CSIRO PUBLISHING Marine and Freshwater Research http://dx.doi.org/10.1071/MF13341 Population structure and biology of shortfin mako, Isurus oxyrinchus, in the south-west Indian Ocean J. C. GroeneveldA,E, G. Cliff B, S. F. J. DudleyC, A. J. FoulisA, J. SantosD and S. P. WintnerB AOceanographic Research Institute, PO Box 10712, Marine Parade 4056, Durban, South Africa. BKwaZulu-Natal Sharks Board, Private Bag 2, Umhlanga Rocks 4320, South Africa. CFisheries Management, Department of Agriculture, Forestry and Fisheries, Private Bag X2, Rogge Bay 8012, South Africa. DNorwegian College of Fishery Science, University of Tromsø, NO-9037, Tromsø, Norway. ECorresponding author. Email: [email protected] Abstract. The population structure, reproductive biology, age and growth, and diet of shortfin makos caught by pelagic longliners (2005–10) and bather protection nets (1978–2010) in the south-west Indian Ocean were investigated. The mean fork length (FL) of makos measured by observers on longliners targeting tuna, swordfish and sharks was similar, and decreased from east to west, with the smallest individuals occurring near the Agulhas Bank edge, June to November. Nearly all makos caught by longliners were immature, with equal sex ratio. Makos caught by bather protection nets were significantly larger, males were more frequent, and 93% of males and 55% of females were mature. Age was assessed from band counts of sectioned vertebrae, and a von Bertalanffy growth model fitted to sex-pooled length-at-age data predicted a À1 birth size (L0) of 90 cm, maximum FL (LN) of 285 cm and growth coefficient (k) of 0.113 y . Males matured at 190 cm FL, aged 7 years, and females at 250 cm, aged 15 years. Litter sizes ranged from nine to 14 pups, and the presence of gravid females in bather protection nets suggested that some pupping occurred in shelf waters. Teleosts (mainly Trachurus capensis) occurred in 84% of stomachs collected on longliners, whereas elasmobranchs (63.5%) were most common in samples collected from bather protection nets, followed by teleosts (43.1%) and cephalopods (36.5%). Larger prey size may be a factor that attracts large makos to coastal waters. Additional keywords: demography, GLM, pelagic longline fisheries, shark bycatch, stomach contents. Received 23 December 2013, accepted 1 May 2014, published online 24 September 2014 Introduction In the north-west (NW) Atlantic, juveniles migrate seasonally The shortfin mako (Isurus oxyrinchus Rafinesque, 1810, family between offshore wintering grounds and summer feeding Lamnidae), hereafter called mako, is a fast-swimming and grounds over the shelf (Casey and Kohler 1992). Mean lengths active shark with a wide distribution in tropical and temperate of makos caught by longliners in the North Atlantic and waters above 168C(Compagno 2001). It is a highly migratory Mediterranean differed by area and latitude (Buencuerpo et al. species that inhabits the epipelagic zone down to ,500 m depth 1998). Cliff et al. (1990) found more large males than females in and also enters littoral waters (Casey and Kohler 1992; Loefer coastal bather protection nets in eastern South Africa, and et al. 2005). Makos can reach 3.7 m fork length (FL) and a catches increased in winter and spring. The diffusive and highly weight of 600 kg, and they are sexually dimorphic, with females dispersive nature of mako populations is supported by the becoming longer and heavier than males (Compagno 2001). absence of clear genetic population structure (Heist et al. Ontogenetic and seasonal movements of makos give rise 1996; Schrey and Heist 2003). to demographically structured populations with high spatio- The reproductive mode of makos is aplacental viviparity, and temporal variability. Makos exhibit sexual segregation the embryos develop by ingesting unfertilised ova supplied by (Mucientes et al. 2009) and also segregate by developmental their mother, a practice known as oophagy (Gilmore 1993). stages (Nakano and Nagasawa 1996). Juvenile makos spend Information on mature and pregnant females is sparse (Semba 90% of their time in the mixed layer near the surface, whereas et al. 2011), but litter and birth size, size at sexual maturity and adults dive much deeper (Holts and Bedford 1993; Sepulveda parturition season have nevertheless been reported for popula- et al. 2004). Small immature makos observed near the coast of tions in several regions. Litter size mostly ranges from 4 to 26 Chile suggest the existence of a nearshore pupping and nursery embryos (Stevens 1983; Mollet et al. 2000) with a possible area in spring and summer (Bustamante and Bennett 2013). maximum of 30 (Mollet et al. 2002), and larger females produce Journal compilation Ó CSIRO 2014 www.publish.csiro.au/journals/mfr B Marine and Freshwater Research J. C. Groeneveld et al. more offspring. Parturition appears to occur year-round, with a saltatrix, making up .75% of the diet by volume (Stillwell peak in late winter and spring (Stevens 1983; Cliff et al. 1990; and Kohler 1982; Wood et al. 