Eects of vesicular±arbuscular mycorrhizal inoculation on the
yield and phosphorus uptake of ®eld-grown barley
A. Khaliq*
, 1, F.E. Sanders
Department of Plant Sciences, University of Leeds, Leeds LS2 9JT, UK
Accepted 20 April 2000
Abstract
In a ®eld experiment on barley, the eects of soil application of phosphorus fertilizer and inoculum ofGlomus mosseaeon the mycorrhizal colonization of roots, crop yield and P uptake were evaluated in a natural or methyl bromide fumigated soil. Soil fumigation raised the amount of available nitrogen (NH4±N+NO3±N) in the soil by 13 mg kgÿ1. Although both soils were
equalized for the N status by adding N fertilizer to non-fumigated soil, fumigated soil gave higher crop yield. The applied P increased the dry matter yield signi®cantly but suppressed mycorrhizal infection. An overall small increase of 3% in total P uptake and a decrease of up to 2% in grain and straw yield were observed as a result of inoculation, which were statistically non-signi®cant at P<0:05: These trends in results were most likely due to the status of available soil P higher than the
threshold limit for a positive mycorrhizal growth response.72000 Elsevier Science Ltd. All rights reserved.
Keywords:VA-mycorrhiza; Barley; Crop yield; P uptake; Fumigation; Phosphorus; Field study
1. Introduction
Mycorrhizal roots due to their extramatrical hyphae that are capable of absorbing and translocating nutri-ents, can explore more soil volume than the non-mycorrhizal roots (Joner and Jakobsen, 1995), and thus increase the supply of slowly diusing ions, such as phosphate to the plant (McArther and Knowles, 1993). In return, the plant meets the carbon require-ment of the fungus (Jakobsen and Rosendahl, 1990). Although the roots of barley crop in the ®eld are always colonized with vesicular±arbuscular mycorrhi-zal (VAM) fungi, the eect of this association on the crop is variable perhaps due to complexity of the inter-action between symbiont and environment.
To measure the plant response to VAM infection,
mycorrhizal fungi need to be eliminated from the con-trol plants. This can be achieved by the use of certain crop rotations or various soil sterilization techniques (Jakobsen, 1994). It has been reported that VAM in-oculation enhanced the growth of barley signi®cantly in soils with low available P (Clarke and Mosse, 1981; Powell, 1981), while such improvement in barley growth was not signi®cant in an irradiated soil con-taining moderate amounts of available phosphorus (Jensen, 1984). Black and Tinker (1979) produced dierent sizes of VAM fungal populations in an arable ®eld by dierent crop rotations, and on the following barley crop they found a negative relationship between mycorrhizal infection and grain yield. However, the results were considered inconclusive owing to possible dierences in soil properties between the treatments. In a pot experiment conducted in a ®eld environment, Khaliq and Sanders (1998) used methyl bromide fumi-gated soil with adequate available P and noted that mycorrhizal fungi decreased barley yield. Fay et al., (1996) in a sand culture study also observed that when
P was applied in the range of 50±800 mg kgÿ1,
inocu-0038-0717/00/$ - see front matter72000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 3 8 - 0 7 1 7 ( 0 0 ) 0 0 0 8 6 - 9
www.elsevier.com/locate/soilbio
1
Department of Soil Science, University of Agriculture, Faisalabad 38040, Pakistan.
* Corresponding author. Tel.: 627793; fax: +44-92-41-647846.
lation with the mycorrhizal fungus Glomus mosseae consistently depressed the growth of barley plants.
Even very carefully carried out pot experiments can not truly predict the behavior of the ®eld crops. The objective of our study was, therefore, to evaluate the eect of VAM inoculation and P application on the yield and P uptake of ®eld-grown barley in a natural or methyl bromide fumigated soil.
2. Materials and methods
This experiment was conducted in Field 426 (Leeds University Farm, Headly Hall, Tadcaster) in a sandy±
clay±loam soil comprising (g kgÿ1) coarse sand, 133;
®ne sand, 374; silt, 165 and clay 328 of the Wother-some series (Crompton and Matthews, 1970). The soil
had available P (0.5 M NaHCO3 extractable), 25 mg
kgÿ1; nitrogen (2 M KCl extractable), 30 mg kgÿ1
(13.3 mg kgÿ1 as NH4±N and 16.6 mg kgÿ1 as NO3±
N); potassium (1 M NH4NO3 extractable), 136 mg
kgÿ1 and pH 7.4. Extractable N, P and K contents in
soil were determined as given by Gough (1973). The ®eld had been under fallow for 1 year after a kale crop, hence the soil had a low indigenous VAM
popu-lation (Glomusspore density 0.01±0.02 spores gÿ1).
