Offspring survival is generally more variable than adult survival and may limit population growth. Although white-tailed deer neonate survival has been intensively investigated, recent work has emphasized how specific cover types influence neonate survival at local scales (single study area). These localized investigations have often led to inconsistences within the literature. Developing specific hypotheses describing the relationships among environmental, habitat, and landscape factors influencing white-tailed deer neonate survival at regional scales may allow for detection of generalized patterns. Therefore, we developed 11 hypotheses representing the various effects of environmental (e.g., winter and spring weather), habitat (e.g., hiding and escape cover types), and landscape factors (e.g., landscape configuration regardless of specific cover type available) on white-tailed deer neonate survival up to one-month and from one- to three-months of age. At one-month, surviving fawns experienced a warmer lowest recorded June temperature and more June precipitation than those that perished. At three-months, patch connectance (percent of patches of the corresponding patch type that are connected within a predefined distance) positively influenced survival. Our results are consistent with white-tailed deer neonate ecology: increased spring temperature and precipitation are likely associated with a flush of nutritional resources available to the mother, promoting increased lactation efficiency and neonate growth early in life. In contrast, reduced spring temperature with increased precipitation place neonates at risk to hypothermia. Increased patch connectance likely reflects increased escape cover available within a neonate’s home range after they are able to flee from predators. If suitable escape cover is available on the landscape, then managers could focus efforts towards manipulating landscape configuration (patch connectance) to promote increased neonate survival while monitoring spring weather to assess potential influences on current year survival.
Citation: Michel ES, Jenks JA, Kaskie KD, Klaver RW, Jensen WF (2018) Weather and landscape factors affect white-tailed deer neonate survival at ecologically important life stages in the Northern Great Plains. PLoS ONE 13(4): e0195247. https://doi.org/10.1371/journal.pone.0195247
Editor: Marco Festa-Bianchet, Université de Sherbrooke, CANADA
Received: April 13, 2017; Accepted: March 19, 2018; Published: April 5, 2018
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This project was funded by Federal Aid to Wildlife Restoration administered through North Dakota Game and Fish Department (Project W-67-R-57, Study No. C-VIII to WFJ and W-67-R57 to JAJ). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Ungulate population dynamics are influenced by multiple factors including environmental variables [1–3], density-dependent effects of forage resources [4–6], and predation [7,8]. Although adult survival rates are generally stable, juvenile survival may display more variability and thus, potentially drive population dynamics [9,10]. Therefore, understanding how factors influence juvenile survival is necessary to predict annual recruitment and its impact on population size.
Assessing how various factors such as predation and local site habitat composition influence white-tailed deer (Odocoileus virginianus; hereafter, deer) neonate survival has been of recent interest. Many studies have emphasized cause-specific neonate mortality [11–13] and have found coyotes (Canis latrans) to be the primary predator; though, predator management is often an ineffective prescription for improving population response [14, 15]. Other studies also have assessed which local habitat characteristics are associated with neonate survival across the range of white-tailed deer [12, 16–18]. Understanding which habitat characteristics influence neonate survival is important as hiding and escape cover are generally limiting in highly fragmented landscapes . For example, in the Northern Great Plains, native prairie and planted grasslands have been widely converted to row crop agriculture , and Grovenburg et al.  showed that vegetation height in Conservation Reserve Program grasslands positively influenced neonate deer survival. Unfortunately, most studies assessing these relationships are site specific and inconsistencies occur [11, 22, 23]. Nevertheless, the highly fragmented landscape of the Northern Great Plains results in patches of habitat that vary in size and distribution. This landscape allows for a unique opportunity to test specific hypotheses regarding the effect of environmental (e.g., winter and spring weather), habitat (e.g., hiding and escape cover types), and landscape level factors (e.g., landscape configuration regardless of specific cover type available) on deer neonate survival and to identify general patterns at a regional scale.
Neonatal deer behaviour is heavily dependent upon age and may interact with landscape or environmental variables. For example, newborns display various predator avoidance strategies as they age, moving from a hider strategy (that incorporates fear bradycardia) early in life to a flight response from predators later in life [24, 25]. Similarly, studies show that neonate survival also varies by age [11, 26, 27], indicating that extrinsic factors likely influence neonate survival during the first weeks of life. Our objective was to assess how environmental, habitat, and landscape factors influenced deer neonate survival up to one-month (28 days) and between one- and three-months (90 days) of age using a long-term dataset collected from three states that largely comprise the Northern Great Plains (Minnesota, South Dakota, and North Dakota, USA).