2009). Cliff et al. (1990) found a Duffy and Francis 2001; Semba et al. 2011). Size at birth is large proportion of elasmobranchs in the stomachs of makos around 60 cm fork length (FL), based on the size of near-term caught close to the shore in eastern South Africa. Large pelagic embryos (58–67 cm FL; Duffy and Francis 2001; Semba et al. fish and cephalopods were dominant in the stomachs of juvenile 2011) and the smallest observed free-swimming specimens and subadult makos sampled off southern Australia (Rogers (57–61 cm; Cliff et al. 1990; Mollet et al. 2000; Bustamante et al. 2012). and Bennett 2013). Makos are taken as by-catch in commercial longline and gill- Sexual maturity is reached at 2.5 to 2.9 m FL in female net fisheries that target tuna and swordfish, as well as in directed makos, based on measurements of reproductive organs, the shark fisheries and artisanal and recreational fisheries in tropical presence of fertilised ova and indications of past pregnancy or and temperate oceans (Francis et al. 2001; Campana et al. 2005; mating (Stevens 1983; Cliff et al. 1990; Mollet et al. 2000; Petersen et al. 2009; Bustamante and Bennett 2013). In the south- Francis and Duffy 2005; Joung and Hsu 2005; Bishop et al. west (SW) Indian Ocean, they are also taken as an incidental 2006). Males mature at a smaller FL of 1.7 to 2.0 m, based on the catch in bather protection nets, set in parallel and close to the calcification and development of claspers, which indicates the shore (Dudley and Cliff 2010; Cliff and Dudley 2011). Statistics capacity to copulate (Maia et al. 2007; Semba et al. 2011; of mako catches made by international fishing fleets in the SW Bustamante and Bennett 2013). Estimates may vary according Indian Ocean are collected by the Indian Ocean Tuna Commis- to sample size and the length range sampled, as well as the sion (IOTC), but records probably under-represent actual catches choice of maturity criteria and models (Francis and Duffy 2005; because of inaccurate or incomplete reporting (IOTC 2012). The Semba et al. 2011). Real geographical or between-hemisphere total reported mako catch from South Africa was 581 tonnes differences in length-at-maturity have been suggested (Mollet dressed weight in 2011 (DAFF 2012). Makos are characterised et al. 2000), but estimates may have been affected by different by low rates of population increase and high fishing mortality length estimation methods used by individual researchers throughout their range, and hence are considered to be vulnerable (Francis and Duffy 2005; Francis 2006). to overfishing (Dulvy et al. 2008; IUCN 2013). Previous studies on makos have mainly used counts of Data collected by fisheries observers stationed on pelagic alternately deposited calcified (opaque) and less-calcified longliners and from makos caught in bather protection nets were (translucent) bands in vertebral centra to estimate age (Pratt used to assess spatio-temporal size and sex distribution in the and Casey 1983; Bishop et al. 2006; Natanson et al. 2006; SW Indian Ocean. Biological information was used to estimate Semba et al. 2009 ; Wells et al. 2013). Annual deposition of age and growth, size at sexual maturity, and the number and size bands has been validated for many shark species, and is now of pups carried by pregnant females. Stomach contents were widely regarded as the norm (Francis et al. 2007). Nevertheless, used to compare the diets of makos caught near the shore in some uncertainty remains over band deposition rates in makos. bather protection nets with those caught in oceanic waters by Most studies on makos have demonstrated the deposition of a longliners. This study provides new information on makos from single annual band pair (Campana et al. 2002; Ribot-Carballal the SW Indian Ocean, a part of the world where they have been et al. 2005; Natanson et al. 2006; Semba et al. 2009), but some little studied, and as such it complements studies from other have shown biannual deposition (Pratt and Casey 1983), at least oceans. in young age classes where growth is rapid (Wells et al. 2013). Uncertainty over band deposition rates can affect estimates of Materials and methods growth rates, longevity and age-at-maturity. Apart from the Study area deposition rate, band widths in large makos become narrower and more compacted towards the periphery of the corpus The study area off South Africa extended from Cape Point 8 calcareum, and this may lead to underestimation of ages when (18 E) to the Mozambique border, and 200 nautical miles (nm) they become too narrow to distinguish, as found for porbeagle offshore to the boundary of the exclusive economic zone (EEZ) sharks (Francis et al. 2007). Various mechanical, chemical and (Fig. 1). This part of the SW Indian Ocean is characterised by a digital methods have been developed to enhance vertebral narrow and steep continental shelf, which gradually broadens to , bands, including the use of whole vertebrae or sections thereof, form the shallow ( 200 m deep) Agulhas Bank in the west.