The current method of systematically triennially surveying the entire refuge (Game Management Units 45A, B, and C – not including the New Water Mountains, which are surveyed with their respective hunt unit, 44B) by helicopter was initiated in 1992 as a result of funding shortfalls that precluded the continuation of the annual aerial surveys started in 1981. The Kofa and Castle Dome mountains are divided into blocks using geographic features in order to standardize areas flown and level of effort in each area. Sheep populations are estimated using the “Kofa Group Size Estimator” developed by Arizona Game and Fish in 1988.
Surveys from 1987-1991 were based on the same block system as is used now, but a sample of 50% of the refuge was surveyed annually and the estimate was extrapolated to the whole. This tended to result in more variable estimates than complete surveys. From 1981-1987, complete aerial surveys were conducted but the current survey blocks had not been defined. Before 1980, population estimates were based on animals observed during foot and aerial lamb surveys and water hole counts. Thus, data from before 1992 may not be comparable to data obtained after 1992. Population estimates for 1981-2006 are given in Figure 1. The 2006 survey was the first time since 1980 that the population estimate was below 600 bighorn and represents the sharpest drop recorded. Confidence intervals for these population estimates are wide.
If as research indicates, drought during the past 10 years is the primary cause of sheep decline, it must have affected adult mortality equal to lamb production or lamb survival, since lamb to adult ewe ratios did not decline. Population and age/sex ratios are estimated during each survey. Because surveys take place when the majority of lambs are close to one year old, the counts of lambs and yearlings are considered estimates of recruitment. However, actual mortality rates and productivity are unknown. Long-term lamb-to-ewe ratios, determined during fall aerial surveys when lambs are 6 to 12 months old, have averaged about 19 lambs per 100 ewes. During the recent decline, lamb ratios were 22 in 2003 and 25 in 2006, higher than the long-term average. Drought should also have affected bighorn habitats in game management units surrounding the Kofa equal to that on the refuge. Population estimates in those areas do not exhibit the same decline as on the Kofa
Determining the cause of mortality is vital to understanding which factors are contributing the most to population decline so that those factors can be addressed. Estimating survival rates and determining the cause of bighorn mortality can be accomplished by radio-collaring sheep. Although collaring 20-30% of the entire population would be ideal, (E. Rubin, pers. comm.) expending the effort and resources to collar 80-100 sheep would not be practical and might cause unnecessary stress on the population. Restricting the monitoring to just the Kofa Mountains, where declines have been most marked, would limit the target population to about 140 estimated adult females. Collaring 20-30% of these would represent 30 to 40 sheep and should allow for adequate mortality estimation. A combination of satellite GPS and VHF radio collars would be the most efficient way to obtain a wide range of data.
Kofa bighorns viewed from survey helicopter
Figure 1: Bighorn sheep population estimates
for Kofa NWR (Unit 45), 1981-2008*.
* Population estimates from 1981-1986 and 1987-1991
were obtained using different survey methods than
estimates from 1992-2008. Data from before
1992 are not directly comparable to data obtained
after 1992 but can be used to show trends.