root surface area by diameter class and soil depth. Diameter class (mm) Depth (cm) <0.2 0.2-0.40.4-0.60.6-0.80.8-1.01.0-1.2 >1.2 Total 0-12 39.4% 39.8% 10.5% 5.7% 2.6% 1.1% 1.0% 31.0% 93.38* 94.33 24.89 13.51 6.16 2.61 2.37 237.01 12-24 29.7% 45.7% 14.4% 5.3% 2.5% 1.2% 1.0% 63.0% 142.93 219.92 69.30 25.51 12.03 5.77 5.29 481.23 24-36 22.9% 45.4% 17.6% 7.7% 3.4% 1.5% 1.5% 6.0% 10.48 20.77 8.05 3.52 1.56 0.69 0.69 45.75 Root Length* m root m2 soilTotal 32.3% 43.9% 13.4% 5.6% 2.6% 1.2% 1.1% 100% 246.65 335.24 102.14 42.77 20.04 8.93 8.23 763.99 m) Diameter class (m Depth (cm) <0.2 0.2-0.40.4-0.60.6-0.80.8-1.01.0-1.2 >1.2 Total 0-12 25.82* 8.7 % 64.32 27.4% 50.0 21.3% 42.72 18.2% 24.41 10.4% 11.74 5.0% 15.73 6.7% 234.73 28.0% 12-24 43.74 8.1% 189.00 35.0% 130.14 24.1% 77.22 14.3% 45.36 8.4% 25.38 4.7% 29.70 5.5% 540.00 64.4% 24-36 3.88 6.1% 21.65 34.0% 16.62 26.1% 10.00 15.7% 5.54 8.7% 2.67 4.2% 3.44 5.4% 63.68 7.6% Root Surface Area* cm2 root m2 soil Total 73.14 8.7 % 274.99 32.8% 196.62 23.5% 130.03 15.5% 75.37 9.0% 39.41 4.7% 48.86 1.0% 838.41 100.0% Root Length Density No significant differences in RLD were observed under the three N rates applied to potato. Significant differences in RLD among environments (p<0.0001) and between slice contaiing the plant had higher RLD than t slices (p<0.05) were observed (Fig. 3-4). The n he slice between plants (Fig. 3-4). RLD values in the environments M2 and M3 were higher than all of the values observed in other environments except from the valuesof T2 and T3 where the difference was not significant. M2 and M3 were the environments where the seed piece was located making them the center of origin of the potato root system (Fig. 3-4). There were no differences among the environments covering M2 and M3. The environments in the bottom layer (B1 to B6) had significantly

PAGE 74

61 lower values than the environments in the middle layer (M1 to M4), suggesting that soil exploration by the potato root system is evidently reduced at more that 24 cm dep th (Fig. 3-4). T2 T3 T4 T1 M1 0.467bc M2 0.838a M3 M 4edee B10.019 B20.070 B30.038 B4 B5 B6 T20.65614 abd T30.58 5T8bcM1M2M30.812aM40.497bc119830eP387a4b oh disn t soil). P= Slice containing the anicnT=nenonithowlenoticffenvent(p<0.0001); Slices (<0.025). Root Surface Area Density No significant differences on RSAD under N rates were observed. This variable was highly correlated with RLD (0.95, p<0.0001). Differently from RLD, a significant interaction environment by slice was observed (Fig. 3-5). RSAD had higher values for environments M2 and M3 in the slice containing the plant but the difference was not significant with T2. The difference between M3 and T3 was significant. The differences in RSAD in T2 and T3 as compared with M2 and M3 could be the result of the position of the seed piece in the two central M environments and the position of the stem in the two central T environments. A lot of adventitious roots were originated from the stem ab T40.304 cde 0.3 0.467bc0.838a B 0.0e B2 0.070 de B3 0.03 e B4 0.007 e B5 0.0 e B6 0.053 d W 0. BP0.29 Fig ure 3-4. Ro t lengt ensity d tributio (cm roo cm -3 plw t; B= Sl e betwee plants. 1 to B6 Enviro ment id tificati Means same l ercase tter are t statis ally di rent. E ironm s

PAGE 75

62 when part of it was soil covered at the time of hilling and N side-dressing. That means that in terms of root surface the potato plant could have its maximum uptake capacity in T29.70abcT37.63abT41.85fghT13.53efghM15.08cdedfghM210.39abM313.06aM45.61bcdefg B10.27hB21.68ghB31.05ghB40.14hB50.59ghB60.66gh T24.34defghT33.82efghT43.06efghT12.53efghM14.22efghM29.60abcdM37.04bcdefM44.08efghB10.15hB20.37ghB30.25hB40.07hB50.24hB60.50ghBP density distribution (mm2 root cm-3 soil). WP= Slice containing the plant; BP= Slice between plants. T1 to B6 = Environment t (p<0.0059). Possible comparisons between slices. the central two M environments. Along with the greater RSAD in the two central M environments, higher soil moisture and higher NO3 concentration were observed in the middle layer than in the top layer (Fig. 3-5 ). In chapter 2 was shown that the residual effect of the side-dressed fertilizer applied 40 DAP remained in the soil even after potato harvest. The dryness of the top layer could be limiting NO3 uptake by the potato roots. Even though the difference among environments in different slice was not significant except when the environments were compared with M2 and M3 in both slices it is WP Figure 3-5. Root surface area identification. Means with same lowercase letter are not statistically differen-

PAGE 76

63 possible suggest that roots in the region between plants could have less capacity to scavenge NO3from the soil. In the slice between plants all of the environments around M2 and M3 were not different among them, even though high differences among the RDA values for environments in the bottom layer compared with the two environments at both sides of the middle layer and all of the environments in the top layer were observed but not statistically detected by the model (Fig. 3-5). Specific Root Length There were no significant differences by slice and N treatment in SRL. Highly significant differences were detected among environments. No significant differences were observed for the interactions among factors. The highest values for SRL were observed in the environments T1 and T4 without significant difference between them (Fig. 3-6). The high values of SRL in the environments T1 and T4 are indication of very fine roots in this part of the ridge. The stem coverage with loose soil at N side-dress and hilling time (40 DAP) could be the explanation for the development of a system of fine lues observed inater d n environments T2 and T4 are the places where the potato plant has its most fine fraction of adventitious roots in these environments (Fig. 3-6). On the contrary, the low SRL va B3 and B4 (Fig. 36) could be the result of the increase in diameter caused by the combination of the high bulk density in B3 and B4 environments and by the wrising by capillary effect from the water table. Saturated soil conditions could increase root diameter by stimulation of aerenchyma formation (Drew et al. 1979), and saturateconditions could be motivated by a high capillary flux from the water table as reported in presence of a plow pan (van Loon and Bouma, 1978). The higher SRL values observed ithe environments beside B3 and B4 suggest that the compacted zone could be localizedjust in the two central environments of the bottom layer. In terms of SRL the

PAGE 77

64 roots and probably where high nutrient uptake could take place but it also could be limited by the fluctuant soil moisture on the top of the ridge. If RLD values (Fig. 3-4) for T1 and T4 are compared with the values for M2 and M3 (Fig. 3-4 ) it is possible to nothat the two environments on the top of the ridge have significantly lower values than M2 and M3 in the middle of the ridge. It is explained by the high concentration of roots in thecentral environments (M2 and M3) where the seed pie tice ce was placed. On the other hand, T2 14.23bcT313.34bcdT419.56abT121.98a15.59bc14.12bc17.01abc15.59bcB1cdB2abcB3deB4eB5cdB6bcFigure 3-6. Specific root length distribution (cm root mg -1 root dry weight). T1 to B6 = Environment identification. Means with same lowercase letter are not statistically different. (p<0.0001). the significant lower SRL values observed in M2 and M3 compared with T1 and T4 are indication of the presence of roots with high diameter close to the root piece. Generallyfinest roots are more efficient in nutrient uptake than coarse roots, therefore would be possible suggest that roots in T1 and T4 could be more active in NO3 uptake when soil moisture is not limiting it. Specific Root Surface Area environment by slice was significant. The scarce difference observed among SRSA M1M2M3M413.2917.127.215.7312.0616.51 -SRSA was analyzed using RDW as a covariate (p <0.0004). The interaction

PAGE 78

65 suggests that the potato root system has a very homogeneous configuration in terms of.the amount of biomass allocated by unit of root surface area (Fig. 3-7). Significant low SRSA values were observed in B3 and B4 environments located in the slice between plants, however, these values were not statistically different from many higher values in the same slice and in the slice including the potato plant (Fig. 3-7). The lack of sensitivity of the model to detect high differences among SRSA could be attributed to the high T21.791bcT31.444bcdT41.291cdT11.542bcd M11.782bcM21.816bcM33.698aM41.849 bc B11.392cdB21.916bcB32.036bcB41.225cdB52.206bcB61.517bcd T21.313cdT31.268cdT41.538bcdT11.575bcdM11.666bcM21.905bcM31.741bcM41.426bcdB11.291cdB22.700abB30.415dB40.315dB51.266cdB61.356cdWPBP relatively low resolution (300 dpi) used to capture the root on maybe affected more area than length estimation. Based on theents e Figure 3-7. Specific root surface area distribution (cm 2 root mg -1 root dry weight). WP= Slice containing the plant; BP= Slice between plants. T1 to B6 = Environment identification. Means with same lowercase letter are not statistically different (p<0.0395). Comparisons possible between slices. variability produced by the images; therefore, low resoluti results of SRL where the lowest values were observed in the same environm(B3 and B4), it would be possible suggest that the low values of SRSA observed in th

PAGE 79

66 corresponding environments in the slice between plants could be really different fromhigher values in both of the slices. If this idea is correct the theory of increased thickness due to soil impedance and/or flooded conditions could be supported by these results. Conclusions This study was developed to propose a new methodology to study the potato root system under Hastings conditions and to describe spatially the potato root distribution as affected by N rates currently used in commercial potato production. The proposed sampling methodology performed satisfactorily and is suitable for future root studies in the TCAA or agricultural areas with similar conditions. The ridge slicing methodology allowed to recover most of the roots from a representative sample of the potato root system and spatially classify sub-samples around the potato plant. The potato root distribution was not affected by N rate. RLD, SRL, RSAD, and SRSA were affected by their spatial position in the soil profile. Soil exploration capacity (RLD and RSAD) was maximized into 15 cm around the potato plant at a depth of 24 cm from the top of the ridge. Significant lower soil exploration capacity was detected under 24 cm from the top of the ridge. The invested biomass by unit of root length or unit of surface area (SRL and SRSA) were relatively homogeneous except in the environments thickening ons. High bulk iting t under the potato plant (B3 and B4) where low values were probably caused by root caused by a combination of increased soil strength and saturated conditidensity and water saturation at the bottom layer of the ridge could be a lim factor for potato root development. Observed differences in root parameters could be used as base to plan fertilizer placement strategies in order to enhance nutrient uptake. To optimize fertilizer placemena detailed study of the water movement in the root zone should be done.

PAGE 80

CHAPTER 4 USIONS AND F SUMMARY, OVERALL CONCLUTURE RESEARCH effect of nutrient enriched water bodies is exchange capacity. The TCAA in northeasent. Inclusion of legumes as green manure or as a second cash crop in rotation with potatoes Some years ago nutrient loses from agricultural production systems were considered as a purely agronomic problem limited productivity and negatively affected the economic profitability of agricultural producers. Currently in Florida, the negative turning on a great concern in endangered zones where nutrient mobility is expedited by the low capacity of the soil to retain them in exchangeable form and for the sandy nature of the soils. This is the case of most of the Florida soils where the low content of mineral colloids results in a pH dependent t Florida is a region where potatoes and cabbage are the dominant crops. Due to the relatively high N demand by potato, farmers apply more N fertilizer than is needed for optimum production generating nitrate non-source pollution from the potato fields. The over-fertilization problem could be enhanced by the current irrigation system used for agricultural production. It consist in a perched water table that is controlled by pumping water from deep wells into irrigation-drainage furrows running parallels to the raised potato beds. The concern by the potential environmental impact of agricultural practices have generated a growing interest to develop Best Management Practices (BMPs) that allow sustainable crop productivity without negatively affecting the environm has been considered a viable way to supply part of the relatively high N requirement of soluble fertilizers applied to potatoes. 67

PAGE 81

68 Synchronizing N availability in the soil and plant uptake is another proposed approach that require good knowledge of the root system in order to enhance fertilizer uptake by an optimum placement in the region of the root zone where uptake will be maximized. Basequed plant Concentrations of NO-N in the water table under potato In the first year NO-N concentration in the water table was increased up to 54.6 mgL under effect of fertilization at planting. Predicted values for non-fertilized plots were in the range of 8.2 to 18.2 mgL-1. Decreasing patterns of the regressions describing NO3-N concentration in the water table were observed from potato planting to harvest. In the second year predicted values for NO-N concentration were below 10 mgL and a light increase were predicted close to the harvest time. There was no noticeable difference among fertilized and unfertilized plots. The low rainfall after potato planting could be addressed as the cause of the low NO-N in the water table. In the third year a similar pattern to the first year was observed with the difference that unfertilized plots exhibited lower NO-N concentration than the first year. Side-dressed nitrogen at 40 DAP did not affect the decreasing pattern of NO3-N concentration sed on the previous considerartation included in the Chapter 2 the tions this disse ntial study of NO3-N concentration in the water table, soil solution, soil, antissue as affected by four different crop rotations and five N rates over a three-year period. Chapter 3 reports the study of the potato root system in order to describe quantitatively and spatially its distribution in the soil in terms of root length and root surface area under three rates of nitrogen. Each one of these chapters had specific objectives and their results are summarized below. Summary Crop Rotation and N rates 33-13-133

PAGE 82

69 in the water table in the first an d third years suggesting fertilization at potato planting athe main source of the increased NO3-N concentration in the water table. Concentrations of NO3-N in the soil solution s r cowpea and sorghum Concentration of NO3-N in the water table under cowpea plots under effect of ed with 280 kg.ha-1 was significantly higher than s no difference in NO3-N conceConchum and cowpea in summotato It was possible to study this variable just in the second year due to insufficient information from the first year and problems with the installation of the lysimeters in thethird year. Even though 2002 was a year with low rainfall after potato planting NO3-N concentration values as high as 156.0 mgL-1 were observed after fertilization. Theunaffected decreasing pattern after side-dress fertilization suggests low effect of it on NO3-N concentration in the water table. Concentrations of NO3-N in the water table unde residual N from potato fertiliz concentration under plots fertilized with 168 kg.ha-1. There wa ntration in the water table under cowpea planted in unfertilized plots and under cowpea planted in plots fertilized with 168 kg.ha-1 of N. NO3-N concentration in the water table under sorghum was significantly lower than under cowpea in plots fertilized with 280 kg.ha-1 except in the second sampling when high variability was observed. entrations of NO3-N and NH4-N in the soil Soil NO3-N concentration was increased after side-dressing N and exhibited a rising pattern across the N rates applied. The same pattern persisted in the soil for 3 months and was detected before potato harvest and at planting of sorg er. The concentration of NO3-N at planting of the cover crops showed variation among years. This variation could be attributable to the amount of rainfall between pharvest and sorghum and cowpea planting. Soil concentration of NH4-N across the N

PAGE 83

70 rates was described by a quadratic pattern 3 weeks fertilizer side-dressing and there was no difference between unfertilized plots and plots with only fertilization at planting oincident high concentration of NO3-N and lotion ed plots. Nitrogen accumulation in sorghum ation after N sidulation ts e tato is fertilized with more than 168 kg.hatic r suggesting that nitrification occurred. In 2003 c w concentration of NH4-N 3 weeks before potato harvest suggest a late nitrificaof the side-dressed fertilizer. Increased soil NO3-N concentration in plots planted with green bean in fall as compared with plots left as fallow was noticed. Nitrogen recovery by the summer cover crops Residual N from the fertilizer applied to potato produced increased nitrogen recovery by the sorghum from the fertiliz biomass was described by similar regressions describing soil NO3-N concentr e-dress, at potato harvest, and at summer cover crops planting. Cowpea N accumulation was not affected by the N rate applied to potato. Cowpea N accumin biomass in fertilized plots was not different from nitrogen accumulation in unfertilizedplots. This event supports the observation that the residual fertilizer from potato side-dressed fertilization could be generating a second flux of nitrate leaching from the ploplanted with cowpea in summer. These results suggest sorghum as an effective N catch crop that should be planted as close as possible to the potato harvest. Low soil N uptak by cowpea could generate leaching-runoff when po -1 of N. Potato N uptake at full flowering Nitrogen rate affected significantly potato N uptake over the 3-year period. Nitrogen uptake was described by linear regressions in 2001 and 2003, and by a quadraregression in 2002. The continuous increment in N uptake even after potato reach its maximum yield suggest luxury consumption of N when higher N rates than needed fo

PAGE 84

71 maximum marketable yield were applied. In 2002, the overall potato N uptake showed closer to N rate needed for maximum yield. Potato yield Quadratic increase by effect of N rate was observed in each year. In 2001, there was a relatively similar response of the four rotations to N rate. This year was considerea transition year where the effect of crop rotation with legumes was n d ot evident. In 2002 elded more than rotations with crop e kg -s observed explaining the relatively unaffected exportation of N by the Potatm and 2003 rotations including sorghum as summer cover crop yi owpea. The difference in the N rate required for maximum yield by the four crotations was increased from 2002 to 2003. A period with heavy rainfall at the end of 2002 maybe propitiated leaching of the mineralized N from the cover crop residues generating different potato N requirement for maximum yield in 2003. Exported N as total yield of tubers in fertilized plots remained statistically the samwithin year throughout the three-year period without statistically differences among N rates except in the third year when the treatment with just fertilization at planting (112ha-1 of N) exported significantly less N than plots with side-dressed fertilizer. These results suggest that fertilization at planting could be enough to supply the N needed to reach optimum yield under Hastings conditions. A generalized increase of %TKN along with total yield decreasing during the threeyear period wa decrease in yield. o Root Distribution Study Root sampler The root sampler device was an efficient method to sample the potato root systein the ridges used for potatoes production in Hastings.

PAGE 85

72 Soil strength, bulk density, and NO3-N by depth Significant increase in the soil strength from the top to the bottom layer of the ridge Bulk density followed the same trend but there were no differences betwee alue as compared with t d by nitrogen rate. Differences wereoues were observed in the bottom layer. RLD in the slice was e slice containing a plant. Specific Root Length (SRL) by N rate. There were no differences between slices. The highest values for SR erage with soil at side-dress was observed en the middle (12-24 cm) and the bottom (24-36 cm) layers. Soil NO3-N concentration by layer was significantly higher in the top and middle layers as compared with the bottom layer. This observation coincide with the observed persistence of the side-dressed fertilizer in the soil in the top of the ridge. Rooting depth, root length, and root surface area The deepest roots were observed between 24 and 36 cm. This is a very shallow rooting depth as compared with 60 cm reported for Atlantic in well drained soils. Thtotal root length observed was 0.763 km root m2 soil This is also a low v he reported values for some potato varieties. Ninety four percent of the total root length was in the 2 top layers. The total root surface area was 838.41 cm2m2 soil and 92%of the root surface area was in the 2 top layers. There were no reported values of potato root surface area to compare. Root length density and specific root length Root Length Density (RLD) was not affecte bserved among different environments and between slices. Significant higher values of RLD were observed in the central environments of the middle and top layers. Very low RLD val significantly lesser than RLD in th was not affected L were observed in the external environments of the top layer maybe caused by thefine adventitious root system generated by the stem cov

PAGE 86

73 fertilization. In order to predict placement of fertilizer to enhance nitrogen uptake values of RLas p layers of the slice containing ecause in those environments was locateOverall Conclusions increase in NO3-N concentration the water table aluated. ven dent in trate D should be compared with SRL in order to select the environments with coincident highest values for these two variables. Root surface area density and specific root surface area RSAD and SRSA were not affected by, nitrogen rate and were highly correlated with RLD and SRL. Differently from RLD significant interaction environment by slice wobserved. RSAD in the central environments in the two to the plant exhibited the highest RSAD values maybe b d the seed piece and the stems generated from it. SRSA was relatively homogeneous among environments and between slices suggesting that potato has a homogeneous configuration in terms of the amount of biomass allocated by unit of root surface. Significant low SRSA values were observed in the central environments of the bottom layer in the slice between plants. This is the result of the thickening observed in some of the root samples recovered from those environments. Fertilization at potato planting generated with a steady decreasing pattern through the time under the four rotations evThe little difference among fertilized plots (2-3a to 2-5b) was maybe caused by the eN rate applied at planting to fertilized plots (Table 2-1). The influence of the rainfall on the magnitude of the increase in NO3-N concentration in the water table was evi2002 when an atypical low rainfall period between potato planting and N side-dress (Table 2-2) caused lower NO3-N concentration in the water table before N side-dressing (Fig. 2-4a and 2-4b) compared with the same period in 2001 and 2003. Fertilization at planting, when potato has not established its root system could increase the risk of ni

PAGE 87

74 leaching/runoff. Therefore, fertilization at crop emergence could be more effective to reduce the chance of N leaching/runoff by increasing the N uptake. Fertilizer plashould be considered also in order to enhance N up cement take. Based in the description of the 3 as the optimal envirot in these t ant movement of nitrates due to water evaporation and planiods. Therefore, low placement of theed potato root distribution is possible suggest T2, T3, M2, and M nments to locate fertilizers at planting or potato emergence. These environments comprise a zone of 15 cm around the potato plant with a depth of 24 cm from the top of the ridge. In these environments the potato seed piece was located; therefore, in this soil region nutrient uptake will take place first. The environments T2, T3, M2, and M3 comprise the root zone where coincided high RLD and high SRL showing thaenvironments was concentrated the maximum exploratory capacity of the potato roosystem (Fig. 3-4) along with high density of fine roots (Fig. 3-6). A detailed study of thewater movement in the potato root zone should be done in order to optimize nutrient placement into the root zone. Previous reports in the region (Elkashif and Locascio, 1983) and in lysimeters (Patel et al. 2001) have reported ascend t uptake during dry per fertilizers on dry periods like after potato side dressing, or placement of controlled release fertilizers under the potato seed at planting, programmed to liberate nutrients when the dry period after N side-dressing begins could be an alternative to enhance plant uptake N uptake. Controlled release fertilizers could be placed closer to the potato seed with lesser risk of seed damage than conventional soluble fertilizers. Optimized fertilizer placement at potato planting could help to minimize the neof side-dress N fertilization turning it in an optional practice when seasons with heavy rains occur. Minimal side-dressed fertilization would diminish the risk of a potential

PAGE 88

75 second flux of NO3-N leaching-runoff originated in the residues of the side-dressed application of N by effect of the rainfall peak in summer (Fig. 2-7). Summer cover crops with high demand of N planted as soon as possible after potato harvest should be planted in order to recover any residual N from the pofertilization. In this study sorghum exhibited higher capacity to recover residual N than cowpea when the N rate applied to potato was higher than 168 kg.ha-1. Potato in rotation with sorghum in summer yielded more than rotations with cowpea suggesting sorghum asa better crop to storage N to be used in the next potato season specially when heavyoccur at the end of the year. Bush green bean as fall crop increased soil N at potato planting but the amount orainfall between harvest of the fall crop and potato planting in spring (Table 2-2) determined how much N persisted in the soil until potato planting (Table 2-3). The relatively large difference in the N rate required for maximum potato yield of the rotaPSG between 2002 and 2003 (Fig. 2-12) suggests that green bean could be decreasing theoverall C/N ratio of the soil organic matter, therefore increased N mineralization and nitrification could enhance N availability for the next potato crop but also could increase the risk of nitrate leaching/runoff in years with heavy rains between green bean harveand potato planting. Finally it is possible to conclude that an improved N management of the potato production system in order to reach optimum marketable yield minimi tato rains f tion st zing at the same time t he risk of N leaching/runoff could be achieved by improved fertilization practices including optimized fertilizer timing, placement, and N sources. These practices should

PAGE 89

76 be complemented by rotation with cover crops that allow recover residual N from thepotato season and transfer it in an efficient manner to the next potato crop. and the sub-irrigation system should be studied in order to estimate its role on in order to ensure a maximum recovery of the possible residual nitrogen fertilizer. Depending on the catch crop and its C/N ratio different incorporation times in order should be tested. Future Research Based on the results of this dissertation is possible suggest some topics to be addressed by future research work. The water movement toward and from the water table as influenced by the rainfall nitrate movement. Different non-leguminous nitrogen catch crops after potato harvest should be tested to match available nitrogen from the mineralized residues with potato uptake Based on the results of this dissertation the effect of reduced nitrogen rates or controlled nitrogen sources allocated in the sites of the root zone with high scavenging capacity should be studied. Determination of the fraction of potato root diameters implied in nitrogen uptake should be investigated.

