Evidence for climatic niche and biome shifts between native and novel ranges in plant species introduced to Australia

Authors


Correspondence author. E-mail: rachael.gallagher@mq.edu.au

Summary

1. The potential invasive success of exotic plant species is thought to be associated with similarity in climate and biome between the original and novel range. We tested this assumption by quantifying the match between the realized climatic niches and biomes occupied in the exotic and native range of 26 plant species introduced to Australia. We then explored correlations between the propensity to shift climatic niche with residence time, invasion status, geographic range size, and species traits.

2. Occurrence data from the native and exotic range of 26 species introduced to Australia were obtained, and the overlap between native and exotic climate niches was calculated using between-class analysis. Shifts into novel biomes were assessed using a Geographic Information System (GIS). Correlations between introduction, distribution and species traits and the degree of climate matching were examined using nonparametric statistical tests.

3. Exotic species frequently occurred in climatic conditions outside those occupied in their native range (20 of 26 species). Nineteen species inhabited biomes in Australia not occupied in the native range and in some instances this shift represented the establishment of populations in novel biomes not present in the native range. No single-species trait, introduction or distributional characteristic was significantly associated with the degree of climatic niche shift.

4.Synthesis. Exotic species are able to occupy climate niches in the new range that differ substantially from those of the native range, and generally do not show biome conservatism between their native and introduced ranges. This implies that novel climatic conditions are not a major obstacle for exotic species establishing populations outside their native range. These results have important implications for the use and interpretation of ecological niche models used to predict the distribution of species in novel climates in time or space. The results also highlight the importance of alternate mechanisms, such as enemy release, phenotypic plasticity or rapid evolution, in the establishment of naturalized and invasive populations.

Introduction

Understanding the factors that lead to the successful establishment of organisms outside their native range is a fundamental, but complex, goal of invasion biologists. Although only a fraction of introduced species make the transition to become naturalized and even fewer become invasive, those species capable of successful integration into new habitats can have devastating effects on native ecosystems. Invasive plant species have been linked to reductions in native species richness (Greenwood, O’Dowd & Lake 2004; Lake & Leishman 2004; Fridley et al. 2007), changes in soil nutrient dynamics (Mack, D’Antonio & Ley 2001; Allison & Vitousek 2004; Yelenik, Stock & Richardson 2004; Ashton et al. 2005), declines in limiting resources such as light (Asner et al. 2008; Iponga, Milton & Richardson 2008), and global and regional biotic homogenization (Qian & Ricklefs 2006; Schwartz, Thorne & Viers 2006; La Sorte, McKinney & Pysek 2007). Hence, invasive plants are recognized as one of the most pervasive agents of global change, and escalating rates of introduction, linked to the expansion of global trade, suggest they will continue to pose conservation challenges in the future.

The invasion process is characterized by three distinct, but not discrete, stages: introduction, naturalization and spread (Richardson et al. 2000). Naturalization is defined as the ability to establish self-sustaining populations following introduction into a new range, whereas invasion is only achieved by a subset of naturalized species that spread away from founding populations to become widespread and abundant (Pyšek et al. 2008). Species progress through these stages by overcoming a series of abiotic and biotic barriers, and the suitability of the climate in the introduced range is cited as one of the main abiotic barriers to the establishment of exotic plants (Williamson 2006; Hayes & Barry 2008).

The term ‘climate matching’ refers to the match between native and exotic regions based on either a single variable such as temperature (Chown, Gremmen & Gaston 1998), a suite of climatic variables (Thuiller et al. 2005; Richardson & Thuiller 2007), or on indirect measures of climate such as latitude (Pyšek 1998; Blackburn & Duncan 2001; Maron 2006; Jimenez et al. 2008). This technique has been applied to a range of issues in invasive species management. These include the identification of high-risk source areas (Richardson & Thuiller 2007), pre-emptive projection of the spatial extent and magnitude of risk posed by introductions both now (Panetta & Mitchell 1991; Thuiller et al. 2005; Ficetola, Thuiller & Miaud 2007; Mgidi et al. 2007; Crossman & Bass 2008) and under future climates (Broennimann & Guisan 2008; Beaumont et al. 2009), prioritization of search areas for potential biocontrol agents (Senaratne, Palmer & Sutherst 2006; Robertson, Kriticos & Zachariades 2008) and the identification of suitable areas for release of biological control agents across invasive species ranges (Rafter et al. 2008).

