Larger but younger fish when growth outpaces mortality in heated ecosystem
Curation statements for this article:-
Curated by eLife
eLife assessment
The work by Lindmark et al. provides us with an important natural experiment on fish that challenges current literature on relationships between temperature, growth rate, and size. The strength of their results is compelling, as Lindmark et al. mixed a unique warming setup with a large battery of models and statistics. The work will be of interest to ecologists and physiologists interested in the impacts of global warming on natural communities.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Ecology (eLife)
Abstract
Ectotherms are predicted to ‘shrink’ with global warming, in line with general growth models and the temperature-size rule (TSR), both predicting smaller adult sizes with warming. However, they also predict faster juvenile growth rates and thus larger size-at-age of young organisms. Hence, the result of warming on the size-structure of a population depends on the interplay between how mortality rate, juvenile- and adult growth rates are affected by warming. Here, we use two-decade long time series of biological samples from a unique enclosed bay heated by cooling water from a nearby nuclear power plant to become 5–10 °C warmer than its reference area. We used growth-increment biochronologies (12,658 reconstructed length-at-age estimates from 2426 individuals) to quantify how >20 years of warming has affected body growth, size-at-age, and catch to quantify mortality rates and population size- and age structure of Eurasian perch ( Perca fluviatilis ). In the heated area, growth rates were faster for all sizes, and hence size-at-age was larger for all ages, compared to the reference area. While mortality rates were also higher (lowering mean age by 0.4 years), the faster growth rates lead to a 2 cm larger mean size in the heated area. Differences in the size-spectrum exponent (describing how the abundance declines with size) were less clear statistically. Our analyses reveal that mortality, in addition to plastic growth and size-responses, is a key factor determining the size structure of populations exposed to warming. Understanding the mechanisms by which warming affects the size- and the age structure of populations is critical for predicting the impacts of climate change on ecological functions, interactions, and dynamics.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
The authors studied Eurasian perch in an experimental setup facilitated by a nuclear cooling plant to provide a natural laboratory. The heated area of the ecosystem raised in temperature by 8 degrees centigrade, while a reference area remained unheated. The authors provide a thorough and convincing description that the two areas are segregated such that individuals could not escape from one area to another prior to 2004, and such use data only until 2003 to test their hypotheses. The authors used both length-at-catch and age-increment data in a series of Bayesian mixed effects models to estimate the growth rate and length-at-age. They find that in the warmed area, both younger, smaller fish and older adults grew faster, contrary to the prediction of the temperature-size rule as well as …
Author Response
Reviewer #1 (Public Review):
The authors studied Eurasian perch in an experimental setup facilitated by a nuclear cooling plant to provide a natural laboratory. The heated area of the ecosystem raised in temperature by 8 degrees centigrade, while a reference area remained unheated. The authors provide a thorough and convincing description that the two areas are segregated such that individuals could not escape from one area to another prior to 2004, and such use data only until 2003 to test their hypotheses. The authors used both length-at-catch and age-increment data in a series of Bayesian mixed effects models to estimate the growth rate and length-at-age. They find that in the warmed area, both younger, smaller fish and older adults grew faster, contrary to the prediction of the temperature-size rule as well as many predictions and observations from other systems that fish reach smaller terminal body sizes in warmer environments due to increased metabolic demands. The authors furthermore combine the estimated body sizes with a mortality rate to determine the size-spectrum slope for both areas and determine the increased growth and increased mortality combine to essentially leave the size-spectrum slope observed in the ecosystem unchanged.
This is a thorough and interesting paper presented clearly and succinctly. These authors present a strong and thorough analysis of how temperature affects growth when all other ecosystem factors remain unchanged in a population. The dataset is a powerful one to support this type of analysis, and the statistical analysis methods the authors used appear to be robust and thorough. The diagnostics and visualizations are complete and inspire confidence in the convergence and accuracy of the modeling approach. The use of the size spectrum exponent to roll up individual-level changes across the population into a single metric was useful and interesting.