We developed four general groups of hypotheses to examine how various ecological factors affect fawn survival during early life: hiding and escape cover hypotheses, weather hypotheses, a landscape configuration hypothesis, and a nutritional resource hypothesis. Our hiding and escape cover hypotheses examined whether the amount of forested area, grasslands, pasturelands, and wetlands affected neonate survival . We expected that neonate survival would decrease with increasing forested cover and would increase with cover provided by grasslands, pasturelands, and wetlands. Our weather hypotheses examined the influence of winter [28–30] and spring weather [21, 31] and also examined if there was a potential lag effect of prior year weather  on neonate survival. Winter weather may negatively affect maternal body condition, thus influencing her ability to successfully rear offspring. Therefore, we expected that winter weather (represented by a deer winter severity index and also the independent effects of lowest winter temperature and total winter snow accumulation) would negatively affect survival. Similarly, inclement spring weather (represented with the lowest June temperature and total June precipitation) may negatively affect a neonate’s ability to thermoregulate and as a result, decrease survival. However, effects of adverse spring weather may be counteracted if proper habitat is available so we developed a thermal cover hypothesis to examine if the potentially negative effects of low June temperatures and increased total June precipitation were alleviated with increased availability of grassland, pastureland, and wetland cover.
In addition to the type and amount of hiding and escape cover that influences neonate survival, the configuration of hiding and escape cover on the landscape can also affect neonate survival . We expected that neonate survival would increase with increasing patch density, connectivity among patches, and complexity of the patch shape of hiding and escape cover. Finally, row crop agriculture is prevalent in the Northern Great Plains region but is generally unavailable early in the parturition season and can negatively affect neonate survival if its presence results in a decreased amount of other hiding and escape cover , though we would expect that an increased amount of open water would positively affect maternal lactation efficiency  thus positively affecting neonate survival.
Materials and methods
We assessed factors that influenced neonate deer survival from 10 counties comprising 8 study sites across 3 states in the Northern Great Plains region from 2001 to 2015. We captured fawns in Walsh, Grand Forks, Grant, and Dunn counties, North Dakota; Brookings, Edmunds, and Perkins counties, South Dakota; and Lincoln, Pipestone, and Redwood counties, Minnesota (Fig 1). All counties were located within 4 Level III Ecoregions [34, 35]: Lake Agassiz Plain (Walsh and Grand Forks counties, North Dakota), Northwestern Great Plains (Grant and Dunn counties, North Dakota and Perkins County, South Dakota), Northern Glaciated Plains (Edmunds and Brookings counties, South Dakota and Lincoln County, Minnesota), and the Western Corn Belt Plains (Pipestone and Redwood counties, Minnesota). State agencies granted permission for publicly owned properties while we obtained permission from landowners to collect data from privately owned lands.
Study sites where we captured neonate white-tailed deer located in a.) Lincoln, b.) Pipestone, and c.) Redwood counties, Minnesota; d.) Walsh, e.) Grand Forks, f.) Grant, and g.) Dunn counties North Dakota; h.) Brookings, i.) Edmunds, and j.) Perkins counties, South Dakota from 2001 to 2015. Lincoln and Pipestone counties, Minnesota were combined to create the Lincoln-Pipestone study area and Walsh and Grand Forks counties, North Dakota were combined to create the Walsh-Grand Forks study area used for analyses.
There was variation in weather variables and vegetation communities among study sites. For example, thirty-year (1981–2010) mean annual precipitation ranged from 412 mm (Grant County, North Dakota) to 729 mm (Redwood County, Minnesota), and thirty-year mean temperatures ranged from a winter low of -15.1°C (Grant County, North Dakota) to a summer high of 30.3°C (Perkins County, South Dakota ). Vegetation types were generally classified as Northern Wheatgrass-Needlegrass Plains, Northern Mixed Grass Prairie, and the Tallgrass Prairie . As such, there was variation in the vegetation composition for each study area. Specific vegetation types are described in [11, 38, 39].
We used reproductive female postpartum behaviour as an indicator of presence of neonates [40–42], and we also used the aid of Vaginal Implant Transmitters (Advanced Telemetry Systems, Inc., Isanti, MN, USA) to assist in neonate capture . We captured neonates by hand or net after locating them. We wore latex gloves and stored all radio-collars and other equipment in natural vegetation to minimize scent transfer. We fitted neonates with expandable breakaway radio-collars (Advanced Telemetry Systems, Isanti, MN, USA or Telonics Inc., Mesa, AZ). We kept handling time under five minutes when possible to reduce capture-related stress. Fawn capture methods were generally similar among study sites. For additional information regarding capture methods see [11, 22, 39].
We monitored neonates daily for the first 30 days using a truck-mounted null-peak antenna system , hand-held Yagi antennas (Advanced Telemetry Systems), aerial telemetry, and omnidirectional whip antennas and used these same techniques to monitor neonates 2–3 times per week thereafter. We investigated mortalities immediately after detecting a mortality signal and transported carcasses to either the North Dakota Game and Fish Department Wildlife Laboratory in Bismarck, North Dakota, USA or to the Animal Disease Research Diagnostic Laboratory at South Dakota State University, South Dakota, USA depending on study site.
We followed the American Society of Mammalogists guidelines for mammal care and use , and the South Dakota State University Institutional Animal Care and Use Committee approved all handling protocols (Approval numbers: 00-A038, 02-A037, 02-A043, 04-A009, 10-006E, 13-091A).
