The relative abundances of native and non-native emydid turtles across an urban to rural habitat gradient in Central New Jersey.
KEY WORDS: Trachemys scripta, Pseudemys rubriventris, invasive species, Herpetology
Slider turtles, Trachemys scripta, are aquatic emydids native to the Mississippi River Basin, Texas, northern Florida, and the southeastern Atlantic Coastal Plain of the United States (Conant and Collins, 1998). Red-eared sliders, T. s. elegans, have been introduced to many countries all over the world via the pet trade and are thought to negatively impact freshwater ecosystems. They can pose a serious threat to native turtle species by competing (Cadi and July, 2004) or hybridizing with them (Seidel, 2002). Even though they are native to the United States, red-eared and yellow-bellied sliders, T. s. scripta, also have many localized non-native ranges within the country (Ernst et al., 2009). Both subspecies can be found throughout New Jersey (Schwartz and Golden, 2002), but T. s. elegans is more prevalent than T. s. scripta (personal observation). Non-native sliders are thought to be more common in urban or suburban locations in proximity to initial release sites. In Europe, T. s. elegans is typically found inhabiting waterbodies in, or close to, developed areas (Bringsoe, 2006).
The impacts that extralimital populations of sliders have on ecosystems in the U.S. are not well known. Therefore, it is difficult to classify New Jersey sliders as either potentially harmful invasives or benign non-natives. In order to determine their status, long-term studies investigating the possible negative impacts sliders have in New Jersey are necessary. If sliders are found to compete or hybridize with native turtles, limit other native species, or spread throughout New Jersey, then they could be seen as invasive. However, if they stay restricted to areas in the vicinity of their initial release, without significantly limiting populations of native organisms, they would be referred to as non-native. Determining relative abundances of sliders and native emydids is a proper baseline study that will lend insight and direction to future studies.
Sliders should be more common in urban environments that are already devoid of most native emydid turtles and less common in rural settings that have high numbers of native emydids. If sliders are eventually found to impact native turtles in some way, the impact they have will probably not be severe. Many native New Jersey species occur sympatrically with sliders in other parts of their natural ranges without being seriously impacted by them. Additionally, a few species occupy habitats that are inconducive to long-term slider turtle survival. Others share habitats with multiple turtle species and might be more adaptable to a non-native presence, a distinction that has not been afforded to European species impacted by slider invasion. Most European locales possess only a single species of native turtle: the European pond turtle, Emys orbicularis. The isolation of this species for a good portion of its evolutionary history could explain why they have a limited ability to compete with introduced sliders (Cadi and Joly, 2004).
Hybridization has been documented in captive turtles and is also known to occur in declining wild populations. Unlike most "traditional" hybrids in the animal kingdom, hybrid turtles are often fertile and capable of breeding with both parent species. Hybridization can be viewed as a threat to biodiversity, because it produces phenotypically intermediate offspring, thus blurring the distinction between separate evolutionary lineages (Blanck, 2006). Finding hybrids in healthy populations of native New Jersey turtles is unlikely. However, hybridization could possibly occur in habitats that have low numbers of turtles between sliders and a closely related species possessing a similar niche such as the northern red-bellied turtle, Pseudemys rubriventris. The possible interaction between T. s. elegans and P. rubriventris is of special interest to this study because of their similar life histories and naturally allopatric distributions (both species were geographically isolated prior to the human-mediated introduction of T. s. elegans to the Northeast).
This study intends to provide an initial assessment of the possible impact slider turtles exert on native turtles by comparing their relative abundances across a development gradient. The study also attempts to document hybrid individuals (if they exist) in habitats supporting native and non-native species.
[FIGURE 1 OMITTED]
MATERIALS AND METHODS
Data were collected from June-August of 2010 at multiple sites across New Jersey (Fig. 1). The sites visited in this study were state parks/forests, county parks, waterbodies allowing public access, and in some cases roadside ponds. Several types of wetland habitats were surveyed including lakes, ponds, reservoirs, hard-bottomed and soft-bottomed rivers, canals, swamps, marshes, and freshwater ponds adjacent to salt bays/marshes. Each survey site was contained in one of five, 16.10 km wide corridor transects running east to west and each transect consisted of eight to eleven sites. Five corridor transects were investigated to maximize the number of accessible sites and to assure that all target species would be accounted for during the study.
