Many taxa follow a latitudinal diversity gradient (LDG): a pattern of high species richness in the tropics and lower species richness outside of the tropics. One frequently neglected group in LDG studies are the fungi. Lichenized fungi represent a unique group for studying the latitudinal diversity gradient because (1) they are a group defined by lifestyle, rather than phylogenetic relationship, and (2) each lichen is itself a symboisis between a fungus and a mutualistic photosynthetic partner (alga or cyanobacterium). We used georeferenced herbarium records data to investigate whether lichenized fungi follow a latitudinal diversity gradient in the Americas. To do this, we used data from the Consortium of North American Lichen Herbaria (CNALH), the Global Biodiversity Information Facility (GBIF), and the Brazilian Institutos Nacionais de Ciencia e Technologia (INCT) Virtual Herbarium database. Data from all three sources were combined to provide the best possible coverage. From CNALH, which contains only lichen data, we downloaded all data from North, Central and South America. From GBIF and INCT, we searched specifically for classes, orders, families and genera containing lichenized fungi. We removed any records that were not georeferenced, in addition to any records that were not identified to species. We binned records into latitudinal bands of a single degree from 72S to 83N, and calculated species richness, number of observations, and land area (square km) per latitudinal band.
Results/Conclusions
Although latitude was a significant predictor of species richness of lichenized fungi (p<0.0001), in our preliminary analyses we did not find the expected LDG pattern of high species richness in the tropics and lower species richness outside of the tropics at the studied scale. The peak of species richness was in fact between 34N-45N in North America. We hypothesized that this peak may be influenced by sampling bias. In support of this hypothesis, number of observations was a significant predictor (p<<0.0001) of species richness, and accounted for 84% of the variation. Interestingly, land area was not a significant predictor of species richness per latitudinal band, but in a combined model, land area and number of observations became a significant interaction term (p<<0.0001), and together with latitude accounted for 88% of the variation. Until recently, there has been a strong collecting bias of lichens in the northern hemisphere. We discuss how to correct for these sampling biases, and we identify areas that should be the focus of future collecting endeavors.