Response to comment on 'Parasite defensive limb movements enhance acoustic signal attraction in male little torrent frogs'
- CAS Key Laboratory of Mountain Ecological Restoration and Bioresource Utilization & Ecological Restoration and Biodiversity Conservation Key Laboratory of Sichuan Province, Chengdu Institute of Biology, Chinese Academy of Sciences, China
- Ministry of Education Key Laboratory for Ecology of Tropical Islands, Key Laboratory of Tropical Animal and Plant Ecology of Hainan Province, College of Life Sciences, Hainan Normal University, China
- Department of Ecological Sciences, Vrije Universiteit Amsterdam, Netherlands
Abstract
Recently we showed that limb movements associated with anti-parasite defenses can enhance acoustic signal attraction in male little torrent frogs (Amolops torrentis), which suggests a potential pathway for physical movements to become co-opted into mating displays (Zhao et al., 2022). Anderson et al. argue for alternative explanations of our results and provide a reanalysis of part of our data (Anderson et al., 2023). We acknowledge some of the points raised and provide an additional analysis in support of our hypothesis.
Introduction
Physical movements have been viewed as the raw material of visual signals for many years (Harper, 1991; Johnsgard, 1962). However, little is known about how physical displays evolve and become a part of communicative systems. According to the sensory exploitation hypothesis, physical movements may be incorporated into multimodal systems if they enhance the attractiveness of male individuals (Ryan, 1998). Recently we showed that movements associated with anti-parasite behavior can act on female perception and may thereby affect the evolution of a multimodal display in little torrent frogs (Amolops torrentis) (Zhao et al., 2022). Anderson et al. provide some alternative interpretations for our findings, some of which we found useful for discussion and future experiments (Anderson et al., 2023). Their most important argument concerns our finding that females showed a strong preference for limb movements that were not strongly associated with the presence of eavesdropping parasites. Below, we respond and provide an additional analysis based on some of the suggestions made by Anderson et al.
Results and discussion
Underestimations of parasite-associated limb movements
Anderson et al. suggest that the two movements most strongly preferred in playback experiments by females – hind foot lifting (HFL) and arm waving (AW) – are not associated with the presence of eavesdropping parasites, whereas two other movements – wiping (W) and limb shaking (LSA) – do increase as parasite numbers increase, but are less preferred (or not preferred) by females. They base this suggestion on a reanalysis of our original data, but we argue that their approach is problematic for two reasons.
First, their reanalysis used samples that differed from those used in our original paper. Most importantly, they included observations from non-calling males (which had been recorded outside of a breeding context for our control videos), whereas we analysed data from actively displaying males only. Many silent males were observed to be undisturbed by parasites and produced few limb movements (Supplementary file 1 in Zhao et al., 2022). Thus, the inclusion of data from silent males greatly underestimates any effect. Similar to previous results, our new analysis found that all movements were produced around parasites in calling individuals (Table 1).
Table 1
Outcomes of a generalized linear mixed model (GLMM) to investigate the effect of parasite presence on four limb movements in actively displaying males.
The overall movement and noise data used in this analysis are available in Supplementary file 1.
| Response | Predictor | z | P | Parameter estimates |
|---|---|---|---|---|
| Intercept | 0.858 | 0.391 | 1.251±1.457 | |
| Arm waving | Noise intensity | 0.142 | 0.887 | 0.003±0.022 |
| Parasite | 3.840 | <0.001 | 0.053±0.014 | |
| Intercept | 0.442 | 0.658 | 0.745±1.683 | |
| Hind foot lifting | Noise intensity | 1.057 | 0.291 | 0.026±0.024 |
| Parasite | 1.788 | 0.074 | 0.026±0.015 | |
| Intercept | 1.445 | 0.148 | 2.000±1.384 | |
| Limb shaking | Noise intensity | 0.186 | 0.853 | 0.004±0.020 |
| Parasite | 2.277 | 0.023 | 0.034±0.015 | |
| Intercept | 0.145 | 0.885 | 0.220±1.517 | |
| Wiping | Noise intensity | 0.231 | 0.817 | 0.005±0.023 |
| Parasite | 2.792 | 0.005 | 0.035±0.013 |
Second, the proportion of parasite-induced movements is likely to be underestimated in our original study. For parasite-induced displays, we only included movements that occurred when parasites landed on the body of frogs or moved very close to them. Although the exact distances could not be measured from our videos, given the difficult field conditions, we feel that this is an appropriate approach.
