Seasonal Variation in Hematology and Blood Plasma Chemistry Values of the Timber Rattlesnake (Crotalus horridus)

Hematology, biochemical, and body condition index analyses are useful tools for diagnosing diseases at early stages (Strik et al. 2007) and assessing the relationship between a population's health and the quality of habitat (Allender et al. 2006). Furthermore, changes in environmental conditions or disease can cause stress, which can be detected in hematologic data (Oppliger et al. 1998), especially if they change over time (Peterson 2002). The monitoring of individuals' data over time is necessary for detecting significant deviations and is especially useful when making management decisions for species of conservation concern (Christopher et al. 1999). However, baseline hematology, biochemical, and body condition data from healthy, wild individuals are rare or nonexistent in many wildlife species.

The timber rattlesnake (Crotalus horridus) is the most widespread rattlesnake in the eastern US (Conant and Collins 1998), and it is imperiled at some level across much of its range (MacGowan et al. 2009); however, no baseline hematologic and plasma biochemical temporal reference intervals exist for this species. To aid in its conservation, we report baseline and temporal hematologic and plasma biochemical reference intervals for this species.

We repeatedly captured and sampled 13 (five male and eight female) healthy, radiomarked (MacGowan and Walker 2013), free-ranging timber rattlesnakes over the course of an active season (April–September 2012) in south-central Indiana. All individuals had survived at least one winter after transmitters were implanted, indicating they had recovered. We sampled each individual between 6:00 am and 8:00 pm at each of three sampling periods: spring (6–18 April), midsummer (9–25 July), and autumn (7–23 September). We captured and restrained snakes using snake hooks and a snake tube. To minimize effects of handling stress on our results, we collected blood samples immediately upon capture (Kimble and Williams 2012). We drew 0.5–1.0 mL of blood from the caudal vein using a lithium-heparinized 25-gauge, 1.9-cm needle. Immediately after blood collection, we made three blood smears per individual (Sykes and Klaphake 2008) and fixed them with 100% methanol in the field. We stored the remaining blood in lithium-heparinized tubes on ice for later laboratory analysis. We released snakes immediately at the point of capture. Animals were handled according to Purdue University Animal Care and Use Protocol 07-037.

We processed blood within 12 hr of field collection. To determine the packed cell volume (PCV), we centrifuged blood at 38,000 × G for 120 sec and 37,000 × G for 30 sec to separate solids and plasma. We measured total solids using a TS400 refractometer (Reichert, Inc., Depew, New York, USA). We stained blood slides with Giemsa and manually scanned them at 100× magnification to search for parasites and inclusions. Using an Avian/Reptile Profile Plus rotor in a Vetscan classic analyzer (Abaxis, Inc., Union City, California, USA), we analyzed plasma concentrations of 12 plasma chemistry parameters.

We measured snout-vent length (SVL) and mass, and determined sex based on tail length and cloacal probing. We determined body condition indices (BCI) by regressing the cubed root of the mass over SVL and plotted the resulting per-season mean residuals against the average SVL for each season to obtain mean BCI values (Williams et al. 2009). We used a Shapiro-Wilk normality test to confirm the normality of the data. To determine effects of season on hematologic parameters and biochemical concentrations, we used repeated-measures analysis of variance followed by Tukey's honest significant difference test. Finally, we regressed plasma biochemistry values on corresponding BCI values (Baker 1992). We conducted all statistical tests in R (version 2.15.0; R Development Core Team 2012) and accepted significance at α = 0.05.

We screened 39 blood smears (three per individual) and found no blood parasites or inclusions for any individual or sampling occasion. During biochemistry analysis, we found little evidence of reduced plasma sample quality due to lipidemia or hemolysis (1% of analyses failed). The majority of samples yielded complete plasma biochemistry data (92%; Table 1). Calcium was significantly lower in the spring than summer or autumn (F_2,16 = 10.75, P<0.05), whereas total protein (F_2,24 = 5.91, P<0.05) and glucose (F_2,22 = 9.44, P<0.05) were significantly elevated in the summer (Table 1). Calcium levels are often elevated during follicular development (Campbell 2006), which may account for the observed higher plasma calcium levels in summer and autumn as the breeding season progressed, though there were no significant differences between sexes during any season. Elevated levels of total protein can be caused by inflammation or dehydration (Campbell 2006). Our study site experienced a significant drought in summer 2012, but it had subsided by the autumn sampling period (http://droughtmonitor2.unl.edu), making dehydration a more plausible explanation than disease for high protein levels. Likewise, sodium and potassium were within the normal reported levels for snakes (Campbell 2006). Glucose levels in reptiles vary greatly among species and are dependent on nutrition and temperature (Campbell 2006) and can be elevated in response to stress (Moore and Jessop 2002). Thus, the higher glucose levels we observed during summer likely reflect natural variation due to changing resources (e.g., prey availability) and drought.

Levels for BCI, PCV, and total solids were constant across the active season (F_2,24 = 0.24, P = 0.79; F_2,24 = 1.67, P = 0.21; F_2,24 = 2.37, P = 0.12; respectively). No significant relationships between plasma biochemistry values and BCI were found, suggesting that BCI may not be a good predictor of variation in plasma biochemistries or that this method of calculating BCI may not be a good predictor of health in this species (see also Peig and Green 2010). Variations in hematologic and blood plasma parameters among seasons represent natural variation across the active season; however, they also suggest avenues for further research into the health of timber rattlesnakes.

Funding for this project was provided by the Purdue University Undergraduate Research Award from the College of Agriculture and the Indiana Department of Natural Resources, Division of Fish and Wildlife, State Wildlife Improvement Grant T07R09. We thank G. Marin, M. Hamilton, V. Wuerthner, E. Estabrook, B. Zinman, and C. Fella for assistance with fieldwork; J. Doyle and N. Lichti for statistical assistance; Purdue University veterinarians L. Corriveau, D. Huse, and S. Thompson for transmitter implantations; and the Williams Laboratory for reviewing previous versions of this manuscript. We collected all samples under Indiana Department of Natural Resources permit (12-0101) and Purdue University Animal Care and Use Protocol 07-037 and amendments.