Sociability is an individual’s tendency to associate with conspecifics in a non-aggressive manner. Sociability can manifest in the formation of social groups that can reduce predation risk and increase feeding success. Studies of social behaviour in insects are typically through the lens of classically know social insects, however many insect species that have been long thought as non-social have been shown to exhibit social behaviour, in particular Drosophila. A previous experiment evolved lineages of high and low sociable fruit flies (Drosophila melanogaster) following 25 generations of artificial selection, after which RNA and DNA was extracted and sequenced. The main goal of this thesis was to integrate analyses of differential gene expression, transcript usage and population genomics to investigate the genetic architecture of sociability in Drosophila. I developed a pipeline to perform differential gene expression analysis by modelling gene expression using a generalized linear mixed-effect model. Here I found a total of 327 genes differentially expressed and 174 genes differentially expressed between the low and high sociable lineages. Next, I developed a pipeline to perform differential transcript usage analysis using a generalized linear mixed-effect model to model transcript usage. I found 619 genes to have transcripts with differential usage and 190 genes to have transcripts with differential usage between the low and high sociable lineages. Lastly, I developed a pipeline for population genomics to identify regions of the genome under selection. I identified genes that are likely under selection and the overlap between these genes and genes/transcripts found to be differentially expressed/used. Overall, I identified potential genes that are involved in the genetic architecture of sociability and can be further candidate tested. / Thesis / Master of Science (MSc)

No description available.

Com a disponibilidade de tecnologias de genotipagem robustas, fornecendo milhares de marcadores moleculares a um baixo custo por amostra, tornou-se mais acessível escanear o genoma todo. Logo, as abordagens de mapeamento associativo, baseado no cálculo do desequilíbrio de ligação, e a busca de regiões do genoma que apresentam sinais de seleção foram favorecidas. Essas ferramentas apresentam um grande potencial a ser explorado em programas de melhoramento de soja, auxiliando na obtenção de informações valiosas em relação à arquitetura genética dos caracteres de interesse agronômico e à dinâmica dos processos de seleção. Portanto, os objetivos deste estudo foram: (i) verificar os efeitos da relação genética entre genótipos no desequilíbrio de ligação; (ii) obter informações em relação a arquitetura genética para dez caracteres por meio da abordagem de mapeamento associativo; e (iii) identificar regiões sob seleção diferencial em uma população representada por cultivares brasileiras comparadas a de uma população de genótipos exóticos de diferentes origens geográficas. Para isto, 95 genótipos de soja foram genotipados utilizando a plataforma da Affymettrix (180 K Axiom® Soybean Genotyping Array). O efeito da estrutura populacional sobre DL foi investigado utilizando como correções as matrizes resultantes do software STRUCTURE, DAPC, PCA e a matriz de parentesco. Para o mapeamento associativo, os genótipos foram fenotipados para dez caracteres. Foram realizadasvárias abordagens para detecção de sinais de seleção: estimação da diferenciação populacional nas regiões do genoma por meio do índice de fixção (FST) e cross-population composite likelihood ratio test (XP-CLR); identificação de regiões do genoma com redução de diversidade nucleotídica (μ); e a presença de blocos de haplótipos. Ao utilizar medidas de correções para estimação do DL no genoma da soja observamos a influência da estrutura da população e do parentesco nos padrões de DL. Pelo mapeamento associativo, foram identificados 181 marcadores associados para dez caracteres avaliados em soja. Além disso, foi verificada a complexa interação entre regiões envolvidas no controle de caracteres quantitativos (QTL) e ambientes. Regiões diferencialmente selecionadas foram identificadas entre a população de genótipos brasileiros e de materiais de origens diversas, demonstrando que essas passaram por processos de seleção divergente. / The availability of robust genotyping technologies providing thousands of markers with a low-cost per sample, whole-genome scans are becoming more accessible. Hence, associative mapping approaches, based on the calculation of linkage disequilibrium (LD), and the detection of signals of selection were favored. These tools have a great potential to be explored in soybean breeding programs, helping to obtain valuable information about the genetic architecture of the characters of agronomic interest and of the dynamics of selection processes. Therefore, the objectives of this study were: (i) to verify the effects of the genetic relationship among genotypes on soybean linkage disequilibrium; (ii) to obtain information about the genetic architecture in ten soybean traits with an association mapping approach; and (iii) to identify regions under differential selection between a population represented by Brazilian cultivars compared to a population of exotic genotypes from different geographic origins. For this, 95 soybean genotypes were genotyped using an Affymetrix (180 K Axiom® Soybean Genotyping Array) platform. The effect of the population structure on LD was investigated using as corrections the matrices resulting from the software STRUCTURE, DAPC, PCA and a kinship matrix. We phenotyped ten soybean traits in order to carry the association mapping. Several approaches were carried aiming to detect selection signals: estimation of population genetic differentiation in the genomic regions with fixation index (FST) and cross-population composite likelihood ratio test (XP-CLR); identification of genomic regions with reduced nucleotide diversity (μ); and the presence of haplotype blocks.Using the LD corrections, it was showed the influence of population structure and kinship on LD patterns in soybean. By the association mapping, we identified 181 markers associated with the ten traits evaluated in soybean. In addition, complex interactions between QTL and environments were verified. Differentially selected regions were identified between theBrazilian genotypes population and the population of materials from diverse origins, demonstrating that they underwent divergent selection processes.

