The Southwestern Naturalist 48(4):654–660 December 2003 Implications of Hybridization between White-tailed Deer and Mule Deer

Frequency of interspecific hybridization between mule deer and white-tailed deer was investigated using ribosomal DNA and nucleotide sequences from the mitochondrial cytochrome b gene. Two of 15 individuals (collected by hunters) from northwestern Texas (Kent County) were identified as hybrids based on the ribosomal DNA marker. Sequence data from the cytochrome b gene indicated that 1 individual was the result of mating between a mule deer doe and a white-tailed deer buck, whereas the second individual resulted from mating between a whitetailed deer doe and a mule deer buck. These results indicate that hybridization between mule deer and white-tailed deer is not restricted to the Trans-Pecos region of Texas. Given the low levels of genetic divergence and frequency of hybridization, the validity of recognizing mule deer and white-tailed deer as distinct species is discussed. It is concluded that without the morphological and behavioral differences these 2 taxa could be considered subspecies rather than distinct species. RESUMEN La frecuencia de hibridación entre venados bura y venados cola blanca fue investigada usando ADN ribosomal, y secuencias de nucleótidos del gen citocromo b mitocondrial. Dos de 15 individuos (obtenidos a través de cazadores) provenientes del noroeste de Texas (condado de Kent) fueron identificados como hı́bridos basándose en el marcador ADN ribosomal. Los datos de las secuencias del gen citocromo b indicaron que un individuo fue el resultado de un apareamiento entre una hembra venado bura y un macho venado cola blanca, mientras que otro individuo fue el resultado de un apareamiento entre una hembra venado cola blanca y un macho venado bura. Estos resultados indican que la hibridación entre los venados bura y los venados cola blanca no está restringida a la región Trans-Pecos de Texas. Dado los bajos niveles de divergencia genética y frecuencia de hibridación, se discute la validez del reconocimiento de los 2 venados como especies distintas. Se concluye que sin las diferencias morfológicas y conductuales estos 2 grupos podrı́an ser considerados como subespecies y no especies distintas. Hybridization between mule deer (Odocoileus hemionus) and white-tailed deer (O. virginianus) has been well documented in various regions of North America (Cowan, 1962; Kramer, 1973; Wishart, 1980; Carr et al., 1986; Stubblefield et al., 1986; Cronin et al., 1988; Gavin and May, 1988; Cronin, 1991; Derr, 1991; Ballinger et al., 1992; Carr and Hughes, 1993; Cathey et al., 1998). In Texas, it has been hypothesized that hybridization between these 2 species has been a result of white-tailed deer expanding their range into the historical range of mule deer (Carr et al., 1986). This hypothesis is based on an encroachment of woody plant species (preferred habitat of white-tailed deer) into the more open grassland habitat typically associated with mule deer (Wiggers and Beasom, 1986). In addition, it has been inferred that the dynamics of hybridization are driven by behavioral and ecological factors instead of being limited by genetic sterility (Derr et al., 1991). Carr et al. (1986) extensively studied hybridization between these 2 species in areas of western Texas generally referred to as the TransPecos region. Their studies indicated that some specimens identified as mule deer, based on morphological appearance, had mitochondrial DNA normally associated with whitetailed deer. This finding was used as evidence December 2003 655 Bradley et al.—Hybridization between deer of natural hybridization. These authors further hypothesized that the direction of hybridization was a result of white-tailed deer does mating with mule deer bucks. However, Ballinger et al. (1992), Carr and Hughes (1993), and Cathey et al. (1998) provided evidence that a more likely scenario was hybridization between male white-tailed deer and female mule deer, with the white-tailed deer population eventually capturing the mitochondrial DNA genome of mule deer. Similarly, Stubblefield et al. (1986) examined 319 individuals of mule deer and whitetailed deer from 5 counties in the Trans-Pecos region of Texas. Based on allozyme data from the albumin locus, they concluded that hybridization averaged 5.6% (range 0.0 to 13.8%) within the 5 counties. Additionally, hybridization varied greatly between populations, with hybridization as high as 24% within 1 population. In this study, we examined sympatric populations of white-tailed deer and mule deer from Kent County, Texas, for