RESULTS
Models that were painted to color-match in the urban environment had higher COI scores than those that were not color-matched (urban gray vs urban red, t = 2.59, df =18, P = 0.019), and models that were painted to color-match on the ranch had higher COI scores than those that were not color-matched (ranch gray vs ranch red), (t = 4.86, df = 18, P = 0.00013) (Fig. 3).
A total of 61 predation events were recorded over both sampling periods representing 12.1% of all models. We found no difference in the total number of predation events occurring between early summer (N = 30) and late summer (N = 31) (P = 0.89) (Table 1). Four predation events occurred on the controls over both sampling periods, all of which were located on the ranch. Predation events were higher on the models than on the controls, both overall (χ2 = 3.74, df = 1, P = 0.05) and on the ranch (χ2 = 3.24, df = 1, P = 0.04). We observed more predation events on models at the ranch (N = 60) over the two sampling periods than in the towns (N = 1) (P = 1.12x10-18) (Table 1). During the early summer, 1 model was attacked in town and 29 were attacked on the ranch (P = 1.05x10-8) (Table 1). During the late summer, no models were attacked in town and 31 were attacked on the ranch (P = 1.13x10-10). Hatchling models were attacked less than juvenile or adult models (χ2 = 7.08, df = 2, P = 0.029) (Table 1). Models that had evidence of being disturbed by non-predators (i.e. hoof marks) (n=2) or went missing (n=6) were only on the ranch and were not counted as predation events and excluded from the analysis.
We found a significant difference in the number of predation events by predation category: birds, rodent, other, and unknown (χ2 = 33.24, df = 3, P = 2.86x10-7) (Fig. 4). We also found that the number of attacks by birds in the early summer (N = 13), when lizard models were not painted to color-match the red soils on the ranch was significantly higher than during late summer (N = 5) when models were painted to background color-match (χ2 = 3.8, df = 1, P = 0.05) (Fig. 4). We did not see any difference in color-matching in the remaining categories on the ranch rodents:( χ2 =2.9 , df = 1, P = 0.09), other (χ2 =1.5 , df = 1, P = 0.23) or unknown (χ2 = 0.1, df = 1, P = 0.78) (Fig 4). The attack in June on a control piece had distinct peck marks, whereas the three remaining controls that were attacked in August had conspicuous half-moon shape bite marks (Fig. 5). Our controls during the second round were painted to color-match the soils on the ranch and as a result resembled dried prickly pear (Opuntia spp.) pads and fruits; which may have attracted Texas tortoises (Gopherus berlandieri ) (Fig. 5). We frequently encountered Texas tortoises eating both the fruits and pads of Opuntia spp. on the Dimmit County ranch and the bite marks on the controls were similar in size and shape to the tortoise bites on cactus pads (Fig. 5).