To be a useful predictor of invasion potential, climate matching relies on the conservation of climate niches across both space and time. While there are numerous examples of correspondence between the native climate of exotic species and their introduced range (see Wiens & Graham 2005), not all exotic species show high levels of climate matching (e.g. see Broennimann et al. 2007; Fitzpatrick et al. 2007; Loo, Mac Nally & Lake 2007; Beaumont et al. 2009) due to both ecological and evolutionary processes. Historic and geographic constraints may prevent a species from occupying the entire fundamental climatic niche within its native range. Introduction to a new environment may also result in changes to a realized niche as the species is ‘released’ from the biotic constraints imposed by enemies and competitors on its native range margin. Rapid evolutionary adjustments to novel environments may also contribute to the successful spread of invasive species (Broennimann et al. 2007; Lavergne & Molofsky 2007) and can occur within < 20 generations (Prentis et al. 2008). Results of studies exploring these issues have implications beyond invasion ecology, as they can inform us about the factors that constrain the fundamental niche of species. In this way, studies that compare exotic species in their native and novel ranges are analogous to large-scale transplant experiments.

Thus, the assumption that species will maintain the same climate niche across their exotic ranges as they do in their native range does not always hold true. Across multi-dimensional climate space, five patterns of realized native and exotic climate niches can be conceptualized, as illustrated in Fig. 1: (i) both niches may overlap almost completely; (ii) the exotic climate niche may form a subset of the native climate niche; (iii) the native climate niche may form a subset of the exotic climate niche; (iv) there may be a partial overlap between the two niches; or (v) the two niches may be completely disjunct.

Figure 1.

 Five possible relationships between the climate space realized by native (white) and exotic (grey) populations.

While climate matching assumes either near-complete overlap or that the exotic niche represents a subset of the native niche, equivocal evidence from single-species studies indicates that we cannot yet assess how common the ability to shift climatic niche is in species introduced to novel regions. This information is critical for quantifying the uncertainty inherent to climate matching exercises used in invasive species management, such as niche modelling. It also has wider implications for niche theory in its ability to delimit the bounds of the elusive fundamental niche. Using a comparative approach that examines a suite of species and their traits invading into a common location may further our understanding of the need to account for climatic niche shifts when applying climate matching techniques. In this study, we quantified the extent to which the current realized climatic range of naturalized or invasive exotic species overlaps with that of their native range, for 26 plant species introduced to Australia. We then tested hypotheses about the role of species traits, introduction history and attributes of the distribution of invasive species in determining the degree of climate matching between the native and exotic ranges.

Consistent with climate niche theory, invasive species are also generally expected to show strong biome conservatism between their native and exotic ranges (Rejmánek et al. 2005). Biomes are broad-scale ecological regions defined by vegetation structure and climatic conditions, in particular rainfall and temperature. They often closely approximate climatic zones and are used to divide the Earth into areas that share a similar functional or ecological role. Biome conservatism between species’ native and exotic ranges has previously been shown to be associated with invasion success across a variety of taxa (Peterson 2003; Thuiller et al. 2005), but the ability to shift into new biomes following introduction remains unexplored. Therefore, we also examined how often the 26 species used in this study have shifted into a novel biome not occupied in the native range, when introduced to Australia.

Materials and methods

Species selection

This study focused on exotic plant species in Australia. Few ecosystems in Australia have not been invaded (Adair & Groves 1998); since European colonization in 1788, over 29 000 plant species have been introduced with an estimated 3000 species becoming naturalized and around 300 of these becoming invasive (Randall et al. 2007). We collated data for 26 introduced plant species, representing 15 families and a range of functional types (Table 1).