The experimental area was ploughed, rotavated and divided into two halves separated by a 1 m wide path. One half was used as such (non-fumigated) and the other half was sealed under a polythene sheet and sub-jected to methyl bromide fumigation (200 kg methyl
bromide containing 10 g kgÿ1 chloropicrin haÿ1).
After 4 days the polythene sheet was removed and the area was left to aerate. The analysis of soil 15 days after fumigation (i.e., 8 days before sowing) revealed that the extractable N and P were higher by 13 and 5
mg kgÿ1, respectively in fumigated compared to
non-fumigated soil. Most of the N released was in NH4
form. The non-fumigated and fumigated areas were
each divided into four blocks (block size 66 m), and
each block into four plots (plot size 3 3 m). The
plots were either supplied with VAM free sand (M0)
or inoculum of Glomus mosseae (M1) and either given
no P (P0) or supplied with P (P1) at the rate of 100 kg
P haÿ1 as triple superphosphate (P 196 g kgÿ1).
Pot-assium (31 kg K haÿ1) was applied to all the plots as
muriate of potash (K 499 g kgÿ1), whereas N, 28 kg N
haÿ1as nitram (N 345 g kgÿ1) was applied only to the
non-fumigated plots to balance for the N released as a result of soil fumigation. No compensation was made for the P mobilized as a result of soil fumigation.
The mycorrhizal inoculum used was produced in a
glasshouse where inoculated (with sporocarps of
Glo-mus mosseae recovered from the original inoculum
brought from Rothamsted Experimental Station) or uninoculated maize plants were grown in dazomet
ster-ilized sand beds. The rhizosphere sand along with maize roots (after necessary homogenization) from the inoculated beds was used as inoculum and that from the controls as VAM free sand for the present
exper-iment at the rate of 20 kg mÿ2 on oven dry basis.
Inoculum application at this rate raised the Glomus
spore density of soil (to a depth of 15 cm) by 1 spore gÿ1.
The factorial combinations of M and P contents constituted four treatments (i.e., M0P0, M1P0, M0P1, M1P1) which were randomized in each block. Statisti-cally this experiment was treated as two independent sets i.e., one on non-fumigated and the other on fumi-gated area, each with a randomized block design.
Fertilizers and inoculum were broadcast and incor-porated in the soil to a depth of 15 cm by a rotavator;
fertilizers to the whole plot i.e., 33 m but inoculum
only to the central 2 2 m area of the plot. Barley
(Hordeum vulgare L. cv. Lofa Abed) seed was sown at
2.5 cm spacing in rows 14 cm apart using a Nordsten drill. Rotavator and drill were always run from one direction so that the movement of inoculum and fertili-zers across the plots was similar for the whole area.
After allowing for soil displacement, the central 22
m area of each plot was designated as the sampling area. Bird scares were used. The seedlings emerged 10 days after sowing. Both areas were hand-weeded. The fumigated plots had fewer weeds than non-fumigated plots. Plant populations were scored just before tiller-ing. The crop approached anthesis 61 days after sow-ing (DAS) and maturity at 124 DAS.
Root samples were taken from each plot (sampling area) at 40, 55, 84, 99 and 124 DAS. At each
sampling, 10 soil cores (2.5 cm diameter) plotÿ1 to a
depth of 15 cm were taken and bulked to give a com-posite sample from which the roots were washed out, cleared, stained (Phillips and Hayman, 1970) and assessed for mycorrhizal infection by the line-intersect method (Giovannetti and Mosse, 1980). The composite
samples of soil were taken at ÿ17, ÿ8, 25, 39, 54, 84,
102 and 124 days from sowing in the same way as that for root and analyzed for extractable soil N. At
har-vest (124 DAS) the central 2 2 m area (2 m 15
rows) of each plot was harvested and the components of crop yield were recorded. The dried samples of grain and straw after their digestion in a di-acid
mix-ture (HClO4, HNO3; 1:4, v/v) were analysed for P by
the vanadomolybdate method (Cavell, 1955).
