I thought the new paper built very nicely on the old. Both echo a well known problem all ecologist face: systems really are complex and unpredictable. Even in laboratory systems, where virtually everything is controlled. We face the same problem in my lab and only combat it by a lot replication. This problem is becoming more critically to understand given the unrepresented pressures and challenges currently facing all natural systems. I did find it somewhat humorous that the new paper is from PNAS and its biggest result is basically "stochasticity is a thing. "
Park conducted an interesting experiment that examined the competition between 2 flour beetles – Trilobium confusum and T. castaneum. Maybe I missed something, but my major question was why did the author set 3 levels of media with the same population density (population/medium ratio)? Wouldn’t it be better if he could set different population densities to simulate different levels of environmental pressure? Also, the author spent appreciable pages on describing the results of control in which I was confused. Personally, I think it would be easier to follow if he could put the figures of control together with the corresponding figures of experimental treatments. I like the companion paper and I think it is relevant to the actual pest-control practice. The author brought up some general questions at the beginning, and I think the authors answered most of the questions to some extent, although they did not separate the effect of stochasticity and the effect of management approach. I am not familiar with pest/invasive species control, but I assume that “harvesting” is a general term describing some management combos, and the difficulties and success of “harvesting” vary among species. Thus, it would be exciting to see how this modeling method can be applied to some real systems with some known “harvesting” ways and whether the results are as predicted or not if the Pr(success) is high.
I agree with Yuguo that the Park paper was somewhat confusing, both in its setup and connection to theory, and in its communication of results and their significance. I really enjoyed the modern paper. It's really interesting to see applied problems receive a rigorous basic science controlled experimental approach. It's usually not feasible to do randomized controlled experiments in the field, as situations like lionfish invasion are typically treated as "triage" situations. I would be really interested in what would be necessary to expand our ethics and priorities to do true management experiments.
I thought it was interesting for Park's paper to state that "all interpretations of this paper, therefore, are to be regarded as exploratory and suggestive rather than as definitive" (p. 297). Park definitely had quite a lot of data over a long period of time, so his hesitancy in his conclusions is a bit surprising. The companion paper highlights one of the benefits of mathematical modeling. In the Park paper, he tracked the beetle populations for many years, but in the White et al paper, they were able to use mathematical models for the same species and generate similar results for much less time, energy and cost. One thing I did find a bit weird in the companion paper was the data representation in their figures. Throughout their graphs, the (0,0) points were rarely ever at the origin in the bottom left of the graph, so the graphs have these lines of blank areas next to the x- and y-axes. It is a small detail, but I think this form of representation can add a source of bias into the data interpretation.
The Park paper was interesting but as others have said, a little confusing and long-winded. The most helpful part of the paper was the Summary at the end. I do think their results were interesting, especially some of the differences between the "sterile" and "non-sterile" treatments. I enjoyed the modern paper and thought the approach was neat. Both papers did a nice job of highlighting how complex ecological systems are. Try as we may to understand how systems work, there will always be unpredictable outcomes.
The recent paper was a good pairing for the classic paper which also studied the dynamics of flour beetle populations. It raised important points on the complexity and stochasticity of systems, and the challenges this raises in the case of making applied ecological decisions in conservation & management. Their microcosm and model experiments paired nicely to show the significance of variability and randomness.
I found the detail in which Park outlined the life-history of both beetle species at the start of the manuscript to be quite useful, as this is important to understand for insect competition studies, especially given their small environment in this study. I wasn’t entirely sure what was meant by “number of replicates at start” but based on Table 2 it seems this was the number of individual populations of beetles (containers) for each treatment (super impressive!). I would have been curious to know more about how he determined how many replicates to run for each control/experimental treatment (as they varied quite a bit), but I’m guessing this was mostly dictated by funds and time. Also, it was interesting to see that sterilization increased the mean number of beetles per gram, although these results are difficult to translate to a “natural” flour environment that would have parasites present and interacting with competition in effects on body size, fecundity and survival.
I really appreciated the premise of the White et al. paper – while a lot of research aims to assess best management practices, there is considerably less study of what factors affect the efficacy and outcomes of those management decisions. I agree with Elizabeth in that their use of microcosms/paired models was very effective in showing how much stochasticity in the beetle populations ultimately shaped their results. However, as someone who has done research with management applications for controlling beetle populations, I was kind of sad to see their finding that the effects of their two management strategies were inconsistent and variable even in this controlled environment.
Both papers were very informative, but at times difficult for me to understand. I am looking forward to our discussion today in order to gain a better understanding of the methods employed in both papers. The Parks (1948) paper was clearly an intensive study with very interesting conclusions. I did not expect a parasite to play such a vital role in competition between two species. As for the other paper, it seemed to me as if they were more or less presenting a new method more than anything. I look forward to discussing the conclusions further.
