Methods
VAM fungal populations were studied in 14 field sites and 3 forest sites at Cedar Creek Natural History Area, about 50 km north of Minneapolis, Minnesota (Fig. 1). The field sites
included a continuously cropped rye field, a fallow field and 12 abandoned fields (ranging in age from I to 60 years). The rye field, fallow field and the 6 youngest abandoned fields formed a
series of adjacent plots (0.2 ha) that were sequentially abandoned over the past 13 years (Delaney 1988), while the remaining 6 abandoned fields were interspersed throughout the 2200 ha natural history area (Fig. 1). The forest sites included an oak savanna, an upland pin oak (Quercus ellipsoidalis Hill) forest and a northern hardwood forest. Soils of the 17 sites are all well drained upland sands which occur on the same landform and have similar patterns of soil development. Although the sites do differ in soil series (Nymore, Sartell. or Zimmerman), all of them, with the exception of the 60 year old field, belong to the same taxonomic group. Soil in the 60 year old field was an Alfic Udipsamment, whereas the soil of the remaining fields and forests was a Typic Udipsamment (Grigal et al. 1974). Characteristics of these sites are summarized in Table 1. Soil series classifications in Inouye et al. (1987) differ slightly from those of this study because classifications in the former study were based on general soil maps, while classifications in the present study were based on field examination.
Transects were established in each of the 17 sites. The center point of each transect was randomly established and then the transect was oriented in a north-south direction. Three sample points were located at 5m intervals along each transect. In November, 1987, six cores (2.5cm diameter x 15cm deep) were collected from each sample point, three cores were composited for analysis of spore populations, soil P and pH, and the remaining three cores were used for a bioassay and were placed directly into plastic growth tubes called Conetainers (Stuewe and Sons, Inc., Corvallis, Oregon USA). A sampling depth of 15cm was chosen because VAM propagule densities are generally greatest in the surface 15cm. In a study of the vertical distribution of VAM fungal spores, An et al. 1990 found 87% of the spores occurred in the top 15cm and just 13% occurred at greater depths. In June, 1988, composite soil samples were collected for a second bioassay. For this bioassay, a total of 15 cores (6cm diameter x 10cm deep), five cores from each of the three sample points, were collected from each transect and composited.
Spores were extracted, counted, and identified from composite soil samples collected from each of the three sample points along 15 of the 17 transects (spore populations in the 3 and 8 year old fields were not analyzed). Spores were extracted from 25ml aliquots of soil by wet-sieving followed by sucrose centrifugation (McGraw and Hendrix 1984). Spores were placed in a gridded petri dish and counted using a dissecting microscope (40 x ). Permanent slides of randomly selected sub-samples (10 to 20%) of these spores were made and examined at 400 x to 1000 x . Spores were identified based on wall structure (Walker 1983; Schenck and Perez 1990) and comparison with holotypes, paratypes and collections obtained from the Oregon State University herbarium. Voucher numbers were assigned to representative specimens of each of the species we identified. These voucher specimens can be obtained upon request. Spore counts from the three sample locations in each transect were added together (so spore populations were analyzed from a total of 75ml of soil from each site). An aliquot of each soil sample was air dried, and bulk density was determined. Total spore counts were expressed as spores per gram dry soil. The relative abundance (%), of each species at each site, was calculated as: (ni/Nt ) x 100, where ni= number of spores from the"ith'' species and Nt=total number of spores examined from the site. Mycorrhizal fungal diversity (at each site) was calculated by the Shannon-Wiener index. This index combines two components of diversity: numbers of species and the evenness of allotment of individuals among the species (Krebs 1985). The percentage dissimilarity of the VAM fungal community in each site compared to the rye field was computed as in Olffand Bakker (1991): D = 100x(1--(2c/(a+b))); a = the sum of species abundances in the site; b = the sum of species abundances in the rye field; c=the sum of the minimum abundances of species common to both sites.
Mean species abundances were square root x 10 transformed prior to the computation of dissimilarity.
A modification of Moorrnan and Reeves' (1979) infectivity bioassay was used to quantify the relative soil densities of infective propagules of VAM fungi (including spores, mycelium, and infected root fragments). Soil samples were collected from the transects, placed in surface sterilized Conetainers, and planted with a single surface sterilized corn seed (hybrid A639 x A676, University of Minnesota Agronomy Department). It has been shown that early in the colonization process the level of root infection is linearly related to soil densities of VAM fungal propagules (Carling et al. 1979; Smith and Walker 1981). Thus, we assumed that the level of mycorrhizal infection of 30 day old corn roots was proportional to the density of VAM fungal propagules in the soil.
