Southeastern Naturalist 41 J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 22001155 SOUTHEASTERN NATURALIST Vo1l4.( 114):,4 N1–o5. 61 Distribution and Habitat Selection by the Maritime Pocket Gopher Jorge D. Cortez1, Scott E. Henke1,*, Dean W. Wiemers1, Timothy E. Fulbright1, David B. Wester1, and Richard Riddle2 Abstract - Geomys personatus maritimus (Maritime Pocket Gopher) is a genetically distinct subspecies that only occurs in deep sandy soils of Nueces and Kleberg counties of southern Texas. The US Fish and Wildlife Service has considered recommending the Maritime Pocket Gopher for federal listing. Because a large proportion of this gopher’s current range occurs on US Navy (hereafter, Navy) property, active management by the Navy plays a key role in the conservation of this subspecies. Therefore, our objective was to assess the distribution and habitat preferences of Maritime Pocket Gophers on Navy properties in southern Texas. We conducted a strip-transect survey to evaluate the number and distribution of gopher mounds on the Navy base. We created GIS layers of the different disturbance types and plant communities, and overlaid them onto the mapped gopher-mound density layer. If the proportion of available area used by Maritime Pocket Gophers was below or above the corresponding 90% confidence interval, we categorized disturbance types and plant communities as used more often or less often, respectively. The distribution of Maritime Pocket Gophers was influenced by soil-particle size, plant community, type of habitat disturbance, and quantity of leaf litter. Maritime Pocket Gophers favored areas with sandy soil in frequently mowed native prairie, restoration fields, and Cynodon dactylon (Bermuda Grass) with little to no ground litter (i.e., dead vegetation, debris). Introduction Geomys personatus maritimus Davis (Maritime Pocket Gopher) is one of 7 subspecies of Geomys personatus True (Texas Pocket Gopher) and is endemic to the sandy soils of the mainland in Kleberg and Nueces counties of coastal South Texas, between Baffin Bay and Flour Bluff (Williams and Genoways 1981). This subspecies was closely associated with native coastal prairie and deep sandy soils; however, development of agricultural fields and urbanization have minimized and fragmented sandy grasslands, resulting in a patchy distribution of this subspecies. Published data on Maritime Pocket Gophers focused on general characteristics of morphology, distribution, and habitat (Williams 1982). Accordingly, the assumed current geographic distribution of the Maritime Pocket Gopher may be outdated because it was determined more than 30 years ago by Williams and Genoways (1981). Potential threats to persistence of the Maritime Pocket Gopher include habitat loss, degradation, and fragmentation (by urbanization and agricultural conversion) resulting in reduced genetic variability and gene flow. In addition, native 1Caesar Kleberg Wildlife Research Institute, MSC 218, 1150 Engineering Avenue, Texas A&M University-Kingsville, Kingsville, TX 78363-8202. 2US Navy, 8851 Ocean Drive, Corpus Christi, TX 78419-5226. *Corresponding author - scott.henke@tamuk.edu. Manuscript Editor: Allan O’Connel Southeastern Naturalist J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 42 vegetation has been replaced by the sub-tropical exotic grasses Urochloa maxima (Jacq.) R. Webster (Guinea Grass), Pennisetum ciliare (L.) Link (Buffel Grass), and Cynodon dactylon (L.) Pers. (Bermuda Grass) in much of the region (Larson et al. 2010). Exotic grass invasion has been associated with the decline of small-mammal populations (Germano et al. 2001). The threat of predation is relatively small because gophers are subterranean (Andersen 1987) and few endemic predators (e.g., feral cats, snakes) present on the study site have the means and persistence to dig a gopher out of its elaborate tunnel system. The Maritime Pocket Gopher’s restricted distribution makes it particularly vulnerable to adverse human impacts such as urbanization (Hafner 2000). Accordingly, the US Fish and Wildlife Service has labeled the Maritime Pocket Gopher as a species of concern and has considered recommending it for federal listing. Because the potential threats are mainly habitat- related, it is important to determine the habitat preferences of Maritime Pocket Gophers to help managers develop habitat-management practices that are attuned to the requirements for this subspecies. A large proportion of this gopher’s current distribution occurs on Navy property, thus, active management by the Navy plays a key role in the conservation of the subspecies. Therefore, our objectives were to assess the distribution and habitat preferences of Maritime Pocket Gophers on Navy properties in southern Texas. We sought to determine Maritime Pocket Gopher distribution patterns and the preference or avoidance of plant communities and types of habitat disturbance on Naval Air Station (NAS) Corpus Christi, and Navy Landing Airfield (NALF) Waldron Station. We also assessed how the presence of exotic grasses versus native grasses and forbs, the presence of ground litter and roads, and soil chemical and physical properties influenced site selection by Maritime Pocket