33 Chapter 5. Dynamics of rabbitﬁsh (Siganus sutor) spawning aggregations in southern Kenya Melita Samoilys, Nyaga Kanyange, Denis Macharia, George Waweru Maina and Jan Robinson Introduction Overﬁshing, destructive ﬁshing methods and weak governance are widespread and re-occuring problems in eastern Africa’s coastal ﬁsheries (McClanahan and Mangi 2004; UNEP 2009; Burke et al. 2011; Samoilys et al. 2011b), the eﬀects of which are further exacerbated by climate change (Schubert et al. 2006; Graham et al. 2007). e Shoemaker spinefoot rabbitﬁsh, Siganus sutor, is one of the most widely targeted and heavily ﬁshed species on the Kenyan coast (McClanahan and Mangi 2004; Maina et al. 2008; Samoilys et al. 2011b) and therefore likely a signiﬁcant species for food security in coastal communities (Foale et al. 2012). Studies have documented ﬁsher knowledge of reef ﬁsh spawning aggregations in the western Indian Ocean, including those of S. sutor, since 2006 through the World Conservation Union (IUCN) and the Western Indian Ocean Marine Science Association (WIOMSA) funded programmes in Kenya and Seychelles (Samoilys et al. 2006; Kimani 2007; Robinson et al. 2007; Samoilys et al. 2007). Concerns regarding the sustainability of ﬁsheries that target spawning aggregations (Sadovy and Domeier 2005; Robinson et al. 2011; Sadovy de Mitcheson and Colin 2012) highlight the need to determine the management implications of this ﬁshing (Grüss et al. in press). Major factors confound the study of spawning aggregations, notably their ephemeral nature, the remoteness of many sites and the cost and risks of diver-based studies on suﬃciently replicated time and spatial scales (Johannes et al. 1999; Colin 2012). Previous studies on S. sutor in Kenya have not adequately veriﬁed ﬁsher’s knowledge, aggregation sites have not been conﬁrmed and there is inadequate information on their spatial and temporal dynamics. We addressed these diﬃculties by employing four diﬀerent methods: (1) acoustic telemetry at spawning aggregation sites; (2) underwater visual census (UVC) surveys and observations of ﬁsh behaviour, (3) in situ observations of the aggregation ﬁshery; and (4) ﬁshers’ knowledge. is study was also part of a broader multi- disciplinary study on the S. sutor ﬁshery in the same area (see Chapters 3 and 4). e aim of the acoustic telemetry component was to obtain information on the spatial and temporal dynamics of aggregations than cannot be achieved by UVC alone and to estimate remote parameters such as individual site ﬁdelity and residency times which are critical to management development. While acoustic telemetry has been extensively used for studies of pelagic ﬁsh behaviour (e.g. Dagorn et al. 2007), the technology was under-utilised in the WIO in the context of reef ﬁsh resources prior to this WIOMSA-MASMA Programme (see Chapters 6 and 8). To verify that ﬁsh aggregations are reproductive requires the application of strict criteria (Colin et al. 2003; Domeier 2012; SCFRA database: http://www.scrfa.org/database). According to Domeier (2012), a spawning aggregation is deﬁned as “a repeated concentration of conspeciﬁc marine animals, gathered for the purpose of spawning, that is predictable in time and space. e density/number of individuals participating in a spawning aggregation is at least four times that found outside the aggregation. e spawning aggregation results in a masspoint source of oﬀspring”. Reproductive activity within spawning aggregations is deﬁned by either direct or indirect indicators. Direct indicators include observation of spawning, such as release of gametes in a spawning rush, or presence of hydrated eggs and/or post-ovulatory follicles in the gonads. Indirect indicators include observations of spawning-related behaviour, such as courtship, colour changes only known to be associated with reproduction and high catches of gravid ﬁshes conﬁrmed, for example, through increases in gonadosomatic index (GSI).