PAGE 90

APPENDIX A REGRESSION PARAMETERS NO3-N concentration in the water table (Figs. 2-3a to 2-5b) Crop N rate Conc. Yr eaRotation kg.ha-1 mg.L-1 Intercept Linear term Quadratic term R2 2001 3 PCF 0 Ln[NO-N] = 1.750575 0.077125*dss 0.79** 2001 PCF Ln[NO3-N] = 2.310352 0.111002*dss + 0.000350*dss2 0.78** 112 2 001PCF 168 Ln[NO3-N] = 2.335808 0.120305*dss + 0.000710*dss2 0.77** 2 001PCF 224 Ln[NO3-N] = 2.477959 0.118539*dss + 0.000821*dss2 0.73** 2001 PCF 3 280 Ln[NO-N] = 2.871943 0.176226*dss + 0.001600*dss2 0.77** 2 001PCG 0 Ln[NO3-N] = 1.893703 0.102081*dss + 0.000360*dss2 0.81** 2 001PCG 112 Ln[NO3-N] = 2.37892 0.105337*dss + 0.000485*dss2 0.66** 2001 PCG 168 Ln[NO3-N] = 2.36447 0.104694*dss 0.80** 2001 Ln[NO3-N] = 2.79625 0.077589*dss 0.77** PCG 224 2 001PCG 280 Ln[NO3-N] = 2.529325 0.058704*dss 0.67** 2 001PSF 0 Ln[NO3-N] = 2.546013 0.102689*dss 0.76** 2001 0.74** PSF 112 Ln[NO3-N] = 2.592968 0.025896*dss 0.000878*dss2 2001 9 0.040917*dss 0.000700*dss2 0.81** PSF 168 Ln[NO3-N] = 2.80135 2 001PSF 224 Ln[NO3-N] = 3.157429 0.087921*dss 0.72** 2 001PSF 280 Ln[NO3-N] = 3.140889 0.050690*dss 0.78** 2001 3-N] = 2.529126 0.122880*dss 0.64** PSG 0 Ln[NO 2001 PSG 112 Ln[NO3-N] = 3.077667 0.072097*dss 0.87** 2001 PSG 168 Ln[NO3-N] = 3.69064 0.108683*dss 0.79** 2001 PSG 224 Ln[NO3-N] = 3.08599 0.076067*dss 0.84** 2001 PSG 280 Ln[NO3-N] = 3.127979 0.033881*dss 0.000947*dss2 0.81** 2002 PCF 0 Ln[NO3-N] = 0.125753 0.126270*dss + 0.002298*dss2 0.57** 2002 PCF 112 Ln[NO3-N] = -0.74375 0.067456*dss 0.25ns 2002 PCF 168 Ln[NO3-N] = -0.33049 0.098137*dss + 0.001950*dss2 0.46** 2002 PCF 224 Ln[NO3-N] = -0.12305 0.079519*dss + 0.001408*dss2 0.44** 2002 PCF 280 Ln[NO3-N] = -0.13962 0.094500*dss + 0.001928*dss2 0.26ns 2002 PCG 0 Ln[NO3-N] = 0.078867 0.067357*dss 0.37* 2002 PCG 112 Ln[NO3-N] = -0.14726 0.040648*dss + 0.001036*dss2 0.24ns 2002 PCG 168 Ln[NO3-N] = 0.060474 0.088368*dss + 0.001519*dss2 0.50** 2002 PCG 224 Ln[NO3-N] = 0.16497 0.107891*dss + 0.002066*dss2 0.30* 2002 PCG 280 Ln[NO3-N] = 1.123871 0.036187*dss + 0.000097*dss2 0.26ns 2002 PSF 0 Ln[NO3-N] = -1.04924 0.031680*dss + 0.000662*dss2 0.05ns 2002 PSF 112 Ln[NO3-N] = 0.146383 0.068382*dss 0.32* 2002 PSF 168 Ln[NO3-N] = 0.394715 0.080456*dss 0.34* 2002 PSF 224 Ln[NO3-N] = 0.456274 0.092204*dss + 0.001404*dss2 0.46** 2002 PSF 280 Ln[NO3-N] = 0.13171 0.085000*dss 0.25ns 2002 PSG 0 Ln[NO3-N] = -0.42654 0.046118*dss + 0.001040*dss2 0.15ns 77

PAGE 91

78 NO3-N concentration in the water table. Continued. Crop N rate Conc. Year Rotation kg.ha-1 mg.L-1 Intercept Linear term Quadratic term R2 2002 PSG 112 Ln[NO3-N] = -0.12776 0.094905*dss + 0.001872*dss2 0.53* 2002 PSG 168 Ln[NO3-N] = 2 0.33* 0.609595 0.112335*dss + 0.001897*dss 2002 PSG 224 Ln[NO + 0.001258*dss2 0.10ns 3-N] =.055495*dss -0.42545 0 2002 PSG 280 Ln[N+ 0.000418*dss2 0.20ns O3-N] = 0.243151 0.043875*dss 2003 PCF 2 0.22ns 0 Ln[NO3-N] = 0.382195 0.067770*dss + 0.000672*dss 2+ 0.000995*dss2 0.44** 003 PCF 168 Ln[NO3-N] = 2.866326 0.129757*dss 2003 PCF 280 Ln[NO3-N] = 3.289673 0.113775*dss 0.60** 2003 PCG Ln] 2.928474 0.127533*dss + 0.001098*dss2 0.82** 0 [NO3-N = 2003 PCG L0.131558*dss 0.54** 168 n[NO3-N] = 2.873487 2003 PCG 280 Ln[NO3-N] = 4.226966 0.117990*dss 0.81** 2003 PSF 0 Ln[NO3-N] = -0.11973 0.030667*dss + 0.000130*dss2 0.18ns 2003 PSF 168 Ln[NO3-N] = 2.59872 0.093499*dss + 0.000520*dss2 0.47** 2003 PSF 280 Ln[NO3-N] = 2.771621 0.082717*dss + 0.000289*dss2 0.55** 2003 PSG 0 Ln[NO3-N] = 0.73014 0.055740*dss + 0.000460*dss2 0.37* 2003 PSG 168 Ln[NO3-N] = 4.118405 0.166594*dss + 0.001241*dss2 0.84** 2003 PSG 280 Ln[NO3-N] = 3 .925743 0.124762*dss 0.83** dss= Dsince-dre Nnificant Sican.05 Sicant 0.01 ays side ss ns = o sig = ignif t to 0 ** = ignif to

PAGE 92

79 NO3-N concentration in the soil solution (Figs. 2 -6a, b) ss= Days since side-dress s = No significant = Significant to 0.05 NO3-N concentration in soil Fig. 2-8a Year Conc. mg.kg Intercept Linear term Quadratic term R2 Crop N rate Conc. Year n -1 Rotatio kg.ha mg.L-1 Intercept Linear term Quadratic term R2 2002 PCF 0 Ln[NO3-N] = 1.4963 0.1526*dss + 0.0019*dss2 0.78** 2002 PCF 112 Ln[NO3-N] = 2.3555 0.2012*dss + 0.0028*dss2 0.72** 2002 PCF 168 Ln[NO3-N] = 2.8498 0.1464*dss + 0.0018*dss2 0.69** 2002 PCF 22 4 Ln[NO3-N] = 2.3856 0.1461*dss + 0.0021*dss2 0.47** 2002 PCF 280 Ln[NO3-N] = 2.5739 0.1652*dss + 0.0023*dss2 0.76** 2002 PCG 2 0.82** 0 Ln[NO3-N] = 2.4596 0.1420*dss + 0.0019*dss 2002 PCG 2 11 Ln[NO3-N] = 3.3709 0.1716*dss + 0.0023*dss2 0.72** 2002 PCG 2 168 Ln[NO3-N] = 2.8677 0.1660*dss + 0.0024*dss 0.64** 2002 PCG 2 224 Ln[NO3-N] = 2.6941 0.1917*dss + 0.0032*dss 0.77** 2002 PCG 0 28 Ln[NO3-N] = 3.3876 0.1467*dss + 0.0023*dss2 0.59** 2002 PSF 0 Ln[NO3-N] = 1.4101 0.1532*dss + 0.0024*dss2 0.76** 2002 PSF 112 Ln[NO3-N] = 2.1226 0.1455*dss + 0.0018*dss2 0.89** 2002 PSF 16 8 Ln[NO3-N] = 2.6749 0.1821*dss + 0.0025*dss2 0.87** 2002 PSF 224 Ln[NO3-N] = 2.2218 0.1242dss + 0.0015*dss2 0.70** 2002 PSF 280 Ln[NO3-N] = 3.1244 0.1497*dss + dss2 0.81** 0.0020* 2002 PSG 0 Ln [NO3-N] = 2.0610 0.1320*dss + 0.0017* dss2 0.79** 2002 PSG 112 Ln[NO3-N] = 2.8405 0.1818*dss + 0.0024* dss2 0.63** 2002 PSG 168 Ln[NO3-N] = 3.4428 0.1793*dss + 0.0026* dss2 0.58** 2002 PSG 224 Ln[NO3-N] = 2.9767 0.1836*dss + 0.0027* dss2 0.55** 2 002 PSG 280 Ln[NO3-N] = 2.9657 0.1306*dss + 0.0019*dss2 0.49** d n *** = Significant to 0.01 2001 [NO3-N] = 10.8547 + 0.1953 Nrate 0.53** 2003 [NO3-N] = 3.2588 + 0.0049 Nrate + 0.000042 Nrate2 0.57** Fig. 2-8b Year Conc. mg.kg Intercept Linear term Quadratic term R2 2001 [NO3-N] = 1.7392 + 0.0077 Nrate 0.0077 Nrate2 0.51** 2002 [NO3-N] = 2.0927 + 0.0263 Nrate 0.000023 Nrate2 0.27** 2003 [NO3-N] = 5.2414 + 0.0691 Nrate 0.17**

PAGE 93

80 -N concent in s 2-Yenc. .kg erc rQuatR NH4 ration oil Fig. 9a ar Co mg Int ept Lin ea term dra ic term 2 2 00 [NO3-N] .14 0.06 0.0 0.6 1 = 5 98 71 Nrate + 006 Nrate2 4** 2 00 [NO3-N] .16 0.05 0.070.5 3 = 2 59 58 Nrate + 003 Nrate2 5** Fig. 2-Yenc. .kg erc nrQuatR 9b ar Co mg Int ept Li ea term dra ic term 2 2 0 [NO3-N] .03 0.00 0.0 01 = 5 31 52 Nrate 8* 2 00 [NO3-N] .61 + 0.0049 0.2 2 = 0 43 Nrate 5** 2 00 [NO3-N] .08 + 0.01660.1 3 = 1 74 Nrate 3* Nrat e =ertilizn rat) =signit =ifica 0.05 =ificano 0.01s taktake -1Intercept Linear term Quadratic term R2 N f atio e (kg.ha-1 ns No fican Sign nt to ** Sign t t Crop N up e Sorghum. Fig. 2-10 Year N up Kg.ha 2001 TKN = 10.8858 + 0.3293 Nrate 0.36** 2002 TKN = 17.9582 0.0659 Nrate + 0.00060 Nrate2 0..36** 2003 TKN = 8.3331 0.0701 Nrate + 0.00054 Nrate2 0.42** P otato. Fig. 2-11 Intercept Linear term Quadratic term R2 Year N uptake Kg.ha-1 2001 TKN = 111.3804 0.3714 Nrate 0.23** 2002 TKN = 37.2222 + 0.8693 Nrate 0.0018 Nrate2 0.45** 2003 TKN = 28.0746 + 0.4817 0.51** Nrate Nrat fertilizate (kg.hTKNtal Kns o sig = Significant to 0.05 nificant to 0.01 e = N tion ra a-1) = To jeldahl Nitrogen = N nificant ** = Sig

PAGE 94

81 P redicted N rate for maximum potato yield Crop NYmax Fig. 2-12 Year Rotatio n (kg.ha-1) Intercept Linear term Quadratic term R2 2001 Pooled NYm= 20.1574 N+ rate 0.00050 Nrate2 ax rate 0.1640 N 0.38** 2002 PCF NY= 8.4584 N+ rate Nrate2 max rate 0.1455 N 0.00035 0.81** 2002 PCG 1 Nrate2 ** NYmax = 12.7438 Nrate + 0.0884 Nrate 0.0002 0.48 2002 PSF Nrate2 ** NYmax = 7.4520 Nrate + 0.1968 Nrate 0.00047 0.72 2002 PSG NYmax = 8.9293 Nrate + 0.1701 Nrate 0.00043 Nrate2 0.91** 2003 PCF NYmax = 2.9133 Nrate + 0.1385 Nrate 0.00032 Nrate2 0.79** 2003 PCG NYm= 3.9586 Nrate + 0.1506 Nrate 0.00040 Nrate2 0.92** ax 2003 PSF NY= te + rate Nrate2 .84** max 3.9419 Nra 0.1680 N 0.00041 0 2003 PSG 0033 Nrate2 ** NYmax = 4.2412 Nrate + 0.1553 Nrate 0.0 0.86 NY PredNrate = N fer max = icted N rate for maximum yield (kg.ha) -1 tilization rate (kg.ha-1) ns = No significant = Significant to 0.05 ** = Significant to 0.01

PAGE 95

APPE NDXANOVA TBLES FOR POTARAEIn this appendix are reported ANOVA tablesie vgy d. Tblecluoeslasy e,yil y and c gity sizf tey I B A TO YIELD PA M TERS for potato y ld ariables durin the stud perio hese ta s in de rot potat c sification b siz marketable eld, tota ield, specifi rav Due to the e o he tables, th are included beginning in the next page. 82

PAGE 96

83 potato yield parameters Main plot effects tested using main plot error *Significant at Type I error < 0.05 **Significant at Type I error < 0.01 Type III Mean Square 2 Source of Variation DF Rots Damaged B A1 A2 A3 A4Specific gravity Market Yield Total Yield Replication 1 3 15.68** 1.03 5.45* 63.59 609.01* 346.05 0.21 8*2*000030* 249.32*216.80 *0. Rotation 1 3 4.57* 13.60 3.94 42.14 29.65 246.05 0.59 2132000008 .48 31.33 0. Main Plot Error 9 1.11 3.85 2.24 54.33 132.33 109.92 0.61 275259.53 000006 .72 0. N rate 4 5.85** 34.70** 1.04 192.09** 963.69 580.91 1.02 366*9*000174** 2.42*381.34 *0. Rotation N rate 12 0.58 2.69 2.42 36.95 168.39 37.32 0.61 301275.77 008 .46 0. 000 Error 79 1.40 2.75 1.44 32.25 75.78 46.89 0.47 172 900462 .4817.87 000 Table B-1. 2001 ANOVA table for 1 2 0.

PAGE 97

Table B-2. 2002 ANOVA table for potato yield parameters 841 Main plot effects tested using main plot error Type III Mean Square 2 Source of DF Rots Damaged B A1 A4 Market Total Specific gravity Variation A2 A3 Yield Yield Replication 1 2 12 0 13.49 1.63 5.97* 8.51 425.25 .78 .030 604.14 657.55 0.000005 Rotation 1 3 16.31 1.64 4.49* 25.14 231.25 10.46 0.039 211.00 126.12 0.000027 Main Plot Error 6 6.17 5.39 2.80 55.42 313.14 6.71 0.067 191.18 227.68 0.000006 N rate 4 10.08* 12.04** 7.39** 765.81** 980.66** 7.30* 0.043 3676.50**4266.38**0.000234** Rotation N rate 12 5.52 0.74 1.10 49.25** 36.28 1.32 0.063 126.23** 152.94** 0.000004 Error 59 6.96 0.79 1.28 15.31 33.08 2.09 0.06 37.67 29.76 0.000004 2 *Significant at Type I error < 0.05 **Significant at Type I error < 0.01

PAGE 98

Table B-3. 2003 ANOVA table for potato yield parameters 85ze did not allowed to analyze this variable 1 Main plot effects tested using main plot error Type III Mean Square 2 Source of Variation DF Rots Damaged B A1 A2 A3 A43 Market Yield Total Yield Specific gravity Replication 1 3 11.54 3.39 15.76* 101.77 129.12* 72.31* ---188.22 342.64 0.000047* Rotation 1 3 17.10 0.54 2.44 101.99 111.12* 14.65 ---431.07* 346.41 0.000109* Main Plot Error 9 11.75 4.03 2.56 77.01 27.97 12.61 ---102.88 101.65 0.000006 N rate 4 14.84** 3.73* 13.09** 192.09 533.78** 27.23* ---5299.24** 6704.77** 0.000174** Rotation N rate 12 3.17 2.45* 1.00 36.95 29.85 4.75 ---67.15* 65.45* 0.000008 Error 79 2.42 1.08 0.63 14.67 16.29 3.80 --32.18 32.55 0.000038 2 *Significant at Type I error < 0.05 **Significant at Type I error < 0.01 In this year low production of potatoes A4 si 3

PAGE 99

86 APPENDIX C AN YIELD 1 Croring green bean in the crop rotation. 2 N rate applied to potato in spring. p p eced Previous crop S orghum 1 Cowpea 1 Y2 Green bean yield (kg.ha-1) ea r Potato N rate(kg.ha -1) 0 5428 7355 100 57366653 150 617 56351 200 575 46705 20250 51967294 01 0 22053519 100 190 93671 150 163 13467 200 21193220 2002 250 129 92540 0 2860 2888 100 30343095 150 28393077 200 362 23513 20 73269 03 250 328 GREE N BE

PAGE 100

LISTES Asfary, A.F.; Wild, A.; Harris, P.M. Growth, Mineral Nutrition and Water Use By Potato Crops. J. Agri, 100 (1), 8Bishop, J.C.; Grimerecisiffectsoot and Tuer Production. Aotato Campbell, K.L.; RogeS.; Hensel, D.Rr Table ConPotatoes in Florida. Transactions of the ASAE. 1978, 21 1-705. Campbell, K.L.; Rogers, J.S.; Hensel, D.R.age Water Quality from Potato Production. 1985 (6), 1798-1801De Roo, H.C.; Waggoner, P.E. Root Development of Potatoes. Agron. J. 1961, 53, 15-17. Drew, M.C.; Jackson,.; Giffard, S. Et-promoted Adventitious Rooting and Development oical Air Spaceschyma) in Roots May be Adaptive Responses to Flooding in Zea mays 147, 83-88Elkash; Locas.J.; Hensel, D.Rutylidene Did Sulfucoated Urea as N Sources for Potatoes. J. Amer. Soc. Hort. Sci. 198Errebhi, M.; Rosen, C.J.; Gupta, S.C.; Birong, D.E. Potato Yield Response and Nitrate Leaching as Influenced by Nitrogen Management. Agron. J. 1998, 90, 10-15. tin, M.W.; J.B. Bamberg. Evaluation of Tuber-earing Solanum Species for Nitrogen Use Efficiency and Biomass Partitioning. Am. J. Potato Res. 1999, 76, 143-151. Gao, S.; Pan, W.L.; Koenig, R. T. Integrated Root System Age in Relation to Plant utrient Uptake Activity. Agron. J. 1998, 90, 505-510. Goolsby, D.A.; Battaglin, W.A.; Aulenbach, B.T.; R.P. Hooper. N Input to the Gulf of Mexico. J. Env. Qual. 2001, 30, (2) 329-336. Goulding, K. Nitrate Leaching from Arable and Horticultural Land. Soil Use and Management. Special issue: Tackling Nitrate from Agriculture. 2000, 16, (Suppl.) 145-151. Guddanty, S.; Chambers, J.L. GSRoot.-Automated Root Length Measurement Program. Users Manual. Version 5.00. Louisiana State University. Agricultural Center. 1993, OF REFERENC c. Sci. 1983 7-101. s, D.W. P on Tillage E on Potato RJ. 1978, 55, 65-71. b merican P rs, J. Wate trol for (4), 70 Drain 2 8 M.B hylene f Cort (aeren L. 1979 if, M.E. cio, S Isob urea an r3, 108 (4), 523-526. Errebhi, M.; Rosen, C.J.; Lauer, F.I.; Mar B N 87

PAGE 101

88 Griffin, T.S.; Hesterman, O.B.; Potato Response to Legume and Fertilizer Nitrogen Sources. Agron. J. 1991, 83: 6, 1004-1012. Havlin, J.L.; Beaton, J.D.; Tisdale, S.L.; Nelson, W.L. Soil Fertility and Fertilizers. An Introduction to Nutrient Management. 1999, 6th ed. Prentice Hall. Upper Saddle River, N.J. 499 p. Hill, A.R.; McCague, W.P. Nitrate Concentrations in Streams Near Alliston, Ontario, as Florida. S-756. 2000, 22 pp. nsion. Circular 1152. 2000. 8 Honisonse of Surface and Subsurface Water .; ion n Service. Gainesville, Lesczf Root Distribution of Irrigated, LivinApplicants Handbook. St. Johns River Water Management District, Influenced by Nitrogen Fertilization of Adjacent Fields. J. Soil and Water Conserv. 1974, 5, 217-220. Hochmuth,G.J.; Cordasco, K. A Summary of N, P, and K Research on Potato in University of Florida, IFAS. Extension. Circular H Hochmuth,G.J.; Hanlon, E.A. IFAS Standarized Fertilization Recommendations for Vegetable Crops. University of Florida, IFAS. Exte pp. ch, M.; Hellmeier, C.; Weiss, K. Resp Quality to Land Use Changes. Geoderma. 2002, 105, 277-298. Hutchinson, C.M.; Mylavarapu, R.S.; F. Munoz. Evaluation of Nitrogen Best Management Practices for Potato Production in Northeast Florida. ASA-CSSA-SSSA Annual Meeting Abstracts, Indianapolis, Nov 10-14, 2002. Iwama, K.; Hukushima, T.; Yoshimura, T.; Nakaseko, K. Influence of planting Denasity on Root Growth and Yield in Potato. Jpn. J. Crop Sci. 1993, 62:4, 628-635. Jensen, E.S. The Release And Fate Of Nitrogen From Catch Crop Materials Decomposing Under Field Conditions. Journal of Soil Science. 1992, 43: 2, 335-345. Joern, B.C.; Vitosh, M.L. Influence of Applied Nitrogen on Potato. Part I: Yield, Quality and Nitrogen Uptake. Am. Potato J. 1995, 72 (1), 51-63. Kidder, G.; Hochmuth, G.J.; Hensel, D.R.; Hanlon, E.A.; Tilton, W.A.; Dilbeck, J.DSchrader, D.E. Potatoes. Horticulturally and Environmentally Sound Fertilizatof Hastings Area Potatoes. Florida Cooperative Extensio Florida. 1992. ynski, D.B.; Tanner, C.B. Seasonal Variation o Field-Grown Russet Burbank Potato. American Potato J. 1976, 53, 69-78. gston-Way, P. Tri-County Agricultural Area Water Quality Protection Cost Share Program, Palatka, FL. 2000.

