Global patterns of thermal niche filling in ectotherms

doi:10.21203/rs.3.rs-2635490/v1

Download PDF

Article

Global patterns of thermal niche filling in ectotherms

https://doi.org/10.21203/rs.3.rs-2635490/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 06 Nov, 2023

Read the published version in Nature Ecology & Evolution →

Version 1

posted

You are reading this latest preprint version

Understanding how temperature determines the distribution of life is necessary to assess species’ sensitivities to contemporary climate change. Here we test the importance of temperature in limiting geographic ranges of ectotherms by comparing temperatures across occupied ranges to those species could potentially occupy based on their physiological thermal tolerances. Whereas marine and tropical terrestrial species occupy temperatures that closely match their thermal tolerances, high-latitude terrestrial species under-occupy warm temperatures and are absent from thermally tolerable areas towards the equator. This suggests that on land, temperature less often limits the equatorward range edge of temperate species, supporting the hypothesis that their historic expansion ‘out of the tropics’ was associated with tropical biotic exclusion. Our findings predict more direct responses to climate warming of marine ranges and cool range edges of terrestrial species.

Biological sciences/Ecology/Biogeography

Biological sciences/Ecology/Macroecology

Biological sciences/Evolution/Evolutionary theory

Biological sciences/Ecology/Climate-change ecology

Biological sciences/Ecology/Evolutionary ecology

Temperate species underfill their tropical thermal potentials on land.

Climate warming is already altering the distributions of species worldwide¹. Yet sensitivity to climate change varies considerably among species^2,3, calling into question how often temperature limits geographic ranges. If a species occupies all accessible habitat where temperatures suit its tolerances (i.e., the species fills its potential thermal niche), then we expect the species’ range limits to be sensitive to temperature change⁴. Yet in reality, species are prevented from filling their potential thermal niche since ranges are often limited by factors other than temperature^5–7, such as dispersal⁸, species interactions⁹, resource availability⁹, and non-thermal abiotic factors like moisture (on land)¹⁰ or oxygen (in water)¹¹. Understanding where and when temperature directly constrains species ranges can help clarify the mechanisms responsible for historical range shifts and improve projections of species’ sensitivities to contemporary climate warming.

A relevant long-standing hypothesis posits that temperature and abiotic factors are more limiting at species’ poleward range edges compared to their equatorward edges^12,13. This hypothesis is supported by a recent synthesis of empirical studies of the ecological factors limiting species ranges, which shows that biotic interactions consistently influence species' low-latitude and low-elevation range edges more often than their high-latitude and high-elevation edges⁹. One possible cause for this pattern, for which empirical evidence is strong^14–18, is an increased intensity of species interactions toward the tropics^9,12,13 owing to the increased biodiversity, density, or activity levels in the more productive, warmer, and more seasonally-stable tropics (reviewed by ref ¹⁹). However, alternative hypotheses exist for how biotic interactions might override or supersede direct temperature limitation in the tropics.

Two particularly relevant hypotheses offer testable predictions about how biotic interactions in the tropics might limit species’ ranges. The first posits that the role of abiotic factors in limiting species ranges gradually decreases towards the tropics because of an increasing role of biotic interactions^14–17. Under this ‘Reduced-Abiotic-Limitation-in-the-Tropics Hypothesis’, we would expect both poleward and equatorward range limits to be more abiotically-limited at higher latitudes and less so at lower latitudes. The second hypothesis postulates that the reduction of abiotic limitation at warm range edges is more pronounced for species outside of the tropics that evolved tolerance to colder and more variable climates. This is an extension of the Out of the Tropics hypothesis^20,21, which posits that some tropical lineages evolved greater cold tolerance that enabled them to expand into relatively uninhabited temperate and polar environments. If greater cold tolerance came at the cost of lower resistance to competitors and natural enemies because of a trade-off between performance and thermal tolerance breadth (i.e., ‘jack-of-all-trades is master of none’, Principle of Allocation, refs ^22,23), temperate species would fare worse in antagonistic species interactions compared to tropical species, potentially leading to greater biotic exclusion of higher-latitude species at their equatorward range edges. We call this the ‘Out-of-the-Tropics-Trade-Off Hypothesis’.

While not mutually exclusive, these two hypotheses nevertheless make contrasting and testable predictions about how species interactions alter the relative importance of temperature in limiting ranges. The Reduced-Abiotic-Limitation-in-the-Tropics Hypothesis predicts that all range edges will become increasingly biotically constrained toward the equator such that species will be increasingly excluded from environments with tolerable temperatures towards low latitudes. On the other hand, the Out-of-the-Tropics-Trade-Off Hypothesis predicts greater exclusion of high-latitude species at their equatorward range edges specifically.