Gophers. Our ultimate objective was to use our findings to propose management options for this subspecies. Field-site Description The study was conducted on two Navy properties, NAS Corpus Christi (27°41'33.47''N, 97°17'28.36"W) and NALF Waldron (27°38'05.68''N, 97°18'43.90"W) in the Flour Bluff region (16.1 km southeast of Corpus Christi) in Nueces County, TX. Flour Bluff lies in the Gulf Prairies and Marshes ecoregion and is surrounded by water on 3 sides (i.e., Corpus Christi Bay to the north, Oso Bay to the west, and the Laguna Madre to the east). NAS Corpus Christi is 1049 ha in extent and is home to the US Coast Guard, Army Depot, and Navy. The landscape of NAS Corpus Christi mirrors that of the entire Flour Bluff region in that it is highly developed with new construction encroaching on the fragmented areas of grassland and scrubland habitat. NLAF Waldron is 352 ha in extent, 7 km south of NAS Corpus Christi, and consists of an airfield, control tower, and recreational sport fields. This station is less developed and has a more contiguous area of mid-grass prairie and scrubland communities than NAS Corpus Christi. Only 2 soil series occur at both study sites, Galveston (mixed, hyperthermic Typic Udipsamments) and Mustang (siliceous, hyperthermic Typic Psammaquents) Southeastern Naturalist 43 J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 fine sand and dredge spoils (fine-loamy, mixed, superactive, frigid Typic Haploxerolls), which consist of clay loam (NRCS 1960). The percentage of Galveston and Mustang fine sand to clay loam on NAS Corpus Christi are 80% and 20%, respectively. The soil in the NALF Waldron study area is entirely Galveston and Mustang fine sand (NRCS 1960). Both study areas consist of predominantly coastal, mid-grass-prairie grasslands and scrub-dominated, mixed grassland communities. Both communities occur on Galveston and Mustang fine sand and clay loam. Grass species include Aristida purpurea var. wrightii (Nash) Allred (Wright’s Threeawn), Dichanthelium oligosanthes (Schult.) Gould (Scribner’s Panicum), Spartina spartinae (Trin.) Merr. Ex Hitchc. (Gulf Cordgrass), Paspalum plicatulum Michx. (Brownseed Paspalum), Andropogon virginicus L. (Broomsedge Bluestem), Paspalum monostachyum Vasey (Gulfdune Paspalum), Schizachyrium scoparium var. litoralis (Seacoast Bluestem), and Dicanthium annulatum (Forssk.) Stapf (Kleberg Bluestem) (Garcia 2001). The forb species include Baptisia alba (L.) Vent. (Whitestem Wild Indigo), Phlox drummondii Hook. (Annual Phlox), Ambrosia psilostachya (Cuman Ragweed), Physalis spp. (ground cherry), Croton spp. (croton), Rayjacksonia phyllocephala (DC.) R.L. Hartm. & M.A. Lane (Camphor Daisy), Rhynchosia americana (Houst. Ex Mill.) M.C. Metz (American Snoutbean), Cassia fasciculata (Michx.) Greene (Partridge Pea), and Amorpha fruticosa L. (False Indigo) (Garcia 2001). The predominant woody species are Quercus virginiana Mill. (Live Oak), Salix nigra Marshall (Black Willow), and Prosopis glandulosa Torr. (Honey Mesquite). Opuntia engelmannii Salm-Dyck ex Engelm. (Prickly Pear Cactus) is also present. Additionally, 4 nonnative grasses occur on the sites: Bermuda grass, Stenotaphrum secundatum (Walter) Kuntze (St. Augustine Grass ), Buffel Grass, and Guinea Grass. Methods Gopher population assessment We surveyed the occurrence of Maritime Pocket Gopher mounds on both study sites employing strip-line transects (Witmer et al. 1999). We used a 2004 US Department of Agriculture (USDA) aerial photograph of the study sites to create a strip-line transect map. With Using ArcGIS 9.1, we laid a 3-ha-grid matrix over the aerial photograph and subdivided each grid into 1-ha strips. We chose one of the three 1-ha strips from each grid selected by a random number generator in Microsoft Excel. Due to the configuration of the study area, transect lengths ranged from 250 m to 536 m, and all transects were >15 m apart. The surveyor walked each 1-ha strip using a Trimble GPS unit (model GeoExplorer III DGPS with beacon receiver, accurate to within less than 0.5 m; Trimble, Sunnyvale, CA) to stay within the strip area and recorded the location of every mound within the strip. We counted all mounds, regardless of condition—either fresh or old—because old-appearing burrow systems still may have been active (Proulx 2002). We assumed that we detected 100% of mounds present because our transects were relatively narrow, making it unlikely that we missed any Maritime Pocket Gopher mounds. We verified suspected burrow systems by visual estimation of mound proximity, and mapped them using Southeastern Naturalist J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 44 a Trimble GPS unit. We considered mounds > 15 m apart to be associated with separate burrow systems (Witmer et al. 1999). We determined the area of each burrow system using ArcGIS 9.1 XTools analysis and recorded the number of mounds within each suspected burrow system. Macrohabitat assessment We imported gopher-burrow system locations into ArcGIS 9.1 and created a burrow-system map. We delineated plant communities and habitat-disturbance areas with the Trimble GPS unit and then categorized them (Appendices 1, 2). We classified plant communities as salt marsh, Bermuda Grass, Bermuda