Am, Midi, Nat. 160:278-288. Diel Movement Behavior of the Stripe-necked Musk Turtle. (Stemotherus minor peltifer) in Middle Tennessee. JOSHUA R. ENNEN' .
Am, Midi, Nat. 160:278-288
Diel Movement Behavior of the Stripe-necked Musk Turtle (Stemotherus minor peltifer) in Middle Tennessee JOSHUA R. ENNEN' .AND A. FLOYD SCOTT Center jorField Biology. Aus/in Peay State University, Clarksville, Tennessee 37044
ABSTRACT.—Few studies have used radiotelemetry to focus directly on the diel behavior of turtles. We used radiotelemetry to relocate 10 Stripe-necked Musk Turtles {Stemotlierus minor pdtifiir) every 2 h over a 24 h cycle in Middle Tennessee, Movements during the 24 b cycle were monitored on seven occasions between 27 Jul, and 4 Nov. 2004. Stem,otherus minor peltifer were often inactive and remained sedentary throughout a 24 h cycle; 87% of the total relocations revealed no movement. Overall, S. m. peltifer'^ diel movement behavior was restricted to evening and nocturnal hours. Frequency data suggested a uni-modal distribudon of movement during evening and midnight periods (1700-0159 h). Similarly, the mean distance traveled showed a uni-modal distrihution with a peak in the evening period of a 24 h cycle. Based on the data set witli only light and dark categories, males were found to move more frequently in daylight relative to females. A decrease in movement (frequency and distance) occurred with the changing of the .seasons from summer to winter. Probably due to temperature change, nocturnal and crepuscular hehaviors were highest in Jul. and Aug., while during the autumn months, there was little preference for daylight or darkness. This phenomenon has been reported in several Stemotherus species; but, to our knowledge, this is the first report for S. minor.
INTRODUCTION
Chelonian movement patterns are strongly correlated to life history and ecology, and are subdivided into two categories: spatial (intrapopulation and extrapopnlation) and temporal (diel, seasonal and sporadic) movements (Gibbons et al, 1990). A particular diel pattern could reflect an adaptation of timing activities to coincide with temperature regimes that wonld increase efficiency of physiological processes and locomotion (Gourley, 1979). In turdes, this is predominantly evident in dinrnal species, which bask to regulate most of their physiological processes. However, this does not explain nocturnal behaviors of other ttirde .species. The evolution of nocturnal behaviors could have been influenced by compeddon rather than by physiological demands. Within a particular community, different diel behaviors may be the mechanism temporally .separating turtles and other organisms occupying the same guild (Gourley, 1979; Park, 1940), thereby alleviadng interspecific competition pressures (Gourley, 1979). Three diel patterns appear to have evolved to promote this temporal separation within a community: diurnal, crepusculai" and nocturnal. Stemotherus minor peltifer (Smith and Glass, 1947) is a highly aquatic turde restricted to the southeastern United States. Previous studies have reported all categories of diel behavior (ditirnal, nocturnal and creptiscular) in the genus Slemothenis (Smith and Iverson, 2004; Bancroft et ai, 1983; Graham and Hutchinson, 1979; Lagler, 1943; Mahmoud, 1969; Ernst, 1986; Dodd el ai, 1988; Dodd, 1986). However, most of these .studies used trapping techniques restilting in the collection of diel data without the aid of radiotelemetry. Radiotelemetry is a common tool used in studying movement patterns, home-range sizes and habitat use of freshwater turtles, but to our knowledge only two studies (Dodd et al. Corresponding author present address: University of Southern Mississippi, Hattieshurg 39406; e-mail: jo,sh 1 ia.ennen@usm .edu
278
2008
279
ENNEN & SCOTT: MUSK TLIRTLE BEHAVIOR
Legend • Wliileoak Creek Tennessee River
FIG. 1,—Map of study site and the counties encompassing Whiteoak Creek in Middle Tennessee where radio-tracked Stemotherus minor pelifer in 2004. The star indicates location of study site
1988; Dodd, 1986) have applied radiotelemetry to collect movement data on species in the genus Stemotherus. We used radiotelemetry to investigate basic diel movement behavior of S. m. peltifer in western Middle Tennessee. Our objective of tbe study was to describe tbe bi-hourly movement behavior within a 24 h period, Becau.se others have reported a variety of diel behaviors for the genus Stemotherus, we hypothesize that there will be no .significant difference in tnrtle movement behavior between darkness, daylight or crepuscular period.s when con.sidering (1) frequency of movement, (2) distance traveled, (3) gender frequency of movement and (4) gender distance traveled. MATERIAI-S AND METHODS
Study .site.—The distribution of Stemotherus minor (Loggerhead Musk TurUe) is restricted to seven southeastern states: Florida, Georgia, Alabama, Mississippi, Tennessee, North Garolina, Virginia (Iverson, 1977; Ernst i-/a/., 1994; Conant and GoIIins, 1998). However, the subspecies, S. m. peltifer (Stripe-necked Musk Turtle), distribution i.s more restricted to the northern and western portions of the range (Ernst et al, 1994; Iverson, 1977; Conant and GoIIins, 1998), The S. m. peltifer population we sampled occurred in the Whiteoak Greek drainage of Houston and Humphreys counties, Tennessee (Fig. 1), 120 km north of the nearest known populadon in Lauderdale County, Alabama (Mount, 1975). The site for this study (36'13'30"N, 87'46'18"W) was located between Gander Branch and Grice Ford bridges on Wlñteoak Greek (Fig. 1), which flows for 42 km and drains parts of Dickson, Houston and Humphrey.s counties before emptying into Kentucky Lake (impounded lower reaches of the Tennessee River) at river mile 82. The area drained by Whiteoak Greek is located in the Western Highland Rim ecoregion, which is part of the Interior Plateau (Amwine *•/ al, 2000).