Depuis leur domestication, les équidés sont étroitement associés aux activités humaines et ont, au fil des siècles, été façonnés au gré d’exigences morphologiques, esthétiques, d’allure ou de performance. Cette sélection artificielle a fortement impacté leurs traits phénotypiques et fonctionnels, produisant le large panel des races actuelles. Les sources historiques ont abondamment décrit l’importance des équidés et la diversité de leurs usages dans les sociétés passées. Ceci interroge donc sur la potentielle existence de types morphologiques spécialisés à des périodes antérieures à l’émergence des races. Dans ce contexte, les os présentent un intérêt particulier en tant que reflet des caractéristiques morphologiques et fonctionnelles des animaux du passé. En outre, les équidés étant fréquemment retrouvés sous la forme de squelettes complets en contexte archéologique,leurs restes offrent la possibilité d’étudier l’intégralité de la morphologie squelettique et notamment les interactions entre les os. Pourtant, les restes osseux d’équidés restent relativement peu exploités, probablement en raison des limites inhérentes aux méthodes d’étude actuelles. L’objectif de ce travail est de mieux comprendre la diversité phénotypique et fonctionnelle des équidés domestiques par des approches en morphométriegéométrique. La question de leur identification spécifique est tout d’abord abordée via la recherche de critères discriminants, qualitatifs et quantitatifs, applicables à du matériel archéologique. Les patterns de covariation entre les os sont ensuite explorés afin d’aborder des questions fonctionnelles. Les résultats révèlent une forte intégration morphologique entre les os des membres chez les équidés domestiques et mettent en évidence des différences entre les races selon un axe de covariation principalement porté par des interactions fonctionnelles. Ceci tend à montrer que la sélection artificielle, considérée comme le principal acteur de la diversification morphologique chez le cheval domestique, n’influence pas seulement le phénotype mais aussi les facteurs biologiques qui le sous-tendent. Enfin, une première application à des spécimens archéologiques permet dediscuter l’impact de potentielles formes de sélection artificielle et de standardisation morphologique sur des chevaux anciens. Les résultats démontrent l’intérêt d’étudier non seulement les variations de forme des os, mais aussi leurs covariations, afin d’enrichir nos connaissances concernant les traits morphologiques et fonctionnels des animaux passés, ainsi que les pratiques d’élevage qui y sont associées. L’étude des covariations contribueégalement à accroitre notre compréhension des processus micro-évolutifs, tels que la sélection artificielle, et à travers cela, permet de mieux documenter la manière dont la diversité phénotypique est produite. / Equids and humans share a long history of interaction from the first domestication to the standardization of modern breeds. In order to suit human activities, they have been molded through selection for conformation, harmony, gaits, or performance. This artificial selection is known to have largely impacted morphological and functional traits, producing the large range of current breeds. Historical sources extensively described the widepanel of equid usage in different human civilizations, raising the issue of the potential existence of specialized morphological types in the past, prior to the emergence of modern breeds. In this respect, bones prove to be of particular interest, as an image of the phenotypic and functional characteristics of past animals. Moreover, horses being often found as complete skeletons in archaeological contexts, their remains allow for the study of the complete skeletal morphology, especially concerning the interactions between bones. However, equid bone remains are largely under-explored, probably due to the recurrent limitations inherent to existing study methods. The objective of this work is to describe and understand the phenotypic and functional diversity of domestic equids, using geometric morphometrics approaches. Identification issues are firstly addressed through the research of qualitative and quantitative discrimination criteria, applicable to archaeological samples. Next, morphological and functional questions are addressed, especially investigating bone shape covariation. The obtained results reveal strong morphological integration within equid limb bones and evidenced breed specific differences along a covariation axis largely produced by functional interactions between bones. They show thatartificial selection, regarded as responsible of most of the modern diversification of horse breeds, not only targets the phenotype but also impacts the biological factors which underlie it. Finally, a first application to archaeological skeletons allows to question the influence of potential artificial selection and morphological standardization on past horses. The results demonstrate the interest of not only exploring bone shape variation,but also covariation, to increase our knowledge about the morphological and functional traits of past equids and about the related breeding practices. The study of morphological integration may also contribute to enhance our understanding of micro-evolutionary processes, such as artificial selection on domestic taxa, and through that, gain insights into how phenotypic diversity is produced.

