|Table of Contents|
Table of Contents
List of Tables
List of Figures
Chapter 1. Introduction
Chapter 2. Screening vesicular-arbuscular mycorrhizal fungi for aluminum on their spore germination and root penetration
Chapter 3. Effect of isolates of vesicular-arbuscular mycorrhizal fungi selected for aluminum tolerance on field growth of Phaseolus vulgaris
Chapter 4. Characteristics of germination, hyphal growth and root penetration of Glomus manihotis, Glomus mosseae, Acaulospora longula and Scutellospora pellucida
Chapter 5. Conclusions
Appendix A. Soil characteristics of the isolation site of the vesicular-arbuscular mycorrhizal species and isolates
Appendix B. Production of pot cultures
Appendix C. Additional screening data
THE EFFECT OF PHYTOTOXIC LEVELS OF ALUMINUM ON THE GROWTH OF SELECTED SPECIES OF VESICULAR-ARBUSCULAR MYCORRHIZAL FUNGI
ANNE WILLIAMS BARKDOLL
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
The author is grateful to Dr. Norman C. Schenck, chairman of the supervisory committee, for his support and encouragement throughout the period of study. She also thanks Dr. David J. Mitchell, Dr. David H. Hubbell, Dr. Laurence H. Purdy, Dr. Roy D. Rhue and Dr. William Blue (former committee member) for their helpful comments.
She would also like to thank Dr. Hugh Popenoe and the Office of International Programs who provided some of the funds which helped make this research possible.
Special thanks go to Dr. Judy Kipe-Nolt, the microbiologist in
the Bean Program, at the Centro Internacional de Agricultural Tropical (CIAT) in Cali, Colombia, where much of the research was conducted. She facilitated many aspects of the research and always lent a sympathetic ear. The help of the members of the Bean Microbiology Laboratory was always much appreciated as were the good times spent working with them in the field. The work also would not have been possible without the help of Dr. Ewald Sieverding and the members of the Mycorrhizae Project at CIAT.
Finally, the author is very grateful to her friends and family who have been so supportive. She especially wishes to thank Meg Niederhofer, Brigitte Maass, Veronique Schmidt, Libardo Bello Vega and Rebeca Rivera. She also wishes to thank her parents who were always supportive even when pursuing her research meant long absences.
TABLE OF CONTENTS
ACKNOWLEDGMENTS ............................................... ii
LIST OF TABLES ............................................ .... v
LIST OF FIGURES .................... o...........o................ vii
ABSTRACT ......i..... ............. ........ ... ....... x
CHAPTER I GENERAL INTRODUCTION ............................ 1
CHAPTER II SCREENING VESICULAR-ARBUSCULAR MYCORRHIZAL FUNGI
FOR ALUMINUM TOLERANCE AND THE EFFECTS OF
ALUMINUM ON THEIR SPORE GERMINATION AND ROOT
PENETRATION, ....o.....o........ ............ .....o. 10
Introduction ..... ..... -e ee.................. 10
Materials and Methods....... ........ .......... 11
Results ........ ..... 0..... 0 -0.... . ....... 24
Discussion ................. ........ o......o... .. 59
CHAPTER III EFFECT OF ISOLATES OF VESICULAR-ARBUSCULAR
MYCORRHIZAL FUNGI SELECTED FOR ALUMINUM TOLERANCE
ON FIELD GROWTH OF Phaseolus vulgaris ........... 74
Materials and Methods... .. ...... ..o . ..... 75
Results ... o....................... 0.....0...... 84
Discussion ..... o.... ....... ... *... 0...... ...... 102
CHAPTER IV CHARACTERISTICS OF GERMINATION, HYPHAL GROWTH
AND ROOT PENETRATION OF Glomus manihotis, Glomus
mosseae, Acaulospora loneula and Scutellospora
pellucida ..... ................. ...... .. ...... 107
Materials and Methods ... ... ....... .... .... 108
Results, ......-.....o .. .... 108
Discussion, ..... -i .oi ........................ 117
CHAPTER V CONCLUSIONS ............ ..... ............ ..... 122
APPENDIX A SOIL CHARACTERISTICS OF THE ISOLATION SITE OF
THE VESICULAR-ARBUSCULAR MYCORRHIZAL SPECIES AND
ISOLATES ................ ...............o........ 126
APPENDIX B PRODUCTION OF POT CULTURES ...................... 128
APPENDIX C ADDITIONAL-SCREENING DATA ....................... 131
REFERENCES ................................................... 140
BIOGRAPHICAL SKETCH .................................. o..* ..... 148
LIST OF TABLES
1 Origin of isolates of vesicular-arbuscular mycorrhizal
fungi from the collection at the Centro Internacional de
Agricultura Tropical and characteristics of the soil
from which they were originally obtained .................. 12
2 Chemical properties of five Typic Dystropepts from
Mondomito, Cauca, Colombia ...................... ......... 13
3 Chemical analyses of a Typic Haplustox 2 1/2 months after
liming ................0"................................ 18
4 Modified Hoagland's nutrient solution used in the
aluminum solution experiment ........... ...... .......... 22
5 Correlation coefficients for plant parameters with
milliequivalents of soil aluminum and calcium ............. 25
6 Correlation coefficients for fungal parameters with
milliequivalents of soil aluminum and calcium of selected
isolates of vesicular-arbuscular mycorrhizal fungi ........ 26
7 Correlation coefficients of selected isolates for fungal
parameters with plant parameters .......................... 31
8 Mean percent spore germination of vesicular-arbuscular
mycorrhizal (VAM) fungi across all soil types naturally
varying in percent aluminum saturation.................... 32
9 Mean number of hyphal penetration sites formed per 100
centimeter of root as affected by isolates of vesiculararbuscular mycorrhizal fungi and soil .................... 33
10 Correlation coefficients for fungal parameters correlated with soil chemical factors in limed soil .................. 37
11 Correlation coefficients for fungal parameters in limed soil ...................................................... 39
12 Correlation coefficients for fungal parameters associated with plant parameters in limed soil ....................... 40
13 Mean percent spore germination across all aluminum levels for both solution experiments ............................ 55
14 Means of hyphal penetration sites across all aluminum levels for both solution experiments..................... 55
15 Mean number of penetration sites per centimeter of root for each isolate of vesicular-arbuscular mycorrhizal
fungi across all aluminum levels for solution experiment
16 Soil chemical analyses before and after liming of field sites in Mondomito, Cauca, Colombia....................... 76
17 Number of propagules of vesicular-arbuscular mycorrhizal fungi in field soil and inoculum as indicated by the most
probable number test ....................... ............... 85
18 Summary of fungal growth characteristics in regard to spore germination, hyphal growth and root penetration..... 119
A-1 Soil characteristics for the isolation site of the species and isolates of vesicular-arbuscular mycorrhizal
fungi ..................................................... 126
B-1 Analyses of soils used for pot cultures................... 129
C-1 Correlation coefficients for spore germination, hyphal growth and root penetration for the second experiment
with soils naturally varying in milliequivalents of
aluminum .................................................. 131
C-2 Mean values for percent spore germination, hyphal growth and root penetration for the isolates of vesiculararbuscular mycorrhizal (VAM) fungi in the second experiment
with soils which naturally vary in milliequivalents of
aluminum........................... ........ ..... ....... 133
C-3 Correlation coefficients of isolates of vesiculararbuscular mycorrhizal fungi with plant parameters in the
second experiment with soils naturally varying in
milliequivalents of aluminum............................o. 135
C-4 Correlation coefficients for spore germination, hyphal growth and root penetration for the first lime experiment. 136
C-5 Correlation coefficients of isolates of vesiculararbuscular mycorrhizal fungi with plant parameters for
the first lime experiment.............. ..... ............. 137
C-6 Mean values for percent spore gemination, hyphal growth and root penetration for the isolates of vesiculararbuscular mycorrhizal fungi in the first lime experiment. 138
LIST OF FIGURES
1 Illustration of the system used in nutrient solution
studies ........................................ ........... 21
2 Percent spore germination by vesicular-arbuscular
mycorrhizal fungi as affected by milliequivalents of
aluminum in the soil...................................... 28
3 Root penetration by vesicular-arbuscular mycorrhizal fungi
as affected by milliequivalents of aluminum in the soil... 30
4 Effect of lime on shoot and root dry weight of Phaseolus
vulgaris.................. ...... ............. ............ 35
5 Lime effects on spore germination, root penetration and
hyphal growth of Glomus mosseae c-30...................... 42
6 Lime effects on spore germination, root penetration and
hyphal growth of Scutellospora pellucida c-44-............ 43
7 Lime effects on spore germination, root penetration and
hyphal growth of Scutellospora pellucida c-3-7............ 44
8 Lime effects on spore germination, root penetration and
hyphal growth of Acaulospora longula c-17-2 .............. 45
9 Lime effects on spore germination, root penetration and
hyphal growth of Acaulospora longula c-12-1................ 46
10 Lime effects on spore germination, root penetration and hyphal growth of Acaulospora longula c-94-................ 47
11 Lime effects on spore germination, root penetration and hyphal growth of Glomus manihotis c-17-................... 49
12 Lime effects on spore germination, root penetration and hyphal growth of Glomus manihotis c-20-................... 50
13 Lime effects on spore germination, root penetration and hyphal growth of Glomus manihotis c--1.................. 51
14 Effect of solution aluminum on shoot dry weight of Phaseolus vulgaris................ ....... ................. 52
15 Effect of solution aluminum on root aluminum of Phaseolus vulgaris. ..................... ...... .......... 54
16 Hyphal growth of vesicular-arbuscular mycorrhizal fungi as affected by solution aluminum ......... ................. 57
17 Percent spore germination of vesicular-arbuscular mycorrhizal fungi as affected by solution aluminum ........ 58 18 Aborted germ tubes of Scutellospora pellucida............ 61
19 Number of root penetration sites by vesicular-arbuscular mycorrhizal fungi as affected by solution aluminum ........ 62 20 Bean shoot weight in field one at three sampling dates as affected by mycorrhizal treatment and lime ................ 86
21 Bean shoot weight in field two at three sampling dates as affected by mycorrhizal treatment and lime................ 87
22 Bean root length in field one as affected by mycorrhizal treatment, lime and sampling date .............-........... 89
23 Bean root length in field two as affected by mycorrhizal treatment, lime and sampling date ............o......... 90
24 Dry weight of bean roots in field two as affected by mycorrhizal treatment, lime and sampling date ............. 91
25 Dry weight of bean roots in field one as affected by mycorrhizal treatment, lime and sampling date............ 92
26 Percent colonization of bean roots by vesicular-arbuscular mycorrhizal fungi in field one at three sampling dates.... 93 27 Percent colonization of bean roots by vesicular-arbuscular mycorrhizal fungi in field two at three sampling dates.... 94 28 Total bean shoot phosphorus in field one as affected by mycorrhizal treatment, lime and sampling date ............. 96
29 Total bean shoot phosphorus in field two as affected by mycorrhizal treatment, lime and sampling date ............. 97
30 Total bean leaf aluminum in field one as affected by mycorrhizal treatment, lime and sampling date............. 98
31 Total bean leaf aluminum in field two as affected by mycorrhizal treatment, lime and sampling date ............. 99
32 Bean plant population in the field at harvest as affected by vesicular-arbuscular mycorrhizal treatment and lime.... 100 33 Bean yield of Phaseolus vulgaris at harvest as affected by vesicular-arbuscular mycorrhizal treatment and lime.... 101 34 Germination of spores of vesicular-arbuscular mycorrhizal fungi through the subtending hyphae....................... 109
35 Characteristics of hyphal growth by vesicular-arbuscular mycorrhizal fungi............................ ............. 111
36 Spore germination and root penetration by Scutellospora pellucida ................................................. 113
37 Differences among spore walls of isolates of Acaulospora longula mounted in lactophenol............................ 115
38 Germination of Acaulospora longula c-12-1................. 116
39 New spore formation by Acaulospora longula................ 118
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
THE EFFECT OF PHYTOTOXIC LEVELS OF ALUMINUM ON THE GROWTH OF
SELECTED SPECIES OF VESICULAR-ARBUSCULAR MYCORRHIZAL FUNGI
ANNE WILLIAMS BARKDOLL
Chairman: N. C. Schenck
Major Department: Plant Pathology
Aluminum in soil and nutrient solution was evaluated for its
effects on the germination, hyphal growth and root penetration by 10 isolates of the following four species of vesicular-arbuscular mycorrhizal (VAM) fungi: Glomus mosseae, Glomus manihotis, Scutellospora pellucida, and Acaulospora longula. In greenhouse experiments germination, hyphal growth and root penetration were not always affected to the same degree by aluminum. Germination was usually less susceptible to increasing aluminum concentrations than were hyphal growth and penetration. Isolates within a species did not respond in the same manner always. Spore germination and root penetration of Glomus manihotis c-1-i and c-17-1 were positively correlated with milliequivalents of soil aluminum in repeat experiments while G. manihotis c-20-2 was not. Spore germination, hyphal growth and root penetration by G. mosseae c-30-1 were positively correlated consistently with milliequivalents calcium. Other isolates usually showed no consistent correlation with soil aluminum and calcium.
These fungi were also screened for tolerance to aluminum in soil and nutrient solution procedures for the purpose of field inoculum application. In soil, some isolates like G. manihotis c-1-1 had root penetration positively correlated with milliequivalents aluminum but in solution the addition of 0.5 ppm aluminum depressed root penetration. The two soil procedures, one with soils which naturally varied in milliequivalents aluminum and the other with limed soils, gave similar results. The results of the nutrient solution studies often did not support the soil procedures.
All species tested produced auxiliary cells after germinating and all produced appressoria before root penetration. Acaulospora longula appears to form compartments in the spore wall during germination, similar to the genus Scutellospora.
Field experiments were conducted to determine the effects of
isolates, which differed in their responses to aluminum, on the growth and yield of Phaseolus vulgaris under conditions of aluminum stress. Most probable number (MPN) values of native VAM fungi in both fields decreased with the addition of 3 tons of lime. In the field, at 0 tons lime/ha, introduced VAM fungi did not increase yields above those produced in plots with only native VAM fungi. As lime levels increased to 3 and 6 tons/ha, A. longula c-12-1 and G. manihotis c-1-1, selected for growth correlated to milliequivalents calcium and aluminum, respectively, increased bean seed yields when compared to the native VAM fungi in only one field. The MPN values of the inoculum of these two isolates were the highest of the field-tested isolates.
