This experiment was conducted in two phases. The first phase assessed the influence of chemical fertilizers on the species composition of VAM fungal communities in long-term field plots at the Cedar Creek Natural History Area (CCNHA) in central Minnesota. The second phase measured the effects of VAM fungal assemblages from the fertilized and unfertilized field plots on Big bluestem grass (Andropogon gerardi Vitm.) grown in a greenhouse.

Phase 1

On May 6, 1990, composite soil samples were collected from twelve 4m X 4m experimental plots in a field abandoned from agriculture for 22 years. Six of these plots had been fertilized annually for 8 years with: 50g/m2 NH₄NO₃; 20 g/m2 P₂O₅ ; 20 g/m2 K₂O; 40 g/m2 CaCO₃; 30 g/m2 MgSO₄; 18ug/m2 CuSO₄; 37.7ug/m2 ZnSO₄; 15.3ug/m2 CoCl₂; 322ug/m2 MnCl₂; and 15.1ug/m2 NaMoO₄. The remaining six plots had been left unfertilized. The fertilized (FERT) soils contained an average of 8.6 PPM NO₂-NO₃ and 62.2 PPM P (Bray-1), while the unfertilized (UNFERT) soils contained an average of 1.6 PPM NO₂-NO₃ and 26.5 PPM P (Bray-l). Each soil sample was a composite created by combining nine 25cm deep X 5cm diameter cores taken at 90cm intervals along two transects diagonally crossing each pot. Samples were mixed gently by hand and stored for 6 weeks in plastic bags at ca. 5 degrees C.

Species of VAM fungi are identified by the morphology of their soil-borne spores. To assess the species composition of VAM fungal communities, spores were extracted from 25g subsamples of each of the 12 inoculum soils and spread evenly on a membrane filter using the technique of McKenney and Lindsey (1987). All spores within the central 27% of the filter paper were removed using a fine forceps and a dissecting microscope (50X) and mounted on permanent

slides. These slides were examined with a compound microscope (100 to 1000 X) and spores were counted and identified to species based on wall structure (Schenck and Perez, 1990). The relative density (%) of each species in each sample was calculated as: ni/Nt, where ni = the number of spores from the ith species and Nt = the total number of spores examined from the sample.

It is likely that many of the infective VAM propagules in the inoculum soils collected in May were in the form of hyphae or VAM root fragments and would not be accounted for in spore analysis. To more effectively assess the species composition of the VAM fungal community, cultures of the 12 inoculum soils were established in an effort to induce all VAM propagules (spores, hyphae and VAM roots) to infect a host plant and subsequently sporulate. These cultures were established in sterilized 23cm diameter clay pots by mixing 100g of inoculum soil with a 1:1 mixture of steamed CCNHA sand + river washed sand. Each pot was sown with surface sterilized (70% EtOH for 3 minutes) seeds of Sorghum sudanense (Piper) Stapf. and kept in a greenhouse (22 to 27 degrees C) for 4 months. Plants were watered on alternate days with a nutrient solution which was modified from Sylvia and Hubbell's (1986) formulation for aeroponic culture of mycorrhizae, and contained (in mg/l): 433 KNO3; 8.4 Ca(NO3)2-4H2O; 199 CaSO4-2H2O; 130 K2SO4; 72 MgSO4; 0.86 H3BO3; 0.54 MnCl-4H2O; 0.07 ZnSO4-7H2O; 0.02 CuSO4-5H2O; 0.03 NaCl. Watering was stopped and tops of the sorghum plants were cut to a height of ca. 10cm one week before terminating the cultures. Spores were extracted from duplicate 25g subsamples of each culture, counted and identified as described above.

Phase 2

Big bluestem seedlings were inoculated with three types of soil: fertilized (FERT), unfertilized (UNFERT), and a steam sterilized control (NONVAM), and were grown with and without supplemental P and with and without supplemental N. This 3 X 2 X 2 factorial experiment had 10 replicates. Five randomly selected replicates were harvested at 1 month and the remaining 5 replicates were harvested at 3 months. Big bluestem was chosen as the indicator plant in the bioassay because it is native to CCNHA and is well adapted to soils in the area. Thus, if coadaptated mycorrhiza-soil complexes exist at CCNHA, then big bluestem should be coadapted with the VAM fungi present in the unfertilized soils.

Subsurface sand from CCNHA (containing 0.7 PPM N02-N and 32 PPM P (Bray-l) was mixed (1:1) with river washed sand (#19 granusil, Unimin Corp. LeSuer, MN) and steam sterilized (ca. 85 degrees C) for one hour on two consecutive days. Square pots (11.5 cm on a side) were filled with 980g of this sand mixture and leached with deionized water. A 6cm deep X 3cm wide hole was pushed into the center of each pot and filled with 30 g of "inoculum-soil."

Thirteen different inoculum soils were used: 6 FERT, 6 UNFERT, and one NONVAM composed of a steamed composite of soils from all 12 plots. To equalize the fertility of FERT and UNFERT inocula, 15g of "live" FERT soil was combined with 15g of steamed composite of

UNFERT soil, and 15g of "live" UNFERT soil was combined with 15g of a steamed composite of FERT soil. Communities of soil microbes (<25um in size) were equalized across treatments by applying 15ml of a "microbial-wash" to each pot. To prepare this microbial-wash 800ml of a composite of all 12 soils was blended with 4000ml of deionized water and filtered thrice through a 25um mesh screen.

Big bluestem seeds were surface sterilized in 5% sodium hypochlorite for 10 minutes and sown in flats of steamed sand. Five days following germination, uniform seedlings were

transplanted into the inoculum-soil plugs in the center of each pot. Plants were randomly arranged in a greenhouse and maintained at 27 degrees C/22 degrees C (day/night) without supplemental lighting from 16 June through 12 September, 1990. On alternate days every plant was watered with Sylvia and Hubbell's (1986) modified nutrient solution (without N or P) which was described above. Enough nutrient solution (50ml to 150ml, depending on plant size and weather) was applied to each pot so that excess solution drained from the bottom. The +N treatments had 433mg/l KNO3, and the +P treatments had 44mg/l KH2PO4 added to the background nutrient solution. The pH of all solutions was brought to 6.5 using .01 N NaOH.

Five randomly chosen plants from each treatment were harvested after 4 weeks and the remaining plants were harvested when inflorescences were fully mature and the plants had begun to senesce after 12.5 weeks. Plant heights and shoot and root weights (oven dry) were measured at both harvests and the number of inflorescences per plant were counted at the second harvest. At the one month harvest, a subsample of roots (< 0.25g) from each plant was stained with trypan blue in lactoglycerin using the technique of Phillips and Hayman (1970), and specific root length

of root length colonized with VAM magnified line-intercept method of (m/g root) and the percentage fungi were assessed using the Ambler and Young (1977).

Statistical analysis

Two sample t tests were used to compare the relative spore densities of individual VAM fungal species in FERT and UNFERT field soils and culture soils. Plant and fungal responses to the experimental treatments were assessed with two-way ANOVA with the 12 inoculum sources nested into two (FERT and UNFERT) inoculum treatments. NONVAM plants were not included in the ANOVAs but within each of the four nutrient treatments, mean responses of the controls were compared with mean responses of plants inoculated with FERT and UNFERT soils using least significant difference multiple comparisons at a 95% confidence level. Prior to statistical analysis, shoot and root weights were log transformed, number of inflorescences per plant were square-root transformed, and percent VAM root length and relative spore densities were arcsine square-root transformed. Analyses were performed with SAS (SAS, 1988) and Statgraphics (STSC, 1986).