2011 SOUTHEASTERN NATURALIST 10(2):357–364 Correlated Morphological, Ecological, and Behavioral Aspects of the Microhabitat Associations in Geolycosa Wolf Spiders of Florida (Araneae, Lycosidae) Samuel D. Marshall1,* and Katherine A. Martin1 Abstract - Species of Geolycosa (burrowing wolf spider) found in the eastern USA exhibit one of two species-specific microhabitat associations with associated burrow architectures: “turricolous” species burrow in areas covered in leaf litter and construct a distinct turret of silk and debris at the burrow mouth, while “aturricolous” species locate their burrows in more barren areas and do not construct a turret. Our research goal was to document the relationship between spider coloration, burrow-site selection, and burrowconstruction behaviors in five Geolycosa species in Florida (G. micanopy, G. hubbelli, G. xera, G. patellonigra, and G. escambiensis). Discriminant analysis of morphological data supports our contention that each ecomorph has a distinct pattern of pigmentation, with the barren-substrata species being paler. Field studies of species pairs at three sites in Florida (Archbold Biological Station, Ocala National Forest, and the Panhandle) documented the species-specific association with one of the two microhabitats. Field enclosure tests with five of these species demonstrated that the observed microhabitat association was due to active burrow-site choice. Finally, tests we conducted in the laboratory documented that these species will build the predicted species-specific burrow types under controlled conditions. Introduction Habitat selection is the behavioral process by which dispersing animals identify and settle in a location meeting their resource needs (e.g., food, nest sites, moisture). Spiders have been the subject of extensive research in habitat selection (Uetz 1991, Wise 1993) in part because most are sedentary, even sessile, building webs and/or silk-lined retreats. Because of the investment in site choice implicit in investing resources in a trap or retreat, habitat choice is unambiguous (Uetz 1991). The “ecomorph” concept (short for ecological morphology) proposes an adaptive correlation of habitat use and functional characteristics of an organism’s phenotype. Ecomorphs are suites of developmentally stable species-specific traits functionally linked to the organism’s niche (Losos 1990, Schulte et al. 2004). This study examines the correlation of the state of each of three ecological traits in a suite of Geolycosa (burrowing wolf spiders): habitat selection, burrow construction, and color pattern. The specific combinations of the trait states should be consistent between the two ecomorphs we propose. Geolycosa wolf spider species, unlike most other burrowing lycosids, are obligate burrowers and spend their entire lives in and around a burrow they construct themselves. All foraging activity is centered on the burrow mouth, and 1Northwestern State University of Louisiana, Department of Biological Sciences, 317 Bienvenu Hall, Natchitoches, LA 71497. *Corresponding author - marshalls@nsula.edu. 358 Southeastern Naturalist Vol. 10, No. 2 the spiders only leave their burrow to make strikes at passing prey. An exception to this rule is when males become vagrant upon molting into their terminal instar. The burrows of North American Geolycosa are similar to each other in structure: a circular, vertical tube terminating in an enlarged chamber (Wallace 1942). Burrow depths vary with species, soil, and latitude, but can range from 15–100 cm (S.D. Marshall, pers. observ.). Most North American Geolycosa species exhibit a highly localized pattern of distribution and dispersion (Marshall 1995a, b; Marshall et al. 2000). Geolycosa from the eastern US appear to exhibit clear microhabitat preferences; some species burrow in barren areas, while others are found associated with soil covered with a litter layer (Carrel 2003, Edwards and Edwards 1989, Marshall 1997, Marshall et al. 2000, Wallace 1942). “Turricolous” Geolycosa species are those that construct their burrows in microhabitats in which the soil surface is covered in leaf litter or other vegetative debris. They construct a turret around the burrow mouth by binding leaves, stems, and other plant debris together with silk. The turret is an unambiguous feature that entails specific behaviors to construct and can exceed 2 cm in height. In contrast, “aturricolous” Geolycosa species build no turret at the burrow mouth, and are restricted to microhabitats in