Fluctuations in hydrogen ion concentration are known to affect vesicular-arbuscular mycorrhizal (VAM) fungi and the subsequent development of mycorrhizae. Researchers have examined the effects of pH, often controlled by lime, on spore germination (Daniels and Trappe, 1980; Green et al., 1976; Siqueira et al., 1982; Siqueira et al., 1984), germ tube growth (Siqueira et al., 1984; Siqueira et al., 1982), mycorrhizal efficiency (Davis et al., 1983; Graw, 1978; Hayman and Tavares, 1985), root colonization (Hayman and Tavares, 1985; Killham and Firestone, 1983; Wang, 1984) and mycorrhizal populations (Wang, 1984). Different species of VAM fungi respond differently to acid and alkaline soils. Glomus mosseae is notoriously sensitive to an acid pH (Green et al., 1976; Siqueira et al., 1984; Davis et al., 1983; Hayman and Tavares, 1985) while Glomus fasciculatum functions well over a broad pH range but functions best under more acid conditions (Davis et al., 1983; Hayman and Tavares, 1985). Consequently, plant response to inoculation with mycorrhizal fungi at a given pH depends not only on the plant species but also the species of VAM fungi (Davis et al., 1983; Graw, 1978; Hayman and Tavares, 1985).
Acidic soils may be caused by several factors which include leaching and soil parent material low in bases. Acidic soil conditions are often adverse to plant growth. The plant growth
problems in acid soils may be caused by factors such as toxicities of aluminum (Al) and manganese (Mn) and deficiencies of phosphorus (P), magnesium (Mg), calcium (Ca) and molybdenum (Mo). The pH must be less than 3.0 before the H+ concentration is itself phytotoxic to most plants (Bohn et al., 1979). Although cadmium (Cd), nickel (Ni), copper (Cu), and lead (Pb) may be present in toxic quantities in certain soils, such as mine spoils, the principal elements which cause toxicity to plants in acid soils are Al and Mn. Of the two, Al is more toxic (Bohn et al., 1979) and in addition causes an increase in acidity through hydrolysis reactions and subsequent release of protons (Bohn et al., 1979; Evans and Kamprath, 1970). Raising soil pH using lime is primarily done to neutralize exchangeable and solution Al and thus relieve Al toxicities to the plant (Cochrane et al., 1980; Haynes, 1984). Since Al, Mn, Ca, P, Mg, iron (Fe) and the levels of other nutrients fluctuate with the addition of lime, it is not possible to separate empirically the effects of Al from those of other elements in the soil. The question arises whether the effects of hydrogen ion concentration on VAM fungi observed by researchers are really effects of Al or other soil elements whose levels vary with pH; hydrogen ion concentration per se between pH 4.1 and pH 8.0 does not limit the growth of most crops (Arnon and Johnson, 1942; Foy, 1974; Moore, 1974).
Aluminum toxicity may occur at a pH as high as 5.5 but usually occurs below pH 5.0. Plants suffering from Al toxicity usually have stubby, thickened roots, inhibition of lateral root development and no fine roots (Hecht-Buchholz and Foy, 1981; Lee and Pritchard, 1984).
Leaves may be curled, small and yellowing or purplish. The following mechanisms of action of Al toxicity have been reported: reduction of P uptake by fixation on root surfaces, in root-free space and in soil; interference in cell division, cell wall deposition and nutrient uptake; reduced DNA replication; increased cell wall rigidity and altered root cell membrane structure (Klimashevskii et al., 1979; Hecht-Buchholz and Foy, 1981; McCormick and Borden, 1974). Phosphorus and Al apparently are adsorbed at the same sites on the root and it is thought that P may precipitate as insoluble aluminum phosphates in the
root or be adsorbed by hydroxy-Al polymers already precipitated in the root-free space (McCormick and Borden, 1974; Naidoo et al., 1978). Imbalance of plant nutrient uptake is reported most commonly as reductions in Ca and P uptake but may also occur as reduced uptake of Mg, potassium (K), Cu, zinc (Zn) and Mn (Andrew et al., 1973; Clark, 1977; Lee, 1971; Long and Foy, 1970).
Aluminum induced growth reduction of plants has been demonstrated in both nutrient solution and field soil (Foy et al., 1967; Foy et al., 1972; Foy et al., 1974). McCain and Davis (1985) showed that increasing Al concentration of a solution increased the number of tillers of Agrostis capillaris L. that failed to form roots. Microscopic examination of roots of Al tolerant and susceptible barley cultivars after exposure to 9 ppm Al showed destruction of the root cap, epidermal and cortical cells. Electron microscopy showed incipient disorganization of the plasmalemma in resistant roots and more mucigel around resistant roots than around susceptible roots. Changes also occurred later in the resistant variety than in the
susceptible variety. The authors proposed that the greater mucigel production may be responsible for resistance to Al (Hecht-Buchholz and Foy, 1981). Aluminum is not confined to the root; translocation to the shoots varies with different species. Translocation is thought to occur both symplastically and apoplastically. Aluminum seems to accumulate first in the roots and then after a threshold is reached an increase in translocation to the shoot occurs (Wagatsuma, 1984). Wagatsuma (1984) found a high Al accumulation in the epidermis, hypodermis, cortex and endodermis with little in the stele. Varieties susceptible to Al often seem to have more Al in the roots and less Ca in the shoots than varieties resistant to Al (Foy et al., 1967; Foy et al., 1972).
Aluminum has been shown to be toxic to organisms other than plants. Somers (1961) examined the fungitoxicity of metal ions to Alternaria tenuis and Botrytis fabae. The median effective doses (Ed50) for A13+ were 1.3 x 10-5 and 1.4 x 10-4 depending on the fungus tested. Ko and Hora (1972), working with a soil fungistatic to Neurospora tetrasperma, extracted a fungitoxin that was deactivated by raising the pH or by treatment with cation and anion resins. They identified the toxin as Al and found that a solution with 0.65 ppm Al at pH 4.8 completely inhibited ascospore germination. Aluminum was fungicidal at the concentration found in the soil unless the spores were dormant when exposed to it. In the case of Aphanomyces euteiches, Al and other metal ions reduced fungal growth and inhibited zoospore formation both by 95%. Aluminum was toxic to growth at 1 Ug m1- and to zoospore formation at 10 Ug ml-I. On the other hand,
zoospore germination was not inhibited by these Al levels (Lewis, 1973). Orellana et al. (1975) examined the effect of Al on the growth of Verticillium albo-atrum and Sclerotinia sclerotiorum and their pathogenicity on sunflower. Aluminum caused complete inhibition of growth of V. albo-atrum above 8 ppm Al and below that caused very thin pigmentless threadlike hyphae to form. Sclerotinia sclerotiorum was quite tolerant of Al in the range of 4 to 32 ppm Al and produced moderate to abundant growth. Disease induced by V. albo-atrum was more severe at higher soil pH levels while just the reverse was true for S. sclerotiorum (Orellana et al., 1975). Muchovej et al. (1980) examined the effects of Al on Phytophthora capsici and its pathogenicity in green pepper. Increasing amounts of A13+ from 0.4 to
0.6 meq A13+ caused sequential decreases in mycelial growth while pH alone of 4.1 or greater did not affect mycelial growth. Disease severity was found to increase when soil was limed with Ca(OH)2 of CaCO3 (Muchovej et al., 1980). Finally, Wang (1984) studied the effects of Al and Mn on infection produced by the VAM fungus Glomus caledonicum. She used a sand culture with four levels of Al and Mn. The oat hosts were preinfected and then planted in sand. Infection by G. caledonicum was strongly inhibited by Al; Mn was five times less inhibitory and the hydrogen ion concentration tested did not inhibit infection (Wang, 1984). Since Al also destroys plant roots, it is difficult to determine if the Al directly affected the fungus or if the effect was indirect and reduced infection occurred due to host damage.
The effects of heavy metals on ericoid and ectomycorrhizae have been researched (Bradley et al., 1981; Brown and Wilkins, 1985; Jones and Hutchinson, 1986). For plants forming ericoid and ectomycorrhizae, the fungi seem to provide some protection for the plants by concentrating the toxic metal in the roots rather than allowing it to be translocated to the shoots. This is of interest when compared to the work of Wagatsuma (1984), which indicated Al was not translocated to the shoots until root absorption sites were saturated. Brown and Wilkins (1985) found that ectomycorrhizal Betula spp. had a higher Zn concentration in the roots and a lower concentration in the shoots than the nonmycorrhizal control. The same response was found with mycorrhizal Betula papyrifera with regard to Ni but not Cu. In the case of Cu, mycorrhizal inoculation did not increase growth over the nonmycorrhizal control at low Cu levels and at high Cu levels inoculation reduced growth without increasing Cu uptake (Jones and Hutchinson, 1986). With Calluna vulgaris, which forms ericoid mycorrhizae, Zn and Cu were found to reduce growth of nonmycorrhizal plants significantly compared to the growth made by mycorrhizal plants. Levels of these metals in the mycorrhizal roots were significantly higher than the levels in the shoots. At high metal levels, not enough nonmycorrhizal roots were formed to be analyzed (Bradley et al., 1981).
In the case of VAM fungi, metals other than Al have been studied; these include Zn and Cu, and less commonly, Fe, Ni, Pb, Mn and Cd (Mcllveen et al., 1975; Lambert et al., 1979; Killham and Firestone, 1983; Hepper and Smith, 1976; Gildon and Tinker, 1983b; Gildon and
Tinker, 1983a; Gildon and Tinker, 1981; Benson and Covey, 1976). The situation in regard to heavy metals with VAM fungi seems to be different from the ecto- and ericoid mycorrhizal fungi. Several researchers have found that, under conditions of heavy metal toxicity, fungal germination and infection, and mycorrhizal plant growth were reduced; metal concentrations in the mycorrhizal shoots were greater than in nonmycorrhizal shoots (Killham and Firestone, 1983; Gildon and Tinker, 1983a; Hepper and Smith, 1976; McIlveen et al., 1975). Killham and Firestone (1983) found that at pH 3.0 mycorrhizal shoots of Ehrhata calycina had about a 300% greater concentration of Ni and Cu than did the nonmycorrhizal shoots. Research by Gildon and Tinker (1983a) showed Zn and Ni values to be nonsignificantly higher in mycorrhizal plants than in nonmycorrhizal plants and that heavy metal additions strongly reduced root colonization. In the course of the work with heavy metals, differences in the behavior of isolates of the same species have been observed (Gildon and Tinker, 1981; Gildon and Tinker, 1983a; Hepper and Smith, 1976). Hepper and Smith (1976) observed that Mn and Zn could inhibit spore germination and hyphal growth. The inhibitory effect of Zn on germination depended on the isolate source. They suggested that differences in isolate reactions may reflect adaption to different soil conditions over time. In support of this idea, Gildon and Tinker obtained isolates of Glomus mosseae from heavy metal contaminated sites (Shipham strain). One of these isolates was compared to a G. mosseae isolate (Rothamsted strain) from a noncontaminated site which was sensitive to Ni, Cu, Cd, and Zn. As levels of Cd and Zn increased, the percent colonization by
the Rothamsted isolate decreased significantly while the Shipham isolate was significantly affected only at the highest levels of heavy metals. At 1000 pg Zn g-1 soil, plants inoculated with the Shipham strain had a significantly higher shoot weight than those inoculated with the Rothamsted strain. The greater shoot weight was not accompanied by a higher shoot metal concentration (Gildon and Tinker, 1981; Gildon and Tinker, 1983a).
In the utilization of VAM fungi in agriculture, it is necessary to understand the edaphic factors affecting these fungi and have isolates adapted to particular edaphic conditions. Several researchers have discussed the problems of field inoculations and the need for appropriate screening methods to select effective and adapted VAM fungi (Abbott and Robson, 1982; Abbott and Robson, 1981a; Kucey and Diab, 1984; Howeler and Sieverding, 1983; Azcon-Aguilar et al., 1986a; Powell, 1980; Powell, 1982). This is particularly important in view of the observed differences of isolates of the same species (Haas and Krikum, 1985; Hepper and Smith, 1976; Gildon and Tinker, 1981; Gildon and Tinker, 1983a). It is erroneous to assume, for example, that because some isolates of Glomus manihotis perform well under acid soil conditions all isolates will perform the same. Considerable research has involved P uptake by VAM fungi in low P soils where plants need VAM fungi for good growth. In these situations fertilizer requirements can often be reduced through the utilization of VAM fungi. Many low P soils are also acid and have Al toxicity problems. Since Al may cause problems of P fixation in the soil and P precipitation as insoluble aluminum phosphates in the root, it is
necessary to understand the effect of Al on VAM fungi and its subsequent effect on P nutrition. There is some evidence that Al3+ is quite effective at reducing percent colonization of the root by VAM fungi (Wang, 1984). Gildon and Tinker (1981, 1983a) found that isolates of VAM fungi tolerant to Zn produced greater shoot weights than nontolerant isolates under high levels of Zn. In view of these results mycorrhizal isolates tolerant to Al may prove to enhance plant
growth and Al tolerance may be an important factor in the selection of VAM fungi adapted to many acid soil conditions.
The objectives of this research were to determine 1) if aluminum rather than hydrogen ion concentration affected the growth of VAX fungi; 2) if spore germination, hyphal growth and root penetration of the VAX fungi were all affected by aluminum in the same manner; 3) if some species or isolates were adapted to aluminum; 4) which screening procedure for aluminum tolerance functioned best; and 5) if isolates of VAM fungi selected for aluminum tolerance benefitted the host under conditions of aluminum stress.
SCREENING VESICULAR-ARBUSCULAR MYCORRHIZAL FUNGI
FOR ALUMINUM-TOLERANCE AND THE EFFECTS OF ALUMINUM
ON THEIR SPORE GERMINATION AND ROOT PENETRATION
Considerable research to date has examined the effects of pH or lime on VAM fungi (Daniels and Trappe, 1980; Green et al., 1976; Siqueira et al., 1984; Siqueira et al., 1982; Graw, 1978; Hayman and Tavares, 1985), but only that of Wang (1984) has specifically separated the effects of Al from those of hydrogen ion concentration on VAM fungi. She used infected host roots exposed to various concentrations of Al and Mn in solution. Increasing either Al or Mn reduced root colonization by VAM fungi. Metals other than Al have also been shown to affect mycorrhizal root colonization and spore germination. Percent colonization by G. mosseae was reduced with increasing Zn and Cd (Gildon and Tinker, 1983a).
Aluminum is toxic to plants, fungi and other organisms (Somers, 1961). Agricultural soils are generally limed to reduce exchangeable Al (Evans and Kamprath, 1970) and thus relieve toxicity to the plant. In soil, Al, Ca and other elements fluctuate in availability as pH fluctuates.
The following experiments were designed to a) determine if
differences in tolerance to Al occur within and between species of VAM fungi, b) determine if spore germination and host penetration are
affected differentially by Al, c) separate the effects of pH and Al by using a nutrient solution, and d) develop a rapid screening method for VAM fungi adapted to phytotoxic Al levels.
