Materials and Methods
We investigated relationships between chigger mite loads and lizard growth rates in yearling juveniles of Sceloporus undulatus in the New Jersey Pinelands National Reserve during June and July of 2016 and 2019. We conducted our studies in Colliers Mills Wildlife Management Area (Burlington County, New Jersey, USA; 40.1° N, 74.4° W), which we chose because of its relative abundance of lizards. Our study site was a 0.045 km2 patch of oak-pine forest bounded by open, mowed fields and man-made sand paths, with a variably open canopy, sparse understory due to prescribed burns, and abundant tree debris on the forest floor.
Eggs in this population are laid from late May into June and begin to hatch in late July. Depending on weather, the neo-natal activity season for hatchlings can extend into mid-November. Lizards emerge from winter inactivity as yearlings in March of the next year and grow rapidly to reach the size of reproductive maturity by the end of the first full activity season. Following a second winter of inactivity, lizards re-emerge as reproductively mature adults at approximately 20 months of age, in the spring of their second full activity season (Haenel & John-Alder, 2002). Males and females are initially indistinguishable in body size, but sexual divergence in body size occurs most rapidly in yearlings during June/July of the first full activity season, coinciding with rising levels of plasma testosterone in males and faster growth rates in females (Skelly & John-Alder 2002; Cox et al. 2005). With the attainment of the size of reproductive maturity at the end of the first full (yearling) activity season, females are approximately 10% larger than males (Haenel & John-Alder, 2002). In both years of our study, we applied mark-recapture techniques to sample the same lizards repeatedly for measurements of body size and mite loads. In 2016, we found 55 yearling females and 70 yearling males. In 2019, we found 42 yearling females and 41 yearling males. Of these, we made measurements and counted mites on 39 females and 55 males in 2016, and 28 females and 20 males in 2019. All measurements each year were collected by the same individual (NBP in 2016, HC in 2019) to prevent possible measurement biases in SVL, body mass, or mite counts. We captured lizards with a slip-noose or by hand. Upon capture, we measured snout-vent length (SVL, mm) using a ruler and body mass to 0.1 g using a Pesola spring scale. We determined sex by the presence (male) or absence (female) of enlarged post-cloacal scales, and we unambiguously distinguished yearlings from older age-classes by their smaller size during the month of June (Haenel & John-Alder, 2002). Each lizard was given a unique toe clip for permanent identification and a unique dorsal paint mark just anterior to the base of the tail (a region not preferred by mites) for visual identification.
To determine mite loads, we used a 10X hand lens in the field to count the number of mites on each lizard at the time of capture. In 2016, we counted mites on recaptured lizards at weekly intervals from June 9 through July 14, although not all lizards were recaptured every week. In 2019, we counted mites on recaptured lizards on a less structured schedule, with averages of 1.6 ± 0.1 counts per lizard in June (median: 2; range: 1 to 3) and 1.9 ± 0.1 counts per lizard in July (median: 2; range 1 to 4). Average intervals between counts were 10 ± 0.7 days in June 2016 (median: 12 days) and 13 ± 1.0 days in July 2016 (median: 13 days).
In 2019, we assessed the accuracy of mite counts by comparing the number of mites counted by eye in the field to the number of mites counted by using ImageJ (National Institutes of Health, USA) to analyze photographs of the same lizards taken at the time of mite counts in the field with a Nikon D7200 digital camera and a Sigma 105mm F2.8 EX DG OS HSM Macro lens. Mite loads determined by these two methods ranged from 7 to 424 mites/lizard and were highly correlated (Supporting Figure 1; n = 19; slope: 1.06 ± 0.05, 95% CI: 0.94–1.17, R2: 0.997).
For a direct comparison of 2016 versus 2019, we selected comparable three-week periods: 22–24 June through 13-14 July 2016 versus 24–27 June through 15–17 July 2019. These three-week periods included measurements of SVL and mite counts taken in weeks 3, 4, 5, and 6 of the 2016 study (Pollock & John-Alder, 2020) and in two intensive recapture efforts of 2019. Intervals between measurements of SVL, thus time intervals for calculating growth rate, invariably differed due to uneven recapture success. In 2016, we demonstrated that growth rate calculated as ((SVL2 -SVL1) / (time in days)) was equal to growth rate computed as the slope of SVL plotted as a function of date of recapture for each lizard (Supporting Figure 2; slope: 1.04 ± 0.07, 95% CI: 0.91–1.17, F1,54 = 253.28, p< 0.001). In 2019, we demonstrated that growth rate calculated as ((SVL2 -SVL1) / (time in days)) was independent of the number of days between measurements of SVL (Supporting Figure 3; slope: -0.001 ± 0.003, F1,41 = 0.12, p = 0.736). To analyze relationships between growth rate and mite load separately for the months of June and July, we averaged each lizard’s mite counts taken within the focal periods for each month. For example, June’s mite count in 2016 was the average of mites counted in weeks 3 and 4 of that study, or the single mite count for lizards captured in only one of those weeks.