Table 1.   Characteristics of 26 species introduced to Australia and between-class inertia ratios comparing the overlap between climate niches of the native range and the exotic Australian range. The P-value of 99 Monte Carlo randomizations of the between-class inertia ratios was <0.05 for all species
TaxonFamilyBetween-class inertia ratioSpecies traitsIntroduction traitsDistribution traits
Dispersal modeGrowth formLongevitySeed mass (g)Weed statusAustralian residence time (years)Province of native rangeExtent of occurrence (cells occupied in native range)
Thunbergia alata SimsAcanthaceae0.13UnassistedVinePerennial0.0258Invasive92Afrotropic48
Hedera helix L.Araliaceae0.026AssistedVinePerennial0.0236Naturalized109Palearctic3749
Cryptostegia grandiflora R.Br.Asclepiadaceae0.083AssistedVinePerennial0.00857Invasive67Afrotropic18
Gomphocarpus fruticosus (L.) R.Br.Asclepiadaceae0.29UnassistedShrubPerennial0.00738Naturalized206Afrotropic65
Chrysanthemoides monilifera (L.) Norl. subsp. moniliferaAsteraceae0.052AssistedShrubPerennial0.1012Invasive100Afrotropic67
Onopordum acanthium L.Asteraceae0.159AssistedHerbAnnual0.0083Naturalized109Palearctic1256
Parthenium hysterophorus L.Asteraceae0.035AssistedHerbAnnual0.0006Invasive44Neotropic174
Xanthium spinosum L.Asteraceae0.092AssistedHerbAnnual0.104Naturalized126Neotropic27
Macfadyena unguis-cati (L.) A.H. GentryBignoniaceae0.127AssistedVinePerennial Invasive109Neotropic212
Hypericum perforatum L.Clusiaceae0.078UnassistedHerbPerennial0.0002Naturalized151Palearctic2440
Erica lusitanica RudolphiEricaceae0.202AssistedHerbPerennial Invasive63Palearctic14
Euphorbia paralias L.Euphorbiaceae0.243AssistedHerbPerennial0.0056Naturalized81Palearctic229
Calopogonium mucunoides Desv.Fabaceae0.052UnassistedVinePerennial0.01137Naturalized56Neotropic260
Macroptilium lathyroides (L.)Urb.Fabaceae0.09UnassistedVineAnnual0.00703Naturalized97Neotropic123
Mimosa pigra L.Fabaceae0.028AssistedTreePerennial0.017Invasive117Neotropic415
Parkinsonia aculeata L.Fabaceae0.073UnassistedTreePerennial0.107Invasive100Ne arctic, Neotropic213
Senna obtusifolia (L.) H.S.Irwin & BarnebyFabaceae0.126UnassistedShrubAnnual0.025Naturalized47Neotropic226
Fumariamuralis Sond, ex W.D. J.KochFumariaceae0.383UnassistedVineAnnual0.00308Naturalized116Palearctic819
Passiflorafoeiida L.Passifloraceae0.084AssistedVinePerennial0.008Invasive74Ne arctic, Neotropic773
Eragrostis curvula (Schrad.) NeesPoaceae0.068AssistedGrassPerennial0.00025Invasive101Afrotropic32
Nassella neesiana (Trin. & Rupr.) BarkworthPoaceae0.1AssistedGrassPerennial Invasive74Neotropic23
Pennisetum polystachion (L.) Schult.Poaceae0.091AssistedGrassAnnual0.0004Naturalized84Afrotropic49
Sporobolus pyramidalis P.Beauv.Poaceae0.314AssistedGrassPerennial0.00016Invasive198Afrotropic45
Reseda luteola L.Resedaceae0.148UnassistedHerbAnnual0.0003Naturalized115Palearctic39
Rosa rubiginosa L.Rosaceae0.235AssistedShrubPerennial0.0152Naturalized109Palearctic685
Cardiospermum grandiflorum Sw.Sapindaceae0.172AssistedVinePerennial0.039Naturalized206Neotropic, Afrotropic77

We selected species according to the following criteria: (i) they are listed on a national priority list of introduced species in Australia (the shortlist of 71 Weeds of National Significance (WONS), or the Alert List for Environmental Weeds (http://www.weeds.gov.au/publications/guidelines/alert/index.html); (ii) they are a distinct taxonomic entity following the nomenclature of the Australian National Herbarium, Canberra; (iii) their native range and exotic Australian range can confidently be defined spatially through descriptive information and species occurrence records; and (iv) they are restricted to terrestrial habitats.

