TY - JOUR
T1 - Shaping the niche of Taxus baccata, a modelling exercise using biologically meaningful information
AU - Sanchez-Martinez, Pablo
AU - Marcer, Arnald
AU - Mayol, Maria
AU - Riba, Miquel
N1 - Funding Information:
We particularly thank E. Berganzo for the field work and many local botanical experts, among them X. Rotllan, R. Bernal, F. Carrillo and “La Asociación de Amigos del Tejo”, for providing information on T. baccata occurrences, and M. Batalla and I. Regalado for their work in building the final data set. J. Ferreiro and P. Ramil-Rego (IBADER) kindly provided information on T. baccata occurences for northern Spain (obtained within the framework of the project LIFE-BACCATA, LIFE15 NAT/ES/000790). We are also indebted to M. Ninyerola for helpful discussions in the early phases of the work. This work was supported by the Spanish Ministry of Science and Innovation (grants CGL2007-63107/BOS, CGL2011-30182-C02-01/02 and PID2020-112738 GB-I00). P.S.-M. acknowledges an FPU predoctoral fellowship from the Spanish Ministry of Science, Innovation and Universities (grant FPU18/04945).
Publisher Copyright:
© 2021 Elsevier B.V.
PY - 2021/12/1
Y1 - 2021/12/1
N2 - Widely used Correlative Species Distribution Models (C-SDMs) usually make some simplifying assumptions, often failing to consider important ecological and evolutionary attributes potentially hindering the characterization of the species niche. Here, we use the tree species Taxus baccata to explore the effects of including biologically meaningful information on processes and features beyond purely abiotic factors that are expected to determine its niche and range size. To elucidate how these often neglected factors affect C-SDM results, we modelled the current niche in the species’ southernmost European range using Maxent. More specifically, we included available basic information regarding biotic interactions, local adaptation and non-equilibrium demographic dynamics. The potential effect of biological interactions was introduced using habitat suitability of co-occurring tree species as predictive variables. Local adaptation was included modelling two distinct regional adaptive groups. We also used individual growth estimated under field conditions as a surrogate for demographic behaviour to control for the quality of model predictions and empirically assess the effect of biotic interactions. Including information on co-occurring tree species improved model performance and decreased the projected range size in most cases. These effects were not a result of biological interactions per se, but instead a consequence of co-occurring species accounting for fine-scale environmental variability not described by any of the climatic variables used. Considering local adaptation allowed detecting the role of different climatic variables in shaping the niche of each adaptive group that could potentially also act as selective pressures in the near future. Finally, and more importantly, we found that including populations that are probably currently found under non-equilibrium suboptimal conditions might largely overestimate the species niche.
AB - Widely used Correlative Species Distribution Models (C-SDMs) usually make some simplifying assumptions, often failing to consider important ecological and evolutionary attributes potentially hindering the characterization of the species niche. Here, we use the tree species Taxus baccata to explore the effects of including biologically meaningful information on processes and features beyond purely abiotic factors that are expected to determine its niche and range size. To elucidate how these often neglected factors affect C-SDM results, we modelled the current niche in the species’ southernmost European range using Maxent. More specifically, we included available basic information regarding biotic interactions, local adaptation and non-equilibrium demographic dynamics. The potential effect of biological interactions was introduced using habitat suitability of co-occurring tree species as predictive variables. Local adaptation was included modelling two distinct regional adaptive groups. We also used individual growth estimated under field conditions as a surrogate for demographic behaviour to control for the quality of model predictions and empirically assess the effect of biotic interactions. Including information on co-occurring tree species improved model performance and decreased the projected range size in most cases. These effects were not a result of biological interactions per se, but instead a consequence of co-occurring species accounting for fine-scale environmental variability not described by any of the climatic variables used. Considering local adaptation allowed detecting the role of different climatic variables in shaping the niche of each adaptive group that could potentially also act as selective pressures in the near future. Finally, and more importantly, we found that including populations that are probably currently found under non-equilibrium suboptimal conditions might largely overestimate the species niche.
KW - Biotic interactions
KW - Local adaptation
KW - Plant performance
KW - Range size
KW - Species Distribution Models (SDM)
KW - Taxus baccata
UR - http://www.scopus.com/inward/record.url?scp=85115244720&partnerID=8YFLogxK
U2 - 10.1016/j.foreco.2021.119688
DO - 10.1016/j.foreco.2021.119688
M3 - Article
AN - SCOPUS:85115244720
VL - 501
M1 - 119688
ER -