which the soil surface is barren (Carrel 2003, Marshall et al. 2000). The goal of this research was to document the morphological, ecological, and behavioral attributes correlated with the species-specific microhabitat associations in three Geolycosa species pairs across Florida. Methods General methodology All spiders used in the research below were collected in the field by digging them from their burrows. Because it is not possible to determine the sex of these spiders before the penultimate instar (when males begin to show an enlargement of the pedipalps prior to maturity), all research detailed here was conducted with either adult females or immature males and females. Morphometric analysis We measured 10 morphological attributes on each preserved specimen of Geolycosa. Carapace width was recorded and used to scale the following 3 features: carapace height (measured at the anterior edge), and the lengths of patella-tibia I and IV. We also quantified the degree of black pigmentation on ventral body segment surfaces: sternum, venter, coxa I, femur I, patella I, and tibia I. Pigmentation was noted as lacking on the anatomical feature (0), partially covering the feature (0.5), or entirely covering the feature (1.0). We used ventral color pattern as a proxy for overall degree of pigmentation because these ventral pigmentation marks remain distinct, even after the specimen has been preserved in alcohol for decades (S.D. Marshall, pers. osberv.). The more subtle dorsal patterns become difficult to quantify after a similar period of preservation as the specimens fade. We further justify the use of ventral pigmentation noting that Wallace in his 1942 revision of the genus relied heavily on ventral pigmentation as a key character, indicating its stability as a species-specific trait. A total of 51 subadult male and 2011 S.D. Marshall and K.A. Martin 359 female and adult female specimens were measured, including aturricolous species: G. escambiensis Wallace (n = 12), G. patellonigra Wallace (n = 13), G. xera xera McCrone (n = 12), and G. xera archboldi McCrone (n = 2); and turricolous species: G. hubbelli Wallace (n = 2) and G. micanopy Wallace (n = 10). Discriminant analysis (SPSS ver. 14.0) was used with “turricolous” or “aturricolous” being used as categories. Habitat analysis We used a line-intercept method to study the soil-surface conditions in the vicinity of Geolycosa burrows. In all cases, burrows were located by walking, head down, scanning the soil surface for burrows during daylight hours. At Archbold Biological Station, data were collected on the burrows of 24 G. x archboldi and 29 G. hubbelli. At Ocala National Forest, data were collected in the Big Scrub area on the burrows of 25 G. patellonigra and 21 G. hubbelli. In the Florida Panhandle, data were collected in coastal scrubs west of Apalachicola and east of Pensacola on the burrows of 20 G. escambiensis and 15 G. micanopy. Habitat was quantified by laying a 1-m transect in a north–south direction to standardize data collection across sites. At each site, the meter stick was centered on the burrow (i.e., at the 50-cm mark) and substratum type was recorded for each 10-cm interval as: (1) barren, (2) sparse litter, or (3) completely litter covered. These determinations were made by visual inspection. We collected spiders from all burrows surveyed to confirm species identity. These data were analyzed using a MANOVA of counts of the categories of substratum cover along each transect (leaf litter, leaf litter + sand, barren sand). These data were analyzed with SPSS (ver. 14.0) using locality and species as factors. Habitat choice We conducted microhabitat choice experiments in field enclosures in which recently collected spiders were offered a choice of barren or litter-covered substrata for settlement. Geolycosa movement is restricted to relocation after burrow abandonment. Research has documented that Geolycosa do settle in self-constructed burrows immediately upon dispersing from the maternal burrow (Marshall 1995a, b). We have never observed Geolycosa wandering or captured them in pitfall traps, apart from sexually mature males. It is assumed that these spiders quickly construct a new burrow if they find themselves without a burrow. Observed causes for burrow abandonment in the field include invasion by ants, territorial conflicts, and movement of litter (Marshall 1995a, b). The enclosures were built of 30-cm-wide aluminum sheeting (wolf spiders lack adhesive tarsi and so are unable to climb