280
T H E AMERICAN MIDLAND NATLÍRAI^IST
160(2)
TABLE 1.—Mass, number of days tracked, number of days inactive during tracking period, the percentage of days that each indi\idtial was inactive, tlie longest distance traveled during a 2 h interval, lotal number of relocation points for each individual, the number of relocation points without movement and the percentage of the total relocation points involving no movement for Stemotherus minoj peltifer smdied in Whiteoak Creek, Humphreys County, Tenne.ssce
11) (gender)
Mass (^)
No. davs tracked
11-3 F 1-11 F 10-8 F 6-0 F 2-8 F 2-1 M Il-IOM 22M 8-8 M 8-1 M
146 237
7 7
162
6
158 169 169 110 105 162 166
7 7
Mean
158.4
No, days inactive 3 1 3
% Days inactive
longest No, relocation distance without (m) No. relocations movemenl
5 4
4 1 4 4 2 I
42.9 14,3 50 28.6 57.1 25 57.1 57.1 40 25
5 28 35 16 30 33 36 52 61
6.1
2.5
39.71
30,3
4 7
7
2
7
83
% Nonmovement rt'lotalion
87 87 74 87 87 53 87 87 65 53
64 73 80 42 74 74 56 44
92 79.2 85.1 85.1 86.2 83
76.7
66.4
86,6
74
95.4 85.1 86.5
83.9
Telemetry study.—We hand-taught turtles during daylight hours (0700-1800 h) while wading, snorkeling or canoeing. We fitted 10 individuals (5 males, 5 females) with radio transtnitters (Wildlife Materials International model: SOPB-2190), and we monitored with a telemetry receiver (AVM Instruments Company, Ltd mode!; LA12-DS) and receiving anienna (AVM Instmments Company, Ltd model; M-Yagi). The particular mode! of trati,smitter we tised was selected becaase of its compact size (33 mm X 13 mm X 8 mm), weight (4.6-5.0 g) and battery longevity (-248 d). In addition to tbe SOPB-2!90 transmitter, we attached a Thermochron iButton® (Model DS 192K^F5) thermal sensor to the carapace of each study animal. We applied both the tbermal sensors ¡md uansmitters to the carapace with PC^uperepoxy® adhesive and allowed them to dry for at least 12 h. The total package weight (8 g) ranged from 3.2% to 7.6% of the mass of the individual turdes studied. We used these dime-sized (-17 mm diam X 6 mm height and -3.1 g) thermometers to record temperature (accurate to 0.1 G) and dme (±2 s per mo) and were used concurrendy with another study. The sample size was influenced by the natural density of this species in Wliiteoak Creek. Furthermore, the logistics of relocating individuals every otber hotir restricted the number of ttirltes we could monitor. The limited number of days tracked was based on uansmitter failure and the difficulty of capturing individuals tofitwith a transmitter. We monitored six individuals for 7 d, one for 6 d, one for 5 d and two for 4 d (Table 1). This effort involved obtaining radio fixes of each individual every 2 h over the 24 h cycle. Once we determined the position of a turde. the position was recorded with a Trimb!e^^' GPS receiver and a Tiimble Survey Controller'^'^. Initially the purpose of this study was to determine U" there was a movement preference for dark or light hours. If the turtle had moved, we categorized the tnovement as dark or ligbt by determining the amount (if light in a given 2 h intenal and the time-of-day. If a 2 h interval had more than an hour of darkness or daylight, it was considered a nocturnal or diurnal interval, respectively. After we analyzed the two categories (light and dark), we added a crepuscular categoiy to include 2 h intervals that encompassed periods of dusk and dawn.
2008
ENNEN & SCOTT; MLÎSK TURTI.E BEHAVIOR
281
In the lab, we determined straight-line distance between relocation points using ESRI ARGGIS 9 software and verified witli an online distance calculator. To further clarify the did behavior, we tested for a dme-of-day preference for movement by grouping the time of each relocation fix into one of six categories; early morning (0200-0559 h), morning (0600-0959 h), midday (1000-1359 b), afternoon (1400-1759 h), evening (1800-2L59 b) and midnight (2200-0159 h). On 27 Jul. and 12 Atig. no data were taken between 0700 and 0900 h; on 16 Aug. and 2 Sept. no data were taken between 0600 and 0800 h. Statistics.—^We used binar> logistic regression model (BLR) and general linear mode! (GLM) analyses taking into account (he repeated measures design ttsing an overall alpha of 0.05. To assess the population's overall and gender-specific frequency of movement data, we tised two binary logistic regression models with the individual turtle nested witbin gender, gender and lighting period and an interaction term between gender and lighting period as predictors. We used another BLR model with the individual turtle nested witbin gender, gender and dme-of-day as predictors and an interaction term between gender and time-ofday was run to determine the propensity to move at a certain time-of^lay. To determine if movement frequencies sbifted with time, we used two BLR models witb three predictors (turde ID, date and lighting period) and an interaction term between date atid ligliling period. The non-normal nature of the distance data was due to the large number of times animals exhibited no movement-s, tbus skewing the data. To account for tbese zeros and to lessen the skevvness, we calculated mean distance of each individual turde within lighting intervals (light, dark, crepuscular) and time-of-
3 O (D
0)
: an Introductory Biology of Amphibians and Reptiles, 2nd edition. Academic Press, San Diego, California. Ei) 9 NOVEMBER 2007
ACCEPTÏD 2 APRIL 2008
