Estimating eastern wild turkey productivity using trail camera images

Introduction

The eastern wild turkey (Meleagris gallopavo silvestris; hereafter, wild turkey) is an economically and culturally important upland game bird (Keck and Langston 1992) with a widespread distribution across eastern United States of America (McRoberts et al. 2020). Once extirpated from many regions, the species was successfully recovered following extensive restoration efforts within and beyond their historical range (Chamberlain et al. 2022), including Wisconsin, USA. Wild turkey populations continued to prosper post-restoration (Tapley et al. 2005), and many state management agencies increased harvest opportunities in response to increased demand (Isabelle et al. 2018). However, by the mid-2000s, reports suggested potential regional declines in wild turkey abundance (Tapley et al. 2011; Casalena et al. 2016) and corresponding reductions in spring harvest (Ericksen et al. 2016; Parent et al. 2016). Furthermore, many wildlife managers reported declining trends in annual indices used to monitor wild turkey productivity (Byrne et al. 2016).

Annual production is an important parameter that affects abundance and can have a profound influence on population growth and sustainability in wild turkeys (Vangilder and Kurzejeski 1995; Roberts and Porter 1996; Pollentier et al. 2014a). Therefore, robust and reliable estimates of productivity (i.e. nest success, hatch rates), recruitment (i.e. rate at which individuals are added to a population) and population trends are helpful for establishing effective management objectives for wild turkey populations (Pollentier et al. 2014a; Tyl et al. 2020). For example, in Wisconsin, harvest authorization levels are reviewed by an advisory committee prior to spring and autumn harvest seasons, and adjustments to the allocation of harvest authorizations may be recommended based upon trends in harvest success rates in conjunction with indices to productivity, hunter densities, and information on population trends from local field staff, conservation partners and hunters (Wisconsin Department of Natural Resources 2015).

Radio telemetry and brood flush counts have been used to estimate wild turkey productivity metrics (Porter et al. 1983; Roberts et al. 1995; Vangilder and Kurzejeski 1995; Paisley et al. 1998; Pollentier et al. 2014b). However, such techniques are often reserved for short-term research projects and inferences are typically limited to the spatial scope of the study area. To monitor population trends and productivity at broader spatial and temporal scales, managers often rely on less intensive annual surveys to infer population dynamics in relation to management needs. For wild turkeys, brood surveys have been widely used by state wildlife managers to obtain annual indices of productivity (Johnson and Danks 2022; Oakley and Glick 2022; Backs and McCallen 2023).

Wild turkey productivity had been regularly monitored in Wisconsin via two statewide surveys: the wild turkey landowner brood survey (Dhuey 2012) and the 10-week brood observation survey (Rees Lohr 2018). The landowner brood survey was a mailed postcard survey of rural landowners of 40 or more acres; recipients were requested to record the number of wild turkey broods, poults per brood and adult hens (i.e. adult females) with or without a brood seen from June to August (Dhuey 2012). The 10-week brood observation survey was a Wisconsin Department of Natural Resources (WDNR) staff-only opportunistic statewide survey of upland game bird broods, including gray partridge (Perdix perdix), ring-necked pheasant (Phasianus colchicus), ruffed grouse (Bonasa umbellus) and wild turkey, observed during normal field operations during a 10-week period from June to August (Rees Lohr 2018). Even though both surveys were helpful for monitoring annual wild turkey productivity, multiple factors contributed to declines in survey participation including fewer staff, shifting work duties and waning interest from public participants (Dhuey et al. 2020). Reduction in survey participation eroded confidence that data reflected wild turkey productivity, and thus the landowner brood survey and 10-week brood observation survey were suspended in 2012 and 2019 respectively.

Since the discontinuation of the 10-week brood observation survey, wildlife managers in Wisconsin have sought alternative methods to monitor annual wild turkey productivity and recruitment such as the use of trail cameras. Trail cameras have emerged as a promising management tool to monitor wildlife continuously and noninvasively (Steenweg et al. 2017; Townsend et al. 2021). Advances in trail camera technology and increased interest in using wildlife images for community (or citizen) science and for outreach and education have contributed to rapid growth in the use of trail cameras (Burton et al. 2015). In contrast to traditional field methods, trail cameras provide researchers the ability to remotely study animals at increased temporal and spatial scales, and often with less effort. Trail cameras have been increasingly used in ecological studies to determine occupancy, habitat and activity patterns, species richness, distribution, population dynamics and abundance (O’Brien and Kinnaird 2008; O’Connell et al. 2011). Trail cameras have also been used to identify wild turkey roost sites and to investigate nest predators and nest parasitism (Pharris and Goetz 1980; Dreibelbis et al. 2008; Melville et al. 2014; Ruttinger et al. 2014; Brautigam et al. 2016), although none to our knowledge have attempted to estimate wild turkey productivity across large spatial scales.