To establish a development gradient, individual sites were evaluated for local patterns of land-use using four, 3.22 km long line transects running in all cardinal directions from the center of each site. With Google Earth, an aerial view of each site was obtained and four line transects were drawn using the straight line ruler from the Google Earth toolbar (Google, 2010). Landscape types intersected by each line were classified as forest, field, wetland, park/garden, or developed area. The distance a line ran through a specific landscape type was measured and converted to a percent value. Percent values for each landscape type from all four lines were averaged together to produce landscape percent values for each site. A site's percentage of developed land dictated its placement along the development gradient. The authors of this study are confident that this method is more accurate than using the predefined landscape classes available in GIS software.
East to West Gradient
An east to west gradient was also established for all five transects using Google Earth (Google, 2010). In each transect, a line running east to west was drawn equidistantly between the northern and southernmost sites, and the ends of the line were aligned longitudinally with the eastern and westernmost sites. The distances between individual sites and the eastern end of a transect were measured along the line and converted to percent values. In central New Jersey, eastern locations were areas of ample development and disturbance while western locations were presumed to be more rural. The east to west gradient roughly parallels the development gradient and indicates where in the state some species are most abundant.
At each site, turtles were observed with binoculars while basking or bobbing at the water's surface. When surveying, possible basking substrates were scanned for turtles first, then the surface of the water was repeatedly scanned in an undulating motion starting near shore and eventually out towards the center of the waterbody. When surveying relatively small waterbodies (roadside and park ponds) that possessed clear views of the water from most angles and a perimeter that could easily be traversed on foot, turtles were observed from the water's edge. When surveying large waterbodies (large ponds, lakes, reservoirs, canals, and rivers) with limited access and viewing opportunities from land due to thick surrounding vegetation, turtles were observed from a canoe. Since the waterbodies surveyed varied in size, the amount of time spent searching for turtles at each site varied. Searching ceased after an entire waterbody was scanned. The number of turtles seen in a specific area was recorded on a tally sheet only once to prevent counting the same individuals twice. While this method does not reveal exact population densities, it does indicate the relative abundances of native verses non-native species at each site. Only six common basking emydids were targeted during the study due to their conspicuous nature. In order to correctly identify turtles from a distance, the observer acquired a basic familiarity with each species' morphology from guide books and several years of field experience. Hybrid individuals, if present, may or may not be readily identifiable as one of the six species sought in this study, but they would be noticeable in populations because hybrids tend to exhibit a mixture of morphological features from both parental lineages (Luttershcmidt et al., 2007). Turtles that could not be confidently identified due to visual obstruction or excessive distance were excluded from analysis.
Linear regression analysis for species abundance and development gradients was conducted for Trachemys scripta, Pseudemys rubriventris and Chrysemys picta using SPSS Vers. 17. Species association in relation to environmental factors (predictors) was explored using Canonical Correspondence Analysis (CCA) (Canoco Vers. 4.5 Ter Braak and Smilauer, 2002). Environmental factors were the landscape types derived from the spatial analysis described above (forest, field, wetland, park/gardens, or developed area) and the position along the east to west transects and transect identity (number). The derived ordination diagram of the first two axes (Fig. 3) shows the scores of species and environmental factors. For species scores the distance between the symbols in the diagram approximates the dissimilarity of distribution of relative abundance of those species across the samples as measured by their Chi-square distance. Environmental factor arrows show the expected direction in which the factor increases while relative length indicates the strength of the factor. In addition, the angles between arrows indicate correlations between individual environmental factors (e.g., 180 degrees indicates negative correlation). Species scores can be projected perpendicularly onto the arrow line of a given environmental variable and indicate the estimated optima of individual species of that environmental variable.
RESULTS AND DISCUSSION
Slider turtles, Trachemys scripta, were found to be more abundant in developed areas and rare or non-existent in rural areas. The majority of sliders identified during the study were red-eared sliders, T. s. elegans. Yellow-bellied sliders, T. s. scripta, only accounted for less than two percent of all identified sliders. Where sliders cohabitated with natives, they were usually outnumbered by any one species of native emydid. The two species that occurred with T. scripta, most frequently, were Pseudemys rubriventris and Chrysemys picta. In mixed communities of slider, red-bellied, and painted turtles, on average, only 20% of identified turtles were sliders. With the exception of a single red-eared slider inhabiting a canal with northern map turtles, Graptemys geographica, sliders were not found with the other three species counted during the study.
