Granite Gap is a desert scrub habitat located in the Chihuahuan Desert in southwestern New Mexico about 200 km from the Texas border. In May 2016, we collected specimens of eight lizard species, six of which range into Texas: Callisaurus draconoides, Cophosaurus texanus, Uta stansburiana, Urosaurus ornatus, Gambelia wislizenii, and Aspidoscelis sonorae (a parthenogenetic species), plus two species not found in Texas: Sceloporus clarkii and Aspidoscelis tigris. We analyzed stomach contents of the preserved lizards and computed diet breadth and overlap for each. All lizard species consumed arthropods only. Considering the entire sample, there was a significant correlation between lizard snout-vent length (SVL) and total volume of arthropods consumed, and between lizard SVL and mean volume per prey item, but not between lizard SVL and number of arthropods consumed. This suggests larger lizards did not consume more arthropods than smaller lizards, but they did consume larger ones. Overall, A. sonorae was the most specialized lizard species at Granite Gap whereas C. texanus seemed to be the most generalized owing to its high numeric and high volumetric diet breadth. The dietary overlap data suggest there were two groups of lizards at Granite Gap: one that consumes a large number and volume of termites (Callisaurus, Cophosaurus, and the two species of Aspidoscelis) and the other that consumes a large number and volume of ants (Uta, Urosaurus, small Sceloporus) plus cicadas when lizard size is large enough to consume them (large Sceloporus and Gambelia).
The southwestern United States supports a diverse lizard fauna, the study of which has contributed enormously to herpetology’s understanding of lizard biology. Many studies of southwestern lizard biology have described lizard diet through analysis of lizard stomach contents. These studies have established that nearly all southwestern lizard species are arthropodivorous (Milstead & Tinkle 1969; Mitchell 1979). The bulk of studies of southwestern lizard diets have been single species studies (e.g., Parker & Pianka 1975; 1976; Vitt & Ohmart 1977; Maya & Malone 1989; Punzo 2007). However, populations of one lizard species often coexist with populations of several other species forming a diverse lizard community. A major factor in determining lizard community composition is the level of niche partitioning along the diet dimension. Therefore, an understanding of the diet of lizard species within a community is necessary for understanding community structure. Study of the diets of all the lizard species within a community makes it possible to compare diet breadths among species to determine which species are generalists and which are specialists. It also allows computation of diet overlap statistics to determine which species are most similar to one another. Such results provide a more detailed picture of the ecology of a lizard species than that which can be obtained by study of a single species in isolation.
Granite Gap, Hidalgo County, New Mexico, is a desert scrub habitat that is home to a lizard community comprising eight species, six of which occur in Texas: Callisaurus draconoides (zebra-tailed lizard), Cophosaurus texanus (greater earless lizard), Uta stansburiana (common side-blotched lizard), Urosaurus ornatus (ornate tree lizard; these four are in the family Phrynosomatidae); Gambelia wislizenii (long-nosed leopard lizard; family Crotaphytidae); and Aspidoscelis sonorae (Sonoran spotted whiptail-a parthenogenetic species, family Teiidae). Granite Gap also supports populations of two other lizard species not found in Texas: Sceloporus clarkii (Clark’s spiny lizard, a phrynosomatid) and Aspidoscelis tigris (tiger whiptail; a teiid). The diets of C. draconoides, C. texanus, U. stansburiana, G. wislizenii, and A. tigris have been studied extensively (Kay et al. 1970; Pianka & Parker 1972; Tanner & Krogh 1974; Parker & Pianka 1975; 1976; Vitt & Ohmart 1977; Best & Gennaro 1984; 1985; Smith et al. 1987; Maya & Malone 1989; Quijada-Mascareñas 1992; Maury 1995; Palacios-Orona & Gadsden-Esparza 1995; Durtsche et al. 1997; Gadsden & Palacios-Orona 2000; Lemos-Espinal et al. 2000; Dibble et al. 2007; Punzo 2007). All of these species are arthropodivorous with the smallest species, U. stansburiana consuming large numbers of ants and the largest species, G. wislizenii, occasionally consuming lizards in addition to arthropods. However, the diets of the other three species, U. ornatus, A. sonorae, and S. clarkii have been described in only a few studies (U. ornatus: Aspland 1964; Bergeron & Blouin-Demers 2020; A. sonorae: Mitchell 1979; Wojnowski 2010; S. clarkii: Brooks & Mitchell 1989). There is a need, then, for additional data to provide a more complete description of the diet of these less-studied species within the context of the larger lizard community to gain a better understanding of the feeding habits of the species that inhabit Texas as well as the rest of the American southwest. The purpose of this study is to describe and analyze the diet of a sample of lizards collected during a class field trip led by CEM with special focus on the three species that have been studied the least, to compute the diet breadths of all eight lizard species to determine which species are specialists and which are generalists, and to examine diet overlap among lizard species in this community.
