Spatial pattern changes in aboveground plant biomass in a grazing pasture

Abstract

Using gamma distribution and spatial autocorrelation, it was demonstrated that plant biomass per unit area of a pasture grazed by cattle exhibited two kinds of spatial heterogeneity: small‐scale heterogeneity caused by grazing and large‐scale heterogeneity caused by topography, land aspect, etc. For each of the 10 measurement times from May to August, 100 quadrats 50 cm × 50 cm were arranged along a straight line 50 m long in a pasture, and the plants within the quadrats were harvested at the height of 3 cm above the ground surface to measure the dry weight. The data were aggregated into frequency distributions, and gamma distribution and the parameter values were estimated. This analysis showed that with the progression of grazing the amount of biomass decreased and the degree of spatial heterogeneity in biomass, measured per 0.25 m2, increased, and due to plant regrowth the trends were reversed. By rearranging the 100 biomass data in order of weight, it was suggested that plots with an extremely large biomass were not grazed by cattle and remained in the pasture. For the same data, variations of biomass along the straight line were divided into two parts based on the moving average: the spatial trend and the residuals which cannot be explained by the trend. In this analysis, 48–75% of the total spatial variation was explained by the trend along the straight line. Analysis using spatial autocorrelation for the actual biomass changes showed that the biomass changes within a range of about 10 m on the straight line gave a positive correlation, which indicates a topographical trend in biomass. Spatial autocorrelation for residuals suggested that the spatial changes in biomass along the straight line followed a wave‐like or checker‐board pattern. Small‐scale spatial heterogeneity in plant biomass may be caused by the uneven deposition of excreta by grazing animals, uneven use of the grassland by grazing animals, and uneven dispersal of plant seeds through faeces over the grassland. The possibility that such unevenness might accelerate energy flow in the grassland ecosystem and contribute to grassland sustainability is discussed.

The assumption of optimality in foraging models: A simulated experiment with dairy cows grazing grass pasture

“…11). This conflicts with the observations made by Shiyomi et al (1983b) on sheep, who found that heterogeneity increased as a result of grazing (as observed in the RR, RS and NS strategies in this study). This may indicate that the sheep in Shiyomi et al's (1983b) study used a strategy similar to NS, which produces good results at low stocking density (Fig.…”

“…The results given in Shiyomi et al (1983b) indicate that both the mean µ and the parameterp decrease with time over a day's grazing. This implies that the field becomes "patchier", or more heterogeneous, under grazing.…”

“…Based on observations of the spatial distribution of grass in grazed paddocks, Shiyomi et al (1983a) proposed using a gamma frequency distribution to describe the heterogeneity of grass biomass density. The density of grass in each patch x (g dm m-2 ) was therefore initialised for each simulation run by sampling from a gamma distribution…”

“…9 is that bite mass can be estimated directly from µ, providedp, and foraging strategy, are known (or assumed). For example, if it was assumed that real pastures have heterogeneityp = 3 (Shiyomi et al 1983a) and that cows forage using a strategy equivalent to NF, then average bite size can be estimated as b 1.24 kLµ, where 1.24 is the selection efficiency from Fig. 9.…”

“…11. Based on Shiyomi et al 's (1983b) observations, we would expect thep f /p ratio to be less than one. in all simulations, the random patch selection strategies RR and RS reduced p over 1 h of grazing.…”