Materials and Methods
The Effect of Soils Naturally Varying in Percent Al Saturation on Spore Germination of VA Mycorrhizal Fungi and Hyphal Penetration of Roots
For these and succeeding experiments 10 isolates of VA
mycorrhizal fungi representing four species were selected from the mycorrhizal fungal collection at the Centro Internacional de Agricultura Tropical (CIAT) in Cali, Colombia (Table 1). The four species selected were Acaulospora longula Spain & Schenck, Glomus mosseae (Nicol. & Gerd.) Gerd. & Trappe, G. manihotis Howeler, Sieverding & Schenck, and Scutellospora pellucida (Nicol. & Schenck) Walker and Sanders (Walker & Sanders, 1986; Koske & Walker, 1986). The isolates were selected on the basis of soil characteristics of their respective collection sites. For each species an isolate was selected from a low pH soil and a moderate to high pH soil. The species and CIAT isolate identification numbers are shown in Table 1 (see Appendix A for further information on species in Table 1).
Five soils that varied naturally in meq A13+ and percent Al saturation were selected from farms in the Mondomito region, Department of Cauca, Colombia. The meq of A13+ ranged from 0.30 to 4.74 meq Al 100 g-I soil, and the percent Al saturation from 9.6 to 69.2 (Table 2). The soil was a Typic Dystropept (Instituto Geografico Agustin Codazzi, 1976).
Table 1. Origin of isolates of vesicular-arbuscular mycorrhizal fungi
from the CIAT collection and characteristics of the soil
from which they were originally obtained.
Isolate Soil meq 1 100
Species Number Site and Origin pH g soila
Glomus mosseae c-30-1 Palmira, Valle, 7.6 0
G. mosseae c-131-2 Cajica, Cundinamarca, 6.6 0
Scutellospora c-44-2 Piangua Grande, 3.8 NAb
pellucida Buenaventura Valle,
S. pellucida c-3-7 Popayan, Cauca, 4.7 1.93
Acaulospora c-17-2 Media Luna, Magdalena, 5.7 0
A. longula- c-12-1 Carimagua, Meta, 4.3-4.5 1.7-2.4
A. longula c-94-5 Santander de 4.0 4.8
Glomus c-17-1 Media Luna, Magdalena, 5.7 0
G. manihotis c-20-2 Carimagua, Meta, 4.3-4.5 1.7-2.4
G. manihotis c-1-1 Santander de 4.0 4.8
aMeq exchangeable soil aluminum was determined by using a IN KC1 extracting solution.
bNA = Information not available
Table 2. Chemical properties of five Typic Dystropepts from Mondomito, Cauca, Colombia.
meg 100 g-1 soil d Concentration (ppm)
Soil ppm P Effective % Al Exchangeable Total
# OM Bray lIb pH Al Ca Mg K CECc saturation [H+] Mn Fe N
1 4.8 2.7 5.5 0.70 3.99 2.38 0.24 7.31 9.57 44.0 21.4 1904
2 9.6 2.9 5.7 0.30 1.12 0.43 0.36 2.21 13.57 23.9 1.7 4592
3 6.0 4.0 4.6 2.96 1.06 0.62 0.25 5.09 58.15 0.2 31.1 43.0 2576
4 5.0 4.6 4.9 2.52 0.92 0.32 0.12 .4.08 61.18 0.2 24.7 32.8 2016
5 5.2 3.1 4.2 4.74 0.04 0.43 0.14 6.85 69.20 0.5 36.0 165.0 2240
aoM = organic matter.
bBray II extraction method.
cCEC = cation exchange capacity.
percent Al saturation = (meq Al/ECEC) X 100.
Soil analysis data presented were determined by the methods
described below. Exchangeable soil aluminum and [H+] were determined by a 1N KCl extraction and then titration with 0.1 N NaOH to separate [H1+] from aluminum (McLean, 1965). If the soil pH was greater than pH
5.5, the soil was extracted with 1N ammonium acetate to determine Ca, Mg, and K (Chapman, 1965). If the soil pH was less than 5.5, the soil was extracted with 1N KCl to determine Al, Ca and Mg (Salinas and Garcia, 1985). In the latter case K was run on the modified Bray II extract; P was always determined using the modified Bray II extraction (Salinas and Garcia, 1985; Olsen and Dean, 1965). Soil pH was determined in a 1:1 soil-water ratio. Effective cation exchange capacity (ECEC) is the sum of the milliequivalents of H and Al, Ca, Mg and K. Percent Al saturation was calculated by dividing the meq of Al by the ECEC and multiplying by 100 (Evans and Kamprath, 1970). Total N was determined by the Kjeldahl method using the H 2so04 and salicylic acid digestion, (Bremner, 1965) and Mn and Fe were extracted with H 2SO04 and HCl (Salinas and Garcia, 1985). The Walkley-Black method was used in the analysis of percent organic matter (Allison, 1965).
The soils from Mondomito were taken to CIAT-Palmira, Department of Valle del Cauca, Colombia, where they were used in greenhouse studies. The soils were air dried, ground, coarsely sieved, wetted and steam pasteurized at approximately 80*C (Sylvia and Schenck, 1984) during two 4-hour periods between which the soil was allowed to cool overnight. The soil was then air-dried and 1-kg pots were filled half way with the respective soils.
Inoculum was prepared by wet-sieving spores (Gerdemann and
Nicolson, 1963) from single-species pot cultures of the mycorrhizal fungi grown on kudzu [Pueraria phaseoloides (Roxb.) Benth.] that had been initiated 4 mo previously (see Appendix B). The sieve aperture sizes used were 350, 106, and 63 m, depending on the spore diameter of the mycorrhizal fungus desired. The spores were removed from soil using a modified sucrose centrifugation method used at CIAT (Sieverding, 1983). The spores were then mixed into 40 g of the respective finely ground, sieved soil (2-mm sieve) at the rate of 15 spores g-1 soil. Spore numbers were achieved by applying an aliquot of a spore suspension to the soil and mixing thoroughly by hand in plastic bags. The soil carrier was the same type as the soil treatment in the pot. Each pot received 40 g of this inoculum. The control received 40 g of the appropriate soil type that had been mixed with spore washings. This was spread in an even layer over the soil surface and covered with more soil of the same type.
The pots were then watered to field capacity with double
deionized water and planted with pregerminated surface sterilized seeds of Phaseolus vulgaris L. var Carioca G 4017 at the rate of two seeds per pot. These were later thinned to one plant per pot. After planting, approximately 100 spores of each isolate were placed on 47mm-diameter millipore Gelman grid filters. These were folded twice and buried equidistant between the seeds at the rate of two per pot in the first experiment and one per pot in a subsequent experiment. All pots were also inoculated with a suspension of a strain mixture of Rhizobium phaseoli (CIAT strains 144, 899, and 985) in sterile
distilled water (106 cells ml-1) applied with a pipette at the rate of
1 ml at the base of each plant. The experiment contained 10 mycorrhizal isolates, one nonmycorrhizal control, five soil types with five replications and was repeated once.
After 11 and 23 days the millipore filters were removed from the pots and stained with 0.01% trypan blue in lactoglycerol (one part lactic acid, one part glycerol and one part water) to determine percent spore germination. A minimum of 30 spores was counted per filter. On day 23, the shoots and roots were also harvested. Roots were thoroughly washed and fresh weights were taken, and a 0.5-g subsample of each root was placed in a Tissue-tek holder. These were cleared with hot 10% KOH and stained in 0.1% trypan blue in lactophenol (Phillips and Hayman, 1970). Root length was determined using the grid-line intersect method (Giovannetti and Mosse, 1980), and total number of penetration sites of the endomycorrhizal fungus into the root was counted in the 0.5-g sample. Dry shoot and root weights were also measured and samples were ground and analyzed for P, Ca, Al, and N contents. Phosphorus, Ca and Al were all extracted using a HNO3 and HC1O4 acid digestion (Salinas and Garcia, 1985). Aluminum was then determined using the aluminon method which is a colorimetric analysis based on the reaction of Al with aluminon (Barnhisel and Bertsch, 1982); Ca and P were determined by atomic absorption spectrometry and colorimetrically, respectively (Salinas and Garcia, 1985). The micro Kjeldahl method was used for N (Bremner, 1965).
Statistical analyses of data from this and all succeeding experiments were done using SAS (SAS Institute Inc., 1985).
Lime Effects on VA Mycorrhizal Spore Germination and Hyphal Penetration of Roots
In this second series of experiments, the soil, a Typic Haplustox (Guerrero, 1975) from the CIAT research station at Carimagua, Meta, Colombia, was limed to determine the effect of modifying percent Al saturation by lime on the capacity of endomycorrhizal fungi to germinate, grow through the soil, and infect a host. The soil was air-dried, ground and coarsely sieved. Six hundred kilograms were stored in large plastic bags until needed and 400 kg were weighed into 25 kg portions and mixed with CaCO 3. The rates of CaCO3 applied were 0, 21.75, 40.75, and 52.5 g CaCO 3 25 kg1_ of soil (0, 1.7, 3.3 and 4.2 metric tons CaCO 3ha-1)(al 3). For each lime treatment 100 kg of
soil were limed. For the first experiment the soil and lime were mixed thoroughly while dry in a cement mixer. While still in the cement mixer, 2 L of water 25 kg- soil were thoroughly blended into the limed soil. The treated soil was then stored in plastic bags for
2 1/2 mo to allow the lime to react with the soil. For the second experiment this method was modified. The reserved soil was limed at
the same rates as in the first experiment but the lime was mixed dry with the soil in the cement mixer. One-kilogram pots were then filled with this dry soil and watered with double deionized water for 2 1/2 mo to allow the lime to react. For both experiments, after the incubation period, the soils were steam pasteurized at 80*C for two 4hour periods with an overnight period of cooling between. Lime treatments in the first experiment were placed in burlap bags and then
Table 3. Chemical analyses of a Typic Haplustox 2 1/2 months after liming.a
-1l f Exchange- Concentration (ppm) b c eg 10 g oil% A1f ableToa
meqi00-i oll Effective Satur- Toa
Incubation Lime % ppm P d pH Ca g sEce ation ae N
treatment level OM Bray II pH Al Ca Mg K CEC action [Mn Fe N
In plastic 0 2.8 1.8 4.8 1.82 0.25 0.08 0.06 2.23 81.6 0.02 5.7 214.3 1120
bags 21.75 2.7 1.7 5.3 1.02 1.04 0.22 0.07 2.35 43.4 0 3.1 60.2 1008
40.75 2.8 1.8 5.8 0.22 2.31 0.39 0.07 2.99 7.4 0 3.2 42.2 896
52.5 2.8 1.9 6.1 0.12 2.97 0.44 0.07 3.60 3.3 0 3.1 42.2 1008
In pots 0 2.7 1.6 4.3 1.96 0.42 0.10 0.07 2.71 72.3 0.16 2.7 190.3 1008
21.75 2.7 2.0 5.4 0.58 1.43 0.28 0.06 2.35 24.7 0 2.7 48.0 1008
40.75 2.7 1.8 5.6 0 2.70 0.45 0.06 3.21 0 0 2.2 39.0 1008
52.50 2.7 1.9 6.0 0 3.70 0.47 0.06 4.23 0 0 2.1 34.0 1008
aAnalyses performed after pasteurization. b, -1
g CaCO3 25 kg soil.
cOM = organic matter.
dBray II extraction method.
eCEC = cation exchange capacity.
f% Al saturation = (meq Al/ECEC) X 100.
steamed. For the second experiment the pots were placed in a steam wagon and the pasteurization took place in the pots.
In these experiments all of the isolates of mycorrhizal fungi
(Table 1) were used except G. mosseae c-131-2 which was eliminated due to insufficient spore production for inoculum use. The treatments consisted of nine species of mycorrhizal fungi and one nonmycorrhizal control plus four lime levels, with five replications of each treatment. The experiment was repeated once.
The soil inoculum carrier of each lime level was finely ground and sieved (2-rn sieve openings) and then 50 samples of 40 g each of each lime treatment were placed in plastic bags. The soil in each bag was infested at the rate of 15 spores g1 of soil with an aliquot from a spore suspension constantly stirred by a Nalgene floating stirbar to prevent spores from being damaged. The spores and soil were thoroughly mixed by hand. The spores, obtained from 4-mo-old pot cultures of P. phaseoloides, were separated from the soil by wet sieving and sucrose centrifugation. The control was prepared in the same manner but spore washings, rather than spores, were used to infest the bags of soil.
One-kilogram pots were half filled with steamed soil. The soilspore inoculum mix was spread evenly over the surface, then the pots were filled with steamed soil and watered to field capacity with double deionized water. Germinated, surface disinfected seeds of field bean, P. vulgaris L. var Carioca G. 4017, were then planted two per pot and later thinned to one per pot. Gelman 47-mm-diameter grid millipore filters containing approximately 100 spores of each isolate
were twice folded, and one filter was placed between the bean seeds for each fungal species in each pot.
After 23 days the millipore filters were removed and roots and
shoots were harvested. The millipore filters were unfolded in a Petri dish, stained with 0.01% trypan blue in lactoglycerol applied with an eye dropper and observed for spore germination with a dissecting microscope. Percent spore germination was calculated based on observations of. at least 30 spores per filter. Root fresh weight was
taken and at the same time a 0.5-g subsample for the first experiment and a 0.25-g subsample for the second experiment were taken. The subsamples were placed in Tissue-tek (Miles Laboratories) holders and processed as in the previous experiment. Root and shoot dry weight were determined and analyzed for P, Ca, Al, and N as previously indicated.
The Effect of Varying Al- Concentrations in Nutrient Solutions on VA Mycorrhizal Spore Germination and Hyphal Penetration of Roots
A nutrient solution experiment was designed to separate the
effects of Al concentration and pH. Four different concentrations of Al were evaluated using a system of four tables with four troughs per table into which a nutrient solution flowed every half hour (Fig. 1).
Each trough contained pots with holes in the sides and bottoms to allow entry and exit of the nutrient solution. The pots were filled
*with acid washed quartz sand. Each table had a plastic drum from which the nutrient solution was automatically pumped into the troughs every half hour. The nutrient solution was prepared in 120 L of double deionized water per drum and the pH was adjusted to 4.5 with
Figure 1. Illustration of the system used in nutrient solution studies. Drawing represents
one of four tables, drums and pumps used per experiment; drum (D) containing nutrient
solution, pump (P) and trough (T). Each drum and table of four troughs contained
one aluminum treatment.
