Kettlewell, 1955; Nosil et al., 2018
Papers:
Kettlewell HBD. 1955. Selection experiments on industrial melanism in the Lepidoptera. Heredity 9: 323–342.
Nosil P, Villoutreix R, de Carvalho CF, Farkas TE, Soria-Carrasco V, Feder JL, Crespi BJ, Gompert Z. 2018. Natural selection and the predictability of evolution in Timemastick insects. Science 359: 765–770.
Blogpost: Yuguo Yang
Author background:
Henry Bernard Davis Kettlewell was a British geneticist who worked on Lepidoptera (moth and butterfly). He was a researcher in Department of Zoology in Oxford University who conducted the classic study about natural selection – the influence of industrial melanism on peppered moth coloration.
Patrik Nosil is a independent research fellow in University of Sheffield, UK. His research is mainly about evolutionary processes that drives speciation.
Kettlewell 1955
The peppered moth (Biston betularia) has 3 phenotypes: 1.typical, which is white color with black dots. 2.carbonaria, which is a melanistic type with completely black color. 3. insularia, which is a less complete melanism – dark color with white scales. The purpose of this study was to figure out whether the different colored moths can take advantages (i.e. “cryptic”) from the background color and whether these moths were eaten by the predators selectively due to their visibility.
To study the conspicuousness of moth on there background, different moths were judged by walking away from them until the moths became invisible. The results showed that the crypsis of moths from human eyes were highly dependent on whether the moths’ color matched the background. To test whether the moths were taken by predators differently due to the camouflage, an aviary experiment was conducted using 2 great tits (Parus major) in a cage. As a result, moths on incorrect background had higher chance to be eaten.
Then the author conducted a release-recapture study in an industrially smoked area where the proportion of birch (all moths protected) to oak (dark moths protected) was less than 1:10. 27.5% carbonaria, 13% typical, and 17.4% insularia males were recaptured after releasing. In addition, local population had 85.03% carbonaria, 10.14% typical, and 4.83 insularia. From the author’s observation, although the detection of conspicuous moth would make other moths nearby at a disadvantage due to the active searching of predators, there still were a evidence that predators had biased predation due to crypsis.
Nosil et al. 2018
The aim of this paper was to study the predictability of evolution in walking stick insects (Timema cristinae). T. cristinae has 3 morphs: 1. striped green that is cryptic on the leaves of Adenostoma fasciculatum, 2. unstriped green that is cryptic on the leaves of Ceanothus spinosus, 3. melanistic form that is cryptic on the stems of both plants above. The morphs are controlled by 3 alleles where green is dominant to melanistic, and unstriped is dominant to striped.
The authors used sequences from 3 datasets to quantify the changes in allele frequency overtime. The result showed that Mel-stripe (the loci controlling body color and strip) had the highest temporal allele frequency change in all the datasets, which was far beyond what can be explained by dispersal. The authors thus attributed this change to selection.
Predictability of evolution was quantified using autoregressive moving average models (ARMA). The model analysis indicated that the temporal change in color-morph frequency (green vs. melanistic) can be only moderately predicted because complex and contradictory factors were driving it, whereas the temporal change in pattern-morph frequency were highly predictable possibly due to negative frequency-dependent selection (NFDS). Then the authors conducted a transplant experiment to verify the NFDS. 1000 insects (green striped and green unstriped) were transplanted into 20 A. fasciculatumwith either 4:1 or 1:4 ratio. Partly congruent with the hypothesis, the striped morph had strong selection preference when the initial proportion was low.
My thoughts
The kettlewell’s moth paper is one of the most famous study in intro-biology textbook. I like the flow of the paper as well as the step-by-step experimental design. The aviary experiment after human visual experiment is reasonable but the sample size was small, and I think it would be better if more than one bird species are tested. Overall, this paper provided solid results of evolution driven by selection. I especially like how the author justified the conclusion by refuting the potential alternative explanations such as biased recapture, different lifespan, migration, etc. The companion paper studied another insect species that had similar types of morph as peppered moth. This paper is highly compacted and I am not familiar with the model in the paper, but I find it interesting that temporal change of color morph is driven by counteracting sources. It will be great if a field study can be conducted to address the relative effect of those selection sources.