PAGE 102

89 Millburn, P.; Richards, J.E.; Gartley, C.; Pollock, T.; ONeill, H.; Bailey, H. Nitrate Leaching from Systematically Tiled Potato F ields in New Brunswick, Canada. J. Environ. Qual. 1990, 19, 448-454. Mitcht pean Journal of Agronomy. 2001, 15, 17-29. ostate.edu/pubs/crops/00541.html (accessed September 2003). TL) nsion. Circular 1248. 2002, 18 pp. eetesen and K. Dilz (Eds). Institute for Soil Fertility, Haren (Gn) Netherlands. 1985, pp101-110. Natio pril 2004). itrate with Peralta, J.M.; Stockle, C.O. Dynamics of Nitrate Leaching under Irrigated Potato en Porter, G.A.;Opena, G.B.; Bradbury, W.B.; Mcburnie, J.C.; Sisson, J.A. Soil Management and Supplemental Irrigation Effects on Potato: Soil Properties, Tuber Prunty, L.; Greenland, R. Nitrate Leaching using Two Potato-Corn Fertilizer Plans on Rabalais, N.N.; Turner, R.E.; Scavia, D. Beyond Science Into Policy: Gulf of Mexico Hypoxia and the Mississippi River. Bioscience 2002, 52, (2) 129-142. ell, R.; Webb, J.; Harrison, R. Crop Residues Can Affect N Leaching Over at LeasTwo Winters. Euro Mortvedt, J.J.; Soltanpour, P.N.; Zink, R.T.; Davidson, R.D. Fertilizing Potatoes. 2001. (Web page) Colorado State University. http://www.ext.col Mueller, D.K.; Dennis, R.H. Nutrients in the Nations Waters. Too Much of a Good Thing? US Geological Survey. Circular 1136, 1996. Mylavarapu, R. S.; Kennelley, E. D. UF/IFAS Extension Soil Testing Laboratory (ESAnalytical Procedures and Training Manual. University of Florida, IFAS. Exte Nielsen, N.E.; Jensen, H.E. Soil Mineral Nitrogen as Affected by Undersown CatchCrops. In: Assessment of Nitrogen Fertilizer Requirement. J.J. N nal Oceanic and Atmospheric Administration. (NOAA). National Weather ServiceWeb-page: www.nws.noaa.gov Weather database (Accessed on A Omernik, J.M. Nonpoint Source-Stream Nutrient Level RelationshipsA Nationwide Study. EPA /3-P77-105 USEPA, Corvallis, Or. 1977, 151 pp. Patel, R.M.; Prasher, S.O.; Broughton, R.S. Upward Movement of Leached NSub-Irrigation. Trans. ASAE. 2001, 44 (6), 1521-1526. Rotation in Washington State: A Long-Term Simulation Study. Agric, Ecosystems and Environ. 2002, 88, 23-34. Porter, G.A.; Sisson, J.A. Response of Russet Burbank and Shepody Potatoes to NitrogFertilizer in Two Cropping Systems. American Potato J. 1991, 68, 425-443. Yield and Quality. Agron. J. 1999, 91,(3), 416-425. Sandy Soil. Agric, Ecosystems and Environ. 1997, 65, 1-13.

PAGE 103

90 Randall, G.W.; Mulla, D.J. Nitrate Nitrogen in Surface Waters as Influenced by ClimConditions and Agricultural Practices. J. Envir. Qual. 2001, 30, (2), 337-344. atic Richards, J.E.; Milburn, P.H.; Maclean, A.A.; Demerchant, G.P. Intensive Potato Ruser, R.; Flessa, H.; Schilling, R.; Steindl, H.; Beese, F. Soil Compaction and il il. 1, 59, 177-191. SattelMorphology of Two Potato Varieties Differing in Nitrogen Aquisition. Plant and Sanderson, J.B.; Macleod, J.A. Soil Nitrate Profile and Response of Potatoes to Fertilizer f Influence on the Following Crops. Biological Agriculture and Horticulture. 1999, 17(2), 159-170. Schroction System. Netherlands J. of Agric. Science. 1996, 44, (4), 293-315. Scott-oot to mechanical impedance. 1974, Neth. J. Agric. Sci. 22, 305-318. ShephCrop Rotation. J. Agric. Sci. 1996, 127, 215-229. Singhs of Nitrates Beyond Potential Rooting Zone. I. Proper Coordination of Nitrogen Splitting with Water Production Effects on Nitrate-N Concentrations of Rural New Brunswick Well Water. Canadian Agricultural Engineering. 1990, 32, (2) 189-196 Robinson, D.; Linehan, D.J.; Caul, S. What Limits Nitrate Uptake from Soil? Plant, Celland Environment. 1991, 14, 77-85. Fertilization Effects on Nitrous Oxide and Methane Fluxes in Potato Fields. SoSci. Soc. Am. J. 1998, 62,1587-1595. Ruser, R.; Flessa, H.; Schilling, R.; Beese, F.; Munch, J.C. Effect of Crop Specific Field Management and N Fertilization on N2O Emissions from a Fine-Loamy SoNutrient Cycling in Agroecosystems. 200 SAS Institute, The SAS system for windows. Release 8.02 1999-2001. macher, B.; Klotz, F.; Marschner, H. Influence of the Nitrogen Level on Root Soil. 1990, 123, 131-137. N in Relation to Time of Incorporation of Lupin (Lupinus Albus). Canadian J. oSoil Sci. 1994, 74, (2), 241-246. Schmidt, H.; Phillipps, L.; Welsh, J.P.; Fragstein, P.V. Legume Breaks in Stockless Organic Farming Rotations: Nitrogen Accumulation and der, J.J.; van Dijk, W.; de Groot, W.J.M. Effects of Cover Crops on the Nitrogen Fluxes in a Silage Maize Produ Russell, R.; Goss, M.J. Physical Aspects of Soil Fertility The Response of r erd, M.A.; Lord, E.I. Nitrate Leaching from a Sandy Soil: The Effect of Previous Crop and Postharvest Soil Management in an Arable B.; Sekhon, G.S. Some Measures of Reducing Leaching Los Management. Plant and Soil. 1976a, 44, 193-200.

PAGE 104

91 Singh, B.; Sekhon, G.S. Some Measures of Reducing Leaching Loss of Nitrates Beyond Potential Rooting Zone. II. Balanced Fertilization. Plant and Soil. 1976b, 44, 3 91-395. Singhd 977, 47, 585-591. Stalhaty, Irrigation Regime and Planting Date on depth, Rate, Duration and Density of Root Growth in the Potato (Solanum St. Jo (SJRWMD). Nutrient Pollution How much is a Lot? Tri-County Crop Talk. Newsletter Summer/Autumn. Palatka, FL. 1996, St. Jone tml Nitrogen Nutrition of Succeeding Crops. Fertilizer Research. 1994, 37, 227-234. Thoruing Crops. Plant and Soil. 1998, 203, 79-89. Tyler. oil van Loon, C.D.; Bouma, J. A case study on the effect of soil compaction on potato 6, Vos, J.; Groenwold, J. Root Growth of Potato Crops on a Marine-clay soil. Plant and Vos, J.; van der Putten, P.E. L. Nutrient Cycling In A Cropping System With Potato, B.; Sekhon, G.S. Some Measures of Reducing Leaching Loss of Nitrates BeyonPotential Rooting Zone. III. Proper Crop Rotation. Plant and Soil. 1 m, M.A.; Allen, E.J. Effect of Varie tuberosum) Crop. J. Agric. Sci. 2001, 137, 251-270. hns River Water Management District. hns River Water Management District. (SJRWMD). Web-page. (accessed in Ju2004) www.sjrwmd.com/programs/outreach/pubs/index.h Thorup-Kristensen, K. The Effect of Nitrogen Catch Crop Species on the p-Kristensen, K.; Nielsen, N.E. Modeling and Measuring the Effect of Nitrogen Catch Crops on the Nitrogen Supply for Succeed K.B.; Broabent, F.E.; Bishop, J.C. Efficiency of Nitrogen Uptake by Potatoes. AmPotato J. 1983, 60 (4), 261-269. U.S. Department of Agriculture (USDA). Soil survey of St. Johns County, Florida. SConservation Service. 1981, 196 pp. growth in a loamy sand soil. 2. Potato plant responses. Neth. J. agric. Sci. 1978, 2421-429. Soil. 1986, 94 17-33. Spring Wheat, Sugar Beet, Oats and Nitrogen Catch Crops. I. Input and Offtake ofNitrogen, Phosphorus and Potassium. Nutrient Cycling in Agroecosystems. 2000, 56: 87-97. Waddell, J.T.; Gupta, S.C.; Moncrief, J.F.; Rosen, C.J.; Steele, D.D. Irrigation and Nitrogen Management Effects on Potato Yield, Tuber Quality, and Nitrogen Uptake. Agron. J. 1999, 91, 991-997.

PAGE 105

92 Waddell, J.T., Gupta, S.C.; Moncrief, J.F.; Rosen, C.J.;Steele, D.D. Irrigation and Nitrogen Management Impacts on Nitrate Leaching under Potato. J. Env. Qual. 2000, 29, 251-261. arcel Dekker, Inc. New York, NY. pp 323 331. Webb, J.; Sylvester-Bradley, R.; Seeney, F.M. The Fertilizer Requirement of Cereals Westra, J.V.; Boyle, K.J.; Porter, G.A. Net Revenues of Potatoes Rotated With Other Waisel, Y. Aeroponics: A Tool for Research Under Minimal Environmental Restrictions.In: Plant RootsThe Hidden Half. Y. Waisel; A. Eshel; U. Kafkafi (Eds). 2002, 3rdEd. M Weaver, J.E. Root Development of Field Crops. McGraw-Hill Book Co., New York. 1926. Grown in Sandy Soils. J. Sci. Food and Agric. 2000, 80, 263-274. Crops. Am. Potato J. 1995, 72, 99-117.

PAGE 106

BIOGRAPHICAL SKETCH Fernando Muoz was born in Cali, Colombia, on December 19, 1959. In 1976, h e completed high school at Colegio de Santa Librada. He attended the Universidad Naciohe opportunity Agriculture (CIAT) in Cali, Colombia. After earned his B.S. in agronomy he was hired t in the Cassava Program. In 1991, he was warded with a scholarship from the Dutch government to attend the Centro Agronomico ropical de Investigacion y Enseanza (CATIE) in Turrialba, Costa Rica. There, he eveloped his research work with the CATIE-GTZ Agroforestry Project earning his .Sc. degree in agricultural production systems with emphasis in agroforestry. He went back to CIAT to work as Research Associate in the Plant Nutrition Section of the Bean Program. In 1998 he joined the research team of the Pathology Section at the Tropical Forages Program in CIAT. He joined the University of Florida, Gainesville, Florida, in summer 2001 to pursue a Doctor of Philosophy in soil and water science. After getting his degree, he plans to continue his professional career on agricultural research. nal de Colombia, and at the end of his undergraduate work he had t to conduct his thesis in the Agronomy section of the International Center for Tropical by CIAT to work as Research Assistan a T d M 93


Permanent Link: http://ufdc.ufl.edu/UFE0006202/00001

Material Information

Title: Improving Nitrogen Management in Potatoes through Crop Rotation and Enhanced Uptake
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0006202:00001

Permanent Link: http://ufdc.ufl.edu/UFE0006202/00001

Material Information

Title: Improving Nitrogen Management in Potatoes through Crop Rotation and Enhanced Uptake
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0006202:00001


This item has the following downloads:


Full Text












IMPROVING NITROGEN MANAGEMENT IN POTATOES THROUGH
CROP ROTATION AND ENHANCED UPTAKE
















By

FERNANDO MUNOZ


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004


































Copyright 2004

by

Fernando Munoz

































This work is dedicated to Luz Marina, Jose Fernando, and Isabela















ACKNOWLEDGMENTS

I would like to express my gratitude to Dr. Rao S. Mylavarapu, my advisor, for his

support, time and confidence, and for allowing me develop my ideas about root systems

and their relationship with nutrient uptake. My appreciation is also extended to my

committee members, Drs. Chad Hutchinson, Kenneth Portier, Thomas Obreza, and

Yuncong Li, for their support, useful comments, suggestions, and patience.

I thank the staff at the Hastings Research and Education Center's Yelvington Farm

in Hastings, Florida especially Pam Solano and Douglas Gergela; it would have been too

hard to do this work without their opportune aid.

I am so grateful to Dr. Kenneth Portier for all the time devoted to support me with

the statistical analysis of my research data. His kindness and patience were so valuable to

get this study done.

Special thanks go to Dr. Jim Chambers from Louisiana State University and Dr.

Nick Comerford at University of Florida for authorizing me to use GSRoot, the

automated root length measurement software used here to estimate the potato root

parameters. Their kindness facilitated my work.

My appreciation and gratitude go to Joseph Nguyen and Martin Anderson; their

excellent assistance helped me get this study successfully done.

Thanks go to my friends, Jaime Sanchez, Daniel Herrera, and Leighton Walker, for

their unconditional friendship throughout these three years in UF. We shared funny talks

that softened rough times.









I acknowledge the faculty, students, and staff in the Soil and Water Science

Department for allowing me to be here. For sure these three years sharing classrooms,

seminars, and informal talks will be significant in my professional life.

I would like thank my parents and brothers for their support from a distance. I also

thank my aunt Nubia for her unselfish and opportune support.

Finally, I want to say thanks to Luz Marina, my wife, and to Jose Fernando and

Isabela, my children, for their unconditional support and love.
















TABLE OF CONTENTS

page

ACKNOW LEDGM ENTS ........................................ iv

LIST OF TABLES .................................... ....... ..................... ix

LIST OF FIGURES ............................................... ..............x

ABSTRACT................................. .............. xii

CHAPTER

1 INTRODUCTION ................... .................. .............. .... ......... .......

Nitrates as a Non-Point Source of Water Pollution.................................................1
The N itrate Problem in N northeast Florida ............................................... ......2
O organization of this D issertation ....................................................................................4

2 NITRATE-N LOSSES FROM POTATOES GROWN UNDER DIFFERENT
CROP ROTATIONS AND NITROGEN RATES.....................................................6

Introduction ...................................... .......... ......... .. .6
M materials and M methods .............................................................9
Site Description..................................................... .........9
Soils .............. ........... ......... ................................ 9
Statistical Design and Experimental Layout ................................... ...............10
Crop Production Practices ................. ...................................... ...... .........10
P lant m materials ................... ............................................... .............. 10
Pest m anagem ent............... ............................. ....... .. ...... .............. 12
Irrigation m anagem ent ............................................................................ 12
Nutrient management .................................. ........................... ..... 12
Water and Soil Sampling Schedule ........................................13
Soil sam pling ...................................... ............................................ 13
W ater sampling .............................................. ...... ..14
Tissue Sam pling Schedule.................................................... 16
Potato tissue....................................... ........ ........... .16
Sorghum and cow pea tissue .......................... .................... .......... 17
Potato yield ................ ................. .... .........17
Statistical Analysis .................. ... ....... ........................... 18
Concentrations of NOs-N in the water table ..................................... 18










Concentrations of NO3-N and NH4-N in the soil ............... ................18
N itrogen uptake by the crops ....................................................... 19
Potato yield................................... ........ 20
Results and D discussion .................................. .... .... ...... ..............20
Concentrations of NO3-N in the Water Table under Potato............................20
Concentrations of NO3-N in the Soil Solution ......................... ...............30
Concentrations of NO3-N in the Water Table under Cowpea and Sorghum.......33
Concentrations of NO3-N and NH4-N in the Soil............... ..................34
Nitrogen Uptake by the Crops ....... ....... ........................ 39
Sorghum .......................................... ........39
Cowpea ............................................ ......... .. .40
Potato .......... .. ............... .................. 41
Potato Yield .............. ....... ............................ ........ 42
Conclusions....................................... ........ .47

3 POTATO ROOT DISTRIBUTION STUDY .......................................49

Intro du action ................ ......... .................................. 4 9
M materials an d M eth o d s .......................................................................................... 5 1
Site Description....................................... ........ ...........51
Statistical Design........................................ .................51
Root Sam pling M ethodology ................................... ...... ............... 51
Root W washing ....................... .. ................. ........ 53
Root Scanning and Image Analysis............................. ............... 53
Root Variables ................. ......... .......... .... ......... ................. 53
Soil Strength, Bulk Density, and NO3-N concentration by Depth ......................54
Statistical Analysis ...................... ........ ........ .... ...55
R oot sam pler ................ ........................55
Soil strength, bulk density, and NO3-N concentrations by depth..............55
R oots...................................... ............................... ......... 56
Results and Discussion ........................ ................. .......... 56
Root Sam pler D evice ................... .................. .................... .. ... .. ... ..... 56
Soil Strength and Bulk density .................................................................... 57
Rooting Depth, Root Length, and Root Surface Area.......................................58
R oot Length D ensity................. ...... ...........................................60
R oot Surface A rea D ensity ....................................................................... ...... 61
Specific R oot Length ................... ..................................................................... ....... 63
Specific R oot Surface A rea ........................................... ............... 64
Conclusions..............................................................66

4 SUMMARY, OVERALL CONCLUSIONS AND FUTURE RESEARCH ..............67

Sum m ary ............... ........ .. ......... .......................... 68
Crop Rotation and N rates .......... ... ........... ..... .........................68
Concentrations of NO3-N in the water table under potato ...........................68
Concentrations of NO3-N in the soil solution ......................................... 69
Concentrations of NO3-N in the water table under cowpea and sorghum ...69










Concentrations of NO3-N and NH4-N in the soil .......................................69
Nitrogen recovery by the summer cover crops ........................................ 70
Potato N uptake at full flow ering ...................................... ........... ....70
Potato yield.......................................... ........ 71
Potato Root Distribution Study ............................... ............... 71
R oot sam pler ......... .. .. .......................................... .............71
Soil strength, bulk density, and N03-N by depth ......................................72
Rooting depth, root length, and root surface area ......................................72
Root length density and specific root length............... .................72
Root surface area density and specific root surface area..............................73
Overall Conclusions.............................. .............. 73
Future Research ......................................... .........76

APPENDIX

A REGRESSION PARAM ETERS ....................................................... 77

B ANOVA TABLES FOR POTATO YIELD PARAMETERS...............................82

C G R E E N B E A N Y IE L D ......................................................................................... 86

L IST O F R E FE R E N C E S ............................................................................... 87

BIOGRAPHICAL SKETCH ............................................... 93
















LIST OF TABLES


Table page

2-1 Nitrogen rate treatments and nutrients applied to potato crop each year..............13

2-2 Precipitation during crops and intercrops periods during the study period...........29

2-3 Effect of crop rotation on soil NO3-N concentration at potato planting ..................38

2-4 Nitrogen uptake of cowpea by year at incorporation time. ...................................41

2-5 ANOVA of the annual potato N uptake over the 3 year period..............................41

2-6 Marketable and Total yield of potato under rotation by N rate...............45

2-7 Tuber exported N, % TKN, and specific gravity of potato under year by N rate
interaction....................................... ................................ ......... 47

3-1 Soil strength, Bulk density, and Soil NO3-N by depth....................................57

3-2 Root length and root surface area by diameter class and soil depth....................60

B-1 2001 ANOVA table for potato yield parameters ...................................... 83

B-2 2002 ANOVA table for potato yield parameters ..................................... 84

B-3 2003 ANOVA table for potato yield parameters ..................................... 85
















LIST OF FIGURES


Figure page

1-1 Localization map of the tri-county agricultural area (TCAA) in northeast Florida.. .3

2-1 Field plot plan, crop rotations and N rates .......................................................11

2-2 Crop cycles, sampling schedules, and precipitation pattern over the three-year
period of study ...................................... ..... ........... .15

2-3a Predicted NO3-N concentration in the water table under two crop rotations and
five N rates during the first year.............................................................. ........... 21

2-3b Predicted NO3-N concentration in the water table under two crop rotations and
five N rates during the first year.............................................................. ........... 22

2-4a Predicted NO3-N concentration in the water table under four crop rotations and
five N rates during the second year................................................................. 24

2-4b Predicted NO3-N concentration in the water table under two crop rotations and
five N rates during the second year................................................................. 25

2-5a Predicted NO3-N concentration in the water table under two crop rotations and
five N rates during the third year ................................. ............... 27

2-5b Predicted NO3-N concentration in the water table under two crop rotations and
five N rates during the third year ................................. ............... 28

2-6a Predicted NO3-N concentration in the soil solution under two crop rotations and
five N rates during the second year................................................ .......... 31

2-6b Predicted NO3-N concentration in the soil solution under two crop rotations and
five N rates during the second year................................................ .......... 32

2-7 NO3-N concentrations in the water table under sorghum and cowpea ....................33

2-8 Predicted N03-N concentration in the soil by effect of five N rates .......................35

2-9 Predicted NH4-N concentration in the soil by effect of five N rates ..................36

2-10 Nitrogen uptake of sorghum by residual effect of the fertilizer applied to potato ...40









2-11 Potato N uptake at full flowering during the study period..................... ..........42

2-12 Predicted N rate for maximum marketable potato yield .......................................43

3-1 Procedure to take the root sam ples................................ .................. 52

3-2 Distribution and sample size into the slice. .........................................57

3-3 D etail of potato roots thickening............................................................ ....... 59

3-4 R oot length density distribution .......................................................... ......... 61

3-5 Root surface area density distribution.............................................62

3-6 Specific root length distribution ........................................................ 64

3-7 Specific root surface area distribution.............................................65
















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

IMPROVING NITROGEN MANAGEMENT IN POTATOES THROUGH CROP
ROTATION AND ENHANCED UPTAKE

By

Fernando Munoz

August 2004

Chair: Rao S. Mylavarapu
Major Department: Soil and Water Science

Impact on water quality due to nutrient loading of our finite water resources is a

growing concern. Nitrate and phosphates originating from agricultural operations have

been identified as the primary contributors to such non-source point pollution of water.

The Tri-County Agricultural Area (TCAA) in north-east Florida is predominantly

dedicated to intensive potato production system with producers often resorting to over-

fertilization, particularly with N sources. Best management practices (BMPs) including

crop rotation and enhanced N uptake have been proposed as options to minimize the risk

of nitrate contamination of the surface and groundwater in the St. Johns River watershed.

To evaluate the effect of N03-N concentration on shallow water table and to determine

optimum N requirement for an economically successful potato crop (var. Atlantic), a 3-

year field study with four potato-cover crop rotations and five N rates was conducted in

the TCAA. Increased N03-N concentrations in the shallow water table were recorded

immediately after N application at planting but not after the side dressing. At the end of









potato season an increase in soil N03-N resulting from residual effect of N side-dress

was found. While nitrate concentrations produced linear and quadratic increase of

sorghum N uptake across N rates, cowpea did not exhibit any response. Rotation with

green bean in the fall increased soil N03-N concentration at potato planting in the next

season, and a high rainfall between during the intervening period increased the potential

for nitrate leaching and run-off Potato in rotation with sorghum produced higher yields

than in rotation with cowpea. Sorghum was found to be a more efficient catch crop for

residual N03-N.

A study of potato root distribution under the effect of three N rates was done in the

third year. There was no effect of N rate on root length density, specific root length, root

surface area density, and specific root surface. Observations suggested that potato root

system was constrained due to soil compaction, and increased soil bulk density recorded

at the study site. A soil region where the potato root system showed highest activity was

identified which could guide precise fertilizer placement.














CHAPTER 1
INTRODUCTION

Nitrates as a Non-Point Source of Water Pollution

Worldwide, agriculture has been identified as a major contributor of nitrate-N to

surface water (Randall and Mulla, 2001). A recent study showed that the mean annual

input of N to the Gulf of Mexico from the Mississippi-Atchafalaya River Basin, the third

largest river basin in the world, was estimated to be 1,568,000 t yr1 in the period 1980-

1996; the mean annual flux of N was about 61% nitrate-N, 37% organic-N, and 2%

ammonium-N (Goolsby et al. 2001). The high N load, mainly as nitrate, has been related

to increased eutrophication and hypoxia in the Mississippi River Delta (Rabalais et al.

2002). A US nationwide study reported nine times greater total N concentration

downstream from agricultural lands than downstream from forested areas (Omernik,

1977). Along with several major horticultural crops, potato (Solanum tuberosum L.)

cropping regions have been suspected of adding excess nitrate to surface and

underground waters in Germany (Honisch et al. 2002) and Canada (Milbum et al. 1990).

The nitrate concentration in rural well water in New Brunswick (Richards et al. 1990)

and the nitrate concentration in streams flowing through the cropped fields in Ontario

(Hill and McCague, 1974) have been associated with the proportion of the land cropped

to potatoes. Therefore, emphasis on how intensive management practices impact the

environment has been the primary focus of current day research (Ruser et al. 1998;

Honish et al. 2002).









Application of higher rates of N than are required for optimum potato production

coupled with the susceptibility of NO3 leaching by rainfall and irrigation have been the

primary focus of current day research (Milbum et al. 1990). Furthermore, studies

characterizing N uptake by potato roots have shown that N fertilizers are generally

applied in excess of the optimal rate for maximum yield (Tyler et al. 1983; Joern and

Vitosh, 1995a; Prunty and Greenland, 1997). Over-fertilization increases the risk of

environmental pollution because N not taken up by the plant can move out of the root

zone. Synchronization between N availability in the soil and N uptake by the potato root

has been proposed as a core strategy to maximize yield and minimize N loss (Ruser et al.