The association between limb displays and parasites
Anderson et al. also argue that the association between limb displays and parasites could be confounded by other environmental variables, such as local climatic conditions, densities of calling males or ambient noise levels. We agree on this issue and have performed a new analysis of our observational data that includes background noise level as a covariate (see Table 1).
When taking additional environmental variables into account, we now find that all four movements tested on female preferences increased with parasite presence. It is important to note, however, that it is not clear how effective the limb movements, in particular a movement such as hind-foot lifting, are as an anti-predator strategy. As suggested by Anderson et al., it would be interesting to control the amount and type of parasites around calling frogs experimentally in order to obtain a better understanding of limb movements as an anti-parasite function. Mechanistically, different limb movements may be produced (somewhat) simultaneously (for example, when they are controlled by the same neuronal and/or physiological pathway). Theoretically, some movements may have evolved after females started to pay attention, and thus immediately functioned in mate attraction. If such movements recruited (part of) the same mechanism, then we would also expect to find a relation with parasite pressure, even if the movement lost or never possessed an anti-parasite function.
The role of male–male interactions is another interesting topic for future studies
The role of limb displays in male–male competition can be easily observed and recorded in the wild, whereas observing or manipulating male–female interactions is challenging in many species. Anderson et al. summarize studies on limb displays in torrent frogs and show that data on female choice behavior is rare, but we disagree with the statement that this means that female choice behavior has a smaller role than male–male competition. Only very few studies have addressed multimodal displays in the context of intra- and intersexual interactions simultaneously, but when they do, they typically show that males and females pay attention to similar cues (Halfwerk et al., 2014; James et al., 2022). In animal communication, it is generally assumed (although perhaps wrongly) that male responses to playback reflect female responses and thus can be used to study signal evolution in a broad comparative framework. It would be interesting to test whether male torrent frogs respond similarly to limb movements that are associated with parasite pressure, as this would provide a much easier way to assess how these movements and responses evolved across the phylogeny.
Conclusions
Our data on males is observational and does not provide evidence for causal relationships, a point on which we agree with Anderson et al. The novelty of our paper lies in the fact that we propose a novel hypothesis of how and why multimodal displays evolve, namely via a process of co-option and possibly sensory exploitation. Studying the evolutionary history of multimodal displays, such as the limb movements produced during calling in torrent frogs, requires a phylogenetic comparative approach. For such a study, data on male–male interactions might be more useful than data on male–female interactions, given the challenges of collecting such data in the field. Finally, our data on female choice is highly novel because we show that several limb movements enhance the attractiveness of a calling male, which is a prerequisite for the co-option of cues into multimodal displays.
Materials and methods
Anderson et al. pointed out that different males had great variation in the number of limb movements, which may bias our original statistics. They tested whether the probability of limb movements increased to a significant extent in response to parasite presence. However, they included silent males outside of a breeding context in their reanalysis, and may therefore underestimate the effect size. This study focused on breeding behaviors in which only calling males are relevant. Moreover, parasite presence was mediated by calling behaviors because parasites often used acoustic signals to find frogs. Thus, only calling individuals were included in our analysis.
We performed a new analysis using generalized linear mixed models (GLMMs) in R (v.4.2.0) to test the effect of parasites on limb display. Models using a Poisson distribution and log-link function were constructed using the package lme4. In each model, the frequency of one of four types of limb display was added as the dependent variable, while parasite number was included as fixed factor and overall movements as random factor. The overall movements represented the total number of all types of limb display. Anderson et al. proposed that individuals with the same proportion of a specific movement may have different impact if the absolute total number of limb movements is different. So the overall movements were treated as random factor to avoid potential statistical bias. Furthermore, to control for environmental variation, noise level at the male display site was added as covariate. The noise intensity of the habitat may vary in different locations and could affect the production of limb movements, as pointed out by Anderson et al. After finishing video recording, we used a sound level meter (AWA 5661, Hangzhou Aihua Instruments Co., Hangzhou, China) to measure the background noise level for each individual at the position of its head, and used these data in our new analysis. A likelihood ratio test (χ2=0, df = 1, P=1) did not show a significant reduction in explanatory power when we compared a full model (which included all terms of interest) with a collapsed model (which excluded a random factor of male ID). Only overall movement was treated as the random factor in order to avoid singularities (Table 1). Individual and environmental variation may affect parasite distribution and limb display. The GLMM evaluated the relationship between the frequency of parasite visit and the number of limb movements, while controlling for individual and environmental variation, and was therefore appropriate to test whether the presence of parasites increased limb movement.
