Of the ~80 EPBC-listed Threatened and Migratory marine species known to occur in the North Marine Bioregion, 16 were identified as priority species through consultation with research end-users and experts. The priority group consisted of three sawfishes, two river sharks, Dugong, two inshore dolphins, six shorebirds and two turtles. Dwarf and then Green Sawfish had the most data gaps, indicating that these were the most poorly-known of the selected priority species in the North Marine Bioregion, and as such are a priority for research. These were followed (in order of data gaps) by the other river sharks and sawfishes, inshore dolphins, Hawksbill Turtle, Dugong, Olive Ridley Turtle, and shorebirds. Research assessing the relevance and impact of pressures was identified as a gap for all species. New data identified during the project can fill data gaps for all 16 species, and the analysis of these datasets can improve the accuracy of distribution maps, but new data collection is still required for all sharks and sawfishes, Hawksbill Turtle, and inshore dolphins to improve data coverage for distribution modelling and mapping. The gap analysis identified numerous new datasets, both published and unpublished, that are currently not incorporated into SPRAT profiles and distributions (see Table 5). This provided an opportunity to begin compiling and analysing this information to fill current data gaps, as well as identify targeted research needs for the future.
Due to time constraints, the completion of a knowledge gap analysis for the full list of Threatened and Migratory marine species in the North Marine Bioregion (~80 species) was not possible. After consultation with research end-users and project partners, the list of species was reduced to those considered as priority due to their Threatened EPBC Status, while retaining a diversity of taxa to guide future needs.
The main resources used by the DoEE to assess referrals under the EPBC Act are Species Recovery Plans and the Species Profile and Threats Database (SPRAT). We thus used the SPRAT profile and distribution maps as the main basis for the gap analysis. However, this approach was complicated by the reference list as provided on the SPRAT profile not being linked to the distribution maps. Thus, it is unclear which references (if any) relate to the distribution map. The species distribution maps are considered indicative only and in general combine the specific habitat type or geographic feature that contains observed locations of the species (known to occur), the suitable or preferred habitat occurring in close proximity to these locations (likely to occur); and the broad environmental envelope or geographic region that encompasses all areas that could provide habitat for the species (may occur) (DoEE). As the SPRAT database is not open access we requested access to the data in order to undertake the gap analysis for the 16 priority species but it was not able to be provided due to licensing restrictions. However, the internal DoEE high resolution distribution maps were provided as was as a spreadsheet with the names of the specific government departments, atlases, museums and conservation organizations that had contributed the data. Although this would theoretically have enabled us to identify where or if, there was new data not included in the SPRAT database, this information did not identify the original data sources (published/unpublished study or simple observations) and thus details of the nature and quality of the data were unknown.
The first part of the gap analysis process was to review the information in the SPRAT profile (using the SPRAT profile reference list) and the distribution maps for the North Marine Bioregion. As we could not access the data in the SPRAT database and had no information regarding its original source (as mentioned above), our assessment of the data behind the distribution maps was based on our understanding of the data using the information provided to us by DoEE and our understanding of the data generally available for these species (from the SPRAT profile reference list). A score was assigned against a range of categories according to the resolution (spatial, temporal and quality) of the data (high: 3; medium: 2; low: 1). We then summed the score in order to understand in relative terms how good the knowledge and data were for each species and then averaged and rounded the score for each species to provide an overall score of high (3), medium (2) or low (1).
The second part of the gap analysis set out to uncover what new data and information exists for the priority species to update SPRAT profiles and distributions and fill the data gaps identified by the above process. This consisted of a review of the peer-reviewed and grey literature using Google Scholar, enquiries to species experts, government departments, conservation organisations (e.g. Queensland Wader Study Group (QWSG), BirdLife Australia), industry contacts (e.g. INPEX, ConocoPhillips) and searching free, online data repositories (e.g. Zoatrack, eBird, ALA, Global Biodiversity Information Facility, Australian Ocean Data Network). The classification score for each of the categories in the table was then updated for each species by taking into account the new information and datasets identified.
The final step of the gap analysis was more quantitative. We contacted owners/custodians of the new georeferenced datasets identified and ask them to contribute to the project by sharing their data. New data obtained in this way, and from open access databases, was then plotted over the SPRAT high resolution distribution maps. Where the new data was not provided, either due to time or licensing constraints or a nil or negative reply, we attempted to simply place a point on the map where the study took place (obtained from the literature).
As we could not access the data in the SPRAT database nor comprehensive metadata for it, quantitatively assessing spatial gaps was not simply a comparison of the data used versus data available. In addition, the dataset that we compiled would contain data used in the SPRAT distribution map (i.e. not new data). Thus our approach consisted of gridding the area that contained the ‘new data’ (that compiled here) and the SPRAT high resolution distribution (0.1 degree grid cells). For each species we then calculated the proportion of grid cells in each occurrence category (known, likely and may) and in previously un-categorised grid cells (i.e. areas within the NMB that were not included as part of the species distribution in the SPRAT distribution maps) that contained at least one new data point. As only the ‘known’ category contains actual data points in the SPRAT distribution, any overlap of data in the ‘likely’ and ‘may’ categories was thus considered new data not yet included in the SPRAT distribution. To summarise where the new data came from, we defined five sub-regions within the North Marine Bioregion (Top End, Arnhem, Western Gulf, Southern Gulf, and Cape York; see Section 7.7) and for each species we calculated the proportion of grid cells in only the ‘likely’ and ‘may occur’ categories and in previously un-categorised grid cells that contained at least one new data point within each of those sub-regions. These results cannot be compared among species (only within) as the proportions are relative to the size of each species total distribution. The combination of the knowledge and spatial gap analyses allowed the identification of true gaps (no or limited data) for the priority species and assisted with recommendations to guide future research effort.