Grass/ Guinea Grass, debris-native vegetation/Guinea Grass, golf course, Guinea Grass, Guinea Grass/Buffel Grass, Gulf Cordgrass, Live Oak, Honey Mesquite/Guinea Grass, native prairie, restoration field, softball field or park, St. Augustine Grass, and vegetated runway (Appendix 1). We categorized disturbance types as construction, developed, manicured, mowed, previously disturbed, shrub removal, or untouched (Appendix 2). Differentiation of plant communities and habitat-disturbance types are explained in Appendix 1 and Appendix 2, respectively. We used the geoprocessing tool, Dissolve, on the gopher-burrow -system layer to combine overlapping gopher systems (ERSI) and clip analysis in ArcGIS 9.1 on both the plant community layer and habitat-disturbance layer overlaid upon the gopherburrow layer. This analysis enabled us to determine the area (m2) for each burrow system in each plant community and habitat-disturbance type. We converted burrow-system area into proportion of area observed in each plant community and habitat-disturbance type. We determined gopher use of plant community and habitat-disturbance types as described by Neu et al. (1974) using chi-square analysis and Bonferroni Z-statistics to control experiment-wise error at P = 0.10. Plant communities and habitat-disturbance types were considered used more often or less often, respectively, if the proportion of available area used by Maritime Pocket Gophers was below or above the corresponding 90% confidence interval. Microhabitat assessment We used a restricted random sampling design to examine microhabitat selection by Maritime Pocket Gophers (Keating and Cherry 2004). We randomly placed 10-m-long transects within each vegetation community at NAS Corpus Christi and NALF Waldron using Hawth’s Analysis Tools 3.8 in ArcGIS 9.1 (Beyer 2004). Transects were proportionally allocated to communities based on their spatial extent; we placed a minimum of 5 transects in the communities that comprised the smallest proportion of the study sites. We established a total of 196 transects resulting in a sampling density of 1 transect/3.5 ha. We recorded the presence or absence of Maritime Pocket Gopher mounds within a 4 m x 10 m belt along each transect. We chose a 4-m-wide belt because we could easily observe pocket gopher mounds on either side of the center of the belt. We calculated the distance of each transect from roads using ArcGIS 9.1 software. We estimated percent cover of exotic grass species, native grasses and grass-like Southeastern Naturalist 45 J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 plants, forbs, bare ground, and ground litter within three 20 cm x 50 cm quadrats placed at 0, 5, and 10 m along each of the 196 transects using methods of Daubenmire (1959). We collected soil from the upper 12 cm at 0 m, 5 m, and 10 m along each transect using a 2-cm-diameter soil probe. Samples from within a transect were combined, placed in pans, spread out uniformly, and dried at room temperature to a constant mass. We sieved the samples using a 5-mm sieve to remove roots and other debris and analyzed the composite soil samples from each transect to determine soil texture, pH, nitrate nitrogen (NO3-N), phosphorous (P), and potassium (K). We used the hydrometer procecure to estimate soil texture (Day 1965, Murphy and Riley 1962), a 1 N KCl solution to evaluate NO3-N, a cadmium column and spectrophotometer measurement to estimate the reduction of nitrate to nitrite (Keeney and Nelson 1982), a hydrogen-selective electrode to determine soil pH, a conductivity probe to estimate conductivity (Schofield and Taylor 1955), and the inductively coupled plasma (ICP) method and Mehlich III extractant d to determine P and K (Mehlich 1978, 1984). We employed t-tests to compare percent cover of ground litter, Bermuda Grass, Buffel Grass, Guinea Grass, Kleberg Bluestem, native forbs, and native grasses; % clay; NO3-N(μg/ml); K (μg/ml); pH; P (μg/ml); % sand; % silt; % bare ground; and distance from roads (km) between transects with ≥1 Maritime Pocket Gopher mound and transects that did not contain Maritime Pocket Gopher mounds . We tested distributions of residual errors for normality via the Shapiro-Wilk test (Steel and Torrie 1980). Variables that differed significantly (P < 0.05) between used and non-used sites were included as predictor variables in further analyses using logistic regression models to estimate 3rd-order resource-selection functions (Johnson 1980, Manly et al. 2002). The response variable in logistic regression was the log odds of presence of Maritime Pocket Gopher mounds (Hosmer and Lemeshow 1989, Moruzzi et al. 2002). Predictor variables in logistic regression models included litter percent cover, percent clay, Kleberg Bluestem percent canopy cover, and native grass cover. We used combinations of the 4 variables to produce 15 candidate models. We used Akaike weights to compare models (Burnham and Anderson 2002) using program MuMIn (Barton 2013) in Program R. Results Gopher-population assessment Of the 557 transects surveyed on NAS Corpus Christi and NALF Waldron, we observed Maritime Pocket Gopher mounds on 308 (55.3%) transects. We documented 14,627 Maritime Pocket Gopher mounds constituting 