Data availability
The overall movement and noise data used in the new analysis reported in this article are available in Supplementary file 1. The other data used in this analysis are available in Zhao et al., 2022.
-
Dryad Digital RepositoryThe data of parasite-induced and spontaneous displays in each limb movement for calling males, silent males and males that have females nearby.https://doi.org/10.5061/dryad.f1vhhmgzg
References
-
BookCommunication, in Behavioural Ecology: An Evolutionary ApproachOxford: Blackwell Scientific Publications.
-
Cross-modal facilitation of auditory discrimination in a frogBiology Letters 18:20220098.https://doi.org/10.1098/rsbl.2022.0098
Decision letter
-
Ammie K KalanReviewing Editor; University of Victoria, Canada
-
Christian RutzSenior Editor; University of St Andrews, United Kingdom
In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.
In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.
Thank you for submitting your article "Response to comment on 'Parasite defensive limb movements enhance acoustic signal attraction in male little torrent frogs' " to eLife for consideration as Scientific Correspondence.
Your article, and the original comment by Fuxjager and colleagues, have been reviewed by two peer reviewers (who have opted to remain anonymous), and the evaluation was overseen by Ammie Kalan as the Reviewing Editor and Christian Rutz as the Senior Editor.
We have decided to accept for publication both the original comment and your response (although your response will require some revisions; please see below). The substantive comments in this decision letter will also be published as part of your article (and likewise for the comment from Fuxjager).
The reviewers supported publication of your response to the comment article, subject to the following points being addressed in a revised version. In particular, please respond to the points raised by Reviewer #1.
Essential Revisions:
1) Please respond to the concerns regarding your mixed model analysis. Although use of mixed models is not a problem per se, there needs to be some careful explanation of the variables and structure of the model and how they address the particular criticisms raised in the comment (e.g., the sampling-effort issue described by Reviewer #1 below).
2) Please make sure the revised data you analyzed for your response are included in your revised submission.Reviewer #1:
This paper responds to some criticisms of the authors' original article about limb movements and their origin and function in a frog species. The commenters had made an alternative analysis of the data, and here the authors provide yet a different analysis of the data that comes to a different conclusion. I think the authors here do a good job of discussing the main criticisms and I appreciate their efforts to take them seriously and reconsider their own data. I think there are a few places where the rebuttals aren't particularly strong, or are still a bit unclear, which I detail below. The main one is this reanalysis of the data, which needs more explanation and justification.
First paragraph in the “Underestimations of parasite-associated limb movements” subsection: This could be an important point, but it would help to explain why it would be wrong to include recordings from non-calling males. If the question is about whether limb movements are a response to parasite presence, why does it matter whether the male is calling or not? I don't see a good justification for excluding those individuals, but if the authors have one, it would strengthen their argument. (see also first paragraph of “Materials and Methods”)
First paragraph of “Results and Discussion”: It would be helpful to reference Table 1 and the new analyses already here. When I first read this it raised the question of what would happen if a reanalysis was actually done. Later it becomes clear that there was indeed a reanalysis. But it would have made it easier to follow this argument if the results were referred to right away.
Second paragraph in the “Materials and Methods”: I didn't follow what "overall movements of different males" means, or why that would be a random effect. In general, this description of the mixed model is unclear, and this is a pretty important point since it's meant to be refuting the argument put forth in the comment. So this really needs to be clarified. I don't see why a mixed model makes sense here. It actually is pretty difficult to see how to test the hypothesis that limb displays are produced more when parasites are present than when they're absent because the null is unclear (as pointed out by the authors of the comment, who I think make a reasonable attempt to do some stats with an appropriate null). Really what we'd need is to weigh this somehow by observation effort. How many limb displays per minute are produced in the presence of parasites, and how many limb displays per minute are produced in the absence of parasites? If you don't standardize it to some kind of measure of sampling effort then it's really difficult to compare what is happening with and without parasites (it's almost an apples and oranges situation). So to summarize, it would be helpful to be very explicit about what these variables are that are included in the mixed model, to justify the use of a mixed model as opposed to some other kind of statistical test (specifically, justifying the random term), and to explain why this model is appropriate to answer the question of whether limb movements truly are more common in the presence of parasites.Reviewer #2:
- Does the Response respond to the criticisms made in the Comment in a way that is convincing enough to merit publication?