1998; Waddel et al. 2000; Mortvedt et al. 2001). Furthermore, the efficiency of N uptake

can be improved by placing fertilizer in bands and/or applying fertilizer in multiple,

relatively small applications (Ruser et al. 1998; Waddel et al. 1999).

Measures such as suppression of tile-drainage systems, construction of wetland

restoration areas before discharge into ditches and rivers, application of correct rate of N

at the optimum time, development of improved soil N testing methods to predict

availability of mineralized N and carryover from previous crops, improved management

of green and animal manures (Randall and Mulla, 2001), and increasing N uptake, and

use efficiencies (Errebhi et al. 1998, 1999) are accepted strategies to reduce nitrate losses

to surface and groundwater.

The Nitrate Problem in Northeast Florida

The TCAA (Tri-County Agricultural Area), located in Flagler, Putnam and St.

Johns counties in northeast Florida (Fig. 1-1), contains approximately 15,000 ha of

irrigated cropland, predominantly potato and cabbage farms. For more than a century,

agricultural production and services have contributed significantly to the economic










viability of the basin. Non-point source pollution resulting from storm water runoff from

agricultural systems is considered to be one of the major sources of pollution in the low




The
~nA-l SA Fe-and!-- St. Johns River

Water

BAAER Management
1i A A :.Lj-ck Wie
District


S 9ring % tL. Augustine



F LAGLE
B. ell Ick




r1~ 1r~ L. rltfwiI
Foiai A i laion s e.
j-,~
V 0 L U S KA
A I~


0 R










1 0 1. 1 Vero
FI r ER Belas-l






Figure 1-1. Localization map of the tri-county agricultural area (TCAA) in northeast
Florida. Approximate location of the study site.

St. Johns River Basin (LSJRB). It has been estimated that nonpoint pollutant sources

induced by human activities may account for as much as 36% of the amount of pollutants

entering the basin today (SJRWMD, 2004). Past work has identified the TCAA as a









possible source of non-point pollution from enriched nutrient runoff. Enhanced algal

growth in the St. Johns River coinciding with peak runoff associated with the TCAA

potato season have been reported (SJRWMD, 1996). Best management practices (BMPs)

to control water runoff and to improve fertilizer management by reduced rates and

improved timing lowered the phosphorus runoff into the St. Johns River watershed 22.5

% the first year and about 55 % in 15 years (SJRWMD, 2004). A study comparing the

traditional water furrow irrigation-drainage system with a system using subsurface

irrigation and drainage showed that surface runoff nitrate concentrations were often

higher than concentrations in the subsurface drains (Campbell et al. 1985). Best

management practices to minimize potential runoff and leaching of nitrates from potato-

cropped lands including nutrient management, irrigation and sediment management, and

crop rotation with low N requiring crops have been implemented in the TCAA

(Hutchinson et al. 2002).

This three year study was initiated to document the impact the impact of legume

rotation crops on N03-N concentrations in the water table under potato beds, N uptake by

cover crops, potato tuber production, and possible reduction of the N rate of soluble

fertilizers currently used in the TCAA for potato production. Along with this study, a

study of the potato root system addressed to describe its spatial distribution under a

sufficient N fertilization regime. The knowledge of the spatial distribution of potato

would allow in the future optimize fertilizer placement in order to increase potato N

uptake.

Organization of this Dissertation

This work is organized in four chapters. First chapter is a general introduction

addressing the leaching/runoff problem from potato fields with emphasis in northeast






5


Florida potato production. The second chapter reports results of an 3-year experiment

with four crop rotations including legume crops as supplemental N source to five N rates

applied to potato. Effect of rotation and N rate on N03-N concentration, N uptake by the

crops and potato yield are discussed. The third chapter describe and report results of a

study of the potato root distribution under effect of three sufficient N rates. The fourth

chapter is addressed to summarize, state general conclusions, and suggest future research

topics.














CHAPTER 2
NITRATE-N LOSSES FROM POTATOES GROWN UNDER DIFFERENT CROP
ROTATIONS AND NITROGEN RATES

Introduction

Potato has been reported as a crop with high potential to generate groundwater

contamination due to its high N demand to produce optimum yields. Farmers usually

apply N fertilizer in excess of the recommended rate as an insurance against N losses due

to N deficiency. In northeast Florida, potato is grown as a late winter or early spring crop

using seepage irrigation. Seepage irrigation is a subsurface irrigation system where the

depth to or the level of water in the perched water table is controlled taking advantage of

the naturally existing hardpan averaging at a depth of 1.5 meters. Seepage irrigation is

very unique to this region of the country in contrast with the overhead irrigation used in

the rest of the potato producing states in the US. The sandy nature of Florida soils, excess

nitrate in the potato beds, and seasons with high rainfall can create conditions resulting in

N leaching and/or runoff with the subsequent eutrophication of water bodies. In northeast

Florida, most of 9000 ha under potato production are located in the Tri-County

Agricultural Area (TCAA) a major agricultural region under the St. Johns River Water

Management District (SJRWMD) boundaries. The SJRWMD being responsible for both

the water quality and quantity is particularly concerned about the potential nutrient losses

from the TCAA to the St. Johns River (Livingston-Way, 2000).

The agricultural practice of alternating crops on a particular during a defined period

of time (crop rotation) has several benefits, including interrupting plant disease and insect









cycles, increasing mineral or organic content, and improving soil structure. This practice

can be used to store N in a catch crop during fallow periods, decreasing potential nitrate

leaching (Nielsen and Jensen, 1985; Schroder et al. 1996). Cover crops fit the definition

of a rotational crop and are used extensively in production systems that utilize organic

fertilizers. Cover crops also provide soil coverage during the periods when no

commercial crops are grown, serve as a N sink after the commercial crop is harvested,

and act as a green manure for subsequently planted cash crops (Goulding, 2000). Using a

cover crop as a green manure lessens the impact of nitrate leaching by maintaining the N

in a slow-release form (Thorup-Kristensen and Nielsen, 1998). However, the

mineralization of green manures under specific environmental conditions must be known

to insure that mineralization of residues is synchronized with crop need (Griffin and

Hesterman, 1991; Porter et al. 1999; Schmidt et al. 1999).

In the US, crop rotation with cover crops is used in areas such as the Great Plains,

Florida, the northeast and along the west coast where well drained soils and intensive

farming of row crops like potato cause high nitrate concentrations in groundwater

(Mueller and Dennis, 1996). It has been further reported that plant residues of arable

crops such as potato increased N leaching over at least two winters in free drained soils

(Mitchell et al. 2001). The great challenge in establishment of a successful crop rotation

with catch crops is synchronizing the availability of nutrients from mineralized residues

with nutrient demand of the target crop in the rotation. Generally, most commercial crops

are fertilized with close to optimum rates and the primary objective of catch crops would

be to replace only part of the soluble fertilizer N (Thorup-Kristensen, 1994).









Factors and processes such as content of nitrate in the soil, N mineralization and

biomass production of the catch crop (Jensen, 1992), rooting depth of the catch crop,

climate, soil type, and rooting depth of the succeeding crop (Thorup-Kristensen and

Nielsen, 1998) must be taken in account to select the most appropriate catch crop species

under specific environmental conditions. Nitrogen residues from fertilizers in the soil

after potato harvest are generally high and depending of the irrigation applied, could be

deep in the soil profile (Webb et al, 2000). Therefore, the catch crop used in rotation with

potatoes would require an efficient root system in order to recover the N not taken up by

the potato (Thorup-Kristensen and Nielsen, 1998; Singh and Sekhon, 1977). C/N ratio

and incorporation time of catch crop must also be considered and adapted to the specific

environmental condition (Thorup-Kristensen and Nielsen, 1998).

Legume cover-crops such as lupin (Sanderson and MacLeod, 1994), green bean

(Hutchinson et al. 2002) and red clover (Porter and Sisson, 1991), and cereal crops wheat

and maize (Singh and Sekhon, 1976a, 1976b, 1977), sorghum (Hutchinson et al. 2002),

winter rye, wheat, and barley (Shepherd and Lord, 1996; Vos and van der Putten, 2000)

have been used as green manures and/or additional cash crops in rotation with potato.

Modification of potato production practices including rotation with cash crops and green

manure crops adapted to specific environmental situations could help producers increase

net revenues by reduction of purchased inputs and increase tuber quality and/or yield

(Westra et al. 1995). This research studied the effect of replacing sorghum sudangrass,

the traditional summer cover crop used in the TCAA, with cowpea and/or the

introduction of green bean as a fall cash crop.









The objectives of this study were i) to identify and to quantify elevations in nitrate

concentrations in the perched water table under potato beds due to leaching and ii) to

study the effect of legume crops as possible rotation with potato on nitrate leaching and

potato yield.

Materials and Methods

Site Description

The study included three potato seasons in the spring of the years 2001, 2002, and

2003 at the UF/IFAS Plant Science and Education Unit, Yelvington Farm in Hastings,

FL. The mission of the research farm is to conduct applied and adaptive production

research in vegetables, agronomic, and other crops produced in north Florida and to

disseminate new information to the academic and agricultural communities. The facility

contains 20 ha of land for university research in plant and soil science. Weather data is

continuously recorded at the research site through the Florida Automated Weather

Network (FAWN)

Soils

The soil at the experimental site was classified as sandy, siliceous, hyperthermic

Arenic Ochraqualf belonging to the Ellzey series (USDA, 1981). This soil has a sandy

loam layer about 1 m deep with a clayey restricting layer below. As a result the profile is

very poorly drained profile even though the saturated hydraulic conductivity in the top 1

m is about 10 cm-h-1. This soil is about 94% sand, 2.5% silt, and 3.5% clay (Campbell et

al. 1978). In its natural state, the water table is within 25 cm of the surface for 1 to 6

months in most years and slopes are less than 2% (USDA, 1981).









Statistical Design and Experimental Layout

The study was conducted at the same location all the three years of the study. The

experiment was replicated four times, and arranged in a split-plot design where the main

plot factor was crop rotation and the sub-plot factor was N fertilizer rate applied to potato

(Table 2-1). Each block consisted of one bed 16.32 m wide (16 rows) by 160 m long (Fig.

2-1). The block was divided into two strips of 8.16 m (8 rows) by 160 m. Each of the

strips (8.16 m x 160 m) was divided again in two parts resulting 4 areas (main plots) of

8.16 m wide (8 rows) by 80 m long in each block (Fig. 2-1). To each of these main plots

a crop rotation was randomly assigned for summer and fall. In summer 2000 main plots

planted with sorghum and cowpea as summer cover crops (SCC) were randomly

assigned. Green bean and fallow main plots were randomly assigned in Fall 2000. In

spring 2001, before potato planting each of the main plots was divided into five subplots

(8 rows x 16m) and the five nitrogen rates (Table 2-1) were randomly assigned to each of

the subplots. Randomization of both the main plot subplot treatments was maintained

during the 3-year study period. A code consisting on three letters to identify the four

rotations is detailed in Fig. 2-1.

Crop Production Practices

Plant materials

Potatoes (Solanum tuberosum, var. 'Atlantic') were planted in early February each

year. Seed pieces were cut to a target size of 57 g and planted at an in-row spacing of

20.3 cm with a 101.6 cm between-row spacing resulting in a crop density of 48,216

plants-ha-1. The sorghum/sudan grass hybrid (Sorghum vulgare x Sorghum vulgare, var.

Sudanese, var. SX17) and cowpea (Vigna unguiculata, var. Iron Clay) were seeded in a

single row on a 101.6 cm between-row spacing and a within-row seed spacing of 2.5 cm





















Block 1

3 2
0 kg/ha N 224 kg/ha N



224 kg/ha N 112 kg/ha N



112 kg/ha N 0 kg/ha N



280 kg/ha N 168 kg/ha N



168 kg/ha N 280 kg/ha N

1 4
224 kg/ha N 168 kg/ha N



280 kg/ha N 0 kg/ha N



0 kg/ha N 224 kg/ha N



112 kg/ha N 112 kg/ha N



168 kg/ha N 280 kg/ha N


Rows 1 -8


Rows 9 -16


Block 2

1 *4
168 kg/ha N 280 kg/ha N



280 kg/ha N 0 kg/ha N



112 kg/ha N 224 kg/ha N



224 kg/ha N 168 kg/ha N



0 kg/ha N 112 kg/ha N

3 2
280 kg/ha N 168 kg/ha N



168 kg/ha N 112 kg/ha N



0 kg/ha N 280 kg/ha N



224 kg/ha N 0 kg/ha N



112 kg/ha N 224 kg/ha N


Main Plot Spring Summer cover crop Fall crop Code
1 Potato Sorghum sudan Fallow PSF
2 Potato Sorghum sudan Green beans PSG
3 Potato Cowpea Fallow PCF
4 Potato Cowpea Green beans PCG




Figure 2-1. Field plot plan, crop rotations and N rates


Block 3

4 3
0 kg/ha N 168 kg/ha N



224 kg/ha N 112 kg/ha N



280 kg/ha N 0 kg/ha N



112 kg/ha N 224 kg/ha N



168 kg/ha N 280 kg/ha N

2 1 *
280 kg/ha N 280 kg/ha N



224 kg/ha N 112 kg/ha N



112 kg/ha N 168 kg/ha N



168 kg/ha N 0 kg/ha N



0 kg/ha N 224 kg/ha N


Block 4

3 1 *
280 kg/ha N 224 kg/ha N



168 kg/ha N 112 kg/ha N



0 kg/ha N 168 kg/ha N



112 kg/ha N 280 kg/ha N



224 kg/ha N 0 kg/ha N

4 2
112 kg/ha N 0 kg/ha N



280 kg/ha N 168 kg/ha N



0 kg/ha N 224 kg/ha N



224 kg/ha N 112 kg/ha N



168 kg/ha N 280 kg/ha N


Irrigation furrow



Plots used for the root study









resulting in a density of 392,000 plants-ha-1. Cowpea and sorghum were planted in June

after potato harvest in each production year as summer cover-crops (SCC). Bush green

bean (Phaseolus vulgare, var. Espada) was planted in September in a single drill per row.

with in-row spacing of 2.5 cm thus resulting in a density of 392,000 plants-ha-1. Cowpea

and bush green bean were inoculated with specific Rhizobium strains. Plots not planted

with bush green bean in fall were maintained as a weed-free fallow

Pest management

All ground was fumigated with 1, 3-dichloropropene (56 L-ha-1) 4 weeks before

planting potatoes for nematode control.

Irrigation management

The seepage irrigation system uses high row beds to plant potatoes and shallow

water furrows to supply irrigation water and to remove drainage water. The natural high

water table combined with the sandy and flat nature of the soils and landscape in this area

make the irrigation and drainage by the subsurface seepage method possible. The rows

were raised to 40 cm in height with a between-row spacing of 102 cm. Each bed

consisted of 16 rows separated by irrigation-drainage furrows. The water furrow was

about 20 cm deeper than the alley between rows in the bed. During dry periods water was

added continuously to the head of the water furrow to maintain the water table 60 cm

beneath the top of the bed.

Nutrient management

Each year at potato planting, 112 kg-ha-1 N as ammonium nitrate (NH4N03), was

banded and incorporated in the plots receiving N. However, the control plot was not

fertilized with any source of N. Two weeks after emergence of potato plants, the

remaining N rates of 56, 112, and 168 kg-ha-1 were side-dressed using the same NH4N03









at hilling to complete the final treatmental dosages of 0, 112, 168, 224, and 280 kg-ha-1

of N respectively (Table 2-1). Recommended N rate for potatoes in the region is 224

kg.ha-1 (Hochmuth and Hanlon, 2000). Based on pre-plant soil tests 34 and 168 kg-ha-1 of

P and K, respectively, were incorporated I all the plots prior to potato planting in each

production year (Table 2-1). No fertilizers or nutrients were applied at any time to either

cowpea or sorghum crops. In each production year, around mid August both sorghum and

cowpea (before the cowpea set seed) were chopped and incorporated into the soil as

green manure. Bush green beans were fertilized pre-plant and a month after planting with

400 kg-ha-1 of 14-3-10 and 600 kg-ha-1 of 14-0-10 fertilizer blends respectively. Total N

applied to bush green bean through the split application amounted to 140 kg-ha-1.

Table 2-1. Nitrogen rate treatments and nutrients applied to potato crop each year.
N (kg.ha-1) P (kg.ha-1) K (kg.ha-1)
At planting Side-dress Total At planting
0 0 0
112 0 112
112 56 168 56 168
112 112 224
112 168 280


Water and Soil Sampling Schedule

Soil sampling

Soil samples were taken with an open side tubular auger (diam. 1.9 cm) to a depth

of 20 cm. A composite soil sample from eight randomly selected points located in the two

central rows of the plots was taken at each soil sampling time. A margin of at least 1 m

was left at both ends of each sampled row in order to avoid border effects such as a rare

sample contamination with soil or fertilizer movement between adjoining plots. Samples

were collected in labeled plastic bags, air-dried, and sieved in preparation for subsequent









laboratory analyses. Soil samples were tested for N03-N, NH4-N, P, K, Ca, and Mg at the

UF/IFAS Analytical Research Laboratory (ARL) in Gainesville per the procedures

manual (Mylavarapu and Kennelley, 2002). Both N03-N and NH4-N in solution were

analyzed in an air-segmented continuous autoflow analyzer. Mehlich-1 extraction

procedure also referred to as the "dilute double acid" extractant (0.05M HCI + 0.0125M

H2S04) was used to extract soil samples and determine P, K, Ca, and Mg by the ICP

method EPA 200.7. Results from analyses of only soil N03-N and NH4-N concentrations

analyses were interpreted and reported in this chapter as other nutrients will be outside

the scope of this manuscript. The soil sampling schedule throughout the 3-year period is

described in Fig. 2-2.

Water sampling

A 10 cm diameter PVC pipe was buried in each potato plot to a depth of 70 cm

from the top of each ridge. The pipe acted as a well casing providing easy access to

sample the water table below each plot. Additionally, a lysimeter was buried at 30 cm

depth on one of the central ridges to sample water from the soil solution at the root zone

during the potato season. During the first year (2001) water was sampled from wells and

lysimeters on weekly basis, and on a bi-weekly basis during the second (2002) and third

(2003) year (Fig. 2-2). In summer 2003, sampling wells were installed to sample the

water table under cowpea and sorghum in the plots where potato was fertilized with 0,

168, and 280 kg.ha&1 ofN. Three bi-weekly samplings were taken following the same

procedure for water sampling during the potato season. At each sampling time water

samples from wells and lysimeters were collected in plastic vials and frozen immediately

to measure N03-N and NH4-N concentrations.






































2002




Potato






II


Sorghum& Cowpea

T^1----


2003
?0 -------- -------- -------J-------J--------- ^-------J- -------- ------J--j-------- --------I -------- --------
Side-dress N

Potato Sorghum &Cowpea Bush green bean Water (well)

1o ---- L -- -- -- -- - -- I------ -1Water --y----
Soil
Tissue
S.1 Root sampling
0 -1, A


Figure 2-2. Crop cycles, sampling schedules, and precipitation pattern over the three-year period of study


10


0


j:: 0


0


Bush green bean


- ____ __ Y _________









In the first year and second years wells were installed 3 and 8 days before N side-

dressing respectively, and 2 days after side-dressing in the third year. Lysimeters were

installed at the end of the potato season (early May), and 1 week before N side-dressing

in the second and third years. Due to the late installation of the lysimeters in the first year

(early May) and problems with recovery of the water samples from lysimeters during the

third year, only samples from the second year were analyzed and included in this

dissertation. The wells and lysimeters sampling schedule is illustrated in Fig. 2-2. Water

samples from wells and lysimeters were tested at the UF/IFAS ARL in Gainesville.

Water samples were tested in an air-segmented continuous autoflow analyzer for N03-N

and NH4-N (Mylavarapu and Kennelley, 2002). Only N03-N concentrations are

analyzed, reported and discussed in this dissertation.

Tissue Sampling Schedule

Tissue samples from the crops were tested for Total Kjeldahl Nitrogen (TKN), P,

and K at the UF/IFAS ARL in Gainesville. Digestates from the Kjeldahl digestion

process were tested using semi-automated colorimetric analysis (EPA Method 351.2) to

determine nitrogen. The instrument (Alpkem autoanalyzer) was set up and calibrated.

Calibration standards and quality control samples were digested in the same manner as

the samples as per the procedures described in manual (Mylavarapu and Kennelley,

2002).

Potato tissue

Potato tissue was sampled at full flowering 60 to 65 days after planting (DAP) in

each of the three-years (Fig. 2-2). Shoots of two plants from the two central rows in each

plot were cut at ground level and divided in leaves and stems, oven-dried, weighed,

ground, and individually labeled. At harvest time a sample of marketable potato tubers









from each plot was taken. Tubers were diced, fresh weighed, oven-dried, dry weighed,

ground, and individually labeled. Leaves, stems, and tuber samples were sent to the

UF/IFAS ARL in Gainesville for analysis. Pooled TKN concentrations of leaves and

stems were analyzed and reported as shoots TKN.

Sorghum and cowpea tissue

Sorghum and cowpea plants were chopped and incorporated into the soil in August

(Fig. 2-2). Before incorporating sorghum and cowpea, a tissue sample of each to estimate

biomass and N added to the soil was taken following the same procedure described for

potato plant sampling. Ground tissue samples were sent to the UF/IFAS ARL in

Gainesville.

Potato yield

In all potato plots, two-row segments 6.1 m long from the central rows were

harvested with a mechanical harvester 100 DAP. A margin of at least 1 m was left at both

ends of each sampled row in order to avoid the border effect of adjoining plots. Potato

tubers were graded by size (tuber diameter) into five classes (B= <4.8, Al= 4.8-6.4, A2 =

6.4-8.3, A3 = 8.3-10.2, A4 = >10.2 cm). Marketable yield was comprised of classes Al to

A3 only. Total yield was calculated by the summation of all of the classes. Tubers were

rated for general appearance, disease, and nematode damage. In this dissertation, total

and marketable yield, specific gravity, and % TKN in the tubers were analyzed and

reported. Agronomic or physiological interpretation of the yield results was not

attempted.









Statistical Analysis

Concentrations of N03-N in the water table

The assumptions of homogeneity of variance and normality of the population were

checked for the water data sets (wells and lysimeters). Normality was checked by residual

analysis using the Shapiro-Wilk test as implemented in the CAPABILITY procedure of

SAS (SAS Institute, 2004). Concentration of NO3-N in water was log transformed and

checked again for normality. An exploratory analysis consisting of a time series plot to

see the general pattern of NO3-N concentration in water, and a general linear model

analysis on the log transformed data set were performed. Using orthogonal polynomial

contrasts the N rate factor on the subplot was partitioned into linear, quadratic, cubic, and

quartic components to determine which regression best described N03-N concentrations

in water with-time. Subsequently, the RSREG procedure of SAS (SAS institute, 2004)

was used to estimate the regression parameters and predicted values. The number of days

after or before N side-dressing were used as the independent variables.

A general linear model by individual sampling times was performed to check the

effect of residual fertilization on the N03-N concentration in the water table under

sorghum and cowpea planted in summer 2003. The analysis was implemented as a split

plot design where crops cowpeaa and sorghum) were the main plot factor and residual

effect from the five nitrogen rates applied to potato were the sub-plot factor. Means were

separated by the Tukey test as implemented in the GLM procedure of SAS (SAS

Institute, 2004) and confidence limits of 95% were calculated for each mean.

Concentrations of N03-N and NH4-N in the soil

The soil sampling schedule was not the same for each year (Fig. 2-2) therefore a

general linear model by individual soil sampling times was performed and therefore was









not combined over years. The Tukey multiple comparison procedure was used to separate

rotation and N treatment means when overall significant effects were observed. The N

treatment effect was decomposed into linear, quadratic, cubic and quartic components

using orthogonal polynomials in order to identify the response of soil N03-N to the five

N rates applied to potato. Soil N03-N content at the time of the previous soil sampling

was used as a covariate to check the effect of past N03-N concentration on selected

sampling times. Since N treatment had either linear or quadratic effect on soil N03-N and

NH4-N concentrations, regression analysis was used to test hypothesis and estimate

regression parameters. A general linear model combining soil samplings at potato harvest

was used to compare residual N03-N before SCC planting.

