VEGETATION DYNAMICS OF THREE TALLGRASS

                                    PRAIRIE SITES


                                    Jon K. Piper and Mark C. Gernes
                     The Land Institute, 2440 East Water Well Road, Salina,
Kansas 67401



Abstract. As part of research to develop a sustainable agriculture that
incorporates many aspects of the North American Prairie, I) the seasonal
phytomass levels supportable by native prairie and 2) how plant taxa shift
in temporal importance within the community were examined. This report
summarizes the first two years of a study documenting community patterns
on three tallgrass prairie sites in Saline County, Kansas that differ in
soil
type and annual productivity. Average August phytomass at the three sites
ranged from 284 to 682 g/m' in 1986 and from 377 to 1077 g/m2 in 1987.
Diversity declined with standing crop biomass and trom spring to summer
at all sites. Although big bluestem (Andropogon gerardii Vitman) domi-
nated all sites, legumes represented as much as 26% of total phytomass
on the poorest site. Some composite species were ubiquitous, but never
constituted more than 5% of vegetation. Spring forbs and cool-season
grasses were of greatest importance in April and May. The results, in
concert with other ongoing research, have implications for the design of
perennial seed crop mixtures suited for the Great Plains.
Key Words. diversity, legumes, production, tallgrass prairie, soil type,
sustainable agriculture, Kansas

                       INTRODUCTION
  Presently, agriculture on the North American Great Plains is
characterized by extensive monocultures of annual grain crops
which are subsidized largely by petroleum, synthetic fertilizers,
and pesticides. The environmental and social consequences of such
large-scale industrialized farming include high levels of soil loss,
pesticide and fertilizer contamination of soil and groundwater,
complete dependence upon finite fossil fuel resources, loss of
cultural knowledge, and the depopulation of rural communities.
  During settlement native species were commonly replaced with
wheat [Triticum aestivum (L.) L.], soybean [Glycine max (L.)
Merr.], and sorghum (Sorghum spp. Moench) introduced from
other continents. Agroecosystems in North America were then
modified to accommodate the biological requirements of these new
crops.
  These agroecosystems differ ecologically from the climax grass-
lands they replaced in many ways. The most apparent difference
between the two is in degree of diversity in both space and time.
Within the prairie, species use different portions of the soil volume,
have different ecological roles, and the demand on both biotic and
abiotic resources, is spread out over the growing season (Weaver
and Fitzpatrick 1934, Parrish and Bazzaz 1976, Rabinowitz et al.
1981). Monocultures of annual grain crops, on the other hand, use
the soil volume less efficiently, and plants' demands on environ-
mental resources occur simultaneously. Secondly, the prairie dis-
plays tight nutrient cycles, as most nutrients are tied up in living
biomass and soil organic matter (Woodmansee 1979, Knapp and
Seastedt 1986). But most nutrients in monocultures are supplied
externally, and are rapidly removed from the system via harvest,
leaching, and erosion. Thirdly, the prairie has the sun as its primary
energy source, in contrast to most temperate zone agricultural
sYstems which rely heavily also upon human-applied fossil fuels.
Lastly, because the climax prairie represents a later, rather than
m0 earlier, successional stage, its biotic components are likely more
lltegrated (Odum 1969, Risser et al. 1981).
  The objectives of this research were to determine 1) relative
PhYtomass contributions by grasses, legumes, and composites; 2)
seasonal and site related variation in productivity, richness, di-
jorsity, and evenness; and 3) phenological differences among ma-
JOr taxonomic groups.



                          METHODS
Productivity and Species Composition
  Research was conducted on prairie sites, Wauhob, Corner, and
Hill, located in Saline County, Kansas (S5 T15S R2W, Hutchinson
Quadrangle), within the western edge of the tallgrass prairie region
[transition between Bluestem Prairie and Bluestem-Grama Prairie
(Kuchler 1974)]. The Wauhob site is on a west-facing 6- 12% grade
with thin Kipson shaly silt loam soil. The soil of the Corner site
is a Geary silt loam and the Hill site is on a deep Longford silt
loam. The Wauhob site lies approximately 640 m southwest of the
Corner site; the hilltop site is located 110 m north of the Corner.
The sites were grazed seasonally prior to this study. The Wauhob
site was burned in April 1984 and in March 1987. The Corner and
Hill sites were burned during April in 1982, 1984, and 1985 and
in March 1987. Growing season (March through August) precip-
itation at the Salina reporting station (approximately 7 km NW of
the field sites) was 54.9 cm in 1986 and 61.0 cm in 1987 (National
Oceanic and Atmospheric Administration 1986 and 1987).
  Aboveground live phytomass was sampled within 12 quadrats
(50 x 50 cm) at each site in April, May, June, and August, 1986;
and May, June, and August in 1987. The spring sampling dates
coincided with the time that ephemeral forbs flower and set seed
while most grasses were just emerging. August sampling corre-
sponded to the flowering period of the dominant tall grasses and
probably represents peak live phytomass on the prairie (Risser et
al. 1981). Vegetation within each sample frame was clipped and
separated by taxon, then dried at 60 C to constant mass and weighed
to the nearest 0.01 g. In 1986, vegetation was divided into grasses
(Gramineae), composites (Compositac), legumes (Leguminosae),
and the rest (e.g. Liliaceae, Cyperaceae, Labiatae, Umbelliferae,
etc.) was combined. In 1987, plants were separated by species in
the field prior to drying. Species determinations followed Great
Plains Flora Association (1986).
  Biomass was compared between years by Student's t test, and
among sites by ANOVA followed by Duncan's Multiple Range
Test. In all analyses, the 0.05 significance level was used. Means
in text and tables are given with + one standard error.
  From  the 1987 data, richness, diversity, and evenness were
calculated for each site. Richness is the cumulative number of
species sampled over the season. Alpha diversity [exp(H')], a
function of richness and evenness, was calculated using the Shan-
non-Weiner Index (H'):
                        H' =       pIn p,
Evenness, a measure of biomass equitability among species, is
expressed:
                  E  - exp(H')- I/(1/E pi2)  1
where pi is the relative biomass of each species i (Collins 1987).

Phenology
  To examine temporal differences in resource use among plant
families, the sampling area and surrounding prairie was sampled
biweekly from 5 April to 22 October 1987, noting during each
census whether a species had emerged, flowered, or was setting
seed. Plants were recorded as emerged on the first date that the
species was recognizable. Thus, unfamiliar taxa may have been
up for some time before their emergence was recorded.



PROCEEDINGS OF THE ELEVENTH NORTH AMERICAN PRAIRIE CONFERENCE 1989 9