The latitudinal gradient in phylogenetic diversity of seed plants in Asia during the Holocene

Tropical Niche Conservativism (TNC) postulates decreasing phylogenetic age and diversity with higher latitudes and lower temperatures due to environmental filtering. Phylogenetic diversity patterns along extensive elevation gradients do not corroborate with TNC. Alternatively, the Out of the Tropics Hypothesis (OTT) posits that some clades have expanded to the extratropical region and diversified slowly compared to the tropics, hence the average phylogenetic age and diversity will increase with latitude due to evolutionary convergence.  We use palaeoecological data of seed-plant taxa spanning the Holocene to test predictions deduced from TNC and OTT.

Regression analyses revealed the variation in phylogenetic relatedness along the latitudinal gradient from 25 to 65 degrees N (Central Asia). This was expressed as the Net Relatedness Index (NRI) and Nearest Taxon Index (NTI) for each taxon assemblage of the fossil pollen data sets. Phylogenetic relatedness was also regressed against climate variables.

The positive (TNC) and negative (OTT) predicted trends in phylogenetic relatedness were rejected regardless of the metric used. The Holocene-wide latitudinal pattern of phylogenetic relatedness is non-linear with a bimodal hump around 45 degrees north. The phylogenetic clustering increases again from 55 to 65 N, but this pattern is not detected for late glacial (12kBP) and 8-10K BP. In general, are the response of NRI and NTI along the latitudinal gradient quite consistent concerning the time interval analyzed. The increasing trend in mean phylogenetic age with increasing latitude is the only pattern that verifies one of the predictions (OTT).

Falsification of the prediction deduced from both TNC and OTT indicates that these theories may represent an oversimplification. These patterns show a good correlation with elected climatic variables, and we hypothesize that drivers of evolutionary divergence and environmental filtering may be better explained by climate variables.