Nitrogen uptake by the crops

Statistical analysis of N as TKN in the biomass of the summer cover crops at the

time of incorporation into the soil was done to estimate N uptake and N provided by the

soil incorporation of each of the crops. The assumptions of homogeneity of variance and

normality of the population were checked. Normality was checked by residual analysis

using the Shapiro-Wilk test as implemented in the CAPABILITY procedure of SAS

(SAS Institute, 2004). A general linear model by individual years was used to determine

the effect of crop rotation and N rate on N uptake by the SCC. Using orthogonal

polynomial contrasts the N rate factor on the subplot was partitioned in linear, quadratic,

cubic, and quartic components to determine the regression that best described total N

uptake of the cover crops over-time. Cowpea did not respond to residual N from the

potato fertilization therefore no regression was adjusted.

Similarly, homogeneity of variance and normality of residues were checked prior to

performing the statistical analysis for N uptake values for potatoes. A general linear









model by individual years on potato N uptake at full flowering was performed. N rate

factor was partitioned in its linear, quadratic, cubic, and quartic components by using

orthogonal polynomial contrasts in order to determine the best regression to describe the

effect of N rate on potato N uptake. Finally, the REG procedure as implemented in SAS

(SAS Institute, 2004) was used to estimate regression parameters (Appendix A).

Potato yield

The assumptions of homogeneity of variance and normality of the population were

individually checked for potato size classes. A correlation analysis including all of the

variables was performed. Normality was checked by residual analysis using the Shapiro-

Wilk test as implemented in the CAPABILITY procedure of SAS (SAS Institute, 2004).

A general linear model by individual years was used to study yield response to crop

rotation and N rate. The LSVIMEANS with Tukey adjustment as implemented in SAS

(SAS institute, 2004) was used to separate individual factor means and/or interaction

means when they resulted significant. Subsequently, a general linear model decomposing

N rate in its linear, quadratic, cubic, and quartic components was used to determine the

best regression explaining the response of marketable potato yield to N rate. The

quadratic model explained the response of yield to fertilizer rates. The first derivative of

the quadratic regression equation was solved for N rate to determine the N rate required

to reach maximum marketable yield.

Results and Discussion

Concentrations of N03-N in the Water Table under Potato

The time of application of side-dressed ammonium nitrate, approximately 40 DAP

(days after planting) was used as the reference point to describe the effect of the five rates

of applied fertilizer on the N03-N concentration in the water table (Figs. 2-3a to 2-5b).






















S28 R =0.77
6


2
Z

O
Z
05




01

0 05

0025


-10 0 10 20 30 40 50 60 7
150-
100- N rate (kg.ha-1) o R2 =0.81 PC(

50: 112 R2 =0.66

168 R2 =0.80
20
"-- 224 R2 =0.77

10 280 R2 =0.67
j 6
cJ) 4

E -



2-
z


05





0 05-

0 025

-10 0 10 20 30 40 50 60

Days since side-dress

Figure 2-3a. Predicted N03-N concentration in the water table under two crop rotations
and five N rates during the first year. PCF=Potato-Cowpea-Fallow,
PCG=Potato-Cowpea-Green bean













150-
100oo- N rate (kg.ha-1)- 0 R2 =0.76 PSF

50 112 R2 =0.74

-- 168 R2 =0.81
20

10-
8 280 R2 =0.78
S6



E 12
z

05-




01-

005-

0 025-


-10 0 10 20 30 40 50 60 70
150
100- N rate (kg.ha-1) 0 R2 =0.64 PSG

50 112 R2 =0.87

168 R2 =0.79
20
2"20" 224 R2 =0.84
10:
8: 1 280 R2 =0.81
6


Z 2

0 1
Z
05




01

0 05-

0 025-


-10 0 10 20 30 40 50 60 70

Days since side-dress

Figure 2-3b. Predicted N03-N concentration in the water table under two crop rotations
and five N rates during the first year. PSF=Potato-Sorghum-Fallow,
PSG=Potato-Sorhum-Green bean









The concentration of NO3-N in the water table was explained by linear and

quadratic regressions (Appendix A). During the 3 year study period, N03-N

concentration in the water table under the four rotations exhibited decreasing trend from

high values before N side-dressing to values close to zero two months after N side-

dressing(Fig. 2-3 to 2-5).

In the first year, at the time of the first water sampling (3 days before side-dressing)

the predicted values for fertilized plots were in the range of 13.5 to 54.5 mg-L- of NO3-

N. The predicted values for the unfertilized plots were between 8.2 and 18.2 mg-L-1 of

N03-N. In the PSF rotation the predicted value for the control plot was higher than the

value for the 112 kg-ha-1 N rate suggesting a possible cross-contamination of the control

plot from the adjoining plots. The declining trend was explained by a significant linear

regression model with coefficients of determination ranging between 0.64 and 0.87 (Fig.

3-2). N03-N in the water table under plots with the rotations PCF and PSG with 280

kg-ha-1 of N, and plots with the rotation PSF and 112 kg-ha-1 of N were described by a

quadratic model. In spite of these exceptions, an overall similar pattern was observed for

each one of the 4 rotations studied. No differences between N treatments including the

control were observed (Fig. 2-3a, b).

In the second year, coefficients of determination ranged from 0.57 to 0.32 (Fig 2-

4a, b). The predicted values at the first sampling (8 days before N side-dressing) were in

a range of 0.3 to 5.0 mg-L-1 of NO3-N in the water table. No differences between

fertilized and unfertilized plots were observed. The relative low concentrations ofNO3-N

in the water table in this year produced tendency for lower values of R2 and low

adjustment of the models suggesting high variability when N03-N concentration values





















L 6






05-






S05-

0025-


-10 0 10 20 30 40 50 60 70
150-
100- N rate (kg.ha-1) 0 R2 =0.37 PCG

50: 112 R2 =0.24

168 R2 =0.50
20
S224 R2 =0.30
10:
8 : 280 R2 =0.25
6






05




01-

0 05

0 025


-10 0 10 20 30 40 50 60 70

Days since side-dress
Figure 2-4a. Predicted N03-N concentration in the water table under four crop rotations
and five N rates during the second year. PCF=Potato-Cowpea-Fallow,
PCG=Potato-Cowpea-Green bean




















_____ LOU rM- -U.L/
L 6

E

O
z

0.5




0.1-

0.05-

0.025-


-10 0 10 20 30 40 50 60 70
150
100- N rate (kg.ha-1)- 0 R2 =0.15 PSG

50: 112 R2 =0.53

S 168 R2 =0.33
20
--224 R2 =0.10
10
8: 280 R2 =0.20
S6
Cj) 4










0.1 -
2-










0.05

0.025-


-10 0 10 20 30 40 50 60 70

Days since side-dress

Figure 2-4b. Predicted N03-N concentration in the water table under two crop rotations
and five N rates during the second year. PSF=Potato-Sorghum-Fallow,
PSG=Potato-Sorghum-Green bean









were close to the analytical method detection limit (0.5 mg-L-1 of NO3-N). ). Even

though, the low coefficients of determination a similar overall pattern was observed in the

water table N03-N concentration under the four crop rotations (Fig. 2-4a, b). A probable

explanation for the low N03-N concentration values was the low rainfall in the period

between potato planting and N side-dressing in 2002 compared with the high rainfall in

2001 and 2003 (Table 2-2).

In the third year, water samples from wells were taken only from the plots fertilized

with 0, 168, and 280 kg-ha-1 ofN. As in the first year, the N03-N concentration in the

fertilized plots followed a decreasing pattern. The predicted values for fertilized

plots in the first sampling (2 days after side-dressing of N) were between 11.0 and 54.6

mg-L-1 ofNO3-N (Fig. 2-5). These values are similar to the predicted values at the first

sampling during the first year. The only difference being sampling time which was 3 days

before side-dressing in the first year (Fig. 2-3). Predicted N03-N concentration values in

the water table under fertilized plots treated with different fertilizer rates tended to be

similar and unfertilized plots showed low N03-N concentration values (0.8 to 2.0 mg.L'1)

except in PCG rotation (Fig. 2-5a) where the effect of the two legumes in rotation with

potato could explain the relatively high N03-N concentration in the water table under the

control plot for this rotation. Similarly to the second year, the relatively low N03-N

concentration in the water table under unfertilized plots produced low R2 and low

adjustment to the regression model (Fig. 2-5a, b). The main difference between the first

and third year is that the predicted values for the unfertilized plots in the first year were

higher compared with the low values for the same plots in the third year. The difference











150-
100- N rate (kg.ha-1) 0 R2 =0.22 PCF

50; 168 R2 =0.44
280 R2 =0.60
20

10
8
6

2-

z

O 1
0
0.5




0.1-

0.05-

0.025


-10 0 10 20 30 40 50 60 70
150
100 N rate (kg.ha-)- o R2 =0.82 PCG

50 168 R2 =0.54
280 R2 =0.81
20

10
8
6
4
E
2
z

O
Z
0.5




0.1

0.05

0.02


-10 0 10 20 30 40 50 60 70

Days since side-dress


Figure 2-5a. Predicted N03-N concentration in the water table under two crop rotations
and five N rates during the third year. PCF=Potato-Cowpea-Fallow,
PCG=Potato-Cowpea-Green bean













N rate (kg.ha-1) 0
--168
280


PSF


R2 =0.18
R2 =0.47
R2 =0.55


-10 0 10 20 30 40 50 60 70
150-
100- N rate (kg.ha-1) 0 R2 =0.37 PSG

50o -- 168 R2 =0.84
S 280 R2 =0.83
20-

7 8:
J 6:

20



05-



01



0 025-

-10 0 10 20 30 40 50 60 70
Days since side-dress

Figure 2-5b. Predicted N03-N concentration in the water table under two crop rotations
and five N rates during the third year. PSF=Potato-Sorghum-Fallow,
PSG=Potato-Sorghum-Green bean









among unfertilized plots in the first and third year was due to high residual effect of N in

the field from the past season.

In each year, a declining pattern in N03-N concentration starting with higher values

prior to side-dressed application of N was a common observation under all rotations and

N rates. This was evidently due to the application of 112 kg ha-i of N through readily

soluble ammonium nitrate to all the potato plots that received N treatments at planting

when the root development is limited leading to inefficient fertilizer uptake and higher

N03-N concentration in the water table. This hypothesis is supported by the steady

decline in N03-N concentration after side-dress fertilization through the potato season in

the fertilized plots. In the 3-year period N03-N concentration in the water table under the

four different crop rotations exhibited a similar pattern within season but different trends

were observed during different seasons. In the second year, N03-N concentrations in the

water table between planting and side-dressing N fertilizations under all rotations were

lower than 5 mg-L- compared with the same period on the first and third years when the

highest concentrations were close to 50 mg-L-1 of NO3-N (Figs 2-3 and 2-5). High

amounts of rainfall received during those years (Table 2-2) have resulted in higher

leaching ofNO3-N. During the first and third years, precipitation received during the

month of planting was 47% and 51% higher respectively than the historic 25-yr average

(NOAA, 2004) with occurrence of leaching rain Kidder et al. 1992) before the first

Table 2-2. Precipitation during crops and intercrops periods during the study period.
Potato Crops in rotation with potato
Year Planting to N side-dress Inter Sorghum/ Inter Green Inter
N side-dress to harvest crop Cowpea crop bean crop
Rainfall (cm)
2001 15.98 5.86 16.32 27.37 5.97 54.24 5.87
2002 8.24 7.56 1.79 30.05 13.10 23.81 26.55
2003 22.30 11.18 12.19 38.75 14.48 22.46 6.32









water sampling. The second year was not a typical year with a received precipitation 45%

below the average. Therefore, less NOs3 could have leached from the potato beds in the

second year compared to the first and third years.

Concentrations of N03-N in the Soil Solution

The concentration of N03-N in the soil solution under four crop rotations and five

N rates during the second year (Fig. 2-6a. b), was explained by the quadratic regression

model with coefficients of determination ranging between 0.47 and 0.89 (Fig. 2-6a,b) .

Eight days before N side-dressing, predicted values fluctuating between 156.0 and 63.4

mg-1-1 of NO3-N were estimated for fertilized plots planted with PCG (Fig. 2-6a) and PSG

(Fig. 2-6b) rotations, and predicted values for the non-fertilized plots planted under the

same rotations were 40.4 and 25.0 mg-1-1 of NO3-N respectively. In the plots with crop

rotations including fall as a fallow period predicted values for fertilized plots were

between 27.1 and 85.6 mg-1-1 of N03-N. Predicted values for non-fertilized plots with the

same rotations were around 16.0 mg-L- of NO3-N (Fig. 2-6a, b). Also, a comparison of

predicted N03-N concentrations at first sampling in PSF and PSG under fertilization

showed that the concentration in plots with green bean in fall was twice as the

concentration in fallow plots. A similar trend was observed when PCF and PCG were

compared. Factors such as residual N from green bean fertilization, increase in the soil N

due to the biological N fixation and/or mineralization of crop residues from green bean

could have resulted in such a difference in the N03-N concentration in the soil solution.

An elevation in N03-N concentration in the soil solution was detected between 40

to 60 days after N side-dress (Fig. 2-6a, b). A similar elevation in N03-N concentration

in water samples from the water table beginning 30 days after N side-dress (Fig. 2-6a, b)

was detected. These elevations in N03-N concentration in the soil solution and water











300- N rate (kg.ha-1) 0 R2 =0.78 PUb

150- -- 112 R2 =0.72
100- 168 R2 =0.69

50: ---- 224 R2 =0.47

280 R2 =0.76
J 20 -
cJ)
E 10:
z 8
6-
O 4-
z
2-

1-

0.5


0.2-


-10 0 10 20 30 40 50 60 70

300- N rate (kg.ha-1) 0 R2 =0.82 PCG

150- 112 R2 =0.72
100- 1 168 R2 =0.64

50: 2 224 R2 =0.77

280 R2 =0.59
20

10:
Z 8:

O 4-
Z
2-



0.5-


0.2-


-10 0 10 20 30 40 50 60 70
Days since side-dress


Figure 2-6a. Predicted N03-N concentration in the soil solution under two crop rotations
and five N rates during the second year. PCF=Potato-Cowpea-Fallow,
PCG=Potato-Cowpea-Green bean







32





300- N rate (kg.ha-1) 0 R2 =0.89 PSF

150- 112 R2 =0.68
100- 1 168 R2 =0.87

50 ---- 224 R2 =0.70
S 280 R2 =0.81
j 20



O -

2-



05


02-


-10 0 10 20 30 40 50 60 70
300- N rate (kg.ha-1) 0 R2 =0.79 PSG

150- 112 R2 =0.63
100- 168 R2 =0.58

50: -\ 224 R2 =0.55
S280 R2 =0.49
S20 -
E
S10

O 4-
Z






02-


-10 0 10 20 30 40 50 60 70
Days since side-dress


Figure 2-6b. Predicted N03-N concentration in the soil solution under two crop rotations
and five N rates during the second year. PSF=Potato-Sorghum-Fallow,
PSG=Potato-Sorghum-Green bean










table close to potato harvest coincide with the suggestion of increased nitrification of the

NH4F applied at N side-dress fertilization (Fig. 2-8 and 2-9). The coincidence in the

elevation of N03-N concentration in the soil solution at 30 cm from the top of the ridge

with the elevation in N03-N concentration in the water table 40 cm below of the

lysimeter cup suggest leaching of NO3-N.

Concentrations of N03-N in the Water Table under Cowpea and Sorghum

In summer 2003, three bi-weekly samplings from the water table under plots

planted with cowpea and green bean were taken in order to study the residual effect of the



40
I 0 kg.ha-1 N
35-
S168 kg.ha-1 N

30 I 280 kg.ha-1 N

25

20

0 155-

10





Cowpea Sorghum Cowpea Sorghum Cowpea Sorghum
First sampling Second sampling Third sampling

Figure 2-7. N03-N concentrations in the water table under sorghum and cowpea. Vertical
whiskers are confidence intervals at 0.95 level.

fertilizer applied to potato in spring. At first sampling N03-N concentrations in the water

table under cowpea showed a significant increase when the N rate applied to potato in

spring was increased from 168 to 280 kg.ha-1 and there was no difference between plots









fertilized with 0 and 168 kg.ha-1 ofN (Fig. 2-7). Also at the first sampling there was no

difference in the N03-N concentrations in the water table under plots planted with

sorghum. The N03-N concentrations in water table under sorghum were significantly

lower than the N03-N concentrations under cowpea in plots fertilized with 280 kg.ha-1 of

N but there was no difference with cowpea planted in plots where potato was no-

fertilized or fertilized with 168 kg.ha-1 of N (Fig. 2-7). These results showed that

sorghum was a better N03-N catch crop compared to cowpea. Including sorghum as a

summer cover crop in the potato production rotation was found to have a significant

potential in reducing applied N rates, improving N efficiency and minimizing leaching.

Concentrations of N03-N and NH4-N in the Soil

In 2001 and 2003, 3 weeks after side-dress N was applied soil N03-N concentration

across N rates showed a positive relationship described by linear and quadratic

regressions respectively (Fig 2-8a) Twelve weeks later, at planting of sorghum and

cowpea a corresponding trend was detected (Fig 2-8b). The persistence of the N fertilizer

in the soil indicated N over-fertilization to potato. Similarly to 2003, at seven weeks after

N side-dressing soil N03-N concentration increased linearly across N rates (Fig 2-8b)

Ammonium-N concentration 3 weeks after N side-dressing exhibited a positive

relationship described by a quadratic regression in both years (Fig. 2-9a). Prior to side-

dressing, concentrations of NH4-N in unfertilized plots (0 N) and plots with fertilization

only at planting (112 kg.ha-1 N rate) showed similar values possibly due to both high

nitrification rate and/or leaching and runoff due to the high amount of rainfall after potato

planting in 2001 and 2003 (Table 2-2). Predicted soil NH4-N concentrations at planting of

sorghum and cowpea (12 weeks after N side-dressing) in 2001 and 2002, and 7 weeks

after N side-dress in 2003 (Fig. 2-9b) decreased to similar concentration values observed











70

0 2001 R2=0.53** A
60--


50 2003 R2=0.57**


E 40
z
'" 30
z

20


10


0-

0 112 168 224 280







2001 R2=0.51** B
60
2002 R2=0.26**

50 -- 2003 R2=0.17**


C 40

Z
M 30-


O 20
UO

10


0

0 112 168 224 280


N rate (kg.ha -1)



Figure 2-8. Predicted N03-N concentration in the soil by effect of five N rates. A- Three
weeks after N side-dressing. B-At sorghum and cowpea planting in 2001 and
2002, and three weeks before potato harvest in 2003. Coefficient of
determination significance: p <0.05 and **p<0.01















2001 R2=0.64- A

2003 R2=0.55"














5-


0-

0 112 168 224 280






S- 2001 R2=0.08* B

2002 R2=0.25**

2003 R2=0.13**











5


0-


3 112


168


224


280


N rate (kg.ha1)



Figure 2-9. Predicted NH4-N concentration in the soil by effect of five N rates. A-Three
weeks after N side-dress. B-At sorghum and cowpea planting in 2001 and
2002, and 3 weeks before potato harvest in 2003 Coefficient of
determination significance: p <0.05 and **p<0.01.









in unfertilized plots 3 weeks after side-dressing (Fig. 2-9a). In 2003, the decrease in soil

NH4-N concentration between N side-dressing (Fig. 2-9a) and potato harvest (Fig. 2-9b),

during a relatively low rainfall period (Fig. 2-2), along with increased N03-N in the

seventh week after N side-dressing in 2003 (Fig. 2-8b), showed that increase in N03-N

concentration at the end of the potato season was caused by nitrification of the NH4F

contained in the ammonium nitrate side-dressed to potato. The increase in N03-N

concentration in the soil solution at 30 cm from the top of the ridge between 40 and 60

days after N side-dressing in 2002 (Fig. 2-6a, b) supports the suggestion of increased

nitrification before potato harvest. The accumulation of NO3-N in the potato beds at the

end of the potato cycle could be explained by the relatively low rainfall between time of

side-dress fertilization and harvest of potato compared with the high rainfall between

potato planting and side-dress N application (Table 2-2). Furthermore, as seepage

irrigation relies on upward capillary movement of water instead of the downward

movement of the water under conventional overhead irrigation NOs3 accumulated in the

relatively dry soil on the surface potato beds. Also aerobic conditions on top of the potato

beds could result in an increased nitrification favoring build up of soil N03-N.

In 2003, the soil sampling at 3 weeks prior potato harvest revealed increased NO3-

N concentration along with the increase in N rate (Fig. 2-8b). Three months later, at

sorghum and cowpea incorporation, soil N03-N concentration dropped close to 0.5

mg-kg-1 and no statistical differences were detected among crop rotation or N treatments.

Most of the residual N03-N detected before potato harvest was evidently captured up by

sorghum as was found in its increased N uptake values (Fig. 2-10) and/or leached to the

water table from cowpea plots as depicted in Fig 2-7.









In 2001 at green bean harvest, there was no difference in soil N03-N concentration

among crop rotations (Table 2-3). Two months later, at potato planting 2002, soil N03-N

concentration in plots planted with green bean in fall was significantly higher than plots

left as fallow (Table 2-3). Soil N03-N concentration in the rotation with cowpea in

summer and fallow in fall (PCF) was significantly higher than soil N03-N concentration

Table 2-3. Effect of crop rotation on soil N03-N concentration at potato planting
G. bean harvest Potato planting G. bean harvest Potato
Crop planting
Rotation 11-29-01 02-04-02 11-15-02 02-03-03
mg.kg-1
PCG 3.90 a 10.58 a 5.11 a 1.94 a
PCF 3.00 a 5.29 b 3.90 a 1.30 c
PSG 2.96 a 9.91 a 2.86 a 1.56 b
PSF 2.55 a 3.99 c 2.12 a 0.95 d
Means in the same column followed by different letter are statistically different (p<0.01)


in plots with sorghum in summer and left as fallow in fall (PSF) suggesting

increased soil N03-N concentration due to mineralization and nitrification of cowpea

residues incorporated in summer into the soil (Table 2-3). Cowpea residues have an

average C/N ratio of 15:1 (Havlin et al. 1999) which should enhance N mineralization. In

summer 2001, soil incorporation of sorghum in PSF rotation could result in an initial

immobilization of N due to its high C:N ratio (60:1) (Havlin et al. 1999). Six months later

at potato planting in 2002, soil N03-N concentration in PSF plots increased suggesting

late mineralization of sorghum residues incorporated in summer 2001 (Table 2-3).

In 2002 at green bean harvest, similarly to 2001, there were no statistical

differences in soil N03-N concentration (Table 2-3). However, in a couple of months at

potato planting 2003, a decreased soil N03-N concentration was observed (Table 2-3).

Declining in soil N03-N concentration at potato planting in 2003 was due to rainfall









during the intercrop period at the end of 2002 which was five times higher than the

rainfall during the similar intercrop period at the end of 2001 (Table 2-2). These results

show that the beneficial effect of enhanced N availability by rotation with legumes easily

can be eliminated in sandy soils by effect of rainy periods. Therefore, the increase in N

available to potato grown in the following spring season from the rotation with green

bean should be taken in account to calculate the fertilization rate for the following potato

crop but the amount of rainfall should be considered also.

Nitrogen Uptake by the Crops

Sorghum

Rotation did not significantly affect N uptake by sorghum. Significant linear

response in the first year and quadratic in the second and third years ofN uptake to

residual soil NO3-N from the five N rates applied to potato 70 days before summer cover-

crops (SCC) planting were observed (Fig 2-10). Overall decreasing sorghum N uptake

during the 3-year period was observed (Fig. 2-10). An initial data exploratory study

showed sorghum biomass highly correlated (0.91, p<0.0001) with sorghum N uptake,

therefore, vegetative growth determined its N uptake. Decreasing N uptake by sorghum

in spite of the trend to increased residual NO3-N across N rates through the study period

suggests that growth of sorghum was limited by other factors, possibly in combination.