Com a disponibilidade de tecnologias de genotipagem robustas, fornecendo milhares de marcadores moleculares a um baixo custo por amostra, tornou-se mais acessível escanear o genoma todo. Logo, as abordagens de mapeamento associativo, baseado no cálculo do desequilíbrio de ligação, e a busca de regiões do genoma que apresentam sinais de seleção foram favorecidas. Essas ferramentas apresentam um grande potencial a ser explorado em programas de melhoramento de soja, auxiliando na obtenção de informações valiosas em relação à arquitetura genética dos caracteres de interesse agronômico e à dinâmica dos processos de seleção. Portanto, os objetivos deste estudo foram: (i) verificar os efeitos da relação genética entre genótipos no desequilíbrio de ligação; (ii) obter informações em relação a arquitetura genética para dez caracteres por meio da abordagem de mapeamento associativo; e (iii) identificar regiões sob seleção diferencial em uma população representada por cultivares brasileiras comparadas a de uma população de genótipos exóticos de diferentes origens geográficas. Para isto, 95 genótipos de soja foram genotipados utilizando a plataforma da Affymettrix (180 K Axiom® Soybean Genotyping Array). O efeito da estrutura populacional sobre DL foi investigado utilizando como correções as matrizes resultantes do software STRUCTURE, DAPC, PCA e a matriz de parentesco. Para o mapeamento associativo, os genótipos foram fenotipados para dez caracteres. Foram realizadasvárias abordagens para detecção de sinais de seleção: estimação da diferenciação populacional nas regiões do genoma por meio do índice de fixção (FST) e cross-population composite likelihood ratio test (XP-CLR); identificação de regiões do genoma com redução de diversidade nucleotídica (μ); e a presença de blocos de haplótipos. Ao utilizar medidas de correções para estimação do DL no genoma da soja observamos a influência da estrutura da população e do parentesco nos padrões de DL. Pelo mapeamento associativo, foram identificados 181 marcadores associados para dez caracteres avaliados em soja. Além disso, foi verificada a complexa interação entre regiões envolvidas no controle de caracteres quantitativos (QTL) e ambientes. Regiões diferencialmente selecionadas foram identificadas entre a população de genótipos brasileiros e de materiais de origens diversas, demonstrando que essas passaram por processos de seleção divergente. / The availability of robust genotyping technologies providing thousands of markers with a low-cost per sample, whole-genome scans are becoming more accessible. Hence, associative mapping approaches, based on the calculation of linkage disequilibrium (LD), and the detection of signals of selection were favored. These tools have a great potential to be explored in soybean breeding programs, helping to obtain valuable information about the genetic architecture of the characters of agronomic interest and of the dynamics of selection processes. Therefore, the objectives of this study were: (i) to verify the effects of the genetic relationship among genotypes on soybean linkage disequilibrium; (ii) to obtain information about the genetic architecture in ten soybean traits with an association mapping approach; and (iii) to identify regions under differential selection between a population represented by Brazilian cultivars compared to a population of exotic genotypes from different geographic origins. For this, 95 soybean genotypes were genotyped using an Affymetrix (180 K Axiom® Soybean Genotyping Array) platform. The effect of the population structure on LD was investigated using as corrections the matrices resulting from the software STRUCTURE, DAPC, PCA and a kinship matrix. We phenotyped ten soybean traits in order to carry the association mapping. Several approaches were carried aiming to detect selection signals: estimation of population genetic differentiation in the genomic regions with fixation index (FST) and cross-population composite likelihood ratio test (XP-CLR); identification of genomic regions with reduced nucleotide diversity (μ); and the presence of haplotype blocks.Using the LD corrections, it was showed the influence of population structure and kinship on LD patterns in soybean. By the association mapping, we identified 181 markers associated with the ten traits evaluated in soybean. In addition, complex interactions between QTL and environments were verified. Differentially selected regions were identified between theBrazilian genotypes population and the population of materials from diverse origins, demonstrating that they underwent divergent selection processes.

