Field Experiment
Study Sites
The long-term experiment will be established at two contrasting study sites; a upland ridge top and a level bottomland. Historically, both sites were forested, but were cultivated following European settlement. The upland ridge is located at Touch of Nature Environmental Center, 16 km south of the main SIUC campus. Following purchase of the land by SIUC in 1949-54, the 3 ha area was used as a horse pasture until 1977. The field was plowed in Spring 1981 and mown annually until 1987 when it was abandoned. It is presently being invaded by shrubs, especially the exotics Autumn Olive (Elaeagnus umbellata) and multiflora rose (Rosa multiflora).
The site is on Hosmer silt loam (a fine-silty, mixed mesic Typic Fraguidalf) soil, a moderately well drained soil on 2 - 7 % slopes. The bottomland site is a 1.21 ha old-field <1 km from the main SIUC campus. It is on land owned by the SIUC College of Agriculture, Horticultural Research Center, and was abandoned from crops in 1987. The last crop was Sorghum bicolor var. sudax) in 1987, since that time the field has been mowed biennially in the spring and summer. The soil is a Bonnie Silt Loam, a fine-silty, mixed, acid, mesic Typic Fluvaquent (Herman et al. 1979). These soils are poorly drained on low areas of flood plains. Seasonal flooding is common and wetness-tolerant grasses and legumes are recommended for planting.
Long-term experiment
The success and practical simplicity of the mowing treatments used in the Nettlecombe Grassland Experiment (Crothers & Lucas 1982; 1991) prompts us to use it as a model for our planned long-term experiment. In addition, a fertilizer pplication treatment is planned to investigate the role of soil nutrients in old-field succession (Reed 1977; Carson & Barrett 1988; Van Auken et al. 1992). The general patterns of old-field succession are well known nationally (Inouye et al. 1987; Carson & Barrett 1988; Vankat & Carson 1991) and for southern Illinois in particular (Bazzaz 1968; Dunbar 1981). Dunbar (1981) conducted an experimental study into the effects of burning, plowing, discing, mowing, and fertilizer on vegetation, arthropods, and soil fertility in fescue pasture (Festuca arundinacea) and two old-fields. All treatments had an effect upon one or more aspects of old-field succession. The goal of the study was to maximize the presence and abundance of game species, especially cottontail rabbits and quail. The general conclusion was that a variety of treatments should be applied to adjacent sites to supply a variety of resources for these species.
A split-plot experimental design has been established at the Touch of Nature site. Seventy-two 0.01 ha plots have been established in eight blocks of 9 plots each. Within each block, three subblocks corresponding to annual mowing treatments of once, and twice during the season, and control (no mowing) has been established. Plots within the subplots have been randomly assigned to one of three fertilizer treatments; annual fertilizer, year one only fertilizer, and control (no fertilizer). All plots were disced and in early spring 1996 with treatments (mowing and fertilizer) coming on-line in late spring. The Autumn Olive infestation at the upland site was removed during initial site preparation by physically uprooting the small trees. The treatments are similar to those used in the Nettlecombe Grassland Experiment (Crothers & Lucas 1982; 1991), with the addition of the fertilizer treatment.
The two predictions of the Intermediate Disturbance Hypothesis regarding species richness response to disturbance will be addressed (see background). The contrast of mowing frequencies is a press experiment and will allow the effect of disturbance frequency to be addressed. The fertilizer treatments contrast press and pulse experiments through the annual fertilizer application versus the year one only fertilizer application. The pattern of species richness on the plots fertilized only in year one, and in those not mown following site preparation will address the prediction of the IDH that species richness reaches a maximum before declining with time since the last disturbance. Annual measurements of plant abundance will be determined by estimating late season (late summer - early fall) species canopy cover according to the Daubenmire scale (Daubenmire 1959) in ten 1m square quadrats randomly located within each plot. Quadrats will be at least 1m from the plot edge. After estimations of canopy cover, the plants will be clipped to soil level, sorted by species, dried to constant biomass at 80° C, and weighed.