Given Caprinae life history and plausible combinations of mean recruitment and adult female survivorship, we evaluated population persistence and estimated population MVP. The values describing adult female survivorship and recruitment, plus the variability we employed match values found in other populations of Caprinae. We do not pool data across different Caprinae populations or species. Our approach and results directly inform the conservation and management of many Caprinae, especially those for which the acquisition of demographic data remains beyond reach.
Our work embodies the characteristics of a high-quality PVA: clear objectives, appropriate demographic data, model structure matching species life histories, stochasticity, examination of extinction probability, appropriate time interval, use of mean values and associated variability6. As with most ecological models, the quest for more data remains problematic, not debilitating, and is addressed by creatively and aptly using existing information to generate meaningful results3.
Wildlife agencies generate lamb:adult female ratios from Caprinae surveys, recognizing that yearlings can be mistaken for adult females, causing miscounts. Excluding yearlings from the ratio’s denominator assumes that no miscounts are occurring, yet an unknown and inconsistent number of yearlings remain in the adult female category across survey events. For these reasons, surveyors of other species, like Dall’s sheep and caribou, pool counts of yearlings and adult females, generating lamb:“adult female-like” ratios instead15,23,24,25.
Managers of Caprinae populations can follow these precedents and produce lamb:(adult female + yearling) ratios. Consistency would help standardize methods for building comparisons and meta-analyses across populations of Caprinae, while reducing variability across surveys due to differing techniques.
Typically, metrics like elasticity (proportional) and sensitivity (additive) describe the influences of demographic parameters on population growth13,14,22,26. For Caprinae, when adult female survivorship is 0.90 and recruitment 0.30, the elasticity in survivorship and recruitment are 0.61 (90% CIs 0.40–0.75) and 0.24 (90% CIs 0.13–0.40) respectively (elasticity in young adult survivorship is 0.16 (90% CIs 0.12–0.21). For ungulates in general, the elasticity values for survival tend to be higher than those for recruitment27. Our results match this pattern, as the elasticity results indicate that a change in adult survival has a 2.5 times greater effect on λ than an equivalent change in recruitment. Relatedly, other theoretical work reports that demographic parameters with more temporal variability have lower elasticities, indicating less impact on population fitness (e.g.28,29).
Our work centers on applications. Since most management actions affect these demographic parameters simultaneously, at issue is the practicality (e.g. feasibility and affordability) of management to increase these parameters, and understanding how such changes could impact λ. For example, imagine a population with mean recruitment of 0.30 and adult survival 0.85, with a biologist interested in increasing recruitment or adult female survival to acquire λ ≥ 1. The answer is to increase either value by 0.02 (Fig. 1, Supplementary Data S1). Similarly, one can set a λ target and determine the amount of recruitment and adult female survival necessary for acquiring it (Fig. 1, Supplementary Data S1).
Minimum abundance target
A minimum population of 50 adult females meets the persistence criteria, given intermediate levels of recruitment and survival producing λ ~ 1 (Table 2). The risk of population collapse wanes as populations increase above the minimum threshold (Table 2; Fig. 1). For example, a population of ~ 100 adult females always meets persistence criteria (Table 2). Populations of adult females should be somewhat larger than 50 when modest declines (λ ~ 0.97) are suspected, providing a cushion to address the causes of decline, and mitigate further reductions.
Translocation of 5 adult females during each of 5 years, or 10 in each of 3 years, requires a starting abundance of 70 adult females for the population to maintain the persistence criteria, never reach a lower confidence interval of 0, and for the population to return to the starting population size within 30 years. If managers mistakenly target a population having < 50 adult females, the population mean is unlikely to recover to pre-removal levels within 30 years (Fig. 4). The more adult females removed per year will reduce a populations abundance and elevate stochastic effects.
Applications
These survival and recruitment parameters have high temporal and geographical variability. This uncertainty originates from the variability inherent to survivorship and recruitment data within Caprinae populations and demographic stochasticity operating over all age classes, making it difficult to predict exact values of λ and project abundances (Fig. 3). The causes of such variability are complex (e.g. factors such as variation in environmental drivers, predation, management types and timing, the distribution of female reproduction across populations and ages through time), and are rarely monitored. Without data, these factors cannot be explicitly examined. The effects of these factors, however, are contained in the empirically-derived estimates for survival and recruitment, and their associated variability.
We examined the effects of time and management actions on reducing variation in recruitment and adult female survivorship. We began by halving the variance in recruitment, given its empirical basis. For all recruitment data, when the temporal window is short (3 years), recruitment variability declined to nearly half the overall mean variability. Also, the NMDGF operate a fenced facility (6.2 km2) for desert bighorn sheep, to translocate surplus animals elsewhere. These sheep remain wild, yet receive year-round water and rare predation events. This population has biological variance in recruitment approximately half that of the variance of other populations of desert bighorn in the southwest U.S. (variance = 0.015). The mean recruitment value (L:EY) recorded within this facility (0.56; N = 18 years; CV = 0.22) is nearly double the amount of recruitment calculated from the other populations of desert bighorn sheep.
For survival, most of these data represented desert bighorn sheep populations whose abundance were declining (disease, predation, habitat loss22). Survivorship of adult females in endangered or declining populations of ungulates could be lower or more variable than in demographically viable populations30,31,32.
Management actions can reduce variation in recruitment and survival. However, the simulations halving variance in recruitment or survival received minimal increases in precision in λ or abundance (Table 4; Fig. 3). Halving variability in adult female survivorship and recruitment, or eliminating it, moved the probability of λ < 1 to 0.39 and 0.31 (from 0.41) while reducing QE(30) from 0.05 to 0.01 and 0 respectively (Table 4). Therefore, in application, halving variance in recruitment (i.e., calculated from predator-free populations) and survival (i.e., from predator control) does not generate much change in population trajectories.