A rich diversity of social behaviours exists in the animal kingdom, and these behaviours have evolved to perform a variety of adaptive functions. Social behaviours show variation both among and within species, however the mechanisms that give rise to this variation are not well understood. Using fruit flies (Drosophila melanogaster), my goal was to uncover the genetic and behavioural mechanisms that underpin natural variation in two different social behaviours: sociability and sexual aggression. First, I showed that sociability, which is the tendency of animals to engage in friendly activities together, is influenced by indirect genetic effects (IGEs), and that encounters among individuals drive these effects (Chapter 2). I then showed that sociability and social plasticity have low-moderate heritability (Chapter 3), and sociability is not correlated between the sexes or with activity. I then generated lineages of flies with high and low sociability using artificial selection (Chapter 4). The evolved lineages had significantly diverged sociability which was not associated with fitness measures or nearest-neighbor distances, but was negatively correlated with intrasexual aggression (Chapter 4). Finally, in sexual aggression, which I quantified as male forced copulation rate, I showed that evolved differences and differences due to social plasticity were both associated with the differential expression of many genes, but only a few of these genes were significant in both (Chapter 5). I also showed that these sets of genes are enriched in neuropeptide hormone and serotonin gene ontology categories, and that 4 of 7 chosen genes were validated for their effects on sexual aggression. Overall, this thesis sheds light on the complex mechanisms that underlie variation in these social behaviours, and it paves the way for future research to further elucidate some of these mechanisms, especially on the genetic basis of sociability using the evolved lineages I generated. / Thesis / Doctor of Philosophy (PhD) / Individual animals tend to vary in many traits including social behaviours. Using fruit flies, my goal was to understand what causes individuals to vary in two social behaviours: sociability and sexual aggression. I found that highly sociable flies tended to influence other flies to become more sociable due to a change in how much these flies interacted. I also found that individual differences in sociability are moderately heritable, and the genetic variation contributing to this is different between the sexes. Also, less sociable flies tended to be more aggressive than highly sociable flies. Finally, for sexual aggression, I showed that variation in a male’s success in forcibly mating with a female was associated with changes in the expression of hundreds of genes, but these changes were mostly unique for evolved versus environmentally induced variation. Future work will similarly look to identify genes involved with individual differences in sociability.

In the present study, the ease of rearing, short generation time and molecular research tools available for the zebrafish model (Danio rerio, Hamilton) were exploited to investigate transcriptional regulation in relation to feeding, photoperiod and experimental selection. Chapter 2 describes transcriptional regulation in fast skeletal muscle following fasting and a single satiating meal of bloodworms. Changes in transcript abundance were investigated in relation to the food content in the gut. Using qPCR, the transcription patterns of 16 genes comprising the insulin-like growth factor (IGF) system were characterized, and differential regulation between some of the paralogues was recorded. For example, feeding was associated with upregulation of igf1a and igf2b at 3 and 6h after the single-meal was offered, respectively, whereas igf1b was not detected in skeletal muscle. On the other hand, fasting triggered the upregulation of the igf1 receptors and igfbp1a/b, the only binding proteins whose transcription was responsive to a single-satiating meal. In addition to the investigation of the IGF-axis, an agnostic approach was used to discover other genes involved in transcriptional response to nutritional status, by employing a whole-genome microarray containing 44K probes. This resulted in the discovery of 147 genes in skeletal muscle that were differentially expressed between fasting and satiation. Ubiquitin-ligases involved in proteasome-mediated protein degradation, and antiproliferative and pro-apoptotic genes were among the genes upregulated during fasting, whereas satiation resulted in an upregulation of genes involved in protein synthesis and folding, and a gene highly correlated with growth in mice and fish, the enzyme ornithine decarboxylase 1. Zebrafish exhibit circadian rhythms of breeding, locomotor activity and feeding that are controlled by molecular clock mechanisms in central and peripheral organs. In chapter 3 the transcription of 17 known clock genes was