Habitat associations of Freshwater Sawfish (Pristis microdon) and Northern River Sharks (Glyphis sp. C): including genetic analysis of P. microdon across nothern Australia: report to May 2008

About this document

This study investigated the ecology, morphology, habitat utilisation and population genetics of the vulnerable (Environment Protection and Biodiversity Conservation Act (EPBC) 1999) or critically endangered (IUCN) Freshwater Sawfish (Pristis microdon). It also examined the distribution and utility of satellite tags in tracking the movements of the endangered (EPBC Act 1999) or critically endangered (IUCN) Northern River Shark (Glyphis sp. C) in the Kimberley.

Section I of this project was a tagging-based investigation into the movements and habitat utilisation of the Freshwater Sawfish (Pristis microdon) and Northern River Shark (Glyphis sp. C) in the Kimberley region of Western Australia. In the case of P. microdon, between June and November 2007 (middle to late dry season) 38, nine, and two individuals from the Fitzroy River were tagged using Rototags, acoustic tags and satellite (SPOT) tags, respectively. Satellite tags were also fitted to two individuals of Glyphis sp. C from King Sound during this period.

The Fitzroy River is an important nursery area for P. microdon, with all individuals recorded from the river being immature. During 2007 there was an extremely high level of recruitment of the species into this river. The presence of a large number of small individuals with umbilical scars, i.e. new recruits, in the lower reaches of the river (estuary) during June suggests that the females may pup in the vicinity of the river mouth. A significant correlation between the strength of recruitment and high water levels during the late wet (i.e. April) between 2002 and 2007 was evident; suggesting that pupping may occur in the late wet or that either releasing their litter at this time may provide offspring with more favourable conditions, or that a longer wet season facilitates higher survivorship of offspring. A later or longer wet season may allow small individuals to migrate upstream relatively unimpeded to safer (fewer predators) freshwater pools for a protracted period of time. It is predicted that there might be some synchronisation in the timing of parturition and/or an initial period of high site fidelity among new recruits prior to their upstream migration. In any case, presumably any propensity of new recruits to travel in groups, would increase an individual's chance of avoiding predation on their journey up river.

The use of acoustic tracking demonstrated a high degree of habitat partitioning between different age classes of P. microdon, with the new recruits (0+ fish) clearly spending most of their time in shallow water (<0.6 m) compared to the larger 1+ individuals that mainly occurred in waters deeper than 0.6 m. Following Simpfendorfer (2006), who observed similar behaviour in Pristis pectinata, possible reasons for the prevalence of smaller individuals in shallow water include: (i) predator avoidance, noting that smaller individuals are potentially more susceptible to predation and (ii) maximisation of growth rate due to the relatively warm water temperatures in the shallows. Alternatively, the occurrence of the larger individuals in the slightly deeper water allows the animal more space to manoeuvre while also maintaining a close proximity to potential prey. The +1 individuals also showed a marked tendency to occur at deeper depths by day compared to night. The ontogenetic and diurnal differences in habitat utilisation by different age classes of P. microdon are also likely to be reflected in differences in diet and could, in fact, be partly driven by foraging strategies related to prey availability. For example, the prey that is available in the shallows differs significantly between the day and night, with the known prey of larger sawfish being more abundant in the shallows at night compared to the day. The diversity of potential prey in the shallows also increases substantially at night. If small P. microdon are feeding in the shallows during the day then their diet is likely to be different to older individuals that occupy different depths and primarily venture into the shallows at night.

The acoustic data also demonstrated that small individuals of P. microdon in the estuarine reaches of the Fitzroy River readily move between pools, even though most pools at low tide are separated by long stretches of shallow waters. Furthermore, on the incoming tides, ~98% of the movements of the 0+ sawfish between pools was in an upstream direction, i.e. they moved with the tide. In contrast, the 1+ sawfish moved to another pool only when tidal waters reached the site and this movement was both with the tide (50% of the time) and against the tide (50% of the time). The ability to swim between pools and utilise the shallow runs and riffle zones between both tidally influenced and riverine pools is potentially beneficial because it allows the 0+ fish to avoid deeper bodied predators and also to forage in areas not being exploited by larger fishes. It also allows the new recruits to continue to migrate upstream relatively unimpeded until late into the dry season; to at least the Barrage, a substantial unnatural barrier, in Fitzroy River. While there was no obvious change in the behaviour of 0+ sawfish during the various moon phases, 1+ P. microdon typically utilised very shallow waters (<20 cm) only during the full moon, and water depths greater than 1.5 m only during the new or half moon. The greater visibility at night during the full moon may make them more vulnerable to deep water predators (e.g. Bull Sharks and Estuarine Crocodiles), or may make their shallow water prey more visible. Alternatively, they may follow prey species that move shallower, perhaps in an effort to avoid predators during this time.

Satellite tracking of P. microdon using Wildlife Computer's Smart Position or Temperature Transmitting (SPOT) tags was not successful, possibly due to animals not remaining in the shallows for enough time to allow for the calculation of an accurate location. In this regard, it is relevant that the two individuals with the SPOT tags were both relatively large (1780 and 2580 mm TL), which, on the basis of the acoustic tracking data, may rarely go within 0.5 m of the surface. Nevertheless, satellite tags in general still have the potential to provide important information on the movement patterns and habitats of individuals of P. microdon that leave the river. Pop-off archival tags may be a more effective tool to examine this aspect of their life-history because the collection of data via these tags is not dependent on the tagged individuals spending significant amounts of time at the surface.

