Limits...
Biological and environmental influences on parturition date and birth mass of a seasonal breeder.

Wolcott DM, Reitz RL, Weckerly FW - PLoS ONE (2015)

Bottom Line: Random effects revealed considerable variation among mothers and years.This study demonstrates that, in long-lived polytocous species, environmental factors may have a greater influence on natal features than previously supposed and the influence from biological factors is also complex.The documented responses to environmental influences provide unique insights into how mammalian seasonal reproductive dynamics may respond to current changes in climate.

View Article: PubMed Central - PubMed

Affiliation: Texas State University, Department of Biology, San Marcos, Texas, United States of America.

ABSTRACT
Natal features (e.g. Julian birth date and birth mass) often have fitness consequences and can be influenced by endogenous responses by the mother to seasonal fluctuations in nutritional quality and photoperiodic cues. We sought to further understand the biological and environmental factors that influence the natal features of a polytocous species in an environment with constant nutritional resources and limited seasonal variation. During a 36-year study we assessed the influence of biological factors (maternal age and litter type [i.e., litter size and sexual composition]) and environmental factors (total precipitation and mean maximum temperature during months encompassing conception, the last trimester of gestation, and the entire length of gestation) on Julian birth date and birth mass using linear-mixed effects models. Linear and quadratic functions of maternal age influenced both natal features with earliest Julian birth dates and heaviest birth masses occurring at prime-age and older individuals, which ranged from 5-9 years of age. Litter type influenced Julian birth date and birth mass. Interestingly, environmental factors affected Julian birth date and birth mass even though mothers were continuously allowed access to a high-quality diet. Random effects revealed considerable variation among mothers and years. This study demonstrates that, in long-lived polytocous species, environmental factors may have a greater influence on natal features than previously supposed and the influence from biological factors is also complex. The documented responses to environmental influences provide unique insights into how mammalian seasonal reproductive dynamics may respond to current changes in climate.

No MeSH data available.


Related in: MedlinePlus

Predicted values from a linear mixed effect model estimating the parturition date (in Julian days) of captive white-tailed deer at Kerr Wildlife Management Area, Kerr County, Texas, USA from 1977–2012.Regression coefficients were obtained from model-averaged parameter estimates of competing models. Predicted Julian birth date was estimated across the range of each variable deemed important while controlling for all other variables (variable constants included: Maternal age = 4, Litter type = female singleton, Study program = study program 1, April–June precipitation = 256.2 mm, October–Jun precipitation = 611.7 mm, June precipitation = 102.6 mm, and November temperature = 20.4°C). The solid lines represent the predicted estimate for Julian birth date and the dashed lines are the standard error envelopes for the estimates. Random effects were treated as categorical variables and included a unique identifier for each mother and the year of birth for each fawn.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4401666&req=5

pone.0124431.g003: Predicted values from a linear mixed effect model estimating the parturition date (in Julian days) of captive white-tailed deer at Kerr Wildlife Management Area, Kerr County, Texas, USA from 1977–2012.Regression coefficients were obtained from model-averaged parameter estimates of competing models. Predicted Julian birth date was estimated across the range of each variable deemed important while controlling for all other variables (variable constants included: Maternal age = 4, Litter type = female singleton, Study program = study program 1, April–June precipitation = 256.2 mm, October–Jun precipitation = 611.7 mm, June precipitation = 102.6 mm, and November temperature = 20.4°C). The solid lines represent the predicted estimate for Julian birth date and the dashed lines are the standard error envelopes for the estimates. Random effects were treated as categorical variables and included a unique identifier for each mother and the year of birth for each fawn.

