Botanic Gardens Trust, Sydney, Australia

Podocarpus elatus - rainforest conifer

The impact of Quaternary climate change on the distribution of a late-successional rainforest conifer

Rohan Mellick - PhD student
Maurizio Rossetto - Principal Research Scientist, Manager Evolutionary Ecology

Podocarpus elatus (R.Br. ex Endl.), is a late successional, mature-phase tree that is associated with drier rainforest types. A previously published microsatellite-based study reported regional genetic differentiation on either side of the Clarence River Corridor (CRC), and suggested that the northern populations of P. elatus were of lower diversity. In the Australian Wet Tropics (AWT) it has been shown that rainforest gymnosperms have expanded during glacial maxima. Unfortunately we do not know the full effect of reduced angiosperm competition during these periods, or if P. elatus is contributing at all to the AWT fossil record, due to a number of co-occurring Podocarpus species and fossil pollen only being classified to the generic level. The use of Ecological Niche Modelling (ENM) may supplement the deficiencies of the fossil record and allow for inference about these cyclic changes in distribution to be made. Combining molecular and ENM techniques can help identify refugal areas and date climate driven expansion and contraction events.

Phylogeographic analysis confirmed previous findings that two genetically divergent regions reside within the range of the species (distributed north and south of the CRC, and that the central and southern ranges harbour the majority of genetic diversity. The northern distributional region persisted through the Last Glacial Maximum (21 Ka; LGM) in a restricted central northern refugial area, and later expanded to its full range during the Holocene Climatic Optimum (6 Ka). This broad expansion from a narrow genetic pool scenario identified by niche modelling is supported by the low genetic variation found in the northern range. For the southern region, historical niche modelling showed a gradual contraction in palaeo-distribution since the LGM from a very large distribution to the current potential range.
However single species are expected to respond to climate change in a relatively uniform way across their range. Variation in regional climatic effects may explain the current genetic structure and the observed difference in range shift response to post-glacial climatic warming. Molecular and ecological niche modelling results are congruent and show a non-uniform response of rainforest vegetation to Quaternary climate change, and further undermine the general assumption that broadly distributed species respond in a uniform way to climate change across their range.

We have complex evolutionary processes occurring in these rainforests, the majority of which are surrounded by an urban matrix with no means of altitudinal or latitudinal range shifts in response to anthropogenic-induced climate change. Understanding the manner of intraspecific divergence within these ‘captive’ rainforests is integral to conservation. For example, is the regional differentiation observed in P. elatus a result of balance between genetic drift and geneflow, or is ancestral polymorphism being maintained in the absence of present day geneflow?  In combination with future modelling onto IPCC climatic estimates of 2100, we will make suggestions on how to best maintain these natural processes to conserve these threatened communities in relation to future climate change. This study notes the need for conservation of the natural range shift processes in these rainforests, and suggests that the design and direction of habitat corridors could take on a broader evolutionary application.





Podocarpus elatus
Podocarpus elatus. Photo: Rohan Mellick

ENMs for P. elatus
ENMs for P. elatus north (a, b and c) and south (d, e and f) of the CRC biogeographic barrier in eastern Australia based on the MIROC global climatic model for 21 Ka Last Glacial Maximum (a and d), 6 Ka Holocene Climatic Optimum (b and e) and the 0 Ka Pre-industrial (c and f) time periods. Dark blue indicates low probability of occurrence and warmer colours indicate higher probability of occurrence.