Linear Regression Analysis for Species Abundance and Development Gradients
The numbers of individual T. scripta, P. rubriventris, and C. picta counted at each study site were plotted over a development gradient (Fig. 2). Trachemys scripta was found to be more common in areas exceeding 40% development and almost non-existent in rural areas with less than 15% development (Fig. 2a). The association between numbers of individual sliders counted at each site and development was significant (Factor Development f = 3.486 p = 0.044 df = 58). Pseudemys rubriventris and C. picta were common in both rural and developed areas. There was a general pattern of evenness for C. picta abundance over a development gradient (Fig. 2c) and no significant association of C. picta with development was found (Factor Development f = 3.146 p = 0.057 df = 58). Pseudemys rubriventris was common in rural and developed areas but less so in areas of intermediate development (Fig. 2c). As with T. scripta, there was a significant association of P. rubriventris with development (Factor Development f = 3.960 p = 0.031 df = 58).
Species Association in Relation to Environmental Factors
Of the environmental factors considered, wetlands and fields, indicated by the lengths of the effect vectors in the CCA (Fig. 3), seem to be the most profound determinants of emydid turtle community composition in central New Jersey. Directional factors like the east to west gradient and transect number (north to south gradient) are secondary determinants. Forests and development, while not as influential as the factors mentioned above, also play a part in determining community composition.
[FIGURE 2 OMITTED]
Most species are closely associated (directly or indirectly) with specific environmental factors (Fig. 3). Trachemys scripta is largely associated with development in the east. Pseudemys rubriventris is also associated with development but in the central, southern portions of New Jersey. The central placement of the painted turtle, Chrysemys picta, (Fig. 3) indicates an equal association with all environmental factors. The close grouping of T. scripta, C. picta, and P. rubriventris (Fig. 3) indicates a close association between these species that is most likely due to their similar habitat preferences. Spotted turtles, Clemmys guttata, are not closely associated with any of the evaluated environmental factors. The northern map turtle, Graptemys geographica, is associated with the rural, northwestern quarter of New Jersey where agriculture is prevalent. The diamondback terrapin, Malaclerays terrapin, is associated with coastal wetlands that coincide with the highly developed eastern portions of the state. In New Jersey, G. geographica and M. terrapin are naturally distributed in regions where extensive anthropogenic landscape alteration (agriculture and development) has occurred. While our analysis clearly shows associations of G. geographica with agriculture and M. terrapin with development, it should be noted that these environmental parameters are certainly not conditions for the presence or absence of these species. Instead, they are merely indicators of how the regions associated with G. geographica and M. terrapin appear in central New Jersey.
[FIGURE 3 OMITTED]
Diamondback terrapins, Malaclemys terrapin, northern map turtles, Graptemys geographica, and spotted turtles, Clemmys guttata were not expected to be seen cohabitating with large numbers of T. scripta because of obvious differences in habitat preference. Malaclerays terrapin is a coastal species that inhabits wetlands with drastic variations in salinity (Holliday et al., 2009). In Georgia and South Carolina, it is possible to find T. scripta living in or near salt marshes (Ernst, et al. 2009); however, they are undoubtedly more suited to life in freshwater, making any interaction with M. terrapin a rare occurrence. In New Jersey, G. geographica is restricted to the northern half of the Delaware River, its neighboring canals, and the Raritan River Watershed (Schwartz and Golden, 2002). During the study, G. geographica was only common in flowing waterbodies with hard rocky bottoms, abundant basking platforms, and sparse vegetation. This is contrary to the habitat preferences of T. scripta who generally prefer still, soft-bottomed waterbodies with abundant aquatic vegetation (Ernst and Barbour, 1989). Spotted turtles, C. guttata, mainly inhabit shallow waterbodies less than half a meter deep such as swamps, bogs, damp pastures, small woodland streams, wet forests, and vernal pools (Meylan, 2006); whereas, T. scripta seem to prefer deeper waters (Thomas, 2006).
The threat non-native T. scripta pose to M. terrapin, G. geographica, and C. guttata in New Jersey is likely minimal. In different portions of their natural range, sliders occur sympatrically with these three species (Ernst et al., 2009). The habitat preferences of these species, most likely, exclude them from being impacted by both native and non-native sliders.