Distribution data

Occurrence records from the native and Australian ranges of each species were collated from the Global Biodiversity Information Facility (GBIF, http://www.gbif.org/), Australia’s Virtual Herbarium (AVH, http://www.ersa.edu.au/avh/public_query.jsp), the UK National Biodiversity Network (http://data.nbn.org.uk/) and the TROPICOS data base (http://www.tropicos.org/) in 2009. Those few records missing latitude and longitude data, but with sufficient information on the known location, were completed using the Alexandria Digital Library (http://www.alexandria.ucsb.edu). The native range of each species was determined through the use of online data bases, e.g. Hawaiian Ecosystems at Risk (HEAR, http://www.hear.org/), Germplasm Resources Information Network (GRIN, http://www.ars-grin.gov/npgs/aboutgrin.html), Global Invasive Species Database, http://www.issg.org/database/welcome/), and floras for each region of the native range. When delimiting the extent of the native range was problematic due to conflicting information, advice was sought from taxonomists at the National Herbarium of NSW in Sydney, Australia, who were familiar with the species. All species records were plotted and coded within ArcGIS (ESRI 2006) and records with clearly incorrect geo-codes were removed. A total of 20 029 location records across the 26 species were collated at a resolution of 10 arcminutes (duplicates were removed), representing an average of 464 (range: 14–3749) occurrences per species in the native range and 92 (range: 16–341) in the introduced Australian range.

Climate and biome data

Nineteen bioclimatic variables representing global climate data were obtained from Worldclim (version 1.4, Hijmans et al. 2005) at a resolution of 10 arcminutes. Values for each of the 19 bioclimatic variables were appended to species occurrence points. These bioclimatic variables represent both mean and extreme values of temperature and precipitation. The data set is derived from average monthly temperature and precipitation from a network of meteorological stations measured between 1950 and 2000. Worldclim data have been widely used in large-scale examinations of climatic constraints on species distributions because of its global coverage and accessibility (e.g. Broennimann et al. 2007; Steiner et al. 2008; Beaumont et al. 2009; Leachéet al. 2009; Puschendorf et al. 2009). See Hijmans et al. (2005) for a thorough explanation of the Worldclim data set.

Although it may be desirable to use more detailed climatic data in an analysis of this kind, such data are not widely available, especially for a comparative study across 26 species. It is important to note that we have approximated the climatic tolerances of each species using their realized climatic niche in their native range, but without experimental manipulation the full complement of a species climatic tolerance – or its fundamental climatic niche – cannot be determined.

A GIS layer of the terrestrial ecoregions of the world (see Olson et al. 2001) was downloaded (http://www.worldwildlife.org/science/data/item6373.html) and the biome in which each occurrence point fell was assessed using the spatial join function in ArcGIS (ESRI 2006). Biomes occupied in the native range were compared to those occupied in the exotic Australian range and novel biomes were identified.

We used the same technique to compare the native and Australian distributions with regard to the five major Köppen climate zones, namely Tropical, Arid, Temperate, Boreal and Cold Snow, based on Peel, Finlayson & McMahon (2007).

Trait data

We collated information on a number of introduction, distribution and life-history traits (hereafter collectively referred to as traits) for each species and tested for associations with the degree of climatic niche overlap, as measured using the between-class inertia ratio (see Calculation of niche shifts and analysis of species traits). The traits examined were:

  • 1Minimum residence time in Australia. This was taken from two sources: (i) published literature and (ii) the date of the first herbarium specimen collected for the species in Australia. For most introduced species, data on the exact time of introduction are scarce. However, Hamilton et al. (2005) showed a strong correlation between known times of introduction and the date of the first herbarium collection in Australia. Therefore, wherever possible, the exact year of the introduction of a species was recorded from historical records, but in cases where this information was not available, we used the date of the first herbarium collection as a surrogate. Minimum residence time ranged between 19 and 206 years (Table 1).
  • 2Biogeographic province of the species’ native range (i.e. Afrotropic, Nearctic, Neotropic, Palearctic). The classification presented in Udvardy (1975) was used to assign the native range of each species to a province.
  • 3The extent of occurrence of the species in its native range represented by the number of 10-arcminute grid cells occupied. Within ArcGIS we created a grid with resolution of 10 arcminutes and used the Count Points in Polygons feature in the Hawth’s Tools Extension (Beyer 2004) to assess the number of polygons.
  • 4Weed status in Australia. We used a comprehensive national data base of introduced species to classify the species as either naturalized or invasive (Randall et al. 2007; available at http://weeds.cbit.uq.edu.au/). This data base lists 29 430 introduced species and provides information on the weed status of each plant, compiled from literature and weed checklists. Of the 26 species examined, 14 were classified as naturalized and the remaining 12 as invasive (Table 1). The method of classifying species as naturalized and invasive species has been applied previously to questions in invasion ecology (see Diez et al. 2009) and represents our best current understanding of the introduction status of exotic species in Australia.
  • 5Physical traits including seed mass (g), dispersal mode (two classes: assisted or unassisted), longevity (two classes: annual or perennial) and growth form (five classes: grasses, herbs, vines, shrubs and trees). Species possessing an assisted dispersal mode were those with seeds spread via wind, water or animals, based on morphological characteristics of the seed. Data on each trait were taken from a multipurpose data base of species traits compiled within the research group from published literature, field observations and the World Wide Web. Whilst the 26 species chosen cover a wide taxonomic breadth (15 families, 26 genera), shared ancestry amongst species may influence our ability to generalize findings related to species traits to other invasive taxa.