Differences in sorghum N uptake among years could be explained in part by the

combined effect of different length of the period between planting and soil incorporation

and the amount of residual N from potato fertilization. In the first, second, and third years

sorghum stayed in the field for 57, 77 and 48 days, respectively. These results reveal

sorghum as an efficient NO3-N catch-crop and showed that sorghum grown for an

extended period in the field leaving only a short time between potato harvest and







40



120
2001 R2=0.36*

100 2002 R2=0.36*
80 2003 R2=0.42*




c 60


40


20



0 112 168 224 280

N rate (kg.ha-1)


Figure 2-10. Nitrogen uptake of sorghum by residual effect of the fertilizer applied to
potato. Coefficient of determination significance: p <0.05 and **p<0.01.

sorghum planting as feasible measures to reduce potential leaching of residual N03-N

from potato fertilization. The effect of a sorghum crop as an efficient N scavenger will be

particularly apparent during the wet summer season (Fig. 2-2) as was showed by the

significant less N03-N leaching under sorghum plots compared with cowpea plots (Fig.

2-7).

Cowpea

Crop rotation and residual N from potato fertilization had no significant effect on N

uptake of cowpea. N uptake of cowpea in control plots (0 N) was similar to N uptake in

plots with residual N from fertilization to potato in spring (Table 2-4). This response was

expected from a leguminous plant like cowpea, which can meet its N requirement

through biological fixation with minimal dependence on soil N availability. Low demand

for soil N by cowpea explains the occurrence of the second flux of N03-N to the water









Table 2-4. Nitrogen uptake of cowpea by year at incorporation time.
N rate (kg-ha1) (kg-ha-1)
2001 2002 2003
0 213a 225a 59 a
112 204 a 214 a 66 a
168 225 a 163 a 68 a
224 199a 221a 68 a
280 191a 186a 61 a
Means followed by the same letter are not statistically different.


table detected after potato harvest in 2003 (Fig. 2-7) when the summer rainfall peak

occurred (Fig. 2-2). Nitrogen uptake in 2003 was significantly reduced compared to the

corresponding values in 2001 and 2002. However, no difference among N rates persisted

(Table 2-4) suggesting that there was another limiting factor at play other than N fixation.

Similar to the reduced N uptake observed in sorghum (Fig. 2-10) and potato (Fig. 2-11)

in 2003 compared with 2001 and 2002, the reduced N uptake of cowpea suggests the

occurrence of a growth limiting factor affecting any crop.

Potato

Nitrogen rate affected significantly N uptake by potato shoots at full flowering all

the three years (Table 2-5). Effect of crop rotation on N uptake of potato was not

significant in any of the 3 years suggesting a short term effect of crop rotation (Table 2-

Table 2-5. ANOVA of the annual potato N uptake over the 3 year period


Source of Type III Mean Square2
variation
variation DF 2001 DF 2002 DF 2003
Replication1 2 22179.39* 2 3861.55 3 5004.21
Rotation1 3 1610.55 3 6456.90 3 7750.47
Main Plot Error 6 3312.78 6 2250.81 6 3818.49
N rate 4 21837.80** 4 24594.45** 4 44526.97**
Rotation*N rate 12 3509.94 12 1310.37 12 1723.61
Error 59 3878.20 59 1466.22 79 1311.33


Main plot effects tested using main plot error.
2 *=significant at type I error <.05. **=Significant at type I error <0.01










5). In 2001, potato N uptake was described by a linear regression and was higher than N

uptake in 2002 and 2003 (Fig. 2-11). In 2002 and 2003 potato crop N uptake was

described for quadratic and linear regressions respectively, with similar overall N uptake

(Fig. 2-11). The linear increase in N uptake across N rates 2001 and 2003 (Fig. 2-11)

without response in marketable yield (Fig. 2-12) confirmed luxury consumption ofN by

potato after it has reach its maximum yield.



250






150



100

S- 2001 R2=0.23**
50 2002 R2=0.45**

2003 R2=0.51**

0 112 168 224 280

N rate (kg.ha-1)

Figure 2-11. Potato N uptake at full flowering during the study period. Coefficient of
determination significance: p <0.05 and **p<0.01

Potato Yield

An initial exploratory analysis showed that marketable yield in the N fertilized

treatments were highly and significantly correlated with yield of potato sizes A2 and A3.

Therefore, these two sizes were the main components of the potato yield. In 2001, N rate

affected significantly most of the yield components (Appendix B). In 2002 and 2003

there was a significant interaction between crop rotation and N rate (Appendix B). Yields







43



35
R2=0.38** 2001
30

25-

20

15- NYmax=1 64

10-

5

0
0 112 168 224 280
35
PCF R2=0.81** 2002
,7 30 PCG R2=0.48**

M- PSF R2=0.72**
S25
2 PSG R2=0.91"
20

15

S10 NYmax=196_
NYmax =208
S 5 NYmax=210
III NYmax =215
0 III
0 112 168 224 280
35
PCF R2=0.79** 2003
30 PCG R2=0.92"

25 PSF R2=0.84"
PSG R2=0.86**
20



10- NYmax =189
NYmax =205
5 -- I I I NYmax =219-
NYmax =234
0-
0 112 168 224 280
N rate (kg.ha-1)
Figure 2-12. Predicted N rate for maximum marketable potato yield. NYmax= N rate for
maximum yield. Coefficient of determination significance: p <0.05 and
**p<0.01.









increased quadratically with increasing N rate. In 2001, there was no interaction between

crop rotation and N rate (Appendix B) and The predicted N rate required for potato to

reach maximum marketable yield (NYmax) was 164 kg.ha-1 of N for all four crop

rotations(Fig. 2-12). In 2002, the predicted NYmax for the four crop rotations was

between the range of 196 to 215 kg.ha-1 ofN (Fig. 2-12). In 2003, the NYmax ranged

between 189 and 234 kg.ha-1 of N (Fig. 2-12). The annual increase in the spread of the

range containing the predicted NYmax indicated the effect of crop rotations over-time. A

decline in yield with N rates higher than NYmax was observed in the study period (Fig.

2-12). These results were similar to reports of 168 to 196 kg.ha-1 for maximum yield and

decreased yields above 280 kg.ha-1 of N in the same site (Hochmuth and Cordasco,

2000). In 2002 and 2003, rotations including sorghum in summer produced higher

marketable yield than rotations with cowpea (Fig 2-12) suggesting late mineralization of

sorghum residues from the past season as a possible source of N during the initial stages

of the potato crop. In 2002, PSG rotation had the lowest NYmax among the rotations

studied (Fig. 2-12). This event could be explained by a decreased overall C:N ratio in the

PSG rotation caused by the inclusion of green bean which resulted in increased

mineralization of sorghum residues and finally more N could be available at potato

planting in 2003 reducing in this way the dependence of soluble fertilizers to reach the

NYmax. Furthermore, in the intercrop period between green bean harvest-2001 and

potato N side-dressing 2002 rainfall was three times less than the same period between

2002 and 2003 (Table 2-2), favoring the persistence in the soil of the mineralized N from

sorghum and green bean residues (Table 2-3). In the same period in 2003, under a three









times increased rainfall, the PSG rotation exhibited the highest requirement of N for

maximum yield (Fig. 2-12) suggesting increased N leaching-runoff (Fig 2-12).

Maximum marketable yield decreased over-time from 32 Mg.ha&1 in 2001 to 22

Mg.ha&1 in 2003 (Fig. 2-12). The NYmax reported here (Fig. 2-12) are close to the

current recommended N rate of 224 kg.ha-1 for maximum yield in the TCAA (Hochmuth

Table 2-6. Marketable and Total yield of potato under rotation by N rate.
Rotation N rate Marketable yield (Mg-ha-1) Total yield (Mg-ha-1)
PCF 0 8.68 d 11.02 d
PCF 112 21.31 be 25.48 ab
PCF 168 25.57 ab 29.73 a
PCF 224 23.45 abc 27.22 ab
PCF 280 22.63 abc 26.30 ab
PCG 0 13.92 d 16.52 c
PCG 112 23.25 abc 26.73 ab
PCG 168 24.34 abc 28.00 ab
PCG 224 24.49 abc 28.21 ab
PCG 280 20.06 c 23.26 b
PSF 0 8.46 e 11.06 d
PSF 112 25.88 ab 29.09 a
PSF 168 26.79 a 30.26 a
PSF 224 24.83 abc 28.31 ab
PSF 280 25.58 ab 28.79 a
PSG 0 12.93 d 15.24 cd
PSG 112 24.62 abc 28.08 ab
PSG 168 26.18 ab 29.83 a
PSG 224 26.71 a 30.52 a
PSG 280 23.25 abc 26.65 ab
Means followed by different letter are significantly different. Marketable yield
(p<0.0004). Total yield (p<0.0004).


and Cordasco, 2000). The small difference in yield observed among the plots with only

fertilization at planting (112 kg.ha-1 N) and plots with additional side-dressed fertilization

plots (Fig 2-12) support previous reports that most of the N used for potato under

subsurface irrigation in Hastings comes from the fertilization at planting (Elkashif and

Locascio, 1983). That report suggests upward movement of soluble salts as water

evaporated during dry periods reducing the need of side-dress fertilization. In this study,









the increased residual effect of the side-dressed fertilization on soil N03-N concentration

close to potato harvest (Fig. 2-8b) produced increase in sorghum N uptake in summer

(Fig. 2-10) supporting the suggestion that a sufficient side-dressed N fertilization would

decrease the risk of N03-N leaching-runoff without negatively affect potato marketable

yield. Higher yield with commercial N fertilization programs using only one side-dress

application compared with experimental fertilization programs including fertilization at

planting and side-dress fertilization have been reported (Hochmuth and Cordasco, 2000).

This observation suggest the possibility of reach higher potato yields with only one side-

dressed fertilization with optimum timing placement and N rate.

The inclusion of green bean in the crop rotations increased significantly marketable

and total yield of unfertilized plots as compared to unfertilized plots planted with cowpea

and sorghum in summer and left as fallow in fall. There was no effect of N rate on

marketable and total yield of potato grown under the different crop rotations (Table 2-6).

The observed increase in yield by inclusion of green bean in fall and the lack of response

of yield to N rates confirmed the potential of incorporating legumes in fall as a possible

measure to reduce the application rates of soluble fertilizers.

The export of N by total yield of potato tubers decreased significantly in

unfertilized plots along the three-year period. There were however no differences in

exported N from plots that received N rates during the three years (Table 2.7). Along

with the decrease in exported N from unfertilized plots a general increase in the % TKN

was observed throughout the period of study (Table 2-7). In a preliminary exploratory

statistical analysis these two variables were negatively correlated (r2=-0.28, p<0.0001)

explaining why in spite of the reduction of yield over the three years the exported N in









tuber remained statistically the same. Specific gravity (Table 2-7) of fertilized plots was

not affected by N rate through the 3-year period.

Table 2-7. Tuber exported N, % TKN, and specific gravity of potato under year by N
rate interaction.
Year N rate N exported (kg-ha') % TKN Specific gravity
1 0 40.39 e 0.99 h 1.071 be
1 100 79.65 cd 1.15 gh 1.078 ab
1 150 89.20 abc 1.23 fg 1.078 ab
1 200 82.15 bcd 1.22 fg 1.079 ab
1 250 85.71 bcd 1.42 ef 1.079 ab
2 0 27.27 e 1.20 fg 1.065 c
2 100 90.07 abc 1.63 de 1.074 abc
2 150 95.51 abc 1.71 de 1.075 abc
2 200 104.68 a 1.84 bcd 1.075 ab
2 250 99.08 ab 1.82 bcd 1.075 ab
3 0 16.93 f 1.75 cd 1.077 ab
3 100 69.01 d 1.79 cd 1.077 ab
3 150 90.85 abc 1.96 be 1.078 ab
3 200 94.84 abc 2.05 ab 1.079 ab
3 250 99.62 a 2.25 a 1.080 a

Means followed by different letter are significantly different. N exported (p<0.0001),
%TKN (p<0.0003), Specific gravity (p<0.0001).

Conclusions

Fertilization at potato planting generated increase in N03-N concentration in the

water table under potato beds. A similar pattern of the N03-N concentration under N

fertilized treatments was caused by the even N rate applied at potato planting.

The amount of rainfall between potato planting and N side-dress determined the

increase in N03-N concentration in the water table. Rainy periods between potato

planting and N side-dressing caused increased N03-N concentration in the water table

under fertilized potato beds.

Nitrate and nitrification of ammonium from ammonium nitrate side-dress

fertilization could generate a potential second flux of N03-N leaching/runoff by effect of

the rainfall peak in summer.









Summer cover crops with high demand of N could act as N catch crops diminishing

the possibility of NO3 leaching in summer. On the contrary, legume crops due to their

low dependence on soil N could not be efficient N catch crops especially when more than

168 kg.ha-1 of N are applied to potato.

Bush green bean as fall crop increased soil NO3-N at potato planting but the

amount of rainfall between its harvest and potato planting in spring determined how

much NO3-N persisted in the soil.

In 2001 and 2003, potato N uptake increased linearly even after potato reached its

maximum marketable yield. Potato yield under the four rotations responded quadratically

to N rate with maximum marketable yield under the recommended N rate for the region.

Rotations including sorghum in summer produced higher marketable yield than rotations

including cowpea.














CHAPTER 3
POTATO ROOT DISTRIBUTION STUDY

Introduction

The potato root system began to be studied many years ago. Pictorial descriptions

of the potato root system (Weaver, 1926) and the effect of fertilizer placement and soil

compaction (De Roo and Waggoner, 1961) have been reported. Research reports on the

relationship between the potato plant root characteristic water use and uptake of nutrients

are scarce. Total root length of potato in the field has been used to calculate inflow rates

of N, P, K and water into the plant. These studies demonstrate that roots below 30 cm

were more active in uptake of nitrate and water than those nearer the soil surface (Asfary

et al. 1982). In another study, two potato cultivars differing in their N acquisition rate

from the soil demonstrated the importance of root length and surface area to estimate N

uptake over root dry weight. This study suggested that the larger root system of one

variety enabled it to absorb more nitrate when it was limited, which allowed it to produce

its optimum yield at lower soil N availability compared with a second variety

(Sattelmacher et al. 1990). However; these approaches may underestimate the influx rate

because not all the root length and root surface area have equal nutrient uptake capacity.

The region immediately behind the meristem has greater uptake activity than older

segments, but these segments maintain some uptake capacity (Gao et al. 1998; Robinson

et al. 1991). The improvements in technology have now enabled roots to be analyzed in

detail. Root characterization studies have great potential to elucidate the role of root

structure in N management.









Enhanced N use and uptake are likely results of proposed strategies to minimize

nitrate leaching, including diminished fertilizer rates. In the short term, this could be

accomplished by timely and precise placement of the fertilizer in soil regions where it

could be quickly and efficiently absorbed by the roots. In the long term, the development

of potato cultivars with high uptake and use efficiencies by plant breeding would be part

of the solution to nitrate leaching from potato fields (Errebhi et al. 1998). In both cases a

detailed knowledge of the root system is required. The precise placement of fertilizer in

time and space is especially important for nutrients such as nitrate that have high mobility

in the soil. Root distribution studies are important to enhance the understanding of

nutrient and water uptake in order to improve irrigation and nutrient management

programs (Asfary et al. 1983).

Nitrate leaching from potato crops has been considered one of the possible causes

of water eutrophication in the St. John's River watershed, therefore improved knowledge

about the potato root system under northeast Florida conditions is needed to enhance N

uptake. In order to begin the collection of information about the potato root system, a

study of potato root distribution was initiated. The information obtained about root

distribution could be used to develop fertilization regimes based on controlled release

fertilizers in order to synchronize N release with plant demand. Knowledge of the root

system distribution could be used as well to model the N cycle in the potato production

system.

Based in the above considerations the following objectives were stated: i) to design

and test a new methodology to obtain a representative sample from the potato root system

in order to study vertical and horizontal root distribution. ii) to study the effect of









sufficient fertilization rates on the root distribution of the Atlantic potato variety used

extensively by the farmers in the TCAA located in the St. John's River watershed. iii)

generate information that could be used as part of strategies to enhance fertilizer

placement in order to maximize N uptake and minimize leaching and/or runoff of nitrate

from the potato beds.

Materials and Methods

Site Description

This study was conducted at the Hastings Research and Education Center's

Yelvington Farm in Hastings, Florida. Soil and other details about the experimental site

were described in chapter 2 of this dissertation.

Statistical Design

The plots containing the potato-sorghum-fallow rotation under the three highest N

rates (168, 224, and 280 kg-ha in the crop rotation experiment (Fig 2-1) were used to

study potato root distribution at full flowering (60 DAP) in the third year. At this time,

Atlantic reaches its maximum vegetative growth and rooting depth (Stalham and Allen,

2001). The plots used in this study were set as a randomized complete block design with

three replications and the treatments were arranged under a split-split plot design where

nitrogen rate was the main-plot (Fig. 2-1), slice was the split-plot, and position into the

slice was the split-split plot (Fig. 3-2).

Root Sampling Methodology

A device to sample a representative portion of the potato root system was designed

based on a sampler used by Vos and Groenwold (1986). The device, a "slicer", was

composed of two sharp-edged metal sheets of 101.6 cm long (distance between rows) x

50.8 cm height (Fig 3-la). Three wooden blocks keep the sheets in place with an even








separation of 10.2 cm (half the distance between plants) and three "C" clamps firmly held
the pieces together (Fig 3-la).












C D
muA









Figure 3-1. Procedure to take the root samples. (a) Root sampler; (b) leveled and buried
into the ridge; (c) slice of ridge; (d) splitting sub-samples.
The "slicer" was driven into the soil (Fig 3-1b), keeping it horizontal (Fig. 3-1c).

Subsequently, part of the ridge was removed and the slicer containing the slice of soil
was flipped on the ground and the upper sheet removed to access the ridge slice (Fig. 3-
Id) Afterward, each slice was split in 14 sub-samples (environments) arranged in three

layers (top, middle, and bottom). Each environment was coded by the first letter of the
layer where it came from followed by a number to identify the position in the layer (Fig.
3-2). Soil samples containing roots where weighed individually. Two slices, one









including a plant and another one between plants, were taken in each plot in order to get a

representative sample of the potato root system under each N rate.

Root Washing

The roots were separated from the soil by washing them on a screen while

removing organic matter debris. The soil samples containing roots were stored in a cooler

at 7 'C during the washing process. In the laboratory, roots were spread on white trays to

enhance contrast and sorted from thin residues of organic matter. Afterward, roots were

refrigerated in plastic bags containing a solution of sodium azide (0.01%) as preservative.

Root Scanning and Image Analysis

In order to enhance the image contrast and prior to scanning, roots were stained by

immersion in a solution of methylene blue (1 g L1) for at least 24 hr. Roots were spread

on translucent Plexiglas trays containing water and scanned on a HP Scanner at

resolution of 300 dpi. Images acquired by this method were loaded and saved in Adobe

PhotoDeluxe in format PDD. Root images stored as PDD were converted to grayscale

BMP format as is required by GSRoot (Guddanti and Chambers, 1993) specialized

software for root analysis. GSRoot was set up to classify roots by diameters, and estimate

root length and surface area.

Root Variables

After the scanning process, the root samples were oven dried at 105 OC and

weighed to estimate root dry weight (RDW) on an analytical scale Sartorius Model 1712

MP8 with readability of 0.00001 g. GSRoot was set up to measure seven diameter classes

(<0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, 0.8-1.0, 1.0-1.2, >1.2 mm) to get root length and root

surface area for the individual root diameter classes. Total root length (TRL) in cm and

total root surface area in cm2 (TRSA) were calculated and reported as the total sum of









length and surface area of the individual root diameter classes and expressed as meters of

root per square meter of soil (m root m-2 soil) and square centimeters of root surface per

square meter of soil (cm2 root surface m-2 soil) respectively. Using the calculated soil

volume (SVOL) for each sample (see next section) root length density (RLD) expressed

as cm root cm-3 soil and root surface area density (RSAD) expressed as mm2 root

surface-cm-3 soil were calculated. RLD and RSAD intended to describe the exploratory

capacity of the root system in terms of length and root surface by unit of soil volume,

respectively. High values of RLD and RSAD mean high capacity of the root system to

explore the soil. Specific root length (SRL) expressed as cm root mg-1 root dry matter

and specific root surface area (SRSA) expressed as cm2 root surface mg1 root dry

matter were calculated in order to estimate the resources invested by the plant in terms of

length and root surface area by unit of root dry matter. Therefore, high values of SRL and

SRSA indicate high efficiency of the root system in the assignment of resources. Total

and marketable tuber yield of the plots where the root study was done were measured and

analyzed in conjunction with all of the plots in the crop rotation experiment (chapter 2).

Soil Strength, Bulk Density, and N03-N concentration by Depth

Two days before root sampling, a study of the soil strength in the plots was done. A

Delmi penetrometer with a 300 cone-shaped tip was pushed into the soil 60 cm deep from

the top of the ridge in each sampled plot. The measure was taken in the same ridge where

the root sampling was performed. The penetrometer provided a continuous plot (on a

standard 7.6 by 12.7 cm card) of soil strength in megapascals (MPa). Using the method of

undisturbed soil cores, soil samples were collected and used to estimate soil bulk density

and soil moisture content in each one of the three layers. The soil volume (SVOL) of

each one of the 14 sub-samples was calculated using their estimated soil dry weight and









the estimated soil bulk density of the layer from which each one of the 14 sub-samples

came from. The calculated values were plugged into the following relationship:

V=W-D

where:

V= Soil volume of each sub-sample, W= Soil Dry Weight of each sub-sample, and D=

Bulk density of the layer (top, middle, or bottom) and applied to the data from each sub-

sample. Along with the sampling to estimate bulk density, soil samples were taken to

estimate N03-N concentration in the three layers of each slice.

Statistical Analysis

Root sampler

The soil wet-weight (SWW) of each sub-sample recorded at sampling time, the soil

dry weight (SDW) estimated from the bulk density sampling, and the estimated sub-

sample soil volume (SVOL) were used as response variables to evaluate the performance

of the designed root sampling device. Replication (bed) and slice were considered

random effects and their contribution to the total variance of the dependent variables was

estimated using the VARCOMP procedure of SAS (SAS Institute, 2004).

Soil strength, bulk density, and N03-N concentrations by depth.

Soil strength values obtained with the penetrometer and N03-N concentration by

depth were analyzed under a Split Plot Design using the GLM procedure of SAS (SAS

Institute, 2004). Nitrogen level was the main plot factor and depth from 0 to 36 cm in

increments of 12 cm from the top to bottom of the ridge was the subplot. Bulk density

was analyzed as a RCB design for the same depths sampled with the penetrometer. The

LSMEANS option with Tukey adjustment was used to separate soil strength and N03-N

concentration by depth. A regular Tukey test was used to separate bulk density means.









Roots

As exploratory analysis, scatter plots of the studied variables to search possible

association were plotted using the GOPTION procedure of SAS/GRAPH (SAS Institute,

2004). Four different models using the MIXED procedure and considering common

pooled variance, heterogeneous variances for bed (replication) by N treatment, spatial

correlations within slice for each slice, and spatial correlations within and between slices

were evaluated. Furthermore, the same four models were re-evaluated using root dry

weight (RDW) and soil volume (SVOL) as covariates. After the evaluation process, RLD

and RSAD were analyzed using the model considering spatial correlations within and

between slices, SRL was analyzed using the model considering heterogeneous variances,

and SRA was analyzed with the model considering spatial correlations within and

between slices. The LSMEANS option was used to perform means separation. Means of

SRL and SRA were separated with the Bonferroni adjustment and SRL and SRA by the

Fisher test with alpha=0.001.

Results and Discussion

Root Sampler Device

The variance attributable to the sampling technique (slice) was relatively low

compared with the variance of environments. This result could be caused by the irregular

shape of the sub-samples coming from the external part of the ridge (Fig. 3-2). Soil

volume was used to evaluate the root sampler performance. This effect was evaluated

using the soil volume of each one of the subsamples taken in each slice. Average soil

volume (cm3) and horizontal and vertical dimensions for each environment are included

in Figure 3-2. Soil volume for each environment is the overall mean estimated through

out all the soil slices sampled.











cm
0 ---------- --




12 ---------- -
M1
2044


24 ------
B1I B2
2044 219(


20 I


15 1


cm

Figure 3-2. Distribution and sample size into the slice. Soil volume (cm3), dimensions
(cm), Tl to B6 = Environment code for identification.