This bioassay was conducted twice, first with samples collected in November, 1987 and again with samples collected in June, 1988. All 17 sites were analyzed in the 1987 bioassay, but the three forested sites were not included in the 1988 bioassay. The two bioassays differed slightly in the sampling and cultural methods used. In the 1987 bioassay, nine intact soil cores (three cores from three sample points) from each transect were placed into small (50 ml) Conetainers, loosely covered with plastic, and stored at 11 degrees C. After 48 days of cold treatment, the Conetainers were sown with corn and maintained in a growth chamber: 16 h, 25 degree C "days"(ca. 340 microeinsteins m-2 s-1 I PAR), and 8 h, 15 degree C nights. Plants were watered daily with one-tenth strength Hoagland's solution (Hoagland and Snyder 1933) minus phosphorus. In the 1988 bioassay, composite soil samples from each transect were thoroughly mixed by hand for 3 min, subsamples were placed in 160 ml Conetainers, and were sown with corn within 24 h following collection. Plants were kept in a greenhouse (20-32 degrees C) and watered daily with deionized water.
Corn plants were harvested in the same manner for both harvests. After 30 days, shoots were cut from roots, oven dried at 70 degrees C and weighed. Root systems were washed, cut into 2.5cm segments, and 0.5g of randomly selected segments were stained with trypan blue in lactoglycerin (Phillips and Hayman 1970). Percent root length containing vesicles and/or arbuscules was assessed using a gridline intersect method (Giovannetti and Mosse 1980).
Composite soil samples collected in November, 1987 were analyzed for pH (1: I water slurry) and P. Bray-l P (NH4F + HCl extractable P) was measured following the procedure of Dahnke (1988), and total P was measured as in Tandon et al. (1988). Total soil N, organic C and H2O soluble organic C are summarized from Zak et al. (1990). However. they did not study the rye field, fallow field and 13 year old field. Total N was determined by digesting air dried soil in a block digester with concentrated H2SO4 and HgO as a catalyst. Organic C was determined by combustion in a LECO automatic C analyzer (LECO Corp. St. Joseph, MI USA) and H2O soluble organic C was measured following the procedure of Burford and Bremner (1975).
Root biomass was measured along transects ca. 5m from those used for VAM analysis (Gleeson and Tilman 1990). Between 13 and 30 July, 1987, three soil cores (4.8cm diameter x
30cm deep) were taken from five 100m transects in each of the field sites except the rye field, fallow field and 13 year old field. Cores were placed on a screen and soil was gently rinsed from the roots. Roots were sorted from litter and adhering debris, but no attempt was made to distinguish between living and dead roots. Roots were oven dried at 70 degrees C for 24 h and weighed (Gleeson and Tilman 1990).
Detailed studies of plant species composition have been conducted across the chronosequence (Delaney 1988; Inouye et al. 1987), and we used this information in the present
study. Plant species' cover data from these two studies were coded according to the putative mycorrhizal status of their families. Species from four families: Brassicaceae, Caryophyllaceae, Chenopodiaceae, and Polygonaceae, reported to contain a large proportion of nonmycorrhizal species (Newman and Reddell 1987), were coded as 'non-hosts". Non-host abundance, calculated as (relative cover of non-host plants/relative cover of host plants), was determined for 11 of the 12 old field sites (no plant community data were available for the 13 year old field).
Spearman rank correlation analysis was used to relate abundances of VAM fungal species with age, or successional rank, and soil or plant properties of the 15 field and forest sites in which spores were analyzed (spore populations were not analyzed in the 3 year and 8 year old fields). The oak savanna, upland pin oak forest and northern hardwood forest were assigned successional ranks of 13, 14, and 15, respectively. Simple and partial correlation analyses were used to examine the relationships between infectivity, total spore counts, field age, soil properties, below-ground biomass, and non-host abundance of the 14 field sites. Prior to statistical analysis, root infection data were arcsine square root transformed and soil C, soil P, soil N and root biomass were ln transformed. Significance of the correlations were accepted at alpha = 0.05. Correlation analyses were performed using Statgraphics (STSC, IN., USA 1986). Plant community data was coded and collated using SAS (SAS IN., USA 1988). Percentage dissimilarity was calculated using VEGROW (Fresco 1989).
| webmaster@cedarcreek.umn.edu | Last updated 6/3/3 11:07 a.m. (Tuesday) |