Calculation of realized niche shifts and analysis of species traits

Following Broennimann et al. (2007), we used Principal Component Analysis (PCA) in the R CRAN library ‘ade4’ to quantify the positions of native and Australian exotic climatic niches in a multi-dimensional space. We reduced the 19 variables to the first two axes of the PCA and then used these to characterize the occurrence clouds for each species. The magnitude and statistical significance of differences between the native and invasive occurrence clouds were assessed using the between-class inertia ratio for each species. The significance of this ratio was tested by conducting 99 Monte Carlo randomizations. It is important to note that similarity in the between-class inertia of species does not indicate that they exhibit similar patterns as outlined in Fig. 1.

Nonparametric statistical analyses testing the association between traits and the degree of climate matching on each individual trait were undertaken in r (http://cran.r-project.org/). We used Spearman’s rank correlation coefficient for continuous data (seed mass, extent of occurrence and residence time), Wilcoxon signed-rank tests on dichotomous categorical variables (weed status, longevity and dispersal) and Kruskall–Wallis one-way analysis of variance on categorical variables with more than two levels (growth form and biogeographical province of the native range).

Results

Shifts in the realized climatic niche

Examination of the PCA bi-plots showed that the centroids of native and exotic Australian climatic niches differed for all species, with between-class inertia ratios ranging from 0.03 (substantial overlap, Hedera helix) to 0.38 (little overlap, Fumaria muralis) (Table 1). These differences were confirmed by the results of 99 Monte Carlo randomizations in which all native and exotic Australian populations were shown to occupy significantly different climatic space (= 0.05).

We visually assessed the convex hulls encompassing the occurrence clouds in the PCA bi-plots and categorized species subjectively into one of the five categories illustrated in Fig. 1 (see Fig. S1 in Supporting Information for PCA bi-plots for each species). Partial overlap of the climatic space occupied (Fig. 1d) was identified as the most common pattern (18 out of 26 species) (Table 2). The area of overlap was quite small for some species (e.g. Hypericum perforatum and Onopordum acanthium; Fig. 2a,b). In the case of F. muralis (Fig. 2c), the small area of overlap resulted from a few outlying native populations: with the exception of these populations, the climatic space occupied by native and exotic Australian populations of this species was entirely different.

