With increased human populations and subsequent pressure to develop or farm land, the rate of fragmentation of tropical rainforests has accelerated in the past several decades. How native organisms respond to such fragmentation has been the subject of intense study in temperate ecosystems and at several tropical sites in Central and South America, but there has been little study of this phenomenon in Mexico, the country bridging the Neotropics and temperate North America. A reason for this neglect is an apparent lack of long-term data; however, such data can be obtained from “non-traditional” sources, such as birders and tour leaders. We make innovative use of such data, combining them with more traditional data (e.g., museum specimens) to create a record of occurrence for Palenque, Mexico, from 1900 to 2009, including a near-continuous presence–absence record since 1970. We analyzed these data using logistic regression and, importantly, recent statistical advances expressly for sighting records. As recently as the 1960s Palenque’s forest was contiguous with that of Selva Lacandona to the east, but the protected area surrounding the famous ruins is now a forested island. As a result, various species formerly known from the site have disappeared, including species both large (Crax rubra, Penelope purpurascens, and Ara macao) and small (Notharchus hyperrhynchos, Malacoptila panamensis, Microrhopias quixensis, and Pachyramphus cinnamomeus). By contrast, several species of open areas or second growth have apparently colonized the area (e.g., Thryothorus modestus, Mimus gilvus, Euphonia affinis). Some species turnover has occurred within particular families, such as Columbidae, Trochilidae, and Troglodytidae. Losses and declines we documented at Palenque correspond with those reported from other sites in Mesoamerica, suggesting the soundness of our approach and the general vulnerability of certain species. Compilation and analysis of sighting record data holds great promise for tracking trends in many regions and across many taxa for which long-term census data are lacking.
Read full abstract