Sociality in a habitat specialist fishAn investigation of the ecological constraints hypothesisMartin Hing†, Marian Wong†, Mark Dowton† and Selma Klanten‡

†University of Wollongong, School of Science, Medicine and Health‡University of Technology Sydney, School of the EnvironmentContact e: mlh913@uowm ail.edu.auw: au.linkedin.com/in/martinhing | martyhing.wordpress.com

ConclusionsThe results of this study show that species of Gobiodon display a variety of social structures, from species such as G. brochus and G. axillaris which are rarely found in groups of more than 2 individuals, to G. citrinus and G. rivulatus which are often observed in groups of up to 12 individuals (pers. observation). The factors contributing to group formation (or dissolution) are not well understood in these species but these results indicate that coral size, Lα and habitat saturation are related to group size in only a few species. Other factors could therefore be influencing sociality in other species such as G. erythrospilus which are often observed in groups of 3 individuals. This suggests that it is unlikely that any single factor has been responsible for the evolution of sociality in these fishes. Cooperative breeding theory contains many other hypotheses which look at characteristics such as life-history and breeder tolerance which were not tested in this study. The study presented here forms one part of a larger project which will further examine the role that such factors have played in the evolution of sociality in these habitat specialist fishes. Watch this space….

Acknowledgements I would like to thank Kylie Brown, Grant Cameron, Kaz Hing and Marian Wong for their help in the field and my supervisors Marian Wong, Mark Dowton and Selma Klanten for their continued support.

References 1 Emlen, S. T. (1982) The evolution of helping 1. An ecological constraints model. American Naturalist 119, 29-39.2 Stacey, P. B. & Ligon, J. D. (1991) The benefits-of-philopatry hypothesis for the evolution of cooperative breeding - variation in territory quality and group-size effects. American Naturalist 137, 831-846.3 Arnold, K. E. & Owens, I. P. F. (1998) Cooperative breeding in birds: a comparative test of the life history hypothesis. Proceedings of the Royal Society B-Biological Sciences 265, 739-745.4 Roberts, R. L., Williams, J. R., Wang, A. K. & Carter, C. S. (1998) Cooperative breeding and monogamy in prairie voles: Influence of the sire and geographical variation. Animal Behaviour 55, 1131-1140.5 Purcell, J. (2011) Geographic patterns in the distribution of social systems in terrestrial arthropods. Biological Reviews 86, 475-491.6 Komdeur, J. (1992) Importance of habitat saturation and territory quality for evolution of cooperative breeding in the seychelles warbler. Nature 358, 493-495.7 Buston, P. (2003) Social hierarchies: Size and growth modification in clownfish. Nature 424, 145-146.8 Wong, M. Y. L. (2011) Group Size in Animal Societies: The Potential Role of Social and Ecological Limitations in the Group-Living Fish, Paragobiodon xanthosomus. Ethology 117, 638-644.9 Wong, M. Y. L. & Buston, P. M. (2013) Social Systems in Habitat-Specialist Reef Fishes: Key Concepts in Evolutionary Ecology. Bioscience 63, 453-463. H UNIVERSITY OF

WOLLONGONGAUSTRALIA

2.5

5.0

7.5

20 30 40

Coral Size (cm)

G. r

ivul

atus

gro

up s

ize

5

10

15

20

0.4 0.6 0.8 1.0

Proportion of inhabited corals

G. u

nico

lor g

roup

siz

e

Figure 1. Relationship between group size and coral size in G. rivulatus. Larger corals contained larger groups of G. rivulatus.

Figure 2. Predicted relationship between group size and habitat saturation in G. unicolor. Group sizes are larger when the degree of habitat saturation was lower.

Is group size constrained by coral size or Lα?Coral size was used in this study as a proxy for habitat quality. Three separate measures along different axes of the coral were taken and used to calculate a mean diameter for the coral. Fishes were removed from the coral and measured. Data were analysed using a general linear model with a log link function.

