Plant and Soil 220: 207218, 2000. 2000 Kluwer Academic Publishers. Printed in the Netherlands. 207
Wheat responses to aggressive and non-aggressive arbuscular mycorrhizalfungi
J. H. Graham1, and L. K. Abbott21University of Florida, Citrus Research and Education Center, Lake Alfred, 33850 USA; 2Soil Science & PlantNutrition, Univ. Western Australia, Nedlands, WA 6907 Australia
Received 28 July 1999. Accepted in revised form 12 January 2000
Key words: carbon costs, host-fungus compatibility, parasitism, plant growth depression, root colonization, soilphosphorus supply
In southwestern Australia fields, colonization of wheat roots by arbuscular mycorrhizal fungi (AMF) is reduceddue to repeated use of phosphate (P) fertilizers. We predicted AMF that aggressively colonize wheat roots atlow P supply would also aggressively colonize at high P supply, but provide no additional P uptake benefit andreduce growth. Wheat (cv. Kulin) seedlings were non-mycorrhizal (NM) or inoculated separately with 10 isolatesof AMF from wheat-belt soils in a glasshouse experiment. Kojonup loamy sand was supplied with P to providesuboptimal and supraoptimal P for growth of NM wheat in this soil. At low P supply, wheat growth was limitedby P availability. All AMF isolates colonized wheat roots at 14 days after emergence of seedlings. At 42 days,percentage root length colonized (%RLC) was highest for two isolates of Scutellospora calospora, WUM 12(2)and WUM 12(3), followed by Glomus sp. WUM 51, G. invermaium WUM 10(1), Acaulospora laevis WUM 11(4)and Gigaspora decipiens WUM 6(1). These isolates, designated as aggressive colonizers, ranged from 50 to89%RLC. A second group of AMF ranged from 1 to 19%RLC at 42 days. This group, termed non-aggressivecolonizers, included Acaulospora spp. WUM 11(1), WUM 46, and WUM 49 and Glomus sp. WUM 44. High soilP supply increased seedling growth 23 fold, but reduced%RLC. Grouping of aggressive and non-aggressive AMFbased on colonization rate at high P supply was similar to that at low P. At low P supply, only the two isolatesof S. calospora increased wheat growth compared to the NM plant. The remaining aggressive and non-aggressiveAMF reduced growth of wheat at low P, while aggressive colonizers reduced growth at high P. At low P supply,the aggressive colonizers increased shoot P concentration, while at high P, shoot P was not affected by AMF.Growth depression by aggressive colonizers was associated with reduced sucrose concentration in roots. Based onthe negative growth response under low and high P fertility in the glasshouse, AMF could be expected to producenon-beneficial effects on wheat in the field depending on the P status of the soil and the aggressiveness of AMF inthe community.
Arbuscular mycorrhizal fungi (AMF) are a major de-terminant of plant growth response in a crop soil.Effective AMF are those that produce the greatest be-nefit in terms of increased P acquisition for the leastcost expressed as C expenditure on mycorrhizas (Gra-ham and Eissenstat, 1994). Species of AMF in the FAX No: 863 956 4631. E-mail: firstname.lastname@example.org
genera Glomus and Acaulospora are considered moreeffective and are more abundant in many agriculturalsoils than species of Gigaspora and Scutellospora, thatmay decrease in abundance under cultivation (Millerand Jastrow, 1992; Siqueira et al., 1989). Obviously,agricultural managements such as tillage, fertilizationand crop rotations exert selection pressure in favorof AMF that are tolerant of such practices (Johnsonand Pfleger, 1992). Although presently a controversialconcept, some researchers have suggested that certain
decline diseases of crops may be related to a shift frombeneficial to non-beneficial or even parasitic fungi incontinuous crop monoculture under moderate to highfertility conditions (Johnson et al., 1992; Johnson,1993; Schenck and Siqueira, 1987).
Glasshouse studies predict AMF that aggressivelycolonize roots and stimulate plant growth at low Psupply (Abbott and Robson, 1981), will also aggress-ively colonize at high P supply, but may provide noadditional P benefit and reduce growth (Graham etal., 1996). Wilson and Trinick (1983) first used theterm aggressiveness to refer to the ability of AMF tocompete with other fungi for colonization space in theroot. In contrast, Graham et al. (1996) defined aggress-iveness as the rate of root colonization in the absenceof competing AMF. Rate of colonization is a usefulpredictor of growth response to increased P uptake byindividual AMF from low P soil (Abbott and Robson,1981). Fungi that aggressively colonize roots com-monly occur in the field (Abbott and Robson, 1991;Brundrett, 1991), but their relevance to mycorrhizaleffectiveness remains to be determined (Abbott andGazey, 1994). Carbon cost analysis of citrus (Grahamet al., 1996; Peng et al., 1993) and observations ofseveral crops in the field reveal that effectiveness ofAMF from managed crop soils ranges from mutual-istic to parasitic depending mostly on the soil P supply(Johnson et al., 1997). In moderate to high P soils,growth of field-grown citrus supports the hypothesisthat within communities of AMF there exist aggress-ive fungi capable of colonization sufficient to producecarbon cost without off-setting benefit (Graham andEissenstat, 1998).