Another good week for papers. This week, we read the famous Kettlewell paper (1955) which is commonly used in intro level biology courses (and evolution courses). I have heard a lot about this study (so so so many times have I heard about this study, I thought I could write a peppered moth paper) but this was my first time reading the paper. I found the paper quite good, especially since Kettlewell seemed to respond to concerns about the study (i.e. the moths not freely picking where they wanted to rest on the trees). The paper had a lot more information and experimentation than is often taught in general biology courses (at least the ones I have been in). I was impressed with all the information about the birds (at one point, Kettlewell even said they had learned about the moth-depositing habits of the humans so they had to change the study to include other insects). In some cases, I wondered why they released such unequal proportions of each moth morph (in some cases, many more of one and less than 10 of another), but I suspect it had something to do with the breeding of the moths. All in all, it was a valuable read.
ReplyDeleteThe companion paper was a nice follow-up, not only because it covers another insect species with different phenotypic morphs, but also because it incorporates a broad question: how can we predict evolutionary change? My one gripe, which may be influenced from my own biases, is how they claim to see a predictable patter in a 25 year span (ignoring randomness). Is a consecutive pattern of 25 years enough to observe and predict an evolutionary pattern driven by natural selection? In saying this, the changes in the moths (from light to dark morph) occurred in a short amount of time (around 50 years I believe), so maybe I am over-zealous in my questioning. This, however, did arise when environmental patterns changed rather dramatically. I also may not fully understand their methods, so I will say my hesitation may not be completely valid, but I still wonder about this study.
I really enjoyed the peppered moth paper, both for its rigorous experimental treatment of evolution and for its writing style. I kind of felt like I was reading a noir detective novel about moth evolution. It's really fascinating to get a historical example of a natural experiment, and makes me wonder where contemporary natural experiments might be hiding.
ReplyDeleteThe companion paper was a fascinating contrast because it gets at some of the complexities of evolution not addressed in the classic peppered moth study. I really appreciated their analytical treatment of non-linear selection patterns and their robust efforts to link genotypic and phenotypic patterns.
Before reading the Kettlewell paper, I did not realize that there were actually three morphs of the peppered moths. In the high school lesson that I used, the information on peppered moths only referenced two morphs: black (what Kettlewell calls carbonaria) and white (Kettlewell’s typical). I did not realize there was a third “insularia” morph. I can understand why this third morph was left out of introductory curriculum because its removal simplifies the concept and makes it easier to model. I liked to use the peppered moth simulation on this website: http://peppermoths.weebly.com/. The simulation helps to demonstrate the cryptic coloration and how difficult it is for people (and birds) to see the moths on the “correct” backgrounds.
ReplyDeleteI’ll admit that I got a bit overwhelmed with the amount of information presented in the Nosil paper. There were so many graphs! I guess the amount of information is useful for supporting Nosil’s claim that “selection is multifaceted”. I don’t think I quite understood what was going on in Figure 6. What the authors exactly meant by “predicting evolution” was a bit fuzzy and it seemed odd to me to be comparing “evolution” between different species.
I was super excited to read the moth paper. That system in one of the reasons I became interested in biology. I would have liked to read the 1937 one as well thought. Overall I feel it was a sound paper for the time. Although a larger sample size and more diverse predators would have made it stronger. I would have also really liked to see a drawing or picture of the different moth morphs.
ReplyDeleteAS for the newer paper, I'm not sure I quite understand "predicting evolution." It makes sense in a limited context, such as a change in food source causes a change in beak size. But I still don't really understand how of why one would try in a complex system full of different selective pressures.
-Miranda
Kettlewell was an engaging paper to read and it was neat seeing how so many lines of evidence were used to answer the question of whether predation drives selection on moth color. Since birch bark can offer cover for both carboneria and typical moths because birch bark has both light and dark patches, does it follow that there may have been higher proportions of carboneria in birch dominated forests prior to industrialization? This could be investigated today, since there is less soot pollution.