smooth surfaces). Each enclosure was 0.5 by 1.0 m. Prior to each trial, we would randomly select (by coin toss) which half of the enclosure would be litter covered, with the other half of the enclosure remaining barren. We would then cover the sand in the selected end with leaf litter collected in the vicinity. A 5-cm wide strip of mesh plastic of the type used to prevent leaves from clogging household rain gutters was used to contain the leaf litter. This strip was necessary as otherwise wind would move the litter around the enclosure. Spiders were collected from burrows in the area and used within 5 days. The open vial containing the spider was 360 Southeastern Naturalist Vol. 10, No. 2 placed along the middle of the enclosure next to the plastic mesh. The side of the mesh for release was selected randomly, by coin toss. The spider was allowed to exit the vial on its own. Because the leaf litter barrier material was low (5 cm) and perforated, it presented no barrier to movement by the spiders. We would note the location of the spider the next morning. Where it was found, with or without a burrow, was considered a choice. Between trials, the leaf litter was replaced and the sand was thoroughly mixed. Each spider was used only once. Burrowing construction behavior The use of debris in turret construction under standardized conditions was tested in a greenhouse at Northwestern State University of Louisiana using spiders collected in Florida and transported to Louisiana in 85- by 30-mm cylindrical plastic vials. The length of time these spiders were held in the vials varied from one to several weeks. During this holding period, all spiders were given moist sand as a substratum and fed crickets. Prior to the test, spiders were allowed to establish burrows in 592-ml plastic drink cups filled to a depth of approximately 10 cm with sand. The test consisted of offering each spider ten 1.0-cm sections of pine needle, placed 1.0 cm from the burrow mouth for turret construction. Pine needles were used because it was relatively easy to standardize the size and shape of the material offered. We have observed Geolycosa using pine needles for turret construction in the field. We examined the containers 48 h later, checking for the incorporation of the pine needle sections into the silk collar at the mouth of the burrow. A spider was scored as turret building if they incorporated more than one pine needle section into the burrow mouth. Results Morphometric analysis The discriminant function was able to separate the two classes of Geolycosa species (turricolous and aturricolous) primarily based on the amount of black pigmentation on the ventral body surfaces (Wilk’s Lambda = 0.015; with sternum, venter, coxa I, femur I, and patella I color being the most important). Habitat analysis When these data were summarized, there were clear differences between the burrow sites of turricolous and aturricolous species in the prevalence of barren sand versus complete litter cover (Fig. 1). The differences between the number of barren intervals and litter-covered intervals at aturricolous burrows vs. turricolous burrows were significant; the difference between the number of mixed-substratum intervals was not (MANOVA: F3, 129 = 40.9, P < 0.001). These results indicate a clear separation based on the extremes of substratum condition (i.e., barren vs. litter) but ambiguity when there are elements of both (scattered litter on barren sand). Habitat choice We found a strong preference for the predicted substrata for all species except the aturricolous G. patellonigra, which settled randomly (Table 1). 2011 S.D. Marshall and K.A. Martin 361 Burrow-construction behavior A total of 27 turricolous and 63 aturricolous spiders were tested. Virtually all spiders constructed or did not construct turrets as predicted by their species (Table 2). Based on a binomial test for each species/locality tested, all P < 0.01. Figure 1. Results of 1-m transect studies of soil-surface conditions around Geolycosa wolf spider burrows at three localities in Florida. “Archbold” represents data collected at Archbold Biological Station, Highlands County; “Ocala NF” represents Ocala National Forest, Marion County; “Port St. Joe” represents coastal scrubs in the vicinity of Port St. Joe, Gulf County. Soil-surface conditions were recorded for each of ten 10-cm intervals along the 1-m transect as barren (no leaf litter present), light leaf litter/debris (litter present and sand visible) and leaf litter/debris (no