Table 4. Modified Hoagland's nutrient solution used in the aluminum
Final solution Final solution
Reagent concentration Reagent concentration
(g L-1 x 10-3) (g L-1 x 10-3)
MgSO4.7 H20 12.033 ZnS04.7 H20 0.22
Ca(N03)2.4 H20 22.79 CuS04.5 H20 0.08
CaC12.2 H20 11.09 NaMo04.2 H20 0.0252
KH2PO4 0.439 FeS04.7 H20 5.0 x 10-3
KC1 26.00 Tartaric acid 4.0
KNO3 0.86 AlC13.6 H20 0.00
(NH 4)2SO4 52.24 AlC13.6 H20 4.47
H3BO3 2.86 AlC13.6 H20 8.94
MnCl2.4 H20 1.81 AlC13.6 H20 89.4
HCl or NaOH before planting the experiment. The pH was adjusted every
2 days and each nutrient solution was completely changed after 11 days. The base nutrient solution was the same for all treatments, only Al varied as 0, 0.5, 1.0, and 10 ppm Al. The nutrient solution used was a modified Hoagland's Solution (Hoagland and Aron, 1950) with very low P based on the work of Howeler et al. (1982) and Ojala and Jarrell (1980) (Table 4).
For this experiment several isolates of the original ten were
selected for evaluation. Initially, G. mosseae c-30-1, G. manihotis c-I-i, and c-17-1 were used. In the repeat experiment three additional isolates were included: G. manihotis c-20-2 and S. pellucida c-44-2 and c-3-7. The treatments then were three or six inoculated treatments, one nonmycorrhizal control, and four Al concentrations, with five replications per treatment. The inoculum treatments were randomized within the table containing the Al treatment.
Plastic bags containing 40 g of quartz sand were infested with a suspension of spores of the appropriate fungal isolate at the rate of 15 spores g-1 of quartz sand. Spore washings were used to infest the nonmycorrhizal control. This inoculum was placed evenly in a circle about 2 to 3 cm from the edge of the pots which had been previously filled 2/3 with quartz sand. The inoculum was then covered with quartz sand to a depth of about 6 cm. Surface sterilized, germinated seed of P. vulgaris var Carioca G 4017 were then planted two per pot and later thinned to one per pot. Approximately 100 spores of each isolate were placed on a Gelman millipore, 47-mm-grid filter which was folded twice and buried equidistant between the germinated seeds.
The first experiment was harvested after 21 days and the repeat
experiment was harvested after 26 days to allow further time for fungal development. In both experiments the millipore filters were removed and stained with 0.01% trypan blue in lactoglycerol, applied with an eye dropper, to determine percent spore germination as previously described. The roots were rinsed in double deionized water, fresh weight was determined, and subsamples of 0.21 and 0.25 g were taken in the first and second experiments, respectively. The subsamples were placed in Tissuetek holders and processed as in the previous experiments. Root and shoot dry weight were also determined and samples were ground and analyzed for P, Ca, Al and N content by the methods previously mentioned.
The Effect of Soils Naturally Varying in Percent Al Saturation on Spore Germination of VAN Fungi and Hyphal Penetration of Roots
In both of the greenhouse screening experiments, shoot dry weight, total root dry weight and total root length were negatively correlated with meq of aluminum in the soil across all mycorrhizal treatments (P<0.05) (Table 5). These same plant parameters showed a positive correlation with meq of calcium in the soil (P<0.05) (Table 5).
Species and isolates of the VA mycorrhizal fungi reacted in
different manners to soil aluminum and soil calcium (Table 6). Percent spore germination of G. mosseae, c-131-2, had a positive correlation with meq of soil calcium. In the repeat experiment spore germination of c-131-2 was negatively correlated with soil Al but not positively correlated with soil Ca (Table C-i). Isolate c-12-1 of A. longula had percent spore germination negatively correlated with soil Ca, and
Table 5. Correlation coefficientsa for plant parameters with
millequivalents of soil aluminum and calcium.
Shoot dry Total root Total root
weight dry weight length
Exp. 1c Exp. 2c Exp. 1 Exp. 2 Exp. 1 Exp. 2
meq Al 100 g- soil -0.43 -0.54 -0.74 -0.33 -0.68 -0.45
meq Ca 100 g- soil 0;36 0.76 0.24 0.61 0.60 0.47
a Correlation coefficients were determined using the SAS Institute Statistical Analysis System CORR Procedure. bp
Table 6. Correlation coefficients for fungal parameters with
milliequivalents of soil aluminum and calcium of selected
isolates of vesicular-arbuscular mycorrhizal fungi.a
Mycorrhizal fungi meq Al 100 g-1 soil meq Ca 100 g-1 soil
% Spore germination
Glomus mosseae c-131-2 ns 0.46 (0.09)*
Acaulospora longula c-12-1 ns -0.50 (0.06)
Scutellospora pellucida c-44-2 0.45 (0.09) ns
Number of penetration
A. longula c-12-1 ns 0.91
Glomus manihotis c-17-1 0.80 -0.52 (0.06)
G. manihotis c-20-2 ns 0.43 (0.11)
G. manihotis c-1-1 0.65 ns
aAll isolates not shown in table showed no significant correlation with soil chemical characteristics.
*P<0.01 unless otherwise indicated in parentheses. ns = not significant
percent spore germination of S. pellucida c-44-2 had a positive correlation with meq of soil aluminum. Different germination trends occurred within isolates of G. manihotis and A. longula in soils with high and low meq Al (Fig. 2). At 0.70 meq Al there were no differences in spore germination among the isolates of A. longula, but at 4.74 meq Al germination of isolate c-12-I was significantly greater than that of isolate c-17-2. Spore germination of G. manihotis isolate c-1-1 was significantly greater than that of c-20-2 at 0.70 meq Al. At 4.74 meq Al spore germination of G. manihotis isolate c-17-1 was significantly greater than that of c-1-1, which was significantly greater than that of c-20-2 in the first experiment only. In the repeat experiment spore germination of c-20-2 was usually greater or not different from spore germination of c17-i and c-i-i (Table C-2).
Several isolates of VA mycorrhizal fungi had significant correlations between the number of hyphal penetration sites formed in the host and either soil aluminum or calcium. Acaulospora longula c-12-i had a positive correlation (0.91) between the number of hyphal penetration sites formed per centimeter of host root and meq of soil calcium (Table 6). In the repeat experiment the number of penetration sites formed by c-12-1 was negatively correlated with soil Al but not positively correlated with soil Ca (Table C-i). Two isolates of G. manihotis, c-17-1 and c-i-1, showed a positive correlation between the number of penetration sites found per centimeter of root and meq of soil aluminum in both experiments (Tables 6 and C-1). The third isolate of G. manihotis, c-20-2, had no significant correlation with soil aluminum but had a positive correlation with soil calcium in the first experiment only. Germination and penetration of the
0 0.70 meq RI I 4.74 meq RI
Mos 3DMos 131 Pel 44 Lon 17 Lan 12 Lon 94 Man 17 Man 20 Man 1
VAM Frmaql Isolate
Figure 2. Percent spore germnation by vesicular-arbuscular mycorrhizal
(VAM) fungi as affected by milliequivalents aluminum in the
ie I eA
0MOS 30OoS 131 Pel1 44 Lon 17 Lon 12 Lon 94 Man, 17 Mani 20 Man I
VAM Furnqal Isolate
Figure 2. Percent spore germination by vesicular-arbuscular mycorrhizal
(VAM) fungi as affected by milliequivalents aluminum in the
soil. Within a soil type (0.70 and 4.74 meq Al) VA
mycorrhizal isolates with the same small letters are not
different according to Duncan's multiple range test. Within a mycorrhizal isolate differences between soil types with the
same capital letters are not significantly different
according to Duncan's multiple range test. Mos 30 = Glomus
mosseae c-30-1, Mos 131 = G. mosseae c-131-2, Pel 44 =
Scutellospora pellucida c-44-2, Lon 17 = Acaulospora longula c-17-2, Lon 12 = A. longula c-12-1, Lon 94 = A. longula c-945, Man 17 = Glomus manihotis c-17-1, Man 20 = G. manihotis c20-2, Man 1 = G. manihotis c-1-1.
same isolate were sometimes affected in a different manner by soil Al and Ca as shown by A. longula c-12-1 where germination was negatively correlated with Ca and penetration was positively correlated with Ca in the first experiment (Table 6, Figs. 2 and 3).
Some fungal parameters were correlated with plant parameters (Tables
7 and C-3). Percent spore germination of G. mosseae c-30-1 was positively correlated with shoot dry weight and root length in the first experiment only. Of all the VAN fungal isolates and species, only one isolate of S. pellucida, c-44-2, showed a positive correlation between percent spore germination and the number of hyphal penetration sites formed per centimeter of root. In the repeat experiment hyphal growth of c-44-2 was positively correlated with both germination and penetration sites but germination was not correlated with penetration (Table C-i). One of three isolates of G. manihotis, c-17-1, had a significant negative correlation between number of penetration sites formed per centimeter of root and shoot dry weight and root length for both experiments (Tables 7 and C-3). For one isolate of A. longula, c-12-1, there was a positive correlation between the number of penetration sites formed per centimeter of root and root length in the first experiment only.
In this first experiment there were significant differences in mycorrhizal species for percent spore germination and there was a significant mycorrhizae by soil interaction for number of hyphal penetration sites formed per centimeter of root (Tables 8 and 9). In the repeat experiment there were significant mycorrhizal treatment by soil interactions for spore germination, hyphal growth and root penetrations (Table C-2). One isolate of S. pellucida, c-44-2, and two isolates of G.
0 0.70 meq RL 1 4.74 meq RI
Onl 30pl4 e on1 o 2ln9 ma !a 2 a
0 FAbf imos 30 pe1 44 pe1 3 Ion 17 Ion 12 Ion 94 man 17 mon 20 man I
VAM Flungol Iso I rates
Figure 3. Root penetration by vesicular-arbuscular mycorrhizal (VAM)
fungi as affected by milliequivalents aluminum in the soil.
Within a soil type mycorrhizal isolates with the same small
letters are not different in penetration according to
Duncan's multiple range test. Within a mycorrhizal isolate differences in penetration between soil types with the same
capital letters are not significantly different according to Duncan's multiple range test. Mos 30 = Glomus mosseae c-301, Mos 131 = G. mosseae c-131-2, Pel 44 = Scutellospora pellucida c-44-2, Pel 3 = S. pellucida c-3-7, Lon 17 =
Acaulospora longula c-17-2, Lon 12 = A. longula c-12-1, Lon
94 = A. longula c-94-5, Man 17 = Glomus manihotis c-17-1, Man
20 = G. manihotis c-20-2, Man 1 = G. manihotis c-1-1.
Table 7. Correlation coefficients of selected isolates for fungal
parameters with plant parameters.a
Correlation coefficients Shoot dry Root Penetration sites
Mycorrhizal fungi weight length per cm root
% Spore germination
Glomus mosseae c-30-1 0.65* 0.49 (0.08) ns
Glomus manihotis c-17-1 ns -0.51 (0.06) ns
pellucida c-44-2 ns ns 0.46 (0.10)
sites per cm root
S. pellucida c-44-2 -0.57 (0.03) ns
G. manihotis c-17-1 -0.74 -0.78
longula c-12-1 ns 0.66
aAll isolates not shown in table showed no significant correlation with plant parameters.
*P<0.01 unless otherwise indicated in parentheses. ns not significant.
Table 8. Mean percent spore germination of vesicular-arbuscular
mycorrhizal (VAM) fungi across all soil types naturally
varying in percent aluminum saturation.
VAM fungal isolate % Mean spore germinationa
Scutellospora pellucida c-44-2 55 A
Glomus manihotis c-17-1 44 B
G. manihotis c-1-1 40 BC
Glomus mosseae c-131-2 35 C
G. manihotis c-20-2 24 D
Acaulospora longula c-12-1 23 D
G. mosseae c-30-1 12 E
A. longula c-94-5 9 E
A. longula c-17-2 1 F
aMeans with the same letter are not significantly different.
Table 9. Mean number of hyphal penetration sites formed per 100 centimeter
of root as affected by VAM fungal isolate and soil.a
Acaulospora longula c-12-1 Glomus manihotis c-17-1 Glomus manihotis c-i-i
Mean no. of Mean no. of Mean no. of
penetration meq Al1 penetration meq Al penetration meq Al
sites per 100 100 g sites per 100 100 g-1 sites per 100 100 g
cm root soil cm root soil cm root soil
1.0b B 0.30 22.0 B 0.30 2.0 C 0.30
29.0 AB 0.70 17.0 B 0.70 22.0 BC 0.70
36.0 A 2.52 79.0 A 2.52 156.0 A 2.52
4.0 B 2.96 72.0 A 2.96 34.0 B 2.96
1.0 B 4.74 97.0 A 4.74 26.0 BC 4.74
aSee Table 2 for additional soil characteristics. bMeans with the same letter are not significantly different.
manihotis, c-17-1 and c-i-i, gave the highest percent spore germination across all soils in experiment one (Table 8). Since there was a soil by spore germination interaction in experiment two, germination could not be looked at across all soils. In this experiment S. pellucida c-44-2 and c3-7 had the highest percent germination in every soil except at 0.30 meq Al when germination of A. longula c-12-1 was intermediate between the two (Table C-2). Although G. mosseae c-131-2 germinated significantly better than G. mosseae c-30-1, it was dropped from further experiments because it produced insufficient spore numbers in pot culture to provide adequate inoculum. Of the above isolates, c-17-1 and c-i-i of G. manihotis had the greatest number of penetration sites per cm root at 4.74 meq Al in the first experiment only (Fig. 3). Only those isolates (c-17-1, c-12-1, c-i1) which showed a significant correlation with P<0.01 with soil aluminum or calcium and number of penetration sites formed per centimeter of root in Table 6 had penetration significantly affected by soil type that varied in meq Al (Table 9).
Effect of Lime on Percent Spore Germination and Hyphal Penetration of
The effect of added lime and the consequent reduction in the amount
of exchangeable soil aluminum produced a quadratic increase in shoot dry weight and a linear increase in root dry weight in the second experiment
(Fig. 4). This was true for shoot dry weight across all mycorrhizal
treatments W0(.01) as determined by orthogonal polynomial contrasts. In
the first lime experiment shoot and root dry weight increased
quadratically with increasing time.
A shoot dry weight 2 0 root dry weight 2
0.5-o shoot dry weight I root dry weight I -.
0 0.4 A
O. ........... .........
L 0.2 ....o
0 1 2 3 4
Tons L I me/Ho
Figure 4. Effect of lime on shoot and root dry weight of Phaseolus
vulgaris. Shoot and root dry weight 1 are from first lime
experiment. Shoot and root dry weight 2 are from second lime
There were several correlations between fungal and soil parameters.