Yes, the responses to the Comment are helpful and somewhat convincing. But they certainly do not show that the Comments are unfounded. However, the reanalysis of their data that shows all four limb movements of the frogs are associated with parasites does tend to reject one of the comments.
- If yes, are any revisions required before the Response can be accepted for publication.
No
[Editors' note: further revisions were requested as described below.]
Clarify in the text if the model results shown in Table 1 include BOTH random effects of Male ID and Overall Movements. If not, please provide a statistic, such as Likelihood Ratio Test, to show that the models with Male ID were not significantly different from the ones with Overall Movements.
https://doi.org/10.7554/eLife.90404.sa1Author response
Essential Revisions:
1) Please respond to the concerns regarding your mixed model analysis. Although use of mixed models is not a problem per se, there needs to be some careful explanation of the variables and structure of the model and how they address the particular criticisms raised in the comment (e.g., the sampling-effort issue described by Reviewer #1 below).
Thank you for your instructions. We have responded to these concerns and added more careful explanation and justification according to below comments. Please see the point-point responses below.
2) Please make sure the revised data you analyzed for your response are included in your revised submission.
Confirmed.
Reviewer #1:
This paper responds to some criticisms of the authors' original article about limb movements and their origin and function in a frog species. The commenters had made an alternative analysis of the data, and here the authors provide yet a different analysis of the data that comes to a different conclusion. I think the authors here do a good job of discussing the main criticisms and I appreciate their efforts to take them seriously and reconsider their own data. I think there are a few places where the rebuttals aren't particularly strong, or are still a bit unclear, which I detail below. The main one is this reanalysis of the data, which needs more explanation and justification.
Thank you for your insightful comments. We have clarified them and replenished explanation and justification according to your detailed comments below. Please see the point-point responses below.
First paragraph in the “Underestimations of parasite-associated limb movements” subsection: This could be an important point, but it would help to explain why it would be wrong to include recordings from non-calling males. If the question is about whether limb movements are a response to parasite presence, why does it matter whether the male is calling or not? I don't see a good justification for excluding those individuals, but if the authors have one, it would strengthen their argument. (see also first paragraph of “Materials and Methods”)
This study focused on breeding behaviors in which only calling males were relevant. Moreover, parasite presence was mediated by calling behaviors because parasites often used acoustic signals to find frogs. Thus, only calling individuals were included here. This has been clarified in “Materials and Methods”. Please see revised p. 6 lines 117-120.
First paragraph of “Results and Discussion: It would be helpful to reference Table 1 and the new analyses already here. When I first read this it raised the question of what would happen if a reanalysis was actually done. Later it becomes clear that there was indeed a reanalysis. But it would have made it easier to follow this argument if the results were referred to right away.
Thank you very much. We have included the new analyses and referred to Table 1 in text.
Second paragraph in the “Materials and Methods”: I didn't follow what "overall movements of different males" means, or why that would be a random effect. In general, this description of the mixed model is unclear, and this is a pretty important point since it's meant to be refuting the argument put forth in the comment. So this really needs to be clarified.
The overall movements represented the total number of all types of limb displays. Anderson et al. proposed that individuals with a same proportion of specific movement may have different impact if the absolute total number is quite different. So the overall movements were treated as random factor to avoid potential statistical bias. This has been clarified.
I don't see why a mixed model makes sense here. It actually is pretty difficult to see how to test the hypothesis that limb displays are produced more when parasites are present than when they're absent because the null is unclear (as pointed out by the authors of the comment, who I think make a reasonable attempt to do some stats with an appropriate null).
Individual and environmental variation may affect parasite distribution and limb display. The GLMM evaluated the relationship between the frequency of parasite visit and the number of limb movement, while controlled for these factors, and therefore was appropriate to test whether the presence of parasite increased limb movement. We have explained this in text.
Really what we'd need is to weigh this somehow by observation effort. How many limb displays per minute are produced in the presence of parasites, and how many limb displays per minute are produced in the absence of parasites? If you don't standardize it to some kind of measure of sampling effort then it's really difficult to compare what is happening with and without parasites (it's almost an apples and oranges situation). So to summarize, it would be helpful to be very explicit about what these variables are that are included in the mixed model, to justify the use of a mixed model as opposed to some other kind of statistical test (specifically, justifying the random term), and to explain why this model is appropriate to answer the question of whether limb movements truly are more common in the presence of parasites.