Management actions can also change the mean values of demographic parameters. For instance, populations of grazing mammals lacking predation have higher growth rates with less variability33. Therefore, when populations experience predation, predator control can improve population sizes of Caprinae by raising the survival of adult females and recruitment12,34,35. Survival of adult females can increase 5–12% with predator control, and survival with predation is often twice as variable as populations without predation12. Imagine predator control boosted mean adult female survival from 0.90 to 0.9512 and recruitment from 0.30 to 0.50 (similar to the NM penned facility), while reducing the survival and recruitment variance by half (12; NM penned facility). Population growth would increase 20%, from λ = 1.02 to 1.22. Mean abundance would climb from 56 adult females to 135 by year 5.
Predators are important components of ecosystem health36,37, so predator management requires conscientious approaches. Predator control (or perhaps reduction in conspecific, alternative prey) could be applied to improve Caprinae populations with < 50 adult females. Larger populations should withstand background levels of predation, when stable or growing. Curbing declines in larger populations, or attaining targets in population growth or abundances could also warrant predator control, to accomplish management objectives. Monitoring the abundance and recruitment of Caprinae populations helps identify conditions suggesting management actions like predator control, while indicating when the desired project objectives are achieved (i.e. Caprinae population responses). Factors like environment, climate and disease influence population abundance too, and are worth considering during such monitoring38.
Provision of supplemental water can increase the distributions of Caprinae39,40. Supplemental water, however, could increase predator distributions (directly or indirectly), potentially aggravating predation issues affecting adult female survival and recruitment41,42. Other hypotheses, like temporal reductions in water supplementation, may also reduce predator presence within Caprinae ranges43,44. Alternative management tools for improving Caprinae abundances are supplemental feeding and disease control, challenging and costly to pursue at large scales.
The variability in adult female survival and recruitment means that 1 year’s recruitment or survival cannot accurately predict the next, and historical trends in population abundance may have little bearing on future performance (Fig. 3). Revealing management contributions to population growth is also challenged by the variability inherent within the demographics of Caprinae populations. This situation makes Caprinae management reactive, with management decisions based on the short-term monitoring results for a population. The repeated acquisition of accurate abundance data, however, builds a more proactive paradigm.
Monitoring Caprinae
Population vital rates and their importance to population growth often varies across populations and within a population over time (our results31,32,45,46). Indeed, within and across populations of Caprinae, the annual differences in survival and recruitment are substantial47,48. Some temporal and geographical variation is biological, and some stems from the use of different methodologies for quantifying variability. Hence, analytical projections of future abundance based on survival and recruitment data from one population are unlikely to apply to other populations for the same or different species. Understanding a given populations status requires directly monitoring that population.
Frequent monitoring of many populations requires simple and affordable methods for identifying a population’s status (abundance, growth). We discovered the threshold of 50 adult females with three types of demographic parameters: abundance, survival and recruitment. Monitoring the threshold is achievable with two of them.
Adult female survival is the most challenging and costly parameter to acquire, unattainable for most biologists managing Caprinae populations. Indeed, we struggled to locate data describing adult female survival given a species extensively studied. Future monitoring and analyses, therefore, do not require survival data, but can rely solely on measures of abundance and recruitment.
Minimum counts are the primary method for monitoring Caprinae species. This method is appropriate if precision measurements are not required and acknowledgement that data are biased low. With minimum counts, management actions dependent upon abundance triggers (i.e., 50 adult females) would be identified early, providing a conservative approach to population management. In the short term, populations can be prioritized for conservation using minimum abundances, but these data make it difficult to generate longer-term conclusions about population sustainability.
When minimum counts or mean abundance estimates for adult females are well above 50 (i.e. ~ ≥ 75), the population is likely to persist (unless conditions change). When populations occur at or near the threshold value, recruitment estimates associated with these abundance estimates (and trends) help biologists infer if the population is stable or growing, and if the minimum abundance threshold were reached. For example, imagine a population with known abundance of 70 adult females. A survey was conducted, producing an estimate of 65 adult females (90% CI 44–86 (20% CV)), with a recruitment estimate (Lamb:(adult female + yearling)) > 0.30. Trends in abundance from prior surveys indicate an increase. A reasonable interpretation of these data is that the population is likely at or above the threshold and growing.
We used abundance trends because of the high variability in recruitment trends (demonstrated herein). If either abundance, abundance trend or recruitment shows a concerning decline, it should initiate work to discern the cause, as the population may no longer be sustainable.
While the acquisition of accurate abundance data can be challenging48,49,50, researchers have posed tenable solutions for meeting these challenges50,51. These methods rely on methodological changes in aerial surveys, or replacing aerial surveys with motion activated cameras, that collaterally identify predators and their relative abundances.
For annually-reproducing Caprinae that begin reproduction in their second to third year, the minimum abundance threshold of 50 adult females should apply. Source populations for translocation have a 70 adult female abundance threshold, assuming translocations of 5 or 10 adult females per year, in each of 5 and 3 years respectively. For Caprinae with different life histories, our model code is easily modified to accommodate those changes.
Hence, our work extends beyond Caprinae. The population projection matrix model we developed can be applied to any species reproducing on an annual cycle. Users specify parameters such as the number of stage-classes, mean and variance of recruitment and adult survival, starting population size, correlation between recruitment and survival, sex-ratio at recruitment, and number of translocations (Supplementary Data S1). As exemplified herein, results will identify a population’s’minimum threshold for persistence, which subsequently informs threat assessments, harvest quotas and the triage of conservation activities aimed at recovering ailing populations.
Source: Ecology - nature.com