Hyphal growth, spore germination and number of penetration sites per cm of root were negatively correlated with meq Al and positively correlated with meq Ca for G. mosseae c-30-1 (Table 10). The same results were obtained in the first experiment for G. mosseae c-30-1 except penetration was not negatively correlated with Al (Table C-4). For S. pellucida c-3-7 hyphal growth and penetration were both positively correlated with meq Ca in one experiment only. Isolate c-44-2 of S, pellucida had spore germination and hyphal penetration negatively correlated with meq Ca in both experiments; penetration was positively correlated with meq Al in both experiments. Hyphal growth of A. longula c-12-1 was positively correlated with meq Al but not negatively or positively correlated with meq Ca. Percent spore germination of A. longula c-12-1 was negatively correlated with meq Al and positively correlated with meq Ca. The other two isolates of A. longula, c-94-5 and c-17-2, had both penetration and germination negatively correlated with meq Ca and penetration for both isolates was positively correlated with meq Al. None of the correlations with Ca and Al for the isolates of A. longula were supported by the repeat experiment. In the second experiment all three isolates of C..manihotis showed a significant negative correlation for spore germination with meq Ca and only c-17-1 and c-i-i were positively correlated with meq Al. Penetration sites formed by c-i-i were positively correlated with meq Al and negatively correlated with meq Ca. Neither of the other two isolates had significant correlations for penetration sites. In the first experiment germination of c-17-1 and c-i-i was negatively correlated with Ca and positively correlated with
Table 10. Correlation coefficients for fungal parameters correlated
with soil chemical factors in limed soil.
Mycorrhizal fungi meq Al 100 g-1 soil meq Ca 100 g-1 soil
Hyphal growth rating
Glomus mosseae c-30-i -0.83a 0.79
Scutellospora pellucida c-3-7 ns 0.53 (0.02)
Acaulospora longula c-12-1 0.48 (0.03) ns
% Spore germination
G. mosseae c-30-1 -0.80 0.90
S. pellucida c-44-2 ns -0.50 (0.02)
A. longula c-17-2 0.65 -0.59
A. longula c-12-1 -0.70 0.80
A. longula c-94-5 ns -0.58
G. manihotis c-17-1 0.59 -0.73
G. manihotis c-20-2 ns -0.61
G. manihotis c-1-1 0.73 -0.87
per cm root
G. mosseae c-30-1 -0.57 0.63
S. pellucida c-44-2 0.72 -0.62
S. pellucida c-3-7 -0.56 0.53 (0.02)
A. longula c-17-2 0.67 -0.53 (0.03)
A. longula c-94-5 0.66 -0.55 (0.02)
G. manihotis c-1-i 0.81 -0.69
aP<0.01 unless otherwise indicated in parentheses.
Al (Table C-4). Penetration by G. manihotis c-i-i and c-20-2 was positively correlated with Al and negatively correlated with Ca.
In five out of nine VAM fungi, there were significant correlations between fungal parameters (Table 11). Hyphal growth of G. mosseae c-30-1 was positively correlated with percent spore germination and the number of penetration sites formed per cm of root. Percent spore germination was also positively correlated with the number of penetration sites. Hyphal growth of A. longula c-12-1 and G. manihotis c-1-i were negatively and positively correlated with percent spore germination, respectively. Germination of c-i-i was also positively correlated with penetration sites formed. In the second experiment only hyphal growth of S. pellucida c-44-2 showed a significant positive correlation with the number of penetration sites formed. In the first experiment hyphal growth of G. mosseae c-30-i and G. manihotis c-i-1 was positively correlated with spore germination but not penetration sites (Table C-4).
Plant parameters were often significantly correlated with percent
germination, hyphal growth and number of penetration sites formed per cm root (Table 12). Glomus mosseae c-30-I showed a significant positive correlation for root weight and root length with percent spore germination, hyphal growth and number of penetration sites formed. Root weight of plants inoculated with S. pellucida c-44-2 was significantly negatively correlated with germination and penetration sites in the root. Root length was negatively correlated with penetration in the root. Germination of A. longula c-12-i was positively correlated with root weight, and shoot weight was positively correlated with root penetration. Isolates c-94-5 of A. longula and c-20-2 of G. manihotis had root length
Table 11. Correlation coefficients for fungal parameters in limeda soil.
Hyphal growth % Spore sites per
Mycorrhizal fungi rating germination cm root
Hyphal growth rating
Glomus mosseae c-30-1 0.69b 0.51 (0.02)
Glomus manihotis c-1-i 0.43 (0.06) ns
Scutellospora pellucida c-44-2 ns 0.73
Acaulospora longula c-12-1 -0.42 (0.07) ns
% Spore germination
G. mosseae c-30-1 0.69 0.67
G. manihotis c-1-1 0.43 (0.06) 0.55
A. longula c-17-2 ns 0.47 (0.06)
aLime applied at the rate of 0, 1.7, 3.3 and metric 4.2 tons CaCO3 ha-1 bp
Table 12. Correlation coefficients for fungal parameters associated with
plant parameters in limeda soil.
Dry shoot Dry root Root
Mycorrhizal fungi weight weight length
Hyphal growth rating
Glomus mosseae c-30-1 ns 0.75b 0.55
% Spore germination
G. mosseae c-30-1 -0.45 0.66 0.49
Scutellospora pellucida c-44-2 ns -0.47 ns
Acaulospora longula c-12-1 ns 0.60 ns
per cm root
G. mosseae c-30-1 ns 0.52 0.76
S. pellucida c-44-2 ns -0.47 -0.46
A. longula c-12-1 0.48 ns ns
A. longula c-94-5 ns ns -0.46
G. manihotis c-20-2 ns ns -0.50
aLime applied at the rate of 0, 1.7, 3.3 and 4.2 metric tons CaCO03 ha-1 bP<0.05.
negatively correlated with penetration sites per cm root. Of all of these results only the negative correlation of root length with penetration sites for A. longula c-94-5 was supported by data in the repeat experiment (Table C-5).
In the case of percent spore germination and number of hyphal
penetration sites formed per centimeter of root, there were significant lime by mycorrhizal treatment interactions in the second lime experiment. In the first lime experiment there were significant lime by mycorrhizal treatment interactions for spore germination, hyphal growth and number of penetration sites (Table C-6). For percent spore germination orthogonal polynomial contrasts indicated that reducing exchangeable meq Al with lime had a positive linear effect on spore germination of G. mosseae c-30-1 in both experiments (Fig. 5 and Table C-6). Hyphal growth of c-30-1 increased linearly in the second experiment and quadratically in the first experiment with increasing
lime. The linear increase of penetration sites formed by c-30-1 was significant only in the second experiment. Germination of S. pellucida c-44-2 decreased linearly and quadratically in the second and first experiments, respectively (Fig. 6 and Table C-6). Germination of S. Pellucida c-3-7 was unaffected by lime in either experiment (Fig. 7 and Table C-6). Lime had a positive linear effect on spore germination of A. longula c-12-1 (Fig. 9) and a negative linear effect on germination of A. longula c-94-5 and c-17-2 (Figs. 8 and 10). These effects on the spore germination of isolates of A. longula were not supported by significant results in experiment one (Table C-6). Lime had a negative linear effect on spore germination of the isolates c-17-1, c-20-2 and
6.0 100 -- -15.0
80 ./ 12.0
0 0 i
3 4.0 E
0 2.0 o 0
CL 1, "
:D ,a / # penetration sites
1.0 -., o Z germination m
Sa hyphal growth index
0.0 -""' 0.0
0 I 2 3 4 5
Tons L I me/Ha
Figure 5. Lime effects on spore germination (g), root penetration (p)
and hyphal growth (hg) of Glomus mosseae c-30-1. Regression
equations for lime effects:
g = 10.358 59.165 lime + 47.355 lime2 7.099 lime3
p = -1.62 + 3.75 lime, hg = 1.31 + 0.95 lime.
5.0 90 --- -- ----------- 50.0
-OE, O Soegem o i:on
es g0 io
2st entraio si..tes
0.0 0 0.0
0 I 2 3 4 5
Tons L i me/H-o
Figure 6. Lime effects on spore germination (g), root penetration (p)
and hyphal growth (hg) of Scutellospora pellucida c-44-2.
Regression equations for lime effects:2
g = 71.976 2.646 lime, p = 40.86 24.45 lime + 3.62 ale2,
hD = 5.870 0.813 lime.20.0
1.- 20 a hujhol growth index -10.0
0.0- 0a .
0 12 3 45
Figure 6. Lime effects on spore germination (g), root penetration (p)
and hyphal growth (hg) of Scutellospora vellucida c-44-2.
Regression equations for lime effects:2
g =71.976 2.646 lime, p = 40.86 24.45 lime + 3.62 lie,
hg =5.870 0.813 lime.
6.0- 1100 .50.0
840 --- ------------ -.30.0
L. 4o. 2o.o
Eg o Z germination
.... at penetration si tes
20. 10.0 30.0
1.0 a hyphal growth index
0.0 01 0.0
0 2 3 4 5
Tans L i me/Ho
Figure 7. Lime effects on the spore germination (g), root penetration
(p) and hyphal growth (hg) of Scutellospora pellucida c-7-3.
Regression equations for lime effects:
g = 80.889 18.053 lime + 12.747 lime2 2.067 lime3, p =
3.30 + 5.80 lime.
6 .0 30 -.- -.. .- -.-
I 1.50 E
o Z germination u
# g at penetration sites 0
3o a hyph l growth index
0.0- 0 0.00
0 I 2 3 4 5
Tons LI me/Ha
Figure 8. Lime effects on spore germination (g), root penetration (p)
and hyphal growth (hg) of Acaulospora "longula c-17-2.
Regression equation for lime effects:
g = 19.709 4.645 lime.
5. o Z germination .
4.0 [" At penteration Si tes ... ,18O0
x U "" "
1.0 20 ....................:- \- 2.0
a hpha growth index x C
0.0 Q 0 0.0
o tujhal growth index
0 I 2 3 4 5
Tons Li me/Ha
Figure 9. Lime effects on spore germination (g), root penetration (p)
and hyphal growth (hg) of Acaulospora longula c-12-1.
Regression equation for lime effects:
g = )4.424 + 8.991 lime, hg = 6.955 3.389 lime + 0.680
5.0 70 11.50
o Z germination a # penetration sites
ck -- -- -a-
C) e:. 0.,
3.0 2 a
20 hyph Igrow th index
0.0 0 ..' I 0.00
0 1 2 3 4 5
T.. L m/H
Figure 0. Lime effects on spore germination (g), root penetration (p)
and hyphahyphl growth (hg) of Acaulospora lonsula dex-94-5.
0 1 2 3 4 5
Tons LI ma/Ha
Figure 10. Lime effects on spore germination (g.), root penetration (p)
and hyphal growth (hg) of Acaulospora ~la c-94-5.
Regression equations for lime effect:
g = 55.300 +23.832 lime .789 lime hg = 4.6 7.754 lime
+ 5.554 lime 0.910 lime
c-i-i of G. manihotis in the second experiment (Figs. 11, 12 and 13). In the first experiment germination of c-17-1 and c-i-i decreased cubically and linearly, respectively, with increasing lime (Table C-6). Penetration by G. manihotis c-i-i decreased quadratically in the first experiment and linearly in the second experiment as lime levels increased (Table C-6 and Fig. 13). Penetration by S. pellucida c-3-7 increased linearly with the reduction of meq Al in the second experiment only (Fig. 7).
Nutrient Solution Studies With Varying Amounts Of Aluminum
Two nutrient solution studies were conducted. The second experiment was a repeat of the first but also contained three additional treatments. Due to the experimental design, aluminum was treated as a block effect in the individual experiments. When the duplicated portions of experiments one and two were combined, aluminum effects could be analyzed. Experiment one and the corresponding portion of experiment two will be discussed together and then experiment two alone in its entirety will be discussed. Nutrient solution experiments one and two
Plant parameters were affected by aluminum concentration
irrespective of VA mycorrhizal treatment. Shoot dry weight was reduced by increasing aluminum concentrations (Fig. 14). The most dramatic reduction occurred from 0 ppm Al to 0.5 ppm Al. Although orthogonal polynomial contrasts indicated that there was a significant (P<0.05) quadratic effect of aluminum concentration on shoot dry weight,
6.0 50 5.00
5.0 0 .00 0
20 o germination 2.00
S# penetration sites 10 1.00
1.o 0- a hyphal growth index
0.0 0 -1' 0.00
0 I 2 3 4 5
Tons Li m/Ho
Figure 11. Lime effects on spore germination (g), root penetration (p)
and hyphal growth (hg) of Glomus manihotis c-17-1.
Regression equation for lime effects:
g = 44.195 3.631 lime.
5.0 40 2.00
o Z germination A penetration sites 0
4.0 a hyphaI growth index
3.0 "./ .15
L e 20 1 4 1... -- .. ', 1.00
0 - % :J
10 --0.50 a
1.0 .. A.
0.0 0 0.00
0 I 2 3 4 5
Figure 12. Lime effects on spore germination (g), root penetration (p)
and hyphal growth (hg) of Glomus manihotis c-20-2.
Regression equations for lime effects:
g = 2.643 2.494 lime, hg = 2.200 + 10.860 lime 6.214
lime + 0.875 lime3.
6.0 70 20.0
o Z germination A # penetration sites
hyphal growth index E
O 3.0- -- ..-"---.... 10.0
.. ....................................., ....... .. . . .
0.0 0 0.0
0 I 2 3 4 5
Tons Li m/Hc
Figure 13. Lime effects on spore germination (g), root penetration (p)
and hyphal growth (hg) of Glomus manihotis c-1-1.
Regression equations for lime effect :
g = 58.193 0.119 lime 1.662 lime p = 17.48 7.05 lime
+ 1.07 lime.
I o experiment 2
4 experiment 182
"a 0.3 0o 0.
0 1 2 3 4 5 6 7 8 9 10 II 12
PPM Riminum In Solution
Figure 14. Effect of solution aluminum on shoot dry weight of Phaseolus
there appeared to be little further reduction in shoot weight above 0.5 ppm Al. Root tissue aluminum increased with increasing solution aluminum (Fig. 15). The effect of solution aluminum concentration on root tissue aluminum concentration was linear across all mycorrhizal treatments according to the orthogonal polynomial contrast (P
The VA mycorrhizal isolates used in the combined experiments were G. mosseae c-30-1, G. manihotis c-i-I and G. manihotis c-17-1. In the combined experiments there was no significant effect of aluminum on percent spore germination and number of hyphal penetration sites formed per centimeter of root. There were, however, significant differences among VA mycorrhizal isolates for both percent spore germination and number of hyphal penetration sites formed per cm of root (Tables 13 and 14). Isolate c-i-i of G. manihotis formed more penetration sites per cm of root than did the other fungi. Orthogonal polynomial contrasts were performed for species and these indicated that G. manihotis formed significantly higher numbers of penetration sites than did G. mosseae. Glomus mosseae was not significantly different from the control. For percent germination across all aluminum levels, G. mosseae was significantly higher than both isolates of G. manihotis. Nutrient solution experiment two
Plant parameters in the second nutrient solution experiment
responded to aluminum similarly to plant parameters in the combined experiments. Shoot weight decreased with the addition of 0.5 ppm Al; little further reduction occurred in shoot weight with the addition of
1.0 and 10 ppm aluminum (Fig. 14). Root concentrations of aluminum
300 &experiment 18 20
Cc 0aI 000...