Thank you for your good advices. In fact, we had already made such observation efforts. We observed all individuals for same duration (i.e. 10 min) and quantified which was in the absence of parasites and which was in presence of parasites. If parasites were found, we also counted the number of parasite appearance. We controlled for different factors and used GLMMs to test the relationship between parasite and limb movement. This is more comprehensive than simple classification of “absence” or “presence” category.
[Editors' note: further revisions were requested as described below.]
Clarify in the text if the model results shown in Table 1 include BOTH random effects of Male ID and Overall Movements. If not, please provide a statistic, such as Likelihood Ratio Test, to show that the models with Male ID were not significantly different from the ones with Overall Movements.
Only overall movements were included in the model. We also added a statistic, Likelihood Ratio Test, to show that the removal of ID did not show a significant reduction in explanatory power. The following text has been added:
“A likelihood ratio test (χ2 = 0, df = 1, P = 1) did not show a significant reduction in explanatory power when compared a full model (included all terms of interest) with a collapsing model (excluded a random factor of male ID). Only overall movement was treated as the random factor in order to avoid singularities (Table 1).”
https://doi.org/10.7554/eLife.90404.sa2Article and author information
Author details
Funding
National Natural Science Foundation of China (32101240)
- Longhui Zhao
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Senior Editor
- Christian Rutz, University of St Andrews, United Kingdom
Reviewing Editor
- Ammie K Kalan, University of Victoria, Canada
Version history
- Received: June 28, 2023
- Accepted: September 12, 2023
- Version of Record published: October 9, 2023 (version 1)
Copyright
© 2023, Zhao et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 138
- Page views
-
- 14
- Downloads
-
- 0
- Citations
Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Response to comment on 'Parasite defensive limb movements enhance acoustic signal attraction in male little torrent frogs'
eLife 12:e90404.
https://doi.org/10.7554/eLife.90404
Further reading
-
- Ecology
- Evolutionary Biology
Many animals rely on complex signals that target multiple senses to attract mates and repel rivals. These multimodal displays can however also attract unintended receivers, which can be an important driver of signal complexity. Despite being taxonomically widespread, we often lack insight into how multimodal signals evolve from unimodal signals and in particular what roles unintended eavesdroppers play. Here, we assess whether the physical movements of parasite defense behavior increase the complexity and attractiveness of an acoustic sexual signal in the little torrent frog (Amolops torrentis). Calling males of this species often display limb movements in order to defend against blood-sucking parasites such as frog-biting midges that eavesdrop on their acoustic signal. Through mate choice tests we show that some of these midge-evoked movements influence female preference for acoustic signals. Our data suggest that midge-induced movements may be incorporated into a sexual display, targeting both hearing and vision in the intended receiver. Females may play an important role in incorporating these multiple components because they prefer signals which combine multiple modalities. Our results thus help to understand the relationship between natural and sexual selection pressure operating on signalers and how in turn this may influence multimodal signal evolution.
-
- Ecology
- Evolutionary Biology
Groups of animals inhabit vastly different sensory worlds, or umwelten, which shape fundamental aspects of their behaviour. Yet the sensory ecology of species is rarely incorporated into the emerging field of collective behaviour, which studies the movements, population-level behaviours, and emergent properties of animal groups. Here, we review the contributions of sensory ecology and collective behaviour to understanding how animals move and interact within the context of their social and physical environments. Our goal is to advance and bridge these two areas of inquiry and highlight the potential for their creative integration. To achieve this goal, we organise our review around the following themes: (1) identifying the promise of integrating collective behaviour and sensory ecology; (2) defining and exploring the concept of a ‘sensory collective’; (3) considering the potential for sensory collectives to shape the evolution of sensory systems; (4) exploring examples from diverse taxa to illustrate neural circuits involved in sensing and collective behaviour; and (5) suggesting the need for creative conceptual and methodological advances to quantify ‘sensescapes’. In the final section, (6) applications to biological conservation, we argue that these topics are timely, given the ongoing anthropogenic changes to sensory stimuli (e.g. via light, sound, and chemical pollution) which are anticipated to impact animal collectives and group-level behaviour and, in turn, ecosystem composition and function. Our synthesis seeks to provide a forward-looking perspective on how sensory ecologists and collective behaviourists can both learn from and inspire one another to advance our understanding of animal behaviour, ecology, adaptation, and evolution.
