Community structure across grazing regimes
Wet/dry, grazing intensity, and year were all able to explain a significant amount of the variation in overall vascular plant community structure in the PERMANOVA (P <= 0.002 for each; Table 2; Figure 2). Post-hoc analysis revealed that the effect of grazing intensity on community structure was significant between plots with light and pulse grazing (Bonferroni adjusted p = 0.003) and between plots with light and press grazing (Bonferroni adjusted p = 0.003), but not between plots with press and pulse grazing (Bonferroni adjusted p = 0.27; Figure 2b). The effect of year on community structure was only significant between plots in years 2004 and plots 2007 (Bonferroni adjusted p = 0.006; Figure 2c).
In addition to the variation in each of the above metrics of overall plant richness, there was also substantial variation among plots in the mean species richness of shrubs (mean ± SD (range): 3.55 ± 1.84 (0.11 – 9.00) species), herbs (11.05 ± 5.22 (1.56 – 22.56) species), and graminoids (6.15 ± 2.63 (1.44 – 12.56) species). The number of shrub species was independent of year (F3,128 = 1.87, P = 0.14) and the grazing:year interaction (F6,122 = 0.11, P = 0.99), but varied with grazing intensity (F2,128 = 30.27 P < 0.0001) and wet/dry (F1,128 = 44.63, P < 0.0001); lightly grazed plots had more shrub species than plots with pulse (P < 0.0001) and press grazing (P < 0.0001; Figure 3a), and dry plots (lsmean ± SE: 4.23 ± 0.33 species) had more shrub species than wet plots (2.77 ± 0.33 species). The number of graminoid species varied with year (F3,128 = 13.11 P < 0.0001), as plots from 2004 (lsmean ± SE: 7.84 ± 0.54 species) had more species of graminoids than plots from 2005 (6.00 ± 0.54 species; P = 0.0005), 2006 (5.74 ± 0.54 species; P = 0.0001) and 2007 (5.08 ± 0.55 species; P < 0.0001), but was independent of wet/dry (F1,128 = 0.05, P = 0.82) and the grazing:year interaction (F6,122 = 0.30, P = 0.93). Complementing the effect of grazing intensity on shrub species richness, the species richness of graminoids also varied with grazing intensity (F2,128 = 23.24, P < 0.0001) as lightly grazed plots had fewer graminoid species compared to plots with pulse (P < 0.0001) and press grazing (P < 0.0001; Figure 3b). In contrast, the species richness of herbs did not vary with wet/dry (F1,128 = 2.70, P = 0.10), year (F3,128= 2.30, P = 0.08), grazing (F2,128 = 0.15 P = 0.86; Figure 3c) or the grazing:year interaction (F6,122 = 0.07, P = 0.99).