1155 suspected burrow systems on Navy property. The greatest mound density observed on a transect for NAS Corpus Christi and NALF Waldron was 402 mounds/ha and 208 mounds/ha, respectively. Most transects that did not contain observable Maritime Pocket Gopher mounds were composed of large swaths of concrete or live oak. Maritime Pocket Gopher mounds comprised 10 ha (1.5%) and 0.4 ha (0.2%) of useable space on NAS Corpus Christi (669.5 ha) and NALF Waldron (190 ha), respectively. Southeastern Naturalist J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 46 Macrohabitat assessment We identified 15 plant communities and 7 habitat disturbance types on NAS Corpus Christi and NALF Waldron (Tables 1, 2). The composition of plant communities on NAS Corpus Christi and NALF Waldron was 3.7% salt marsh, 19.0% Bermuda Grass, 3.3% Bermuda Grass and Guinea Grass mixture, 1.0% native prairie and Guinea Grass intermixed with debris, 8.5% golf course grasses, 0.8% Guinea Grass, 0.3% Guinea Grass and Buffel Grass mixture, 0.9% Gulf Cordgrass, 8.9% Live Oak, 4.6% Honey Mesquite and Guinea Grass, 43.2% native prairie, 2.9% restoration field, 1.0% softball field or park, 1.4% St. Augustine Grass, and 0.4% abandoned vegetated runway (Table 1). The composition of habitat disturbances on NAS Corpus Christi and NALF Waldron was 1.4% construction, 32.0% developed, 7.3% manicured, 36.4% mowed, 5.3% previously disturbed, 0.1% shrub removal, and 17.5% untouched (Table 2). Macrohabitat selection Gopher mounds were not distributed proportionately to habitat type (χ2 = 92,454; P < 0.0001) or disturbance type (χ2 = 121,585; P < 0.0001). Nearly 92.5% of Maritime Pocket Gopher mounds were observed in native prairie (48.9%), Bermuda Grass (20.7%), restoration fields (16.6%), and Bermuda and Guinea grass associations (6.2%) (Table 1). Maritime Pocket Gophers used all other habitat types less often. The restoration fields accounted for nearly 74% of the observed χ2 value; the Table 1. Occurrence of Maritime Pocket Gopher mounds within various habitat types on Naval Air Station-Corpus Christi and Waldron Airfield, Corpus Christi, TX. χ 2 = 92,453.9; P < 0.0001. Outcomes: P = preferred, A = avoided. % of Area of % observed Total total gopher mounds in each Confidence Disturbance area (m2) area Observed Expected area interval Outcome Salt Marsh 231,923.7 3.70 1885.3 3861.4 1.81 0.0170–0.0192 A Bermuda Grass 1,191,263.5 19.00 21,561.5 19,828.9 20.66 0.2031–0.2101 P Bermuda Grass/ 209,457.5 3.33 6455.3 3475.3 6.19 0.0597–0.0639 P Guinea Grass Debris-Native/ 64,655.1 1.03 347.2 1074.9 0.33 0.0028–0.0038 A Guinea Grass Golf course 535,332.6 8.51 2992.7 8881.3 2.87 0.0273–0.0301 A Guinea Grass 50,604.3 0.81 133.4 845.3 0.13 0.0010–0.0016 A Guinea Grass/ 18,129.9 0.29 0.0 302.7 0.00 0.0000–0.0000 A Buffel Grass Gulf Cordgrass 58,679.8 0.93 0.0 970.6 0.00 0.0000–0.0000 A Live Oak 562,467.9 8.94 2081.8 9330.0 1.99 0.0187–0.0211 A Honey Mesquite/ 286,113.1 4.55 155.6 4748.5 0.15 0.0012–0.0018 A Guinea Grass Native prairie 2,715,941.2 43.20 51,058.4 45,032.5 48.92 0.4847–0.4933 P Restoration field 180,390.1 2.87 17,296.0 2995.2 16.57 0.1625–0.1689 P Softball park 65,506.0 1.04 278.4 1085.4 0.27 0.0023–0.0031 A St. Augustine Grass 91,182.4 1.45 117.1 1513.3 0.11 0.0008–0.0014 A Vegetated runway 27,406.6 0.40 0.0 417.4 0.00 0.0000–0.0000 A Total 6,289,053.7 100.00 104,362.7 104,362.7 100.00 Southeastern Naturalist 47 J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 other frequently used habitat types cumulatively accounted for only 4.6% of the observed χ2 value. Of the remaining 21% of the total χ2 value, Live Oak and the Honey Mesquite and Guinea Grass association accounted for half (5.4% and 4.8%, respectively) of the χ2 value. Maritime Pocket Gophers frequently used areas that were mowed or disturbed (i.e., construction, shrub removal, or previously mowed) and apparently did not use developed, manicured, and untouched areas (Table 2). We observed the greatest number of gopher mounds in mowed and previously disturbed areas (72.2% and 21.2%, respectively). These same 2 disturbance types accounted for 62.5% of the observed χ2 value. The remaining 2 disturbance types that were used relatively more often (i.e., construction and shrub removal) accounted for only 0.5% of the observed χ2 value. Of the disturbance types that were not apparently used by Maritime Pocket Gophers, developed and untouched habitat disturbances accounted for the majority (23.1% and 11.2%, respectively) of the remaining observed χ 2 value. Microhabitat assessment Maritime Pocket Gopher mounds were present on 28% (n = 196) of the transects on NAS Corpus Christi and NALF Waldron. Bare ground; percent cover of Guinea Grass, Buffel Grass, native grasses, and native forbs; soil pH, nitrate nitrogen, and phosphorus did not differ significantly (P > 0.05) between transects with pocket gopher mounds and those with no pocket gopher mounds (Appendix 3). Percent soil potassium, sand, and silt were strongly related to percent clay (r ≥ 0.86) and were not included in further analyses. Percent clay, percent litter cover, percent Kleberg Bluestem cover, and percent native grass cover differed (P < 0.031) between transects with and without Maritime