Ecologists have long been captivated by the influence of climatic drivers (e.g., temperature and rainfall regimes) upon the global distribution, abundance, and diversity of ecosystems (Holdridge 1967, Whittaker 1970, Woodward 1987, Davis and Shaw 2001). Within the context of climate change, studies of the ecological effects of climatic drivers are especially important, because they can help scientists and environmental managers better anticipate and prepare for the ecological consequences of climate change. In this study, we examined the influence of climatic drivers upon the distribution, abundance, and species richness of mangrove forests. Mangrove forests are tidal saline wetland ecosystems located along sheltered tropical and subtropical coasts (Tomlinson 1986, Saenger 2002, Alongi 2009, Spalding et al. 2010, Twilley and Day 2012). Mangrove forests support ecosystem goods and services that have been valued at up to $194,000 per hectare per year (Costanza et al. 2014). In addition to providing fish and wildlife habitat, mangrove forests protect coastlines, support coastal fisheries, store carbon, provide timber, improve water quality, and provide recreational opportunities (Ewel et al. 1998, Barbier et al. 2011, Lee et al. 2014). Despite the tremendous ecological and societal value of mangrove forests, the influence of climatic drivers upon mangrove forest distribution, structure, and function has not been well quantified in many parts of the world. The mangrove literature contains many valuable observations and well-articulated hypotheses regarding the influence of temperature and rainfall regimes upon on the distribution, abundance, and diversity of mangrove forests (e.g., Davis 1940, Lugo and Patterson-Zucca 1977, West 1977, Woodroffe and Grindrod 1991, Duke et al. 1998, Saenger 2002, Ross et al. 2009, Saintilan et al. 2014). Unfortunately, a lack of relevant and easily accessible climate and/or ecological data has meant that many of these relationships have not been fully tested or quantified. Data constraints have sometimes resulted in the use of the best-available surrogate variables. For example, latitude and sea surface temperatures have often been used as proxies for winter air temperatures (e.g., Duke 1992, Twilley et al. 1992, Saenger and Snedaker 1993, Ellison 2002, Alongi 2009, Twilley and Day 2012), where sea surface temperatures have been used as proxies for precipitation (e.g., Duke 1992, Duke et al. 1998), and mean monthly air temperatures have been used a proxies for extreme minimum daily air temperatures (e.g., Quisthoudt et al. 2012, Record et al. 2013, Hutchison et al. 2014, Jardine and Siikamäki 2014, Rovai et al. 2016). Although these surrogate variables are helpful for showing that general relationships are present, the use of proxies can potentially be misleading without adequate discussion and characterization of the relevant physiological mechanisms responsible. Using proxies also makes it difficult to identify ecologically-relevant climatic thresholds, which are needed to anticipate and prepare for future change. In recent years, the quality and availability of global-scale climate and mangrove data has been improving (Polidoro et al. 2010, Spalding et al. 2010, Giri et al. 2011, Osland et al. 2013, Cavanaugh et al. 2014, Armitage et al. 2015). As a result, there is potential to use more meaningful climatic and ecological data to more directly quantify the influence of climatic controls upon mangrove forests. In this study, we used recent climate and mangrove ecological data to investigate the following questions: (1) at the global and distributional range limit scale, how do temperature and rainfall regimes influence the distribution, abundance, and species richness of mangrove forests; and (2) how and where is climate change, in the form of warmer temperatures and altered precipitation regimes, expected to affect the distribution, abundance, and diversity of mangrove forests? We expected that global and region-scale climate-mangrove linkages would be well characterized by nonlinear sigmoidal equations with abrupt ecological thresholds across climatic gradients (e.g., Osland et al. 2016). We also expected that, due to climatic variation and the range limit-specific role of extreme events (Stuart et al. 2007), climatic thresholds would be range limit-specific (i.e., lower in some range limits and higher in others). We presumed that climatic thresholds would be lower for mangrove presence than for abundance and/or species richness. We provide a generalized illustration of the hypothesized relationships between climatic drivers and mangrove forest presence, abundance, and species richness (Fig. 1).
