Limits...
Early warning signals detect critical impacts of experimental warming

View Article: PubMed Central - PubMed

ABSTRACT

Earth's surface temperatures are projected to increase by ~1–4°C over the next century, threatening the future of global biodiversity and ecosystem stability. While this has fueled major progress in the field of physiological trait responses to warming, it is currently unclear whether routine population monitoring data can be used to predict temperature‐induced population collapse. Here, we integrate trait performance theory with that of critical tipping points to test whether early warning signals can be reliably used to anticipate thermally induced extinction events. We find that a model parameterized by experimental growth rates exhibits critical slowing down in the vicinity of an experimentally tested critical threshold, suggesting that dynamical early warning signals may be useful in detecting the potentially precipitous onset of population collapse due to global climate change.

No MeSH data available.


Related in: MedlinePlus

Linear regressions for estimating r and K from experimental data. Here, we show the linear models fit to exponential rate of increase (rt) and algal cell density (Nt) data for each temperature treatment (15°C (A), 20°C (B), 25°C (C), 30°C (D), 35°C (E)). It should be noted that the r and K for 30°C and 35°C treatments cannot be accurately estimated from these regressions because it can be argued that these temperatures are beyond the thermal tolerance of the algae populations, in which case population growth rate is no longer limited by density, but rather physiological constraint. Only one of three regressions is displayed for each treatment temperature; however, the mean of three replicate r and K estimates were used to construct corresponding thermal performance curves. In the above plots, r is the y‐intercept and K is the x‐intercept.
© Copyright Policy - creativeCommonsBy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5016634&req=5

ece32339-fig-0003: Linear regressions for estimating r and K from experimental data. Here, we show the linear models fit to exponential rate of increase (rt) and algal cell density (Nt) data for each temperature treatment (15°C (A), 20°C (B), 25°C (C), 30°C (D), 35°C (E)). It should be noted that the r and K for 30°C and 35°C treatments cannot be accurately estimated from these regressions because it can be argued that these temperatures are beyond the thermal tolerance of the algae populations, in which case population growth rate is no longer limited by density, but rather physiological constraint. Only one of three regressions is displayed for each treatment temperature; however, the mean of three replicate r and K estimates were used to construct corresponding thermal performance curves. In the above plots, r is the y‐intercept and K is the x‐intercept.

Mentions: Intrinsic rates of population increase (or instantaneous per capita growth rates), rt, were calculated using the exponential growth rate equation, rt=logeNt+1/Nt/τ, where τ is the number of days between consecutive density measurements (Fryxell et al. 2014). The exponential growth rate equation is convenient when the time between measurements is irregular, as was the case with our population data (Fryxell et al. 2014). rt for each time series (each replicate) were estimated between consecutive measures of population density at a given time (Nt). We then plotted rt for each time interval against the corresponding Nt (density at the beginning of said interval) and used a linear model to describe the relationship of population growth with increasing abundance, assuming normal distribution (Fig. 3). It is assumed that maximum population growth (denoted here as r) occurs at densities close to zero; thus, we estimate r to be rt when Nt = 0. Further, the carrying capacity (K) exists at the maximum density that can be sustained by limiting resources, such as nutrients (the x‐intercept of the regression line for rt vs. Nt; i.e., Nt when rt = 0).


Early warning signals detect critical impacts of experimental warming
Linear regressions for estimating r and K from experimental data. Here, we show the linear models fit to exponential rate of increase (rt) and algal cell density (Nt) data for each temperature treatment (15°C (A), 20°C (B), 25°C (C), 30°C (D), 35°C (E)). It should be noted that the r and K for 30°C and 35°C treatments cannot be accurately estimated from these regressions because it can be argued that these temperatures are beyond the thermal tolerance of the algae populations, in which case population growth rate is no longer limited by density, but rather physiological constraint. Only one of three regressions is displayed for each treatment temperature; however, the mean of three replicate r and K estimates were used to construct corresponding thermal performance curves. In the above plots, r is the y‐intercept and K is the x‐intercept.
© Copyright Policy - creativeCommonsBy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5016634&req=5

ece32339-fig-0003: Linear regressions for estimating r and K from experimental data. Here, we show the linear models fit to exponential rate of increase (rt) and algal cell density (Nt) data for each temperature treatment (15°C (A), 20°C (B), 25°C (C), 30°C (D), 35°C (E)). It should be noted that the r and K for 30°C and 35°C treatments cannot be accurately estimated from these regressions because it can be argued that these temperatures are beyond the thermal tolerance of the algae populations, in which case population growth rate is no longer limited by density, but rather physiological constraint. Only one of three regressions is displayed for each treatment temperature; however, the mean of three replicate r and K estimates were used to construct corresponding thermal performance curves. In the above plots, r is the y‐intercept and K is the x‐intercept.
Mentions: Intrinsic rates of population increase (or instantaneous per capita growth rates), rt, were calculated using the exponential growth rate equation, rt=logeNt+1/Nt/τ, where τ is the number of days between consecutive density measurements (Fryxell et al. 2014). The exponential growth rate equation is convenient when the time between measurements is irregular, as was the case with our population data (Fryxell et al. 2014). rt for each time series (each replicate) were estimated between consecutive measures of population density at a given time (Nt). We then plotted rt for each time interval against the corresponding Nt (density at the beginning of said interval) and used a linear model to describe the relationship of population growth with increasing abundance, assuming normal distribution (Fig. 3). It is assumed that maximum population growth (denoted here as r) occurs at densities close to zero; thus, we estimate r to be rt when Nt = 0. Further, the carrying capacity (K) exists at the maximum density that can be sustained by limiting resources, such as nutrients (the x‐intercept of the regression line for rt vs. Nt; i.e., Nt when rt = 0).

View Article: PubMed Central - PubMed

ABSTRACT

Earth's surface temperatures are projected to increase by ~1–4°C over the next century, threatening the future of global biodiversity and ecosystem stability. While this has fueled major progress in the field of physiological trait responses to warming, it is currently unclear whether routine population monitoring data can be used to predict temperature‐induced population collapse. Here, we integrate trait performance theory with that of critical tipping points to test whether early warning signals can be reliably used to anticipate thermally induced extinction events. We find that a model parameterized by experimental growth rates exhibits critical slowing down in the vicinity of an experimentally tested critical threshold, suggesting that dynamical early warning signals may be useful in detecting the potentially precipitous onset of population collapse due to global climate change.

No MeSH data available.


Related in: MedlinePlus