Importance of species replication in understanding plant invasions into North American grasslands - Invasive Plants: Ecological and Agricultural Aspects - page 51

and Simoes and Baruch [32]. I found that out of seven studies, all but one
found support for the hypothesis that exotics are more tolerant of grazing or
clipping than native species (Fig. 2). Support of the hypothesis was found by
Caldwell et al. [8], Pyke [10], Fan and Harris [36], Simoes and Baruch [32],
Schierenbeck et al. [37], and Holmgren et al. [38]. Only one study [39] report-
ed inconsistent results, with exotics being more tolerant to grazing only in sit-
uations where water availability was high. Thus, the overall data set seemed to
support the hypothesis that exotics have higher grazing tolerance than natives.
The higher tolerance of exotics to grazing may be important in their spread in
grasslands, which could be due to the commonness of grazing by native and
domestic animals throughout grasslands of the world. The processes underly-
ing this phenomenon deserve much further research.
Figure 2. Proportion of studies cited by Daehler [24] that found higher tolerance to grazing or clip-
ping in exotics than natives (E > N), no significant difference (n.s.), or higher rates in natives than
exotics (N > E). Studies that either did or did not replicate species were combined due to small sam-
ple sizes.
An experimental example on the importance of replication
In an ongoing experiment in the Texas Blackland Prairie region [40], we are
comparing growth characteristics of the common exotic and native C 4 grasses
in the region. Plants are being compared in common garden monoculture plots,
as well as in 2, 4, and 8 species mixtures. Small equal-sized transplants were
planted into monoculture 1 m 2 plots (96 transplants per plot) in spring 2001
within three blocks (block term, F 1,15 = 2.4, P = 0.143) and allowed to grow for
two growing seasons before harvest. Plots were weeded when necessary. An
estimate of average difference between exotics and natives is possible because
species identity is replicated, with 3 exotic and 5 native species being repre-
sented (Fig. 3). A fixed effects model (that uses plots as the error term) is
appropriate here because we have all the major native and exotic species in this
system, and therefore, the choice of species would not change if we were to