We tested among these hypotheses in a spatially-explicit global analysis of the potential and realized thermal niches of ectotherms. We used critical and lethal thermal tolerance limits from experimental assays to define the fundamental thermal niche (Fig. 1). In current climates, some temperatures within the fundamental thermal niche might not occur anywhere or might only occur far outside a species’ current distribution. To describe the tolerable thermal niche space available to a species given climatic and geographic constraints, we delineated the potential thermal niche as the “encounterable” warm and cool extreme body temperatures within the fundamental thermal niche (see Methods). We assessed how well species fill their potential thermal niches by comparing each species’ potential thermal niche to its realized thermal niche, defined by the warm and cool extreme body temperatures across its current estimated geographic distribution (i.e., its realized range). We measured niche filling as the difference between the potential and realized thermal niche extremes. Negative values denote niche underfilling (i.e., cases in which species do not occupy all extreme temperatures available to them, sensu Sunday et al. 2012²⁴; Fig. 1a,c) and positive values denote niche underprediction (i.e., cases in which species’ thermal tolerance limits are less extreme than the extreme body temperatures expected across their current geographic distributions; Fig. 1a,c).

Whereas thermal niche filling describes offsets between occupied and tolerable temperatures, how these offsets play out in geographic space depends on the spatial distribution of temperature²⁵ (Fig. 1b,d). We therefore additionally assessed how patterns of potential thermal niche filling differ when analyzed in geographic space (Fig. 1b,d). We measured range filling as the proportion of a species’ potential thermal range that it occupies and assessed how range underfilling differs between species’ equatorward and poleward range halves.

We used relationships between niche filling and range filling with latitude to test the alternative expectations from the Out-of-Tropics-Trade-Off Hypothesis vs. the Reduced-Abiotic-Limitation-in-the-Tropics Hypothesis. We interrogated our hypotheses within terrestrial, intertidal and marine realms under the expectation that thermal niche filling differs across marine and terrestrial environments^3,24,26. We also asked if species with lower dispersal potential⁴ and species that are more ecologically specialized²⁷ have greater niche or range underfilling according to hypotheses in Table S1. We additionally assessed the sensitivity of latitudinal relationships to variation in thermal tolerance and encountered temperatures owing to species’ capacity to thermoregulate and to adjust their thermal limits through acclimatisation (see Methods).

According to their thermal tolerance limits, most terrestrial species could live in places with warmer extreme temperatures than those they currently experience. Hence, they underfill the warm ends of their potential thermal niche (dashed red lines in Fig. 2, Fig. 3a). This warm niche underfilling was lowest at the equator and increased with latitude (Figs. 2 and 3, fig. S1, table S2). By contrast, intertidal and subtidal marine species underfilled their warm thermal niche less than terrestrial species (closer to zero, i.e., perfect filling), and their amount of underfilling did not change significantly with latitude (Figs. 2 and 3, fig. S1, table S2).

In geographic space, total range filling did not change with the latitudinal midpoint of a species range (Fig. 4a, fig. S1, table S3), but underfilling was generally biased towards the equatorward edges of ranges in terrestrial species, and this bias increased towards higher latitudes (Fig. 4b and red-blue color scale, fig. S1, fig. S2, table S3). By contrast, although marine species underfilled large proportions of their potential thermal ranges, indicating they do not occupy all thermally-tolerable areas, this range underfilling showed little latitudinal or thermal bias (white points and model fits close to zero, Fig. 4b, fig. S1, table S3).

Terrestrial taxa therefore support the Out-of-the-Tropics-Trade-Off Hypothesis. In terrestrial species, both niche and range underfilling are greatest at the warm or equatorward range edges of species that live outside of the tropics (i.e., temperate species underfill their tropical potentials). The observed patterns were opposite to the expectation of increased underfilling of the potential thermal niche and range in species closer to the tropics (Fig. 4b), as predicted under the Reduced-Abiotic-Limitation-in-the-Tropics Hypothesis. By contrast, intertidal and subtidal marine species did not show latitudinal patterns consistent with either hypothesis.

At the cool extremes of the thermal niche, thermal tolerance limits tended to underpredict species’ realized thermal niches (i.e., species appear to occupy places where temperatures are colder than their cold tolerance limits; solid blue lines in Fig. 2, Fig. 3b). Cool niche underprediction increased with the latitudinal midpoint of a species’ range in terrestrial and intertidal marine species, but not in subtidal marine species, for whom cool niche filling was closer to zero (i.e., perfect filling) and did not change with latitude (Figs. 2 and 3, fig. S1, table S2). We explore variation in niche underprediction below as a means to understand the mechanisms responsible (e.g., cold season dormancy).

The observed latitudinal patterns of niche and range filling were generally robust to taxonomic non-independence, variation in thermal limit assay method, and acclimatisation to local conditions, and remained after we accounted for some degree of phenotypic plasticity and thermoregulatory behaviours. We relaxed the assumption that species’ thermal tolerance limits are fixed over space and time by simulating acclimatisation of species to seasonal temperatures across the landscape (see Supplementary Methods). This generally led to broader potential thermal niches and reduced the extent of both warm and cool niche underprediction (grey shadow compared to coloured density distributions in Fig. 3a-b, fig. S3a; warm niche underprediction on land reduced by ~ 10°C, cool niche underprediction reduced by ~ 5°C). However, the relationships with latitude did not change (fig. S3b-e, table S4). Additionally, simulating thermoregulatory behaviour in a subset of terrestrial species (n = 219) by relaxing the assumption that animals always prefer shaded habitat accounted for some portion of warm niche underfilling (see Supplementary Methods), but patterns across latitude remained (fig. S5a-c, table S5).