Mentions: All biological factors tested (maternal age and litter type) influenced Julian birth date, as well as the nuisance variable (study program, Table 1). Inclusion of environmental factors, in the model-selection analysis, indicated that several models fit the data equally well. Five models were within two ΔAICc of each other (Table 2). Model averaged estimates of the five models indicated that maternal age had a negative relationship on Julian birth date, with every 1 year increase in maternal age decreasing Julian birth date by 4.1 days (CI = –5.7 to—2.5, Table 3). The quadratic term for maternal age increased Julian birth date as maternal age increased, with the earliest predicted birth dates occurring at 9 years of age (Fig 3). The post-hoc analysis assessing the possibility of a threshold or senescent effect after prime age demonstrated that linear and quadratic terms were not significant (F1,59 = 0.434, P = 0.513 and F1,59 = 0.398, P = 0.531, respectively), thus, there was a threshold effect. On average, mothers in the youngest age class (2 years of age) gave birth on the latest dates (171, 20 June) and the oldest mothers (13 years of age) gave birth to individuals around the same time as a 6-year old mother (160, 9 June). Predicted values of Julian birth date for each litter type indicate that mixed-sex litters were born earlier than all other litter types (Julian birth date = 161.9, SE = 1.5). Julian birth dates for all other litter types (F = 165.5, SE = 1.7; FF = 166.1, SE = 1.7; M = 165.6, SE = 1.6; and MM = 164.8, SE = 1.6) were not significantly different from each other (Fig 3). While the summary from the model-averaged regression included parameter estimates of environmental influences for November temperature, June precipitation, and October–June precipitation, the only 95% CI that did not overlap 0 was April–June precipitation. Julian birth date was influenced by April–June precipitation with every 1 mm increase in precipitation decreasing Julian birth date by 0.02 days. April–June precipitation was highly variable during the study (min = 85.2 mm, max = 543.6 mm, average = 253.9 mm, CV = 0.51). Variance components for the random effects in the Julian birth date analysis could not be derived from the model-averaged analysis, thus, we reported values derived from the model with the lowest AICc (M10, biological factors and April–June precipitation as main effects). Variance components for this model were dam id (SD = 11.26), birth year (SD = 4.85), and residual error (SD = 15.22). The marginal R2 for this model was 0.09 and the conditional R2 was 0.45. Inclusion of the random effects (dam id and birth year) and fixed effects explained more variation in birth mass than fixed effects alone.


Biological and environmental influences on parturition date and birth mass of a seasonal breeder.

Wolcott DM, Reitz RL, Weckerly FW - PLoS ONE (2015)

Predicted values from a linear mixed effect model estimating the parturition date (in Julian days) of captive white-tailed deer at Kerr Wildlife Management Area, Kerr County, Texas, USA from 1977–2012.Regression coefficients were obtained from model-averaged parameter estimates of competing models. Predicted Julian birth date was estimated across the range of each variable deemed important while controlling for all other variables (variable constants included: Maternal age = 4, Litter type = female singleton, Study program = study program 1, April–June precipitation = 256.2 mm, October–Jun precipitation = 611.7 mm, June precipitation = 102.6 mm, and November temperature = 20.4°C). The solid lines represent the predicted estimate for Julian birth date and the dashed lines are the standard error envelopes for the estimates. Random effects were treated as categorical variables and included a unique identifier for each mother and the year of birth for each fawn.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4401666&req=5

