Evaluating population viability and efficacy of conservation management using integrated population models

Abstract

Predicting population responses to environmental conditions or management scenarios is a fundamental challenge for conservation. Proper consideration of demographic, environmental and parameter uncertainties is essential for projecting population trends and optimal conservation strategies. We developed a coupled integrated population model‐Bayesian population viability analysis to assess the (1) impact of demographic rates (survival, fecundity, immigration) on past population dynamics; (2) population viability 10 years into the future; and (3) efficacy of possible management strategies for the federally endangered Great Lakes piping plover Charadrius melodus population. Our model synthesizes long‐term population survey, nest monitoring and mark–resight data, while accounting for multiple sources of uncertainty. We incorporated latent abundance of eastern North American merlins Falco columbarius, a primary predator of adult plovers, as a covariate on adult survival via a parallel state‐space model, accounting for the influence of an imperfectly observed process (i.e. predation pressure) on population viability. Mean plover abundance increased from 18 pairs in 1993 to 75 pairs in 2016, but annual population growth (λ¯t) was projected to be 0.95 (95% CI 0.72–1.12), suggesting a potential decline to 67 pairs within 10 years. Without accounting for an expanding merlin population, we would have concluded that the plover population was projected to increase (λ¯t = 1.02; 95% CI 0.94–1.09) to 91 pairs by 2026. We compared four conservation scenarios: (1) no proposed management; (2) increased control of chick predators (e.g. Corvidae, Laridae, mammals); (3) increased merlin control; and (4) simultaneous chick predator and merlin control. Compared to the null scenario, chick predator control reduced quasi‐extinction probability from 11.9% to 8.7%, merlin control more than halved (3.5%) the probability and simultaneous control reduced quasi‐extinction probability to 2.6%. Synthesis and applications. Piping plover recovery actions should consider systematic predator control, rather than current ad hoc protocols, especially given the predicted increase in regional merlin abundance. This approach of combining integrated population models with Bayesian population viability analysis to identify limiting components of the population cycle and evaluate alternative management strategies for conservation decision‐making shows great utility for aiding recovery of threatened populations.

Disentangling data discrepancies with integrated population models

“…, Saunders et al. ), we were able to estimate annual per‐capita fecundity ( F t , p ) by combining the data sets and exploiting our model structure (i.e., Eq. ).…”

Integrated modeling predicts shifts in waterbird population dynamics under climate change

“…Furthermore, IPMs also allow for the estimation of latent demographic parameters without specific data (Besbeas et al 2002, Tavecchia et al 2009, Schaub et al 2013, which is important because the estimation of demographic parameters often requires long-term, individual-based data that are costly to collect and, hence, often limited or unavailable. Due to these advantages, IPMs have been increasingly applied in ecological research (Péron and Koons 2013, Lee et al 2015, Schaub and Fletcher 2015, McDonald et al 2016, Abadi et al 2017, Ahrestani et al 2017, Weegman et al 2017, and are considered an important modeling framework for conservation decision-making (Arnold et al 2018, Zipkin andSaunders 2018). Another type of model that can be used to understand population demography and dynamics are dynamic N-mixture models (Dail and Madsen 2011).…”

The sensitivity of seabird populations to density‐dependence, environmental stochasticity and anthropogenic mortality

“…Understanding how population persistence is affected by variable combinations of density-dependence and environmental regulation is unknown (Saether, Engen, Lande, Both, & Visser, 2002). To increase the credibility of predicted outcomes from population assessments it is essential to represent biological reality whilst capturing uncertainty in estimates of understudied processes (Saunders, Cuthbert, & Zipkin, 2018). Failure to faithfully reflect accuracy and precision in predictions leads to one of two detrimental outcomes for human-wildlife conflict; over-precaution due to an inflated estimate of risk may curtail economically important activities, whilst undetected risks to population viability, could set sensitive populations on a path to extinction.…”