Some cool niche underprediction remained after simulating acclimatisation and behaviour. This might be explained by organisms’ abilities to become seasonally dormant, to seek warmer habitats during extreme cold, to locally adapt, or to vary their cold thermal limits via physiological plasticity that was not accounted for by our simulation (e.g., rapid cold hardening²⁸). Although we attempted to use only temperatures during active time periods for species with known dormancy, limited information on the timing and duration of dormancy may lead to underestimates of dormancy periods (see Supplementary Methods). Additionally, within winter burrows, cooling rates experienced are typically much slower than those used in experiments²⁹, potentially weakening the connection between laboratory-assayed cold tolerance and in situ survival in microhabitats. Moreover, individuals in experimental trials were often collected from warmer parts of a species’ range (fig. S4), meaning our analysis might underestimate cold tolerance in colder parts of the range. Population differentiation in cold tolerance might be expected to be greater on land where gene flow across temperature gradients are likely lower than in the ocean^30,31. Importantly, these mechanisms (seasonal dormancy, acclimatisation, and local adaptation) that could explain cases where thermal tolerance is less extreme than temperatures across the realized niche cannot reasonably explain cases when the potential thermal niche exceeds the realized niche, suggesting the tendency for niche underfilling in terrestrial species is driven by other mechanisms.

We found no relationship between either dispersal distance or body size and how well species filled their potential thermal niche or range (fig. S1, table S2, table S3). We note that our analysis did not explicitly consider extrinsic barriers to dispersal (other than the land/sea interface and ecological realms), which might help to further identify relationships between niche underfilling and dispersal limitation. We did find that thermal niche filling was greater in species with larger geographic ranges (fig. S1, table S2), consistent with the hypothesis that larger-ranged species are less ecologically-specialized and thus more temperature-limited²⁷. However, this finding is likely tautological, as larger ranges take up a greater proportion of somewhat fixed thermal niche breadths.

Under the interpretation that the increase in warm niche underfilling on land is linked to biotic interactions, as would be consistent with previous findings⁹, it is intriguing to consider why marine species do not show the same pattern. It might indicate that biotic interactions are less limiting to marine species ranges, perhaps because trophic interactions in marine systems are thought to be more body size-based than species-based, reducing the specificity of some species interactions³². The lack of latitudinal pattern in warm underfilling within marine realms suggests biotic interaction gradients differ between land and ocean. Although there are examples of biotic interaction intensity decreasing with latitude in marine systems, these are mostly intertidal (reviewed in ref¹⁴). Alternatively, differences in thermal limit testing in air versus water might lead to an overestimation of heat tolerance limits in terrestrial species that might contribute to warm underfilling; air has a lower thermal conductivity compared to water, which could cause a delay in the responses of air-immersed animals to rapid heating in comparison water-immersed animals. Even so, empirical evidence that marine ranges are more responsive to climate change suggests there is a biological signal in the difference in warm underfilling between species on land versus in the ocean.

Mechanisms other than biotic exclusion could be responsible for warm niche underfilling. For example, other abiotic niche requirements may be more limiting or co-limiting in warm areas (e.g., moisture in the hot desert belts, oxygen in warmer ocean regions). Alternatively, warm underfilling might occur because species’ ecological limits to population growth are more limiting than an individual organisms’ capacity to function under heat stress (as generally measured in experiments). Temporal variability in temperatures and a history of thermal stress can, for example, reduce heat tolerance at the population scale^33–35. Similarly, if early life stages are more heat-sensitive than the adults typically assayed (e.g., refs ^36,37), or if sublethal temperatures limit critical life-history functions (e.g., mate-finding), wild populations might not be able to persist in sublethal temperatures. Identifying among the possible mechanisms of warm niche underfilling is important to understand species’ temperature sensitivities under climate warming (see Supplementary Materials).

Observational evidence of variation in species’ range shifts in response to climate warming already indicates greater sensitivities in marine compared to terrestrial species^3,26, consistent with the finding that marine species more closely fill their thermal niches. Observed range shifts can be used to test additional hypotheses stemming from results presented here; namely, if thermal niche underfilling is associated with less sensitivity to temperature, we predict broad-ranging species and species in the terrestrial tropics to be more sensitive to temperature change. We also predict warm range edges of extra-tropical species to be less sensitive to temperature change, with contractions more tied to drought or climate-related increases of antagonistically interacting species. Our results show that general patterns of temperature limitation among species emerge despite the existence of many complex factors that likely shape individual distributions. The shared evolutionary history of biodiversity might likewise lead to general patterns in how biodiversity and ecosystem services respond to contemporary climate change.