Pocket Gopher mounds. The logistic regression model with litter and percent clay as independent variables and presence of pocket gopher mounds as the dependent variable was the best model to explain habitat selection by Maritime Pocket Gopher based on AICc weights (Tables 3, 4). Pocket gophers were most likely to be found on sandy soils with ≤15% litter. Relative probability of use of a site by pocket gophers declined dramatically with increasing soil-clay content (Fig. 1). Table 2. Occurrence of Maritime Pocket Gopher mounds on various habitat disturbances on Naval Air Station Corpus Christi and Waldron Airfield, Corpus Christi, TX. χ 2 = 121,585; P < 0001. Outcomes: P = preferred, A = avoided. % of Area of % observed Total total gopher mounds in each Confidence Disturbance area (m2) area Observed Expected area interval Outcome Construction 183,131.0 1.4 1970.0 1280.9 2.2 0.0211–0.0232 P Developed 4,047,333.6 32.0 577.1 29,278.0 0.6 0.0054–0.0066 A Manicured 919,021.1 7.3 1,979.4 6679.1 2.2 0.0208–0.0232 A Mowed 4,599,146.9 36.4 66,035.5 33,303.8 72.2 0.7184–0.7256 P Previously disturbed 665,795.5 5.3 19,425.4 4849.2 21.2 0.2087–0.2153 P Shrub removal 18,799.7 0.1 233.7 91.5 0.2 0.0016–0.0024 P Untouched 2,211,130.4 17.5 1272.8 16,011.4 1.4 0.0130–0.0150 A Total 12,644,358.0 100.0 91,493.9 91,493.9 100.0 Southeastern Naturalist J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 48 Discussion The distribution of Maritime Pocket Gophers was influenced by soil-particle size, plant community, type of habitat disturbance, and quantity of ground litter. Distribution of pocket gophers typically is limited by the quality, type, and nutrient content of soil (Huntly and Inouye 1988, Miller 1964). Because of their fossorial nature, Geomys spp. prefer sandy to loamy soils (Baker et al. 2003) presumably to facilitate the digging of burrows. The main habitat type where dredge spoils (i.e., clay soils) occurred was within the salt-marsh plant community, which was apparently avoided by pocket gophers in our study. Although some gopher mounds were located within this habitat type, most of the mounds occurred on the perimeter of the area. Our finding that Maritime Pocket Gophers apparently avoid areas with clay soils agrees with other researchers’ results, suggesting that clay soil represents a geographic barrier for immigration by this species (Connior et al. 2010; Davis 1940, 1974; Hoffman et al. 2007). Rezsutek and Cameron (1998) found a direct Table 3. Multiple logistic regression models with AICc values for predicting the relative probability of use by pocket gophers (n = 196 observations), Nueces County, TX, June 2006. AICc -2 Log Independent variables k AICc ΔAICc weight likelihood Clay, litter 3 205.82 0.00 0.430 199.70 Clay, Kleberg Bluestem, litter 4 206.83 1.01 0.260 198.62 Clay, litter, native grass 4 207.37 1.55 0.199 199.16 Clay, Kleberg Bluestem, litter, native grass 5 208.56 2.74 0.110 198.24 Clay, native grass 3 218.61 12.79 0.001 212.48 Clay 2 220.06 14.24 0.000 216.00 Clay, Kleberg Bluestem, native grass 4 220.29 14.46 0.000 212.08 Clay, Kleberg Bluestem 3 221.42 15.60 0.000 215.30 Kleberg Bluestem, litter 3 225.62 19.80 0.000 219.50 Kleberg Bluestem, litter native grass 4 225.71 19.89 0.000 217.50 Litter, native grass 3 226.87 21.05 0.000 220.75 Litter 2 228.21 22.38 0.000 224.14 Native grass 2 230.72 24.90 0.000 226.66 Kleberg Bluestem, native grass 3 230.76 24.94 0.000 224.64 Kleberg Bluestem 2 233.33 27.51 0.000 229.27 Null model 1 234.68 28.86 0.000 232.66 Table 4. Model-averaged parameter estimates, standard errors, an d confidence intervals. Adjusted Parameter Estimate Standard error standard error Lower 95% CI Upper 95% CI Intercept 0.420 0.359 0.361 -0.288 1.127 Clay -0.088 0.023 0.023 -0.134 -0.042 Litter -0.038 0.012 0.012 -0.062 -0.015 Kleberg Bluestem -0.019 0.021 0.021 -0.060 0.022 Native grass 0.003 0.004 0.004 -0.006 0.012 Southeastern Naturalist 49 J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 relationship between abundance of pocket gophers and dicots. Similarly, most Maritime Pocket Gopher mounds in our study were located in habitat types with the greatest abundance and diversity of forbs (i.e., native prairie and restoration fields). Although pocket gophers in our study more often used the Bermuda Grass monocultures, this habitat type was maintained in an early successional condition by frequent mowing. Habitat disturbance is potentially disruptive to mammalian populations, yet periodic mowing may be necessary to maintain the early successional vegetation that some mammals prefer or require (Slade and Crain 2006). Pocket gophers tend to be found in areas with high nutritional quality due to the high energy requirements of fossorial animals (Inouye et al. 1997). Mowing maintained habitat types at an early successional stage, and thus, kept vegetation at a young growing state with higher nutritive value. Therefore, Maritime Pocket Gophers in our study apparently preferred habitats that had a frequent mowing regime. In addition, more pocket gopher mounds occurred in areas where there was less litter cover, which coincided with mowed areas as opposed to undisturbed habitat types. Buechner (1942), Connior et al. (2010), and Philips (1936) obtained similar results for other species of pocket gophers and reported many more of them found on lawns than on areas free from anthropomorphic influences. Previous studies have noted the increase of pocket gopher density in grazed areas (Buechner 1942, Reichman and Baker 1972). However, the apparent preference of Maritime Pocket Gophers for grass areas due to their having been mowed has not been previously reported. The effects of mowing and grazing on grassland communities can be similar. Consequently, if the effects are similar and pocket gophers preferred grazed areas in previous studies (Buechner 1942, Reichman and Baker Figure 1. Relative probability of occurrence of a pocket gopher mounds (n = 196 transects) in relation to percent litter cover and percent clay, Nueces County, TX, June 2006. Southeastern Naturalist J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 50 1972), then it is reasonable to assume that pocket gophers also should prefer mowed over unmowed areas. Buechner (1942) assumed that grazing provided more food for gophers by keeping the vegetation in a productive stage of development, which in turn, increased the number of gophers. Pocket gophers may also be attracted to mowed or grazed habitats because of the greater root biomass of plants in those areas. However, this effect is dependent on the intensity of grazing or mowing. Below-ground biomass increases with increased mowing intensity (Dickinson and Polwart 1982) and grazing (Milchunas and Lauenroth 1993, Pearson 1965, Sims and Singh 1978). The proportion of roots in Maritime Pocket Gophers’ diet is unknown, but roots do dominate the diet of a related gopher species, Geomys attwateri Merriam (Attwater’s Pocket Gopher), for which up to 71% of its total diet consisted of roots (Williams and Cameron 1986). The apparent avoidance of shrubland and forested habitats (i.e., Honey Mesquite and Live Oak, respectively) by Maritime Pocket Gophers is probably due to the loss of on-ground primary productivity because of tree-canopy cover. Pocket gophers tend to prefer areas of high primary productivity (Inouye et al. 1997). Additionally, root mass of browse species may increase the difficulty of burrow construction. Because our study area had a history of shredding such habitat within the recent past (i.e., >15 years earlier), root-mass diameters of Live Oak and Mesquite often exceeded 2 m, as witnessed from the excavation of similar root masses from the nearby restoration habitat type. Maritime Pocket Gophers apparently avoided the developed, manicured, and untouched habitat disturbance types. Developed areas (i.e., asphalt, concrete, buildings, etc.) were used less often for the obvious reasons of lack of available food and reduced oxygen turnover rate. Small swaths of open ground occurred within developed areas and we occasionally found gopher mounds in such areas. However, it appeared that Maritime Pocket Gophers used developed areas mainly as travel corridors to get to an adjoining hospitable habitat. When we found mounds in developed areas, they usually occurred in a straight line along a concrete curb, sidewalk, or led directly underneath a road and continued on the other side within available open ground. Reasons for the avoidance of manicured areas are unclear. Habitat fragmentation and urbanization also may have played a role in the lack of Maritime Pocket Gophers in some areas. Some regions of potentially good habitat for Maritime Pocket Gophers (i.e. deep sandy soils, mowed native prairie) were devoid of gophers, but were surrounded on all sides by development (i.e., concrete, buildings, and asphalt). Such areas may have been too far removed for pocket gophers to gain access. Management implications Several management recommendations can be inferred from our research. Although the Navy must maintain the Intercoastal Waterway, placement of dredged soils—typically >70% clay, on top of sandy soils—should be avoided because the presence of clay will limits Maritime Pocket Gopher use. Second, abandoned airplane runways and dense oak secondary growth should be restored to native prairie to Southeastern Naturalist 51 J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 accommodate the overall habitat needs of this species. Corridors of suitable habitat should be established and maintained to encourage use by Maritime Pocket Gophers; however, minimum width and lengths of such corridors that will encourage use by Maritime Pocket Gophers wil need to be determined. Although this gopher species is under consideration for federal threatened or endangered status, the location of gophers on DOD lands places them in conflict with naval operations. As a result, additional research is needed to determine if Maritime Pocket Gophers can tolerate relocation in the event a “pest-management” relocation program becomes necessary. Acknowledgments Funding was provided by the US Navy. This is publication number 11-115 of the Caesar Kleberg Wildlife Research Institute. Literature Cited Andersen, D.C. 1987. Geomys bursarius burrowing patterns: Influence of season and foodpatch structure. Ecology 68:1306–1318. Baker, R.J., R.D. Bradley, and L.R. McAliley, Jr. 2003. Pocket gophers (Geomyidae). Pp. 276–287, In G.A. Feldhamer, B.C. Thompson, and J.A. Chapman. (Eds.). Wild Mammals of North America: Biology, Management, and Conservation. Johns Hopkins University Press, Baltimore, MD. 1216 pp. Barton, K. 2013. Package ‘MuMIn’. Available online at http://ftp.hearnet.ie. Accessed 29 June 2013. [Current version available at http://cran.r-project.org/web/packages/Mu- MIn/MuMIn]. Beyer, H. 2004. Hawth’s analysis tools for ArcGis 9. Available online at http://www.spatialecology. com/htools. Accessed 28 February 2013. Buechner, H.K. 1942. Interrelationships between the pocket gopher and land use. Journal of Mammalogy 23:346–348. Burnham, K.P., and D.R. Anderson. 