Guerra-Coss et al. Modelling and validation of the spatial distribution of suitable habitats for the recruitment of invasive plants on climate change scenarios: an approach from the regeneration niche

Published: 22 December 2020| Version 1 | DOI: 10.17632/99r49328cb.1
Ernesto I. Badano


The study focuses on modelling and validating the current and future distribution of suitable habitats for the recruitment of invasive plants, using the biological invasion of the Peruvian peppertree (Schinus molle L., Anacardiaceae) in Mexico as example. To model the distribution of suitable recruitment habitats, we used 77 occurrence data corresponding to sites where seedlings and saplings are naturally established. The WorldClim geodatabases were used to obtain the current values of 19 bioclimatic variables associated to each of these sites. This information is provided in the Excel file “Data 01 - Occurrence data and bioclimatic variables”. The first sheet of this file (Occurrence data) contains the coordinates of occurrence points (latitude-longitude) and the values of the 19 bioclimatic variables associated to each point. The second sheet (Correlation results) contains the correlation coefficients of the Spearman rank-order tests used to define what variables must be included in the model. In both pages, these variables are indicated in green. The resulting model was projected on the current climate and climate change scenarios and, to validate these predictions, field experiments were conducted at three sites (San Nicolas Tolentino, Armadillo de los Infante and La Salitrera). In these experiments, we assessed the effects of higher temperature and lower rainfall on the emergence and survival of peppertree seedlings. Climate change conditions were simulated with open-top chambers and rainout shelters, while control plots were exposed to the current climate. Additionally, we included experimental plots in which only temperature or rainfall was manipulated. There were five replicates for each climate treatment. The results of experiments are included in the Excel file “Data 02 - Data of field experiments”. In all cases, codes are: CONTROL = control plots under the current climate; SCC = climate change simulations plots with open-top chambers and rainout shelters; OTC = warming plots with open-top chambers only; RSO = rainfall reduction plots with rainout shelters only. The first sheet of this file (Microclimate data) contains data of temperature and rainfall recorded at the experimental units of each site. These data are averages that integrate information from 09 July 2019 to 01 July 2020 for each experimental unit. The second sheet (Seedling emergence and survival) contains the data used to estimate seedling emergence and survival rates for each climate treatment. These databases indicate the date of shoot emergence and seedling dead for each of the 200 seed we sowed at each experimental unit.


Steps to reproduce

We were interested in calibrating a habitat suitability model for the distribution of habitats that meet the regeneration niche of peppertrees. Thus, we used occurrence data of seedlings and saplings only. These occurrence data were obtained between 2012 and 2018 with a series of field surveys conducted in Mexico. In these surveys, we track peppertrees across main roads, secondary roads and rural trails, and performed intensive searches of seedlings and saplings at sites where adult individuals were visualized. If seedlings and saplings were found, we determined whether they were naturally established or planted by man. Samplings were never conducted within human settlements because it is difficult to differentiate natural and human-subsidized occurrences at these sites. Climatic data associated with occurrences of seedlings and saplings were obtained from the WorldClim geodatabases. Because repeated information within the same spatial unit may cause overfitting of HSM, a 5-km diameter plot was draw around each occurrence point and a single record was retained when two or more points were overlapped. AS overfitting can also occur when models include redundant variables, we conducted Spearman rank-order correlation tests for all pairwise combinations of climatic variables. When correlation coefficients above 0.6 were detected, we retained the variable with higher correlation coefficients with the others. The validation experiments focused in assessing the emergence and survival rates of peppertree seedlings under the current climate and under simulated climate change conditions. Biological material (seeds) for the experiments was collected in March 2019 at the three sites in which they were conducted. To estimate the climate change conditions that must be simulated at each site, we looked for normal values (averages over the last 30 years) of air temperature and annual rainfall of the experimental sites and computed the differences between these values and those predicted by the CanESM5 climate change model for the period 2041-2060. Using these values as reference, temperature was increased with hexagonal open-top chambers and rainfall was reduced with rainout shelters. A total of 200 seeds corresponding to each site were sowed at each experimental unit and the emergence of aerial shoots and survival of seedlings was monitored every 21 days between 09 July 2019 and 01 July 2020.