Fundamental thermal niches

We defined the fundamental thermal niche as a 1-dimensional Grinellian niche³⁸ with limits set by experimental estimates of species’ upper and lower thermal tolerance limits. We retrieved thermal tolerance data from the Supplementary Material of Sunday et al. 2019³⁹, which represents a curated subset of the Globtherm database⁴⁰ that contains an estimate of a critical or lethal upper or lower (or both) thermal tolerance limit per species. We subset data to include only subtidal marine, intertidal marine, and terrestrial ectothermic animal species (we excluded species from freshwater locations as we lack appropriate global temperature data for freshwater environments). We included terrestrial species with a freshwater larval stage so long as the thermal limit assays were performed on adult organisms. Despite its use to test other hypotheses related to thermal tolerance evolution^41,39, this study represents a novel use of the Globtherm dataset.

Realized ranges and traits

For species retained from Globtherm, we extracted geographic range maps in the form of polygons (extents of occurrence) from two sources: the International Union for the Conservation of Nature (IUCN) spatial data portal⁴² (n = 318), and the Global Assessment of Reptile Distributions (GARD) data repository⁴³ (n = 51). When a range map was available from both sources, we used the IUCN range map in analyses. Additional range maps for species retained from Globtherm were inferred by fitting convex hulls to carefully filtered occurrence records obtained from the Global Biodiversity Information Facility⁴⁴ (n = 225, see Supplementary Methods), following IUCN methods (iucnredlist.org/resources/mappingstandards). For each species, we then converted range map polygons to 1° × 1° resolution presence/absence grids and searched the literature for a suite of traits chosen in accordance with our hypotheses (table S1, see Supplementary Methods) to include as predictor variables in models. Results were not sensitive to the source of the realized range (fig. S6).

Estimating body temperatures

We used climate data to estimate the warmest and coolest extreme body temperatures each species would experience in each grid cell. For marine and intertidal species, we used coarse-grained macroclimatic data to characterise the span of daily body temperatures a species would experience across the globe. We calculated daily maximum and minimum air surface temperature climatologies at a 1° × 1° grid resolution over the period 1950–2000 from Berkeley Earth land datasets (berkeleyearth.org/data/; accessed on 14 October 2020). We obtained daily maximum and minimum mean sea surface temperature climatologies over the period 1982–2020 from the NOAA Optimum Interpolation (OI) Sea Surface Temperature V2 High Resolution Dataset (psl.noaa.gov/data/gridded/data.noaa.oisst.v2.highres.html; accessed on 15 June 2021) and converted the higher resolution grid to a 1° × 1° resolution by aggregating cells, selecting the maximum or minimum value of aggregates. To account for the fact that intertidal species often experience both wet and dry temperatures, we considered both air and sea surface temperatures when defining their experienced temperatures.

Temperature estimates from coarse-grained macroclimatic data are not the actual temperatures terrestrial organisms experience⁴⁵. To better characterise the extreme body temperature that terrestrial organisms experience, we used species traits and microclimatic data from NicheMapR⁴⁶ to simulate species-specific hourly operative temperatures in the sun and shade across a year in each grid cell (see Supplementary Methods). To ensure we did not overestimate how close species live to their thermal tolerance limits, we assumed in our main analysis that species’ body temperatures would be equilibrated to the shade when experiencing hot extremes and to the sun when experiencing cold extremes, and so defined the species’ extreme body temperatures using the hottest hourly shaded operative temperature and the coldest hourly exposed operative temperature. For species with seasonal dormancy, we masked body temperatures experienced during the six hottest and/or coldest months of the year before selecting the hottest and coolest extremes (see Supplementary Methods; fig. S7). To ensure our operative temperature estimates reasonably represented possible body temperatures of terrestrial species, we used empirical data on species’ field body temperatures from the literature to validate them (fig. S10). We also tested the sensitivity of our operative temperature estimates to variation in chosen model parameter (fig. S11; see Supplementary Methods).

Potential and realized thermal niches

For each species, we first inferred the potential thermal niche in geographic space in the form of a 1° × 1° resolution presence/absence grid by applying a series of species-specific restrictions to reduce available habitat. First, to avoid overestimating niche underfilling by including habitat in the wrong environment (e.g., including areas of ocean in the potential niche of a terrestrial reptile) or that is uninhabited due to large-scale, historical dispersal barriers (e.g., continental divides), we restricted habitat by the inhabited species’ realm. We did this by removing grid cells that were not within habitat in areas of the species’ realm (marine species: ocean; intertidal species: ocean within 200 km of the coastline and land within a 1° grid cell of the coastline; terrestrial species: biogeographic realm(s), as determined by Holt et al. 2013, ref ⁴⁷) that were contiguous with its realized range. We then removed remaining grid cells where the species’ extreme body temperatures exceeded one (or both) of a species’ fundamental thermal niche limits. For species with only one known thermal tolerance limit, we used only one fundamental thermal niche limit to restrict the available habitat. We then used global grids of elevation (earthenv.org/topography) and depth (gebco.net/data_and_products/gridded_bathymetry_data/), to remove remaining cells where habitat was of an inhabitable depth (marine) or altitude (terrestrial) for the species, when information on species depth or elevational distributions were available in the literature (i.e., known elevation or depth range, whether a marine species is pelagic or benthic-associated; see Supplementary Methods).