2002. Model Selection and Multimodal Inference: A Practical Information-theoretic Approach. Springer, New York, NY. 353 pp. Connior, M.B., A.A. Kershen, R.E. Medlin, Jr., D.A. Elrod, D.B. Sasse, and T.S. Risch. 2010. Distribution and habitat attributes of an endemic subspecies of pocket gopher. American Midland Naturalist 164:217–229. Daubenmire, R. 1959. A canopy-coverage method of vegetational analysis. Northwest Science 33:43–64. Davis, W.B. 1940. Distribution and variation of pocket gophers (genus Geomys) in the United States. Bulletin of Texas Agricultural Experiment Station 590:5–37. Davis, W.B. 1974. The mammals of Texas. Bulletin of Texas Parks and Wildlife Department 41:1–294. Day, P.R. 1965. Particle fractionation and particle-size analysis. Pp. 545–567, In C.A. Black (Ed.). Methods of Soil Analysis: Part 1. Agronomy Monograms 9. American Society of Agronomy and Soil Science Society of America, Madison, WI. 860 pp. Dickinson, N.M., and A. Polwart. 1982. The effect of mowing regime on an amenity grassland ecosystem: Above- and below-ground components. Journal of Applied Ecology 19:569–577. Garcia, V. 2001. Ecological site characteristics of coastal sand. US Department of Agriculture, Natural Resources Conservation Service, Washington, DC. 160 pp. Southeastern Naturalist J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 52 Germano, D.J., G.B. Rathbun, and L.R. Saslaw. 2001. Managing exotic grasses and conserving declining species. Wildlife Society Bulletin 29:551–559. Hafner, D.J. 2000. Geomys personatus ssp. maritimus. In International Union for Conservation of Nature (IUCN), 2006 red list of threatened species. Available online at http:// www.iucnredlist.onz. Accessed 6 June 2008. Hoffman, J.D., J.R. Choate, and R. Channell. 2007. Effects of land use and soil textures on distribution of pocket gophers in Kansas. Southwestern Naturalist 52:296–301. Hosmer, D.W., and S. Lemeshow. 1989. Applied Logistic Regression. John Wiley and Sons, New York, NY. 307 pp. Huntly, N., and R. Inouye. 1988. Pocket gophers in ecosystems: Patter ns and mechanisms. Bioscience 38:786–793. Inouye, R.S., N. Huntly, and G.A. Wasley. 1997. Effects of Pocket Gophers (Geomys bursarius) on microtopographic variation. Journal of Mammalogy 78:1 144–1148. Johnson, D.H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65–71. Keating, K.A., and S. Cherry. 2004. Use and interpretation of logistic regression in habitatselection studies. Journal of Wildlife Management 68:774–789. Keeney, D.R., and D.W. Nelson. 1982. Nitrogen: Inorganic forms. Pp. 643–687, In A.L. Page (Ed.). Methods of Soil Analysis: Part 2. Agronomy Monograms 9. 2nd Edition. American Society of Agronomy and Soil Science Society of America, Madison, WI. 1159 pp. Larson, J., T.E. Fulbright, L.A. Brennan, F. Hernandez, and F.C. Bryant. 2010. Texas Bobwhites: A Guide to Their Foods and Habitat Management. University of Texas Press, Austin, TX. 294 pp. Manly, B.F.J., L.L. McDonald, D.L. Thomas, T.L. McDonald, and W.P. Erickson. 2002. Resource Selection by Animals: Statistical Design and Analysis for Field Studies. Kluwer Academic Publishers, Dordrecht, The Netherlands. 292 pp. Mehlich, A. 1978. New extractant for soil-test evaluation of phosphorus, potassium, magnesium, calcium, sodium, manganese, and zinc. Communications in Soil Science and Plant Analysis 9:477–492. Mehlich, A. 1984. Mehlich-3 soil-test extractant: A modification of Mehlich-2 extractant. Communications in Soil Science and Plant Analysis 15:1409–1416. Milchunas, D.G., and W.K. Lauenroth. 1993. Quantitative effects of grazing on vegetation and soils over a global range of environments. Ecological Monographs 63:327–366. Miller, R.S. 1964. Ecology and distribution of pocket gophers (Geomyidae) in Colorado. Ecology 45:256–272. Moruzzi, T.L., T.K. Fuller, R.M. DeGraaf, R.T. Brooks, and W. Li. 2002. Assessing remotely triggered cameras for surveying carnivore distribution. Wildlife Society Bulletin 30:380–386. Murphy, J., and J. P. Riley. 1962. A modified single-solution method for determination of phosphates in natural water. Analytica Chimica Acta 27:31–36. Natural Resources Conservation Service (NRCS). 1960. Soil survey of Nueces County, Texas. US Department of Agriculture, Washington, DC. 65 pp. Neu, C.W., C.R. Byers, and J.M. Peek. 1974. A technique for analysis of utilization-availability data. Journal of Wildlife Management 38:541–545. Pearson, L.C. 1965. Primary production in grazed and ungrazed desert communities of eastern Idaho. Ecology 46:278–285. Philips, P. 1936. The distribution of rodents in overgrazed and normal grasslands of central Oklahoma. Ecology 17:673–679. Southeastern Naturalist 53 J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 Proulx, G. 2002. Effectiveness of trapping to control Northern Pocket Gophers in agricultural lands in Canada. Proceedings of 20th Vertebrate Pest Conference 20:26–31. Reichman, O.J., and R.J. Baker. 