Given the pervasive over-application of P to cropsoils, decreasing input of P fertilizer is most often citedas the approach to increase symbiotic effectiveness.However, such changes in fertilizer management mustbe based on knowledge of the functional attributes ofAMF species and populations. These attributes are:1. rate and development of colonization of roots (Ab-bott and Robson, 1981); 2. proliferation of externalhyphae in relation to P acquisition (Jakobsen et al.,1992); 3. carbon cost to support growth and mainten-ance of the mycorrhizal root (Peng et al., 1993). Rootsof wheat become colonized by AMF to different ex-tents depending on climate, soils, cropping practicesand fertilizer history. In Australian wheat fields, col-onization of wheat roots by AMF is reduced due torepeated use of phosphate fertilizers (Ryan et al.,1994;C. Blackburn, L. K. Abbott and A. D. Robson, un-published data). The impact of AMF on wheat growth
under high P fertility is not established. However,Thompson (1987) showed that reduction in mycor-rhizal colonization by fallowing had a minimal effecton growth of wheat in cracking clay soil with mod-erate levels of bicarbonate extractable P (1016 mgkg1 soil) from the northeastern area of the Australianwheat belt.
The primary objective of this study was to eval-uate the rate and extent to which AMF isolated fromsouthwestern Australia soils colonize wheat roots. Thesecondary objective was to determine how this col-onization affected plant growth, P nutrition and car-bohydrate status of roots at soil P availability similarto levels that have accumulated after fertilization ofwheat fields.
Materials and methods
Unfertilized Kojonup loamy sand with a pH of 6.0(1 mM CaCl2) and bicarbonate extractable P of4.0 mg.kg1(Colwell, 1963), was collected fromnative bush near Clackline, Western Australia. Soilwas passed through a 1.4 mm sieve to removecoarse organic matter and then pasteurized withaerated steam for 1 h on two consecutive daysat 70 C. Nutrients were mixed with dry soil atthe following rates (mg.kg1dry soil): K2SO4, 71;CaCl2, 71; MGSO4.7H2O, 20; MnSO4.H20, 10;ZnSO4.7H2O, 5; CuSO4.H2O, 2; CoSO4.7H2O, 0.35;NaMoO4.2H2O, 0.18 to prevent nutrient deficienciesin wheat except for P and N. Nitrogen was suppliedevery 2 weeks at 37.5 mg kg1 as NH4NO3. Phos-phorus was amended with soil at 10 and 60 mg.kg1P, as KH2PO4. The two rates of P provided 4 and42 mg.kg1extractable P for either suboptimal andmaximal growth of non-mycorrhizal (NM) wheat.Phosphorus supply levels were based on the extract-able P range from southwestern wheat-belt fields (2045 mg.kg1) and on the extractable P in Kojonuploamy sand that produced a beneficial response toAMF (approx. 810 mg kg1) in a preliminary glass-house trial of wheat (Triticum aestivum cv. Kulin).
AMF and colonization potential calibration
Ten isolates of the four major genera of AMF, Glomus,Acaulospora, Gigaspora and Scutellospora, were ob-tained from agricultural soils, revegetated land and
Table 1. Description of arbuscular mycorrhizal fungi (AMF) studieda
AMF species Isolate no. Origin
Acaulospora laevis WUM 11(4) Badgingarra, WA, low pH, under pastureA. laevis WUM 11(1) Mt. Barker, WA, low pH, under wheatScutellospora calospora WUM 12(2) Badgingarra, WA, low pH, under native bushS. calospora WUM 12(3) Badgingarra, WA, low pH, under pastureGigaspora decipiens WUM 6(1) Harvey, WA, clover pastureGlomus invermaium WUM 10(1) Merredin, WA, capeweed/cloverGlomus sp. WUM 44 Jarrahdale, WA, sandy loam, pH 6.0, revegetatedAcaulospora sp. WUM 46 Jarrahdale, WA, sandy loam, pH 4.8, Jarrah forestAcaulospora sp. WUM 49 Jarrahdale, WA, sandy loam, pH 5.0, undisturbed heathlandGlomus sp. WUM 51 Capel, WA, sand, pH 5.8, under pasture
aAll fungi in pot culture at the University Western Australia, Soil Science and Plant Nutrition.
native bush representative of wheat belt soils in south-western Australia (Table 1). To compare wheat re-sponses to several AMF, it was necessary to verifyand calibrate the colonization potential of each isol-ate prior to the study. Pot cultures of the 10 fungipropagated on subterranean clover (Trifolium subter-raneum cv. Seaton Park) in Lancelin sand (Gazeyet al., 1992) were assayed on subterranean clover atthree inoculum densities: 10, 25 and 75 g pot culturesoil per 0.5 kg Kojonup loamy sand and screened onwheat at 75 g per 0.5 kg. Phosphorus supply (20 mgkg1 P added as KH2PO4) was low for subterraneanclover and moderate for wheat. Nitrogen was suppliedas NH4NO3 for wheat and rhizobial inoculation forN2-fixing subterranean clover. The trial was conduc-ted in the glasshouse from July to August 1996 with600 mol.m2. s1 photosynthetically-active radi-ation (PAR) and a controlled soil temperature of 20 C.Seedlings were harvested at 28 days after emergenceand percentage root length colonized (%RLC) was de-termined by clearing and staining, and measurementin roots of hyphae, arbuscules or vesicles dependingon the AMF by the line intersect method (Giovanettiand Mosse, 1980).