We then derived the potential and realized thermal niches in environmental space from estimates of the potential and realized range. To do this, we quantified the span (maximum and minimum) of extreme body temperatures occurring across the 1° × 1° resolution presence/absence grid. Similarly, we quantified the span of extreme body temperatures occurring across the species' realized range to define the realized thermal niche in environmental space. For species with only one known fundamental thermal niche limit, the potential thermal niche was inferred from only one limit, which left us with 460 potential thermal niches (both limits: n = 197; upper limit only: n = 219; lower limit only: n = 44).

Measuring thermal niche filling

In environmental space, we calculated warm and cool filling of the potential thermal niche as the difference between the potential and realized warm and cool thermal niche extremes, respectively. This metric is in units of degrees Celsius. Negative values indicate a species is underfilling its cool or warm potential thermal niche, whereas positive values indicate underprediction, and a value of zero indicates the species’ perfectly fills its thermal niche. Some species did not have temperature data available across their realized range (n = 24), allowing us to measure niche filling in thermal space for 436 species. For species with available body temperature data and both warm and cool thermal limits (n = 186), we calculated both warm and cool thermal niche filling, while only one niche filling value was calculated for species with available body temperature data and only one thermal limit (warm niche filling only: n = 206; cool niche filling only: n = 44).

For species with potential thermal niches inferred using both fundamental limits (n = 197), we also calculated filling of the potential thermal niche in geographic space (range filling). We calculated range filling as the proportion of cells in the species’ potential range that the species occupied, which ignores areas of geographic underprediction. While in environmental space a niche filling value of 0 indicates perfect niche filling, in geographic space a range filling value of 0 indicates complete range underfilling (i.e., occupies no areas of its potential range) while a value of 1 indicates perfect range filling (i.e., occupies all areas of its potential range). We also analysed the asymmetry in range underfilling by calculating the difference between the proportion of the potential range that is underfilled in the equatorward and poleward range halves. To do this, we split the potential range in half latitudinally at the mid-latitude of occupied cells, calculating range underfilling in either half (i.e., the proportion of cells in either half of the range that were not occupied), and subtracting range underfilling in the equatorward half from range underfilling in the poleward half. We did not analyse range filling or asymmetry in underfilling for species whose realized range and potential thermal range did not overlap (n = 37), leaving range filling estimates for 160 species.

Analyses

To test our hypotheses, we fit linear mixed-effect models to warm niche filling, cool niche filling, range filling, and asymmetry in underfilling separately using the nlme package⁴⁸. We included the following traits as fixed effects: realm: categorical; absolute latitudinal midpoint of realized range: continuous; dispersal distance: continuous; log(maximum body size): continuous; log(realized range size): continuous. We excluded log(realized range size) from the range filling model as we recognized that shared geometric constraints imposed by continental barriers on potential ranges might result in a circularity between range size and geographic range filling (see Supplementary Methods). As we expected the relationship between niche filling and latitude would differ across realms based on prior findings²⁴, we included an interaction term between realm and absolute latitude. We only modelled species for which all traits were known, which slightly reduced our sample sizes (warm niche filling: n = 382; cool niche filling: n = 227; range filling and asymmetry in range filling: n = 156; final list of species included in models is found in Appendix A). Since most species underfilled their potential thermal range, range filling values were right-skewed and so were log transformed before modelling. To ensure no variables covaried, we checked for collinearity between explanatory variables using variance inflation factors and removed one of the variables causing the collinearity when values > 3 were found⁴⁹. No variables were removed on account of collinearity.

Using data from diverse sources and on diverse taxa introduces variation that can be accounted for through linear mixed effect modelling techniques. To account for differences caused by experimental procedures, we included thermal limit metric type (critical or lethal) as a fixed effect in our niche filling models, however this was not possible for the range filling model as metric type was found to be collinear with the term “realm”. To account for covariance of data due to taxonomic non-independence and methodological differences in measuring the fundamental thermal niche across taxonomic groups, taxonomic categories from Class through to the Species level were included in all models as nested random effects on the intercept. We acknowledge that this approach does not take into account different divergence times between taxa.

We then performed model averaging rather than selecting a single top model so as to avoid introducing uncertainty through the model selection process. We used the MuMIn package⁵⁰ to run all candidate models, which included all possible combinations of terms. We performed multi-model averaging with maximum likelihood estimation to identify the confidence set, or the models comprising the top 95% of model weight. We report full averages of coefficients. To ensure the assumption of normality was met, model residuals were visually inspected. To ensure the exclusion of certain explanatory variables from some models did not dramatically affect the estimation of other model parameters (one possible problem introduced by model averaging, ref ⁵¹), parameter estimates from models in the confidence set were visually compared to each other and to the model average (see fig. S8).