investigated in skeletal muscle in relation to the photoperiod and food content in the gut. The hypothesis that myogenic regulatory factors and components of the IGF-pathway were clock-controlled was also tested. Positive (clock1 and bmal1 paralogues) and negative oscillators (cry1a and per genes) showed a strong circadian pattern in skeletal muscle in anti-phase with each other. MyoD was not clock-controlled in zebrafish in contrast to findings in mice, whereas myf6 showed a circadian pattern of expression in phase with clock and bmal. Similarly, the expression of two IGF binding proteins (igfbp3 and 5b) was circadian and in phase with the positive oscillators clock and bmal. It was also found that some paralogues responded differently to photoperiod. For example, clock1a was 3-fold more responsive than clock1b. Cry1b did not show a circadian pattern of expression. These patterns of expression provide evidence that the molecular clock mechanisms in skeletal muscle are synchronized with the molecular clock in central pacemaker organs such as eyes and the pineal gland. Using the short generation time of zebrafish the effects of selective breeding for body size at age were investigated and are described in chapter 4. Three rounds of artificial selection for small (S-lineage) and large body size (L-lineage) resulted in zebrafish populations whose average standard length were, respectively, 2% lower and 10% higher than an unselected control lineage (U-lineage). Fish from the L-lineage showed an increased egg production and bigger egg size with more yolk, possibly contributing to the larger body size observed in the early larval stage (6dpf) of fish from this lineage. Fish from S- and L-lineage exposed to fasting and refeeding showed very similar feed intake, providing evidence that experimental selection did not cause significant changes in appetite control. Investigation of the expression of the IGF-axis and nutritionally-response in skeletal muscle after fasting and refeeding revealed that the pattern of expression was not different between the selected lineages, but that a differential responsiveness was observed in a limited number of genes, providing evidence that experimental selection might have changed the way fish allocate the energy acquired through feeding. For example, a constitutive higher expression of igf1a was recorded in skeletal muscle of fish from the L-lineage whereas igfbp1a/b transcripts were higher in muscle of fish from the S-lineage. These findings demonstrate the rapid changes in growth and transcriptional response in skeletal muscle of zebrafish after only three rounds of selection. Furthermore, it provides evidences that differences in growth during embryonic and larval stages might be related to higher levels of energy deposited during oogenesis, whereas differences in adult fish were better explained by changes in energy allocation instead of energy acquisition. In chapter 5 the main findings made during this study and their impact on the literature are discussed.

Competition over access for mates has led to the evolution of many striking examples of morphological traits and behaviour in animals. The rapid development of the sexual selection field in recent decades have dramatically advanced our understanding of what traits make individuals more successful in attracting mates and how preferences for mates evolve over time. However, till now, research in this field has put less emphasis on the mechanisms that underlie variation in mate choice and sexual behaviour. Cognitive processes could potentially be key drivers of individual variation in mating preferences and sexual behaviours and therefore deserve further investigation. In this thesis, I used guppies artificially selected for relative brain size as the model system to study the association between brain size, cognitive ability and various aspects of mate choice. Previous studies in this model system showed that large-brained individuals of both sexes outperformed small-brained individuals in cognitive tests. Here I quantified their sexual behaviours and mating preferences to provide novel empirical data concerning the association between brain size, cognitive ability and sexual selection. In dichotomous choice preference tests based on visual cues, comparisons between large-brained and small-brained guppies showed important differences in their assessment of mate quality. These results are not driven by pre-existing visual biases caused by the artificial selection since further investigation of the visual capacity of these fish detected no differences between large-brained and small-brained individuals in their sensitivity to colour or in their capacity to resolve spatial detail. I also quantified sexual behaviour in male guppies artificially selected for relative brain size and found no difference in the behaviours of large-brained and small-brained males in a single male-single female non-competitive scenario. On the contrary, in a more