Biology Reference
In-Depth Information
known targets of deer herbivory. We simulated herbivory with four unrelated
species and asked three questions: (1) Does leaf removal influence plant repro-
duction? (2) If so, is there a threshold level of leaf removal that must be reached
before plant reproduction is influenced? (3) Does a combination of leaf removal
and meristem removal have a greater impact on a plant than random leaf removal?
We report on our findings for each of these questions.
Results
There was a significant influence of the exclosures on the abundance of three out
of the four species, Phlox, Rhus, and Lathyrus (Figure 1). Flower number strongly
correlated with leaf biomass for Phlox, Rosa, and Rhus (Figure 2). The number of
flowers produced by Phlox (P = 0.6), Rosa (P = 0.13), and Rhus (P = 0.3) was not
significantly affected by the leaf removal treatment when accounting for plant size.
We did not detect an effect in Lathyrus because the few flowers produced were
all aborted. Seed mass positively correlated with both flower number (Figure 3a)
and leaf biomass. The strong, collinear relationship between flower number and
leaf biomass (Figure 2) prevented us from using both variables in the seed mass
analyses. We chose to use flower number as a covariate since it has a more direct
impact on seed set and the number of seeds produced by an individual plant.
Figure 1.
Abundance (mean ± 1 SE) of Lathyrus, Rosa, Phlox, and Rhus inside and outside fenced enclosures.
The effect
of fencing was compared within each species using oneway ANOVAs (Lathyrus n = 4, Rosa n = 4,
Phlox n = 8, Rhus n = 12). Overall, the abundance of all species was significantly higher inside the fence (Type
III GLM F = 14.7, df= 7, P < 0.001, R
2
= 0.837; species F = 21.3, df = 3, P < 0.001; enclosure F = 24.2, df = 1.
P < 0.001; interaction F = 6.72, df = 3, p = 0.003).