Acclimatisation

While our main analysis assumes that a single upper and lower thermal tolerance defines a species’ fundamental thermal niche across its range, we used reduced datasets (warm niche filling: n = 212; cool niche filling: n = 134; range filling: n = 90) to simulate the potential plasticity of fundamental thermal limits to local seasonal temperatures. We compiled data on acclimation response ratios from published^52–55 (47–50) databases and augmented these with individual studies with a new literature search (see Supplementary Methods). We used these to simulate the fundamental thermal niche limits of field-acclimatised individuals based on recently-experienced temperatures in each grid cell (see Supplementary Methods, fig. S9). We then inferred species’ potential thermal niches by comparing their extreme body temperatures in a grid cell to their locally-acclimatised fundamental thermal niche limits. We modelled acclimatised range filling and warm and cool thermal niche filling using the same linear mixed effect modelling framework as in our main analysis and fit our original models to only species included in the acclimation subset (fig. S3, table S4, table S6).

Behavioural thermoregulation

For a subset of terrestrial species (n = 219) for which we could estimate a preferred body temperature, defined as the body temperatures species maintain in nature⁵⁶ either from measurements of preferred temperature in a thermal gradient (T_pref) or from measurements of field body temperature (T_b) from the literature, we adjusted realized upper thermal niche limits to reflect the hottest temperature species would experience in their realized range if allowed to behaviourally thermoregulate towards their weighted mean preferred temperature by moving between the sun and shade (see Supplementary Methods). We modelled warm thermal niche filling in this subset using the same linear mixed effect modelling framework as in our main analysis and fit our original model of warm thermal niche filling to only species included in the behaviour subset (fig. S4, table S5). Although marine species can also regulate their body temperature by moving to different depths of the water column, we could not carry out a similar analysis for marine species due to the limitations imposed by our small sample size.

Acknowledgments: We are grateful for contributions to geographic range validation by Rafael Moreira Lima.

Funding: This paper is a joint effort of the working group sWEEP supported by sDiv, the Synthesis Centre of iDiv (DFG FZT 118, 202548816). N.A.M. and J.M.S. were supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) and N.A.M. was also supported by the Quebec Centre for Biodiversity Science and the Rubin Gruber Science Undergraduate Research Award. I.M-C. acknowledges funding from the Spanish Ministry for Science and Innovation (grant no. PID2019-109711RJ-I00 to I.M-C.) as well as from Comunidad de Madrid and University of Alcalá (funders of IFW through grant CM/BG/2021-003 to I.M-C.). F.V. was supported by CONACyT and INECOL, Mexico. P.C. was supported by a NSERC Discovery Program grant (No. 2020-05627). J.G.R. was supported by the European Commission’s Marie-Skłodowska Curie Individual Fellowship (H2020-MSCA-IF-2018, 843094).

Author contributions: J.M.S., I.M-C., M.A.O.T., F.V., P.C., S.C-T., A.C.A., B.M., L.R., J.M.B., C.R., and M.B.A. conceived the project. I.M-C., M.A.O.T., J.M.S., N.A.M., and A.L.H. helped secure funding to develop the project. N.A.M. collated climate and thermal preference data, combined datasets of realized ranges, collated thermal preference data, collated acclimation response data, ran analyses and data visualizations, and developed scripts. J.M.S. and N.A.M. conceived of specific analyses, data products and visualizations. G.C.V., I.M-C., and J.G.R. developed scripts for specific data products. N.A.M., J.M.S., I.M-C., A.L.H., M.A.O.T., F.V., P.C., S.C-T., A.C.A., B.M., L.R., S.G., and J.M.B. collected species trait data, N.A.M., A.L.H., and S.G. collated and checked the quality of traits data. A.L.H. and J.M.S. filtered occurrence records, and G.C.V. created range polygons from filtered occurrence records under the supervision of M.A.O.T. and I.M-C. J.R.B. collated thermal tolerance plasticity data. All authors participated in consultations about how to run analyses, while J.G.R. and F.V. were additionally consulted about specific analyses. N.A.M. created all figures in consultation with J.M.S. Authors J.M.S. and N.A.M. wrote the manuscript and all authors provided feedback, edits, and revisions to the manuscript. A.L.H. provided thorough feedback on the manuscript and helped majorly with revisions. J.M.S. supervised the project.

Competing interests: Authors declare that they have no competing interests.

Data and materials availability: Data and code needed for analyses associated with the current submission are available in the public Living Up to Thermal Potentials GitHub Repository, with large data files downloadable from Figshare (figshare.com/s/83bba4f780f53b6f21f5). Upon publication, the full repository and associated files will be archived and made available.