barren sand visible). Table 1. Results of habitat choice test. Number of Geolycosa selecting the two different substratum types in field enclosures. Significance levels based on a binomial test with an expectation of random (50/50) settlement. Archbold = Archbold Biological Station, Ocala = Ocala National Forest, and St. Joseph’s = St. Joseph’s Peninsula State Park. Turricolous/ n settle n settle Taxon aturricolous Locality n tested in litter in sand P G. hubbelli Turricolous Archbold 15* 11 2 0.0095 G. xera archboldi Aturricolous Archbold 15 0 13 0.0001 G. hubbelli Turricolous Ocala 10 9 1 0.0092 G. patellonigra Aturricolous Ocala 10 5 4 0.2461 G. escambiensis Aturricolous St. Joseph’s 10 0 10 0.0001 *Differences between the number tested and the number scored represents missing spiders. 362 Southeastern Naturalist Vol. 10, No. 2 Discussion Our results support our hypothesis that Geolycosa wolf spiders actively select burrow sites based on the presence or absence of a litter layer covering the soil surface and that this choice is correlated with degree of pigmentation and burrow-construction behavior. Thus, there are two alternative species-specific ecomorphs in the eastern US. Geolycosa species which primarily exploit areas where the soil is covered in leaf litter are dark in color and construct turrets at the burrow mouth, whereas those species that are restricted to barren areas tend to be paler in color and do not build a turret. The repeated pattern of pair-wise co-distribution of the ecomorphs suggests that competitive interactions may explain why only one of each ecomorph species may be found at a given site. Geolycosa is unusual among spiders in that most species have a very restricted geographical range, and that historical biogeography has led to the accumulation of endemic species in Florida (Marshall et al. 2000). In contrast, in more northern regions with a shorter post-glacial history, only single-pairs of species may be found across a large geographical area (e.g., G. missiouriensis and G. wrightii in the Midwest and Great Lakes region, G. turricola and G. pikei along the Atlantic coastal plain to Georgia). The repeated co-occurrence of these contrasting ecomorph species pairs raises the question of their origin. In Florida, the high rates of endemism of plants and invertebrates on the inland sand ridges is generally attributed to vicariance biogeography (Marshall et al. 2000). To date, there is no definitive answer to the question of evolutionary origins. Unpublished phylogenetic analyses based on mtDNA sequence data (S.D. Marshall et al., unpubl. data) indicate that the aturricolous ecomorph has evolved several times, and that ecomorph status reflects adaptation rather than common descent. The adaptive significance of the turret constructed by Florida’s Geolycosa is untested. Research into the turret-construction behavior of other burrowing lycosids has provided differing hypotheses for the function of the turret. Framenau and Baehr (2007) propose that the turret may function as a “look out” for Dingosa species wolf spiders in Australia by providing them with an elevated perch to sit and wait for prey. Williams et al. (2006) found experimental evidence that the turret of Lycosa tarantula (L.) (Spanish Wolf Spider) functions to deter burrow Table 2. Results of a laboratory test of burrow-construction behavior in which Geolycosa inhabiting self-constructed burrows in laboratory containers were offered turret-building material (ten 1-cm lengths of pine needle). Predictions are based on species identity. Ocala = Ocala National Forest, Archbold = Archbold Biological Station, and Panhandle = Central Panhandle Proportion Proportion Collecting Turricolous/ Number building expected to Binomial Species locale aturricolous tested turrets build turrets probability G. hubbelli Ocala Turricolous 11 0.8 1.0 P = 0.0 G. hubbelli Archbold Turricolous 16 1.0 1.0 P = 0.0 G. xera Archbold Aturricolous 24 0.0 0.0 P = 0.0 G. patellonigra Ocala Aturricolous 15 0.2 0.0 P = 0.0 G. escambiensis Panhandle Aturricolous 24 0.0 0.0 P = 0.0 2011 S.D. Marshall and K.A. Martin 363 invasion by scorpions. We did not test for turret function. However, the observation that these turricolous-aturricolous species pairs are often found within meters of each other segregated by microhabitat may indicate that the turret (or lack thereof) is specifically related to microhabitat structure rather than foraging or anti-predator behavior. Of course, none of these hypothesized