We did not calculate home ranges for each fawn as we did not have relocation data available. Instead, we quantified habitat characteristics within a 352.3-m circular buffer around a neonate’s capture location, which represented a 39-ha mean core (50%) home range for a one-month old neonate . For three-month old neonates, we quantified habitat characteristics in a 669.7-m circular buffer around a neonate’s capture location, which represented a 140.9 ha (95%) summer home range . We quantified habitat characteristics within a larger area for three-month old neonates to represent the increased area available to them once they began following their mother at heel. We overlaid buffered areas with the 2011 National Land Cover Database  and calculated habitat composition (% composition of each cover type in each buffer) using ArcGIS 10.4.1 (ESRI, Inc., Redlands CA). We then reclassified land cover data into 4 cover habitat categories (grassland/herbaceous, pastureland, wetland, and forest) and two nutritional resource categories (cultivated crop and open water).
The relative influence of the configuration of specific habitat types on neonate survival varies (e.g., [11, 22]). Our goal was to simply assess whether the amount of a specific habitat type (e.g., percent grassland/herbaceous, percent wetland) or the general landscape configuration of all habitat types found within a neonate’s home range influenced survival. Therefore, we did not test how landscape configuration of specific habitat types (e.g., number of wetland patches) may influence neonate survival; rather, we investigated how general landscape configuration, regardless of available habitat (e.g., total number of patches of all habitat types), influenced survival. We assessed how patch density, connectance, and shape index influenced survival as these variables represent the amount of escape cover available (patch density), connectivity of escape cover (connectance), and shape complexity (shape index; ). Examining these variables better informed us as to whether general landscape configuration or the amount of specific habitat types available more affected neonatal survival.
We obtained specific weather data from the National Oceanic and Atmospheric Administration website . We used weather stations that were located closest to fawn capture locations when possible. However, full datasets were not always available from these weather stations. If unavailable, we used the next closest weather station with fully available datasets. We also calculated Deer Winter Severity Indices (DWSI, ) from 2000 to 2015. We awarded one point each day the mean temperature was ≤ -7°C from November to April. The index received an additional point for each day mean snow depth was ≥ 350mm during this same time period.
We developed 11 hypotheses that represent the effects of hiding and escape cover, weather, landscape configuration, and nutritional resources on neonate survival (Tables 1 and 2). We chose not to include intrinsic models (e.g., fawn body mass, fawn age, maternal body mass, maternal age) because the variation in which neonates were caught would introduce bias into the analysis . Additionally, deriving specific survival rates would have been informative, but although neonates were monitored on a weekly basis, we did not have weekly encounter histories available for all neonates and thus, we only knew their ultimate fate. Therefore, we assessed factors that influenced survival up to one- and from one- to three-months using a logistic regression via the glm function in Program R (version 3.3.1; ). We adopted the Information–Theoretic Approach, and after analyzing each model, ranked them according the Akaike’s Information Criterion corrected for small sample size (AICc) and considered models within 2 ΔAICc units as competing . We derived AICc values, number of parameters, and model weights using the AICc and Weights functions in the MuMIn package in Program R . We assessed correlation among explanatory variables using a Pearson’s correlation within the cor.test function in Program R and included multiple variables in a single model when |r| ≤ 0.50. We considered variables important when their 95% Confidence Intervals (95% CIs) excluded 0 [52, 54]. We also estimated as a measure of goodness of fit of our global models . We present means ± standard deviation.
Amount of hiding and escape cover in our analyses varied because of our wide geographic range of study areas; magnitude among variables also varied (i.e., weather variables compared to cover variables). Therefore, we scaled the amount of hiding and escape cover and landscape variables by study site using the scale function in Program R with the mean centered on 0 and standard deviation of ± 1. We attempted to scale variables by county of capture; however, sample size precluded us from doing so for all counties. Therefore, we combined adjoining counties within a single state to scale variables for counties with small sample sizes. We scaled all weather variables across the dataset following the same procedure.
We captured 370 neonates from 2001 to 2015. From our one-month analysis, we censored 39 neonates, comprising 3 dropped collars, 1 lost signal, and 35 due to missing data. From our three-month analysis, we censored 105 neonates, comprising 20 dropped collars, 6 lost signals or collar malfunctions, and 66 that did not survive past one-month of age. Overall, we observed a high proportion of neonates surviving (one-month survival = 82%, one- to three-month survival = 92%). However, we observed a wide range of summer survival among populations (0.35, ; 0.94, ). Variables used within our models were not correlated (|r| ≤ 0.34). Mean distance to weather stations from neonate capture locations was 39.0 ± 24.4 km.