The estimates of the von Bertalanffy growth parameters in the results and discussion are less convincing than the growth increment and length-at-age estimates which seem much more robust. The presentation of estimates of the von Bertalanffy growth parameters in Figure S6 exhibit the high negative correlation between the k and L infinity parameters that are typical whenever multiple VBGF models are fit to subsets of data. It is difficult to determine which changes in parameters correspond to actual differences in early vs late life stage growth when, in any given year, if k is estimated low, L infinity will skew high simply due to the model structure. An example of this can be seen in 1995-1997 where L infinity is quite high but k is estimated quite low concurrently - in this case, it seems more reasonable to conclude the likelihood surface is quite flat between different parameter values than that fish suddenly reached a larger asymptotic size in these three years than all of the rest. The data in this case so strongly show larger growth in the heated area even without the VBGF results, and it would be more credible to base the discussion and results of this paper on the growth rate or observed length-at-age (e.g. Figure S4) estimates which are so clear.
We agree with the limitations of the von Bertalanffy growth equation (VBGE), and we agree with you and with Reviewer #2, that the estimated parameters for cohorts 1995–1997 are different, in particular for the L_infinity parameter in the heated area (see also reply to Reviewer#2 for a longer reply to that issue). The main reason for the size-at-age analysis in addition to growth-at-size is because the growth rates in theory could become similar between the areas for a given size, but if the initial growth rates were higher, there would still be a difference in the size-at-age, and size-at-age is an important trait in the context of the temperature-size rule (TSR). We could overcome the issues with the 3-parameter VBGE model by fitting multiple linear models to size-at-age for one age at the time. However, such models would not account for that cohorts may share similar growth trajectories. Therefore, we suggest instead to still use the VBGE growth equation, but put less emphasis on the specific parameter estimates, and instead present the results of the predictions of length-at-age only in that figure. We also wish to clarify that the size-at-age figure referred to here (Figure 2-figure supplement 4) is the predicted size-at-age from the VBGE model, rather than just the data or predictions from some other model.
In summary, we have downplayed the role of the specific parameter estimates and instead focused on the predicted size-at-age. Part of Figure 2 has been made a supporting figure (Figure 2-figure supplement 8). We have also conducted sensitivity analysis with respect to cohorts 1995–1997. This extra analysis shows that omitting these cohorts still results in a clear difference in size-at-age between the areas but reduces the predicted difference in size-at-age by a few percentage points. See first paragraph of the results, and lines 373–378. a
-
eLife assessment
The work by Lindmark et al. provides us with an important natural experiment on fish that challenges current literature on relationships between temperature, growth rate, and size. The strength of their results is compelling, as Lindmark et al. mixed a unique warming setup with a large battery of models and statistics. The work will be of interest to ecologists and physiologists interested in the impacts of global warming on natural communities.
-
Reviewer #1 (Public Review):
The authors studied Eurasian perch in an experimental setup facilitated by a nuclear cooling plant to provide a natural laboratory. The heated area of the ecosystem raised in temperature by 8 degrees centigrade, while a reference area remained unheated. The authors provide a thorough and convincing description that the two areas are segregated such that individuals could not escape from one area to another prior to 2004, and such use data only until 2003 to test their hypotheses. The authors used both length-at-catch and age-increment data in a series of Bayesian mixed effects models to estimate the growth rate and length-at-age. They find that in the warmed area, both younger, smaller fish and older adults grew faster, contrary to the prediction of the temperature-size rule as well as many predictions and …
Reviewer #1 (Public Review):
The authors studied Eurasian perch in an experimental setup facilitated by a nuclear cooling plant to provide a natural laboratory. The heated area of the ecosystem raised in temperature by 8 degrees centigrade, while a reference area remained unheated. The authors provide a thorough and convincing description that the two areas are segregated such that individuals could not escape from one area to another prior to 2004, and such use data only until 2003 to test their hypotheses. The authors used both length-at-catch and age-increment data in a series of Bayesian mixed effects models to estimate the growth rate and length-at-age. They find that in the warmed area, both younger, smaller fish and older adults grew faster, contrary to the prediction of the temperature-size rule as well as many predictions and observations from other systems that fish reach smaller terminal body sizes in warmer environments due to increased metabolic demands. The authors furthermore combine the estimated body sizes with a mortality rate to determine the size-spectrum slope for both areas and determine the increased growth and increased mortality combine to essentially leave the size-spectrum slope observed in the ecosystem unchanged.
