The fin blue line : quantifying fishing mortality using shark fin morphology

Overfishing is a major global concern. Many of the worlds fish stocks are currently over exploited and require immediate action toward effective management and recovery strategies. Sharks are especially susceptible to overexploitation as they are generally slow growing, late maturing and produce few young. As large predators, sharks play an important, but poorly understood, role in marine food webs. As such, the ongoing exploitation of shark stocks is likely to cause detrimental and lasting ecological shifts within many marine systems. Within numerous fisheries, sharks are primarily targeted for their highly priced fins, and in many cases, they are the only body part retained by fishermen. This has created many issues for management as no practical methodologies currently exist to allow for the proper identification and quantification of individual species from fins alone. The high price of fin has resulted in an increased take of sharks, while also increasing the likelihood of illegal activity such as under-reporting and foreign fishing. Consequently, a large proportion of the total fishing mortality (from both commercial and illegal, unreported, and unregulated (IUU) fishing) appears to be unaccounted for, exemplified by an investigation of Australian shark fin export figures (Chapter 1). Confounding this, shark management receives low priority and limited funding. As a result, this has highlighted the immediate need for cost effective tools to quantify shark catch for both legal and illegal fisheries and, in the case of Australian fisheries, validate logbook data. Therefore, the major challenge is to develop cost effective methods for use in the field to identify sharks from fins alone, and to use these methods to generate data on catch composition. Morphological methods for identifying sharks from fins, if accurate, may be the most appropriate tool for such data collection. This premise is tested in this thesis; a major component is the development of methodologies to identify shark species from isolated fins. These techniques were then trialled successfully on specimens from illegal confiscated catch from northern Australian waters to demonstrate the applicability of these protocols for assessing the status of shark species. The majority of the methods investigated in the thesis rely on the analysis of shark fins from digital photographs. This is because digital images provide a cost effective and easy method to collect information about the morphological features of each specimen, and can be used both in field and lab situations. In order to justify the core methodologies used and to evaluate if robust methods could be developed, bias associated with this method were first investigated (Chapter 2). Fins can be wet (fresh) or in varying stages of dryness when identification is needed. As the majority (91.35%) of the confiscated IUU fins available to this study were wet, and there was a limited degree of drying in the foreign fishing vessel (FFV) catch, the identification protocols were developed using wet fins. In order to develop the identification protocols in Chapter 4, morphometric measurements, measured from digital images of the fin specimens, were used. On all fins, substantial changes in camera angle (from 0-20º) did not significantly affect any of the examined measurements. This result validated the use of a handheld camera as a practical tool for capturing images which are to be used for identify species of shark from isolated fins. Dermal denticles, (minute tooth-like structures which cover the body and fins of sharks) have been used as a tool for species identification of whole sharks in many shark taxonomic studies and species guides. Quantitative criteria were assessed in order to test the hypothesis that the morphological characters of the denticles on the dorsal and pectoral fins can be used to distinguish species (Chapter 3). These criteria described denticle crown variation at four specific areas on the dorsal and pectoral fins of 13 species of shark that are common to northern Australian waters. Skin samples from a total of 56 individuals from these 13 species were examined. All but three (Carcharhinus amblyrhynchoides, C. limbatus and C. tilstoni) could be distinguished from all other species investigated by the denticles at one or more areas using dorsal fins, and all but two (C. limbatus and C. tilstoni) using pectoral fins. Galeocerdo cuvier could be distinguished from all other species investigated at all areas on both dorsal and pectoral fins. The most useful area for dorsal and pectoral fins, in terms of percentage of species pairs distinguished (the proportion of all species pair combinations that could be differentiated) were identified. Using the character descriptions devised in Chapter 3, most species show differences in crown morphology at one area, or a combination of areas. Therefore, denticle crown morphology, when described