Wild turkey recruitment is a dynamic process including nesting, nest depredation, renesting, formation of crèches (multiple hens with commingled poults) and poult mortality that we aimed to track throughout summer using trail camera photos. Trail camera photos provide a novel opportunity to monitor wild turkey recruitment. However, similar to in-person surveys, the complication that detection of poults changes as poults grow and become more mobile, is present. In late spring and early summer, many hens have made a first nesting attempt, and some have successfully hatched out the first broods that are seen on trail cameras. These poults are still very small and difficult to detect in trail camera photos. In early- to mid-summer time, hens that lost their first nest may renest and successfully hatch a brood, though smaller brood sizes may occur from renesting attempts given smaller average clutch sizes of renests (Vangilder and Kurzejeski 1995; Pollentier et al. 2014b). During this time, trail cameras are detecting more poults and poults of different ages, occasionally in the same photos, indicating the formation of crèches (Tyl et al. 2020). Throughout summer, more hens become successful, and broods often merge to form additional crèches. At the same time, individual poults experience mortality, thus reducing brood size, but become more detectable on trail cameras as their body size and mobility increase.

The WDNR launched Snapshot Wisconsin (SSWI), a community science trail camera wildlife monitoring project, in 2015 and by 2018 the program expanded statewide with >1500 volunteers placing WDNR-purchased SSWI trail cameras on their private land (Wisconsin Department of Natural Resources 2025). Currently, SSWI has >2000 trail cameras placed by volunteers on private and public land within SSWI survey blocks that evenly cover the state. SSWI trail cameras have been used to estimate male abundance at lek sites of greater prairie-chicken (Tympanuchus cupido) and sharp-tailed grouse (T. phasianellus; Stenglein et al. 2023). Annually, SSWI provides spatial and temporal trends for 11 Wisconsin species including black bear (Ursus americanus), bobcat (Lynx rufus) and fisher (Pekania pennati). SSWI provides an opportunity to develop novel techniques for acquiring robust estimates of wild turkey productivity metrics statewide. Our research objectives were to determine if SSWI trail camera data could be used to: (1) estimate wild turkey productivity metrics, (2) model the effects of temporal and spatial covariates on these metrics, and (3) provide regional wild turkey recruitment information that could be used by managers to inform wild turkey management decisions.

Materials and methods

Study area

The study area was the state of Wisconsin (~169,600 km²) located in the Upper Midwest, USA (Fig. 1). Glaciation has strongly influenced the state’s geology and natural communities. Much of northern Wisconsin was characterized by extensive forest cover composed of northern mesic hardwoods and mixed hardwood–coniferous forests with many freshwater glacial lakes connected by meandering streams, though portions of northwest Wisconsin consisted of dry-mesic pine (Pinus spp.) and oak (Quercus spp.) forests, barrens and grasslands, along with limited areas suitable for row-crop agriculture. Central Wisconsin landscapes were highly variable with a mosaic of pine–oak–aspen (Populus spp.) forest cover, grassland, open wetlands and agriculture. Much of central Wisconsin was influenced by the tension zone, an ecoclimatic transition region that crosses the state from northwest to southeast and represents a shift from southern to northern species assemblages (Attig et al. 2011). Southwest Wisconsin was characterized by highly eroded, unglaciated topography with a mix of oak–hickory (Carya spp.) and maple (Acer spp.)–basswood (Tilia spp.) forests, grasslands and herbaceous fields, as well as row-crop agriculture and dairy farming. Southeast Wisconsin was densely populated with many of the state’s residents residing in the region and the primary land use was agricultural cropland. Statewide, forests constituted ~40% of land cover and cropland–agriculture comprised ~25% of land cover (Fig. 1).

Fig. 1.

Primary land cover types of Wisconsin, USA, from Wiscland 2.0 and Wisconsin’s 7 turkey management zones (Wisconsin Department of Natural Resources 2016).

Most of Wisconsin had a continental climate; the average minimum temperature was −15.7°C in January and the average maximum temperature was 26.9°C in July. Annual precipitation varied across regions with a statewide mean of 794 mm. Winter snowfall totals vary along a latitudinal gradient and ranged from 78.7 to 421.6 cm in southern and northern Wisconsin respectively (Wisconsin State Climatology Office 2023). Turkey Management Zones (TMZs) have been implemented in Wisconsin to regulate harvest opportunities in accordance with wild turkey population status and habitat suitability in certain areas (Fig. 1; Wisconsin Department of Natural Resources 2015).

Productivity metrics

We calculated commonly reported productivity metrics for wild turkeys using trail camera data: the percentage of successful hens and two versions of poult-to-hen ratio (i.e. poult-to-total-hen ratio and poult-to-successful-hen ratio) for each camera site by calendar week and year (Fig. 2). Generating productivity metrics from trail camera data is novel and therefore it was valuable to consider multiple manipulations of trail camera data that approximated productivity metrics from observation surveys (National Wild Turkey Federation Technical Committee 2018). Additionally, different metrics may capture separate components of the complex wild turkey behavior and reproduction process. For example, a poult-to-hen ratio that considers all hens (successful and unsuccessful) may best track the process of initially unsuccessful hens renesting and becoming successful and therefore adding more poults to the landscape. A poult-to-hen ratio that considers only successful hens may better track poult survival and changes within crèches (i.e. smaller broods in second nesting attempts) throughout the summer brood-rearing season.

Fig. 2.

Diagram of methods used to calculate three productivity metrics for eastern wild turkeys in Wisconsin, USA, from Snapshot Wisconsin trail camera data. Step 1 subset the turkey photo classifications into groups by each camera site, week, and year. Step 2 categorized photo classifications into three distinct categories based on whether there are only hens photographed, only poults photographed, or both hens and poults photographed. The hens photographed with poults are considered successful hens. Step 3 calculated three productivity metrics where the numerator and denominator consider different subsets of the photo categories.

We calculated percentage of successful hens as the sum of hens in triggers that included ≥1 poult divided by the sum of hens in all triggers, poult-to-total-hen ratio as the sum of all poults divided by the sum of all hens, and poult-to-successful-hen ratio as the sum of all poults divided by the sum of hens in triggers that included ≥1 poult (Fig. 2, Table S1 of the Supplementary material). We calculated estimates for weeks within the months of June–August because few triggers of poults were captured prior to June, and after August it becomes difficult to distinguish poults from hens as poults have grown to nearly adult size.

Results

Covariate modeling

The top tier 1 model for percentage of successful hens included quadratic effect of week and random effects of year and site (w_i = 0.81; Table 1). The top tier 2 model for percentage of successful hens included the top model from tier 1, TMZ, and the quadratic effect of proportion of conifer cover (w_i = 0.56; Table 1). The model with a quadratic effect of grassland had some support as well (w_i = 0.22; Table 1). Model results indicated that percentage of successful hens increased by 2.71% (Table 2) each week on average (Fig. 3a). The linear and quadratic effects of conifer cover were weak (Table 2), though the trend was positive (Fig. 3b, S1). The average percentage of successful hens by TMZ for the last week of August was 36.0% (range: 32.4–40.5%; Fig. 3c).

Table 1.Top mixed-effects models to explain percentage of successful hens (PSH), poult-to-total-hen ratios (PTTH) and poult-to-successful-hen ratios (PTSH) for eastern wild turkeys as derived from trail camera photos collected from June to August of 2016–2020 in Wisconsin, USA.

Metric	Tier	Model structure	Model statistics
			K	AIC c	ΔAIC c	w i	Cum. w i
PSH	1	week + week2 + year + site	6	66,855.24	0.00	0.81	0.81
	2	TMZ + conifer + conifer2	14	66,804.49	0.00	0.56	0.56
	2	TMZ + grassland + grassland2	14	66,806.33	1.84	0.22	0.78
PTTH	1	week + year + site	5	22,030.98	0.00	0.92	0.92
	2	TMZ + grassland + grassland2	13	22,028.90	0.00	0.28	0.28
	2	TMZ + conifer	12	22,029.61	0.72	0.20	0.48
	2	TMZ + conifer + conifer2	13	22,030.82	1.92	0.11	0.58
PTSH	1	month + year + site	5	7404.82	0.00	0.69	0.69
	2	TMZ + oak + oak2	13	7398.55	0.00	0.36	0.36

Tier 1 models include the temporal covariates and tier 2 includes the spatial covariates. We included the top tier 1 model in each of the tier 2 models. For Model structure, the following represent the random effects: year, 2016–2020; site, trail camera location; and the following represent the fixed effects: week, calendar week 22–35; month, June–August; grassland, conifer, oak, proportion of land cover within 308-ha buffers around each camera location; Turkey Management Zone (TMZ). Furthermore, we only included models within 2 ΔAIC_c in the model set. The full model set is available in Table S2 of the Supplementary material. For Model statistics, model fit is described by the number of parameters (K), Akaike information criterion corrected for small sample sizes (AIC_c), change in AICc as compared with the top model (ΔAIC_c), AICc model weight (w_i) and cumulative weight (Cum. w_i).

Table 2.Covariate estimates (beta values) for fixed effects (mean and standard error, s.e.) and 95% confidence interval upper and lower limits from top models of eastern wild turkey percentage of successful hens (PSH), poult-to-total-hen ratios (PTTH) and poult-to-successful-hen (PTSH) ratios as determined by trail camera photos from June to August 2016–2020 in Wisconsin, USA.

Metric	Covariate	Estimate	s.e.	Lower limit	Upper limit
PSH	week	2.71	0.10	2.52	2.90
	week2	−0.09	0.03	−0.14	−0.03
	conifer	3.02	9.62	−15.84	21.89
	conifer2	11.54	13.03	−14.00	37.08
PTTH	week	0.07	0.00	0.06	0.08
	grassland	−1.02	0.44	−1.88	−0.16
	grassland2	1.41	0.88	−0.33	3.14
PTSH	month	−0.34	0.06	−0.45	−0.22
	oak	−1.13	1.00	−3.08	0.83
	oak2	3.18	1.85	−0.45	6.81

For covariates, week, calendar week 22–35; month, June–August; grassland, conifer, oak, proportion of land cover within 308-ha buffers around each camera location. All random and fixed effects from top models are available in Table S3 of the Supplementary material.

Fig. 3.

Patterns in eastern wild turkey productivity metrics as percentage of successful hens (row 1: panels a–c), poult-to-total-hen ratios (row 2: panels d–f) and poult-to-successful-hen ratios (row 3: panels g–i) in Wisconsin, USA, using Snapshot Wisconsin trail camera data from June to August 2016–2020. Each metric is shown varying through time in column 1 and with proportion of landcover in column 2 with other variables held at their average values (shaded light blue areas are 95% confidence intervals). Column 3 shows predicted metrics at an average site and year for the last time period (week for panels c and f, or month for panel i) and using each turkey management zone’s average proportion of coniferous forest (panel c), grassland (panel f) or oak (panel i). The time and landcover predictors were selected from each metric’s top tier 2 model (see text for model details).

The top tier 1 model for poult-to-total-hen ratio included week and random effects of year and site (w_i = 0.92; Table 1). The top tier 2 model for poult-to-total-hen ratio included the top model from tier 1, TMZ, and the quadratic effect of proportion of grassland cover (w_i = 0.28; Table 1). Models with linear (w_i = 0.20; Table 1) and quadratic effect (w_i = 0.11; Table 1) of conifer cover were competing models. In the top tier 2 model, poult-to-total-hen ratio increased by 0.07 poults-per-hen each week on average (Table 2; Fig. 3d). Linear and quadratic effects of grassland (Table 2) were weak and poult-to-total-hen ratio was highest with less variability at higher grassland proportion (Fig. 3e, Fig. S1). Average poult-to-total-hen ratio by TMZ for the last week of August was 1.08 (range: 0.89–1.31; Fig. 3f).

The top tier 1 model for poult-to-successful-hen ratio included month and random effects of year and site (w_i = 0.69; Table 1). The top tier 2 model for poult-to-successful-hen ratio included the top model from tier 1, TMZ, and the quadratic effect of proportion of oak cover (w_i = 0.36; Table 1). From the top model, poult-to-successful-hen ratio decreased by 0.34 poults each month on average (Table 2; Fig. 3g). The linear and quadratic effects of oak (Table 2) were weak, though the trend was for higher poult-to-successful-hen ratios at sites with a higher proportion of oak cover (Fig. 3h, S1). The average predicted poult-to-successful-hen ratio by TMZ for August was 2.59 (range: 2.19–3.01; Fig. 3i).

Discussion

Our research demonstrates that SSWI provides a novel and viable technique for monitoring wild turkey productivity and recruitment. Through the review of 5 years of SSWI wild turkey trail camera triggers, we estimated the percentage of successful hens, poult-to-total-hen ratios, and poult-to-successful-hen ratios and explored temporal and spatial factors affecting these metrics. We also estimated wild turkey recruitment metrics for each TMZ as wild turkeys enter autumn, a time of year where practitioners are interested in population trends for informing harvest recommendations for the upcoming spring season. A timeline, where decision-makers meet in autumn to make decisions for spring turkey management (e.g. number of available harvest authorizations per TMZ), aligns well with the trail camera methods we outlined here. Further, the focus on August-only data decreases the photo classification task significantly. It is possible that wild turkey productivity estimates from this research are biased low, primarily due to poults being more difficult to detect and moving less than hens. If these metrics are biased low, they can still be useful indices of wild turkey productivity for wildlife managers, especially for considering trends in these metrics across TMZs and years.

Our productivity metrics were lower than those from brood observation survey reports from other Midwestern states, including Indiana, Iowa and Nebraska (Backs 2020; Coffey 2021; Meduna 2021). We are cautious to assume that Wisconsin’s wild turkey population has lower productivity because brood observation surveys and trail cameras use different data collection methods, even though both rely on opportunistic interactions with broods. Both methods can be affected by several sources of variation, including repeat counts and observations of crèches, and observation surveys can further be influenced by surveyor bias (i.e. skill or experience of the observer; Healy and Powell 1999; Wakeling et al. 2022). We believe the use of trail cameras reduces this observational bias (i.e. inaccurate counts which can result in over- or under-estimation) because trail cameras record observations when motion is detected irrespective of whether there is a single or multiple hens or a hen with or without a brood. Moreover, if an observer has previously encountered birds at a particular location (or even has prior knowledge of birds at a site from another source), then they may be more likely to record birds at subsequent surveys in a repeated counts framework. This phenomenon has been demonstrated in several species at a variety of locations and can lead to biased estimates of detection probability, occupancy and abundance (Riddle et al. 2010). This bias is minimized in the Snapshot Wisconsin approach because the triggers are randomized, and observers are not provided information about the location from which the triggers were taken. However, trail cameras may not be detecting hens and poults with equal probability as poults are smaller and more difficult to observe in the trail camera photos, especially young poults, potentially leading to lower productivity metrics from trail cameras. Ideally, future projects should evaluate SSWI trail camera detection rates for hens and poults throughout the summer brood-rearing season and pair trail camera photos of marked hens and broods with in-person flush counts and GPS tracking.

We observed strong positive relationships between a week and percentage of successful hens as well as poult-to-total-hen ratio. The denominator of total hens is the same for both metrics and there is high hen survival in summer (Pollentier et al. 2014b). As such, there is little reason to expect much change in total hens throughout summer. Therefore, the biological processes that increase the numerators of these metrics are what is most likely driving the positive trend through time. First, hens who have lost their first nest can have a successful renest attempt, adding poults to the landscape (i.e. increasing poult-to-total-hen ratios) and transitioning hens from unsuccessful to successful throughout summer (i.e. increasing percentage of successful hens). An additional reason for an increase in poult-to-total-hen ratio is explained by increased movement and detection of poults as they grow, become more mobile and expand their home range (Porter 1980; Anderson et al. 2011; Chamberlain et al. 2020). Initially, it may seem counterintuitive that we observed a strong negative relationship between month and poults-per-successful-hen ratios. However, the decrease in this metric through time becomes clearer when we understand what is happening both mathematically and biologically. First, the denominator is successful hens, and successful hens are increasing throughout summer because of successful renesting attempts. Also, the numerator (i.e. the total number of poults) is increasing through time because successful renest attempts are adding poults and all poults grow and become more detectable. So, although both the numerator and denominator are increasing, a decrease in this metric means that the proportional increase in poults is slower than the proportional increase in successful hens. We expect this could be due to a couple of factors. First, it is established that successful renest attempts typically result in fewer poults within a brood compared to initial nesting attempts. Therefore, later nests may be adding proportionally fewer poults compared to earlier nests (Vangilder and Kurzejeski 1995, Pollentier et al. 2014b). Also, poults experience high mortality until they reach near-adult size in late summer or early autumn (Pollentier et al. 2014b). It is not possible with the methods that we used to disentangle whether smaller broods later in summer or poult mortality was more important for the decline in poult-to-successful-hen ratio over summer. Seemingly, wild turkey productivity metrics from trail camera photos were able to capture the detection and survival process of poults throughout the summer brood-rearing period, thereby highlighting the utility of trail cameras to understand and monitor wild turkey populations. Our formulas for calculating productivity metrics from trail camera data appear to reflect productivity processes that we would expect for wild turkeys in summer.

