1.0 Site Description and Methods
The Luquillo Mountains (LM) have a long history of research on their flora and fauna, with the earliest work published shortly after the US acquired Puerto Rico from Spain in 1898. The pace of research increased with establishment of the Luquillo National Forest by the US Forest Service in 1907 (Lugo et al. 2012), and publications increased in the 1920s with seminal papers on vegetative composition and succession (Gleason and Cook 1926). In the 1930-40s, the US Forest Service International Institute of Tropical Forestry (IITF) conducted silvicultural studies that served as the basis for managing tropical forest production, and led to the designation of the entire National Forest as the Luquillo Experimental Forest in 1956. An emphasis on ecosystem science at the site began in 1963 with initiation of the Rain Forest Project (Odum and Pigeon 1970). Long-term watershed-scale hydrology and biogeochemistry began in 1983 (McDowell et al. 1990; McDowell and Asbury 1994) and continued with U.S. National Science Foundation funding of the Luquillo Long-Term Ecological Research (LUQ-LTER) site in 1988. LUQ-LTER is a multi-faceted research program focused on long-term ecosystem response to disturbance and is jointly led by the University of Puerto Rico and IITF. The University of New Hampshire, the US Geological Survey (USGS) and IITF have led efforts to understand catchment-scale biogeochemical cycles (e.g. Scatena 1989, 1990; McDowell 1998; Schaeffer et al. 2000; Crook et al. 2007; Heartsill-Scalley et al. 2007; Murphy and Stallard 2012; McDowell et al. 2013; Clark et al. 2017; Wymore et al. 2017). This long history of catchment research has resulted in one of the best understood tropical forest systems in the world, including forest plots that span almost 100 years of measurement (Harris et al. 2012) and the longest continuous record of tropical stream chemistry on Earth (McDowell 2017b). These observations provide an important basis for understanding tropical montane forests in a global context (Lugo et al. 2012).