LB 14-167 - The other side of Darwin’s abominable mystery: why do so many families of angiosperms have low species richness?

Background/Question/Methods:

Background. The majority of extant land plant species are angiosperms (90% of species), which emerged within the last 200 Mya. The phenomenal expansion of angiosperm species richness (SR) has prompted many investigations into the factors influencing their diversification rate (DR), but there remains significant gaps in our understanding including the reasons why some angiosperm families have low SR. Methods. We used five published phylogenies employing different molecular markers and fossil data to estimate the age of most of the 454 families of angiosperms as recognized by APG IV. Next, we estimated shift points in the rate of origination of the families using Bayesian methods, and partitioned each dataset into the number of intervals exhibiting an approximately uniform DR. Next, within each time interval we calculated the MLE and confidence intervals for the DR parameters (r, λ, μ, and ε) assuming a birth-death model. We used these confidence intervals to assign families as having low, expected or high SR for each calibrated phylogeny and designated families as having poor, predicted or rich SR using majority and strict consensus rules. Finally, we examined the association between plant family SR designation (poor, predicted or rich) with diverse characteristics of the families.

Results/Conclusions:

Results. Using majority consensus criterion, we find that most families (62.5%) had a SR that falls within the mean +/- 95% CI of that expected for the geological interval in which they originated, while 32.2%, had lower SR and 4.4 % higher SR than expected. Further, we find no correlation between family SR category and the crown age of the family, and families with poor SR did not show phylogenetic clustering and were not nested in families with high SR. Families with poor SR were less likely to be associated with a whole genome duplication event, were found in all realms and biomes, but were more likely to exhibit restricted geographic range.

Conclusions: We find that ~ 1/3 of angiosperm families have poor SR, and that neither family crown age nor phylogenetic position is a predictor of SR. Families with poor SR are widely distributed among orders of angiosperms and in all realms and most biomes, but tend to have restricted ranges. These results are discussed within the context of the literature describing “depauperons” (sensu Donoghue and Sanderson). Further investigation into the characteristics of these families may shed light on the factors contributing to low biodiversity in some angiosperm clades.