We observed two competing models that best described one-month survival (Table 3). Spring weather was our top model, which accounted for a moderate amount of model weight (wi = 0.37). This model included the effects of total June precipitation (β = 0.313, 95% CI = 0.029–0.601, n = 331) and lowest June temperature (β = 0.334, 95% CI = 0.048–0.625, n = 331; Table 4). Our thermal cover model seemed to be competing (ΔAICc = 0.11, wi = 0.35) with our top model. This model included the effects of hiding and escape cover (percent grassland/herbaceous, wetland, and pastureland cover), total June precipitation, and the lowest recorded June temperature. However, 95% CIs overlapped 0 for cover variables but excluded 0 for total June precipitation (β = 0.318, 95% CI = 0.031–0.608, n = 331) and lowest June temperature (β = 0.302, 95% CI = 0.013–0.596, n = 331; Table 4); therefore, we concluded that hiding and escape cover variables were uninformative  and only interpret our spring weather model. McFadden’s R2 for our spring weather model was 0.02 and was 0.92 indicating overdispersion of the data did not occur . All other models were ≥3.32 ΔAICc units away from our top model.
Models within 2 ΔAICc are competing, wi indicates model weight, and K indicates number of parameters calculated within a model.
95% CIs excluding 0 indicate variable influenced survival.
Both lowest recorded temperature in June and total June precipitation positively influenced one-month neonate survival. The mean lowest recorded temperature in June for surviving neonates ( = 4.7 ± 1.8°C, n = 268; Fig 2) was about 11% warmer than for neonates that perished ( = 4.2 ± 1.7°C, n = 63). For mean total June precipitation, surviving neonates ( = 111 ± 27 mm, n = 268; Fig 3) experienced about 7% more rainfall than neonates that perished ( = 104 ± 28 mm, n = 63).
We observed one top model describing survival from one- to three-months (Table 5). Our top model was our landscape model (wi = 0.50), which included the influence of patch density, connectance, and shape index, with 95% CIs excluding 0 for connectance (β = 0.450, 95% CI = 0.079–0.843, n = 265; Table 6). McFadden’s R2 for our spring weather model was 0.06 and for our global model was 0.49 indicating overdispersion of the data did not occur . All other models were ≥ 2.44 ΔAICc units from the top model. Surviving neonates ( = 77.0 ± 13.3%, n = 245; Fig 4) were exposed to about 13% more connectance among patches compared to neonates that perished ( = 68.1 ± 24.3%, n = 20).
wi indicates model weight, and K indicates number of parameters calculated within a model.
We acknowledge that our low R2 values indicate low predictive power for our top models, though R2 values associated with logistic models should be interpreted with caution . The low predictive power of our top models may be related to the relatively high proportion of fawns that survived to one-month (82%) and from one- to three-months of age (92%), which likely reduced the amount of variation that could be explained in our models. Conversely, there may have been little variability associated with our explanatory variables and/or our explanatory variables may have served as a proxy for another unmeasured variable. Therefore, although we obtained a large sample size (one-month survival = 331 neonates and three-month survival = 265 neonates) with high spatial (10 counties spanning three states) and temporal (15-years) replication, interpretation and use of our results for management purposes should be done so with caution.
Our results support our spring weather hypothesis and suggest that two spring weather variables most influence one-month neonate survival, though the direction of relationships differed from what we expected. We predicted that lowest June temperature and total June precipitation would negatively influence one-month survival due to neonates being unable to thermoregulate at a young age [21, 31]. However, we found positive relationships between lowest June temperature and total June precipitation and one-month survival, indicating that survival increased with increasing temperature and precipitation during June. This positive effect indicates that neonates experiencing warmer and wetter springs may be better able to thermoregulate compared those experiencing cooler and dryer springs. Alternatively, neonates may be more affected by maternal care related to forage quality and quantity during spring months than the inability to thermoregulate. For example, increased temperatures and precipitation were related to increased plant quantity and quality [57, 58], which may indirectly influence neonate survival through improved maternal lactation efficiency [59, 60]. Our results also support previous literature discussing the importance of available forage during maternal lactation [61–63] and other research reporting increased neonate survival following increased rain events  and warmer temperatures  during spring. Spring weather may therefore be an important factor for understanding neonatal deer survival, though small annual changes (≤ 11% difference in June temperature and precipitation between surviving neonates and those that perished) likely do not greatly impact overall survival as 82% of neonates in our study survived to one-month.
Our results describing neonate survival between one- and three-months supported our landscape hypothesis with patch connectance being the only important variable in the model. Our findings support Grovenburg et al.  who found that the probability of deer neonates eluding predators decreased with increased distance to grassland and wetland patches (i.e., areas with less connectance). Landscape metrics also have been found to influence neonate survival in other studies [22, 27, 66]. Grovenburg et al.  found that patch density of grassland and wetland cover types positively influenced neonate survival, while Rohm et al.  reported that surviving neonates inhabited home ranges that contained few large and irregular shaped forest patches. Gulsby et al.  also reported that coyote depredation on neonate white-tailed deer decreased with increasing amount of edge found within neonate’s home ranges. Increased escape cover and its proximity to a neonate increases probability of survival, but again, the magnitude of its impact is likely not great considering the relatively high proportion (92%) of fawns surviving from one- to three-months of age.