The capture of large numbers of 0+ individuals of P. microdon from the Fitzroy River during 2006 and 2007 allowed for an examination of natal sex ratios. Males and females occurred in equal proportions in these individuals. In contrast, the sex ratio of older (i.e. >1+) fish was skewed towards females. The females also attained a considerably greater size than males whilst in the river. This suggests that either the males are more heavily predated on, which seems unlikely unless growth rates are slower, or that males mature at a smaller size and younger age and so leave the river earlier. Males appear to leave the river at ~2400 mm TL and females at ~2800 mm TL; presumably prior to attaining maturity.

There were notable differences in the rostral teeth counts between P. microdon from Western Australia (Fitzroy River) and the Gulf of Carpentaria in Queensland (Qld), albeit that some of these differences may be attributable to low sample sizes from Qld. Importantly, although the analysis included a considerably smaller number of males from Qld (n = 14) than from Western Australia (n = 47), the former exhibited a greater variability in left rostral teeth (18 to 24) compared to the latter (19 to 23). The range in the number of left rostral teeth in females was narrower in Qld samples (17 to 21) compared to the Western Australian samples (17 to 22), and may be attributable to differences in sample sizes.

Six individuals of Glyphis sp. C were captured in marine tidal creeks of King Sound, ranging from Doctors Creek to the north-eastern side of the Robinson River estuary. This extends the known distribution of the species in Western Australia by approximately 60 km to the north and east. Although SPOT tags proved slightly more effective with Glyphis sp. C than P. microdon, too few transmissions were made to calculate a location. Most of the transmissions occurred in the early morning and on an incoming tide, which may indicate either peak foraging times for the species or that the species moves over shallows during this period to move in with the tide. Morphometric data increased the previously known extent of morphological variation in Glyphis sp. C. For example, the ratio of the height of the 2nd dorsal fin to the 1st dorsal fin in each of two individuals (1088 and 1237 mm TL) were determined as 0.57 and 0.67, which are both marginally outside the reported range for the species (cf. 0.58-0.66).

Section II of the project investigated the genetic diversity and population structure of P. microdon in Australian waters. It was based upon the analysis of variation in the nucleotide sequence of a 353- 351-bp portion of the control region of the mitochondrial genome. Samples of P. microdon were obtained from a total of 90 individuals from across the entire Australian range of this species, but were mainly from the Fitzroy River on the west coast of Australia (n = 37) and from a number of sites within the Gulf of Carpentaria and associated rivers (n = 47) on the north coast. In order to help place the results for P. microdon into perspective, comparable genetic data were produced for a total of 30 individuals of the Dwarf Sawfish (Pristis clavata), from several sites along the west coast and in the Gulf of Carpentaria and associated rivers.

The results indicate that P. microdon contains a 'healthy' amount of genetic diversity over a range of spatial scales in Australian waters. This was indicated by the fact that the haplotype and nucleotide diversities in the samples of this species were not unusually low either in absolute terms or in comparison with those of other elasmobranchs, including P. clavata. Since the evolutionary potential of a population is proportional to the amount of adaptively significant genetic variation therein, and assuming that control region diversity provides a reflection of genome-wide diversity in P. microdon, this finding is encouraging regarding the prospects for the long-term survival of P. microdon in Australian waters. However, there is a note of caution in that it can take several generations for a decline in the genetic diversity of a long-lived species with overlapping generations, like P. microdon, to become evident. Furthermore, most of the control region diversity in P. microdon was present in rare haplotypes (i.e. in rare alleles), which would be highly susceptible to loss via genetic drift if the abundance of this species should decline in the future. Thus, in order to conserve current levels of genetic diversity, it is vital to develop management plans for P. microdon such that its current population sizes are maintained.

The analysis of the population structure of P. microdon in Australian waters was focused on the Fitzroy River on the north-west coast of Western Australia and in the Gulf of Carpentaria and associated rivers in Queensland. The results indicate that P. microdon is genetically sub-divided between these two locations. This finding is significant because it almost certainly means that the demographics of the assemblages of this species in these two locations are independent of each other. Thus, these assemblages should be regarded as separate management units (populations). However, the smallest spatial scale at which demographically independent units occur in P. microdon in each of these locations could not be resolved. Hence, it is possible that all of the P. microdon in the Gulf of Carpentaria region belong to a single population, as might all of the individuals of this species on the west coast. Alternatively, this species may typically comprise multiple populations within each of these geographic regions. Since they are demographically independent of each other, the demise of the assemblage of P. microdon in either one of these locations will not be offset by immigration from the other location. Thus, the conservation of the assemblages in both the Fitzroy River and in the Gulf of Carpentaria region should be a high priority. In addition, data relating to the demographics (e.g. abundance, mortality, recruitment) of P. microdon should be collected and interpreted at the level of individual populations. Additional work is required to elucidate the relationships between these assemblages and those in intervening waters in the Northern Territory and also those on the east coast of Australia. The above conclusions about the population structure of P. microdon are based upon the patterns of variation in a mtDNA marker (which reflects female-mediated gene flow) and on the assumption that the patterns of patterns of gene flow in males are broadly similar to those of the females. In this regard, it is important to recognise that P. microdon might exhibit sex-biased dispersal with female philopatry and male dispersal, in which case the above conclusions would apply to the female component of populations only.

Finally, analysis of the evolutionary relationships among the control region haplotypes of P. microdon indicates that the amount of evolutionary divergence in this species in Australian waters is moderate. There does not appear to be any particular subsets of individuals or population that require special conservation priorities on their basis of their distinctive evolutionary histories.