Why the brown ghost chirps at night

Reviewer #1 (Public Review):

The authors investigate the role of chirping in a species of weakly electric fish. They subject the fish to various scenarios and correlate the production of chirps with many different factors. They find major correlations between the background beat signals (continuously present during any social interactions) or some aspects of social and environmental conditions with the propensity to produce different types of chirps. By analyzing more specifically different aspects of these correlations they conclude that chirping patterns are related to navigation purposes and the need to localize the source of the beat signal (i.e. the location of the conspecific).

The study provides a wealth of interesting observation of behavior and much of this data constitute a useful dataset to document the patterns of social interactions in these fish. Some data, in particular the high propensity to chirp in cluttered environments, raises interesting questions. Their main hypothesis is a useful addition to the debate on the function of these chirps and is worth being considered and explored further. However, the data they provide does not support strong conclusion statements arguing that these chirps are used for localization purposes and is even less convincing at rejecting previously established hypotheses on the communication purpose of the chirps. I would suggest thoroughly revising the manuscript to provide a neutral description of the results and leaving any speculations and interpretations for the discussion where the authors should be careful to separate strongly supported hypotheses from more preliminary speculations. I detail below several instances where the argumentation and/or the analysis are flawed.

- They analyze chirp patterning and show that, most likely, a chirp by an individual is followed by a chirp in the same individual. They argue that it is rare that a chirp elicits a "response" in the other fish. Even if there are clearly stronger correlations between chirps in the same individual, they provide no statistical analysis that discards the existence of occasional "response" patterns. The fact that these are rare, and that the authors don't do an appropriate analysis of probabilities, leads to this unsupported conclusion.
- One of the main pieces of evidence that chirps can be used to enhance conspecific localization is based on their "interference" measure. The measure is based on an analysis of "inter-peak-intervales". This in itself is a questionable choice. The nervous system encodes all parts of the stimulus, not just the peak, and disruption occurring at other phases of the beat might be as relevant. The interference will be mostly affected by the summed duration of intervals between peaks in the chirp AM. They do not explain why this varies with beat frequency. It is likely that the changes they see are simply an artifact of the simplistic measure. A clear demonstration that this measure is not adequate comes from the observation in Fig7E-H. They show that the interference value changes as the signal is weaker. This measure should be independent of the strength of the signal. The method is based on detecting peaks and quantifying the time between peaks. The only reason this measure could be affected by signal strength is if noisy recordings affect how the peak detection occurs. There is no way to argue that this phenomenon would happen the same way in the nervous system. Furthermore, they qualitatively argue that patterns of chirp production follow patterns of interference strength. No statistical demonstration is done. Even the qualitative appraisal is questionable. For example, they argue that there are relatively few chirps being produced for DFs of 60 or -60 Hz. But these are DF where they have only a very small sample size. The single pair of fish that they recorded at some of these frequencies might not have chirped by chance and a rigorous statistical analysis is necessary. Similarly, in Fig 5C they argue that the position of the chirps fall on areas of the graph where the interferences are strongest (darker blue) but this is far from obvious and, again, not proven.
- They relate the angle at which one fish produces chirps relative to the orientation of the mesh enclosing. They argue that this is related to the orientation of electric field lines by doing a qualitative comparison with a simplified estimate of field lines. To be convincing this analysis should include a quantitative comparison using the exact same body position of the two fish when the chirps are emitted.
-They show that the very vast majority of chirps in Fig 6 occur when the fish are within a few centimeters (e.g. very large first bin in Fig6E-Type2). This is a situation when the other fish signal will be strongest and localization will be the easiest. It is hard to understand why the fish would need a mechanism to enhance localization in these conditions (this is the opposite of difficult conditions e.g. the "cluttered" environment).
- The argumentation aimed at rejecting the well-established role of chirp in communication is weak at best. First, they ignored some existing data when they argue that there is no correlation between chirping and behavioral interactions. Particularly, Hupe and Lewis (2008) showed a clear temporal correlation between chirps and a decrease in bites during aggressive encounters. It could be argued that this is "causal evidence" (to reuse their wording) that chirps cause a decrease in attacks by the receiver fish (see Fig 8B of the Hupe paper and associated significant statistics). Also, Oboti et al. argue that social interactions involve "higher levels of locomotion" which would explain the use of chirps since they are used to localize. But chirps are frequent in "chirp chamber" paradigms where no movement is involved. They also point out that social context covaries with beat frequency and thus that it is hard to distinguish which one is linked to chirping propensity and then say that it is hard to disentangle this from "biophysical features of EOD fields affecting detection and localization of conspecific fish". But they don't provide any proof that beat frequency affects detection and localization so their argument is not clear. Last, they argue that tests in one species shouldn't be extrapolated to other species. But many of the studies arguing for the role of chirps in communication was done on brown ghost. In conclusion of this point, they do not provide any strong argument that rejects the role of chirps as a communication signal. A perspective that would be better supported by their data and consistent with past research would be to argue that, in addition to a role in communication, chirps could sometimes be used to help localize conspecifics.
-The discussion they provide on the possible mechanism by which chirps could help with localization of the conspecific is problematic. They imply that chirps cause a stronger response in the receptors. For most chirps considered here, this is not true. For a large portion of the beat frequencies shown in this paper, chirps will cause a de-synchronization of the receptors with no increase in firing rate. They cannot argue that this represents an enhanced response. They also discuss a role for having a broader frequency spectrum -during the chirp- in localization by making a parallel with pulse fish. There is no evidence that a similar mechanism could even work in wave-type fish.
-They write the whole paper as if males and females had been identified in their experiments. Although EOD frequency can provide some guess of the sex the method is unreliable. We can expect a non-negligible percentage of error in assigning sex.

