A Skunk By Any Other Name...

Historically, skunks have been classified as weasels in the family Mustelidae. Members of the Mustelidae are grouped on the basis of the loss of the carnassial notch on the upper fourth premolar, the loss of the upper second molar, as well as the enlarged scent glands ( Bryant et al, 1993; Martin, 1989; Wozencraft, 1989). The problem with these characters is that they may be associated with a high degree of homoplasy (trait not derived from common ancestor, but rather from convergence, for example). For instance, Wozencraft (1989) has suggested that the loss of the notch on the carnassial occurred more than once in independent lineages of carnivores. All carnivores have scent glands, and they are especially enlarged in mustelids, with skunks taking that enlargement to an extreme. The association of a nipple with the scent gland, rather than a duct as in mustelids, suggests that the scent gland in skunks is apomorphic. Other characteristics shared between skunks and mustelids may represent retained primitive traits (e.g. auditory bulla type; Hunt, 1974). Bryant et al. (1993) reported that of the characters uniting the skunks and the mustelids, two of the shared character states are plesiomorphic, and the polarity of two additional characters is uncertain. Four of these characters that are shared between skunks and otters are based on tooth morphology, and the fifth (auditory bullae) was plesiomorphic. The relationship of the skunks to members of the rest of the family however has been controversial. The close relationship of the North American skunks is well supported (Anderson, 1989; Dragoo et al., 1993; Wozencraft, 1989, 1993), with three living genera recognized including Mephitis (hooded and striped skunks), Spilogale (spotted skunks), and Conepatus (hog-nosed skunks). Other taxa had been included at various times in the subfamily Mephitinae of the weasel family, for instance, two Old World genera, Ictonyx (African zorilla) and Mydaus (Oriental stink badger).

The African zorilla has a color pattern that converges on that of the North American skunks and is very similar in appearance to the spotted skunks. Many of the early naturalists confused the two genera (Nowak, 1991). In fact, Coues (1877) recognized the African “Zorillinae” as the Old World representative nearest to the Mephitinae. O'Brien et al. (1989), in their protein electrophoresis study of black-footed ferrets and other weasels in the genus Mustela, used skunks as an outgroup (close relative of the group being studied). These authors included the zorilla, (referred to as the African striped skunk) with the skunks. They suggested an ancient split between the “skunks” and the genus Mustela.

Traditionally, the Oriental stink badger has been placed within the badger subfamily Melinae, yet some researchers consider the Melinae to be composed of individuals not related by a common ancestor, primarily based on the inclusion of the stink badger (Bryant et al., 1993; Petter, 1971; Pocock, 1921; Radinsky, 1973; Schmidt-Kittler, 1981; Simpson, 1945). Earlier researchers (Petter, 1971; Pocock, 1921; Simpson, 1945) suggested a sister group relationship between the skunks and the badgers based on similar cranial characters shared by the stink badger and the Mephitinae. Radinsky (1973) asked the question: “Are stink badgers skunks?” His answer was inconclusive because the characters he identified as being shared between skunks and the stink badger were retained primitive traits. Even though the shared characters were primitive, Radinsky (1973) argued that the fossil record provides support for the skunks and stink badger being a natural group, with fossil skunks (dating back to the Miocene and Pliocene) occurring in the Old World. Recently, Wolsan (1999) has described what he called the oldest known skunk fossil, which occurred in Germany about 12 million years ago. Schmidt-Kittler (1981) also recognized a close relationship between stink badgers and skunks. Bryant et al. (1993) performed a more comprehensive of the Mustelidae based on cranial, post-cranial, and soft anatomy. Their study supported the stink badgers as a member of the Mephitinae.

More recent morphological data on carnivores have indicated a sister group relationship between the Lutrinae (otters) and the Mephitinae (Bryant et al., 1993; Hunt, 1974; Wolsan, 1999; Wozencraft, 1989; Wyss and Flynn, 1993). This conclusion was based on several features of the cranium, dentition, and soft anatomy.

Skunks have been suggested to be more closely related to three different subfamilies in the Mustelidae. One of the primary difficulties in determining the relationships among subfamilies of mustelids is the diagnosis of monophyletic (derived from a single common ancestor) groups on the basis of shared derived as opposed to shared primitive characteristics. The morphological data uniting the skunks with any particular subfamily of the Mustelidae, or even to the family, have been based on primitive character states and convergent similarity.

Examinations of non-morphological characters have revealed a somewhat different picture of mustelid relationships. For instance, Wurster and Benirschke (1968:374) studied chromosomal data of various carnivores and indicated that the "skunks are remarkably different from the rest of the family". These data suggest that in terms of their chromosomes skunks are apomorphic relative to other mustelids, yet these characters say little about the actual placement of skunks relative to the mustelids. Ledoux and Kenyon (1975) studied serum proteins and suggested that the Mustelinae (weasels), Melinae (badgers), and Lutrinae (otters) shared a common ancestry long after the lineage leading to the modern skunks diverged.