G. rivulatus and G. unicolor had significant relationships between group size and the main effects coral size and Lα, but the interaction Lα x coral size was not significant when tested together. G. erythrospilus, G. rivulatus and G. unicolor all appear to have significant relationships between their group size and host coral size (Figure 1). The relationship between group size and Lα was significant in G. rivulatus, G. unicolor and G. oculolineatus.

How social are coral gobies?There are 16 species of Gobiodon present at Lizard Island displaying a variety of social organisation (Table 1). Gobiodon species were located in their host corals and removed for identification. A sociality index for each species was calculated as the proportion of groups containing three or more individuals from the total number of groups observed.

Table 1. Mean group size ± SE and social index of each species of Gobiodon at Lizard Island. Bold text indicates species with mean group sizes greater than 2 indicating that the species is often found in groups larger than 2.

Is group size constrained by habitat saturation?To test the effects that habitat saturation had on the group size of Gobiodon species, a cross-transect was used to count the number of inhabited and uninhabited corals and their distance from a focal colony of Gobiodon. The proportion of inhabited corals on each transect was used as a proxy for habitat saturation.

Habitat saturation did not appear to affect the majority of species. Only G. unicolor showed a significant relationship between group size and habitat saturation with larger groups occurring at lower levels of habitat saturation (Figure 2).

IntroductionThe animal kingdom contains many examples of species, including our own, which form surprisingly complex social structures. Such immense variation in social structure is intriguing as it suggests that there may be underlying social, ecological or life history factors that influence the evolution of stable groups and their maintenance over many generations. One of the most fascinating aspects of sociality is the tendency of individuals to delay or forgo their own reproductive opportunities in order to join or remain within a group. The reasons for this decision are not universally clear despite being the focus of many behavioural studies. Cooperative breeding theory and its constituent hypotheses attempt to explain a social system in which reproductively mature individuals delay their own independent breeding in order to remain within a group as a non-breeding subordinate member. The ecological constraints1 and benefits of philopatry2 hypotheses look at ecological factors which could restrict individuals from dispersing or encourage them to remain on the natal territory. These factors have been well studied in terrestrial taxa such as birds3, mammals4 and arthropods5. Habitat quality and habitat saturation are thought to influence sociality in some of these vertebrate taxa2,6, but studies of some habitat specialist reef fishes show that colony sizes are limited by the size of the largest individual (Lα) in the colony, but not by the size of their respective habitats7,8. Habitat specialist marine fishes provide a novel system to test hypotheses of social evolution under different evolutionary pressures9. Testing entrenched hypotheses of social evolution under varying conditions is crucial to obtaining a general understanding of these systems. Here, I present an examination of the ecological constraints hypothesis using coral gobies, a group of fishes of the genus Gobiodon, at Lizard Island, Queensland. Gobiodon species are small fishes (2-4 cm) which reside in discrete patches of acroporid coral. They display a variety of social organisation within the genus and even within some species making them an ideal group to study evolutionary principles of social evolution.

Species Colonies Mean GS Sociality

G. acicularis 15 2.13 ± 0.22 0.27

G. axillaris 11 1.65 ± 0.10 0.00

G. bilineatus 2 2.27 ± 0.17 0.50

G. brochus 34 1.73 ± 0.06 0.00

G. ceramensis 19 1.67 ± 0.10 0.00

G. citrinus 18 4.32 ± 0.57 0.67

G. erythrospilus 70 1.85 ± 0.07 0.13

G. histrio 46 1.80 ± 0.05 0.11

G. oculolineatus 44 2.33 ± 0.16 0.20

G. okinawae 22 1.74 ± 0.17 0.14

G. quinquestrigatus 60 1.81 ± 0.05 0.07

G. rivulatus 71 2.33 ± 0.13 0.34

G. spilophthalmus 11 1.83 ± 0.21 0.18

G. sp A 14 2.12 ± 0.21 0.21

G. sp D 4 1.80 ± 0.20 0.00

G. unicolor 53 2.11 ± 0.16 0.21

Gobiodon sp. A

Gobiodon okinawae Cross-transect