Table 2.   Pattern of climatic niche overlap and biome shifts between the native and Australian ranges of 26 species. The asterisks in the final column denote that the species did not inhabit any novel biomes in the exotic Australian range. Biome codes are as follows: 1, Tropical and Subtropical Moist Broadleaf Forests; 2, Tropical and Subtropical Dry Broadleaf Forests; 3, Tropical and Subtropical Coniferous Forests; 4, Temperate Broadleaf and Mixed Forests; 5, Temperate Conifer Forests; 6, Boreal Forests/Taiga; 7, Tropical and Subtropical Grasslands, Savannas and Shrublands; 8, Temperate Grasslands, Savannas and Shrublands; 9, Flooded Grasslands and Savannas; 10, Montane Grasslands and Shrublands; 11, Tundra; 12, Mediterranean Forests, Woodlands and Scrub; 13, Deserts and Xeric Shrublands; 14, Mangroves
SpeciesFamilyClimate niche shiftsBiome shifts
Climate niche overlap patternBiomes in native rangeNovel biomes occupied in exotic Australian range
Thunbergia alataAcanthaceaePartial1,7,10,144,12
Hedera helixAraliaceaePartial4,5,6,11,12*
Cryptostegia grandifloraAsclepiadaceaePartial1,2,137
Gomphocarpm fruticosusAsclepiadaceaePartial1,2,7,9,10,13,144,8,12
Chrysanthemoides monilifera subsp. moniliferaAsteraceaeNative subset of exotic1,10,12,134,8
Onopordum acanthiumAsteraceaePartial4,5,6,128,10,13
Parthenium hysterophorusAsteraceaeExotic subset of native1,2,3,7,8,10,13,144
Xanthium spinosumAsteraceaePartial1,2,7,8,9,10,134,12
Macfadyena unguis-catiBignoniaceaeExotic subset of native1,2,3,7,9,13,144,8
Hypericum perforationClusiaceaePartial4,5,6,8,10,11,12,13*
Erica lusitanicaEricaceaePartial4,12*
Euphorbia paraliasEuphorbiaceaePartial4,1213
Calopogonium mucunoidesFabaceaeExotic subset of native1,2,3,7,9,10,13,14*
Macroptilium lathyroidesFabaceaeExotic subset of native1,2,3,7,9,10,13,144
Mimosa pigraFabaceaeExotic subset of native1,2,3,7,9,10,13,14*
Parkinsonia aculeataFabaceaeExotic subset of native1,2,3,4,5,6,7,8,9,10,12,13,14*
Senna obtusifoliaFabaceaePartial1,2,3,7,9,10,13,14*
Fumaria muralisFumariaceaePartial4,5,6,11,128,13
PassiflorafoetidaPassifloraceaePartial1,2,3,7,8,9,10,13,144
Eragrostis curvulaPoaceaePartial1,7,10,12,134,8
Nassella neesianaPoaceaePartial1,7,8,10,124
Pennisetum polystachionPoaceaePartial1,7,10,1413
Sporobolus pyramidalisPoaceaePartial1,7,10,144,13
Reseda luteolaResedaceaeNative subset of exotic4,5,6,128,10,13
Rosa rubiginosaRosaceaePartial4,5,6,11,128,10
Cardiospermum grandiflorumSapindaceaePartial1,2,3,7,9,13,144,12
Figure 2.

 Comparisons of the climatic niches of five exotic species in Australia (red = native range, blue = Australian range), based on a PCA conducted on 19 bioclimatic variables. A between-class analysis tested the magnitude and significance of the occurrence clouds, yielding a between-class inertia ratio which was further tested using 99 Monte Carlo randomizations. These graphs represent the range of patterns of native and exotic climate niches found across 26 species naturalized or invasive in Australia. The minimum convex polygons that encompass all the occurrence points for each species were used to classify species into the five groups presented in Fig. 1. Partial overlaps between the two realized niches are shown in (a), (b) and (c). Although a partial overlap occurs for Fumaria muralis, this is due to two native outlier populations. In (d), the realized climate space of exotic populations of Macfadyena unguis-cati lies within the bounds of that of native populations, while the opposite occurs in (e) for Chrysanthemoides monilifera subsp. monilifera.

The second most common pattern exhibited across species was where the climatic space occupied by exotic Australian populations was a subset of conditions in the native range (Fig. 1b) (6 out of 26 species, e.g. Macfadyena unguis-cati, Fig. 2d). The final pattern identified was where the climatic space of native populations was within that of exotic Australian populations (Fig. 1c) (2 out of 26 species, e.g. Chrysanthemoides monilifera subsp. monilifera, Fig. 2e). No species were identified as having native and exotic ranges nearly completely overlapping (Fig. 1a) or completely disjunctive (Fig. 1e).

Shifts to novel biomes and Köppen climate zones

Of the 26 species examined, 19 have shifted into novel biomes not present in the native range since being introduced to Australia (Table 2). The most common shift exhibited across species was establishment in Temperate Broadleaf and Mixed Forests (12 of the 19 shifts), followed by shifts into Temperate Grasslands, Savannas and Shrublands (8 of 19 shifts). No species examined in this study had established populations within a novel Köppen climate zone upon introduction to Australia.

Association between traits and degree of climatic niche shift

We tested whether the degree of climatic niche overlap was associated with certain physical attributes of the species in our data set. For instance, we proposed that smaller seed mass or assisted dispersal would affect the ability of species to colonize areas with suitable climates by increasing their potential to spread through the landscape. Similarly we tested the idea that annual plants, with their shorter generation times, will adapt more rapidly to novel conditions than perennial species, resulting in smaller between-class inertia ratios. These hypotheses were not supported by our analyses; none of the traits were found to be significantly correlated with climatic niche overlap: seed mass (= 0.27), dispersal (= 0.67), longevity (= 0.64), and growth form (= 0.87).