using specific locations on the fin, provided an effective method of discriminating shark species from fins alone. Furthermore, denticles show markedly different crown morphologies with location on both the pectoral and dorsal fins, likely due to hydrodynamic and life-history adaptations. Therefore, when comparing denticles on the fin between adult specimens of different species, it is essential to specify the region that is used for comparison. While the use of dermal denticles to differentiate between species of shark may be effective, it is not always the most appropriate method for the field. Differences in denticle morphology are often subtle and require magnification to investigate, while more obvious visual characters may be used for species differentiation in the field, such as fin tip colour, fin colour or distance measurements. In order to investigate such alternative methods, distance measurements, fin tip colour and fin colour were used to develop a protocol to identify 35 shark species, found in northern Australian waters, from their isolated dorsal fins (Chapter 4). A series of discriminant analyses (DA) were conducted using distance measurement and RGB colour data on dorsal fin samples from 541 specimens of known species. These were subsequently used to predict the group (species) membership of 93 dorsal fin samples from the seized catch of IUU fishing boats. The accuracy of this method was then tested by comparison with molecular species identifications from the same dorsal fin. This validation demonstrated a correct classification of 80.4% of these specimens. Furthermore, to predict shark size from the identified dorsal fin, the relationship between shark total length (TL cm) and dorsal fin base length (B mm) was examined using linear regression to generate predictive equations for 35 shark species. Although a high level of accuracy was achieved, the complicated nature of the method resulted in an identification system that is not conducive to use insitu. The key to the future effectiveness of this method might be to incorporate measurements into an automated system (e.g. a computer program) that is applicable for easy use in the field. Ultimately, the goal of developing identification methods for species is to generate data with which to estimate exploitation levels in order to manage these resources sustainably. The denticle and DA identification methods from Chapters 3 and 4, were used to provide the first detailed account of both the number and biomass of sharks from the seized catch (as represented by dorsal fins) of 15 illegal foreign fishing vessels apprehended in northern Australian waters between February 2006 and July 2009. The catch of 13 small Indonesian and two large Taiwanese vessels was quantified, resulting in the identification of 1182 individual sharks with a total estimated biomass of 67.1 tonnes. The catch of the Indonesian fleet, as characterised by the 13 vessels, was mainly composed of smaller inshore and benthic species such as Spot-tail Sharks (Carcharhinus sorrah), Whitecheek Sharks (C. dussumieri) and juvenile Blacktip Sharks (C. limbatus/tilstoni). This species composition was similar to the reported catch from commercial shark fisheries in northern Australia. The Taiwanese fleet, as represented by two vessels, was characterised by a far greater catch of larger, pelagic species such as Blue Sharks (Prionace glauca), Silky Sharks (Carcharhinus falciformis), Oceanic Whitetip Sharks (C. longimanus), and Smooth Hammerheads (Sphyrna zygaena). The catch composition of these vessels was markedly different to the northern Australian commercial shark fishery, due to the fishing activity of these vessels occurring in deeper, offshore waters. Results show that IUU fishing in northern Australia is likely to have detrimental impacts on shark stocks in the region. The estimated level of illegal fishing for sharks by Indonesian vessels for the year 2006 is between 289.6 and 1071.04 tonnes, which is comparable to the largest commercial shark fishery that was operating in northern Australian waters at that time. One of the important distinctions of this assessment was to highlight the inadequacy of current methods, which assess illegal fishing impact based on the number of fishing vessels. In this study, a single Taiwanese vessel was found to be capable of removing the same amount of shark biomass as between 96 and 166 Indonesian vessels. As such, future assessments should include vessel characteristics (e.g. size, holding capacity) as large differences were highlighted both in terms of catch composition and volume of captured species. Ecosystem models often use broad functional groups of species to describe the structure and function of an ecosystem, and predict changes to those ecosystems. Furthermore, species from the same functional group generally exhibit similar morphology, as the ability to move is of crucial importance in many ecological contexts. Therefore, characterization of the morphology of the locomo...