Materials & Methods
Study area.—Lizards were collected from 10–27 May 2016 in the northwest portion of the Chihuahuan Desert at Granite Gap, Hidalgo County, New Mexico (elevation 1372 m; 32°05′36.7″N, 108°58′06.3″W) about 222 km west of El Paso. The study site consists of rocky outcrops, exposed bedrock, and desert vegetation. Vegetation is dominated by desert scrub, cacti (Opuntia sp.), mesquite (Prosopis sp.), yucca (Yucca sp.), acacia (Vachellia sp.), juniper (Juniperus sp.), and ocotillo (Fouquieria splendens).
Collection of specimens.—We located lizards by haphazard visual searching of suitable habitat on foot. Once we detected a lizard, we noted the location of the lizard and attempted to capture it by hand, blow gun, or noose pole. If the lizard was captured, we recorded the date, time, species, sex, mass, snout-vent length, and tail length. We determined sex by probing or identification of sexually dimorphic traits, such as enlarged femoral pores, postanal scales, and the presence of a tail bulge in male lizards. We measured mass with an electronic balance (±0.1 g) and lengths with a ruler (±0.1 cm). We fixed lizards in formalin in the field and transferred them to 70% ethanol after 2 weeks. The specimens are currently housed in the Truman State University Natural History Collection, Kirksville, Missouri (NEMS; Sabaj 2016).
Of the 63 lizard stomachs we examined, 56 had prey items in them (two Cophosaurus texanus, one Uta stansburiana, one Gambelia wislizenii, one Aspidoscelis sonorae, and two Aspidoscelis tigris stomachs were empty). A total of 1269 prey items comprising a total volume of 15047 mm3 was recovered from the lizard stomachs. Analysis of the entire sample showed no significant correlation between lizard SVL and number of prey items per stomach (Pearson correlation coefficient = −0.22; P = 0.657), but there was a significant positive correlation between lizard SVL and total volume of prey consumed (Pearson correlation coefficient = 0.56; P < 0.001) and between lizard SVL and mean volume per prey item (Pearson correlation coefficient = 0.46; P < 0.002). There is a significant positive correlation between lizard SVL and total volume of prey only for Sceloporus clarkii (Pearson correlation coefficient = 0.88; P = 0.012). For the other seven species, all correlations between lizard SVL and diet parameters are either statistically non-significant or lack adequate sample size for a statistical test. The sample of S. clarkii had a greater variation in lizard SVL (ranging from 45 mm to 110 mm) than did the samples of any other species in this study perhaps allowing S. clarkii a sufficient range of lizard sizes to allow the size to total prey volume correlation to reach statistical significance.
Summary of lizard diets.—The overall diet analysis results are presented in online supplemental material available at https://doi.org/10.32011/txjsci_73_1_Article05.SO1; what follows is a brief summary of the diet of each of the eight lizard species studied. The small sample size (n=3) for Callisaurus draconoides makes conclusions tentative, but termites (Isoptera) and caterpillars (Lepidoptera larvae) were numerically most abundant (one lizard ate 49 termites) and caterpillars and ant lions (Myrmeleonidae larvae) were volumetrically dominant. For Cophosaurus texanus, termites, beetle (Coleoptera) larvae, and small ants (Formicidae: small) were numerically dominant (just over 75% of prey items consumed) and ant lions, beetles, and beetle larvae were volumetrically dominant. For Uta stansburiana, ants, termites, and beetle larvae were numerically the most abundant prey (>90% of prey items consumed) and beetle larvae, ants, and grasshopper nymphs (Acrididae) were volumetrically dominant. For Urosaurus ornatus, ants, beetle larvae, and small unidentifiable beetles were numerically the most abundant prey (ants made up nearly 75% of prey items) and a single cicada (Cicadidae) consumed by one individual made up the largest volumetric contribution, followed by beetle larvae and grasshopper nymphs. The sample size for Gambelia wislizenii is small, but it is noteworthy that this species did not consume large numbers of “common” prey such as ants, termites, or beetle larvae; instead newly emergent cicadas were the most important prey volumetrically. For both of the two whiptail lizard species, Aspidoscelis sonorae and A. tigris, the numerically dominant prey item was termites (> 85% or all prey items consumed for both species). Though termites were also ranked high volumetrically, the consumption of a large pupa by both species, ant lions by a few A. sonorae and a large cicada by one A. tigris made these categories important volumetrically. For Sceloporus clarkii, ants were the most abundant prey items consumed (83% of total); these small prey were consumed primarily by smaller lizards. Large cicadas, grasshoppers, and ground beetles (Carabidae) were volumetrically dominant; this was primarily because two large adults consumed these larger prey items. A small amount of plant material was found in the stomachs of one Callisaurus, one A. sonorae, and two Sceloporus, nematode parasites were found in three Sceloporus, but not in any other lizards in the sample. At least one individual of each species had sand grains in its stomach.