evidence of hybridization. Available for study were 15 individuals collected by hunters during a single season at the above locality and 28 individuals from localities in Texas and Colorado. A nuclear marker, 28S ribosomal DNA (rDNA) region, was used to determine if individuals were either of a pure mule deer genotype, pure white-tailed deer genotype, or a mixed genotype. A mixed genotype was viewed to be a result of hybridization between the 2 species. In addition, DNA sequences from the maternally inherited mitochondrial cytochrome b gene were used to establish direction of hybridization for hybrid individuals. METHODS Samples We obtained 43 muscle or liver samples from specimens of mule deer and whitetailed deer collected by hunters from 9 natural populations in Texas (n 5 31) and Colorado (n 5 12) during 1989. In addition, 2 individuals representing a pen-raised F1 hybrid and a white-tailed deer from South Carolina were included as positive controls. Initial identification of specimens was based on ear length, tail coloration, and antlers (if male). Specimen numbers, collection localities, and identification by the rDNA method are listed in Table 1. Ribosomal DNA Data Genomic DNA was isolated from 1 white-tailed deer, 2 mule deer, and the penraised F1 individual using the methods of Bingham and Rubin (1981) followed by a phenol/chlorofom/ isoamyl alcohol extraction. Genomic DNAs were digested with a suite of 37 restriction enzymes following the guidelines of the manufacturer (New England Biolabs, Beverly, Massachusetts; Promega, Madison, Wisconsin; and Boehringer-Mannheim, Indianapolis, Indiana). Digested DNA fragments were electrophoresed on agarose gels and transferred to nylon membranes following Southern (1975). Radioactively labeled probes (labeled via random priming; Boehringer-Mannheim, Indianapolis, Indiana) were constructed from rDNA clones of the 18S (p2546) and 28S (pI19) genes (Arnheim, 1979). Membranes were hybridized to 18S and 28S ribosomal probes and placed against x-ray film. Four restriction enzymes (Bam HI, Bgl II, Hinc II, and Stu I) revealed variation in the 28S fragment between white-tailed deer and mule deer. The remaining samples were then digested with at least 1 of the 4 restriction enzymes and hybridized to the 28S probe. Cytochrome b Sequence Data Using the 28S restriction site data, 2 individuals were identified as hybrids. These individuals, in addition to 1 mule deer and 1 white-tailed deer (from the same locality), were examined by sequencing the first 425 base pairs (bp) of the mitochondrial cytochrome b gene. This sequencing was accomplished by extracting genomic DNA from frozen muscle samples (0.1 g) using the DNeasy Tissue Kit (Qiagen, Valencia, California). The entire cytochrome b gene (1,143 bp) was amplified using polymerase chain reaction (PCR) parameters modified from that described by Saiki et al. (1988): 1 initial cycle (958C for 2 min); 35 cycles of 958C denaturation (30 s), 458C annealing (20 s), 728C extension (1 min 40 s); and 1 final 728C extension cycle (15 min). Primers utilized in the PCR reactions were H15149 and L14724 of Irwin et al. (1991). The resulting PCR product was purified using the QIAquick PCR purification kit (Qiagen, Valencia, California). The following 2 primers were used in cycle sequencing reactions to amplify the forward and reverse strands, respectively: LGL765 (Bickham et al., 1995) and H15149. Cycle sequencing was conducted using the ABI Prism Big Dye version 3 terminator ready reaction mix (PE Applied Biosystems, Foster City, California) and samples were analyzed on an ABI Prism 310 automated sequencer (PE Applied Biosystems, Foster City, California). Vector NTI 7.0 software (Informax, Inc., Bethesda, Maryland) was used to align and proof nucleotide sequences. All cytochrome b sequences obtained in this study were deposited in GenBank (accession numbers AF535863–AF535866). Data Analyses The cytochrome b sequence data generated herein were analyzed using likelihood methods and the HKY algorithm, identified by the MODLETEST program (Posada and Crandall, 2001) as the model best fitting the data. The heuristic search option in the software package PAUP* (Swof656 vol. 48, no. 4 The Southwestern Naturalist TABLE 1—Identification of 44 individuals examined in this study based on gross morphology and 28S ribosomal DNA restriction patterns (Bam HI, Bgl II, Hinc II, and Stu I). W 5 genotype associated with whitetailed deer, M 5 genotype associated with mule deer, and H 5 genotype associated with white-tailed deer and mule deer (hybrid individual). TK# Morphology Bam HI Bgl II Hinc II Stu I Locality