Reviewer #2 (Public Review):

Studying the weakly electric brown ghost knifefish, the authors provide evidence that 'chirps' (brief modulations in the frequency and amplitude of the ongoing electric signal) function in active sensing (specifically homeoactive sensing) rather than communication. This is a behavior that has been very well studied, including numerous studies on the sensory coding of chirps and the neural mechanisms for chirp generation. Chirps are largely thought to function in communication behavior, so this alternative function is a very exciting possibility that could have a great impact on the field. The authors do provide convincing evidence that chirps may function in homeoactive sensing. However, their evidence arguing against a role for chirps in communication is not as strong, and neglects a large body of research. Ultimately, the manuscript has great potential but suffers from framing these two possibilities as mutually exclusive and dismissing evidence in favor of a communicative function.

(1) The specific underlying question of this study is not made clear in the abstract or introduction. It becomes apparent in reading through the manuscript that the authors seek to test the hypothesis that chirps function in active sensing (specifically homeoactive sensing). This should be made explicitly clear in both the abstract and introduction, along with the rationale for this hypothesis.

(2) My biggest issue with this manuscript is that it is much too strong in dismissing evidence that chirping correlates with context. This is captured in this sentence in the introduction, "We first show that the choice of different chirp types does not significantly correlate with any particular behavioral or social context." This very strong conclusion comes up repeatedly, and I disagree with it, for the following reasons:

In your behavioral observations, you found sex differences in chirping as well as differences between freely interacting and physically separated fish. Your model of chirp variability found that environmental experience, social experience, and beat frequency (DF) are the most important factors explaining chirp variability. Are these not all considered "behavioral or social context"? Beat frequency (DF) in particular is heavily downplayed as being a part of "context" but it is a crucial part of the context, as it provides information about the identity of the fish you're interacting with.

In your playback experiments, fish responded differently to small vs. large DFs, males chirped more than females, type 2 chirps became more frequent throughout a playback, and rises tended to occur at the end of a playback. These are all examples of context-dependent behavior.

Further, you only considered the identity of interacting fish or stimulated fish, not their behavior during the interaction or during playback. Such an analysis is likely beyond the scope of this study, but several other studies have shown correlations between social behavior and chirping. In the absence of such data here, it is too strong to claim that chirping is unrelated to context.

In summary, it is simply too strong to say that chirping does not correlate with context. Importantly, however, this does not detract from your hypothesis that chirping functions in homeoactive sensing. A given EOD behavior could serve both communication and homeoactive sensing. I actually suspect that this is quite common in electric fish. The two are not mutually exclusive, and there is no reason for you to present them as such. I recommend focusing more on the positive evidence for a homeoactive function and less on the negative evidence against a communication function.

(3) The results were generally challenging to follow. In the first 4 sections, it is not made clear what the specific question is, what the approach to addressing that question is, and what specific experiment was carried out (the last two sections of the results were much clearer). The independent variables (contexts) are not clearly established before presenting the results. Instead they are often mentioned in passing when describing the results. They come across as an unbalanced hodgepodge of multiple factors, and it is not made clear why they were chosen. This makes it challenging to understand why you did what you did, the results, and their implications. For each set of major results, I recommend: First, pose a clear question. Then, describe the general approach to answering that question. Next, describe the specifics of the experimental design, with a rationale that appeals to the general approach described. Finally, describe the specific results.

(4) Results: "We thus predicted that, if behavioral meaning can be attributed to different types of chirps, as posed by the prevailing view (e.g., Hagedorn and Heiligenberg, 1985; Larimer and MacDonald, 1968; Rose, 2004)..." It should be made clear why this is the prevailing view, and this description should likely be moved to the introduction. There is a large body of evidence supporting this view and it is important to be complete in describing it, especially since the authors seem to seek to refute it.