The statistical analysis of data was carried out by conducting ANOVA. Comparison of the means was made using least signi®cant dierence (LSD) or stan-dard error of the mean (SEM). All the data are
pre-sented as actual (non-transformed) means. The
after arcsine transformation (Snedecor and Cochran, 1980).
3. Results
3.1. Mycorrhizal infection
The mycorrhizal colonization of roots generally increased with time (Fig. 1). Inoculated treatments had
signi®cantly P<0:01, 0.001) higher mycorrhizal
infec-tion than uninoculated treatments in non-fumigated as well as in fumigated soil. Mycorrhizal infection did not
exceed 0.6 cm mÿ1 in uninoculated fumigated plots
during the experiment. The plots given phosphate had lower amounts of infection than their respective con-trol plots not supplied with phosphate. The inhibitory eect of P on VAM colonization of roots was
signi®-cant P<0:001, 0.01, 0.05) at the ®rst sampling in
non-fumigated soil and at the ®rst, third and ®fth
sam-plings in fumigated soil. The M P interaction also
suppressed the amounts of mycorrhizal infection in
both soils but signi®cantly P<0:01, 0.05) only at the
®rst sampling in non-fumigated soil and at ®rst, second and ®fth samplings in fumigated soil.
3.2. Yield
The application of P fertilizer signi®cantly increased
grain, straw and grain plus straw weights mÿ2in
non-fumigated as well as in non-fumigated soil (Table 1; Fig. 2).
Harvest index and number of ears plantÿ1 were also
higher in plots supplied with P fertilizer (M0P1,
M1P1) than in plots with no added P (M0P0, M1P0). The eect was signi®cant in fumigated soil.
3.3. Plant phosphorus
Phosphorus concentrations in grain and straw were increased in plants by P application and by mycorrhi-zal inoculation in both soils, however, the P appli-cation eects were signi®cant. Similarly, the P contents of grain and straw were signi®cantly increased by P Table 1
Eect of mycorrhizal inoculation and P application on barley yield and its components in non-fumigated (NF) and fumigated (F) soila
Parameter Soil Treatment means,n4
M0P0 M1P0 M0P1 M1P1
Grain DW, g mÿ2 NF 477 b 462 b 541 a 521 a F 504 b 487 b 557 a 537 a Straw DW, g mÿ2 NF 721 b 703 b 797 a 776 a F 776 b 756 b 836 a 815 a Harvest index NF 0.398 a 0.397 a 0.404 a 0.402 a
F 0.394 b 0.392 b 0.401 a 0.398 ab Number of plants mÿ2 NF 247 a 249 a 257 a 247 a
F 265 a 253 a 255 a 257 a Number of ears plantÿ1 NF 4.07 a 3.92 a 4.24 a 4.26 a
F 3.95 b 4.12 ab 4.45 a 4.28 ab Number of grains earÿ1 NF 12.5 a 12.5 a 12.8 a 12.9 a
F 12.4 a 12.1 a 12.3 a 12.5 a 1000 grain weight, g NF 38.1 a 37.8 a 38.8 a 38.7 a F 38.9 a 38.6 a 39.7 a 39.5 a
a
Values sharing the same letter(s) in a row do not dier signi®-cantly atP<0:05: DW, dry weight; Harvest index = grain/(grain + straw) dry weight ratio.
Fig. 2. Eect of treatments on dry matter yield and P uptake of bar-ley at harvest. The data bars (means,n4with same letter in each soil do not dier signi®cantly atP<0:05: Error bars show2SEM. DW, dry weight.
application but not by mycorrhizal inoculation (Table 2; Fig. 2).
3.4. Extractable soil nitrogen
Inoculation with VAM fungus and application of P fertilizer had no eect on the amounts of total
extrac-table soil N (NH4±N + NO3±N) in either of the two
soils during the season. The data were, therefore,
pre-sented as mean n16 of all the plots in each soil
(Fig. 3).
3.5. Comparison between soils
A valid statistical comparison between the non-fumi-gated and fuminon-fumi-gated soil could not be made because there was no replication, however, the obvious dier-ences were encountered.
At each P rate the inoculated plants had higher amounts of mycorrhizal infection in non-fumigated than in fumigated soil. The dierence was slight at P0 but considerable at P1 (Fig. 1). On overall basis, the fumigated soil had slightly higher plant populations than the non-fumigated soil (Table 1). The yield and plant P (concentration and contents) were also higher in fumigated than in non-fumigated soil (Tables 1 and 2; Fig. 2). After the application of additional N to the non-fumigated soil, its total extractable N content remained similar to that of the fumigated soil till the
last sampling (124 DAS), however, the NH4±N
remained higher in fumigated than in non-fumigated soil for the period to 39 DAS (Fig. 3).