Diet breadth and overlap.—The greatest numeric diet breadth was recorded for Gambelia wislizenii whereas the smallest was recorded for the two whiptail species Aspidoscelis sonorae and A. tigris (Table 1). This was because no G. wislizenii consumed large numbers of any single prey taxon whereas both whiptails consumed large numbers of a single prey taxon, termites (Isoptera). Interestingly, G. wislizenii showed the lowest volumetric diet breadth (Table 1); this was because most of the volume of prey consumed by G. wislizenii consisted of large cicadas. Aspidoscelis sonorae and Sceloporus clarkii also showed low volumetric diet breadths owing to the large fraction of the volume of prey consumed being of only a few prey taxa (insect pupae for A. sonorae; cicadas for S. clarkii). The greatest volumetric diet breadths were recorded for Urosaurus ornatus and Cophosaurus texanus (Table 1); neither of these two species specialized much on any particular prey taxon (at least in terms of the volume of prey consumed).
The two whiptail species, Aspidoscelis sonorae and A. tigris, showed the highest numeric diet overlap and second highest volumetric diet overlap within the sample (Table 2). This was because both species consumed large numbers of termites (Isoptera) and a large volume of insect pupae. Callisaurus draconoides and Cophosaurus texanus overlapped greatly with each other and moderately so with the two whiptails, primarily because they also consumed a large number of termites. Uta stansburiana, Urosaurus ornatus, and Sceloporus clarkii showed moderate to high overlap with one another primarily due to the consumption of ants (Formicidae) by U. stansburiana, U. ornatus and small Sceloporus (Table 2). The only lizard species that overlapped extensively with Gambelia wislizenii was Sceloporus clarkii because G. wislizenii and large Sceloporus both consumed a large volume of cicadas.
The eight lizard species from Granite Gap followed the general pattern of southwestern lizards in that they consumed a variety of arthropods. The diet of all species other than Gambelia wislizenii was numerically dominated by either termites or ants (see https://doi.org/10.32011/txjsci_73_1_Article05.SO1). These are commonly abundant prey items in the desert scrub of southern New Mexico, but their small size means a lizard can consume many of them. Termites and ants are consumable by small lizards such as Uta stansburiana, Urosaurus ornatus, and small Sceloporus clarkii, but they are also consumed in large numbers by Callisaurus draconoides, Cophosaurus texanus, and both species of Aspidoscelis. The fact that both large and small lizards consumed large numbers of small prey items explains the lack of correlation between lizard SVL and number of prey consumed. Only large lizards, however, consumed large prey such as cicadas, large insect larvae, or insect pupae resulting in a significant positive correlation between lizard SVL and both total volume of prey and mean volume per prey item.
Though the diets of the eight lizard species at Granite Gap are similar to the diets of these species from other areas, some subtle differences exist. For example, C. draconoides at Granite Gap consumed a large number of termites; this was not the case for this species studied in most other places (Kay et al. 1970; Vitt & Ohmart 1977; Smith et al. 1987; Quijada-Mascareñas 1992), though Dibble et al. (2007) found a few termites in the stomach contents of C. draconoides from Sonora. Similarly, C. texanus from Granite Gap consumed many termites, but other studies of this species report little or no consumption of termites (Maury 1995; Durtsche et al. 1997; Punzo 2007). Interestingly though, the high dietary overlap between C. draconoides and C. texanus found at Granite Gap was also documented in a study of these two species coexisting in northwestern Arizona (Smith et al. 1987). Several studies have documented that G. wislizenii consume lizards (Tanner & Krogh 1974; Parker & Pianka 1976). Although G. wislizenii had the largest mean SVL (at 81.7 mm) and consumed the largest mean volume per prey item (241 mm3/prey item) in the Granite Gap sample, there was no evidence of saurophagy by G. wislizenii at Granite Gap. However, the sample was taken at a time of year when neonates of other lizards species were not available. It is possible that a sample of G. wislizenii collected later in the year after hatchlings of the other species had emerged would have had lizards in it.