Lack of hiding and escape cover may potentially influence population dynamics for generalist species such as deer . Our results show that general landscape configuration (connectance) and not specific cover type (percent grassland/herbaceous, wetland, and pastureland) influenced neonate survival; however, reported effects of hiding and escape cover on neonate survival is inconsistent. For example, in northcentral South Dakota, USA, the percent of Conservation Reserve Program grasslands and wetlands found in a neonate’s home range positively influenced survival, while percent forested cover negatively influenced neonate survival . Also, percent forested cover positively influenced neonate survival in southern Illinois . However, percent wetland, cropland, grassland, and forested cover found in a neonate’s home range did not influence survival in South Dakota, USA and Minnesota, USA , nor did percent herbaceous cover influence neonate survival in Pennsylvania, USA . Our cover habitat variables described the amount of hiding cover available to a neonate, not use, and therefore, may only be coarsely related to neonate survival. Similarly, microhabitat characteristics do not seem to better explain neonate survival, as Chitwood et al.  failed to detect any influence of vegetation associated with bed site characteristics on fawn survival in North Carolina, USA. Therefore, general landscape configuration in addition to the type and amount of hiding and escape cover present could be assessed when studying factors that affect neonate survival.
Our results did not support our winter weather hypotheses for one- and between one- and three-month survival and were outperformed by the null model in both candidate sets. Winter weather negatively influenced parturition date, birth mass, and litter size for Soay sheep (Ovis aries; ), fall fawn recruitment of mule deer (O. hemionus; ), and negatively influenced overwinter survival during an offspring’s first year of life (elk, Cervus elaphus, ; mule deer, ). However, winter weather did not influence calf body mass (reindeer, Rangifer tarandus; ) or neonate survival (white-tailed deer, ; elk, ). These discrepancies may be related to an individual species breeding strategy. For example, although white-tailed deer display tendencies of both a capital and income breeder (discussed in ), the greatest portion of fetal development occurs in the third trimester, which generally coincides with spring green-up [71, 72]. This delayed developmental strategy allows female white-tailed deer to avoid negative effects of severe winters on their offspring’s development while in utero.
Our goal was not to assess cause-specific mortality of white-tailed deer neonates; yet, it is important to note that coyotes are generally the leading cause of neonate mortality in this region [11, 16, 22]. Our finding that increased patch connectance is positively related to neonate survival between one- and three-months of age indicates that, within a neonate’s home range, manipulating habitat already found on the landscape may have a negative effect on a coyote’s search efficiency. However, increasing the amount of hiding and escape cover available through active habitat management to promote increased quantity and quality of individual bed sites did not improve fawn survival in the southeastern United States [13, 14] where fawn survival is substantially lower (as low as 14%; ) than what we report for our study. Nevertheless, active habitat management has other benefits such as increasing forage quantity  which, in turn, may positively influence lactation efficiency [53, 54]. Future research in the highly fragmented landscape in the Northern Great Plains could focus on addressing how manipulating the landscape may disrupt a coyote’s search pattern, which may decrease their effectiveness as a predator of neonatal deer.
White-tailed deer are a highly managed species in North America with a current emphasis being placed on assessing factors influencing neonate recruitment [10, 13, 74]. Our results can be used by managers in the Northern Great Plains to improve neonate hiding cover, though managers must be realistic with their expectations regarding the magnitude of increased neonate survival. Regardless, if suitable cover is available on the landscape, then managers could focus efforts on manipulating landscape configuration rather than promoting specific cover types. Managers may accomplish this by increasing the connectance (percent of patches of a corresponding patch type) within 140.9 ha sections representing a neonate’s 95% core summer home range. Doing so will increase the escape cover available to a neonate at a large scale. Additionally, managers could monitor spring weather variables such as lowest recorded June temperature and amount of June precipitation so those parameters can be incorporated into population models allowing for more refined population estimates.
S1 File. Data used to assess factors that influenced white-tailed deer neonate survival up to one-month of age.
We thank the North Dakota Game and Fish Department, South Dakota Department of Game, Fish and Parks, the Minnesota Department of Natural Resources, the Department of Natural Resource Management at South Dakota State University, and numerous private landowners, graduate students, and technicians for their help and cooperation with this project. We also thank M. Apollonio, B. Pokorny, and an anonymous reviewer for their helpful comments. This project was funded by Federal Aid to Wildlife Restoration administered through North Dakota Game and Fish Department (Project W-67-R-57, Study No. C-VIII). Any mention of trade, product, or firm names is for descriptive purposes only, and does not imply endorsement by the U.S. Government.
- 1. Post E, Stenseth NC. Climatic variability, plant phenology, and northern ungulates. Ecology. 1999;80: 1322–1339.
- 2. Forchhammer MC, Post E, Stenseth NC, Boertmann DM. Long-term responses in arctic ungulate dynamics to changes in climatic and trophic processes. Popul Ecol. 2002;44: 113–120.
- 3. Post E, Forchhammer MC. Climate change reduces reproductive success of an Arctic herbivore through trophic mismatch. Philos T Roy Soc B. 2008;363: 2369–2375.