We also hypothesized that the amount of time a species has been present in the exotic range may influence its ability to expand into suitable climatic conditions, resulting in a negative relationship between residence time and the between-class inertia ratio. This idea was not supported by our analyses (= 0.32). Likewise, invasive species did not exhibit a greater degree of similarity in their climate niche between their native and exotic range than naturalized species (= 0.18).

The final tests of association were made between characteristics of the distribution of the species, namely the biogeographic province of the native range and extent of occurrence in the native range, and the between-class inertia ratio. First, we asked if the biogeographic province that the species occupies in its native range, be it Afrotropic, Nearctic, Neotropic or Palearctic, explains a significant amount of the variation in climatic niche overlap. This idea was not supported by our data (= 0.22). Secondly, we tested the hypothesis that there is a positive relationship between is the size of a species’ native range and the degree of climatic niche overlap; this hypothesis was also not supported (= 0.33).

Discussion

The ability of exotic species to occupy novel climatic space and biomes

This study has shown clearly that exotic species have a broad capacity to occupy climatic spaces in novel ranges that differ from those of native populations. Among 26 species exotic to Australia, 20 have extended their exotic climate ranges into ‘new’ space unoccupied by native populations. The between-class inertia ratio (which measures the similarity in the native and exotic climatic niche space) was significantly different for all 26 species, although for some species the exotic populations occurred within the climatic niche space of native populations. For two species, Reseda luteola and C. monilifera subsp. monilifera, the climatic niche of native populations was contained within that of exotic populations.

Whether these shifts represent the occupation of climatic space limited in the native range due to historic or biotic factors, and/or an extension of the fundamental niche, remains to be determined. In fact, the stabilized climatic niche in the invaded range may more accurately reflect the climatic tolerances of a species. Addressing this question across multiple species will be the logical future extension of the research presented here. However, this will require a different approach that tests for differences in the climatic tolerances of species in their native and exotic ranges in a manipulative framework.

At present, single-species studies provide some evidence that the availability of climatic space in the native range restricts the climatic niche. For example, the native range of a mudsnail, Potamopyrgus antipodarum J.E. Gray, 1843, which is invasive in Australia and North America, is limited to New Zealand, and the lack of geographic constraints on larger land masses in its invaded range has enabled the snail to inhabit a larger climatic range in Australia and North America (Loo, Mac Nally & Lake 2007). However, some species, such as the Red Imported Fire Ant (Solenopsis invicta), are reported to be extending their niche through release from competition. The absence of infection by bacteria of the genus Wolbachia and reduced competition may have benefited the fire ant in North America, enabling it to expand into climatic zones not occupied within its native range (Fitzpatrick et al. 2007).

It is worth noting that comparisons made of the climatic space occupied by species in their native range and in their exotic range outside of Australia are consistent with the results presented in this study (R. Gallagher, unpublished data). For brevity, we have not formally presented the results of these analyses here; however, the PCA bi-plots show greater discrepancies between the climatic space of native and exotic ranges when exotic distributions outside Australia are also considered than when compared to Australian populations alone. For instance, nine of the 26 species in our study shifted into novel Köppen climate zones when the global exotic distribution was examined. It is likely, however, that not all regions containing exotic populations from outside Australia were included in this global analysis and therefore these results should be interpreted with caution. Failing to include data for all of these exotic populations means that the degree of overlap in climatic space between native and exotic ranges and the extent to which exotic ranges occupy novel climatic space may be underestimated.

Species in this study were also shown to have consistently established populations in different biomes within Australia than those occupied in their native range. In some instances species occurred in novel biomes that are not available in the native range. For example, Rubber Vine (C. grandiflora) occurs in the biome Tropical and Subtropical Grasslands, Savannas and Shrublands in its introduced Australian range but not in its native range of Madagascar, as this biome does not occur on the island. A lack of biome conservatism in exotic species may indicate that factors other than physiological tolerances for certain climatic conditions, such as competitive interactions between species or geographic barriers, limit the ability of species to occupy certain biomes. For example, Cotton Thistle (O. acanthium) has occupied the Deserts and Xeric Shrublands biome in Australia, but the Mediterranean Sea may have prevented this species extending its native range into this biome in northern Africa from Europe.