(5) I am not convinced of the conclusion drawn by the analysis of chirp transitions. The transition matrices show plenty of 1-2 and 2-1 transitions occurring. Further, the cross-correlation analysis only shows that chirp timing between individuals is not phase-locked at these small timescales. It is entirely possible that chirp rates are correlated between interacting individuals, even if their precise timing is not.

Reviewer #3 (Public Review):

Summary:
This important paper provides the best-to-date characterization of chirping in weakly electric fish using a large number of variables. These include environment (free vs divided fish, with or without clutter), breeding state, gender, intruder vs resident, social status, locomotion state and social and environmental experience, as well as with playback experiments. It applies state-of-the-art methods for reducing dimensionality and finding patterns of correlation between different kinds of variables (factor analysis, K-means). The exceptional strength of the evidence, collated from a large number of trials with many controls, leads to the conclusion that a number of commonly accepted truths about which variable affects chirping must be carefully rewritten or nuanced. Based on their extensive analyses, the authors suggest that chirps are mainly used as probes that help detect beats and objects.

Strengths:
The work is based on completely novel recordings using interaction chambers. The amount of new data and associated analyses is simply staggering, and yet, well organized in presentation. The study further evaluates the electric field strength around a fish (via modelling with the boundary element method) and how its decay parallels the chirp rate, thereby relating the above variables to electric field geometry.

The main conclusions are that the lack of any significant behavioural correlates for chirping, and the lack of temporal patterning in chirp time series, cast doubt on a communication goal for most chirps. Rather, the key determinants of chirping are the difference frequency between two interacting conspecifics as well as individual subjects' environmental and social experience. These conclusions by themselves will be hugely useful to the field. They will also allow scientists working on other "communication" systems to at least reconsider, and perhaps expand the precise goal of the probes used in those senses. There are a lot of data summarized in this paper, and thorough referencing to past work. For example, the paper concludes that there is a lack of evidence for stereotyped temporal patterning of chirp time series, as well as of sender-received chirp transitions beyond the known increase in chirp frequency during an interaction.

The alternative hypotheses that arise from the work are that chirps are mainly used as environmental probes for better beat detection and processing and object localization.

The authors also advance the interesting idea that the sinusoidal frequency modulations caused by chirps are the electric fish's solution to the minute (and undetectable by neural wetware) echo-delays available to it, due to the propagation of electric fields at the speed of light in water.

Weaknesses:
My main criticism is that the alternative putative role for chirps as probe signals that optimize beat detection could be better developed. The paper could be clearer as to what that means precisely. And there is an egg-and-chicken type issue as well, namely, that one needs a beat in order to "chirp" the beating pattern, but then how does chirping optimize the detection of the said beat? Perhaps the authors mean (as they wrote elsewhere in the paper) that the chirps could enhance electrosensory responses to the beat.

A second criticism is that the study links the beat detection to underwater object localization. I did not see a sufficiently developed argument in this direction, nor how the data provided support for this argument. It is certainly possible that the image on the fish's body of an object in the environment will be slightly modified by introducing a chirp on the waveform, as this may enhance certain heterogeneities of the object in relation to its environment. The thrust of this argument seems to derive more from the notion of Fourier analysis with pulse type fish (and radar theory more generally) that the higher temporal frequencies in the beat waveform induced by the chirp will enable a better spatial resolution of objects. It remains to be seen whether this is significant.

I would also have liked to see a proposal for new experiments that could test these possible new roles.

The authors should recall for the readers the gist of Bastian's 2001 argument that the chirp "can adjust the beat frequency to levels that are better detectable" in the light of their current. Further, at the beginning of the "Probing with chirps" section, the 3rd way in which chirps could improve conspecific localization mentions the phase-shifting of the EOD. The authors should clarify whether they mean that the tuberous receptors and associated ELL/toral circuitry could deal with that cue, or that the T_unit pathway would be needed?

On p.17 I don't understand what is meant by most chirps being produced possibly aligned with the field lines, since field lines are everywhere. And what is one to conclude from the comparison of Fig.6D and 7A? Likewise it was not clear what is meant by chirps having a detectable effect on randomly generated beats.

In the section on Inconsistencies between behaviour and hypothesized signal meaning, the authors could perhaps nuance the interpretation of the results further in the context of the unrealistic copy of natural stimuli using EOD mimics. In particular, Kelly et al. 2008 argued that electrode placement mattered in terms of representation of a mimic fish onto the body of a real fish, and thus, if I properly understand the set up here, the movement would cause the mimic to vary in quality. This may nevertheless be a small confounding issue.