One of the species found in abundance with T. scripta, the painted turtle, C. picta, is also naturally sympatric with T. scripta. Chrysemys picta has a large distribution that greatly overlaps much of the slider's natural range and is the only North American turtle with a range extending from coast to coast (Ernst et al., 2009). Chrysemys picta was the species counted most frequently and was found in a variety of freshwater habitats, such as large lakes and reservoirs, woodland and park ponds, swamps, bogs, canals, marshes, and rivers. Painted turtles occurred with all species counted in the study, except M. terrapin. Since C. picta was found in almost any kind of freshwater habitat throughout New Jersey, its placement in the CCA (Fig. 3) was not surprising. It can be concluded that C. picta is a habitat generalist, who is adaptable to anthropogenic landscape alteration and environments that support a multitude of turtle species. This ability to successfully associate with several different turtle species coupled with the fact that C. picta and T. scripta occur sympatrically elsewhere, reduces the likelihood that introduced T. scripta will cause harm to populations of painted turtles. In fact, the adaptability and cold tolerance of C. picta might make it more invasive than T. scripta in Europe, if it were ever released there in abundance. The European Union took this into consideration when they included C picta in a 1997 import ban of T. s. elegans (Bringsoe, 2006).
The other species found in abundance with T. scripta was the northern red-bellied turtle, Pseudemys rubriventris. This species is distributed on the Atlantic Coastal Plain from the southern half of New Jersey, south to northeast North Carolina and westward along the Potomac River to eastern West Virginia. Relict populations also exist in eastern Massachusetts (Ernst et al., 2009).
Pseudemys rubriventris was commonly found in relatively deep bodies of water with soft bottoms and sufficient plant growth such as lakes, ponds, canals, slow flowing rivers, and marshes. These habitats closely match those preferred by T. scripta (Ernst et al., 2009). In locations where both species occurred, P. rubriventris was usually abundant in relation to T. scripta, especially in rural areas. At sites where both species were present, T. scripta accounted for an average of 25.39% of all turtles counted, while P. rubriventris accounted for 47.55%. Out of a total twelve locations where both species were present, only two had more individual T. scripta than P. rubriventris. One of these sites had comparably low numbers of both species; however, T. scripta clearly outnumbered P. rubriventris at the other site (73.69% T. scripta, 21.05% P. rubriventris, 5.26% C picta). This site was adjacent to a heavily developed region of Pennsylvania (Philadelphia County) where red-eared sliders are still commonly sold in pet shops (personal observation). It is certainly possible that this location and others might receive continual replenishment of T. s. scripta from captive releases.
Pseudemys rubriventris was surprisingly associated with development, even more so than T. scripta. This association is most likely an indirect one due to the fact that many waterbodies in central New Jersey are not far from areas of extensive development. If more transects running through the New Jersey Pinelands were investigated, where development is low and numbers of P. rubriventris are high, the association of red-bellied turtles with development would not be as significant.
In the Northeast, large basking turtles of the genera Pseudemys and Trachemys are mostly absent with P. rubriventris being the only naturally occurring member of this group (Hulse et al., 2001). In the southernmost extent of their range (northeast North Carolina and southeast Virginia), red-bellied turtles are sympatric with other large basking turtles, such as eastern river cooters, P. concinna concinna, coastal plain cooters, P. floridana [now recognized as P. concinna floridana (Crother, 2008)], and yellow-bellied sliders, T. s. scripta. This region is the northernmost extent of P. floridana and T. s. scripta while the range of P. c. concinna extends to the piedmont of northern Virginia (Ernst et al., 2009). Since all of these turtles have similar niches and the ranges of both northern and southern species terminate shortly after they overlap, these species might limit each other when they occur together.
In Florida, Aresco (2005) found that niche partitioning occurs between P. floridana and T. s. scripta due to differences in their levels of omnivory. Despite the presence of conspecifics and T. s. scripta, Pseudemys floridana was able to thrive on an almost exclusive diet of plentiful macroalgae, while T. s. scripta could not. In T. s. scripta, the effects of intraspecific competition were more profound than the effects of interspecific competition because of a stronger dependence on an unreliable food source (animal matter) (Aresco, 2005).