This is a thorough and interesting paper presented clearly and succinctly. These authors present a strong and thorough analysis of how temperature affects growth when all other ecosystem factors remain unchanged in a population. The dataset is a powerful one to support this type of analysis, and the statistical analysis methods the authors used appear to be robust and thorough. The diagnostics and visualizations are complete and inspire confidence in the convergence and accuracy of the modeling approach. The use of the size spectrum exponent to roll up individual-level changes across the population into a single metric was useful and interesting.
The estimates of the von Bertalanffy growth parameters in the results and discussion are less convincing than the growth increment and length-at-age estimates which seem much more robust. The presentation of estimates of the von Bertalanffy growth parameters in Figure S6 exhibit the high negative correlation between the k and L infinity parameters that are typical whenever multiple VBGF models are fit to subsets of data. It is difficult to determine which changes in parameters correspond to actual differences in early vs late life stage growth when, in any given year, if k is estimated low, L infinity will skew high simply due to the model structure. An example of this can be seen in 1995-1997 where L infinity is quite high but k is estimated quite low concurrently - in this case, it seems more reasonable to conclude the likelihood surface is quite flat between different parameter values than that fish suddenly reached a larger asymptotic size in these three years than all of the rest. The data in this case so strongly show larger growth in the heated area even without the VBGF results, and it would be more credible to base the discussion and results of this paper on the growth rate or observed length-at-age (e.g. Figure S4) estimates which are so clear.
-
Reviewer #2 (Public Review):
With warming, fishes are generally expected to grow faster to smaller adult body sizes, as described by the temperature-size rule and other similar theories. However, the generality of this shrinking and the patterns among age classes within a species remain major research questions made all the more urgent by the rapid warming faced by many aquatic ecosystems. In this manuscript, the authors take advantage of an artificially heated ecosystem to investigate the impacts of warming on an unharvested population of fish and investigate patterns of growth, size structure, and mortality in an unexploited fish population. Surprisingly, while faster growth rates in juveniles are demonstrated, as would be expected, adult size remains higher in the heated habitat compared to a nearby non-heated habitat. This …
Reviewer #2 (Public Review):
With warming, fishes are generally expected to grow faster to smaller adult body sizes, as described by the temperature-size rule and other similar theories. However, the generality of this shrinking and the patterns among age classes within a species remain major research questions made all the more urgent by the rapid warming faced by many aquatic ecosystems. In this manuscript, the authors take advantage of an artificially heated ecosystem to investigate the impacts of warming on an unharvested population of fish and investigate patterns of growth, size structure, and mortality in an unexploited fish population. Surprisingly, while faster growth rates in juveniles are demonstrated, as would be expected, adult size remains higher in the heated habitat compared to a nearby non-heated habitat. This unexpected result will be of broad interest.
Strengths
The semi-natural experiment provided by the artificial warming from the power plant is a very nice design. While it is not the only place this type of study could be conducted, this system seems to have an unusually high degree of heating, that fact and the unexpected results make for a very interesting study that should be of broad interest. The study is also presented in a clear and concise manuscript and the conclusions are well-supported.
Weaknesses
In certain sections, it seems like the paper would benefit from a more thorough consideration of alternative explanations for the higher body size in the warmed population, like the release from density dependence or altered prey availability, and how those alternative explanations do or do not fit with the result that mortality was higher for the heated population. The consideration of mortality is a strength of the paper, but this result and how it fits with the result that heated adults did not shrink could be discussed in greater depth. It is unfortunate that factors other than the heat that might influence mortality, like predation rates, remain unknown in this system, but then they are rarely well understood in real-world settings like whole ecosystems.
-
-
-