1972. Distribution and movements of two species of pocket gophers (Geomyidae) in an area of sympatry in the Davis mountains, Texas. Journal of Mammalogy 53:21–33. Rezsutek, M.J., and G.N. Cameron. 1998. Influence of resource removal on demography of Attwater’s Pocket Gopher. Journal of Mammalogy 79:538–550. Schofield, R.K., and A.W. Taylor. 1955. The measurement of soil pH. Soil Science Society of America Proceedings 19:164–167. Sims, P.L., and J.S. Singh. 1978. The structure and function of western North American grasslands. IV. Compartmental transfers and energy flow within the ecosystem. Journal of Ecology 66:983–1009. Slade, N.A., and S. Crain. 2006. Impacts on rodents of mowing strips in old fields of eastern Kansas. Journal of Mammalogy 87:97–101. Steel, R.G.D., and J.H. Torrie. 1980. Principles and Procedures of Statistics. 2nd Edition. McGraw-Hill, New York, NY. [PP?]. Williams, L.R., and G.N. Cameron. 1986. Food habits and dietary preferences of Attwater’s Pocket Gopher, Geomys attwateri. Journal of Mammalogy 67:489–496. Williams, S.L. 1982. Geomys personatus. Mammalian Species 170:1–5. Williams, S.L., and H.H. Genoways. 1981. Systematic review of the Texas pocket gopher Geomys personatus (Mammalia: Rodentia). Annals of Carnegie Museum 50:435–473. Witmer, G.W., R.E. Marsh, and G.H. Matschke. 1999. Trapping considerations for the fossorial pocket gopher. Pp. 131–139, In G. Proulx (Ed.). Mammal Trapping. Alpha Wildlife Research and Management Ltd, Sherwood Park, AB, Canada. 241 pp. Southeastern Naturalist J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 54 Appendix 1. Plant community categories for NAS Corpus Christi and NALF Waldron Station Plant community Description Salt marsh Soil with high salinity, >95% salt-tolerant plants Bermuda Grass >95% Bermuda Grass Bermuda Grass/Guinea Grass less than 95% Bermuda + less than 95% Guinea grass, but >95% Bermuda Grass and Guinea Grass combination Debris-native/Guinea Grass less than 95% native prairie + less than 95% Guinea Grass, but >95% native prairie and Guinea Grass combination with scattered concrete and metal debris Golf course >95% golf course grass species Guinea Ggrass >95% Guinea Grass Guinea Grass/Buffel Grass <95% Guinea Grass + <95% Buffel Grass, but > 95% Guinea Grass and Buffel Grass combination Gulf Cordgrass >95% Gulf Cordgrass Live Oak >95% Live Oak forest Honey Mesquite/Guinea Grass less than 95% Honey Mesquite forest + less than 95% Guinea Grass, but >95% Guinea Grass and Honey Mesquite forest combination Native prairie >95% mixture of native grasses and forbs Restoration field >95% mixture of native grasses with Guinea Grass and Bermuda Grass Softball field/park >95% softball field grass species St. Augustine Grass >95% St. Augustine Grass Vegetated runway >95% concrete slab with vegetative growth on top and within cracks of concrete Southeastern Naturalist 55 J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 Appendix 2. Habitat-disturbance categories for NAS Corpus Christi and NALF Waldron Station in southern Texas. Habitat disturbance Description Construction Active construction site Developed Asphalt, concrete, or structure Manicured Planted lawns, fertilized, watered, mowed weekly Mowed Mowed (>2 weeks less than 6 months old) Previously disturbed Previously mowed (6–12 months old) Shrub removal Shredded Live Oak forest with tree bases uprooted Untouched Undeveloped and not mowed (>12 months) Southeastern Naturalist J.D. Cortez, S.E. Henke, D.W. Wiemers, T.E. Fulbright, D.B. Wester, and R. Riddle 2015 Vol. 14, No. 1 56 Appendix 3. Means and lower (LCI) and upper (UCI) 95% confidence intervals and results of t-tests for vegetation and soil variables, bare ground, and distance from roads at transects with ≥1 pocket gopher mound present and transects with no pocket gopher mounds present, Nueces County, TX, June 2006. Used sites (n = 55) Unused sites (n = 141) Difference (n = 194) Variable Mean LCI UCI Mean LCI UCI Mean LCI UCI t P Vegetation Bermuda Grass (%) 17 9 25 12 8 16 5 -4 14 1.04 0.302 Buffel Grass (%) 3 less than 1 7 3 1 4 1 -3 5 0.50 0.618 Guinea Grass (%) 4 less than 1 7 8 4 12 -4 -10 1 -1.63 0.106 Kleberg Bluestem (%) 1 -1 3 5 2 8 -4 -7 1 -2.18 0.031 Native forbs (%) 21 16 26 17 14 21 4 -3 10 1.12 0.263 Native grasses (%) 54 43 65 38 32 45 16 3 29 2.48 0.014 Litter (%) 7 5 10 17 12 21 -9 -14 -5 -3.84 less than 0.001 Soil Clay (%) 8 7 10 15 13 17 -7 -10 -4 -5.19 less than 0.001 NO3-N (μg/ml) 6 4 8 6 4 7 less than 1 -2 3 0.25 0.804 K (μg/ml) 121 98 143 232 199 264 -111 -151 -72 -5.60 less than 0.001 pH 7 7 8 7 7 8 less than 1 less than 1 less than 1 0.48 0.630 P (μg/ml) 24 9 39 15 13 18 8 -7 23 1.10 0.276 Sand (%) 86 84 88 75 72 78 11 7 15 5.92 less than 0.001 Silt (%) 6 5 6 10 9 11 -4 -6 -3 -5.61 less than 0.001 Other Bare ground (%) 4 2 6 6 3 9 -2 -6 1 -1.32 0.188 Distance from roads (km) 65 51 80 57 48 65 8 -8 25 1.03 0.303