Do naturalizations of endemic plant species indicate broader climatic tolerance than what is expected by their native ranges?

Tierney Bocsi1, Bethany Bradley1, Jesse Bellemare2

1Department of Environmental Conservation, UMass Amherst; 2Department of Biological Sciences, Smith College

IntroductionBiogeographic models of potential species distribution typically assume that locations wherespecies exist reflect their climatic tolerance. If so, then naturalized locations where speciespersist outside of their native range should reflect the same tolerance limits. Plants used forhorticultural purposes commonly have naturalized populations (Figure 1). Understanding howwell distribution patterns predict physiological tolerance limits is critical for our interpretationof climatic suitability models and projections of range shifts with climate change. In this study,we explored the differences between climatic conditions in the native and naturalized rangesof 61 woody plant species endemic to the southeastern United States. Woody species arelikely to be limited by dispersal (i.e. they were unable to keep up with warming temperaturesfollowing the last glaciation) rather than climatic conditions within their native range(Svenning & Skov, 2007). Because of this, we hypothesized that species would exhibit broaderclimatic tolerance, particularly for temperature, in their naturalized ranges compared to theirnative ranges.

Methods Downloaded occurrence data from the Global Biodiversity Information Facility (GBIF) for

61 woody plant species endemic to the southeastern United States1

Created point shapefiles based on latitude and longitude and clipped by U.S. boundaries Used county-level range maps2 to classify points as “native” or “naturalized” (Figure 1) Extracted climate data (annual precipitation, annual maximum temperature, annual

minimum temperature, January minimum temperature, and July maximum temperature)3

to native and naturalized points and merged the 61 species for analysis Generated histograms to compare climatic conditions for native and naturalized points Identified spatial extents of naturalized populations for the subset of 15 known naturalized

species (according to BONAP) and counted the number of naturalized species by county1Derived from list of 73 species compiled by the Bellemare lab2Digitized by the Bellemare lab from the Biota of North America Program (BONAP)3Downloaded from PRISM (wwww.prism.oregonstate.edu)

Figure 1. Example species, Aesculus parviflora, with both native(blue) and naturalized (red) point locations. The counties in which itis native are highlighted (light blue).

Results and Discussion

Figure 2. Annual precipitation values in millimeters extracted to thenative (blue) and naturalized (red) point locations for all 61 species.These values indicate that naturalized locations exhibit a broadertolerance for less precipitation than native locations.

Figure 3. Minimum temperature values for January in Celsius (C)extracted to the native (blue) and naturalized (red) point locations forall 61 species. These values indicate that native and naturalizedlocations exhibit similar tolerance for minimum temperatures.

Figure 4. Maximum temperature values for July in Celsius (C)extracted to the native (blue) and naturalized (red) point locationsfor all 61 species. These values indicate that native and naturalizedlocations exhibit similar tolerance for maximum temperatures.

Figure 5. Counties classified by the number of naturalized species they contain.Images depict three of the most common naturalized species in New England. Forthe subset of 15 species known to be naturalized outside their native range, speciesare most likely to be introduced to New England.

Leucothoe fontanesiana

Pieris floribunda

Rhododendron minus

Of the 61 species, we found that 51 had occurrences outside their native county ranges.These totaled to 221 naturalized occurrence points. Because we used county borders todelineate native range boundaries, species occurrences that were located just outside oftheir native counties (and would likely be considered native occurrences) were classified asnaturalized for consistency purposes. There were 57 species with 685 occurrence pointsinside of their native ranges. The combined 906 occurrences were used to compare climatedata for the naturalized and native occurrences of all 61 species. We expected to see similarthresholds in precipitation, and we assumed that the naturalized occurrences woulddemonstrate tolerance for colder temperatures than the native occurrences. Contrary toour hypothesis, naturalized occurrences revealed greater tolerance for less precipitationthan did native occurrences (Figure 2). Furthermore, minimum and maximum temperaturethresholds were consistent between both (Figure 3 and Figure 4, respectively).

Using the subset of 15 species that are known to be naturalized, we examined native andnaturalized locations. We found that on average, naturalized occurrences wereapproximately 500 kilometers from the nearest native county. This value is significantbecause it suggests that species are not simply dispersing, but being introduced withanthropogenic (presumably ornamental) intention. Most of these naturalized species arelikely to be found in New England, where conditions differ distinctly from native ranges inthe southeast (Figure 5). Similar to findings by Van der Veken et al. (2008), this observationindicates that species such as these may be able to persist in the face of climate change, asnorthward range expansion is facilitated by humans.

Aesculus parviflora

Counties with Naturalized Species