Three of the lizard species in the Granite Gap sample are ones whose diet has rarely been studied: Aspidoscelis sonorae, Sceloporus clarkii, and Urosaurus ornatus, The detailed study of the ecology of four whiptail species co-occurring in the desert grassland of southeastern Arizona published by Mitchell (1979) presents the only detailed study of the diet of Aspidoscelis (formerly Cnemidophorus) sonorae. Mitchell (1979) found termites constituted over 95% of the prey items consumed by A. sonorae, a number similar to the nearly 90% found in the Granite Gap sample. Cicadas comprised the second largest fraction of volume of prey consumed by the A. sonorae studied by Mitchell (1979), but were absent from the Granite Gap sample despite their being available (as indicated by their presence in the diet of other Granite Gap lizards). The bulk of the volume of prey consumed by A. sonorae at Granite Gap was comprised of large insect pupae consumed by a couple of lizards; pupae were absent from Mitchell’s Arizona sample. The only detailed study of the diet of Sceloporus clarkii is that of Brooks & Mitchell (1989) who studied this species (along with two others) from a sample collected from a thorn forest in Sonora, Mexico. They found that ants and termites accounted for about two-thirds of the prey items consumed, but that caterpillars and beetles accounted for over 70% of the volume of prey consumed. The Granite Gap S. clarkii did not consume termites or caterpillars even though both were available (as indicated by their presence in the diets of other Granite Gap lizards). Why the Granite Gap S. clarkii avoided termites and caterpillars is unknown. The Granite Gap S. clarkii did consume large numbers of ants and a large volume of beetles, but unlike the S. clarkii reported on by Brooks & Mitchell (1989), the large lizards also consumed a large volume of cicadas (see https://doi.org/10.32011/txjsci_73_1_Article05.SO1). This difference probably reflects that temporary availability of cicadas at Granite Gap. The only detailed studies of the diet of U. ornatus of which we are aware are those that of Aspland (1964) and Bergeron & Blouin-Demers (2020) both of which described diets of populations in Arizona. Both studies found ants were the numerically dominant prey in U. ornatus diet, though Aspland (1964) also documented increased consumption of leafhoppers (Homoptera), termites, bugs (Hemiptera), or adult beetles during seasons when these prey were temporarily abundant. These results parallel what we found at Granite Gap: ants were the dominant prey but several other types of arthropods were consumed as well.
The two whiptail species, Aspidoscelis sonorae and A. tigris, had by far the lowest numeric diet breadth due to their consumption of large numbers of termites. Termite consumption is a hallmark of whiptail lizards in the southwest and frequently results in low numeric diet breadth (Mitchell 1979; Paulissen et al. 2006). Aspidoscelis sonorae also had a very low volumetric diet breadth due to its consumption of a large volume of insect pupae and termites. Thus A. sonorae can be considered to be the most dietarily specialized of the eight Granite Gap lizard species. Identifying which lizard species is the most generalized is difficult. The largest species, Gambelia wislizenii, showed the greatest numeric diet breadth, but the lowest volumetric diet breadth (due to the consumption of large cicadas by two lizards). The smallest species, Urosaurus ornatus, showed the greatest volumetric diet breadth, but one of the lowest numeric diet breadths (due to consumption of large numbers of ants). This suggests that the most generalist species at Granite Gap may be neither of these two, but is probably Cophosaurus texanus since it showed the second highest numeric and second highest volumetric diet breadth in the sample.
The numeric and volumetric diet overlap values show several trends. First, there is extremely high numeric and volumetric diet overlap between the two species of Aspidoscelis as well as between C. draconoides and C. texanus. These four species show very high numeric diet overlap with one another due their consumption of large numbers of termites. Second, Urosaurus ornatus shows high numeric and volumetric diet overlap with both Uta stansburiana and Sceloporus clarkii; the two later species also show high numeric diet overlap with each other. The main factor driving these relationships is the consumption of large numbers of ants by U. ornatus, U. stansburiana, and small S. clarkii. Finally, Gambelia wislizenii shows extensive volumetric overlap with S. clarkii because both large S. clarkii and G. wislizenii consumed a large volume of cicadas. A single Urosaurus ornatus managed to consume a small cicada resulting in a high volumetric diet overlap with G. wislizenii. Overall, the overlap data suggest there were two main groups of lizards at Granite Gap: one that consumes a large number and volume of termites (Callisaurus, Cophosaurus, and the two species of Aspidoscelis) and the other that consumes a large number and volume of ants (Uta, Urosaurus, small Sceloporus) plus cicadas when lizard size is great enough to consume them (large Sceloporus and Gambelia).
Our results describe the diets of lizards during a short interval in late spring of a single year at a single location. Thus, our results do not reveal potential seasonal, annual, or geographic differences in lizard diet, diet breadth, or diet overlap. Nevertheless, our study provides a baseline for comparison of diets of southwestern lizard from different seasons, years, and locations. Since most southwest lizard species range across several states, diet studies in one area can contribute to a better understanding of lizard diets and diet niche metrics in other areas where those lizards occur. Comparison of such studies will aid in developing a broader understanding of diet and diet niche metrics for lizard species throughout the southwest.
We thank P. Goldman for co-offering the course and Truman State University for sponsoring the field course. Research permits were provided to CEM by New Mexico Department of Game and Fish (#3526) and all activities were conducted in accordance with TSU IACUC.
Supplemental material is available for this article online.