4. Discussion
Negligible mycorrhizal infection in the fumigated uninoculated treatments during the season showed that
methyl bromide application at the rate of 200 kg haÿ1
eectively eliminated the indigenous VAM population. Generally, the development of VAM infection seemed to follow a three phase growth curve i.e. a lag phase, a phase of rapid growth and a phase of constancy as observed in previous studies (Saif, 1977; Khaliq and Sanders, 1997). The overall higher VAM colonization levels of plant roots in non-fumigated compared to fumigated soil in the respective treatments were most probably due to the presence of an indigenous VAM population in the non-fumigated soil. Moreover, a
dierential increase (5 mg kgÿ1) in the availability of P
in fumigated soil might be partly responsible for its lower levels of VAM colonization, as mycorrhizal infection was also suppressed in both soils by the ferti-lizer P. This is in accordance with earlier workers who found that the supply of P to plant by soil (Khaliq and Sanders, 1997; Thingstrup et al., 1998) or by foliar application (Sanders, 1975) decreased VAM coloniza-tion. The inhibitory eect of P application on VAM infection was more pronounced in the fumigated soil, indicating that the single population of the introduced
fungus (Glomus mosseae) was more sensitive to
fertili-zer P than the community of indigenous and, indigen-ous and introduced VAM fungi. VAM fungal species can dier signi®cantly in their tolerance to soil P con-centration (Pearson et al., 1994). Abbott and Robson (1978) also observed the better adaptation of native VAM fungi than the added endophyte to higher amounts of soil P. As a result of inoculation, the
increase in the fraction of root mycorrhizal (cm mÿ1)
over control (non-inoculated) was higher in fumigated than in non-fumigated soil. This could be attributed to the reduction of antagonistic (Warnock et al., 1982) and competitive (Powell, 1979) eects of the native soil micro¯ora by soil fumigation. The comparison of a Table 2
P concentration and P contents of barley (on oven dry basis) at har-vesta
Parameter Soil Treatment means,n4
M0P0 M1P0 M0P1 M1P1
Grain P concentration, mg gÿ1 NF 0.34 a 0.35 a 0.38 a 0.39 a F 0.36 b 0.37 ab 0.42 ab 0.43 a Straw P concentration, mg gÿ1 NF 0.12 b 0.13 ab 0.14 a 0.14 a F 0.16 b 0.17 b 0.18 ab 0.20 a Grain P, g mÿ2 NF 1.62 b 1.63 b 2.08 a 2.03 a F 1.83 b 1.81 b 2.32 a 2.30 a Straw P, g mÿ2 NF 0.83 b 0.92 b 1.15 a 1.11 a F 1.23 c 1.30 bc 1.52 ab 1.60 a
aValues sharing the same letter(s) in a row do not dier signi®-cantly atP<0:05:NF, non-fumigated; F, fumigated.
single species versus a mixture of species (i.e. how they respond to a factor) is, however, questionable as it does not preclude the genetic dierences between the two populations.
The non-fumigated soil was supplied with extra N to balance the N concentration in both soils. The ana-lytical results showed that this aim was achieved. An overall higher crop yield from the fumigated compared to non-fumigated area, which was mainly due to the slightly higher 1000-grain weight and plant population of the fumigated area, possibly resulted from the release of P and other nutrients from the lysed micro-organisms (Rovira and Ridge, 1979). The presence of
higher proportion of NH4than NO3nitrogen up to 39
DAS in the fumigated soil might also have contributed to its higher grain and straw yield than that of the non-fumigated soil. This ®nding supports the results of Camberato and Bock (1990) who obtained signi®cantly higher dry matter yield in wheat by maintaining the
NH4 fraction of mineral N in soil higher than in
con-trols. The root system development was observed to be much better in the fumigated than in non-fumigated soil. This was most probably the consequence of the aforementioned nutritional eects of fumigation, while in non-fumigated soil the microorganisms (eliminated by fumigation) might have aected root growth
posi-tively (plant growth promoting rhizobacteria Ð
PGPR) or negatively (parasites or pathogens).
