Within-species responses to abiotic and biotic factors
Within species, vital and population growth rate responses to abiotic heterogeneity in shade and soil separated into three distinct response profiles. One species, P. airoides , exhibited clear evidence of demographic compensation, whereby a convex emergence response to PC1 coupled with a negative seed production response resulted in a non-significant population growth rate response (Fig. 2). For a further five species (all but A. calendula and H. glutinosum ), we found similar compensatory trends along PC1 between emergence and seed production, but relationships were only significant for one of the vital rates (mainly emergence). Hence, most of the species (6/8) had one significant vital rate response (positive or convex) to PC1 yet a non-significant population growth response (Fig. 2). Hyalosperma glutinosum had a positive emergence response to PC1 but no other significant vital or population growth responses (Fig. 2). The final species, A. calendula , had no significant responses of any vital rate or population growth rate to PC1 (Fig. 2).
There were three response profiles for the influence of neighbour abundance (or presence for population growth) on vital and population growth rates within species. Most of the species (6/8) had non-significant vital and population growth rate responses (Fig. 3, Table S2). One species, P. debilis , produced fewer seeds as neighbour abundance increased and had lower population growth rates in the presence of neighbours (Fig. 3). Arctotheca calendula did not have significant survival or seed production responses to neighbour abundance yet had lower growth rates in the presence of neighbours (Fig. 3). For most species, within-species population growth responses to neighbours were in the same direction as seed production responses (for 6/8 species) and survival responses (for 5/8 species, Table S2). Two species had significantly lower population growth rates in the presence compared to absence of neighbours (Fig. 3, Table S2,p <0.05).
The LTRE analysis showed that the life stage transitions contributing most to the difference in population growth rate in the absence versus presence of neighbours differed for the two species assessed. Survival probability and seed production were the only rates allowed to vary between population growth rates as they were the only vital rates measured in response to neighbour removal. For A. calendula , the transition representing a dormant seed remaining viable, emerging, and surviving to reproduction made the largest contribution (52%, Figure S5). This contrasted to P. debilis , for which the transition representing an adult plant producing seeds that stays dormant and viable in the seed bank accounted for 73% of the difference in population growth between neighbour treatments.