Áreas prioritárias para o cultivo de Gracilaria birdiae (Plastino & Olibeira) em cenários de mudanças climáticas

Abstract

The interaction between biotic, abiotic and movement-related factors determines the geographical distributions of species. In response to environmental changes, species can adapt to new conditions, change their distribution, or become extinct. Since the 1970s, global temperatures have been increasing and this warming is expected to continue until the end of the century. Climate change has multiple impacts, threatening biodiversity, and is expected to cause shifts in the distributions of many species towards the poles. Macroalgae are fundamental in providing food and habitat for various organisms. However, the geographic distribution of macroalgae is closely related to temperature, and they are therefore quite susceptible to climate change. Gracilaria birdiae (Plastino & Oliveira, 2002) is an endemic Rhodophyta of Brazil of high economic importance that occurs from the coast of Ceará to Espírito Santo. The objectives of this study were to determine the environmental variables relevant to the geographical distribution of G. birdiae, generate a niche model to predict its current geographical distribution and create projections of future distribution by testing the hypothesis of southward shifts, and build a spatial prioritization portfolio for cultivation of the species under future climate change scenarios. We collected occurrence data for the species in the literature and in geographical species occurrence databases and obtained variables relevant to its physiology in Bio-Oracle 2. First, we build models with all possible combinations between the variables (using 70% of the data for training and 30% for model evaluation). We then selected the best model (combination of variables) based on the AUC, TSS metrics, the importance of each variable, and the species response curves for each environmental variable. The final model was fitted with the variables maximum cloud cover, maximum salinity and maximum temperature. To predict the distribution of G. birdiaewe built an ensemble model based on those generated based on three algorithms (GLM, MAXENT and RF) and 100% of the occurrence data. Niche models were projected into geographic space to predict the distribution of the species at present and in the future. For future predictions, we used the same environmental variables as in the present with predicted climate conditions for the 2050s in RCP 2.6 (optimistic) and RCP 8.5 (pessimistic) scenarios. Finally, we used spatial prioritization analysis for G. birdiaecultivation indicating high priority areas for the activity. The niche model showed good predictive performance, the predicted distribution of G. birdiae matched the known distribution of the species, and the climate projections in both scenarios indicated an increase in habitat suitability in northeastern Brazil and a shift of the distribution southward. The spatial prioritization indicated as priorities several regions where the species is currently cultivated, confirming the feasibility of cultivation in these locations and in accordance with experiments that suggest the tolerance of G. birdiaeto variations in temperature and salinity, including in future scenarios of climate change.