- 4. Gaillard JM, Boutin JM, Delorme D, Laere GV, Dunkin P, Lebreton JD. Early survival in roe deer: causes and consequences of cohort variation in two contrasted populations. Oecologia. 1997;112: 502–513. pmid:28307627
- 5. Coulson T, Milner-Gulland EJ, Clutton-Brock T. The relative roles of density and climatic variation on population dynamics and fecundity rates in three contrasting ungulate species. Proc R Soc Lond B Biol Sci. 2000;267: 1771–1779.
- 6. Coulson T, Guinness F, Pemberton J, Clutton-Brock T. The demographic consequences of releasing a population of red deer from culling. Ecology. 2004;85: 411–422.
- 7. Griffin KA, Hebblewhite M, Robinson HS, Zager P, Barber-Meyer SM, Christianson D, et al. Neonatal mortality of elk driven by climate, predator phenology and predator community composition. J Anim Ecol. 2011;80: 1246–1257. pmid:21615401
- 8. Eacker DR, Hebblewhite M, Proffitt KM, Jimenez BS, Mitchell MS, Robinson HS. Annual elk calf survival in a multiple carnivore system. J Wildl Manage. 2016;80: 1345–1359.
- 9. Gaillard JM, Festa-Bianchet M, Yoccoz NG. Population dynamics of large herbivores: variable recruitment with constant adult survival. Trends Ecol Evol. 1998;13: 58–63. pmid:21238201
- 10. Chitwood MC, Lashley MA, Kilgo JC, Moorman CE, Deperno CS. White‐tailed deer population dynamics and adult female survival in the presence of a novel predator. J Wildl Manage. 2015;79: 211–219.
- 11. Grovenburg TW, Swanson CC, Jacques CN, Klaver RW, Brinkman TJ, Burris BM, et al. Survival of white-tailed deer neonates in Minnesota and South Dakota. J Wildl Manage. 2011;75: 213–220.
- 12. Kilgo JC, Ray HS, Vukovich M, Goode MJ, Ruth C. Predation by coyotes on white-tailed deer neonates in South Carolina. J Wildl Manage. 2012;76: 1420–1430.
- 13. Chitwood MC, Lashley MA, Kilgo JC, Pollock KH, Moorman CE, DePerno CS. Do biological and bedsite characteristics influence survival of neonatal white-tailed deer? PLoS ONE. 2015;10: e0119070. pmid:25734333
- 14. Kilgo JC, Vukovich M, Ray HS, Shaw CE, Ruth C. Coyote removal, understory cover, and survival of white-tailed deer neonates. J Wildl Manage. 2014;78: 1261–1271.
- 15. Conner LM, Morris G. Impacts of mesopredator control on conservation of mesopredators and their prey. PLoS ONE. 2015;10: e0137169. pmid:26361211
- 16. Brinkman TJ, Jenks JA, DePerno CS, Haroldson BS, Osborn RG. Survival of white-tailed deer in an intensively farmed region of Minnesota. Wildl Soc Bull. 2004;32: 726–731.
- 17. Vreeland JK, Diefenbach DR, Wallingford BD. Survival rates, mortality causes, and habitats of Pennsylvania white-tailed deer fawns. Wildl Soc Bull. 2004;32: 542–553.
- 18. Burroughs JP, Campa H, Winterstein SR, Rudolph BA, Moritz WE. Cause-specific mortality and survival of white-tailed deer fawns in southwestern lower Michigan. J Wildl Manage. 2006;70: 743–751.
- 19. Bender DJ, Contreras TA, Fahrig L. Habitat loss and population decline: A meta-analysis of the patch size effect. Ecology. 1998;79: 517–533.
- 20. Wright CK, Wimberly MC. Recent land use change in the Western Corn Belt threatens grasslands and wetlands. Proc Natl Acad Sci. 2013; 110:4134–4139. pmid:23431143
- 21. Grovenburg TW, Jacques CN, Klaver RW, Jenks JA. Bed site selection by neonate deer in grassland habitats on the Northern Great Plains. J Wildl Manage. 2010;74: 1250–1256.
- 22. Grovenburg TW, Klaver RW, Jenks JA. Survival of white-tailed deer fawns in the grasslands of the Northern Great Plains. J Wildl Manage. 2012;76: 944–956.
- 23. Schaffer BA, Jenks JA, Grovenburg TW, Jensen WF. Bed-site selection by neonatal white-tailed deer in central North Dakota. Prairie Naturalist. 2014;46: 34–38.
- 24. Lent PC. Mother-infant relationships in ungulates. In Geist V, Walther F, editors. Symposium on the behavior of ungulates and its relation to management. IUCN Publication 24. Switzerland. 1974. pp. 14–55.
- 25. Carl GR, Robbins CT. The energetic cost of predator avoidance in neonatal ungulates: hiding versus following. Can J Zool. 1988;66: 239–246.
- 26. Nelson TA, Woolf A. Mortality of white-tailed deer fawns in Southern Illinois. J Wildl Manage. 1987;51: 326–329.