complex social setting I found a reduction in large-brained males in the rate of courtship towards females and dominance displays towards other males when exposed to different degrees of predation threat and different numbers of male competitors. However, this reduction in behavioural intensity did not result in a lower access to copulation with females for large-brained males. I likewise evaluated female sexual behaviour and found that large-brained females had higher behavioural flexibility such that they decreased their receptiveness towards males more strongly under higher levels of predation threat. Together, these results provide novel empirical evidence that brain size and cognitive ability are tightly linked to mating preferences and sexual behaviours. These findings suggest that brain size and cognitive ability might be important mechanisms behind variation in mating preferences and in sexually selected traits across and within species. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Manuscript. Paper 3: Manuscript. Paper 5: Manuscript.</p>

Im ersten Teil analysiere ich die Selbstorganisation zur Kritikalität (hier ein Phasenübergang von Ordnung zu Unordnung) und untersuche, ob Evolution ein möglicher Organisationsmechanismus ist. Die Kernfrage ist, ob sich ein simulierter kohäsiver Schwarm, der versucht, einem Raubtier auszuweichen, durch Evolution selbst zum kritischen Punkt entwickelt, um das Ausweichen zu optimieren? Es stellt sich heraus, dass (i) die Gruppe den Jäger am besten am kritischen Punkt vermeidet, aber (ii) nicht durch einer verstärkten Reaktion, sondern durch strukturelle Veränderungen, (iii) das Gruppenoptimum ist evolutionär unstabiler aufgrund einer maximalen räumlichen Selbstsortierung der Individuen. Im zweiten Teil modelliere ich experimentell beobachtete Unterschiede im kollektiven Verhalten von Fischgruppen, die über mehrere Generationen verschiedenen Arten von größenabhängiger Selektion ausgesetzt waren. Diese Größenselektion soll Freizeitfischerei (kleine Fische werden freigelassen, große werden konsumiert) und die kommerzielle Fischerei mit großen Netzbreiten (kleine/junge Individuen können entkommen) nachahmen. Die zeigt sich, dass das Fangen großer Fische den Zusammenhalt und die Risikobereitschaft der Individuen reduziert. Beide Befunde lassen sich mechanistisch durch einen Aufmerksamkeits-Kompromiss zwischen Sozial- und Umweltinformationen erklären. Im letzten Teil der Arbeit quantifiziere ich die kollektive Informationsverarbeitung im Feld. Das Studiensystem ist eine an sulfidische Wasserbedingungen angepasste Fischart mit einem kollektiven Fluchtverhalten vor Vögeln (wiederholte kollektive Fluchttauchgängen). Die Fische sind etwa 2 Zentimeter groß, aber die kollektive Welle breitet sich über Meter in dichten Schwärmen an der Oberfläche aus. Es zeigt sich, dass die Wellengeschwindigkeit schwach mit der Polarisation zunimmt, bei einer optimalen Dichte am schnellsten ist und von ihrer Richtung relativ zur Schwarmorientierung abhängt. / In the first part, I focus on the self-organization to criticality (here an order-disorder phase transition) and investigate if evolution is a possible self-tuning mechanism. Does a simulated cohesive swarm that tries to avoid a pursuing predator self-tunes itself by evolution to the critical point to optimize avoidance? It turns out that (i) the best group avoidance is at criticality but (ii) not due to an enhanced response but because of structural changes (fundamentally linked to criticality), (iii) the group optimum is not an evolutionary stable state, in fact (iv) it is an evolutionary accelerator due to a maximal spatial self-sorting of individuals causing spatial selection. In the second part, I model experimentally observed differences in collective behavior of fish groups subject to multiple generation of different types of size-dependent selection. The real world analog to this experimental evolution is recreational fishery (small fish are released, large are consumed) and commercial fishing with large net widths (small/young individuals can escape). The results suggest that large harvesting reduces cohesion and risk taking of individuals. I show that both findings can be mechanistically explained based on an attention trade-off between social and environmental information. Furthermore, I numerically analyze how differently size-harvested groups perform in a natural predator and fishing scenario. In the last part of the thesis, I quantify the collective information processing in the field. The study system is a fish species adapted to sulfidic water conditions with a collective escape behavior from aerial predators which manifests in repeated collective escape dives. These fish measure about 2 centimeters, but the collective wave spreads across meters in dense shoals at the surface. I find that wave speed increases weakly with polarization, is fastest at an optimal density and depends on its direction relative to shoal orientation.