The literature shows that barley can respond posi-tively to VAM inoculation at available soil P contents
<10±20 mg kgÿ1 (Clarke and Mosse, 1981; Jakobsen
and Jensen, 1981; Powell, 1981; Jensen, 1982). A very small but consistent reduction in the ®nal yield was found in inoculated compared to uninoculated treat-ments. Adequate amounts of available soil P and the dense and ®brous root system of the plant can make mycorrhiza super¯uous, and thus parasitic, because of less bene®t to the plant in terms of improved mineral nutrition for a given exchange of photosynthate to the fungus (Pang and Paul, 1980; Bethlenfalvay et al., 1982). The original concentration of available P (25
mg kgÿ1) and the ecient root system of barley
prob-ably maintained a supply of P to the plants higher than the threshold limit of positive mycorrhizal growth response in the crop. The adequate P status of the ®eld is re¯ected from the grain and straw yields com-parable to the range of yields obtained under normal agricultural practice. Moreover, a high dose of P (100
kg haÿ1) gave a rather small overall increase (up to
11%) in the grain + straw yield.
The inoculum used in the present experiment was from the same lot and applied at the same rate as that for our pot experiments in which VAM infection was
proli®c (70 cm mÿ1) in maize, with an associated
sig-ni®cant enhancement in plant growth (Khaliq and Sanders, 1997); whereas, the amount of infection
obtained was lower in barley (40 cm mÿ1) leading to a
signi®cant reduction in yield (Khaliq and Sanders, 1998). The diering responses in our experiments were most likely due to dierences in the functioning of the symbiosis and not due to dierences in the inoculum. The smaller magnitude of the mycorrhizal eect in the present experiment was probably due to the greater ®eld variability, and the overall lower amounts of mycorrhizal infection as a result of higher status of soil P, compared to that of the barley pot experiment. In this context it is important to add that the barley variety (cv. Lofa Abed) used in our experiments was a high root-shoot ratio type, selected on the basis of a previous VAM inoculation study on barley (A. Khaliq unpub. PhD thesis, Leeds University, 1983), indicating that at an adequate concentration of soil P the de-pression in plant growth resulting from mycorrhizal infection might be greater in a high root±shoot ratio than in a low root±shoot ratio variety. Hetrick et al. (1996) also concluded that this relationship occured in wheat cultivars.
Most of the published studies on mycorrhizal growth responses in plants have described the useful-ness of mycorrhizal symbiosis in P-de®cient soils and very few have reported its adverse eects at adequate P levels. This might lead to an unrealistic picture of the eects of mycorrhiza in normal agriculture. There-fore, there is need to develop a broader view of the nature of mycorrhizal symbiosis which apart from being mutualistic can also be null or even negative (Francis and Read, 1995). A further study is warranted to evaluate the eect of VAM inoculation on barley at a wider range of available soil P contents.
References
Abbott, L.K., Robson, A.D., 1978. Growth of subterranean clover in relation to the formation of endomycorrhizas by introduced and indigenous fungi in a ®eld soil. New Phytologist 81, 575±585. Bethlenfalvay, G.J., Brown, S.M., Pacovsky, R.S., 1982. Parasitic and mutualistic associations between mycorrhizal fungus and soy-bean: development of the host plant. Phytopathology 72, 889± 893.
Black, R., Tinker, P.B., 1979. The development of endomycorrhizal root systems. Part II: Eect of agronomic factors and soil con-ditions on the development of vesicular±arbuscular mycorrhizal infection in barley and on the endophyte spore density. New Phytologist 83, 401±413.
Camberato, J.J., Bock, B.R., 1990. Spring wheat response to enhanced ammonium supply. Part I: Dry matter and nitrogen content. Agronomy Journal 82, 463±467.
Cavell, A.J., 1955. The colorimetric determination of phosphorus in plant materials. Journal of the Science of Food and Agriculture 6, 479±480.
Memoirs of the Soil Survey of Great Britain (England and Wales), pp. 81±144.
Fay, P., Mitchell, D.T., Osborne, B.A., 1996. Photosynthesis and nutrient-use eciency of barley in response to low arbuscular mycorrhizal colonization and addition of phosphorus. New Phytologist 132, 425±433.
Francis, R., Read, D.J., 1995. Mutualism and antagonism in the mycorrhizal symbiosis, with special reference to impact on plant community structure. Canadian Journal of Botany 73, S1301± S1309.
Giovannetti, M., Mosse, B., 1980. An evaluation technique to measure vesicular±arbuscular infection in roots. New Phytologist 84, 489±500.