Pecl, G. T. et al. Biodiversity redistribution under climate change: Impacts on ecosystems and human well-being. Science 355, eaai9214 (2017).
Chen, I. C., Hill, J. K., Ohlemüller, R., Roy, D. B. & Thomas, C. D. Rapid range shifts of species associated with high levels of climate warming. Science 333, (2011).
Lenoir, J. et al. Species better track climate warming in the oceans than on land. Nat. Ecol. Evol. 4, (2020).
Araújo, M. B. & Pearson, R. G. Equilibrium of species’ distributions with climate. Ecography 28, 693–695 (2005).
Gaston, K. J. Geographic range limits of species. Proc. R. Soc. B Biol. Sci. (2009) doi:10.1098/rspb.2009.0100.
Brown, J. H., Stevens, G. C. & Kaufman, D. M. The geographic range: Size, shape, boundaries, and internal structure. Annu. Rev. Ecol. Syst. 27, (1996).
Sexton, J. P., McIntyre, P. J., Angert, A. L. & Rice, K. J. Evolution and Ecology of Species Range Limits. Annu. Rev. Ecol. Evol. Syst. 40, 415–436 (2009).
Hargreaves, A. L., Samis, K. E. & Eckert, C. G. Are Species’ Range Limits Simply Niche Limits Writ Large? A Review of Transplant Experiments beyond the Range. Am. Nat. 183, 157–173 (2014).
Paquette, A. & Hargreaves, A. L. Biotic interactions are more often important at species’ warm versus cool range edges. Ecol. Lett. 24, 2427–2438 (2021).
Buckley, L. B. & Jetz, W. Environmental and historical constraints on global patterns of amphibian richness. Proc. R. Soc. B Biol. Sci. 274, 1167–1173 (2007).
Deutsch, C., Penn, J. L. & Seibel, B. Metabolic trait diversity shapes marine biogeography. Nature 585, 557–562 (2020).
Darwin, C. On the origin of species by means of natural selection, or preservation of favoured races in the struggle for life. (John Murray, 1859).
MacArthur, R. H. Geographical Ecology: Patterns in the Distribution of Species. The Journal of Animal Ecology (Harper & Row, 1972). doi:10.2307/3393.
Schemske, D. W., Mittelbach, G. G., Cornell, H. V., Sobel, J. M. & Roy, K. Is There a Latitudinal Gradient in the Importance of Biotic Interactions? Annu. Rev. Ecol. Evol. Syst. 40, 245–269 (2009).
Peco, B., Laffan, S. W. & Moles, A. T. Global Patterns in Post-Dispersal Seed Removal by Invertebrates and Vertebrates. PLoS ONE 9, e91256 (2014).
Ashton, G. V. et al. Predator control of marine communities increases with temperature across 115 degrees of latitude. Science 376, 1215–1219 (2022).
Roslin, T. et al. Higher predation risk for insect prey at low latitudes and elevations. Science 356, 742–744 (2017).
Zvereva, E. L. & Kozlov, M. V. Latitudinal gradient in the intensity of biotic interactions in terrestrial ecosystems: Sources of variation and differences from the diversity gradient revealed by meta‐analysis. Ecol. Lett. 24, 2506–2520 (2021).
Hargreaves, A. L. et al. Seed predation increases from the Arctic to the Equator and from high to low elevations. Sci. Adv. 5, eaau4403 (2019).
Jablonski, D., Roy, K. & Valentine, J. W. Out of the Tropics: Evolutionary Dynamics of the Latitudinal Diversity Gradient. Science 314, 102–106 (2006).
Jablonski, D. et al. Out of the tropics, but how? Fossils, bridge species, and thermal ranges in the dynamics of the marine latitudinal diversity gradient. Proc. Natl. Acad. Sci. 110, 10487–10494 (2013).
Levins, R. Evolution in Changing Environments: Some Theoretical Explorations. (MPB-2). (Princeton University Press, 1968). doi:10.1515/9780691209418.
Levins, R. Theory of Fitness in a Heterogeneous Environment. I. The Fitness Set and Adaptive Function. Am. Nat. 96, 361–373 (1962).
Sunday, J. M., Bates, A. E. & Dulvy, N. K. Thermal tolerance and the global redistribution of animals. Nat. Clim. Change 2, 686–690 (2012).
Colwell, R. K. & Rangel, T. F. Hutchinson’s duality: The once and future niche. Proc. Natl. Acad. Sci. U. S. A. 106, (2009).
Pinsky, M. L., Eikeset, A. M., McCauley, D. J., Payne, J. L. & Sunday, J. M. Greater vulnerability to warming of marine versus terrestrial ectotherms. Nature 569, 108–111 (2019).
Estrada, A., Morales-Castilla, I., Meireles, C., Caplat, P. & Early, R. Equipped to cope with climate change: traits associated with range filling across European taxa. Ecography 41, 770–781 (2018).
Teets, N. M. & Denlinger, D. L. Physiological mechanisms of seasonal and rapid cold-hardening in insects: Seasonal and rapid cold-hardening in insects. Physiol. Entomol. 38, 105–116 (2013).
Huey, R. B., Ma, L., Levy, O. & Kearney, M. R. Three questions about the eco-physiology of overwintering underground. Ecol. Lett. 24, (2021).
