To finish out the course, we selected two papers on the topic of diversity metrics. First was the 2005 paper by Chao et al., as well as the recent 2017 paper by Hillebrand et al. Both papers proposed improvements to more traditional diversity metrics like the Jaccard index through more comprehensive inclusion of either unseen species or species composition. I think strikingly, these two papers address very different audiences. Chao et al. published in Ecology Letters target ecologists who may adopt this new metric in their own scientific research. Hillebrand et al. rather utilize their new species diversity metric as a way to inform land and biological monitoring program managers on the nuances of stable local richness levels.
I think the adoption of the Chao’s statistical approach to beta diversity has largely been accepted (with >1000 citations). Yet by reading the paper, we began to understand the importance of choosing metrics that best test your questions. Especially with measures as contentious as beta diversity, being clear on the benefits and limitations of certain measures can lead to better selection of complementary metric choices (similar to model stacking). The primary goal of Chao et al. was to draw attention that almost all richness measures will be undersampled, comparisons between sites are rarely equal in sampling sizes, and species occurrence is uneven. The authors compare their revised metric (that incorporates both composition and abundance) to existing frameworks on empirical data and simulations of uneven and incomplete sampling. I think parasite systems fit nicely into datasets that the authors suggest for use of their updated metric.
In many ways the Hillebrand paper seemed to reference key themes we had discussed with island biogeography, though now comparable or connected sites are the islands with various immigration and emmigration rates. While their adoption of species composition as a main reason for mismatches between global and local biodiversity trends, I did not find their results particularly compelling. Also, they fail to fully reach into how the species composition of local sites with stable species richness values may influence the function of these communities. I think by using available datasets that certainly have trait data such as biomass this would have been a simple calculation that could have really hammered home their point. I do think this paper forced me to gain a deeper understanding of immigration credit and extinction debt, and perhaps how quasi-equilibriums from island biogeography may be misleading as to the equilibrium of certain sites. I did appreciate the conceptual diagram given in Figure 1 that shows how both richness and evenness may change in a system, and then the use of their metrics on already established datasets. I’m not sure how much the Dutch phytoplankton added to the analysis, and would have perhaps wanted a more diverse dataset to complement the Iowa phytoplankton or grasslands example.
I think these papers may be best referenced when researching a specific problem with a dataset we are using or when designing a diversity study. I appreciate the recognition of imperfect data but using statistics to overcome imperfect data in logical and repeatable ways. Whenever explaining species richness and its importance, walking through the sampling and analysis approach iteratively, as well as showing alternatives through validated datasets or simulations, strengthens the interpretations and importance of richness measures.
Chao A, Chazdon RL, Colwell RK, Shen TJ. A new statistical approach for assessing similarity of species composition with incidence and abundance data. Ecology letters. 2005 Feb;8(2):148-59.
Hillebrand H, Blasius B, Borer ET, Chase JM, Downing JA, Eriksson BK, Filstrup CT, Harpole WS, Hodapp D, Larsen S, Lewandowska AM. Biodiversity change is uncoupled from species richness trends: Consequences for conservation and monitoring. Journal of Applied Ecology. 2018 Jan;55(1):169-84