- 27. Rohm JH, Nielsen CK, Woolf A. Survival of white-tailed deer fawns in southern Illinois. J Wildl Manage. 2007;71: 851–860.
- 28. Ciuti S, Jensen WF, Nielsen SE, Boyce MS. Predicting mule deer recruitment from climate oscillations for harvest management on the Northern Great Plains. J Wildl Manage. 2015;79: 1226–1238.
- 29. Smith BL, Anderson SH. Juvenile survival and population regulation of the Jackson elk herd. J Wildl Manage, 1998;62: 1036–1045.
- 30. Hurley MA, Hebblewhite M, Lukacs PM, Nowak JJ, Gaillard JM, Bonenfant C. Regional-scale models for predicting overwinter survival of juvenile ungulates. J Wildl Manage. 2017;
- 31. Linnell JDC, Aanes R, Andersen R. Who killed Bambi? The role of predation in the neonatal mortality of temperate ungulates. Wildlife Biol. 1995;1: 209–223.
- 32. Grovenburg TW, Klaver RW, Jenks JA. Spatial ecology of white-tailed deer fawns in the Northern Great Plains: Implications of loss of conservation reserve program grasslands. J Wildl Manage. 2012;76: 632–644.
- 33. Jacques CN, Jenks JA, Grovenburg TW, Klaver RW. Influence of habitat and intrinsic characteristics on survival of neonatal pronghorn. PLoS ONE. 2015. 10:e0144026. pmid:26630484
- 34. Omernik JM. Map Supplement: Ecoregions of the Conterminous United States. Ann Am Assoc Geogr. 1987;77: 118–125.
- 35. Bryce SJ, Omernik M, Pater DE, Ulmer M, Schaar J, Freeouf J, et al. Ecoregions of North Dakota and South Dakota. U.S. Geological Survey, Jamestown, North Dakota. 1998 Available from: ftp://newftp.epa.gov/EPADataCommons/ORD/Ecoregions/sd/ndsd_front.pdf.
- 36. North Dakota Office of Climatology. 2016. North Dakota climate and weather. 2016. Available from: http://www.ndsu.edu/ndsco/.
- 37. Johnson JR, Larson GE. 2007. Grassland plants of south Dakota and the Northern Great Plains. Agricultural Experiment Station Bulletins. Paper 569. Available from: http://openprairie.sdstate.edu/agexperimentsta_bulletins/569.
- 38. Sternhagen KM. An evaluation of life history parameters and management of white-tailed deer (Odocoileus virginianus) in the red river valley of northeastern North Dakota. M. Sc. Thesis, South Dakota State University. 2015. Available from: http://openprairie.sdstate.edu/etd/1130/.
- 39. Moratz KL. Effect of oil and gas development on survival and health of white-tailed deer in the western Dakotas. M. Sc. Thesis, South Dakota State University. 2016. Available from: http://openprairie.sdstate.edu/etd/1122/.
- 40. Downing RL, McGinnes BS. Capturing and marking white-tailed deer fawns. J Wildl Manage. 1969;33: 711–714.
- 41. White M, Knowlton FF, Glazener WC. Effects of dam newborn fawn behavior on capture and mortality. J Wildl Manage. 1972;36: 897–906.
- 42. Huegel CN, Dahlgren RB, Gladfelter HL. Use of doe behavior to capture white-tailed deer fawns. Wildl Soc Bull. 1985;13: 287–289.
- 43. Swanson CC, Jenks JA, DePerno CS, Klaver RW, Osborn RG, Tardiff JA. Does the use of vaginal-implant transmitters affect neonate survival rate of white-tailed deer Odocoileus virginianus? Wildlife Biol. 2008;14: 272–279.
- 44. Brinkman TJ, DePerno CS, Jenks JA, Haroldson BS, Erb JD. A vehicle-mounted radiotelemetry antenna system design. Wildl Soc Bull. 2002;30: 256–258.
- 45. Sikes RS. 2016 Guidelines of the American Society of Mammalogists for the use of wild mammals in research and education. J Mammal. 2016;97: 663–688.
- 46. Homer CG, Dewitz JA, Yang L, Jin S, Danielson P, Xian G, et al. Completion of the 2011 National Land Cover Database for the conterminous United States–Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing. American Society for Photogrammetry and Remote Sensing. 2015;81: 345–354.
- 47. McGarigal K, Cushman SA, Ene E. FRAGSTATS v4: Spatial pattern analysis program for categorical and continuous maps. Computer software program produced by the authors at the University of Massachusetts, Amherst. 2012. Available from: http://www.umass.edu/landeco/research/fragstats/fragstats.html.
- 48. NOAA: National Centers for Environmental Information. Data Tools: Daily Weather Records. 2016. Available from: https://www.ncdc.noaa.gov/cdo-web/datatools/records.
- 49. Brinkman TJ, DePerno CS, Jenks JA, Haroldson BS, Osborn RG. Movement of female white-tailed deer: effects of climate and intensive row-crop agriculture. J Wildl Manage. 2005;69: 1099–1111.
