Realized and potential climate niches:: a comparison of four Rhododendron tree species

Aim This study aims to evaluate the hypothesis that there is no difference between the realized and the ex-situ niches of four selected Rhododendron tree species. If the hypothesis is rejected, the aim is: (1) to evaluate whether the magnitude of discrepancy between the two types of niche is related to competition or external constraint, and (2) to identify which niche dimension is expanded (cold and/or warm limits) and to discuss it in relation to the assumption behind the biogeographical projections related to global warming. Location The four target species (Rhododendron arboreum Sm., R. campanulatum D. Don, R. barbatum Wall., and R. wallichi Sm.) are common evergreen broad-leaved trees in the central Himalayas. Their realized niches are based on data from the elevation-temperature gradient in Nepal (1000-5000 m a.s.l.). The ex-situ data are from botanical gardens and arboreta mainly located in the northern hemisphere (n=43). Method Binary data on these taxa were obtained from 707 geocoded herbarium specimens (elevation and UTM location) and from two elevation transects in Nepal (n =194 plots). Climate conditions from ex-situ locations for each taxon were compared with their realized climate ranges with respect to (i) mean annual temperature (MAT), (ii) mean minimum temperature of the coldest month (MINCM), (iii) mean maximum temperature of the warmest month (MAXWM), and (iv) moisture index (MI). Realized optima were estimated by Generalized Linear Models (GLM), and its non-parametric extension, Generalized Additive Models (GAM), were used to estimate the realized niches. Results All target species have ex-situ individuals outside the realized climate niche, but the number is much higher for maximum and minimum variables than for MAT. The most dominant species (in-situ ), R. arboreum, had very few individuals outside its realized range, indicating congruence between its ex-situ and realized niches. The other taxa had many individuals outside the warm end of their realized temperature ranges, but almost none beyond the cold end of their ranges. All target taxa occur in common gardens under warm temperate conditions, but only R. arboreum grows in the warm temperate zone in the Himalayas. This trend at the warm end of the gradient is interpreted as a result of biotic exclusion. Main conclusion The results demonstrate that an extreme cold temperature may represent an absolute boundary for tree species' survival, whereas warm temperatures do not. This is in agreement with the hypothesis that several tree species may survive global warming in-situ because of high temperature tolerance, but its effect on regeneration is uncertain. In lieu of this there may be a significant time lag between change in climate and transient tree species distribution. Thus the effect of global warming on tree species distribution may be very difficult to predict.