functions are mutually exclusive. Geolycosa wolf spiders are a peculiar component of the fauna of scrubs and barrens across the eastern half of the US. It seems that where there is sand, there will be Geolycosa. We have even discovered Geolycosa in small, isolated barrens in the Midwest and Mid-Atlantic states. As is true with many sand-barrens endemics (e.g., Lycaeides melissa samuelis Nabokov [Karner Blue Butterfly]), fire plays a significant role in the ecology of Geolycosa (Carrel 2008). Fire and post-fire succession in part determines the local distribution and abundance of Geolycosa species pairs. Long-term fire suppression can lead to the local extinction of aturricolous Geolycosa species as a litter layer develops—one way that development pressures in Florida have impacted local species, such as G. escambiensis. Geolycosa escambiensis is only known from the central Panhandle west into Alabama. Much of its habitat has been compromised by development, and there are no protected populations in Florida. We consider it to be the only recognized Geolycosa species at risk. Acknowledgments We thank Archbold Biological Station for allowing one of us (S.D. Marshall) to conduct research into the ecology of Geolycosa over the years. We thank the Florida Department of Environmental Protection for permission to collect spiders and conduct research at St. Joseph Peninsula State Park (Permit # 07060711). We thank the United States Department of Agriculture for a special use permit to conduct research at Ocala National Forest (Authorization ID LAK198). This research was funded in part by Northwestern State University’s Enrichment Fund (NEF Agreement #PS 06-07 002). Finally, we thank the students in S.D. Marshall’s summer 2007 Introductory Biology Laboratory for their enthusiastic contribution to the laboratory studies of turret-construction behavior. We also thank E. Hebets for her editorial improvements to the writing and C. Morrison for comments on an early draft of the manuscript. We especially thank L. Campbell for her critical review of grammar and style. Literature Cited Carrel, J. 2003. Ecology of two burrowing wolf spiders (Araneae: Lycosidae) syntopic in Florida scrub: Burrow/body size relationships and habitat preferences. Journal of the Kansas Entomological Society 76:16–30. Carrel, J. 2008. Differential survival of Geolycosa xera archboldi and G. hubbelli (Araneae, Lycosidae) after fire in Florida scrub. Journal of Arachnology 36:595–599. Edwards, R., and E. Edwards. 1989. Cape Cod’s burrowing wolf spiders. 1989. Cape Cod Naturalist 18:12–19. Framenau, V., and B. Baehr. 2007. Revision of the Australian wolf spider genus Dingosa Roewer, 1955 (Araneae, Lycosidae). Journal of Natural History 41:1603–1629. Losos, J. 1990. Ecomorphology, performance capability, and scaling of West Indian Anolis lizards: An evolutionary analysis. Ecological Monographs 60:369–388. 364 Southeastern Naturalist Vol. 10, No. 2 Marshall, S.D. 1995a. Natural history, activity patterns, and relocation rates of a burrowing wolf spider: Geolycosa xera archboldi McCrone (Araneae, Lycosidae). Journal of Arachnology 23:65–70. Marshall, S.D. 1995b. Mechanisms of the formation of territorial aggregations of the burrowing wolf spider Geolycosa xera archboldi McCrone (Araneae, Lycosidae). Journal of Arachnology 23:145–150. Marshall, S.D. 1997. The ecological determinants of space use by a burrowing wolf spider in a xeric shrubland ecosystem. Journal of Arid Environments 37:379–393. Marshall, S.D., W.R. Hoeh, and M.A. Deyrup. 2000. Biogeography and conservation biology of Florida’s Geolycosa wolf spiders: Threatened spiders in endangered ecosystems. Journal of Insect Conservation 4:11–21. Schulte II, J., J. Losos, F. Cruz, and H. Núñez. 2004. The relationship between morphology, escape behaviour, and microhabitat occupation in the lizard clade Liolaemus (Iguanidae: Tropidurinae: Liolaemini). Journal of Evolutionary Biology 17:408–420. Uetz, G.W. 1991. Habitat structure and spider foraging. Pp. 325–348, In S.S. Bell, E.D. McCoy, and H.R. Mushinsky (Eds.). Habitat Structure: The Physical Arrangement of Objects in Space. Chapman and Hall, London, UK. Wallace, H.K. 1942. A revision of the burrowing spiders of the genus Geolycosa (Araneae, Lycosidae). American Midland Naturalist 27:1–62. Williams, J., J. Moya-Laraño, and D. Wise. 2006. Burrow decorations as anti-predator devices. Behavioral Ecology 17:586–590. Wise, D.H. 1993. Spiders in Ecological Webs. Cambridge University Press, Cambridge, UK.