The 10 isolates colonized wheat and produced2040%RLC on subterranean clover after 28 days atthe intermediate and high inoculum densities. An in-oculum density of 100 g per 1.5 kg soil was chosenfor further characterization of symbiotic effectivenessin the final experiment.
Plant growth conditions and harvest
Pre-germinated wheat seedlings were transplantedinto non-inoculated Kojonup loamy sand or soil in-
oculated with each of 10 AMF contained in pots of1.5 kg soil. Non-mycorrhizal (NM) plants receiveda microbial extract passed through a 38 m sieveof the composite of soil from the 10 inoculant fungi(Table 1). Mycorrhizal treatments were exposed to asimilar combined extract of the 10 AMF. The glass-house experiment was conducted from September toNovember 1996 (spring) under a light intensity of800mol.m2. s1 PAR and controlled soil temper-ature conditions (20 C). The experimental design wascompletely randomized with 2 P levels, 11 AMF (in-cluding NM) and three harvests at 14, 28, and 42 daysafter seedling emergence. There were three replicateseedlings each harvest per AMF and P treatment.
Plant growth and other parameters were determ-ined at each of three harvests. Seedlings were analyzedfor dry weight (70 C, 24 h) and percentage rootlength colonized (%RLC). Leaf nutrient concentrationof P, K, Ca, Mg, Fe, Zn, Mn, Cu was analyzed byinductively coupled plasma atomic emission spectro-scopy after acid digestion of dried and ground tissue.Root starch and sucrose were analyzed in ground roottissue as previously described (Graham et al., 1996)and expressed on a percent dry weight basis. Plantgrowth and carbohydrate responses to AMF were ex-pressed relative to the NM response by calculatingthe mycorrhizal response as [AM dry weight/NM dryweight] - 1. A three-factor analysis of variance (AMF,P supply, harvest) for each parameter was performedusing the general linear models procedure (SAS Insti-tute, Cary, NC USA). The%RLC of groups of AMFwas compared by orthogonal contrast analysis at P 0.05. Comparisons of parameters within groupsof AMF were performed at each harvest time using
Figure 1. Percentage root length colonized (%RLC) and mycorrhizal response of wheat grown at low and high P supply at 14, 28 and 42 daysafter inoculation with six aggressive arbuscular mycorrhizal fungi (AMF) from southwestern Australia soils. Mycorrhizal response is calculatedas [AM dry weight response/non-mycorrhizal response] 1. Significant differences (P0.05) in %RLC among AMF at 42 days are indicatedby unlike letters following each curve. For mycorrizal responses, the bars represent +/ 1 standard error of 3 replications. See Table 1 foridentification of the AMF.
Student-Newman-Keuls multiple range test at P 0.05.
All treatment factors, AMF, P supply and harvest, sig-nificantly affected and interacted with root coloniza-tion (Table 2). At low P supply, AMF colonized wheatroots at 14 days after emergence, the earliest stage
of seedling development evaluated (Figures 1 and 2).Isolates of S. calospora, WUM 12(2) and WUM 12(3),had already produced abundant external hyphae andauxiliary cells at this early stage of colonization. At42 days,%RLC was highest for the two isolates of S.calospora, WUM 12(2) and WUM 12(3), followed byGlomus sp. WUM 51, G. invermaium WUM 10(1),Gi. decipiens WUM 6(1) and A. laevis WUM 11(4)(Figure 1). These six isolates, designated as aggress-
Figure 2. Percentage root length colonized (%RLC) and mycorrhizal response of wheat grown at low and high P supply at 14, 28 and 42 daysafter inoculation with four non-aggressive arbuscular mycorrhizal fungi (AMF) from southwestern Australia soils. Mycorrhizal response iscalculated as [AM dry weight response/non-mycorrhizal response] 1. Significant differences (P0.05) in%RLC among AMF at 42 days areindicated by unlike letters following each curve. For mycorrizal responses, the bars represent +/ 1 standard error of 3 replications. See Table 1for identification of the AMF.
ive colonizers, ranged from 5089%RLC. Orthogonalcontrast analysis separated (P 0.05) the aggressivecolonizers from a second group of AMF that rangedfrom 1 to 19%RLC after 42 days (Figure 2). Thisgroup, identified as non-aggressive colonizers, in-cluded Acaulospora spp. WUM 11(1), WUM 46,WUM 49 and Glomus sp. WUM 44. The%RLC ofonly one isolate in this group, A. laevis WUM 11(1),increased with time.
High soil P supply reduced%RLC, but did notprevent colonization by any of the fungal isolates. Al-though there was a highly significant (P 0.001) Pinteraction with individual AMF (Table 2), groups ofaggressive and non-aggressive colonizers were sim-ilar t...