Integration of physiological knowledge into hybrid species distribution modelling to improve forecast of distributional shifts of tropical corals

Abstract

Aim: Predicting species distributional shifts in future climate scenarios representing conditions that do not exist in the current world is a challenge. Species distribution models may result in misrepresented projections for species living in extreme conditions if based on truncated response functions. Model extrapolation may not detect declines that could occur if future environment conditions exceeded the physiological tolerance of the species. We developed a novel method aimed to overcome this constrain by incorporating the physiological response function of a tropical hydrocoral to temperature as a predictor variable in a Hybrid SDM. Location: Atlantic Ocean. Methods:We conducted ecophysiological experiments simulating heat and cold stress to determine the maximum photochemical efficiency of the hydrocoral's symbiont along a thermal gradient to identify sublethal and lethal conditions. The response curve obtained was then applied to a temperature raster to create a new physio-climatic variable, which was integrated into the Hybrid SDM as a predictor.Simple Physiological and Correlative SDMs were compared with the Hybrid model. Results:The Hybrid SDM outperformed the Correlative SDM allowing predictions without extrapolations in the physio-climatic predictor. It suggested habitat contractions in tropical regions with forecasted temperatures above the coral's physiological tolerance, which were underrepresented by the Correlative SDM. It also incorporated habitat suitability restrictions by other predictors of unknown physiological response by incorporating correlative information (as limitations in river mouths by low salinity). In this way, by integrating mechanistic and correlative knowledge, the Hybrid SDM also predicted a potential expansion to higher latitudes, which agreed with the recent evidence of its expansion into the subtropical Canary Islands. Main conclusion:Integrating physiological knowledge into Hybrid SDMs by adding a physio-climatic predictor improves model transferability resulting in predictions of decline in future climates, which may be misrepresented by SDMs trained at presentday conditions, and therefore are advisable for early warning in conservation management.

Variability in the fundamental versus realized niches of North American mangroves

“…To assess these shortcomings, we need more manipulative experiments characterizing species' fundamental niches (Guisan & Thuiller, 2005;Kearney & Porter, 2009;Peterson & Soberón, 2012). Recent studies have found that integrating mechanistic and correlative models is a key step in predicting species distributions and range dynamics, especially in the complex context of a rapidly changing climate Kotta et al, 2019;Rodríguez et al, 2019).…”

Environmental factors driving the distribution of the tropical coral Pavona varians: Predictions under a climate change scenario

“…Rather, this suggests that the habitat will not be adequate according to the values used within the training data. Obviously, the incorporation of other sources of information, e.g., physiological and genetic adaptation of corals, dispersal and species interactions, or the effects of local anthropogenic disturbances, would improve the reliability of SDMs projections (Diniz‐filho et al., 2019; Donner, 2009; Gardner & Gaston, 2019; Rodriguez, Garcia, et al., 2019; Singer, Schweiger, Kühn, & Johst, 2018). Likewise, corals' symbionts can differ among species and play an important role in corals adaptation (Baker, Starger, McClanahan, & Glynn, 2004; Rodríguez et al., 2019).…”

“…The identification of areas where corals can be most affected by these environmental changes is key for adequate conservation plans. Within this context, only few studies have aimed to predict changes, under climate change scenarios, in the distribution of corals (Chefaoui, Casado‐Amezúa, & Templado, 2017; Guinan, Brown, Dolan, & Grehan, 2009; Tittensor et al., 2009; Woodby, Carlile, & Hulbert, 2009), including tropical corals (Couce, Ridgwell, & Hendy, 2013; Franklin, Jokiel, & Donahue, 2013; Freeman, Kleypas, & Miller, 2013; Rodriguez, Garcia, Carreño, & Martínez, 2019; Rodriguez, Martínez, & Tuya, 2019).…”

“…Predictive modelling is a useful tool for supporting conservation decision making (Guisan et al., 2013; Liu, White, Newell, & Griffioen, 2013). Modelling is especially relevant for species inhabiting habitats difficult to sample, as subtidal marine species, e.g., kelp forests (Franco et al., 2017; Martínez, Viejo, Carreño, & Aranda, 2012; Méléder, Populus, Guillaumont, Perrot, & Mouquet, 2010) and corals (Couce et al., 2013; Franklin et al., 2013; Freeman et al., 2013; Riul et al., 2013; Rodriguez, Garcia, et al., 2019). One of the most common methods to relate the distribution of species with climate and other physical factors is the Maximum Entropy Model (MaxEnt) (Phillips, Anderson, & Schapire, 2006).…”

Climate change impacts on Australia's eucalypt and coral species: Comparing and sharing knowledge across disciplines

“…For eucalypts, these models can include the direct impacts of increasing carbon dioxide levels on growth (Battaglia, Bruce, Brack, & Baker, 2009), which cannot easily be simulated using SDM methods. For instance, Rodriguez, Garcia, Carreno, and Martinez (2019) have used a hybrid model including physiological knowledge to improve predictions of distributional shifts of a coral species under climate change made using the Maxent SDM. The hybrid model predicted potential expansion to higher latitudes, which agreed with some recent observations from the Canary Islands (Clemente et al, 2010; Box).…”