Divergent estimates of herd‐wide caribou calf survival: Ecological factors and methodological biases

Abstract

Population monitoring is a critical part of effective wildlife management, but methods are prone to biases that can hinder our ability to accurately track changes in populations through time. Calf survival plays an important role in ungulate population dynamics and can be monitored using telemetry and herd composition surveys. These methods, however, are susceptible to unrepresentative sampling and violations of the assumption of equal detectability, respectively. Here, we capitalized on 55 herd‐wide estimates of woodland caribou ( Rangifer tarandus caribou ) calf survival in Newfoundland, Canada, using telemetry ( n = 1,175 calves) and 249 herd‐wide estimates of calf:cow ratios (C:C) using herd composition surveys to investigate these potential biases. These data included 17 herd‐wide estimates replicated from both methods concurrently ( n = 448 calves and n = 17 surveys) which we used to understand which processes and sampling biases contributed to disagreement between estimates of herd‐wide calf survival. We used Cox proportional hazards models to determine whether estimates of calf mortality risk were biased by the date a calf was collared. We also used linear mixed‐effects models to determine whether estimates of C:C ratios were biased by survey date and herd size. We found that calves collared later in the calving season had a higher mortality risk and that C:C tended to be higher for surveys conducted later in the autumn. When we used these relationships to modify estimates of herd‐wide calf survival derived from telemetry and herd composition surveys concurrently, we found that formerly disparate estimates of woodland caribou calf survival now overlapped (within a 95% confidence interval) in a majority of cases. Our case study highlights the potential of under‐appreciated biases to impact our understanding of population dynamics and suggests ways that managers can limit the influence of these biases in the two widely applied methods for estimating herd‐wide survival.

Plasticity and repeatability in spring migration and parturition dates with implications for annual reproductive success

“…To ensure that these outliers did not influence our final results, we re-ran our models with these outliers excluded and found qualitatively similar results (see Supplementary Materials S3). We also compared our distribution of birth dates to inferred birth dates from neonate calf capture in Newfoundland (n = 1175; Ellington, 2015), and found no difference in the mean birth dates from our sample and the empirically determined birth dates from Ellington (2015; t = −1.110, p = 0.269).…”

Implementing a novel movement-based approach to inferring parturition and neonate caribou calf survival

“…At the herd-level we compared our IBM and PBM predicted event date distributions to distributions from VHF-collared calves. These distributions have their own limitations, for example parturition dates (as indexed by collaring dates) might not be representative of the entire herd because researchers generally collar animals during a few days of the calving season on any given year due to logistics [ 13 ]. This non-representative sampling could have an obvious effect on generating distribution of parturition dates but could also influence the distribution of calf mortality dates, as parturition date influences calf mortality risk in caribou [ 13 ].…”

“…These distributions have their own limitations, for example parturition dates (as indexed by collaring dates) might not be representative of the entire herd because researchers generally collar animals during a few days of the calving season on any given year due to logistics [ 13 ]. This non-representative sampling could have an obvious effect on generating distribution of parturition dates but could also influence the distribution of calf mortality dates, as parturition date influences calf mortality risk in caribou [ 13 ]. If precision in predicting event date using the IBM and PBM methods can be improved, they would represent non-biased herd-wide distributions of these events, which in turn could improve survival analysis using VHF-collared calves.…”

“…We generated density distributions of the estimated parturition dates and mortality dates derived from the IBM and the PBM. We also converted the estimated calf parturition and mortality events from the IBM and PBM into Kaplan-Meier survival probability curves using the survival package [ 26 ] in R. Following Ellington et al [ 13 ], we generated herd-wide survival curves and parturition and mortality date distributions from the 134 VHF-collared calves from the MR herd and compared them to survival curve and distributions generated using the DeMars IBM and the PBM. The VHF collaring date was used as a proxy for parturition dates.…”

“…Estimates of herd-wide calf survival can be made using recruitment rates from HCS [ 12 ] or from survival analysis using telemetry. These data collection methods can be invasive and prone to biases; particularly, non-representative sampling biases from telemetry and unequal detectability biases from the HCS [ 13 ]. DeMars et al [ 8 ] proposed both an individual-based (IBM) and population-based (PBM) method for determining parturition and calf mortality events using adult movement data.…”