In New Jersey, P. rubriventris and non-native T. scripta might also partition the niche between each other as P. floridana and T. scripta do in Florida. If this is true, then populations of T. scripta probably limit themselves instead of limiting other native turtle species. In southeastern Virginia and northeastern North Carolina where P. rubriventris occurs with P. floridana and T. s. scripta, niche partitioning might occur between three species. Contrastingly, it could occur between T. s. scripta and Pseudemys spp. but not between P. rubriventris and P. floridana, offering a possible explanation for why the ranges of both species terminate shortly after they overlap.
No suspected hybrid individuals were observed during the study. When hybrids exist between two species of dissimilar appearance, they usually possess characteristics of both parent species and are quite noticeable in populations (Luttershcmidt et al., 2007). Because no suspected hybrids were observed, no genetic analyses were attempted. Hybridization does occur in the wild between P. rubriventris, P. concinna concinna, and P.floridana but is seemingly rare (Ernst et al., 2009). It is not known if Pseudemys and Trachemys hybridize in wild populations, but it is certainly possible for T. scripta to hybridize with other species in captivity. Hybrid offspring between T. scripta and Graptemys spp. resembling both parent species are occasionally offered for sale by online turtle breeders (personal observation).
The possibility of introduced T. scripta hybridizing with endemic species elsewhere is a serious threat. The species most likely threatened by hybridization are members of the genus Trachemys that were formerly classified as subspecies of T. scripta. Because many of these "new" species have small restricted ranges and are closely related to T. scripta, wild populations may be especially prone to genetic adulteration from introduced T. scripta (Seidel, 2002).
Introduced slider turtles, Trachemys scripta, are significantly associated with developed areas of lower ecological value and are usually absent in large numbers from most rural areas of higher ecological value. Since sliders are associated with areas of human activity and development, it is logical to assume that these areas are initial and possibly continual release points used by people to rid themselves of unwanted pet turtles. If sliders are remaining in or near these places, then they are not successfully spreading on their own in New Jersey. Most reptiles are not noted for their dispersal abilities; however, sliders are known to move overland in response to changing habitats (Cash and Holberton, 2005), but in the heavily developed regions of New Jersey this can be extremely difficult.
In New Jersey, Trachemys scripta regularly occurs in relatively low numbers with Chrysemyspicta and Pseudemys rubriventris in locations of intermediate to high levels of development. Based on their relative abundances, the five native New Jersey turtle species discussed here do not seem to be under any imminent threat from non-native sliders. This, however, does not mean sliders are completely harmless to New Jersey turtles and their habitats. The release of captive chelonians can potentially harm wild turtle populations by introducing foreign pathogens (Dodd and Seigel, 1991). Also, the addition of a fairly large omnivorous turtle to New Jersey wetlands could still pose a threat to native flora and fauna.
A comparative study of the slider's diet in its non-native Northeast range and native range should be conducted. This will allow us to gain a better understanding of the slider's adaptability to Northeast ecosystems and the species it exploits. The diets of juvenile and adult P. rubriventris and T. scripta should also be compared in New Jersey to elucidate differences in dietary preference between species and various age groups.
An additional study should be conducted to see if the continued sale of Trachemys scripta contributes to higher numbers of slider turtles in Northeast wetlands. The sale of turtles is legal in Pennsylvania but has been illegal in New Jersey since 1972 (NJAC 8:23-2.1). Comparing abundances of introduced sliders between the two states and performing molecular techniques to assess relatedness within populations could determine if introduced sliders depend on natural recruitment or continual replenishment from captive releases to reach and maintain high numbers in the Northeast.
The ecological relationship of Pseudemys rubriventris with T. scripta should be studied further and monitored closely, especially in areas where populations of red-bellied turtles are depleted. New Jersey provides ample opportunity to study interactions between sliders and native turtles in a variety of aquatic environs. A New Jersey Fish and Game sponsored program that encourages citizen volunteers to conduct periodic binocular or spotting scope surveys throughout the state would prove useful, assuming volunteers can be trained to differentiate the six turtle species described above. Such surveys would determine the abundance, distribution, and overall health of turtle populations in New Jersey and provide insightful information about the spread/prevalence and possible impact of T. scripta over time.
ARESCO, M. J. 2005. Ecological Relationships of Turtles In North Florida Lakes: A Study of Omnivory and the Structure of a Lake Food Web (doctoral dissertation). Florida: Florida State University. 158 pp.
