Using population viability analysis to evaluate management activities for an endangered Hawaiian endemic, the Puaiohi (Myadestes palmeri)

Abstract

Evolution in the Hawaiian Islands has produced a unique avian assemblage. Unfortunately, many of these bird species are highly endangered or extinct. Despite numerous and increasing threats and great effort aimed at saving endemic birds, we lack basic science necessary for understanding many species of concern. One such species is the critically endangered Puaiohi (Myadestes palmeri), a rare songbird endemic to the island of Kaua‘i and the only remaining native thrush on the island. At present, the Puaiohi’s breeding population is estimated to be ~500 birds restricted to the Alaka‘i Wilderness Preserve. We collected demographic data from 2007–2012 and supplemented it with published sources. Using Vortex, we developed stochastic population models to represent Puaiohi population dynamics under current and potential management scenarios to determine management’s potential efficacy in aiding species recovery. Management scenarios modeled included rat control, habitat improvement, general survival facilitation, and provision of nest boxes. The model indicated a decline in abundance with a growth rate (r) of -0.267 under baseline conditions. Female and juvenile survival appeared to be the most influential parameters related to population growth and persistence, so management should focus on increasing female and juvenile Puaiohi survival. Rat control, even at more conservative levels, appeared to be the most effective method of increasing Puaiohi abundance. Our results indicate that practical, attainable management activities can increase Puaiohi and bring the species back from the brink of extinction. Such findings provide an example for other endangered species conservation efforts.

An analysis of translocation regimes for the endangered puaiohi Myadestes palmeri

“…Given that the more successful a translocation regime was at establishing a new population, the more detrimental it was to the original population if juveniles were harvested, it may not be possible to build a new population of puaiohi from juveniles within the existing wild flock. Although the wild flock has been presumed stable in recent years (USFWS 2006, Crampton et al 2017, recent models based on best available data indicate that it may in fact be declining (Fantle-Lepczyk et al 2018). Given the profound difference in effects of harvest based on source population, understanding more accurately the actual dynamics of the wild population is of great importance prior to any translocation project that involves regular removal of juveniles.…”

“…To evaluate translocation scenarios, we constructed population viability analysis (PVA) models using Vortex 10 (Lacy & Pollak 2014), which simulates stochastic demographic and environmental processes. Within Vortex we built upon previously developed models (Fantle-Lepczyk et al 2018) using all available information on puaiohi. In addition to running a baseline control model in which no translocation occurred, we considered how translocation regimes affected both the original and the new populations.…”

“…We developed our demographic input parameters from a variety of sources, including previously published information, original data, and expert discussion. Baseline models were previously published in Fantle-Lepczyk et al (2018).…”

“…Under each condition we assumed the release site would be an improvement over current conditions, which at best produce a small and stable, though potentially declining, population. The first of the 3 release conditions reflects a stable population, as extrapolated for the wild stable population (Fantle-Lepczyk et al 2018). Second, we modeled release in a predator-controlled area in which survival and nest success increase due to release from predator impacts.…”

“…With our models exhibiting stable mortality at release sites, we explored the effect of releasing puaiohi into an area managed to increase puaiohi survival at all life stages, while holding steady all other demographics. To model eliminating the effects of malaria, predation, or any other mortality events on the puaiohi population, in the absence of any other changes, we decreased previously estimated mortality at all life stages (VanderWerf et al 2014) by 18%, to a level at which they would provide a stable population projection (Fantle-Lepczyk et al 2018).…”