Soil Strength and Bulk density

Soil strength differences among blocks (beds) were observed. Nitrogen treatment

had no effect on soil strength. Highly significant differences were observed among all of

the three layers studied (Table 3-1). The increase of soil resistance from the top to the

bottom layer coincided with a significant increase in the observed bulk density of the

bottom layer compared with the top and middle layers (Table 3-1). In some root samples

coming from the 12 to 24 and 24 to 36 cm layers a pronounced thickness of the

Table 3-1. Soil strength, Bulk density, and Soil N03-N by depth.
Soil depth (cm) Soil strength (MPa) Bulk density (g-cm-3) Soil N03-N (mg-kg 1)
0-12 0.024 a 1.28 a 10.22 a
12-24 0.105 b 1.45 b 14.15 a
24-36 0.203 c 1.50b 2.71 b
Soil strength (p<0.0001); Bulk density (p<0.0001); Soil N03-N (p<0.0082).

roots were observed (Fig. 3-3). A similar pattern of thickness was reported for tomato

roots impeded by a column of glass beads (Waisel, 2002). Inhibition of the elongation of

the main axis and enhanced formation of lateral roots in young barley plants grown in









fields with high bulk density also have been reported (Scott-Russell et al. 1974). Also,

accumulation of ethylene in roots under mechanical stress and anaerobic conditions have

been reported as the possible cause of aerenchyma formation in the root cortex of

waterlogged maize roots (Drew et al. 1979). In the case of the potato roots, a combination

of these two conditions could have been responsible for the increase in diameter of roots

in the middle and bottom layer (Fig 3-3). Soil N03-N concentration from 24 to 36 cm

was lower than the concentration observed at the 0 to 12 and 12 to 24 cm depths. There

was no difference between the two top layers (Table 3-3). The low concentration of NO3-

N in the deepest layer could be explained by increased denitrification due to the

anaerobic conditions in the bottom layer and/or upward capillary flux of water causing

concentration of nitrate in the two top layers as have been reported in a study with

lysimeters (Patel, 2001).

Rooting Depth, Root Length, and Root Surface Area

The deepest roots were observed in the bottom layer (24-36 cm) and never exceed

36 cm. This is a shallow rooting depth compared to the maximum rooting depth of 60 cm

reported for Atlantic on a deep sandy clay loam soil under overhead irrigation in the UK

(Stalham and Allen, 2001). At that time 16 varieties of varying determinacy were tested

and some of them reached rooting depths close to 1 meter. Atlantic was one of the

varieties with a shallower root system. Even though that observation was under different

conditions, it could be taken as evidence that Atlantic has a restricted root system growth

in Hastings.

The observed total root length (TRL) for Atlantic in Hastings was 0.763 km roo t

m2 soil (Table 3-2). Although in the literature there are no data on TRL of Atlantic it is a

very low value as compared with reported values as high as 24 km root m2 for the









cultivar Norin 1 (Iwama et al. 1999), 11.4 km root m2 for Russet Burbank (Lescyznki

and Tanner, 1976), and even still lower than 1.4 to 4.6 km root m2 for White Rose

(Bishop and Grimes, 1978). The vertical distribution of Atlantic TRL (Table. 3-2 ) shows

that 94% of the TRL was in the two upper layers (0 to 24 cm) but 63% of the TRL was

concentrated in the middle layer (12 to 24 cm) and just 6% was in the bottom layer (24 to

36 cm). The region from 12 to 24 cm could be the place where most of the NOs3 uptake is

taking place under Hastings conditions. Seventy-six percent of the TRL was comprised

by roots less than 0.4 mm in diameter suggesting this fraction was a important part of the

Atlantic root system (Table 3-2).


















Figure 3-3. Detail of potato roots thickening

The observed total root surface area (TRSA) for Atlantic was 838.41 cm rdot -

m2 with 92 % of this area in the upper 24 cm of the ridge (Table 3-2) and 64.4 % in the

middle layer (12 -24 cm). The fraction of roots with diameter less than 0.4 mm

represented 76 % of the TRL but just 45 % of the TRSA. This observation shows one of

the inconveniences when TRL is used to estimate the apparent inflow rates of nutrients.

Therefore, approaches using TRL could underestimate the true values of nutrient uptake









(Robinson et al. 1991). Determination of the root diameter class active in nitrate uptake

will be a key factor to adjust fertilizer placement in order to improve its uptake.

Table 3-2. Root length and root surface area by diameter class and soil depth.
Depth Diameter class (mm)
(cm) <0.2 0.2-0.4 0.4-0.6 0.6-0.8 0.8-1.0 1.0-1.2 >1.2
93.38* 94.33 24.89 13.51 6.16 2.61 2.37 237.01
39.4% 39.8% 10.5% 5.7% 2.6% 1.1% 1.0% 31.0%
S12-24 142.93 219.92 69.30 25.51 12.03 5.77 5.29 481.23
29.7% 45.7% 14.4% 5.3% 2.5% 1.2% 1.0% 63.0%
S 46 10.48 20.77 8.05 3.52 1.56 0.69 0.69 45.75
24-36
22.9% 45.4% 17.6% 7.7% 3.4% 1.5% 1.5% 6.0%
STotal 246.65 335.24 102.14 42.77 20.04 8.93 8.23 763.99
0 ^ Total
32.3% 43.9% 13.4% 5.6% 2.6% 1.2% 1.1% 100%

Depth Diameter class (mm)
(cm) <0.2 0.2-0.4 0.4-0.6 0.6-0.8 0.8-1.0 1.0-1.2 >1.2
25.82* 64.32 50.0 42.72 24.41 11.74 15.73 234.73
g 8.7 % 27.4% 21.3% 18.2% 10.4% 5.0% 6.7% 28.0%
12-24 43.74 189.00 130.14 77.22 45.36 25.38 29.70 540.00
122 8.1% 35.0% 24.1% 14.3% 8.4% 4.7% 5.5% 64.4%
-2 3.88 21.65 16.62 10.00 5.54 2.67 3.44 63.68
24-36
2 6.1% 34.0% 26.1% 15.7% 8.7% 4.2% 5.4% 7.6%
OC 73.14 274.99 196.62 130.03 75.37 39.41 48.86 838.41
Tota 8.7 32.80 23.5% 15.5% 9.0% 4.7% 1.0% 100.0%

Root Length Density

No significant differences in RLD were observed under the three N rates applied to

potato. Significant differences in RLD among environments (p<0.0001) and between

slices (p<0.05) were observed (Fig. 3-4). The slice containing the plant had higher RLD

than the slice between plants (Fig. 3-4). RLD values in the environments M2 and M3

were higher than all of the values observed in other environments except from the values

of T2 and T3 where the difference was not significant. M2 and M3 were the

environments where the seed piece was located making them the center of origin of the

potato root system (Fig. 3-4). There were no differences among the environments

covering M2 and M3. The environments in the bottom layer (B 1 to B6) had significantly











lower values than the environments in the middle layer (Ml to M4), suggesting that soil

exploration by the potato root system is evidently reduced at more that 24 cm depth (Fig.


3-4).





1 T2 T3 T4


M3 M4

1 T2 T3 4
bcd ab ab cde
M1 M2 M3 M4
0.467 0.838 0.812 0.497
bc a a bc
0.294
B1 B2 B3 B4 B5 B6 b
0.019 0.070 0.038 0.007 0.030 0.053
e de e e e de

WP
0.387
a

Figure 3-4. Root length density distribution (cm root cm s3 soil). P= Slice containing the
plant; B= Slice between plants. Tl to B6 = Environment identification. Means
with same lowercase letter are not statistically different. Environments
(p<0.0001); Slices (<0.025).

Root Surface Area Density

No significant differences on RSAD under N rates were observed. This variable

was highly correlated with RLD (0.95, p<0.0001). Differently from RLD, a significant

interaction environment by slice was observed (Fig. 3-5). RSAD had higher values for

environments M2 and M3 in the slice containing the plant but the difference was not

significant with T2. The difference between M3 and T3 was significant. The differences

in RSAD in T2 and T3 as compared with M2 and M3 could be the result of the position

of the seed piece in the two central M environments and the position of the stem in the

two central T environments. A lot of adventitious roots were originated from the stem










when part of it was soil covered at the time of hilling and N side-dressing. That means

that in terms of root surface the potato plant could have its maximum uptake capacity in




T1 T2 T3 T4
2.53 4.34 3.82 3.06
efgh defgh efgh efgh

M1 M2 M3 M4
4.22 9.60 7.04 4.08
efgh abcd bcdef efgh
B1 B2 B3 B4 B5 B6
0.15 0.37 0.25 0.07 0.24 0.50
h gh h h h gh





3.53 9.70 7.63 1.85
efgh abc ab fgh
M1 M2 M3 M4
5.08 10.39 13.06 5.61
cdedfgh ab a bcdefg

B1 B2 B3 B4 B5 B6
0.27 1.68 1.05 0.14 0.59 0.66
h gh gh h gh gh


Figure 3-5. Root surface area density distribution (mm2 root cm-3 soil). WP= Slice
containing the plant; BP= Slice between plants. TI to B6 = Environment
identification. Means with same lowercase letter are not statistically different
(p<0.0059). Possible comparisons between slices.

the central two M environments. Along with the greater RSAD in the two central M

environments, higher soil moisture and higher NOs3 concentration were observed in the

middle layer than in the top layer (Fig. 3-5 ). In chapter 2 was shown that the residual

effect of the side-dressed fertilizer applied 40 DAP remained in the soil even after potato

harvest. The dryness of the top layer could be limiting NOs3 uptake by the potato roots.

Even though the difference among environments in different slice was not significant

except when the environments were compared with M2 and M3 in both slices it is









possible suggest that roots in the region between plants could have less capacity to

scavenge NOs3 from the soil. In the slice between plants all of the environments around

M2 and M3 were not different among them, even though high differences among the

RDA values for environments in the bottom layer compared with the two environments at

both sides of the middle layer and all of the environments in the top layer were observed

but not statistically detected by the model (Fig. 3-5).

Specific Root Length

There were no significant differences by slice and N treatment in SRL. Highly

significant differences were detected among environments. No significant differences

were observed for the interactions among factors. The highest values for SRL were

observed in the environments Tl and T4 without significant difference between them

(Fig. 3-6). The high values of SRL in the environments TI and T4 are indication of very

fine roots in this part of the ridge. The stem coverage with loose soil at N side-dress and

hilling time (40 DAP) could be the explanation for the development of a system of fine

adventitious roots in these environments (Fig. 3-6). On the contrary, the low SRL values

observed in B3 and B4 (Fig. 3- 6) could be the result of the increase in diameter caused

by the combination of the high bulk density in B3 and B4 environments and by the water

rising by capillary effect from the water table. Saturated soil conditions could increase

root diameter by stimulation of aerenchyma formation (Drew et al. 1979), and saturated

conditions could be motivated by a high capillary flux from the water table as reported in

presence of a plow pan (van Loon and Bouma, 1978). The higher SRL values observed in

the environments beside B3 and B4 suggest that the compacted zone could be localized

just in the two central environments of the bottom layer. In terms of SRL the

environments T2 and T4 are the places where the potato plant has its most fine fraction of










roots and probably where high nutrient uptake could take place but it also could be

limited by the fluctuant soil moisture on the top of the ridge. If RLD values (Fig. 3-4) for

Tl and T4 are compared with the values for M2 and M3 (Fig. 3-4 ) it is possible to notice

that the two environments on the top of the ridge have significantly lower values than M2

and M3 in the middle of the ridge. It is explained by the high concentration of roots in the

central environments (M2 and M3) where the seed piece was placed. On the other hand,





T1 T2 T3 T4
21.98 14.23 13.34 19.56
a bc bcd ab

M1 M2 M3 M4
15.59 14.12 17.01 15.59
bce bc abc be

B1 B2 B3 B4 B6 BS
13.29 17.12 7.21 5.73 12.06 16.51
cd abc de e cd bec




Figure 3-6. Specific root length distribution (cm root mg root dry weight). Tl to B6 =
Environment identification. Means with same lowercase letter are not
statistically different. (p<0.0001).

the significant lower SRL values observed in M2 and M3 compared with Tl and T4 are

indication of the presence of roots with high diameter close to the root piece. Generally,

finest roots are more efficient in nutrient uptake than coarse roots, therefore would be

possible suggest that roots in T1 and T4 could be more active in NOs3 uptake when soil

moisture is not limiting it.

Specific Root Surface Area

SRSA was analyzed using RDW as a covariate (p <0.0004). The interaction

environment by slice was significant. The scarce difference observed among SRSA










suggests that the potato root system has a very homogeneous configuration in terms

of.the amount of biomass allocated by unit of root surface area (Fig. 3-7). Significant low

SRSA values were observed in B3 and B4 environments located in the slice between

plants, however, these values were not statistically different from many higher values in

the same slice and in the slice including the potato plant (Fig. 3-7). The lack of sensitivity

of the model to detect high differences among SRSA could be attributed to the high




1 T2 T3 T4
1.575 1.313 10.268 1.538
bcd cd cd bcd
M1 M2 M3 M4
1.666 1.905 1.741 1.426
bc bc bc bcd

B1 B2 B4 B5 B5 BG
1.291 2.700 0.415 0.315 1.266 1.356
cd ab d d cd cdBP




1 T2 T3 T4
1.5421 1.791036 1.444 1.291
bcd bc bcd cd

M1 M2 M3 M4
1.782 1.816 3.698 1.849
bc bc a bc

B1 B2 B3 B4 B5 B6
1.392 1.916 2.036 1.225 2.206 1.517
cd bc bc cd b bcd
WP


Figure 3-7. Specific root surface area distribution (cm2 root mg' root dry weight).
WP= Slice containing the plant; BP= Slice between plants. TI to B6 =
Environment identification. Means with same lowercase letter are not
statistically different (p<0.0395). Comparisons possible between slices.

variability produced by the relatively low resolution (300 dpi) used to capture the root

images; therefore, low resolution maybe affected more area than length estimation. Based

on the results of SRL where the lowest values were observed in the same environments

(B3 and B4), it would be possible suggest that the low values of SRSA observed in the









corresponding environments in the slice between plants could be really different from

higher values in both of the slices. If this idea is correct the theory of increased thickness

due to soil impedance and/or flooded conditions could be supported by these results.

Conclusions

This study was developed to propose a new methodology to study the potato root

system under Hastings conditions and to describe spatially the potato root distribution as

affected by N rates currently used in commercial potato production.

The proposed sampling methodology performed satisfactorily and is suitable for

future root studies in the TCAA or agricultural areas with similar conditions. The ridge

slicing methodology allowed to recover most of the roots from a representative sample

of the potato root system and spatially classify sub-samples around the potato plant.

The potato root distribution was not affected by N rate. RLD, SRL, RSAD, and

SRSA were affected by their spatial position in the soil profile. Soil exploration capacity

(RLD and RSAD) was maximized into 15 cm around the potato plant at a depth of 24 cm

from the top of the ridge. Significant lower soil exploration capacity was detected under

24 cm from the top of the ridge. The invested biomass by unit of root length or unit of

surface area (SRL and SRSA) were relatively homogeneous except in the environments

under the potato plant (B3 and B4) where low values were probably caused by root

thickening caused by a combination of increased soil strength and saturated conditions.

High bulk density and water saturation at the bottom layer of the ridge could be a limiting

factor for potato root development.

Observed differences in root parameters could be used as base to plan fertilizer

placement strategies in order to enhance nutrient uptake. To optimize fertilizer placement

a detailed study of the water movement in the root zone should be done.















CHAPTER 4
SUMMARY, OVERALL CONCLUSIONS AND FUTURE RESEARCH

Some years ago nutrient loses from agricultural production systems were

considered as a purely agronomic problem limited productivity and negatively affected

the economic profitability of agricultural producers. Currently in Florida, the negative

effect of nutrient enriched water bodies is turning on a great concern in endangered zones

where nutrient mobility is expedited by the low capacity of the soil to retain them in

exchangeable form and for the sandy nature of the soils. This is the case of most of the

Florida soils where the low content of mineral colloids results in a pH dependent

exchange capacity. The TCAA in northeast Florida is a region where potatoes and

cabbage are the dominant crops. Due to the relatively high N demand by potato, farmers

apply more N fertilizer than is needed for optimum production generating nitrate non-

source pollution from the potato fields. The over-fertilization problem could be enhanced

by the current irrigation system used for agricultural production. It consist in a perched

water table that is controlled by pumping water from deep wells into irrigation-drainage

furrows running parallels to the raised potato beds. The concern by the potential

environmental impact of agricultural practices have generated a growing interest to

develop Best Management Practices (BMPs) that allow sustainable crop productivity

without negatively affecting the environment. Inclusion of legumes as green manure or as

a second cash crop in rotation with potatoes has been considered a viable way to supply

part of the relatively high N requirement of soluble fertilizers applied to potatoes.









Synchronizing N availability in the soil and plant uptake is another proposed approach

that require good knowledge of the root system in order to enhance fertilizer uptake by an

optimum placement in the region of the root zone where uptake will be maximized.

Based on the previous considerations this dissertation included in the Chapter 2 the

sequential study of NO3-N concentration in the water table, soil solution, soil, and plant

tissue as affected by four different crop rotations and five N rates over a three-year

period. Chapter 3 reports the study of the potato root system in order to describe

quantitatively and spatially its distribution in the soil in terms of root length and root

surface area under three rates of nitrogen. Each one of these chapters had specific

objectives and their results are summarized below.

Summary

Crop Rotation and N rates

Concentrations of N03-N in the water table under potato

In the first year N03-N concentration in the water table was increased up to 54.6

mg-L-1 under effect of fertilization at planting. Predicted values for non-fertilized plots

were in the range of 8.2 to 18.2 mg-L'1. Decreasing patterns of the regressions describing

N03-N concentration in the water table were observed from potato planting to harvest.

In the second year predicted values for N03-N concentration were below 10 mg-L-1

and a light increase were predicted close to the harvest time. There was no noticeable

difference among fertilized and unfertilized plots. The low rainfall after potato planting

could be addressed as the cause of the low N03-N in the water table.

In the third year a similar pattern to the first year was observed with the difference

that unfertilized plots exhibited lower N03-N concentration than the first year. Side-

dressed nitrogen at 40 DAP did not affect the decreasing pattern of N03-N concentration









in the water table in the first an d third years suggesting fertilization at potato planting as

the main source of the increased N03-N concentration in the water table.

Concentrations of N03-N in the soil solution

It was possible to study this variable just in the second year due to insufficient

information from the first year and problems with the installation of the lysimeters in the

third year. Even though 2002 was a year with low rainfall after potato planting N03-N

concentration values as high as 156.0 mg-L1 were observed after fertilization. The

unaffected decreasing pattern after side-dress fertilization suggests low effect of it on

N03-N concentration in the water table.

Concentrations of N03-N in the water table under cowpea and sorghum

Concentration of N03-N in the water table under cowpea plots under effect of

residual N from potato fertilized with 280 kg.ha-1 was significantly higher than

concentration under plots fertilized with 168 kg.ha-1. There was no difference in N03-N

concentration in the water table under cowpea planted in unfertilized plots and under

cowpea planted in plots fertilized with 168 kg.ha-1 of N. N03-N concentration in the

water table under sorghum was significantly lower than under cowpea in plots fertilized

with 280 kg.ha-1 except in the second sampling when high variability was observed.

Concentrations of N03-N and NH4-N in the soil

Soil N03-N concentration was increased after side-dressing N and exhibited a

rising pattern across the N rates applied. The same pattern persisted in the soil for 3

months and was detected before potato harvest and at planting of sorghum and cowpea in

summer. The concentration of N03-N at planting of the cover crops showed variation

among years. This variation could be attributable to the amount of rainfall between potato

harvest and sorghum and cowpea planting. Soil concentration of NH4-N across the N









rates was described by a quadratic pattern 3 weeks fertilizer side-dressing and there was

no difference between unfertilized plots and plots with only fertilization at planting

suggesting that nitrification occurred. In 2003 coincident high concentration ofNO3-N

and low concentration ofNH4-N 3 weeks before potato harvest suggest a late nitrification

of the side-dressed fertilizer. Increased soil N03-N concentration in plots planted with

green bean in fall as compared with plots left as fallow was noticed.

Nitrogen recovery by the summer cover crops

Residual N from the fertilizer applied to potato produced increased nitrogen

recovery by the sorghum from the fertilized plots. Nitrogen accumulation in sorghum

biomass was described by similar regressions describing soil N03-N concentration after

N side-dress, at potato harvest, and at summer cover crops planting. Cowpea N

accumulation was not affected by the N rate applied to potato. Cowpea N accumulation

in biomass in fertilized plots was not different from nitrogen accumulation in unfertilized

plots. This event supports the observation that the residual fertilizer from potato side-

dressed fertilization could be generating a second flux of nitrate leaching from the plots

planted with cowpea in summer. These results suggest sorghum as an effective N catch

crop that should be planted as close as possible to the potato harvest. Low soil N uptake

by cowpea could generate leaching-runoff when potato is fertilized with more than 168

kg.ha1 ofN.

Potato N uptake at full flowering

Nitrogen rate affected significantly potato N uptake over the 3-year period.

Nitrogen uptake was described by linear regressions in 2001 and 2003, and by a quadratic

regression in 2002. The continuous increment in N uptake even after potato reach its

maximum yield suggest luxury consumption of N when higher N rates than needed for









maximum marketable yield were applied. In 2002, the overall potato N uptake showed

closer to N rate needed for maximum yield.

Potato yield

Quadratic increase by effect of N rate was observed in each year. In 2001, there

was a relatively similar response of the four rotations to N rate. This year was considered

a transition year where the effect of crop rotation with legumes was not evident. In 2002

and 2003 rotations including sorghum as summer cover crop yielded more than rotations

with cowpea. The difference in the N rate required for maximum yield by the four crop

rotations was increased from 2002 to 2003. A period with heavy rainfall at the end of

2002 maybe propitiated leaching of the mineralized N from the cover crop residues

generating different potato N requirement for maximum yield in 2003.

Exported N as total yield of tubers in fertilized plots remained statistically the same

within year throughout the three-year period without statistically differences among N

rates except in the third year when the treatment with just fertilization at planting (112 kg-

ha-1 of N) exported significantly less N than plots with side-dressed fertilizer. These

results suggest that fertilization at planting could be enough to supply the N needed to

reach optimum yield under Hastings conditions.

A generalized increase of %TKN along with total yield decreasing during the three-

year period was observed explaining the relatively unaffected exportation of N by the

decrease in yield.

Potato Root Distribution Study

Root sampler

The root sampler device was an efficient method to sample the potato root system

in the ridges used for potatoes production in Hastings.









Soil strength, bulk density, and N03-N by depth

Significant increase in the soil strength from the top to the bottom layer of the ridge

was observed. Bulk density followed the same trend but there were no differences

between the middle (12-24 cm) and the bottom (24-36 cm) layers. Soil N03-N

concentration by layer was significantly higher in the top and middle layers as compared

with the bottom layer. This observation coincide with the observed persistence of the

side-dressed fertilizer in the soil in the top of the ridge.

Rooting depth, root length, and root surface area

The deepest roots were observed between 24 and 36 cm. This is a very shallow

rooting depth as compared with 60 cm reported for Atlantic in well drained soils. The

total root length observed was 0.763 km root m2 soil This is also a low value as compared

with the reported values for some potato varieties. Ninety four percent of the total root

length was in the 2 top layers. The total root surface area was 838.41 cm2 m2 soil and 92%

of the root surface area was in the 2 top layers. There were no reported values of potato root

surface area to compare.

Root length density and specific root length

Root Length Density (RLD) was not affected by nitrogen rate. Differences

wereobserved among different environments and between slices. Significant higher

values of RLD were observed in the central environments of the middle and top layers.

Very low RLD values were observed in the bottom layer. RLD in the slice was

significantly lesser than RLD in the slice containing a plant. Specific Root Length (SRL)

was not affected by N rate. There were no differences between slices. The highest values

for SRL were observed in the external environments of the top layer maybe caused by the

fine adventitious root system generated by the stem coverage with soil at side-dress









fertilization. In order to predict placement of fertilizer to enhance nitrogen uptake values

of RLD should be compared with SRL in order to select the environments with

coincident highest values for these two variables.

