Understanding the spatial relationship between biodiversity and carbon storage is essential for guiding conservation efforts to tackle jointly the biodiversity loss and climate change crises jointly.  Here, we assess tree species diversity and above-ground carbon storage patterns in response to climate, topography, and soil properties. We also analyze the spatial congruence between these cobenefits of the northmost Neotropical dry forest in Mexico and explore to what extent existing protected areas in this region safeguard tree diversity-carbon storage hotspots (i.e., areas where the top quintile of both tree diversity and carbon storage overlap). We used tree diversity and carbon storage data derived from 234 national forest inventory plots to fit machine learning models using a large set of climatic, topographic, and edaphic predictors. We then assessed the spatial congruence of predicted tree diversity and carbon storage with current conservation zones. Low temperatures are a limiting factor for tree diversity and carbon storage. Areas with high evapotranspiration, which is tightly coupled with seasonal water availability and vegetation cover, contain greater values of both predicted variables, but only 10.2% of tree diversity and carbon storage hotspots overlapped. Most of these overlapping hotspots lie outside the boundaries of existing protected areas that protect less than 2% of diversity-carbon hotspots. Our results highlight the low-temperature limitation influence and the key role of water availability –not only for tropical dry forest phenology and productivity, which are well documented, but also for sustaining tree diversity and carbon storage. Future changes in climate regimes have the potential to alter both plant diversity and carbon pools. Synthesis and applications: We identified areas of high conservation value, including prospective ecological reserves, corridors, and stepping stones, to protect hotspots of tree diversity and carbon storage, and to connect them to existing reserves, facilitating species movement and range expansion under changing climatic conditions. Biodiversity and carbon pools should be considered together for more effective spatial planning processes, especially at the regional scale, where management decisions are more easily implemented and evaluated.

We obtained tree species richness and derived the above-ground vegetation carbon pool from all individuals (90% live and 10% dead) in the National Forest and Soils Inventory of Mexico (INFyS, acronym in Spanish), with measurements carried out between 2009 and 2013 (https://snmf.cnf.gob.mx/datos-del-inventario/). This timeframe is the most complete to date. The majority of the data used in our study was collected between 2012 and 2013. The vegetation classification used in the INFyS is based on the Land Use and Vegetation Charter Series III-IV at 1:20.000 scale developed by INEGI. The INFyS follows a systematic stratified sampling scheme using plots called conglomerates (CGLs) as basic sampling units [compatible with Hollinger (2018)]. The CGLs are circular with a 56.42 m radius (1 ha). Each CGL contains four rectangular subplots of 400 m² each, for a total of 1600 m² per CGL, geometrically arranged as an inverted “Y”. These CGLs were distributed in a grid of points with a 10 km spacing in the tropical dry forest. We selected all tropical dry forest CGLs with data from all four sampling units, which included stems ≥7.5 cm in diameter at breast height (DBH) individually recorded, measured, and identified to species-level.

Our study was conducted using a final set of 234 CGLs (i.e., plots with four sampling units) distributed across our study region (Figure 1). Tree richness refers to the number of species per CGL or plot (i.e., species per 1600 m2). We estimated the live tree biomass (Mg ha ^-1) of each CGL using allometric equations with diameter at breast height as a predictor. Biomass for stems with a DBH ≥7.5 and ≤32.6 cm was calculated using the Návar (2009) allometric equation developed for this forest type within this stem width range. For stems with a DBH > 32.6, we used the allometric equation by Bojórquez et al. (2020), appropriate for larger trees. Live tree biomass was converted to tree carbon storage by multiplying the carbon fraction for the tropical dry forest of 0.4809 g C per gram of dry aboveground woody plant biomass (Jaramillo et al., 2003). For each plot, climatic, including hydrologic, information was extracted from TerraClimate (Abatzoglou et al., 2018). Climate variables were extracted by overlapping the climatic layer with the geographic coordinates of each plot using the ‘raster’ package in the R software (R Core Team, 2024). We subsequently calculated the bioclimatic variables using the ‘biovars’ function of the ‘dismo’ package in R (Hijmans et al., 2023). Also, for each plot center, topographic variables were determined using the digital elevation model (DEM) extracted from INEGI (2003) with a spatial resolution of about 15 m². Finally, for each plot center, we extracted the soil information from the SoilGrids 2.0 product (Poggio et al., 2021), with a spatial resolution of 250 m².

References

• Abatzoglou, J. T., Dobrowski, S. Z., Parks, S. A., & Hegewisch, K. C. (2018). TerraClimate, a high-resolution global dataset of monthly climate and climatic water balance from 1958-2015. Scientific Data, 5. https://doi.org/10.1038/sdata.2017.191
• Bojórquez, A., Martínez-Yrízar, A., Búrquez, A., Jaramillo, V. J., Mora, F., Balvanera, P., & Álvarez-Yépiz, J. C. (2020). Improving the accuracy of aboveground biomass estimations in secondary tropical dry forests. In Forest Ecology and Management 474: 118384. https://doi.org/10.1016/j.foreco.2020.118384
• Hijmans, R. J., Phillips, S., Leathwick, J., & Elith, J. (2017). dismo: Species Distribution Modeling (R package version 1.3-14). https://CRAN.R-project.org/package=dismo 
• Hollinger, D.Y. (2008). Defining a Landscape-Scale Monitoring Tier for the North American Carbon Program. In: Hoover, C.M. (eds) Field Measurements for Forest Carbon Monitoring. Springer, Dordrecht. https://doi.org/10.1007/978-1-4020-8506-2_1
• INEGI. (2013). Continuo de Elevaciones Mexicano (CEM) escala 1: 50,000, modelos digitales de elevación del Territorio Nacional. Instituto Nacional de Estadística, Geografía e Informática. Mexico. https://www.inegi.org.mx/app/geo2/elevacionesmex/
• Jaramillo, V. J., Kauffman, J. B., Rentería-Rodríguez, L., Cummings, D. L., & Ellingson, L. J. (2003). Biomass, Carbon, and Nitrogen Pools in Mexican Tropical Dry Forest Landscapes. Ecosystems, 6(7), 609–629. https://doi.org/10.1007/s10021-002-0195-4
• Návar, J. (2009). Allometric equations for tree species and carbon stocks for forests of northwestern Mexico. Forest Ecology and Management, 257(2), 427–434. https://doi.org/10.1016/j.foreco.2008.09.028
• Poggio, L., De Sousa, L. M., Batjes, N. H., Heuvelink, G. B. M., Kempen, B., Ribeiro, E., & Rossiter, D. (2021). SoilGrids 2.0: Producing soil information for the globe with quantified spatial uncertainty. SOIL, 7(1), 217–240. https://doi.org/10.5194/soil-7-217-2021
• R Core Team. (2024). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. https://www.R-project.org/