BLANCK, T. 2006. Introduction to Hybrids. Pp. 148-160 in Vetter H. and P. van Dijk (eds.). Turtles of the World, Vol. 4. East and South Asia. Chimaira Buchhandelsgesellschaft mbH, Frankfurt. 160 pp.
BRINGSOE, H. 2006. NOBANIS--Invasive Alien Species Fact Sheet--Trachemys scripta.--From: Online Database of the North European and Baltic Network on Invasive Alien Species--NOBANIS www.nobanis.org, Date of access 9/9/2010.
CADI, A. and P. JOLY. 2004. Impact of the introduction of the red-eared slider (Tracherays scripta elegans) on survival rates of the European pond turtle (Emys orbicularis). Biodiversity and Conservation 13: 2511-2518.
CASH, B., and R. L. HOLBERTON. 2005. Endocrine and Behavioral Response to a Decline in Habitat Quality: Effects of Pond Drying on the Slider Turtle, Trachemys scripta. Journal of Experimental Zoology 303A:872-879.
CONANT, R. and J.T. COLLINS. 1998. A Field Guide to Reptiles & Amphibians Eastern/ Central North America. Houghton Mifflin Co., Lebanon. 616 pp.
CROTHER, B. I. 2008. Scientific and Standard English Names of Amphibians and Reptiles of North America North of Mexico, With Comments Regarding Confidence in Our Understanding. Society for the Study of Amphibians and Reptiles, Salt Lake City. 84 pp.
DODD, C. K. and R. A. SEIGEL. 1991. Relocation, Repatriation, and Translocation of Amphibians and Reptiles: Are they conservation strategies that work?. Herpetologica 47:336-350.
ERNST, C.H., and R. W. BARBOUR. 1989. Turtles of the World. Washington, D.C.: Smithsonian Institution Press, Washington, D.C. 313 pp.
ERNST, C.H., J.E. LOVICH, and R.W. BARBOUR. 2009. Turtles of the United States and Canada. Johns Hopkins University Press, Baltimore. 827 pp.
HOLLIDAY D.K., A. A. ELSKUS, and W. M. ROOSENBURG. 2009. Impacts of Multiple Stressors on Growth and Metabolic Rate of Malaclemys Terrapin. Environmental Toxicology and Chemistry. 28: 338-345.
HULSE, A.C., C.J. McCOY, and E. J. CENSKY. 2001. Amphibians and Reptiles of Pennsylvania and the Northeast. Cornell University Press, Ithaca. 419 pp.
LUTTERSHCMIDT, W.I., S.A. ESCOBAR, and E.D. WILSON. 2007. Multivariate Analyses of Shell Morphology of Putative Hybrid Box Turtles. Southeastern Naturalist 6:571-576.
MEYLAN, P.A. 2006. Clemmys guttata--Spotted Turtle Pp. 226-234 in Meylan, P.A. (ed.). Biology and Conservation of Florida Turtles. Chelonian Research Foundation, Lunenburg. 376 pp. Sale or distribution of viable turtle eggs and live turtles, NJAC 8:23-2.1 (1972). SEIDEL, M.E. 2002. Taxonomic Observations on Extant Species and Subspecies of Slider Turtles, Genus Trachemys. Journal of Herpetology. 36: 285-292.
SCHWARTZ, V. and D. M. GOLDEN. 2002. A Field Guide to Reptiles and Amphibians of New Jersey. New Jersey Division of Fish and Wildlife, Vineland. 89 pp.
THOMAS, B.R. 2006. Trachemys scripta--Slider or Yellow-Bellied Slider. Pp. 296-312. in Meylan, P.A. (ed.). Biology and Conservation of Florida Turtles. Chelonian Research Foundation, Lunenburg. 376 pp.
THOMAS DUCHAK AND CLAUS HOLZAPFEL
RUTGERS UNIVERSITY NEWARK, DEPT. OF BIOLOGICAL SCIENCES, NEWARK, NEW JERSEY 07102-1811
|Printer friendly Cite/link Email Feedback|
|Author:||Duchak, Thomas; Holzapfel, Claus|
|Publication:||Bulletin of the New Jersey Academy of Science|
|Date:||Sep 22, 2011|
|Previous Article:||Fire and the development of high-elevation pitch pine communities in Northeastern West Virginia.|
|Next Article:||Review of the book titled "Ecology of Estuarine Fishes: Temperate Waters of the Western North Atlantic" by Kenneth Able and Michael Fahay.|