Dataset: Freshwater Fish Biodiversity Hotspots



This dataset and its metadata statement were supplied to the Bioregional Assessment Programme by a third party and are presented here as originally supplied.

Short Description

Freshwater fish sampling sites ranked using a scoring system based on species diversity and abundance, displayed as a weighted score which takes nearby sampling sites into account. Statistical significance is dependent on a site with a high (or low) Biodiversity score being surrounded by other sites with high (or low) scores as well.

This dataset has been provided to the BA Programme for use within the programme only. Third parties may request a copy of the data from DPI Water (previously known as the NSW Office of Water).


To identify regions of high fish biodiversity value in NSW rivers.

Dataset History

This dataset and its metadata statement were supplied to the Bioregional Assessment Programme by a third party and are presented here as originally supplied.

Detailed Description

A biodiversity scoring system was developed that accounts for both species richness (the number of different native species present) and each species' abundance. The scoring system identifies sites which support the highest abundance of the largest number of native fish species. Sites are then scaled to a percentage of the highest ranked site within the reporting region.

Numbers of fish caught from the catch dataset collected by Fisheries NSW using electrofishing at all sites that have been sampled in NSW sampled between 1 January 2002 and 31 December 2011 (a 10 year period) were standardised to catch per minute of electrofishing to account for variation in sampling effort across sites. Estuarine fishes sampled within coastal freshwater habitats were included as being native fish. Prior to analysis, sites were partitioned into meaningful groupings in order to ensure like sites were being compared. For the NSW-wide scale analysis, sites were stratified using the freshwater fish bioregionalisation model of Growns and West (2008), with an additional bioregion for sites within the Lake Eyre Basin - Bulloo catchments in northwest NSW. For CMA scale analysis, sites were stratified into Lowland (3 - 400 m) and Upland (401 - 1,780 m) altitude bands within each CMA area, noting that sites within the Lake Eyre Basin drainage division were analysed separately from the Murray-Darling Basin portion of the Western CMA area.

To calculate the Biodiversity score, each site within each group was ranked for each species in ascending order. Hence, sites where a species was absent had a rank of 0 and the site that had the highest abundance for that species had the highest rank. The ranks for each species were then summed across each sampling site to provide a 'Sum of Ranks'. The site with the highest 'Sum of Ranks' was identified in each zone and then the Biodiversity Score of each site was expressed as a percentage of the 'Sum of Ranks' for the most highly ranked site in each zone.

A cluster analysis was then undertaken on the data for each bioregion or altitude band using the Getis-Ord Hot Spot Analysis tool in ArcGIS (Fischer and Getis 2010). The Getis-Ord statistic (Z score) identifies whether features with either particularly high values or particularly low values tend to cluster in a given area. This tool works by looking at each feature within the context of neighbouring features. If a feature's value is high, and the values for all of its neighbouring features is also high, it is considered a part of a 'biodiversity hot spot'. The local sum for a feature and its neighbours is compared proportionally to the sum of all features; if the local sum is considerably different from the expected local sum, that difference is deemed to be too large to be the result of random chance. When this occurs, a statistically significant Z score is generated.

Suggested display schema

Use 'Quantities/Graduated Colours' on GiZ Score field.

Number of Classes = 8

Classification Method = Geometric Interval (provides weighting to distal part of the distribution curve)

Use a colour ramp from green to red with yellow intermediate (select HSV colour algorithm and full brightness on both black and white sliders)

Symbol size = 9

Flip symbols (ie. so that low values are red and high values are green)

No 'normalisation'

Edit 'label' text to delete numeric values and identify the most negative range as "Low Biodiversity" and the most positive range as "High Biodiversity"



Responsible Person

Dr Dean Gilligan

Senior Research Scientist

NSW Department of Primary Industries (Fisheries)


Fischer and Getis 2010: Handbook of Applied Spatial Analysis. Manfred M. Fischer and Arthur Getis. Baker & Taylor, January 2010

Growns and West 2008: Classification of aquatic bioregions through the use of distributional modelling of freshwater fish, : Ivor Growns, and Greg West. Ecological Modelling, Volume 217, Issues 1-2, 24 September 2008, Pages 79-86.

Dataset Citation

NSW Office of Water (2014) Freshwater Fish Biodiversity Hotspots. Bioregional Assessment Source Dataset. Viewed 18 July 2018,

General Information