Root surface area density and specific root surface area

RSAD and SRSA were not affected by, nitrogen rate and were highly correlated

with RLD and SRL. Differently from RLD significant interaction environment by slice was

observed. RSAD in the central environments in the two top layers of the slice containing

the plant exhibited the highest RSAD values maybe because in those environments was

located the seed piece and the stems generated from it. SRSA was relatively

homogeneous among environments and between slices suggesting that potato has a

homogeneous configuration in terms of the amount of biomass allocated by unit of root

surface. Significant low SRSA values were observed in the central environments of the

bottom layer in the slice between plants. This is the result of the thickening observed in

some of the root samples recovered from those environments.

Overall Conclusions

Fertilization at potato planting generated increase in N03-N concentration the water

table with a steady decreasing pattern through the time under the four rotations evaluated.

The little difference among fertilized plots (2-3a to 2-5b) was maybe caused by the even

N rate applied at planting to fertilized plots (Table 2-1). The influence of the rainfall on

the magnitude of the increase in N03-N concentration in the water table was evident in

2002 when an atypical low rainfall period between potato planting and N side-dress

(Table 2-2) caused lower N03-N concentration in the water table before N side-dressing

(Fig. 2-4a and 2-4b) compared with the same period in 2001 and 2003. Fertilization at

planting, when potato has not established its root system could increase the risk of nitrate









leaching/runoff. Therefore, fertilization at crop emergence could be more effective to

reduce the chance of N leaching/runoff by increasing the N uptake. Fertilizer placement

should be considered also in order to enhance N uptake. Based in the description of the

potato root distribution is possible suggest T2, T3, M2, and M3 as the optimal

environments to locate fertilizers at planting or potato emergence. These environments

comprise a zone of 15 cm around the potato plant with a depth of 24 cm from the top of

the ridge. In these environments the potato seed piece was located; therefore, in this soil

region nutrient uptake will take place first. The environments T2, T3, M2, and M3

comprise the root zone where coincided high RLD and high SRL showing that in these

environments was concentrated the maximum exploratory capacity of the potato root

system (Fig. 3-4) along with high density of fine roots (Fig. 3-6). A detailed study of the

water movement in the potato root zone should be done in order to optimize nutrient

placement into the root zone. Previous reports in the region (Elkashif and Locascio,

1983) and in lysimeters (Patel et al. 2001) have reported ascendant movement of nitrates

due to water evaporation and plant uptake during dry periods. Therefore, low placement

of the fertilizers on dry periods like after potato side dressing, or placement of controlled

release fertilizers under the potato seed at planting, programmed to liberate nutrients

when the dry period after N side-dressing begins could be an alternative to enhance plant

uptake N uptake. Controlled release fertilizers could be placed closer to the potato seed

with lesser risk of seed damage than conventional soluble fertilizers.

Optimized fertilizer placement at potato planting could help to minimize the need

of side-dress N fertilization turning it in an optional practice when seasons with heavy

rains occur. Minimal side-dressed fertilization would diminish the risk of a potential









second flux of N03-N leaching-runoff originated in the residues of the side-dressed

application of N by effect of the rainfall peak in summer (Fig. 2-7).

Summer cover crops with high demand of N planted as soon as possible after

potato harvest should be planted in order to recover any residual N from the potato

fertilization. In this study sorghum exhibited higher capacity to recover residual N than

cowpea when the N rate applied to potato was higher than 168 kg.ha-1. Potato in rotation

with sorghum in summer yielded more than rotations with cowpea suggesting sorghum as

a better crop to storage N to be used in the next potato season specially when heavy rains

occur at the end of the year.

Bush green bean as fall crop increased soil N at potato planting but the amount of

rainfall between harvest of the fall crop and potato planting in spring (Table 2-2)

determined how much N persisted in the soil until potato planting (Table 2-3). The

relatively large difference in the N rate required for maximum potato yield of the rotation

PSG between 2002 and 2003 (Fig. 2-12) suggests that green bean could be decreasing the

overall C/N ratio of the soil organic matter, therefore increased N mineralization and

nitrification could enhance N availability for the next potato crop but also could increase

the risk of nitrate leaching/runoff in years with heavy rains between green bean harvest

and potato planting.

Finally it is possible to conclude that an improved N management of the potato

production system in order to reach optimum marketable yield minimizing at the same

time the risk of N leaching/runoff could be achieved by improved fertilization practices

including optimized fertilizer timing, placement, and N sources. These practices should









be complemented by rotation with cover crops that allow recover residual N from the

potato season and transfer it in an efficient manner to the next potato crop.

Future Research

Based on the results of this dissertation is possible suggest some topics to be

addressed by future research work.

* The water movement toward and from the water table as influenced by the rainfall
and the sub-irrigation system should be studied in order to estimate its role on
nitrate movement.

* Different non-leguminous nitrogen catch crops after potato harvest should be tested
in order to ensure a maximum recovery of the possible residual nitrogen fertilizer.

* Depending on the catch crop and its C/N ratio different incorporation times in order
to match available nitrogen from the mineralized residues with potato uptake
should be tested.

* Based on the results of this dissertation the effect of reduced nitrogen rates or
controlled nitrogen sources allocated in the sites of the root zone with high
scavenging capacity should be studied.

* Determination of the fraction of potato root diameters implied in nitrogen uptake
should be investigated.



















APPENDIX A
REGRESSION PARAMETERS


N03-N concentration in the water table (Figs. 2-3a to 2-5b)


Year Crop N rate Conc.
Rotation kg.ha mg.L'1 Intercept Linear term Quadratic term R2
2001 PCF 0 Ln[N03-N] = 1.750575 -0.077125*dss 0.79**
2001 PCF 112 Ln[N03-N] = 2.310352 0.111002*dss + 0.000350*dss2 0.78**
2001 PCF 168 Ln[N03-N] = 2.335808 0.120305*dss + 0.000710*dss2 0.77**
2001 PCF 224 Ln[N03-N] = 2.477959 0.118539*dss + 0.000821*dss2 0.73**
2001 PCF 280 Ln[N03-N] = 2.871943 0.176226*dss + 0.001600*dss2 0.77**
2001 PCG 0 Ln[N03-N] = 1.893703 0.102081*dss + 0.000360*dss2 0.81**
2001 PCG 112 Ln[N03-N] = 2.37892 0.105337*dss + 0.000485*dss2 0.66**
2001 PCG 168 Ln[N03-N] = 2.36447 -0.104694*dss 0.80**
2001 PCG 224 Ln[N03-N] = 2.79625 -0.077589*dss 0.77**
2001 PCG 280 Ln[N03-N] = 2.529325 -0.058704*dss 0.67**
2001 PSF 0 Ln[N03-N] = 2.546013 -0.102689*dss 0.76**
2001 PSF 112 Ln[N03-N] = 2.592968 0.025896*dss 0.000878*dss2 0.74**
2001 PSF 168 Ln[N03-N] = 2.801359 0.040917*dss 0.000700*dss2 0.81**
2001 PSF 224 Ln[N03-N] = 3.157429 -0.087921*dss 0.72**
2001 PSF 280 Ln[N03-N] = 3.140889 -0.050690*dss 0.78**
2001 PSG 0 Ln[N03-N] = 2.529126 -0.122880*dss 0.64**
2001 PSG 112 Ln[N03-N] = 3.077667 -0.072097*dss 0.87**
2001 PSG 168 Ln[N03-N] = 3.69064 -0.108683*dss 0.79**
2001 PSG 224 Ln[N03-N] = 3.08599 -0.076067*dss 0.84**
2001 PSG 280 Ln[N03-N] = 3.127979 0.033881*dss 0.000947*dss2 0.81**
2002 PCF 0 Ln[N03-N] = 0.125753 0.126270*dss + 0.002298*dss2 0.57**
2002 PCF 112 Ln[N03-N] = -0.74375 -0.067456*dss 0.25ns"
2002 PCF 168 Ln[N03-N] = -0.33049 0.098137*dss + 0.001950*dss2 0.46**
2002 PCF 224 Ln[N03-N] = -0.12305 0.079519*dss + 0.001408*dss2 0.44**
2002 PCF 280 Ln[N03-N] = -0.13962 0.094500*dss + 0.001928*dss2 0.26ns
2002 PCG 0 Ln[N03-N] = 0.078867 -0.067357*dss 0.37*
2002 PCG 112 Ln[N03-N] = -0.14726 0.040648*dss + 0.001036*dss2 0.24ns
2002 PCG 168 Ln[N03-N] = 0.060474 0.088368*dss + 0.001 519*dss2 0.50**
2002 PCG 224 Ln[N03-N] = 0.16497 0.107891*dss + 0.002066*dss2 0.30*
2002 PCG 280 Ln[N03-N] = 1.123871 0.036187*dss + 0.000097*dss2 0.26ns
2002 PSF 0 Ln[N03-N] = -1.04924 0.031680*dss + 0.000662*dss2 0.05ns
2002 PSF 112 Ln[N03-N] = 0.146383 -0.068382*dss 0.32*
2002 PSF 168 Ln[N03-N] = 0.394715 -0.080456*dss 0.34*
2002 PSF 224 Ln[N03-N] = 0.456274 0.092204*dss + 0.001404*dss2 0.46**
2002 PSF 280 Ln[N03-N] = 0.13171 -0.085000*dss 0.25ns
2002 PSG 0 Ln[N03-N] = -0.42654 0.046118*dss + 0.001040*dss2 0.15ns











N03-N concentration in the water table. Continued.

Year Crop N rate Cone.
Rotation kg.ha 1 mg.L1 Intercept Linearterm Quadratic term R2
2002 PSG 112 Ln[N03-N] = -0.12776 0.094905*dss + 0.001872*dss2 0.53*
2002 PSG 168 Ln[N03-N] = 0.609595 0.112335*dss + 0.001897*dss2 0.33*
2002 PSG 224 Ln[N03-N] = -0.42545 0.055495*dss + 0.001258*dss2 0.10ns
2002 PSG 280 Ln[N03-N] = 0.243151 0.043875*dss + 0.000418*dss2 0.20ns
2003 PCF 0 Ln[N03-N] = 0.382195 0.067770*dss + 0.000672*dss2 0.22ns
2003 PCF 168 Ln[N03-N] = 2.866326 0.129757*dss + 0.000995*dss2 0.44**
2003 PCF 280 Ln[N03-N] = 3.289673 0.113775*dss 0.60**
2003 PCG 0 Ln[N03-N] = 2.928474 0.127533*dss + 0.001098*dss2 0.82**
2003 PCG 168 Ln[N03-N] = 2.873487 0.131558*dss 0.54**
2003 PCG 280 Ln[N03-N] = 4.226966 0.117990*dss 0.81**
2003 PSF 0 Ln[N03-N] = -0.11973 0.030667*dss + 0.000130*dss2 0.18ns
2003 PSF 168 Ln[N03-N] = 2.59872 0.093499*dss + 0.000520*dss2 0.47**
2003 PSF 280 Ln[N03-N] = 2.771621 0.082717*dss + 0.000289*dss2 0.55**
2003 PSG 0 Ln[N03-N] = 0.73014 0.055740*dss + 0.000460*dss2 0.37*
2003 PSG 168 Ln[N03-N] = 4.118405 0.166594*dss + 0.001241*dss2 0.84**
2003 PSG 280 Ln[N03-N] = 3.925743 0.124762*dss 0.83**
dss= Days since side-dress
ns = No significant

* = Significant to 0.05
**= Significant to 0.01











N03-N concentration in the soil solution (Figs. 2-6a, b)


Year Crop N rate Conc.
Rotation kg.ha 1 mg.L'1 Intercept Linear term Quadratic term R2
2002 PCF 0 Ln[NO3-N] = 1.4963 0.1526*dss + 0.0019*dss2 0.78**
2002 PCF 112 Ln[N03-N] = 2.3555 0.2012*dss + 0.0028*dss2 0.72**
2002 PCF 168 Ln[N03-N] = 2.8498 0.1464*dss + 0.0018*dss2 0.69**
2002 PCF 224 Ln[N03-N] = 2.3856 0.1461*dss + 0.0021*dss2 0.47**
2002 PCF 280 Ln[N03-N] = 2.5739 0.1652*dss + 0.0023*dss2 0.76**
2002 PCG 0 Ln[N3-N] = 2.4596 0.1420*dss + 0.0019*dss2 0.82**
2002 PCG 112 Ln[N03-N] = 3.3709 0.1716*dss + 0.0023*dss2 0.72**
2002 PCG 168 Ln[N03-N] = 2.8677 0.1660*dss + 0.0024*dss2 0.64**
2002 PCG 224 Ln[N03-N] = 2.6941 0.1917*dss + 0.0032*dss2 0.77**
2002 PCG 280 Ln[N03-N] = 3.3876 0.1467*dss + 0.0023*dss2 0.59**
2002 PSF 0 Ln[N3-N] = 1.4101 0.1532*dss + 0.0024*dss2 0.76**
2002 PSF 112 Ln[N03-N] = 2.1226 0.1455*dss + 0.0018*dss2 0.89**
2002 PSF 168 Ln[N03-N] = 2.6749 0.1821*dss + 0.0025*dss2 0.87**
2002 PSF 224 Ln[N03-N] = 2.2218 0.1242dss + 0.0015*dss2 0.70**
2002 PSF 280 Ln[N03-N] = 3.1244 0.1497*dss + 0.0020* dss2 0.81**
2002 PSG 0 Ln[N3-N] = 2.0610 0.1320*dss + 0.0017* dss2 0.79**
2002 PSG 112 Ln[N03-N] = 2.8405 0.1818*dss + 0.0024* dss2 0.63**
2002 PSG 168 Ln[N03-N] = 3.4428 0.1793*dss + 0.0026* dss2 0.58**
2002 PSG 224 Ln[N03-N] = 2.9767 0.1836*dss + 0.0027* dss2 0.55**
2002 PSG 280 Ln[N03-N] = 2.9657 0.1306*dss + 0.0019*dss2 0.49**


dss= Days since side-dress
ns = No significant

* = Significant to 0.05
**= Significant to 0.01






N03-N concentration in soil


Fig. 2-8a

Year Conc. Intercept Linear term Quadratic term R2
mg.kg
2001 [N03-N] = 10.8547 + 0.1953 Nrate 0.53**
2003 [N03-N] = 3.2588 + 0.0049 Nrate + 0.000042 Nrate2 0.57**


Fig. 2-8b

Year Conc Intercept Linear term Quadratic term R2
mg.kg
2001 [N03-N] = 1.7392 + 0.0077 Nrate 0.0077 Nrate2 0.51*
2002 [N03-N] = 2.0927 + 0.0263 Nrate 0.000023 Nrate2 0.27**
2003 [N03-N] = 5.2414 + 0.0691 Nrate 0.17**













NH4-N concentration in soil


Fig. 2-9a

Year Conc. Intercept Linear term Quadratic term R2
mg.kg
2001 [N03-N] = 5.1498 0.0671 Nrate + 0.0006 Nrate2 0.64**
2003 [N03-N] = 2.1659 0.0558 Nrate + 0.00037 Nrate2 0.55**

Fig. 2-9b

Year Conc. Intercept Linear term Quadratic term R2
mg.kg
2001 [N03-N] = 5.0331 0.0052 Nrate 0.08*
2002 [N03-N] = 0.6143 + 0.0049 Nrate 0.25**
2003 [N03-N] = 1.0874 + 0.0166 Nrate 0.13*
Nrate N fertilization rate (kg.ha-')
ns = No significant
* = Significant to 0.05
** = Significant to 0.01



Crops N uptake

Sorghum. Fig. 2-10

Year 1ae Intercept Linear term Quadratic term R2

2001 TKN = 10.8858 + 0.3293 Nrate 0.36**
2002 TKN = 17.9582 0.0659 Nrate + 0.00060 Nrate2 0..36**
2003 TKN = 8.3331 0.0701 Nrate + 0.00054 Nrate2 0.42**

Potato. Fig. 2-11

Year N Intercept Linear term Quadratic term R2

2001 TKN = 111.3804 0.3714 Nrate 0.23**
2002 TKN = 37.2222 + 0.8693 Nrate 0.0018 Nrate2 0.45**
2003 TKN = 28.0746 + 0.4817 Nrate 0.51**


Nrate
TKN
ns
*
**


=N fertilization rate (kg.ha-1)
Total Kjeldahl Nitrogen
No significant
Significant to 0.05
Significant to 0.01











Predicted N rate for maximum potato yield


Fig. 2-12

Year Crop NYmax
Rotation (kg.ha 1) Intercept Linear term Quadratic term R2
2001 Pooled NYmax = 20.1574 Nrate + 0.1640 Nrate 0.00050 Nrate2 0.38**
2002 PCF NYmax = 8.4584 Nrate + 0.1455 Nrate 0.00035 Nrate2 0.81**
2002 PCG NYmax = 12.7438 Nrate + 0.0884 Nrate 0.00021 Nrate2 0.48**
2002 PSF NYmax = 7.4520 Nrate + 0.1968 Nrate 0.00047 Nrate2 0.72**
2002 PSG NYmax = 8.9293 Nrate + 0.1701 Nrate 0.00043 Nrate2 0.91**
2003 PCF NYmax = 2.9133 Nrate + 0.1385 Nrate 0.00032 Nrate2 0.79**
2003 PCG NYmax = 3.9586 Nrate + 0.1506 Nrate 0.00040 Nrate2 0.92**
2003 PSF NYmax = 3.9419 Nrate + 0.1680 Nrate 0.00041 Nrate2 0.84**
2003 PSG NYmax = 4.2412 Nrate + 0.1553 Nrate 0.00033 Nrate2 0.86**
NYmax = Predicted N rate for maximum yield (kg.ha-1)
Nrate = N fertilization rate (kg.ha-1)
ns =No significant
* = Significant to 0.05
** = Significant to 0.01















APPENDIX B
ANOVA TABLES FOR POTATO YIELD PARAMETERS

In this appendix are reported ANOVA tables for potato yield variables during the

study period. These tables include rot potatoes, classification by size, marketable yield,

total yield, and specific gravity. Due to the size of the tables, they are included beginning

in the next page.












Table B-1. 2001 ANOVA table for potato yield parameters


Type III Mean Square 2
Variation Rots Damaged B Al A2 A3 A4 Yield Yield Specific gravity
Replication 3 15.68** 1.03 5.45* 63.59 609.01* 346.05 0.21 2489.32** 2126.80** 0.000030*

Rotation 1 3 4.57* 13.60 3.94 42.14 29.65 246.05 0.59 213.48 312.33 0.000008

Main Plot Error 9 1.11 3.85 2.24 54.33 132.33 109.92 0.61 275.72 259.53 0.000006

Nrate 4 5.85** 34.70** 1.04 192.09** 963.69 580.91 1.02 3662.42** 3891.34** 0.000174**

Rotation N rate 12 0.58 2.69 2.42 36.95 168.39 37.32 0.61 301.46 275.77 0.000008

Error 79 1.40 2.75 1.44 32.25 75.78 46.89 0.47 172.48 179.87 0.00000462

Main plot effects tested using main plot error
2 *Significant at Type I error < 0.05
**Significant at Type I error < 0.01












Table B-2. 2002 ANOVA table for potato yield parameters


Source of DF Type III Mean Square 2 Total
DF Market Total
Variation Rots Damaged B Al A2 A3 A4 Yield Yield Specific gravity
Replication1 2 13.49 1.63 5.97* 18.51 425.25 2.78 0.030 604.14 657.55 0.000005

Rotation 1 3 16.31 1.64 4.49* 25.14 231.25 10.46 0.039 211.00 126.12 0.000027

Main Plot Error 6 6.17 5.39 2.80 55.42 313.14 6.71 0.067 191.18 227.68 0.000006

N rate 4 10.08* 12.04** 7.39** 765.81** 980.66** 7.30* 0.043 3676.50** 4266.38** 0.000234**

Rotation N rate 12 5.52 0.74 1.10 49.25** 36.28 1.32 0.063 126.23** 152.94** 0.000004

Error 59 6.96 0.79 1.28 15.31 33.08 2.09 0.06 37.67 29.76 0.000004

Main plot effects tested using main plot error
2 *Significant at Type I error < 0.05
**Significant at Type I error < 0.01












Table B-3. 2003 ANOVA table for potato yield parameters


Source of DF Type III Mean Square 2 Total
DF Market Total
Variation Rots Damaged B Al A2 A3 A43 Yield Yield Specific gravity
Replication1 3 11.54 3.39 15.76* 101.77 129.12* 72.31* ---- 188.22 342.64 0.000047**

Rotation 1 3 17.10 0.54 2.44 101.99 111.12* 14.65 ---- 431.07* 346.41 0.000109**

Main Plot Error 9 11.75 4.03 2.56 77.01 27.97 12.61 ---- 102.88 101.65 0.000006

Nrate 4 14.84** 3.73* 13.09** 192.09 533.78** 27.23** ---- 5299.24** 6704.77** 0.000174**

Rotation N rate 12 3.17 2.45* 1.00 36.95 29.85 4.75 ---- 67.15* 65.45* 0.000008

Error 79 2.42 1.08 0.63 14.67 16.29 3.80 ---- 32.18 32.55 0.000038

Main plot effects tested using main plot error
2 *Significant at Type I error < 0.05
**Significant at Type I error < 0.01
3 In this year low production of potatoes A4 size did not allowed to analyze this variable















APPENDIX C
GREEN BEAN YIELD


Potato Previous crop
Year N rate2 Sorghum Cowpea
(kg.ha-') Green bean yield (kg.ha'1)
0 5428 7355
100 5736 6653
2001 150 6175 6351
200 5754 6705
250 5196 7294
0 2205 3519
100 1909 3671
2002 150 1631 3467
200 2119 3220
250 1299 2540


2003


100 3034 3095
150 2839 3077
200 3622 3513
250 3287 3269


1 Crop preceding green bean in the crop rotation.
2 N rate applied to potato in spring.


2860


2888
















LIST OF REFERENCES


Asfary, A.F.; Wild, A.; Harris, P.M. Growth, Mineral Nutrition and Water Use By Potato
Crops. J. Agric. Sci. 1983, 100 (1), 87-101.

Bishop, J.C.; Grimes, D.W. Precision Tillage Effects on Potato Root and Tuber
Production. American Potato J. 1978, 55, 65-71.

Campbell, K.L.; Rogers, J.S.; Hensel, D.R. Water Table Control for Potatoes in Florida.
Transactions of the ASAE. 1978, 21 (4), 701-705.

Campbell, K.L.; Rogers, J.S.; Hensel, D.R. Drainage Water Quality from Potato
Production. 1985, 28 (6), 1798-1801

De Roo, H.C.; Waggoner, P.E. Root Development of Potatoes. Agron. J. 1961, 53, 15-17.

Drew, M.C.; Jackson, M.B.; Giffard, S. Ethylene-promoted Adventitious Rooting and
Development of Cortical Air Spaces (aerenchyma) in Roots May be Adaptive
Responses to Flooding in Zea mays L. 1979, 147, 83-88.

Elkashif, M.E.; Locascio, S.J.; Hensel, D.R. Isobutylidene Diurea and Sulfur-coated Urea
as N Sources for Potatoes. J. Amer. Soc. Hort. Sci. 1983, 108 (4), 523-526.

Errebhi, M.; Rosen, C.J.; Gupta, S.C.; Birong, D.E. Potato Yield Response and Nitrate
Leaching as Influenced by Nitrogen Management. Agron. J. 1998, 90, 10-15.

Errebhi, M.; Rosen, C.J.; Lauer, F.I.; Martin, M.W.; J.B. Bamberg. Evaluation of Tuber-
Bearing Solanum Species for Nitrogen Use Efficiency and Biomass Partitioning.
Am. J. Potato Res. 1999, 76, 143-151.

Gao, S.; Pan, W.L.; Koenig, R. T. Integrated Root System Age in Relation to Plant
Nutrient Uptake Activity. Agron. J. 1998, 90, 505-510.

Goolsby, D.A.; Battaglin, W.A.; Aulenbach, B.T.; R.P. Hooper. N Input to the Gulf of
Mexico. J. Env. Qual. 2001, 30, (2) 329-336.

Goulding, K. Nitrate Leaching from Arable and Horticultural Land. Soil Use and
Management. Special issue: Tackling Nitrate from Agriculture. 2000, 16, (Suppl.)
145-151.

Guddanty, S.; Chambers, J.L. GSRoot.-Automated Root Length Measurement Program.
Users Manual. Version 5.00. Louisiana State University. Agricultural Center. 1993,