pone.0124431.g003: Predicted values from a linear mixed effect model estimating the parturition date (in Julian days) of captive white-tailed deer at Kerr Wildlife Management Area, Kerr County, Texas, USA from 1977–2012.Regression coefficients were obtained from model-averaged parameter estimates of competing models. Predicted Julian birth date was estimated across the range of each variable deemed important while controlling for all other variables (variable constants included: Maternal age = 4, Litter type = female singleton, Study program = study program 1, April–June precipitation = 256.2 mm, October–Jun precipitation = 611.7 mm, June precipitation = 102.6 mm, and November temperature = 20.4°C). The solid lines represent the predicted estimate for Julian birth date and the dashed lines are the standard error envelopes for the estimates. Random effects were treated as categorical variables and included a unique identifier for each mother and the year of birth for each fawn.
Mentions: All biological factors tested (maternal age and litter type) influenced Julian birth date, as well as the nuisance variable (study program, Table 1). Inclusion of environmental factors, in the model-selection analysis, indicated that several models fit the data equally well. Five models were within two ΔAICc of each other (Table 2). Model averaged estimates of the five models indicated that maternal age had a negative relationship on Julian birth date, with every 1 year increase in maternal age decreasing Julian birth date by 4.1 days (CI = –5.7 to—2.5, Table 3). The quadratic term for maternal age increased Julian birth date as maternal age increased, with the earliest predicted birth dates occurring at 9 years of age (Fig 3). The post-hoc analysis assessing the possibility of a threshold or senescent effect after prime age demonstrated that linear and quadratic terms were not significant (F1,59 = 0.434, P = 0.513 and F1,59 = 0.398, P = 0.531, respectively), thus, there was a threshold effect. On average, mothers in the youngest age class (2 years of age) gave birth on the latest dates (171, 20 June) and the oldest mothers (13 years of age) gave birth to individuals around the same time as a 6-year old mother (160, 9 June). Predicted values of Julian birth date for each litter type indicate that mixed-sex litters were born earlier than all other litter types (Julian birth date = 161.9, SE = 1.5). Julian birth dates for all other litter types (F = 165.5, SE = 1.7; FF = 166.1, SE = 1.7; M = 165.6, SE = 1.6; and MM = 164.8, SE = 1.6) were not significantly different from each other (Fig 3). While the summary from the model-averaged regression included parameter estimates of environmental influences for November temperature, June precipitation, and October–June precipitation, the only 95% CI that did not overlap 0 was April–June precipitation. Julian birth date was influenced by April–June precipitation with every 1 mm increase in precipitation decreasing Julian birth date by 0.02 days. April–June precipitation was highly variable during the study (min = 85.2 mm, max = 543.6 mm, average = 253.9 mm, CV = 0.51). Variance components for the random effects in the Julian birth date analysis could not be derived from the model-averaged analysis, thus, we reported values derived from the model with the lowest AICc (M10, biological factors and April–June precipitation as main effects). Variance components for this model were dam id (SD = 11.26), birth year (SD = 4.85), and residual error (SD = 15.22). The marginal R2 for this model was 0.09 and the conditional R2 was 0.45. Inclusion of the random effects (dam id and birth year) and fixed effects explained more variation in birth mass than fixed effects alone.

Bottom Line: Random effects revealed considerable variation among mothers and years.This study demonstrates that, in long-lived polytocous species, environmental factors may have a greater influence on natal features than previously supposed and the influence from biological factors is also complex.The documented responses to environmental influences provide unique insights into how mammalian seasonal reproductive dynamics may respond to current changes in climate.

View Article: PubMed Central - PubMed

Affiliation: Texas State University, Department of Biology, San Marcos, Texas, United States of America.

ABSTRACT
Natal features (e.g. Julian birth date and birth mass) often have fitness consequences and can be influenced by endogenous responses by the mother to seasonal fluctuations in nutritional quality and photoperiodic cues. We sought to further understand the biological and environmental factors that influence the natal features of a polytocous species in an environment with constant nutritional resources and limited seasonal variation. During a 36-year study we assessed the influence of biological factors (maternal age and litter type [i.e., litter size and sexual composition]) and environmental factors (total precipitation and mean maximum temperature during months encompassing conception, the last trimester of gestation, and the entire length of gestation) on Julian birth date and birth mass using linear-mixed effects models. Linear and quadratic functions of maternal age influenced both natal features with earliest Julian birth dates and heaviest birth masses occurring at prime-age and older individuals, which ranged from 5-9 years of age. Litter type influenced Julian birth date and birth mass. Interestingly, environmental factors affected Julian birth date and birth mass even though mothers were continuously allowed access to a high-quality diet. Random effects revealed considerable variation among mothers and years. This study demonstrates that, in long-lived polytocous species, environmental factors may have a greater influence on natal features than previously supposed and the influence from biological factors is also complex. The documented responses to environmental influences provide unique insights into how mammalian seasonal reproductive dynamics may respond to current changes in climate.

No MeSH data available.


Related in: MedlinePlus