Recent molecular studies of carnivore relationships based on DNA hybridization (Árnason and Widegren, 1986; Wayne et al., 1989) have suggested that the procyonids (raccoons and allies) and pinnipeds (the three families of marine carnivores – seals, sea lions, and walrus) group more closely than skunks to other mustelids. Dragoo (1993 examine DNA sequence from 3 genera of skunks, and Ledje and Árnason (1996a, b) examined DNA sequence from a striped and a spotted skunk as well as other carnivores and concluded that skunks should be in their own family. A more complete study of the family was performed by Dragoo and Honeycutt (1997) and included members of all three American (North, Central, and South) genera, in addition to the stink badger and the African zorilla. They also concluded what the genetic data had been suggesting since the late ‘60s that skunks were not mustelids. Flynn et al. (2000) were concerned about the phylogenetic relationship of the red panda to other carnivores. They too concluded that skunks were not members of the Mustelidae, but also that red pandas were closely related to skunks. A more recent study (Eizirik et al. 2010) showed that skunks were at the base of a clade of carnivores called the Mustilida and contains the Musteloidea (Mustelidae and Procyonidae), Ailuridae, and the Mephitidae.

We are striving to use the molecular data to better understand the evolution of morphological traits in carnivores. Part of the discrepancy between the two types of data (morphological and genetic) may reside with the family Mustelidae not being monophyletic. If, as the molecular data suggest, the family is not monophyletic, then the interpretation of some morphological character state changes may be compromised (Anderson, 1989; Bryant et al., 1993).

Part of the answer may be obtained by considering existing information from the fossil record.The earliest true carnivores first appeared during the late Paleocene/early Eocene. These carnivores were small, arboreal, viverrid-like (or weasel-like) forms belonging to the extinct Viverravidae and Miacidae (Martin, 1989). As indicated by Anderson (1989) and Martin (1989), during the late Eocene and into the Oligocene these early forms gave rise to the Caniformia (Mustelidae, Canidae, Procyonidae and Ursidae) and the Feliformia (Felidae, Viverridae, and Hyaenidae). During this same time period, open-land communities were becoming prevalent. Many regions in North Americawere composed of arid-climate biota, and early into the Miocene there was evidence of wide-spread tropical savannas ( Martin, 1989). This period also was marked by the first occurrence of composite flowers, grasses, grazing ungulates with high-crowned teeth, and rodents. The appearance of the first mustelid-like carnivores was coincident with the increase of rodent diversity (Martin, 1989).

By the Oligocene the families Ursidae, Procyonidae, Canidae, and Mustelidae could be distinguished by characters of the basicranium (Baskin, 1982).Many Oligocene and early Miocene taxa had been identified as Mustelidae, but after careful examination it was later concluded that these taxa were actually procyonids (Anderson, 1989; Baskin, 1982). According toAnderson (1989) the fossil record for mustelids is incomplete, and many of the early workers named new taxa without sufficient comparative material, resulting in the placement of many taxa into incorrect subfamilies or families.

The early mustelid-like forms that appeared in the late Eocene cannot be traced to the modern mustelid descendants (Kurtén and Anderson, 1980).What is recognized today as modern mustelids first appeared in the Old World during the mid-Miocene (Kurtén and Anderson, 1980; Martin, 1989). One of the earliest known skunks was Palaeomephitis steinheimensis, which occurred about 11-12 million years ago (Wolsan, 1999). This skunk shares characteristics of the inner ear bones with otters, which Wolsan (1999) considers to be derived rather than primitive. However, Radinsky (1973) suggested that most of the earliest known mustelid brains, dating back 20 to 25 million years ago, are more advanced than those of modern skunks and the stink badger in several cranial features. In addition, Wayne et al. (1989) suggested that the origin of the skunk lineage occurred in the Oligocene approximately 40 million years ago (about 30 million years before the appearance of the first recognized fossil), which is prior to the musteloid stem group that gave rise to the Procyonidae. Because mustelids, in general, retain many primitive traits (Anderson, 1989) and the skunks even more so, it is difficult to find synapomorphies that are diagnostic for a monophyletic Mustelidae as well as other groups of mustelid-like carnivores. As a result, the Mustelidae (especially the Mustelinae) has been a catch-all category for many of the early, poorly differentiated taxa as well as divergent genera of doubtful affinity (Anderson, 1989).

Bonapate (1845) was the first to recognize that skunks were “different”. He called them Mephitina. However, he provided no justification for why they were different. This term only appears in the table of contents of his catalog. There were no specimens of skunks in his museum, so there was no reference to them in text. Often Gill (1872) is credited with the first use of the name which appears as Mephitinae in his work. Gill (1872) goes on to describe the subfamily as….. The first time that skunks were mentioned as a family was in 1894. Samuel Rhoads was discussing the original description by Linnaeus (1758) of what may or may not have been a skunk and stated, “There is little doubt that both described from hearsay of different species, combining interchangeably the habits of a Skunk with the colors of no known member of the Mephitide” (Ord and Rhoads, 1894:Appendix page 11). This likely was a misspelling of the family name because later in the index (Ord and Rhoads, 1894: Appendix page 72) he uses the correct spelling, “Mephitidæ”.