The establishment of species in novel biomes upon introduction to Australia is surprising. Recent work by Crisp et al. (2009) on phylogenetic biome conservatism indicates that the degree of phylogenetic biome conservatism exhibited by species is remarkably high, with some 96% of species examined across 11 000 taxa exhibiting biome stasis during speciation (Crisp et al. 2009). These authors concluded that pre-adaptation to conditions plays a large part in ensuring colonization success in novel regions. We propose that it is the frequency of introduction of propagules to the new biome in the exotic range that facilitates their establishment. Repeated and intentional introductions of exotic species, together with considerable human husbandry in some cases, increase the chance that propagules will encounter a suitable ‘window’ of resources or conditions either in time or space; this mechanism has been identified as a key driver in species invasions (Sakai et al. 2001).

In support of this idea, we found that species in our study most commonly shifted into the Temperate Broadleaf and Mixed Forest biome. This is unlikely to be due to the areal extent of this biome, which represents only 7% of the land mass of Australia. A more likely explanation is that human population density (and hence location of exotic species introductions) is highest on the east coast and in southern areas of Australia which are dominated by this particular biome.

What factors drive shifts to novel climates and biomes in exotic species?

We found no statistically significant relationships between the degree of climate matching between native and introduced ranges and a variety of traits including residence time, extent of occurrence, biogeographic province of the native range, weed status, growth form, longevity, seed mass and dispersal mode. Hence, no single trait emerged as a useful predictor of the propensity to shift to novel climate. These results indicate that these traits offer little a priori predictive power about which species may undergo either realized or fundamental climatic niche shifts when introduced to a novel environment.

Despite strong evidence in the literature for a positive association between the residence time in a novel range and the likelihood of a species becoming invasive (see Pyšek, Richardson & Williamson 2004; Castro et al. 2005; Pyšek & Jarošík 2005; Richardson & Pyšek 2006; Wilson et al. 2007), we found no evidence that residence time was a useful predictor for determining the degree of climate matching between the native and exotic range. This result may indicate that release from natural enemies and co-evolved competitors present in the native range may allow species to expand their range margins.

Factors other than climate, such as novel biotic interactions between species, may play a more important role in shaping the outcomes of species invasions. The lack of association of climatic niche shifts with species traits may also point to the importance of genetic factors or phenotypic plasticity in determining the limits of the fundamental niche. There are several genetic factors that may influence the likelihood of an exotic species extending its climate niche that may be considered in future studies. For example, while the loss of genetic variation due to founder effects is generally expected to restrict the potential for species to evolve, recent research has indicated that genetic bottlenecks do not necessarily limit the capacity of invasive species to rapidly evolve adaptations in key life-history traits such as the number of flowering branches and development of latitudinal clines in flowering time (Dlugosch & Parker 2008; van Kleunen & Fischer 2008). The relative importance of founder effects and phenotypic plasticity in affecting the ability to shift climatic niche may hold important insights for invasion biology and niche theory.

Implications for invasive species predictive schemes

Species distribution models (SDMs) are commonly used to assess the potential threat of exotic species to novel environments and rely on the assumption that the climatic niche is conserved. Our results, along with those of recent studies (e.g. Fitzpatrick et al. 2007; Steiner et al. 2008; Beaumont et al. 2009), clearly suggest that the climate space occupied by exotic populations is not static in space or time. The climate matching procedure is clearly very useful as an initial conservative investigation of the potential threat of an exotic species. However, for exotic species that have already established, the application of SDMs needs to be viewed as iterative: as populations spread, models need to be rerun to identify new threatened areas. Importantly, the ultimate location and extent of a potential invasive range may be underestimated because climate shifts have not been taken into consideration. An alternative approach to modelling the distribution of invasive species is to incorporate empirical data from manipulative experiments of a species’ climatic tolerance in a mechanistic model. This approach more closely approximates the fundamental climatic niche of the invader which may be more informative in species with the potential to undergo climatic niche shifts.

Acknowledgements

R.V.G. was supported by a Macquarie University Research Excellence Scholarship and L.J.B. by an Australian Research Council Postdoctoral Fellowship and Discovery Project (DP0877979). The Australian Research Council also provided funding to M.R.L. and L.H. for this study through the ARC Linkage scheme (LP0776758). We thank Mark Hamilton for providing minimum residence time data for exotic species within Australia and Andres Roubicek for his assistance with obtaining climatic data used in this study. John Wilson provided helpful feedback on earlier versions and comments from three anonymous referees helped to improve the manuscript.

Ancillary