0 1 2 3 4 5 6 7 8 9 10 11 12
PPM Rlininum In Solution Figure 15. Effect of solution'aluminum on root aluminum of Phaseolus
Table 13. Mean percent spore germination across all aluminum levels
for both solution experiments.
VA mycorrhizal % Spore
Glomus mosseae c-30-1 16 Aa
G. manihotis c-i-i 9 B
G. manihotis c-17-1 9 B
aMeans followed by the same letter are not significantly different as indicated by the Duncan's multiple range test.
Table 14. Means of hyphal penetration sites across all aluminum
levels for both solution experiments.
Mean no. penetration sites
VA mycorrhizal -i
isolates 100 cm root
Glomus manihotis c-i-i 2.0 Aa
G. manihotis c-17-1 1.1 B
G. mosseae c-30-I 0.4 BC
Control 0.0 C
aMeans followed by the same letter are not significantly different as indicated by the Duncan's multiple range test.
increased with increasing aluminum in the solution (Fig. 15). For all plant parameters the effect of aluminum was significant (P<0.01).
In the second solution experiment the VAM fungal isolates-were G. mosseae c-30-1; G. manihotis c-i-i, c-17-1 and c-20-2; and S. pellucida c-44-2 and c-3-7. Since aluminum was treated as a block effect, only statistical comparisons within levels but not between aluminum levels could be made. In this experiment there were significant aluminum by mycorrhizal isolate interactions for percent spore germination (P<0.01) and for the hyphal growth rating value (P
Mean hyphal growth values for each aluminum level were usually lowest for G. mosseae c-30-1 and G. manihotis c-17-1 (Fig. 16). Scutellospora pellucida c-3-7 hyphal growth values remained high across all aluminum levels. Glomus manihotis c-i-i and c-20-2 had intermediate levels of hyphal growth. Both of these isolates of G. manihotis grew best at 1 ppm Al and had reduced hyphal growth at 10 ppm Al.
Percent spore germination of S. pellucida c-3-7 and c-44-2 was
consistently significantly higher than all other isolates at all levels of aluminum (Fig. 17). These two isolates behaved differently in regard to percent spore germination with c-3-7 increasing with increasing aluminum concentration and c-44-2 decreasing with increasing aluminum concentration. Although comparisons between aluminum levels cannot be made, aluminum as a block effect was significant and spore germination of the two isolates was significantly different at every level.
8 c-1-1 50 c-17-1
4- c-44-2 9 c-3-7
? 0 c-30-1
L 3 1 c-20-2
ppm Al in solution
Figure 16. Hyphal growth of vesicular-arbuscular mycorrhizal fungi as
affected by solution aluminum. For 0 ppm Al LSD = 1.5, for 0.5 ppm Al LSD = 0.86, for 1.0 ppm Al LSD = 1.4, for 10 ppm
Al LSD = 1.5. Isolates of species used are as follows:
Glomus manihotis c-1-1, c-20-2 and c-17-1; G. mosseae c-301; Scutellospora pellucida c-44-2 and c-3-7.
o O3 c-30-1 I c-i-1 c-17-1
s .1 c-20-2 8 c-44-2 i c-3-7
o U 0 1U 10
ppm AI in solution
Figure 17. Percent spore germination of vesicular-arbuscular
mycorrhizal fungi as affected by solution aluminum. For 0 ppm Al LSD = 5.7, for 0.5 ppm Al LSD = 5.6, for 1.0 ppm Al
LSD = 7.3, for 10 ppm Al LSD = 6.2. Isolates of species
used are as follows: Glomus manihotis c-1-1, c-20-2 and c17-1; G. mosseae c-30-1; Scutellospora pellucida c-44-2 and
Germination of G. mosseae c-30-1 remained quite consistent across all levels of aluminum. Of the three isolates of G. manihotis, only germination of c-20-2 increased at the highest level of aluminum; the other two were depressed and significantly lower than c-20-2 at 10 ppm aluminum.
Averaged over all aluminum levels S. pellucida. c-3-7 had a
significantly higher number of penetration sites per cm of root than did all other isolates (Table 15). This species was also capable of forming multiple germ tubes after the initial germ tube aborted (Fig. 18). Glomus mosseae c-30-1 formed the fewest number of penetration sites per cm of root of all the isolates. Although there was no significant aluminum by mycorrhizae interaction so aluminum effects cannot be determined, aluminum as a block effect was significant (Fig. 19). It appears that almost all isolates were depressed at 0.5 ppm Al.
Wang (1984) showed that high aluminum concentrations in sand culture could eliminate infection of roots p recolonized with Glomus caledonicum. She deliberately avoided the use of spores to eliminate the confounding effects of germination and initial infection. It is precisely these initial steps in the growth of the fungus which would seem to be potentially most sensitive to aluminum and other metals in the soil environment. In the colonization stage the root may provide some protection to the fungus. Certainly germination and ability to penetrate the host are factors most important in determining infectivity of VAM fungi in the field.
Table 15. Mean number of hyphal penetration sites per centimeter of
root for each isolate of vesicular-arbuscular mycorrhizal
fungi across all aluminum levels for solution experiment two.
Mean no. penetration sites
isolates 100 cm-1 root
Scutellospora pellucida c-3-7 5.1 Aa
Glomus manihotis c-1-1 1.1 B
G. manihotis c-20-1 0.9 B
S. pellucida c-44-2 0.9 B
G. manihotis c-17-1 0.5 B
G. mosseae c-30-1 0.3 B
Control 0.0 B
aMeans followed by the same letter are not significantly different (a = 0.05) as indicated by the Duncan's multiple range test.
Figure 18. Aborted germ tubes of Scutellospora pellucida. ab = aborted
germ tube, arrow indicates probable remains of spore
8. 0 c-30-1 c-1-1 0 c-17-1
0 U c-20-2 I c-44-2 0 c-3-7 a
c 4.0 ,--o
D 2.0 b
S1.0- bb b b
0~ L 0. 1.0 10
ppm AI. in solution
Figure 19. Number of root penetration sites by vesicular-arbuscular
mycorrhizal fungi as affected by solution aluminum.
Mycorrhizal fungi within aluminum level are not different if they share the same letter. Isolates of species used are as
follows: Glomus manihotis c-1-1, c-20-2 and c-17-1; G.
mosseae c-30-1; Scutellospora pellucida c-44-2 and c-3-7:
The three screening methods, two with soil and one with nutrient solution, were devised to determine if aluminum affected spore germination, the ability of the fungus to grow through the soil, and the ability to penetrate the host. It was hoped that the techniques would provide a rapid method for the selection of VAM fungi resistant to levels of aluminum found in acid soils.
In the screening method, which used soils that varied naturally in meq of aluminum, a better correlation was found between fungal parameters and soil chemical characteristics if meq and percent aluminum saturation varied in a corresponding manner. If one decreased and the other increased in a particular soil compared to the pattern in the other four soils, that particular soil produced results out of line with the trend in the other soils. For this reason three of the five original soils were selected which had increasing meq of aluminum as well as increasing percent aluminum saturation. The two soils in which percent aluminum saturation increased but meq of aluminum decreased in relation to the other three soils were not included in the correlation data. Germination and penetration by the VAM fungi were often stimulated in these two soils in contrast to the trend in the other three soils, with penetration being more responsive than germination. The actual amount of aluminum seemed more important than the percentage of the exchange complex occupied by aluminum. Another possible confounding factor is that the concentration of Mn (ppm) in these two soils was also lower (Table 2). Manganese reduces percent VAM fungal colonization of the root, although higher concentrations are needed than of aluminum (Wang, 1984). Hepper and Smith (1976) have also shown that
Mn may reduce spore germination and hyphal growth of VAN fungi. Since both Mn and Al toxicity problems often occur in the same soils and they both seem to affect VAN fungi, possibly in a synergistic manner, both elements should be considered when searching for adapted isolates.
Screening in limed and natural soils gave similar results. In the
limed soils there were better correlations between fungal parameters and soil chemical characteristics and among fungal parameters themselves than in the natural soils. Spore germination and root penetration were sometimes positively or negatively correlated with soil Al and Ca. Hyphal growth was rarely correlated with soil chemical characteristics. Two isolates, representing one species, and two isolates, representing two species, had germination and penetration, respectively, positively correlated to meq Al in the soil in the two lime experiments. These same isolates were negatively correlated with meq Ca in the soil.
Many of the isolates used in these two studies originated from
soils with high meq of aluminum (Table 1) and would be expected to show tolerance to soil aluminum. In some cases these isolates showed a strong positive correlation between germination, host penetration and amount of soil aluminum. Two examples of this are A. longula c-12-1, which, although from a soil of 2.4 meq Al, had host penetration
positively correlated with meq Ca; and G. manihotis c-i-i from a soil with 4.8 meq Al, which behaved as expected with strong positive correlations with meq Al. The isolates showed varied activities of germination and penetration in the five natural soils which were not always significantly correlated with aluminum or calcium. This type of variability has also been noted by Azcon-Aguilar et al. (1986a) among
three species in six soils ( pH 6.5 7.9). They found no significant correlation between the density of VAM fungal propagules and concentration of soil P. Dry matter production in response to the VAM fungi varied for each soil as did the yield relationships among the fungi. Abbott and Robson (1981a) found that soil type influenced the effectiveness of several VAM fungi. Glomus fasciculatum grew well irrespective of soil type while growth of Acaulospora laevis and Glomus monosporum depended on soil type. Hayman and Tavares (1985) found that G. fasciculatum was effective across a wide pH range (4 to 7) while the effectiveness of other VAM fungi studied depended on soil type or pH. They found that some VAM fungi did not always show the same optimum pH in different soils. In this study G. manihotis c-1-1 grew.well over a broad pH range while G. mosseae c-30-1 did not.
Some fungi, like G. manihotis c-1-1 and S. pellucida c-44-2 which were isolated from soils with high meq Al, are stimulated to greater growth as meq of soil aluminum increase. The majority of the isolates had germination and penetration depressed as meq Al increased. Isolates of the same species did not always behave in the same manner. Penetration by G. manihotis c-1-1 was positively correlated with increasing Al in all the experiments with soil while that of G. manihotis c-20-2 often had no significant trend. Differences of isolates of VAM fungi of the same species to heavy metals and in their ability to infect and produce a growth response have been found by other researchers (Abbott and Robson, 1981b; Gildon and Tinker, 1981; Haas and Krikum, 1985).
The idea that increasing meq of Al could reduce germination and penetration is not surprising since Wang (1984) has shown that Al can reduce root colonization by VAM fungi and Al is known to be toxic to other organisms. Stimulation of germination and penetration with increasing meq of Al is more difficult to explain. Although germination
studies involving VAM fungi and aluminum concentrations have not been done previously, some research has been done with other fungi. Zoospore germination of Aphanom ces euteiches Drechs. was not decreased compared to the control at 5 Ug Al m:V1 and germination still occurred at 250 pg Al mjSl (Lewis, 1973). Orellana et al. (1975) found little reduction in fung al mass of Scierotinia sclerotiorum with 4 ppm Al compared to the control, and the fungus still grew with as much as 32 ppm Al. Another factor that may enter into the situation is that Siquiera et al. (1982) found there is an interaction between pH and the nutrient content of the medium and their effects on germination of spores of VAM fungi in vitro. At various nutrient concentrations germination could be the same at pH
4.5 and 6.0, at pH 4.5 and 7.5 or different at all pH ranges. In untreated soils the nutrient interactions are more complicated, since nutrient variation is not controlled from soil to soil.
There is an optimal pH range for germination of spores of VAM
fungi, although this may vary for some fungi with soil type (Hayman and Tavares, 1985). The range of pH above and below the optimum over which germination can occur varies depending on the species of VAM fungi. In the present study, it appears that fungi adapted to acid soils may have had a broader range over which germination took place. Green et al. (1976) found that Scutellospora coralloidea germinated best at pH 5 but
gave a minimum of 20 % germination across the pH range of 4 to 8, while G. mosseae germinated best at pH 7 and not at all at pH 4. Wang (1984) found that inoculum from acid soils infected roots almost equally well at all pH ranges, but inoculum from alkaline soils infected roots poorly when grown in acid soils. This wide growth range may be one criterion to look for in VAM fungi adapted to acid soils. Isolates would then still have a chance of being effective in the case that liming occurred. Fungal germination and growth may be reduced from optimal levels but may still be vigorous enough to be competitive with other fungi. This may
have been the case with G. fasciculatum that gave good growth responses irrespective of soil type in the study of Abbott and Robson (1981a). In this present study G. manihotis c-i-i had a range of penetration sites per 100 cm root from about 34 to about 7 while G. mosseae ranged from 0 to about 14 a pH range of 4.2 to 6.0. This is still below the optimal pH range of 7.0 7.2 for G. mosseae (Green et al., 1976; Hepper and Smith, 1976). At higher pH levels G. mosseae may show more vigorous growth than G. manihotis, but for use in acid soils this is immaterial since liming is recommended only to the point of removing available aluminum from the soil exchange sites. This occurs at about pH 5.5.
Spore germination does not guarantee host root penetration. In
this study germination was rarely correlated with number of penetration sites formed in the root. Penetration and germination may be affected differently by different soil factors. Siqueira (1983) found situations where germination and hyphal growth behaved in opposite manners under the influence of one factor. Nitrogen for example increased percent germination but decreased hyphal growth by 55 percent. In the two soil
screening procedures of this present study, spore germination and hyphal growth were rarely correlated with penetration sites. Although this occurred more frequently in limed soil the instances were not substantiated by the repeat experiment. Spore germination and hyphal growth were more frequently correlated, and the correlation was verified in the following three isolates representing three species: S. pellucida c-44-2, G. mosseae c-30-1 and G. manihotis c-i-i. A better correlation in the limed soil than in the natural soils might be expected among the above mentioned factors since soil nutrient levels change in a coordinated manner with lime application. With increased levels of lime both Al and Mn decrease while in the natural soils Al and Mn varied independently of each other. Van Nuffelen and Schenck (1984) in their study involving six species of VAMl fungi found no correlation between percent spore germination and number of penetration sites formed. A possible reason why van Nuffelen and Schenck (1984) found no correlation between germination and penetration sites is that their roots were sampled for penetration at 30 days while those in the present study were sampled at 21 days. At 30 days they may have been counting secondary penetration sites, that is, penetration that results from a hypha that previously penetrated the root.
