Predicted suitable sites for NZ sea lions
IntroductionView a map of sites that have been predicted as suitable for New Zealand sea lion breeding colonies on mainland New Zealand.
On this page:
New Zealand sea lions are returning to the mainland and as populations grow, it's important to understand where they could recolonise and thrive. This information could help direct DOC activities such as monitoring and habitat restoration.
Researcher Veronica Frans identified 395 sites suitable for female New Zealand sea lions to create breeding colonies throughout mainland New Zealand.
This was a desk top study conducted from overseas. The next step is for DOC to assess how we ground truth the findings to increase the potential of this tool.
Each of these sites are highlighted on the map below in black. Aside from identifying these sites, other information about them has also been included. This other information relates to:
- suitability for inland movement
- features for protection or land restoration, for example, forest, tussock
- human impacts such as a site's proximity to roads
More detail of this information and how to interpret it is provided in the table below the map.
You can also find in-depth descriptions of each data fields in the publication 'Integrated SDM database: Enhancing the relevance and utility of species distribution models in conservation management.'
The map below is a large file and may take a moment to load. The map is 40 MB.
Use the table below to interpret the map's data fields. You can download an extended version of this table with more detail (PDF, 214K).
|Data field name
|Site identification tags
|The ID number that corresponds to the row number.
(site identification number)
|A three-letter and number-coded ID for each site.
|The DOC operational region where most of the site is located.
|Name of the NZ region where the site is located.
|North or South Island (includes Stewart Island).
|X centrepoint coordinates (UTM).
|Y centrepoint coordinates (UTM).
|Area of the site in km2 to prioritise or evaluate sites by size/capacity.
(S1 suitable area coverage (%)
|Percentage of a site suitable for the first behavioural state in the breeding season, S1 (breeding).
(S2 suitable area coverage (%))
|Percentage of a site suitable for the second behavioural state in the breeding season, S2 (transition).
(S3 suitable area coverage (%))
|Percentage of a site suitable for the third behavioural state in the breeding season, S3 (dispersion).
(multivariate environmental similarity surface grid mean value class)
|MESS class is the classification of the mean multivariate environmental similarity surface grid value. This reviews potential sites by measuring an area's predicted suitability. This is done by comparing how similar it is to the Auckland Islands.
Sites with a MESS class of none to low are most like the Auckland Islands. Intermediate to very strong MESS classes have qualities unlike the Auckland Islands, so there could be errors in the predictions.
Five classes describe extrapolation levels as follows:
1) none (values 0 - 100)
2) low (values -100 - 0)
3) intermediate (values -500 - -100)
4) strong (values -1000 - -500)
5) very strong (values -1600 - -1000)
(mode of the most dissimilar variable)
|Mode (most common) of the most dissimilar variable for a site, corresponding to the MESS grid. This determines the variable that is the most dissimilar from the training area (Auckland Islands).
(uncertainty for the S1 prediction (%))
|Coefficient of variation (CV) value in percent for the S1 breeding prediction. It estimates how far a prediction deviates from the average prediction.
This is used to gauge uncertainty in the model prediction. Relatively low percentages indicate low uncertainty more reliable predictions. Relatively high percentages indicate high uncertainty less reliable predictions.
(uncertainty for the S2 prediction (%))
|Coefficient of variation (CV) value in percent for the S2 transition prediction. See S1_uncrt for more detail.
(uncertainty for the S3 prediction (%))
|Coefficient of variation (CV) value in percent for the S3 dispersion prediction. See S1_uncrt for more detail.
(mode of limiting factor for the S1 projection)
|Mode (most common) of limiting factor or model variable for the S1 (breeding) projection.
This tests the variable that limits the suitability of a site for each behavioural state. If this variable’s values improve, then the site’s suitability can improve.
(mode of limiting factor for the S2 projection)
|Mode (most common) of limiting factor or model variable for the S2 (transition) projection. See S1_limit for more detail.
(mode of limiting factor for the S3 projection)
|Mode (most common) of limiting factor (or, model variable) for the S3 (dispersion) projection. See S1_limit for more detail.
(potential human impacts coverage (%))
|The percent of a site that faces potential human impacts. This is based on the multi-criteria decision analysis of 3D distances from sealed/unsealed roads and residential areas.
This examines potential NZSL interactions with humans to prioritise areas for community engagement and outreach.
(minimum sealed roads distance (3D; km))
|Minimum 3D path distance (in km) of a site from sealed (paved) roads.
This expands on the information from the multi-criteria decision analysis as vehicle collisions are a threat to NZSLs.
(minimum unsealed roads distance (3D; km))
|Minimum 3D path distance (in km) of a site from unsealed (unpaved) roads. See above entry for more details.
|This assesses the presence or absence of fences within a site. Presence of fences implies there is less suitable area available than predicted.
(grazing grasslands (%))
|Percentage of a site that has high/low producing grasslands for dairy and non-dairy grazing.
|Additional aspects of suitability
(minimum inland water distance (km))
|The minimum straight-line distance between a site and inland water bodies of lakes, ponds, streams. This variable is limited because it excludes inlets, which is important for thermoregulation.
(mean inland water distance (km))
|Mean straight-line distance between a site and inland water bodies of lakes, ponds, streams. This variable is limited as inlets are not included.
(maximum inland water distance (km))
|Maximum straight-line distance of a site from inland water bodies of lakes, ponds, streams. This variable is limited as inlets are not included.
(pine forest (%))
|The percentage of a site that contains planted pine (Pinus radiata) forest as of 2016. This accounts for additional, non-native forest type preferred by NZSL.
|Locations of interest
(current NZ sea lion sites)
|Names of known current sites where females and/or pups have been sighted during the breeding season within a 10 km straight line distance from a site.
(historic NZ sea lion sites)
|Names of known historic (archaeological) breeding sites that are within a 10km straight line distance from a site. This can be used as a reference if future actions lead to proactive conservation measures.
(DOC conservation area codes)
|List of DOC conservation area codes within a site to find if sites have other management priorities.
(DOC conservation area names)
|List of DOC conservation area names within a site.
(DOC conservation areas size (km2))
|Total size of DOC conservation areas within a site to assess how much of a site is already managed for other purposes.
(DOC conservation area coverage (%))
|Percentage of a site that is a DOC conservation area.
Veronica F. Frans, MSc.
Center for Systems Integration and Sustainability
Department of Fisheries and Wildlife
Michigan State University, USA
For any questions on the map or the data model: