Received: 25 June 2025 | Revised: 2 July 2026 | Accepted: 8 July 2026 DOI: 10.1002/jwmg.70264 R E S E A R C H A R T I C L E Vegetation composition and structure influence female wild turkey survival during incubation Mark A. Turner 1 | M. Colter Chitwood 1 | Craig A. Harper 2 | David A. Buehler 2 | Anna K. Moeller 1 | Nicolle Butler 1 | Cody P. Griffin 1 | Cyrena Bedoian 1 | R. Dwayne Elmore 3 | Craig A. Davis 1 | Michael R. Barrett 4 | Evan P. Tanner 4 | Lindsey Phillips 2 | Vincent M. Johnson 2 | Joseph P. Quehl 2 | Marcus A. Lashley 5 1 Department of Natural Resource Ecology and Management, Oklahoma State University, 320G Agricultural Hall, Stillwater, OK 74078, USA 2 School of Natural Resources, University of Tennessee, 401 Agriculture and Natural Resources Building, Knoxville, TN 37996 ‐ 4563, USA 3 Tall Timbers, 13093 Henry Beadel Drive, Tallahassee, FL 32312, USA 4 Caesar Kleberg Wildlife Research Institute, 700 University Boulevard, Kingsville, TX 78363, USA 5 Wildlife Ecology and Conservation, University of Florida, 322 Newins ‐ Ziegler Hall, Gainesville, FL 32611, USA Correspondence Mark A. Turner, Department of Natural Resource Ecology and Management, Oklahoma State University, 320G Agricultural Hall, Stillwater, OK 74078, USA. Email: mark.a.turner@okstate.edu Funding information Tennessee Wildlife Resources Agency; Oklahoma Department of Wildlife Conservation; Florida state legislature; Bollenbach Endowment; National Wild Turkey Federation; Turkeys for Tomorrow Abstract Wild turkey ( Meleagris gallopavo ) populations have experienced recent declines in some areas, particularly in the southeastern United States. Previous research has focused on vegetation effects on reproduction, especially how vegetative cover af- fects nest and brood survival. However, recent meta ‐ analyses have determined adult female survival is the most elastic vital rate, and contemporary decreases in female survival may help explain recent population declines. Female survival typically is lowest during nesting season, particularly when females are incubating eggs. There is limited information available on whether vegetation composition or structure around the nest influences incubating female survival. To address this knowl- edge gap, we monitored 377 turkey nests from 242 females and measured vegetation structure and composition at nest sites in Tennessee and Oklahoma, USA, from 2017 – 2024. We documented 40 instances of female mortality by predators during incubation. Using an analytical framework to jointly consider nest and female fate, we determined that woody vegetation increased nest survival but also led to increased risk of predation for the incubating female, likely because of Journal of Wildlife Management 2026;e70264. wileyonlinelibrary.com/journal/jwmg | 1 of 15 https://doi.org/10.1002/jwmg.70264 All rights reserved, including rights for text and data mining and training of artificial intelligence technologies or similar technologies. © 2026 The Wildlife Society decreased ability to escape predators. Conversely, increased herbaceous vegetation in Oklahoma and visual obstruction 0.5 – 1 m above ground in Tennessee promoted female survival, indicating both vegetation structure and composition play an important role in reducing predation risk for incubating females. We recommend managers interested in increasing turkey vital rates during incubation implement disturbance regimes that promote visual obstruction at least 0.5 m above- ground but limit extensive woody stem development. K E Y W O R D S female survival, galliform, Meleagris gallopavo , nest success, nest survival, nesting cover, predation risk, wild turkey habitat Habitat management influences vegetation composition and structure that provides food and cover for many terrestrial wildlife species. In forested systems, canopy reduction using commercial harvest or noncommercial forest stand improvement practices is used to advance understory growth and increase food and cover resources (Lashley et al. 2011, Harper 2020, Turner et al. 2020). Prescribed fire is used to promote herbaceous vegetation and maintain an open midstory (Harper et al. 2016, Wann et al. 2020, Bones et al. 2026), which improves vegetation structure and composition for several wildlife species, including wild turkey ( Meleagris gallopavo ). Habitat management techniques can enhance nesting and brood ‐ rearing cover for wild turkeys (McCord et al. 2014), and wild turkeys selectively use areas managed with frequent disturbance (Chance et al. 2020, Wann et al. 2020, Turner et al. 2024). Understanding the influence of vegetation composition and structure on wild turkey vital rates is necessary to better inform management. Wild turkey populations recently have declined in several regions of the United States, and previous research has identified low vital rates as a reason for the decline (Londe et al. 2023). Vegetation surrounding the nest and vegetation available for brood ‐ rearing cover has a strong influence on reproduction (Kilburg et al 2014, Johnson et al. 2022, Nelson et al. 2022, Keever et al. 2023). However, population projection models have identified female survival as the vital rate with the greatest population elasticity in wild turkeys (Londe et al. 2023), meaning a small change in female survival has the greatest proportional effect on population trajectory (Taylor et al. 2012). A recent meta ‐ analysis indicated a decrease in female survival over time may explain why some populations are declining (Lashley et al. 2025), given that other key vital rates, such as nest success, have remained relatively stable over the period in which wild turkey populations have declined (Londe et al. 2023). Therefore, vegetation around the nest may be particularly important to female survival, given female survival generally is lowest during the reproductive season (Vander Haegen et al. 1988, McCall et al. 2020, Yarnall et al. 2020, Tyl et al. 2023). Survival may be especially reduced during incubation when females spend con- siderable amounts of time on the nest throughout the day and night (Collier et al. 2009, Low et al. 2010). Wild turkeys generally select nest locations in areas with overhead cover and visual obstruction 0.5 – 1.5 m above the ground (Kilburg et al 2014, Little et al. 2016, Johnson et al. 2022). Many studies have focused on nest success relative to vegetation around the nest, with mixed results based on the study system and vegetation measurements collected (Badyaev 1995, Johnson et al. 2022, Keever et al. 2023). When vegetation has been reported to influence nest success, success generally increases with greater plant cover and visual obstruction, regardless of plant composition. Previous attempts to identify factors influencing incubating female survival pri- marily have focused on the female's behavior. For example, females that spend more of their daily time budget actively incubating the nest generally have reduced survival (Lohr et al. 2020, Yarnall et al. 2020, Carpenter et al. 2023). Lohr et al. (2020) reported female behavior during incubation influenced nest and female survival but 2 of 15 | TURNER ET AL 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License also noted female survival was unrelated to plant height around the nest. However, both nest and female survival decreased during drought conditions in Texas, USA, likely because of decreased vegetation cover around nests (Hohensee and Wallace 2000). There is limited information as to what extent vegetation at the nest site affects survival of the incubating female, and it is unknown if factors promoting nest and female survival are correlated. We evaluated the influence of vegetation surrounding the nest on nest and incubating female wild turkey survival during incubation in Tennessee and Oklahoma, USA. We hypothesized vegetation structure and compo- sition would influence both nest and female survival. We predicted visual obstruction below 1 m would increase nest and female survival during incubation. We also predicted greater coverage of herbaceous vegetation, un- derstory woody plants, and midstory stems would promote nest survival by providing greater concealment cover. Finally, we predicted midstory stems would decrease female survival because they may impede her ability to escape a predation attempt at the nest site. STUDY AREA We conducted our study in Tennessee and Oklahoma, USA (Figure 1). We collected the Tennessee data at 10 sites located in 5 south ‐ central counties that included a mixture of private and public lands. Hardwood forest was the primary land cover, followed by pasture and naturally occurring early successional vegetation. Softwood forests, including loblolly pine ( Pinus taeda ) plantations and eastern redcedar ( Juniperus virginiana ), also were present, along with interspersed row crops. Annual rainfall averaged 145.8 cm, and average annual temperature ranged from 14.4 – 16.1°C (National Oceanic and Atmospheric Administration [NOAA] 2025). Johnson et al. (2022) provided a detailed description of the Tennessee study sites. We collected the Oklahoma data in McCurtain County, which is the southeastern ‐ most county in Oklahoma. We collected data from a mixture of land ownership types, including United States Forest Service, private, and commercial timberland (Weyerhaeuser). Land cover varied but was composed mostly of pine and hardwood forests with interspersed pasture. Dominant tree species included shortleaf pine ( Pinus echinata ), loblolly pine, blackjack oak ( Quercus marilandica ), and post oak ( Quercus stellata ). Annual rainfall averaged 134.5 cm, and average annual temperature was 17.7°C (NOAA 2025). Butler et al. (2025) provided a detailed description of the Oklahoma study site. Potential predators of turkeys and turkey nests at both sites included bobcat ( Lynx rufus ), gray fox ( Urocyon cinereoargenteus ), coyote ( Canis latrans ), northern raccoon ( Procyon lotor ), Virginia opossum ( Didelphis virginiana ), and great horned owl ( Bubo virginianus ). METHO DS Capture and monitoring We captured female wild turkeys in winter flocks with rocket nets prior to the nesting season (December – March from 2017 – 2022 in Tennessee and 2022 – 2024 in Oklahoma. We fitted each captured female with either a very high frequency (VHF) or Global Positioning System (GPS) backpack ‐ style transmitter (Advanced Telemetry Systems, Isanti, MN, USA; e ‐ obs GmbH, Grünwald, Germany; Lotek Wireless, Newmarket, ON, Canada) and released birds at the capture site. The GPS tags obtained locations 1 – 2 times per hour during diurnal hours, and we monitored VHF tags ≥ 2 times per week. We began monitoring females more frequently at the start of the nesting season in late March to detect nesting attempts. We triangulated females with VHF tags at least every other day, and we downloaded data from GPS tags at least once per week to evaluate movements. We considered a female incubating if we recorded the female in the same location for 2 consecutive days. Once we detected an incubating female, we circled the nest 25 – 50 m away to avoid flushing the female and recorded the approximate nesting location. We recorded the initial incubation date VEGETATION INFLUENCE ON INCUBATING TURKEYS | 3 of 15 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License for all nests based on movement data obtained from transmitters. After confirming a nesting attempt, we checked the status of the female every 1 – 2 days to determine nest fate. Once we determined the nesting event was terminated, either by the female frequently being away from the nest site or receiving a mortality signal not associated with her incubating, we approached the nest site to determine the exact nest location, determine fate, F I G U R E 1 Study sites in Tennessee and Oklahoma, USA, where we evaluated wild turkey ( Meleagris gallopavo ) female survival and nest success, 2017 – 2024. 4 of 15 | TURNER ET AL 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License and record vegetation data. We recorded the fate date and calculated the incubation length of each nest attempt. We assigned a female fate during each incubation attempt regardless of the nest fate. When we suspected a female mortality, we evaluated the site to determine cause of death. We collected vegetation measurements at each nest site after hatching, depredation, abandonment, or female mortality to quantify whether structure influenced female survival or nest success. Data collection varied between Tennessee and Oklahoma, which provided us different metrics to consider between sites to evaluate whether vegetation influenced female survival. We used a 2 ‐ m vegetation profile board (Nudds 1977) to measure visual obstruction (VO), and we collected 4 measurements in each of the cardinal directions around the nest. The profile board used in Tennessee had 3 strata: 0 – 0.5 m, 0.5 – 1 m, and 1 – 2 m. We observed the board from 11.3 m away and scored the board on a scale of 1 – 6 based on percent coverage of the board, whereby 1 = <2.5%, 2 = 2.5 – 25%, 3 = 26 – 50%, 4 = 51 – 75%, 5 = 76 – 95%, 6 = >95%. The board used in Oklahoma had 4 equal, 0.5 ‐ m strata, and was observed from 15 m away. We scored the board on a scale of 1 – 5 based on percent coverage of the board, whereby 1 = 0 – 20%, 2 = 21 – 40%, 3 = 41 – 60%, 4 = 61 – 80%, and 5 = 81 – 100%. To allow us to compare structure between studies, we converted each observation to a percentage and calculated the average percent obstruction within each stratum at each nest. With the Oklahoma data, we averaged the 2 strata above 1 m to match the Tennessee data. This approach provided us with 3 standard percent visual obstruction measures at each nest across studies: 0 – 0.5 m (VO low), 0.5 – 1 m (VO medium), and 1 – 2 m (VO high). We also collected additional vegetation covariates at each nest site. In Tennessee, we counted all midstory stems >1.4 m tall and ≤ 11.4 cm diameter at breast height within a 5 ‐ m radius of the nest. In Oklahoma, we estimated the percent coverage of grasses, forbs, and woody plants within a 1 ‐ m 2 plot (Daubenmire 1959) centered on the nest bowl. We combined grasses and forbs into an herbaceous covariate (percent herbaceous) for analysis. Analysis To jointly consider vegetation effects on the competing risks of death of the nesting female and nest failure resulting from other factors, we created a multistate model based on the state ‐ space competing risks model by Servanty et al. (2010; Figure 2). We used a Bayesian implementation for this approach. We created a known ‐ fate, daily nest survival model with 3 outcome states for the nest each day: alive (A), nest failure resulting from the nesting female's death (female mortality; FM), and nest failure resulting from other reasons (other failure; OF). We defined a nest's state (x i,t ) as a multinomial trial taking values (1, 0, 0), (0, 1, 0), (0, 0, 1), if nest i at time t was in state A, FM, or OF, respectively. The state of nests at time t + 1 was dependent on their state at time t and the matrix of transition probabilities ( Ψ ) between states A, FM, and OF (equation 1). x x x 1 Ψ | ~ multinomial( , · ) i t i t i t i t 1 , + , , , (1) where φ fm of Ψ = 0 0 0 0 0 0 i t i t i t i t , , , , In Equation 1, a nest that is alive at time t will survive to time t + 1 with probability φ i,t , will fail at time t + 1 because of the female's death (state FM) with probability fm i,t , and will fail because of other reasons while the female is still alive (state OF) with probability of i,t . Because nests can only survive when the female survives, probability φ i,t is the joint probability that the female survives and the nest survives. We assumed each nest's state at time t was known, and x i,t was only defined for nest i from incubation initiation to the occasion of its VEGETATION INFLUENCE ON INCUBATING TURKEYS | 5 of 15 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License death or hatching (i.e., nest success). To ensure that the probabilities φ i,t , fm i,t , and of i,t were defined on the interval [0, 1] and summed to 1, we used a multinomial logit link function for φ i,t and fm i,t (Choquet 2008, Servanty et al. 2010). To test the differential effects of vegetation on nest survival and female mortality, we included individual nest ‐ varying vegetation covariates on φ i and fm i . Additionally, because timing may affect nest success (Keever et al. 2023), we included day of year of incubation initiation on φ i . We constructed one model for Oklahoma and one model for Tennessee because some of the vegetation metrics differed between sites. Rather than performing model selection, we evaluated support for covariates based on parameters ’ posterior distributions and effect size. We determined that variables were significant if >85% of their posterior distribution was above or below 0, indicating a >0.85 probability that the effect of that covariate was positive or negative, respectively (Arnold 2010, Link and Barker 2010). In Oklahoma, our 6 covariates (VO low, VO medium, VO high, percent woody, percent herbaceous, and day of year of incubation initiation) had low correlation (| r | < 0.51); therefore, we used all 5 vegetation covariates on fm i and all 6 covariates on φ i . In Tennessee, VO high, VO medium, and VO low were highly correlated (| r | > 0.70) but had low correlation with woody stem count (| r | < 0.01) or day of year of incubation initiation (| r | < 0.15). Therefore, we chose to use VO medium and woody stem count as covariates on fm i , and we used VO medium, woody stem count, and day of year of incubation initiation as covariates on φ i At both sites, we pooled nests across years without considering the effect of year because of small sample sizes in individual years in each state. We chose to include all nest attempts from each individual in the analysis as separate observations, as we assumed independence between nest attempts because turkeys generally renest in a F I G U R E 2 Conceptual figure outlining multistate modeling approach to consider risks to wild turkey ( Meleagris gallopavo ) nests and incubating females (hens). Analyses that only consider nest success follow part A, and analyses that only consider hen survival follow part B; a multistate approach links the nests and hens together as shown in part C. 6 of 15 | TURNER ET AL 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License different location (Yeldell et al. 2017, Moscicki et al. 2025). The known ‐ fate model is conditional on incubation initiation, so we removed any nests that failed on the same day incubation initiation began. We excluded females that were killed by humans (e.g., mowing over the nest; n = 20), which makes fm i the probability of natural mortality of females. We used Normal (0, τ = 0.001) priors for the effects of vegetation and time covariates, which we centered and scaled. Because this prior can put strong weight on extreme values for models with a logit trans- formation, we also ran our analyses using a Normal (0, τ = 0.5) prior on all covariates (Northrup and Gerber 2018). We fit the models in JAGS version 4.3.2 (Plummer 2003) using R version 4.5.2 (R Core Team 2025) and the R2jags package (Su and Yajima 2021). We ran 3 chains for 50,000 iterations each and discarded the first 10,000 iterations as burn ‐ in. We checked for chain convergence visually using trace plots and by verifying R ‐ hat values were <1.1 (Gelman and Rubin 1992). While we recognize the importance of assessing model fit in addition to convergence, Bayesian goodness of fit is an active area of research, and no standardized technique exists for a multinomial, time ‐ dependent, hierarchical model such as ours. R E S U L T S Sample size summary Our dataset included nest vegetation data from 206 females in Tennessee and 36 females in Oklahoma, and some females accounted for nesting attempts in multiple years (300 female years total), multiple attempts within a year, or both. These females accounted for 261 initial, 56 second, and 10 third nests in Tennessee and 39 initial, 10 second, and 1 third nests in Oklahoma. We recorded 103 successful nests in Tennessee and 5 successful nests in Oklahoma. Across sites, 40 females (16.5% of monitored females) were predated during incubation; thus, proportional female survival across all incubation attempts was 89.4%. Site ‐ specific proportional female survival during incubation was 84% in Oklahoma and 90.2% in Tennessee. Although our model estimated female mortality, we report female survival (1 − fm ) to maintain consistency for interpreting directional effects of covariates between female and nest survival. The time covariate (day of year of incubation initiation) did not influence nest survival at either of our sites. Changing the prior for covariates did not change the direction or magnitude of effect for any of our analyses (Appendix A, available in Supporting Information). In Tennessee, midstory stem count averaged 20.4 (range = 0 – 319) and VO medium averaged 68.9% (range = 2.5 – 100). In Oklahoma, understory woody vegetation coverage averaged 40.5% (range = 0 – 95) and understory herbaceous coverage averaged 29.0% (range = 2.5 – 75); VO low averaged 96.7% (range = 60 – 100), VO medium averaged 80.3% (range = 35 – 100), and VO high averaged 61.9% (range = 20 – 100). Nest fate In Tennessee, midstory stem count had a positive effect on nest survival (Figure 3; Table 1). At the mean value of all covariates, 28 ‐ day nest survival in Tennessee was 0.358 (95% Bayesian Credible Interval [CRI] = 0.310 – 0.409), and increasing woody stem count by one standard deviation (SD; 28.6 more stems) increased 28 ‐ day nest survival to 0.430 (95% CRI = 0.346 – 0.527). In Oklahoma, understory woody vegetation and VO medium had a positive effect on nest survival, whereas VO low had a negative effect on nest survival (Figure 4; Table 2). At the mean value of all covariates, 28 ‐ day nest survival in Oklahoma was 0.194 (95% CRI = 0.103 – 0.320). Increasing coverage of understory woody vegetation by one SD (28.0 percentage point change) increased 28 ‐ day nest survival to 0.340 (95% CRI = 0.199 – 0.620), increasing VO medium by one SD (19.8 percentage point change) increased 28 ‐ day nest survival to 0.311 (95% CRI = 0.119 – 0.562), and increasing VO low by one SD (8.0 percentage point change) decreased 28 ‐ day nest survival to 0.086 (95% CRI = 0.014 – 0.244). VEGETATION INFLUENCE ON INCUBATING TURKEYS | 7 of 15 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Female fate In Tennessee, VO medium had a positive effect on incubating female survival and midstory stem count had a negative effect on incubating female survival (Figure 5). At the mean value of all covariates, 28 ‐ day female survival in Tennessee was 0.860 (95% CRI = 0.811 – 0.900). Increasing VO medium by one SD (31.6 percentage point change) increased 28 ‐ day female survival to 0.884 (95% CRI = 0.816 – 0.931), and increasing midstory stem count by one SD (28.6 more stems) decreased 28 ‐ day female survival to 0.818 (95% CRI = 0.739 – 0.881). In Oklahoma, percent herbaceous and VO low had positive effects on incubating female survival (Figure 6), whereas understory woody vegetation and VO medium had negative effects on incubating female survival. At the mean value of all covariates, 28 ‐ day female survival in Oklahoma was 0.808 (95% CRI = 0.620 – 0.927). Increasing percent herbaceous by one SD (15.0 percentage point change) increased 28 ‐ day female survival to 0.943 (95% CRI = 0.766 – 0.991), increasing VO low by one SD (8.0 percentage point change) increased 28 ‐ day female survival to 0.920 (95% CRI = 0.723 – 0.984), increasing understory woody vegetation by one SD (28.0 percentage point change) decreased 28 ‐ day female survival to 0.632 (95% CRI = 0.293 – 0.866), and increasing VO medium by one SD (19.8 percentage point change) decreased 28 ‐ day female survival to 0.407 (95% CRI = 0.079 – 0.766). For Tennessee, we also checked that when using VO low or VO high in the model instead of VO medium, results were similar to those obtained with VO medium. The direction of effect of woody stem count held F I G U R E 3 Predicted 28 ‐ day wild turkey ( Meleagris gallopavo ) nest survival (and 95% Bayesian credible interval) in Tennessee, USA, using the range of observed values for variables with a significant effect in the multistate model, 2017 – 2022. All other variables held at their mean value. Woody stem count refers to the number of midstory stems around the nest within a 5 ‐ m radius. T A B L E 1 Effects of covariates on wild turkey ( Meleagris gallopavo ) nest survival ( φ ) and incubating female mortality ( fm ) in Tennessee, USA, 2017 – 2022, including lower (2.50%) and upper (97.5%) confidence limits (CL) of the 95% Bayesian credible interval. We determined that variables were significant if the probability of effect (i.e., proportion of posterior distribution on one side of 0) was above 0.85. Covariates include visual obstruction from 0.5 – 1 m (VO medium), midstory stems, and day of year that incubation initiation occurred. Vital rate Covariate Mean SD 2.50% CL 97.50% CL Probability of effect fm VO medium − 0.203 0.179 − 0.549 0.153 0.870 fm Midstory stems 0.280 0.147 − 0.024 0.561 0.966 φ VO medium − 0.006 0.077 − 0.158 0.142 0.527 φ Midstory stems 0.205 0.101 0.022 0.415 0.987 φ Day of year of incubation initiation 0.056 0.069 − 0.078 0.194 0.790 8 of 15 | TURNER ET AL 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License F I G U R E 4 Predicted 28 ‐ day wild turkey ( Meleagris gallopavo ) nest survival (and 95% Bayesian credible interval) in Oklahoma, USA, using the range of observed values for variables with a significant effect in the multistate model, 2022 – 2024. All other variables in each panel held at their mean value. Percent woody is the percent understory woody vegetation within a 1 ‐ m 2 plot centered on the nest, VO low is percent visual obstruction from 0 – 0.5 m, and VO medium is percent visual obstruction from 0.5 – 1 m. T A B L E 2 Effects of covariates on wild turkey ( Meleagris gallopavo ) nest survival ( φ ) and incubating female mortality ( fm ) in Oklahoma, USA, 2022 – 2024, including lower (2.50%) and upper (97.5%) confidence limits (CL) of the 95% Bayesian credible interval. We determined that variables were significant if the probability of effect (i.e., proportion of posterior distribution on one side of 0) was above 0.85. Covariates include visual obstruction from 0 – 0.5 m (VO low), visual obstruction from 0.5 – 1 m (VO medium), visual obstruction from 1 – 2 m (VO high), percent understory woody vegetation (percent woody), percent understory herbaceous vegetation (percent herbaceous), and day of year that incubation initiation occurred. Vital rate Covariate Mean SD 2.50% CL 97.50% CL Probability of effect fm VO low − 0.945 0.447 − 1.827 − 0.064 0.982 fm VO medium 1.464 0.636 0.284 2.774 0.994 fm VO high − 0.055 0.458 − 0.935 0.863 0.555 fm Percent woody 0.781 0.447 − 0.070 1.673 0.962 fm Percent herbaceous − 1.317 0.546 − 2.407 − 0.275 0.994 φ VO low − 0.426 0.263 − 0.995 0.033 0.963 φ VO medium 0.354 0.257 − 0.147 0.861 0.917 φ VO high − 0.072 0.219 − 0.510 0.346 0.623 φ Percent woody 0.592 0.231 0.145 1.054 0.995 φ Percent herbaceous − 0.196 0.228 − 0.643 0.251 0.807 φ Day of year of incubation initiation − 0.097 0.189 − 0.464 0.277 0.696 VEGETATION INFLUENCE ON INCUBATING TURKEYS | 9 of 15 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License consistent on nest and female survival, but VO low and VO high were not categorized as significant effects on female survival (Appendix B, available in Supporting Information). D I S C U S S I O N Visual obstruction and vegetation composition around the nest influenced nest and incubating female survival with slight differences in vegetation characteristics of importance between sites. Specifically, nest survival in Tennessee increased with greater midstory stem density, whereas incubating female survival increased with greater obstruction 0.5 – 1 m above ground, but decreased with greater midstory stem counts. Nest survival in Oklahoma increased with more understory woody vegetation, greater visual obstruction from 0.5 – 1 m, and less visual obstruction below 0.5 m. Incubating female survival in Oklahoma increased with greater understory herbaceous vegetation, greater visual obstruction below 0.5 m, less visual obstruction from 0.5 – 1 m, and less understory woody vegetation. Although we acknowledge the widening credible intervals at the extremes of covariate ranges, particularly with the smaller sample size in Oklahoma, we believe our results demonstrate a clear pattern of vegetation influencing female turkey survival at 2 different sites in the southeastern United States. Our results highlight the need to consider multiple vital rates in conjunction with vegetation structure near nest sites to better F I G U R E 5 Predicted 28 ‐ day survival (and 95% Bayesian credible interval) of nesting female wild turkeys ( Meleagris gallopavo ) in Tennessee, USA, using the range of observed values for variables with a significant effect in the multistate model, 2017 – 2022. All other variables in each panel held at their mean value. Variables include percent visual obstruction from 0.5 – 1 m (VO medium) and the number of midstory stems within a 5 ‐ m radius of the nest (woody stem count). F I G U R E 6 Predicted 28 ‐ day survival (and 95% Bayesian credible interval) of nesting female wild turkeys ( Meleagris gallopavo ) in Oklahoma, USA, using the range of observed values for variables with a significant effect in the multistate model, 2022 – 2024. All other variables in each panel held at their mean value. Percent woody is the percent understory woody vegetation within a 1 ‐ m 2 plot, percent herbaceous is the percent understory herbaceous vegetation within a 1 ‐ m 2 plot, VO low is percent visual obstruction from 0 – 0.5 m, and VO medium is percent visual obstruction from 0.5 – 1 m. 10 of 15 | TURNER ET AL 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License assess the effects of vegetation characteristics on wild turkey demography, as woody stems in particular had contrasting effects on incubating female and nest survival. Understory woody vegetation and midstory stem density had opposing effects on nest survival and incubating female survival at our sites. Several bird species exhibit tradeoffs associated with nesting cover, where vegetative cover that conceals the nest increases risk to the female (Wiebe and Martin 1998, Amat and Masero 2004, Miller et al. 2007), which may cause females to select nest sites in intermediate cover to balance the tradeoff of survival and nest success (Götmark et al. 1995). A dense woody understory >0.5 m aboveground or a woody midstory may increase the likelihood that a female is unable to escape the nest bowl during a predation attempt, whereas herbaceous vegetation may reduce the likelihood of predators detecting the nest without preventing her from escaping during an encounter. Alternatively, the female may feel more secure in dense woody vegetation and wait longer to flush when a predator approaches, increasing her likelihood of being killed. Predator density also could vary between areas with different vegetation structure and composition, altering female and nest risk. Swafford (2026) reported lower nest survival at nest sites in Tennessee with a greater predator index, and daily nest survival was most strongly related to coyote and bobcat indices. Regardless of the mechanism, our finding of woody vegetation decreasing female survival suggests there may be a need to promote cover dominated more by her- baceous plants, especially when we consider that herbaceous cover increased female survival in Oklahoma. Our results highlight the importance of concurrently considering the effects of vegetation on nest and female survival. Had we only considered nest survival, we would have missed effects of vegetation on female survival, some of which suggest survival tradeoffs. For example, in Oklahoma, visual obstruction measurements that had positive effects on one vital rate negatively affected the other. Similarly, woody vegetation (either in the understory or midstory) had positive effects on nest survival but negative effects on female survival at both sites. Contrasting effects on nest and female survival have been observed with behavior during incubation, as female turkeys that take longer recesses during incubation had greater survival but decreased nest success (Lohr et al. 2020). Even if certain vegetation metrics promote nest survival, they may have a negative effect on overall population trajectory given adult female survival tends to have the greatest effect on population growth rates in longer ‐ lived birds (Clark and Martin 2007). Specifically in turkeys, female survival has the greatest influence on population trajectory (Taylor et al. 2012, Londe et al. 2023, Tyl et al. 2025), and it appears to have declined concurrent with apparent declines in turkey populations (Lashley et al. 2025). Thus, we contend female survival likely is an understudied vital rate, and suggest future studies use a similar multistate model to simultaneously consider nest success and female survival when evaluating vegetation effects during reproduction. Differences in landscape composition, predator context, and vegetation measurements may have contributed to different effects of visual obstruction from 0.5 – 1 m on vital rates between sites. Other studies have commonly reported differences in vegetation metrics of importance and their effects (Kilburg et al. 2014, Johnson et al. 2022, Keever et al. 2023), and several have indicated that vegetation metrics had little influence on survival (Little et al. 2014, Yeldell et al. 2017, Lohr et al. 2020). Thus, it is reasonable that site ‐ level differences contribute to disparate results by making some metrics important at one site but not the other. The Oklahoma site was pre- dominantly pine forests with some interspersed pastures, whereas there was more diversity in landscape com- position at the Tennessee sites, including considerable interspersion of forests, hay or pasture, and agriculture. Turkey habitat selection varies with landscape composition (Ogawa et al. 2025), and the relative importance of various vegetation metrics to vital rates also may change with landscape context. Differences in vital rates could indicate there are different predator contexts, as Oklahoma reported several lower vital rates than Tennessee. Nest survival in 2022 – 2023 in Oklahoma was 17% (Butler et al. 2025) and was 33.9% in 2017 – 2018 in Tennessee (Johnson et al. 2022). Annual female survival was 40% in Oklahoma (Butler et al. 2025) and slightly greater in Tennessee (Buehler and Harper 2024). No poults from monitored females survived to 28 days in Oklahoma (Butler et al 2025), and poult survival was 15 – 20% in Tennessee (Quehl et al. 2025). These differences in vital rates suggest the predator community likely differed between sites. Thus, varying predator species and different predator densities could make overhead cover above the nest (i.e., 0.5 – 1 m in height) beneficial at one site but detrimental at VEGETATION INFLUENCE ON INCUBATING TURKEYS | 11 of 15 19372817, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/jwmg.70264 by University Of Florida, Wiley Online Library on [12/08/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License another. It also is possible that weather differences between sites influenced nest and female survival, though a recent analysis suggests a weaker relationship between weather and nest survival than previously thought (Boone et al. 2023). Finally, it is possible that differences in the vegetation metrics collected influenced our results. Visual obstruction from 0.5 – 1 m had contrasting effects on female survival between Tennessee and Oklahoma, but we did not collect the same understory composition metrics in Tennessee that influenced survival in Oklahoma. Therefore, the positive effect of visual obstruction from 0.5 – 1 m on female survival may have been related to unmeasured herbaceous vegetation at nests in Tennessee, given herbaceous coverage had a positive effect in Oklahoma. Regardless of the mechanism for the differing effect of visual obstruction from 0.5 – 1 m on female survival between sites, the consistent negative effect of woody vegetation highlights the potential influence of vegetation compo- sition and structure on wild turkey vital rates. Assuming that vegetation management can positively influence wild turkey population growth, our results indicate that vegetation metrics must be considered jointly for nest success and survival of the incubating female. Both vital rates are important to population trajectory (Londe et al. 2023), but considering only vegetation influ- ences on nest success ignores an important linkage to the sur