ReplyDeleteI was really impressed with how well Nosil et al. predicted pattern evolution. However, I think the main reason that pattern evolution is more predictable than color change is because the time series is more informative for changes in pattern (oscillating) vs. changes in color (increasing). From my basic understanding of the autoregressive models Nosil et al. used, it is only appropriate to use them when the time series is stationary. So, it’s possible that the poor predictability of color change over time is because the model used by Nosil et al. wasn’t appropriate for that type of data. This makes me wonder what the timeseries from the other evolutionary case studies Nosil et al. referenced look like and what types of models they used.
An interesting development in ecological forecasting has been the development of metrics to determine the intrinsic predictability of systems. The idea here, is that we shouldn’t assess our forecast performance solely on r or r^2, but on how large the “realized” predictive ability of the models are relative to the “intrinsic” predictability of the system. I think that this would be a useful approach to evolutionary forecasting as well. Were evolutionary predictions in those other systems poor because of a lack of data or poor models or because the systems have a low intrinsic predictability?
- David
I enjoyed reading the Kettlewell paper because like most of you, I have seen the moth morphs as a textbook example of evolution. I liked that there were paired field and control studies but wished he had used more replicates for the aviary study, especially because the tits learned to look for the moths so quickly. I also wondered what other species of insects were released in the next trial and how they compared to the moths in terms of palatability and energy/nutrient content. Would the observed bird species normally target these moths as preferred prey or were they only hunting them opportunistically? I think Kettlewell should have addressed predator preferences as well to strengthen his conclusion that birds drove this evolutionary change. I also wondered what the mechanism was for the moths to be able to accurately assess whether they matched their background. This last question might be a bit out of scope for the paper, but color matching did seem to be an important assumption and also I think it's just a cool question.
ReplyDeleteThe predictability of evolution seems inherently difficult (as Nosil et al. are quick to point out) because individuals can be exposed to multiple selective pressures at once. We can't necessarily predict that evolution is the response to selective pressures because animals may disperse or go extinct as well. Also, how do we account for phenotypic plasticity in these models? Plastic responses can also be adaptive, but do not necessarily lead to a change in genotypic frequencies. As the authors conclude, it's complicated..
Coming from an entomology (sort of) background, I was surprised to have never come across the Kettlewell paper before! I had never thought about the idea of “industrial melanism”, and was fascinated by everything from the experimental design, to the author’s enthusiasm about the topic, to the discussion of the particular role that phenotypic variation in color plays for this species.
ReplyDeleteThe Nosil et al. paper was a nice companion, and I liked their efforts to show the comparisons of species-specific evolution predictions in Figure 6 (including the Biston betularia morph!). However, as other folks have mentioned, it is difficult to interpret the biological meaning of these predictions, and whether we can actually meaningfully compare them between species. As David mentioned, the variables used to test predictability (color and pattern) are amenable to the time-series data they had available – would their results be similar if they had tested, for example, trends in body size or fecundity?
Like others, I enjoyed reading the Kettlewell paper because I've heard about it in classes but never learned how in-depth the study actually was. It is nice to see that such a classic example of adaptation is not solely based on one-off observations or speculation like some of the previous papers have been. I also like this paper because we tend to think of adaptation as something that happens quite slowly, whereas in this system all the alleles are already present so changes in color can happen quite quickly. I do remember reading somewhere of a follow up study the moths populations reverting back to the pre-pollution phenotype now that pollution has lessened.
ReplyDeleteI enjoyed the newer paper but agree with others and the authors that predicting evolution is harder than it might seem!
Like many others, I have heard a ton about the peppered moth experiments in my own evolutionary biology courses, but have not gotten a chance to actually read the paper until now. I appreciated how thorough and in depth the paper was, and how well he explained his methods and conclusions. They did only use great tits in their aviary experiment and I would be interested in seeing how other endemic bird species would respond to the same set up.
ReplyDeleteThe new paper was a good follow up to the classic and was an interesting read that was parallel to Kettlewell's. I do agree that predicting evolution is a lofty goal and a bit more difficult that it may seem.
- Elizabeth