Of the fungal parameters measured, the formation of penetration sites within the root seems to be the most sensitive to soil chemical factors. Germination and hyphal growth, although reduced from their optimum level, would occur to some extent, while on several occasions penetration was reduced to zero or almost zero. Siqueira et al. (1982) found that with Gigaspora margarita germ tube growth was less affected
by pH than was germination. In the present study germ tube growth or hyphal growth seemed to be about equally sensitive as germination to pH except in the case of.Glomus mosseae. This species is very sensitive to acid soils. By reducing meq Al 100 g'l soil from 1.96 to 0.58 with lime, hyphal growth increased while germination remained the same (Fig. 5). In a later paper Siqueira et al. (1984) suggested that soil acidity affects the VAM fungus mostly in the rhizosphere before root penetration occurs rather than in the root itself. Results of the present study indicate that penetration is more sensitive to soil chemical factors than either germination or hyphal growth. During penetration a three way or more interaction is possible among the fungus, the host and the soil. The ability of the fungus to penetrate the host may depend to some extent on the host's reaction to soil acidity as well as the reaction of the fungus to soil conditions.
In the solution aluminum study, tissue aluminum increased with increasing solution aluminum concentrations and dry matter decreased. The greatest reduction in dry matter occurred between 0 and 0.5 ppm Al in the solution (Fig. 14). Aluminum effects on the VAM fungi also were usually most pronounced between the range of 0 and 0.5 ppm Al. Wang (1984) found that at high P concentration percent infection was not reduced significantly until 1 ppm Al was reached, while at low P concentration it was reduced by the addition of 0.5 ppm Al. Aluminum had no effect on Glomus mosseae in either solution experiment. Its growth was severely reduced in general in this system. Spore germination for all isolates except S. pellucida was poor in this system. Germination of G. manihotis c-i-i and c-17-1 was reduced only
at 10 ppm Al in both experiments. The two isolates of S. pellucida germinated very well but behaved in different manners, only one of two was suppressed by increasing Al concentrations, Scutellospora pellucida probably maintained a high percent germination in this system because of its manner of germination. Germination of S. pellucida occurs by germ tube growth through the spore wall and not through the subtending hyphal connection as in Glomus spp. If the germ tube of S. pellucida aborted, the spore was able to produce another germ tube. In this system, spores were often seen with several aborted germ tubes and a successful one. Glomus spp., however, seemed incapable of initiating multiple germ tubes after the abortion of the initial germ tube.
Spore germination of G. mosseae was higher than that of G.
manihotis. This was due to the manner in which percent germination was counted. A spore was counted as germinated if new growth of the subtending hyphae could be seen. Glomus mosseae tended to germinate but growth of the germ tube aborted immediately. Glomus manihotis was less likely to germinate, but when it germinated germ tube growth was more likely to continue. If G. mosseae did manage to reach the root surface and form an appressorium, the infection never developed beyond an aborted entry site. Glomus manihotis on the other hand was able to grow enough to produce external hyphae on the root and even an occasional external vesicle.
Aluminum effects on hyphal growth and penetration were usually
consistent in the range of 0 to 1 ppm Al, but not at 10 ppm Al. Both S. pellucida isolates exhibited reduced hyphal growth in up to 1 ppm Al, but at 10 ppm Al hyphal growth was stimulated. Isolate c-1-i of G.
manihotis had hyphal growth reduced at 10 ppm Al but gave hyphal growth as good or better at 1 ppm Al as at 0 ppm Al in both experiments. Penetration by all isolates was reduced at the 0.5 ppm Al level as compared to the 0 ppm Al level but increased at 10 ppm Al. This may indicate that at the higher concentrations aluminum precipitated from the system. Hepper and Smith (1976) found that, although Mn inhibited VAM fungal spore germination, it was not fungicidal. Removal of the spores from the Mn source and washing allowed germination to occur. A similar situation could have occurred here with aluminum.
In comparing the three screening methods, the results from the natural soils and limed soils tended to support each other while the results from the aluminum solution study do not clearly support the results of the first-two methods and sometimes contradicted them. For example spore germination of S. pellucida c-44-2 was positively correlated with Al in the natural soils experiment and spore germination increased quadratically and linearly as meq Al increased in the first
and second lime experiment, respectively. In only one solution experiment germination of c-44-2 was slightly depressed as Al concentration increased (Table 6, Fig. 17). The isolates of S. pellucida were used in only one solution study so these results are not
verified. A confounding factor in the aluminum solution study is that as Al concentration increased so did chloride (Cl), which was the carrier anion for aluminum. Hirrel (1981) found that Cl reduced germination of Gigaspora margarita. Penetration by G. manihotis c-i-i was positively correlated with Al in both of the soil screening methods, but penetration was depressed as Al increased in the solution method.
Aluminum forms not only monomeric but also polymeric complexes in solution and its toxicity is dependent on its form (Bohn et al., 1979). To understand the discrepancies between the results of the soil methods and the solution method, it would be useful to know which chemical forms of aluminum were present in the solution and which were present in the solutions of the soils. This was not examined in this study.
There are certain problems with the screening methods used in this study. It was hoped that the methods could provide a rapid selection technique for VAM fungi adapted to acid soils. Determination of penetration sites is not rapid and slows down the procedure. Initially it was hoped that germination would reflect the trend in penetration and that germination could be used instead of penetration. Unfortunately that is not the case. Only in G. mosseae, the fungus most sensitive to soil acidity, and in G.'manihotis c-1-1, the isolate least sensitive to soil acidity, were there good correlations between germination and penetration.
There are several problems with the use of a solution screening
method. The results of this study do not compare well with the results from screening in soil and ultimately the inoculation with VAM fungi will take place in the soil. It seems logical then that the screening should be done in soil. The solution method is actually more time consuming than the soil methods, since the solution pH must be constantly checked and solution occasionally replenished. In solution, screening is for adaption to one factor while in the soil screening is for adaption to an environment. That environment and screening procedure can be modified by fertilizer application, something that can
not be done in solution culture. Perhaps the most important problem with solution culture is that the VAM fungi do not germinate or develop well in this system. Wang (1984) used precolonized roots and avoided the problem of establishment of the VAM fungi. This fact could also account for some of the discrepancies between the soil and solution results. It is difficult to screen for the effect of some factor on an organism if the system used never gives good growth under supposedly nondetrimental conditions.
If one is to select VAM fungi for use, it would seem best to use growth in soil as the screening method. A soil representative of the area, crop and problem with which one is working should be chosen. Appropriate management practices for the soil and crop could be incorporated into the screening procedure. The use of a series of natur al soils from an area that vary in the desired characteristic should be avoided. This method presents less concise results due to the uncontrolled variation of several factors. As far as selecting an effective VAN fungus is conce rned, it would seem that selection should be geared toward adaptability to a broad soil environment. In other words, a fungus is needed that can grow well, even if not at its maximal level, in diverse soil environments. Glomus fasciculatum, reported on by Hayman and Tavares (1985), and G. manihotis in the present study seem to be broadly adapted fungi. This would be preferable to screening for adaptability to one factor.
EFFECT OF ISOLATES OF VESICULAR-ARBUSCULAR MYCORRHIZAL FUNGI SELECTED
FOR ALUMINUM TOLERANCE ON FIELD GROWTH OF Phaseolus vulgaris Introduction
Many experiments with VAM fungi have been conducted under greenhouse conditions, but results in the greenhouse are not necessarily translated to the field. When attempting field infestation with VAM fungi there are several important aspects to be considered. One, of course, is the potential of the chosen host crop for response to VAM fungi. In choosing the field site one must consider not only soil chemical characteristics such as pH and phosphorus, but also the indigenous population of VAM fungi. Lack of host growth response in the field may be due to competition among introduced and indigenous isolates as well as lack of adaptation of the introduced isolates to field conditions (Powell, 1982; Fitter, 1985; Howeler and Sieverding, 1983). Abbott and Robson (1981a) found that some species were more sensitive than other species to differences between two soils. The ability of Acaulospora laevis and Glomus monosporum to increase plant growth varied depending on soil type while that of Glomus fasciculatum did not. In one soil A. laevis and G. monosporum were capable of replacing some of the indigenous colonization with their own.
Azcon-Aguilar et al. (1986a) found an increase in percent
colonization from introduced VAM fungi only in fields normally low in
infectivity. They also found no correlation between VAN fungal colonization and dry matter, colonization and shoot P, and VAM fungal propagules and shoot P. They suggest that, because of the lack of correlation, a method is needed to screen each site. This suggestion does not seem to be practical. It would seem more practical to screen for an isolate with broad adaptability to the soil and climatic conditions prevailing in a particular area. Glomus fasciculatum in the work of Abbott and Robson (1981a) seems to be such an species.
The VAN fungi used in this field study were selected on the basis of the results of the experiments in Chapter II. The objectives were to evaluate the effect of VAN fungi selected for varying responses to soil aluminum on bean growth parameters, including shoot and root growth, grain yield and total aluminum in the shoot.
Materials and Methods
Based on previous experiments, four isolates of mycorrhizal fungi from two species were selected for evaluation in field trials. The purpose was to evaluate the effect of mycorrhizal fungi, which differed in their response to Al and Ca, on the shoot and root growth and yield of Phaseolus vulgaris.
Two field sites were selected on the farm of Don Fidel Chavez, in Mondomito, Cauca, Colombia, due to their low pH, relatively high percent Al saturation and the similarity of the soil characteristics of the two sites (Table 16). Both sites were fairly level but were bounded by steep slopes planted in cassava or fallow. One field site had been planted previously in cassava and the other had been fallow.
Table 16. Soil chemical analyses before liming and after liming of two field sites in Mondomito, Cauca,
Metric Concentration (ppm)
tons m 0 -1 oi
Field lime % a ppm P b g10g sol Effective % Al d Exchangeable Total
# ha- OMa Bray II pH Al Ca Mg K CECc saturation [H1+] Mn Fe N
1 4.6 1.6 4.2 3.5 0.60 0.11 0.06 4.57 73.7 0.3 29.1 108.6 1792
2 4.4 1.4 4.4 4.3 0.38 0.09 0.05 5.22 82.4 0.4 26.5 70.4 1456
Two months after liming:
1 0 4.9 26.5 4.4 3.10 0.93 0.23 0.24 5.40 57.4 0.9 38.0 129.0 2352
3 5.0 32.4 4.8 1.58 3.48 0.32 0.31 5.91 26.7 0.22 42.0 55.0 2352
6 5.1 39.2 5.6 0.30 8.40 0.39 0.32 9.41 3.2 35.0 20.7 2352
2 0 5.4 10.8 4.6 3.72 0.69 0.38 0.51 6.08 61.2 0.78 -11.0 115.0 2352
3 5.8 26.2 5.0 2.02 3.12 0.32 0.30 5.94 34.0 0.18 6.0 87.8 2352
6 5.6 32.9 5.2 1.14 4.98 0.41 0.3V .6.96 16.4 0.06 16.0 49.9 2352
a OM = organic matter.
b Bray II extraction method.
c CEC =cation exchange capacity.
d % Al saturation = (meq Al/effective CEC) X 100.
Both fields were plowed in January and Aldrin (Shell International Chemical Co., 20% aldrin) was applied at the rate of 25 kg ha-1 and incorporated to control cutworms and whitegrubs. About thirty soil samples from each field were combined and subsamples taken to determine the most probable number (MPN) of propagules of native mycorrhizal fungi. The treatments consisted of three lime levels, 0, 3, and 6 metric tons CaCO03 ha-1, and four treatments of introduced VAM fungi, and one treatment of native mycorrhizal fungi which received autoclaved inoculum. Each field had four replications and the design was a strip plot. All plots received 100 kg P ha-1 as triple superphosphate, 100 kg N ha-1 as urea, 100 kg K ha-1 as KCI, 25 kg Mg ha-1 as MgSO4, 5 kg Zn ha-1 as ZnSO4, and 1 kg B ha-1 as borax (Na2B407). The lime and P were broadcast and incorporated in March. In April the remaining fertilizers were broadcast on the plots and incorporated. At the same time the preemergence herbicide, Treflan (Drexel Chemical Co., trifluralin 0.48 kg emulsifiable concentrate liter-1), was applied to the fields at the rate of 120 ml 16 L-1 by back sprayer. Each field received the contents of one sprayer. The two fields were planted in the two following days. Inoculum Production
Inoculum production for the field experiment was initiated the previous November. Seven isolates of mycorrhizal fungi were multiplied in sterile soil, a typic dystropept (Instituto Geographico Agustin Codazzi, 1976), from Santander de Quilichao, Department of Cauca, Colombia. The legume Canavalia ensiformis L. DC. and cassava,
Manihot esculenta Crantz var M Col 113 were used as hosts. Twelve 10kg-pots of each isolate of mycorrhizal fungi were maintained in a screenhouse. Each pot was infested by placing about 20 g of soil and roots from a single-species pot culture of the selected mycorrhizal isolate below the cassava stakes. In each pot one stake of cassava and two seeds of C. ensiformis were planted. The two host species were used since it was known that some of the mycorrhizal fungal species gave better spore production on one host and some on the other (E. Sieverding, personal communication). The cassava stakes were treated before planting using the following mixture:
Dithane M-45 (Rohm & Haas, 80% mancozeb) 1 g L-1
Manzate D (E. I. du Pont de Nemours & Co. Inc., 80% maneb) 2 g L-1
Vitigran 3 (Hoechst A G, 360 g Cu L-1 2 ml L-1
ZnSO4 20 g L-1
Malathion (Cyanamid, 604 g 0,0-imethyl 2 ml
phosphorodithoate L- ) ml L
The cassava stakes were cut to 20 cm lengths, submerged in the
mixture for 5 min and left to dry in the shade. The following day the stakes were planted.
Five months after initiating the inoculum production and just before planting the fields, the four isolates of mycorrhizal fungi, intended as inoculum, were prepared for field application. All 12 pots of each isolate were used. The shoots were discarded and the root and soil mix were chopped and thoroughly mixed with a shovel. The thickened cassava roots which were beginning to form were cut out and discarded.
