Just Accepted

This article has been peer reviewed and accepted for publication. It is in production and has not been edited, so may differ from the final published form.

Estimating eastern wild turkey productivity using trail camera images

Hannah Butkiewicz, Jennifer L. Stenglein, Jason Riddle, Shelby Truckenbrod, Christopher Pollentier

Abstract

Context Trail cameras provide researchers and managers the ability to study and monitor animals remotely at large temporal and spatial scales and often with less effort than traditional surveys. Trail cameras have been used to identify eastern wild turkey (Meleagris gallopavo silvestris) roost sites and to investigate nest predators and parasitism, although no studies to date have attempted to estimate productivity at larger spatial scales using trail cameras alone. Aims We studied how a statewide network of >2,000 trail cameras operated by community scientists could be used to estimate wild turkey productivity metrics in Wisconsin, USA, and explored the patterns in these metrics throughout the summer brood-rearing season and across Turkey Management Zones (TMZs). Methods Trail cameras from the Snapshot Wisconsin project generated 38,671 turkey triggers (bursts of 3 photos) in June–August 2016–2020. We classified each trigger into counts of hens (breeding-age females) and poults and generated 3 productivity metrics: percentage of successful hens, poult-to-total-hen ratio, and poult-to-successful-hen ratio. For each metric we modeled the effects of temporal and spatial covariates and included site and year random effects. Key results We found positive relationships between time (June – August) and percentage of successful hens and poult-to-total-hen ratio likely due to renesting hens becoming successful and increased detection as poults grow, and a negative relationship with poult-to-successful-hen ratio likely due to renest attempts resulting in fewer poults and mortality of poults. Additionally, we found weak non-linear relationships between each productivity metric and different land cover proportions within 308-hectare buffers of camera sites, including: the percentage of successful hens was largest at sites with high proportions of coniferous forest, poult-to-total-hen ratio was highest and less variable at sites with higher grassland proportion, and poult-to-successful-hen ratio was greatest at sites with high proportion of oak forest. Productivity metrics varied across TMZs with generally higher values in northern and eastern versus southern and western TMZs. For an average year and across TMZs from our modeled results, we estimated 36.0% of hens were successful, a poult-to-total-hen ratio of 1.08, and poult-to-successful-hen ratio value of 2.59. These estimates were for the last time period in the model (i.e., the month of August or the last week in August) and represent our best understanding of Wisconsin wild turkey recruitment for poults that have survived summer and are recruited into the wild turkey population headed into autumn. Conclusions Our research demonstrated that trail cameras can be used to remotely index wild turkey productivity at spatial and temporal scales that would not otherwise be possible with traditional field methods. Our recruitment estimates were low compared to other wild turkey surveys in Wisconsin and surrounding states which could be due to lower detection rates of poults than of hens on trail cameras. Implications We believe that the use of trail cameras warrants further investigation in conjunction with efforts to account for and model detection probability for hens and poults, separately, if trail camera-derived productivity metrics are to be considered for monitoring wild turkey populations

WR24169  Accepted 06 June 2025

© CSIRO 2025