Gough, H.C., 1973. The Analysis of Agricultural Materials. Technical Bulletin 27. Her Majesty's Stationary Oce, London. Hetrick, B.A.D., Wilson, G.W.T., Todd, T.C., 1996. Mycorrhizal
re-sponse in wheat cultivars: relationship to phosphorus. Canadian Journal of Botany 74, 19±25.
Jakobsen, I., 1994. Research approaches to study the functioning of vesicular±arbuscular mycorrhizas in the ®eld. Plant and Soil 159, 141±147.
Jakobsen, I., Jensen, A., 1981. In¯uence of vesicular±arbuscular mycorrhiza and straw mulch on growth of barley. Plant and Soil 62, 157±161.
Jakobsen, I., Rosendahl, L., 1990. Carbon ¯ow into soil and external hyphae from roots of mycorrhizal cucumber plants. New Phytologist 115, 77±83.
Jensen, A., 1982. In¯uence of four vesicular±arbuscular mycorrhizal fungi on nutrient uptake and growth in barley (Hordeum vul-gare). New Phytologist 90, 45±50.
Jensen, A., 1984. Responses of barley, pea and maize to inoculation with dierent vesicular±arbuscular mycorrhizal fungi on irra-diated soil. Plant and Soil 78, 315±323.
Joner, E.J., Jakobsen, I., 1995. Growth and extracellular phospha-tase activity of arbuscular mycorrhizal hyphae as in¯uenced by soil organic matter. Soil Biology and Biochemistry 27, 1153±1159. Khaliq, A., Sanders, F.E., 1997. Eects of phosphorus application and vesicular±arbuscular mycorrhizal inoculation on the growth and phosphorus nutrition of maize. Journal of Plant Nutrition 20, 1607±1616.
Khaliq, A., Sanders, F.E., 1998. Eects of vesicular±arbuscular mycorrhizal inoculation on the growth and phosphorus nutrition
of barley in natural or methyl bromide treated soil. Journal of Plant Nutrition 21, 2163±2177.
McArther, D.A.J., Knowles, N.R., 1993. In¯uence of VAM and phosphorus nutrition on growth, development and mineral nutri-tion of potato. Plant Physiology 102, 771±782.
Pang, P.C., Paul, E.A., 1980. The eects of vesicular±arbuscular mycorrhiza on14C and15N distribution in nodulated faba beans. Canadian Journal of Soil Science 60, 241±250.
Pearson, J.N., Abbott, L.K., Jasper, D.A., 1994. Phosphorus, soluble carbohydrates and the competition between two arbuscular mycorrhizal fungi colonizing subterranean clover. New Phytologist 127, 101±106.
Phillips, J.M., Hayman, D.S., 1970. Improved procedures for clear-ing roots and stainclear-ing parasitic and vesicular±arbuscular mycor-rhizal fungi for rapid assessment of infection. Transactions of the British Mycological Society 55, 158±161.
Powell, C.L., 1979. Spread of mycorrhizal fungi through soil. New Zealand Journal of Agricultural Research 22, 335±340.
Powell, C.L., 1981. Inoculation of barley with ecient mycorrhizal fungi stimulates seed yield. Plant and Soil 59, 487±489.
Rovira, A.D., Ridge, E.H., 1979. The eect of methyl bromide and chloropicrin on some chemical properties of soil and on the growth and nutrition of wheat. In: Mulder, D. (Ed.), Soil Disinfestation. Elsevier, Amsterdam, pp. 231±250.
Saif, S.R., 1977. The in¯uence of stage of host development on ves-icular±arbuscular mycorrhizae and Endogonaceae spore popu-lation in ®eld grown crops. New Phytologist 79, 341±348. Sanders, F.E., 1975. The eect of foliar-applied phosphate on the
mycorrhizal infection in onion roots. In: Sanders, F.E., Mosse, B., Tinker, P.B. (Eds.), Endomycorrhizas. Academic Press, London, pp. 261±276.
Snedecor, G.W., Cochran, W.G., 1980. Statistical Methods. The Iowa State University Press, Ames.
Thingstrup, I., Rubñk, G., Sibbesen, E., Jakobsen, I., 1998. Flax (Linum usitatissimum L.) depends on arbuscular mycorrhizal fungi for growth and P uptake at intermediate but not high soil P levels in the ®eld. Plant and Soil 203, 37±46.