It was desirable to use two species each represented by two isolates to ascertain differences between species and isolates. Isolate c-i-i of G. manihotis was chosen because it showed strong positive correlation with meq Al and negative correlation with meq Ca. This isolate also produced good growth responses with cassava in previous work at CIAT (Centro Internacional de Agricultura Tropical, 1984). Isolate c-20-2 of G. manihotis was chosen as a contrast because it showed an occasional positive correlation with meq Ca but showed no significant negative correlation with meq Al. Two isolates (c-12-1 and c-94-5) of A. longula were chosen because they exhibited opposite tendencies with respect to meq Al and Ca. Spore germination of c-12-I was positively correlated to meq Ca and germination and penetration of c-94-5 were negatively correlated with meq Ca. Penetration of c-94-5 was also positively correlated to meq Al.
Planting and Sampling
The two fields had plots which were 5 and 4 m long, respectively, with six and five rows per plot, respectively, and rows were spaced 50 cm apart. Four rows of each plot received inoculum at the rate of 200 g m' of row. The inoculum was placed in the furrow directly below the seed of P. vulgaris var Carioca G 4017 which was planted at the rate of 20 seeds m-1. The isolates used were G. manihotis c-1-1 and c-20-2 and A. longula c-12-i and c-94-5. To compare the effect of the inoculum carrier to the effect of the introduced VAM fungi, inoculum of S. pellucida c-3-7, which had been prepared in the same manner as the other inoculum but had been sterilized in the autoclave, was applied
to the field in comparison plots at the same rate. At the time of planting, samples of soil from all the lime levels from both fields and samples of the inoculum were taken to perform an MPN test. Due to poor seed germination in field one, all the beans were pulled up and replanted 3 wk after the initial planting. Ascochyta leaf spot (caused by Ascochyta spp.) was controlled in field two by applying 20 g of Benomyl (E.I. du Pont de Nemours & Co. Inc., 50% benomyl) 16 L-1 by backpack sprayer followed by 40 ml of Derosal (Hoechst A G, 20% 2(methoxycarbomylamino)-benzimidazole) 16 L-1 by backpack sprayer. Only one application was necessary. Leaf cutting'ants were controlled in both fields by Aldrin.
Root and shoot samples were taken 1 mo after planting, at
initiation of flowering or about 55 d after planting and at pod fill. All root and shoot samples were taken from the outer two rows which received inoculum of VAM fungi and the inner two rows were left for harvest. The root sampling was done in the following manners. At the first sampling four whole root samples were taken from each plot, two from each outer row where inoculum containing VAM fungi had been placed. The root system was carefully excavated and placed in plastic bags to be later washed free of soil on a sieve. The entire shoots of the corresponding roots were also sampled to determine dry weight and nutrient content. After this first sampling the fields were thinned to one plant 10 cm-I by clipping the plants off at the soil line. At flowering a bipartite root auger (Eijkelkamp Equipment) with a diameter of 8 cm and a length of 15 cm was used. Five soil cores with root samples were taken per plot. One sample was taken directly over
the stem, two samples were taken 5 cm from the stem, one sample was taken 25 cm. from the row and the final sample was taken 12 cm from the row. The samples were always taken toward the next row where the inoculum had been applied. At pod filling four roots and shoots per plot were again sampled by excavating the entire root system and placing the roots in a plastic bag with adhering soil to be washed later. Root fresh weight was determined for all samples and a 0.25 g subsample was removed. Shoot and root dry weight were determined. The subsamples were cleared and stained with 10% KOH and 0.10% trypan blue (Phillips and Hayman, 1970), and percent colonization and root
length were determined using the grid-line method (G iovannetti and Mosse, 1980).
Roots from the second sampling for both fields were stored in the cold room at V0C until they were processed. A dispersing agent was needed to facilitate removal of roots from the cores. Sodium pyrophosphate, Na 2 P20 7, the usual agent was not desirable as the roots were to be analyzed for P. Solutions of NaOH and Na 20CO3 were tested (Kilmer and Alexander, 1949). The concentrations tested were 0.3 N,
0.03 N, and 0.0003 N NaOH and 0.6 N, 0.3 N, and 0.03 N NaCO 3. The
0.03 N NaCO 3 was selected for use because it was the lowest concentration which was effective in dispersing the soil. The 0.03 N solution of Na 2CO 3 was added directly to the soil cores in their plastic bags and allowed to sit overnight. The following day the roots were caught on a 500-rn sieve, washed free of soil and stored again in plastic bags. All the washed root systems were then observed in large petri dishes and any extraneous weed roots were removed with
tweezers. Organic material was removed by flotation in large beakers. At this point fresh weight and a random 0.25 g subsample were taken.
At harvest the first and last 20 cm: of the harvest rows were
discarded to act as a border. The two center rows of each plot were harvested and threshed by hand. Plant population in the two harvest rows was determined at the same time. Seed weight and percent seed humidity were determined. Seed, leaves, and roots from all sampling dates were analyzed for P, Ca, Al, and N by the methods previously indicated in the Materials and Methods of Chapter IL. Most Probable Number Test
As mentioned, samples of field soil and inoculum were taken at planting to be used in MPN tests. Fifty kilograms of soil from the CIAT substation at Santander de Quilichao were mixed in a 1:1 ratio with quartz sand. This was then sieved and sterilized for 2 hours in an autoclave. This soil mix was used to perform the soil dilutions for the MPN tests. The field and inoculum samples were sieved (0.3 cm2 sieve opening) and about 200 g of each sample were stored in plastic bags in the refrigerator until used. The sieve was washed and cleaned with 95% ETOR between each sample. All of the samples were diluted to a factor of 4-8, with 4-1 being used as the lowest dilution.
Seventy-five grams of the sample to be diluted were mixed thoroughly with 225 g of the sterile sand-soil mix to give 300 g of soil. This was the first dilution of 4-1. From this dilution four, 50-g samples were used for planting the MPN test, 75 g were mixed with 225 g of the sterile sand soil mix to give the 4-2 dilution and 25 g were
discarded. This dilution process was continued until a dilution of 48was reached. At the time of planting the MPN experiment, a sample of soil or inoculum was stored in plastic bags until percent humidity could be determined.
The four, 50-g samples from each dilution of the NPN test were planted in the following manner used in the Mycorrhizal Project at CIAT. One hundred grams of the sterile soil-sand mix were weighed into styrofoam cups. The 50-g sample of the diluted soil was placed on top of this and was then capped with another 100 g of the sterile soil-sand mix. In each cup, five, germinated, surface-sterilized seeds of kudzu (Pueria Phaseoloides) were planted and the cups were placed in the greenhouse.
At the end of 2 mo the MPN test was harvested. The cups were cut open with a scalpel and the entire middle cross-section of roots, where the original 50 g of diluted soil had been placed, was excised and saved for processing. The remaining shoots, roots and soil were discarded. The excised roots were washed free of soil, placed in test tubes and cleared and stained with 10% KOH and 0.1% trypan blue, respectively (Philips and Hayman, 1970). The entire sample of roots was then examined with a dissecting microscope for the presence (+) or absence (-) of colonization by mycorrhizal fungi. The number of infective propagules g-1 of soil was determined (Porter, 1979; Fisher and Yates, 1963).
The results of the two field studies were analyzed separately as an initial analysis indicated there was a field by lime interaction. Also shoot and root samples were taken at three periods throughout the growing season excluding harvest. Parameters of shoot and root weight, root length and percent colonization, and tissue analysis were analyzed by sample dates as this was also a significant effect.
The results of the MPN test on the field soil before and after
fertilization and of the inoculum used in the field are shown in Table 17. The number of VAM fungal propagules g-1 of soil were reduced in both fields after fertilization. Both fields had the number of propagules g-1 soil reduced with the addition of lime from 0 to 3 tons CaCO3 ha-I. At 6 tons of lime ha-I there was an increase in the
number of propagules g soil in field two. Of the inoculum used in the field, A. longula c-12-1 and G. manihotis c-i-i had the greatest numbers of propagules g-1 soil.
In both fields lime had a significant positive effect on shoot weight at all sampling times (Figs. 20 and 21). Three tons of lime ha-l usually produced shoot growth significantly greater than at 0 tons lime. Shoot growth at six tons lime was rarely significantly greater than at 3 tons of lime. In field one where water was limiting even with supplemental irrigation, there was a significant lime by mycorrhizae interaction effect on shoot weight at podfilling (Fig 20). In field one A. !ongula c-12-1 and G. manihotis c-1-i produced the greatest shoot weight at 6 tons of lime at flowering and podfilling. This was not verified in field two where a lime by mycorrhizal
Table 17. Number of propagules of vesicular-arbuscular mycorrhizal
fungi in field soil and inoculum as indicated by the most
probable number test.
No. propagules No. propagules
Source (field soil) g-1 soil Source g-1 soil
Field 1a before
fertilization 125.5 Control inoculumb 0.00
(36 to 248)e
Field 2a before
fertilization 24.12 Glomus
(7 to 49) manihotis c-l-1c 187.8
(76 to 514)
Field d 0 tons lime ha-1 11.89 Acaulospora
(4 to 30) longula c-94-5 11.76
(4 to 30)
3 tons lime ha 2.13 A. longula c-12-1 375.60
1(0.8 to 5.3) (163 to 1112)
6 tons lime ha- 4.04 G. manihotis c-20-2 94.64
(2 to 11) (36 to 248)
Field 2 0 tons lime ha-1 8.79
1 (4 to 24)
3 tons lime ha- 1.52
1(0.8 to 5.3)
6 tons lime ha 34.33
(13 to 86)
aSoil samples taken before fertilization or lime application.
bSample of autoclaved inoculum used as control inoculum taken at planting. The autoclaved inoculum consisted of roots and soil from pot cultures of S. pellucida c-3-7. cSamples taken from soil-root inoculum of each species at time of planting.
dSoil samples taken at planting 1 month after lime application.
eFiducial limits indicated in parentheses for number of propagules per gram soil according to Fisher and Yates, 1963.
SHOOT WEIGHT AT 30 DAYS AS FFFECTED BY LIME al.0
0OTIR L I C-I-I B C-94-5
3 C-12-1 C-20-2
0.0- 1ItS LDE t
SHOOT WEIGHT RT FLOWERING AFFECTED BY LIME S WIG4T AT rP FILLING S FECTED BY LIME
PMa I Pla
0 rlmL 8 C.- -1 C-04-5 0 O fil 5 C-1-1 9 C-94-5
[I C-12-1 9 C-20-2 0 C-12-1 0 C-20-2
ID LIEAt TCl LZPEM
Figure 20. Bean shoot weight in field one at three sampling dates as
affected by mycorrhizal treatment and lime. a) According to ANOVA no effect of mycorrhizae, contrast statements indicate lime effect is quadratic. b) Lime effect is linear, LSD for
mycorrhizal treatment averaged over all lime levels = 3.46
a = 0.05. c) ANOVA indicates a lime by mycorrhizal
treatment interaction, LSD = 7.99 for lime, LSD = 6.40 for
mycorrhizal treatment. Isolates of species used are as
follows: Glomus manihotis c-1-i and c-20-2; Acaulospora
longula c-12-1 and c-94-5. The control consisted of
autoclaved inoculum containing Scutellospora pellucida
SHOT WEIGHT AT 30 OYS AS IFFECTE BY LIME a ~ 1mI a
0 CONTROL I C-1-1 8 C-94-5
SAP r3 C-12-t 1 C-20-2
SICOT WIGHT AT FLOWERING AS S FFECTE Y LIM SOT WEIGHT FT PO FILLING AS AFFECTED BY LIIE maU a FUmD a
0 CONTROL I C-I -1 C-94-5 OINROL C--1 C-94-5
a C-12-1 C-20-2 MO 8 C-12-1 a C-20-2
TCM LXIE/M TO& L04EA#
Figure 21. Bean shoot weight in field two at three sampling dates as
affected by mycorrhizal treatment and lime. a) ANOVA
indicates a lime by mycorrhizal interaction LSD = 0.75 for
lime, LSD = 0.56 for mycorrhizal treatment a = 0.05. b) Contrast statements indicate lime effects are quadratic,
ANOVA indicates no effect of mycorrhizal treatment. c) Lime
effects are linear, no effects of mycorrhizal treatment
according to ANOVA. Isolates of species used are as
follows: Glomus manihotis c-1-1 and c-20-2; Acaulospora
longula c-12-1 and c-94-5. The control consisted of
autoclaved inoculum containing Scutellospora pellucida
treatment interaction occurred at 30 days (Fig. 21). Plants in plots which received sterilized inoculum (native mycorrhizal fungi) had the highest shoot weight at 3 tons of lime ha-1 and lowest at 6 tons lime ha-I while shoot weight produced by all other isolates except c-20-2 increased from 0 to 6 tons lime (Fig. 21).
Lime significantly increased root length quadratically and
linearly in fields one and two, respectively, during flowering but there was no significant effect of mycorrhizal treatment in either field (Figs. 22 and 23). Root length was also significantly increased by lime in field two, but not field one, at 30 days and podfilling. At 30 days the lime effect was quadratic and at podfilling it was linear. Root weight was not significantly affected by mycorrhizal treatment at 30 days or at flowering in either field but there was a lime by mycorrhizal treatment interaction during podfilling in field one (Figs. 24 and 25). Isolate of c-i-i of G. manihotis and isolate c-12-1 of A. longula produced the greatest root weights at 0 and 6 tons lime at pod filling (Fig. 25).
For percent root colonization in field two there was a lime by
mycorrhizal treatment interaction during podfilling in addition to the lime by mycorrhizal treatment interaction in both fields during flowering (Figs. 26 and 27). At flowering, percent colonization in the plots with native mycorrhizal fungi and plots with G. manihotis c20-2 increased with increasing lime in both fields. In field two colonization generally increased over the sampling dates, but in field one percent colonization often decreased at podfilling. Although percent colonization usually increased or remained the same with
FM LBEGh AS FFE 13Y VAN RHtF ISCIATE RG LIMO
OCMrROL NI -I- 8 C-94-5
MM UOC- 12-t
RT LB6ThM JU DFEM Y WF#4 FU48F ISORIN FRO LIPS RW LSNWH FI FFTDy BY YM UN. ISO Irs AM LIME
rPas pjn b UU
b 0 MNTRU. NI -I- 8 C-94-5 C Mm ONQTO C- I-I EC-94-5
O1C- 12-1 0 C-20-2 8 C-20-2 aOC- 12-1
IM LUIA T". LIG,
Figure 22. Bean root length in field one as affected by mycorrhizal
treatment, lime and sampling date. ANOVA indicates no
significant effect of lime in a or c and no significant
effects of mycorrhizal treatment in a, b or c. Contrast
statements indicate lime effects are quadratic in b.
Isolates of species used are as follows: Glomus manihotis
c-i-i and c-20-2; Acaulospora longula c-12-i and'c-94-5.
The control consisted of autoclaved inoculum containing
Scutellospora pellucida. c-3-7.