Advancements in ecological niche models for forest adaptation to climate change: a comprehensive review

Abstract

Climate change poses significant challenges to the health and functions of forest ecosystems. Ecological niche models have emerged as crucial tools for understanding the impact of climate change on forests at the population, species, and ecosystem levels. These models also play a pivotal role in developing adaptive forest conservation and management strategies. Recent advancements in niche model development have led to enhanced prediction accuracy and broadened applications of niche models, driven using high-quality climate data, improved model algorithms, and the application of landscape genomic information. In this review, we start by elucidating the concept and rationale behind niche models in the context of forestry adaptation to climate change. We then provide an overview of the advancements in occurrence-based, trait-based, and genomics-based models, contributing to a more comprehensive understanding of species responses to climate change. In addition, we summarize findings from 338 studies to highlight the progress made in niche models for forest tree species, including data sources, model algorithms, future climate scenarios used and diverse applications. To assist researchers and practitioners, we provide an exemplar data set and accompanying source code as a tutorial, demonstrating the integration of population genetics into niche models. This paper aims to provide a concise yet comprehensive overview of the continuous advancements and refinements of niche models, serving as a valuable resource for effectively addressing the challenges posed by a changing climate.

Keywords: assisted migration; climate niche shift; forest adaptation; genecology; landscape genomics; species distribution models (SDMs).

Fig. 1

Conceptual graph depicting the concept and rationale of niche models. (A) Biotic–Abiotic–Mobility (BAM) model and concept of fundamental niche (FN) and realized niche (RN) (adapted from Soberon & Peterson, 2005). Green area A represents abiotic factors and can be regarded as the fundamental niche of a species, blue area B represents biotic factors and red area M represents migration capacity of the species. The realized niche can be represented by RN = A∩B. The area P = A∩B∩M represents areas with observed species presence. Dots in the RN area represent species presence points, the grey‐filled dots represent the observed presence points while the empty dots represent the potential species occurrence that has not been observed. (B) The relationship between fundamental niche, realized niche and species occurrence points. (C) The biased prediction niche (BPN, pink region) due to partial occupancy of occurrence points. Filled dots represent species occurrence points that were documented or recorded, while empty dots represent potential occurrence points that have not been observed or sampled. (D) The green area represents the current climate niche for a species; the blue area represents its predicted future niche. The pink area represents the future biased‐prediction niche; as some niche models relate species presence to environmental factors, the future predictions can be outside of the current range.

Fig. 2

Applications of three types of climate niche models – occurrence‐based, trait‐based and genomic‐based. (A) An occurrence‐based niche model can be used at the community level by stacking the predictive distribution maps for multiple tree species. (B) Example of a trait‐based model where the transfer function and response function are used to predict the universal response function (Wang et al., 2010). (C) A genomics‐based model in which information on a single nucleotide polymorphism (SNP) can be applied to delineate seed zones for assisted migration (Yu et al., 2022).

Fig. 3

Algorithm, climate scenario, research content, and the number of tree species used in each study for the 338 studies included in the literature review. (A) The model types used, divided into ten categories based on the chosen algorithm. GAM, Generalized Additive Model; GLM, Generalized Linear Model; CLIMEX, mechanistic model; BIOCLIM, correlative model based on bioclimatic envelopes; SDM, Species Distribution Model; ENM, Ecological Niche Model. (B) Climate scenarios used in research. (C) Research content categorized into six types. (D) The number of tree species used in each study. (E) Suitable habitat change in future; ‘decrease’ and ‘increase’ represent predicted future suitable habitat loss and expansion respectively, ‘varies’ indicates where the predicted trend includes both increase or decrease among different species or different periods. (F) Direction of centroid shift or species range shift in future; ‘upslope’ represents higher elevation, ‘mixed’ indicates that different directions were predicted across different species or different periods.

Fig. 4

The development of trait‐based models and the key functions. All the functions are based on transfer or response data sets. The transfer data set involves diverse populations transferred to a single test site, while the response data set comprises a single population tested across multiple sites. (A) Individual Transfer Function (ITF) relates populations' fitness (e.g. height) to the climatic distance between provenance climates and a test site; (B) Individual Genecology Function (IGF) relates populations' fitness to provenance climates within one test site; (C) Individual Response Function (IRF) relates the fitness of one population to climates in multiple test sites; (D) Pooled Transfer Function (PTF) is an ensemble of ITFs across two or more test sites; (E) Universal Transfer Function (UTF) relates populations' fitness to both climate distance and test site climates; (F) Universal Response Function (URF) links populations' fitness to climates in both test sites and provenance locations. Each tree in the map symbolizes a population or provenance, while differently coloured circles in the plots depict distinct test sites.

Fig. 5

Application of a gradient forest climate niche model to data for lodgepole pine in response to climate change. (A) Ranking of 20 climate variables driving the local adaptation pattern across 281 lodgepole pine populations in British Columbia and Alberta, Canada. See Table S1 for definitions of climate variables. (B) Cumulative importance plot showing the significant allele frequency change for extreme minimum temperature (EMT). (C) Seed zones cluster based on principal component analysis (PCA) results shows the correlation between allele frequency and environmental variables. (D) Genetic offsets of lodgepole pine under the predicted climate RCP8.5 2041–2070.

Fig. 6

(A) The 20 tree species studied most frequently in our database. (B) The number of tree species studied on each continent and the percentage of studies for each continent (N = 338 studies) included in the literature review.