In addition to exploring temporal covariate relationships, the SSWI data set was robust enough to support land cover modeling efforts. Our findings suggest that land cover has some effect on productivity metrics because models were stronger with the addition of land cover covariates. Both conifer and grassland cover appear to be important factors for predicting hen success and poult-to-total-hen ratios. Grassland and herbaceous vegetation provide nesting cover for hens to avoid ground predators and consistent foraging opportunities for growing poults (Pollentier et al. 2014b; Tyl et al. 2020), and conifers typically offer abundant overnight roosting opportunities and thermoregulation (Adey et al. 2024). Conversely, oak cover was more important for predicting the size of broods, or the recruitment potential of the population, and oak woodlands are regularly associated with numerous benefits for wildlife including wild turkeys (Steffen et al. 2002). While our research shows a relationship between single cover types and productivity metrics, the relationship is likely more complex. We did not explore interactions or other covariates as part of our research because the goal of our project was to explore the efficacy of trail camera data to generate wild turkey productivity metrics rather than a detailed assessment of the patterns in these metrics across the landscape. Future research could consider interactions between land cover types, spatial pattern and fragmentation of land cover types, and more detailed forest characteristics such as age class, tree density, stand size and canopy cover.

We observed variation in each productivity metric across TMZs with, in general, larger estimates in northern Wisconsin (TMZs 4–7) compared to southern portions of the state (TMZs 1–3), and larger estimates in eastern Wisconsin (TMZs 2, 3, 5 and 7) compared with western Wisconsin (TMZs 1, 4 and 6). Higher productivity indices in northern TMZs support anecdotal reports that the wild turkey population may be increasing in northern portions of Wisconsin where wild turkeys were more recently reestablished (Pollentier et al. 2021, 2023), whereas wild turkey populations in southern Wisconsin may be at a later stage of population growth and closer to carrying capacity. Alternatively, wild turkey populations in southern Wisconsin may have a different (i.e. lower) growth rate compared to northern Wisconsin. Similar observations in productivity metrics across TMZs have been made in Wisconsin upland game bird brood observation surveys (Pollentier et al. 2022, 2023).

Statewide trail camera networks can be used as a tool by wildlife managers to monitor wild turkey productivity and recruitment across time and space. SSWI uses community-based efforts to increase data input and reduce agency staff workloads to ultimately provide wildlife data at spatial and temporal scales that would be challenging to acquire otherwise. SSWI maintains high volunteer retention with >90% of volunteers staying with the program for ≥1 year, which is promising for considering collection of community-based wild turkey productivity metrics into the future. Improving efficiency by reducing the number of photos that need to be classified, and especially by volunteers, could further promote long-term sustainability for trail camera-based metrics. Based on this work, WDNR staff now use SSWI trail camera detections of turkeys in August (instead of June, July and August) to generate annual trail camera-based productivity and recruitment metrics. Further, future work should explore the relationship between productivity metrics determined from classifications of photos made by trained observer(s), as was the case for our research, and those identified from community-based classifications of photos. A strong predictive relationship would suggest that volunteer-classified wild turkey data could be effectively used to monitor wild turkey productivity, thereby saving costs associated with photo classification by paid staff. SSWI’s potential will continue to grow with additional years of data collection so that wildlife trends and productivity metrics can be tracked across years. We encourage other states to consider using statewide trail camera data to monitor wild turkeys and wildlife in general. If other states implemented programs like SSWI, these efforts could evolve into regional monitoring for more robust trend information and productivity metrics for wild turkeys and other wildlife.

Supplementary material

Supplementary material is available online.