- 50. Haskell SP, Ballard WB, Butler DA, Tatman NM, Wallace MC, Kochanny CO et al. J Mammal. 2007;88: 1482–1487.
- 51. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2015. Available from: https://www.R-project.org/.
- 52. Burnham KP, Anderson DR. Model selection and multimodel inference: a practical information-theoretic approach. 2nd edition. New York: Springer-Verlag New York, Inc; 2002.
- 53. Barton K. MuMIn: Multi-model inference. R package version 1.15.6. 2016 Available from: https://CRAN.R-project.org/package=MuMIn. Accessed 30 Dec 2016.
- 54. Arnold TW. Uninformative parameters and model selection using Akaike’s Information Criterion. J Wildl Manage. 2010;74: 1175–1178.
- 55. Cooch E, White G. Using MARK–a gentle introduction. 2016 Cornell University, Ithaca, NY, USA.
- 56. Hosmer DW, Lemeshow S, Sturdivant RX. Applied logistic regression. 2013. John Wiley & Sons, New York, NY, USA.
- 57. Pettorelli N, Pelletier F, von Hardenberg A, Festa-Bianchet M, Côté SD. Early onset of vegetation growth vs. rapid green-up: impacts on juvenile mountain ungulates. Ecology. 2007;88: 381–390. pmid:17479756
- 58. Hao F, Zhang X, Ouyang W, Skidmore AK, Toxopeus AG. Vegetation NDVI linked to temperature and precipitation in the Upper Catchments of Yellow River. Environ Model Assess. 2012;17: 389–398.
- 59. Landete-Castillejos T, García A, López-Serrano FR, Gallego L. Maternal quality and differences in milk production and composition for male and female Iberian red deer calves (Cervus elaphus hispanicus). Behav Ecol and Sociobiol. 2005;57: 267–274.
- 60. Therrien JF, Côté SD, Festa-Bianchet M, Ouellet JP. Maternal care in white-tailed deer: trade-off between maintenance and reproduction under food restriction. Anim Behav. 2008;75: 235–243.
- 61. Cook JG, Johnson BK, Cook RC, Riggs RA, Delcurto T, Bryant LD, et al. Effects of summer-autumn nutrition and parturition date on reproduction and survival of elk. Wildlife Monographs. 2004;155: 1–61.
- 62. McArt SH, Spalinger DE, Collins WB, Schoen ER, Stevenson T, Bucho M. Summer dietary nitrogen availability as a potential bottom-up constraint on moose in south-central Alaska. Ecology. 2009;90: 1400–1411. pmid:19537559
- 63. Parker KL, Barboza PS, Gillingham MP. Nutrition integrates environmental responses of ungulates. Funct Ecol. 2009;23: 57–69.
- 64. Hegel TM, Mysterud A, Ergon T, Loe LE, Huettmann F, Stenseth NC. Seasonal effects of Pacific-based climate on recruitment in a predator-limited large herbivore. J Anim Ecol. 2010;79: 471–482. pmid:20002863
- 65. Grovenburg TW, Monteith KL, Klaver RW, Jenks JA. Predator evasion by white-tailed deer fawns. Anim Behav. 2012;84: 59–65.
- 66. Gulsby WD, Kilgo JC, Vukovich M, Martin JA. Landscape heterogeneity reduces coyote predation on white-tailed deer fawns. J Wildl Manage. 2017;
- 67. Forchhammer MC, Clutton-Brock TH, Lindström J, Albon SD. Climate and population density induce long-term cohort variation in a northern ungulate. J Anim Ecol. 2001;70: 721–729.
- 68. Pettorelli N, Weladji RB, Holand Ø, Mysterud A, Breie H, Stenseth NC. The relative role of winter and spring conditions: linking climate and landscape-scale plant phenology to alpine reindeer body mass. Biol Lett. 2005;1: 24–26. pmid:17148119
- 69. Carstensen M, Delgiudice GD, Sampson BA, Kuehn DW. Survival, birth characteristics, and cause-specific mortality of white-tailed deer fawns. J Wildl Manage. 2009;73: 175–183.
- 70. Michel ES, Demarais S, Strickland BK, Belant JL. Contrasting the effects of maternal and behavioral characteristics on fawn birth mass in white-tailed deer. PLoS ONE. 2015;10: e0136034. pmid:26288141
- 71. Armstrong RA. Fetal development of the northern white-tailed deer (Odocoileus virginianus borealis). Am Midl Nat. 1950; 43: 650–666.
- 72. Pekins PJ, Smith KS, Mautz WW. The energy cost of gestation in white-tailed deer. Can J Zool. 1998;76: 1091–1097.
- 73. Edwards SL, Demarais S, Watkins B, Strickland BK. White-tailed deer forage production in managed and unmanaged pine stands and summer food plots in Mississippi. Wildl Soc Bull. 2004;32: 739–745.
- 74. Adams KP, Hamilton RJ. 2011. Management history in Biology and management of white-tailed deer (ed. Hewitt D. G.). CRC Press, Boca Raton, FL, USA.