I thought the new paper built very nicely on the old. Both echo a well known problem all ecologist face: systems really are complex and unpredictable. Even in laboratory systems, where virtually everything is controlled. We face the same problem in my lab and only combat it by a lot replication. This problem is becoming more critically to understand given the unrepresented pressures and challenges currently facing all natural systems.
ReplyDeleteI did find it somewhat humorous that the new paper is from PNAS and its biggest result is basically "stochasticity is a thing. "
-Miranda
Park conducted an interesting experiment that examined the competition between 2 flour beetles – Trilobium confusum and T. castaneum. Maybe I missed something, but my major question was why did the author set 3 levels of media with the same population density (population/medium ratio)? Wouldn’t it be better if he could set different population densities to simulate different levels of environmental pressure? Also, the author spent appreciable pages on describing the results of control in which I was confused. Personally, I think it would be easier to follow if he could put the figures of control together with the corresponding figures of experimental treatments.
ReplyDeleteI like the companion paper and I think it is relevant to the actual pest-control practice. The author brought up some general questions at the beginning, and I think the authors answered most of the questions to some extent, although they did not separate the effect of stochasticity and the effect of management approach. I am not familiar with pest/invasive species control, but I assume that “harvesting” is a general term describing some management combos, and the difficulties and success of “harvesting” vary among species. Thus, it would be exciting to see how this modeling method can be applied to some real systems with some known “harvesting” ways and whether the results are as predicted or not if the Pr(success) is high.
I agree with Yuguo that the Park paper was somewhat confusing, both in its setup and connection to theory, and in its communication of results and their significance. I really enjoyed the modern paper. It's really interesting to see applied problems receive a rigorous basic science controlled experimental approach. It's usually not feasible to do randomized controlled experiments in the field, as situations like lionfish invasion are typically treated as "triage" situations. I would be really interested in what would be necessary to expand our ethics and priorities to do true management experiments.
ReplyDeleteI thought it was interesting for Park's paper to state that "all interpretations of this paper, therefore, are to be regarded as exploratory and suggestive rather than as definitive" (p. 297). Park definitely had quite a lot of data over a long period of time, so his hesitancy in his conclusions is a bit surprising.
ReplyDeleteThe companion paper highlights one of the benefits of mathematical modeling. In the Park paper, he tracked the beetle populations for many years, but in the White et al paper, they were able to use mathematical models for the same species and generate similar results for much less time, energy and cost. One thing I did find a bit weird in the companion paper was the data representation in their figures. Throughout their graphs, the (0,0) points were rarely ever at the origin in the bottom left of the graph, so the graphs have these lines of blank areas next to the x- and y-axes. It is a small detail, but I think this form of representation can add a source of bias into the data interpretation.
The Park paper was interesting but as others have said, a little confusing and long-winded. The most helpful part of the paper was the Summary at the end. I do think their results were interesting, especially some of the differences between the "sterile" and "non-sterile" treatments. I enjoyed the modern paper and thought the approach was neat. Both papers did a nice job of highlighting how complex ecological systems are. Try as we may to understand how systems work, there will always be unpredictable outcomes.
ReplyDeleteThe recent paper was a good pairing for the classic paper which also studied the dynamics of flour beetle populations. It raised important points on the complexity and stochasticity of systems, and the challenges this raises in the case of making applied ecological decisions in conservation & management. Their microcosm and model experiments paired nicely to show the significance of variability and randomness.
ReplyDelete- Elizabeth
I found the detail in which Park outlined the life-history of both beetle species at the start of the manuscript to be quite useful, as this is important to understand for insect competition studies, especially given their small environment in this study. I wasn’t entirely sure what was meant by “number of replicates at start” but based on Table 2 it seems this was the number of individual populations of beetles (containers) for each treatment (super impressive!). I would have been curious to know more about how he determined how many replicates to run for each control/experimental treatment (as they varied quite a bit), but I’m guessing this was mostly dictated by funds and time. Also, it was interesting to see that sterilization increased the mean number of beetles per gram, although these results are difficult to translate to a “natural” flour environment that would have parasites present and interacting with competition in effects on body size, fecundity and survival.
ReplyDeleteI really appreciated the premise of the White et al. paper – while a lot of research aims to assess best management practices, there is considerably less study of what factors affect the efficacy and outcomes of those management decisions. I agree with Elizabeth in that their use of microcosms/paired models was very effective in showing how much stochasticity in the beetle populations ultimately shaped their results. However, as someone who has done research with management applications for controlling beetle populations, I was kind of sad to see their finding that the effects of their two management strategies were inconsistent and variable even in this controlled environment.
Both papers were very informative, but at times difficult for me to understand. I am looking forward to our discussion today in order to gain a better understanding of the methods employed in both papers. The Parks (1948) paper was clearly an intensive study with very interesting conclusions. I did not expect a parasite to play such a vital role in competition between two species. As for the other paper, it seemed to me as if they were more or less presenting a new method more than anything. I look forward to discussing the conclusions further.
ReplyDelete