In her seminal paper, Rabinowitz (1981) proposed that describing species along three axes of rarity would result in direct links between biological and/or ecological factors and

species distributions. The literature citing the rarity matrix is primarily conservation-oriented. Therefore, the dataset VS-4718 includes only species defined as “rare” on at least one axis. Thus, GDC-0994 we cannot use this dataset to answer general questions about rarity and how it is different than commonness. However, we can utilize this dataset to determine the value of categorizing the structure of rarity. The internal structure of the range is an important characteristic of species distributions (Brown et al. 1996), so we ask if this frequently used typology of rarity BX-795 mw leads to alternative conclusions regarding the causes and consequences of rarity. Much of the data available in this literature set are taxonomic and often include reproductive ecology (mating system, pollination syndrome and seed dispersal vector) as these characters

often distinguish closely related species from one another and can be determined without extensive field surveys. We therefore undertook an investigation of the association among reproductive ecology traits and species distribution patterns within the rarity matrix. Methods We performed a Web of Science search for journal articles on plants find more citing Rabinowitz (1981) on 12 February 2007 and updated this search on 5 June 2009. Of the 365 references retrieved, most cited the seven forms of rarity as a general concept without classifying species of interest into a rarity type. Only 101 species, referenced in 27 articles, were classified on at least

two axes of the three-axis rarity grid (Appendix 1). We utilized the rarity categorization reported by the authors of these articles (Fig. 1) and recorded reproductive ecology data from these primary articles (Table 1 and Appendix 1, bold type). Additional data on reproductive ecology were acquired by performing further species-specific literature searches (Appendix 1). Landscape and environmental gradient data were not included in these searches. We categorized the pollination syndrome and seed dispersal vector as either abiotic (not mediated by insects, birds, or mammals) or biotic (mediated by insects, birds, or mammals). We specified the seed dispersal agent if known (ant, bird/bat, wind, water, or ballistic/gravity) and categorized the mating system as selfing (includes clonal reproductive strategies as well as apomictic species), outcrossing (dioecious or self-incompatible species), or mixed (for example, outcrossed flowers and clonal reproduction). We did not categorize reproductive ecology characteristics except when they were available in the literature for the particular species in question.