Dawson, M. N. & Hamner, W. M. A biophysical perspective on dispersal and the geography of evolution in marine and terrestrial systems. J. R. Soc. Interface 5, 135–150 (2008).
Sotka, E. E. Natural Selection, Larval Dispersal, and the Geography of Phenotype in the Sea. Integr. Comp. Biol. 52, 538–545 (2012).
Webb, T. J., Dulvy, N. K., Jennings, S. & Polunin, N. V. C. The birds and the seas: body size reconciles differences in the abundance-occupancy relationship across marine and terrestrial vertebrates. Oikos 120, 537–549 (2011).
Martin, T. L. & Huey, R. B. Why “Suboptimal” Is Optimal: Jensen’s Inequality and Ectotherm Thermal Preferences. Am. Nat. 171, E102–E118 (2008).
Huey, R. B. & Kingsolver, J. G. Evolution of thermal sensitivity of ectotherm performance. Trends Ecol. Evol. 4, 131–135 (1989).
Bernhardt, J. R., Sunday, J. M., Thompson, P. L. & O’Connor, M. I. Nonlinear averaging of thermal experience predicts population growth rates in a thermally variable environment. Proc. R. Soc. B Biol. Sci. 285, 20181076 (2018).
Dahlke, F. T., Wohlrab, S., Butzin, M. & Pörtner, H.-O. Thermal bottlenecks in the life cycle define climate vulnerability of fish. Science 369, 65–70 (2020).
Schiffer, M. et al. Temperature tolerance of different larval stages of the spider crab Hyas araneus exposed to elevated seawater PCO2. Front. Zool. 11, 87 (2014).
Soberón, J. Grinnellian and Eltonian niches and geographic distributions of species. Ecol. Lett. 10, 1115–1123 (2007).
Sunday, J. M. et al. Thermal-safety margins and the necessity of thermoregulatory behavior across latitude and elevation. Proc. Natl. Acad. Sci. 111, 5610–5615 (2014).
Bennett, J. M. et al. GlobTherm, a global database on thermal tolerances for aquatic and terrestrial organisms. Sci. Data 5, (2018).
Bennett, J. M. et al. The evolution of critical thermal limits of life on Earth. Nat. Commun. 12, 1198 (2021).
IUCN. The IUCN Redlist of Threatened Species. (2020).
Roll, U. et al. The global distribution of tetrapods reveals a need for targeted reptile conservation. Nat. Ecol. Evol. 1, 1677–1682 (2017).
GBIF. The Global Biodiversity Information Facility. (2020).
Suggitt, A. J. et al. Conducting robust ecological analyses with climate data. Oikos 126, 1533–1541 (2017).
Kearney, M. R. & Porter, W. P. NicheMapR – an R package for biophysical modelling: the microclimate model. Ecography 40, 664–674 (2017).
Holt, B. G. et al. An Update of Wallace’s Zoogeographic Regions of the World. Science 339, 74–78 (2013).
Pinheiro, J., Bates, D., & R-core. nlme: Linear and Nonlinear Mixed Effects Models. (2022).
Zuur, A. F., Ieno, E. N. & Elphick, C. S. A protocol for data exploration to avoid common statistical problems: Data exploration. Methods Ecol. Evol. 1, 3–14 (2010).
Bartoń, K. MuMIn: Multi-Model Inference. (2020).
Banner, K. M. & Higgs, M. D. Considerations for assessing model averaging of regression coefficients. Ecol. Appl. 27, 78–93 (2017).
Rohr, J. R. et al. The complex drivers of thermal acclimation and breadth in ectotherms. Ecol. Lett. 21, 1425–1439 (2018).
Comte, L. & Olden, J. D. Climatic vulnerability of the world’s freshwater and marine fishes. Nat. Clim. Change 7, 718–722 (2017).
Gunderson, A. R. & Stillman, J. H. Plasticity in thermal tolerance has limited potential to buffer ectotherms from global warming. Proc. R. Soc. B Biol. Sci. 282, 20150401 (2015).
Morley, S. A., Peck, L. S., Sunday, J. M., Heiser, S. & Bates, A. E. Physiological acclimation and persistence of ectothermic species under extreme heat events. Glob. Ecol. Biogeogr. 28, 1018–1037 (2019).
Light, P., Dawson, W. R., Shoemaker, V. H. & Main, A. R. Observations on the Thermal Relations of Western Australian Lizards. Copeia 1966, 97 (1966).

There is NO Competing Interest.

rangethermsupplementarymaterialsnatureecoevo.docx
Supplementary Material for Global patterns of thermal niche filling in ectotherms

Download PDF

Journal Publication

published 06 Nov, 2023

Read the published version in Nature Ecology & Evolution →

Version 1

posted

You are reading this latest preprint version

Global patterns of thermal niche filling in ectotherms

Status:

Journal Publication

Version 1

Abstract

Figures

One-Sentence Summary

Main

Results And Discussion

Methods

Fundamental thermal niches

Realized ranges and traits

Estimating body temperatures

Potential and realized thermal niches

Measuring thermal niche filling

Analyses

Acclimatisation

Behavioural thermoregulation

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1