4. Description of individual GIS data layers
This section provides a breakdown summary of the all the GIS data provided to Ministry for the Environment as part of this project. Each GIS layer or layer group is listed in bold, along with the shape file name and the following descriptors (where known):
- description of what information the layer contains
- classifications of data (where relevant)
- data source
- data age
- geographic coverage of the Catchment
- any caveats on use of the data
- related layers
- ecological values (for ecological layers).
4.1 Background GIS layers
Topographical maps: NZMS_262_Topo
Description: New Zealand Map Series (NZMS) 262 1:250,000 topographical map.
Classes: None.
Data source: Land Information New Zealand.
Data age: Unknown.
Coverage: Entire catchment.
Caveats: None.
Related: Rivers_250 layer.
Waitaki Catchment outline: Waitaki_Catch
Description: Outline of the Waitaki River Catchment.
Classes: None.
Data source: Ministry for the Environment.
Data age: Unknown.
Coverage: Entire Catchment.
Caveats: None.
Related: None.
Hill shading relief: Hillshade
Description: Hill shading from 1:250,000 GIS data. Used as a base layer for all ecological GIS plots.
Classes: None.
Data source: Land Information New Zealand.
Data age: Unknown.
Coverage: Entire Catchment.
Caveats: None.
Related: NZMS_262_Topo.
River Environment Classification: REC
Description: Streamlines classified according to the River Ecosystem Classification (REC).
Classes: Length of stream reach (m), stream order, climate, source of flow, geology, landcover, network position, valley landform.
Data source: REC from NIWA. Details of use of the REC can be found in Snelder et al. (2004).
Data age: Recent - 2004.
Coverage: Coverage of all Catchment.
Caveats: Does not include lakes and excludes a number of small tributaries shown on 1:50,000 topographical maps.
Related: REC_Fish species layers and REC_highorder layer.
REC excluding small streams: REC_highorder
Description: REC layers, excluding first and second order streams (smallest streams). Good river base layer. Used as a base layer for all ecological GIS plots.
Classes: Length of stream reach (m), stream order, climate, source of flow, geology, landcover, network position, valley landform.
Data source: REC from NIWA. Details of use of the REC can be found in Snelder et al. (2004).
Data age: Recent - 2004.
Coverage: Coverage of all Catchment.
Caveats: Does not include lakes and excludes a number of small tributaries shown on 1:50,000 topographical maps. Excludes REC first and second.
Related: REC_Fish species layers and REC.
Lakes
Description: Polygons showing location of lakes. Used as a base layer for all ecological GIS plots.
Classes: Lake name, lake area.
Data source: Environment Canterbury "Lakes" layer.
Data age: Unknown.
Coverage: Major lakes shown on NZMS 262 series maps.
Caveats: Some small unnamed lakes not shown.
Related: NZMS_262_Topo.
Water permits from Otago and Canterbury: Consents
Description: Records and associated information for water permits and certificates of compliance within the Catchment.
Classes: Source (water permit or certificate of compliance), record type (eg, existing or new application), state of application, client name and address, compliance monitoring details, water use, maximum consented rate of abstraction, sub-catchment.
Data source: Environment Canterbury and Otago Regional Council Consents database.
Data age: September - 2004.
Coverage: There is coverage of the whole Catchment, but water permits are concentrated in the lower catchment, especially the Waitaki River mainstem and the Hakataramea River. There are also numerous records around the Twizel-Ahuriri-Ohau area, and tributaries of the Tekapo River.
Caveats: Limited to coverage of Otago Regional Council and Environment Canterbury Consents database in September 2004; more recent applications will not be shown.
Related: None in particular, but of interest in relation to proximity of GIS layers showing ecological values.
4.2 Ecological data
Department of Conservation provided conservation status and threatened species data of from their national "Bioweb" database. Bioweb is a database administered by Department of Conservation that holds data about New Zealand's natural and historic heritage that is of importance to Department of Conservation. The following caveats were provided by Deb Zanders (Department of Conservation Canterbury) and are relevant to all the Lizards, Threat_Plants, and Birds_Pt DoC layers that follow:
- data was extracted from the Bioweb database in October 2004
- non 7-figure NZMS grid references were excluded (by Department of Conservation), because of the extra time required to convert them
- Bioweb was still being populated at the time of writing and so the data does not include information from Tenure Review inspections in Canterbury and includes only some Otago records
- fish, rock wren, and plant data is considered to have reasonable coverage
- invertebrate and some bird data is the weakest, and there is no data for many threatened species
- within pastoral leases that have not been field-inspected by Department of Conservation, an absence of threatened species may be misleading
- the data represents the best at the time of writing, and was drawn together from a variety of databases (some not held by Department of Conservation).
Geckos and skinks: Lizards
Description: Combined lizard (herpetofauna) data records from Otago and Canterbury.
Classes: Includes threat status, locality, habitat and collector notes.
Data source: Department of Conservation Bioweb.
Data age: Unknown.
Coverage: Best coverage around Twizel, Lakes Pukaki, Tekapo and Benmore, and the upper reaches of the Otematata River (tributary of Lake Aviemore). Few records form the lower catchment, especially the Hakataramea River and Waitaki River mainstem.
Caveats: Records do not include sites where lizards were searched for and not found.
Related: None.
Values: Many occur in riparian habitats and there are many threatened species.
Threatened native plants: Threat_Plants
Description: Threatened plants collected by Department of Conservation botanists.
Classes: Species names, threat status.
Data source: Department of Conservation "Biotheme" GIS data layer.
Additional data from Otago and Southland pastoral lease inspections.
Data age: Obtained from Department of Conservation in September 2004; variable original data age.
Coverage: Good coverage around the lakes and the tributaries of Lake Ohau. Patchy for the remainder of the upper catchment. Little data for the lower catchment.
Caveats: Records do not include sites where plants were searched for and not found.
Multiple species records occur on individual points.
Related: None.
Values: All records are for threatened species, suggesting high ecological value.
Weed plant species: Weeds_Poly and Weeds_Pt
Description: Areas (Weeds_Poly) and points (Weeds_Pt) where weed pest plant species have been recorded by Department of Conservation.
Classes: Species names, abundance, area (for Weeds_Poly).
Data source: Department of Conservation "Bioweed" GIS data layers.
Data age: Obtained from Department of Conservation in September 2004; variable original data age.
Coverage: Very sparse and patchy data, and no data for the lower catchment.
Caveats: Records do not include sites where plants were searched for and not found.
Multiple species records occur on individual points.
Related: None.
Values: Weed species may exclude native species.
Areas of Department of Conservation Natural Significance: Nat_Signif
Description: A compilation of Department of Conservation data, including historic Recommended Areas for Protection (RAP), Wetlands of Ecological and Representative Interest (WERI) and Special Sites of Wildlife Importance (SSWI) data. In polygons, rather than point format.
Classes: Area (m2 and ha), name of area, Department of Conservation identifier, ecological district, extensive descriptive notes on ecological values.
Data source: Department of Conservation "Nat_sig" GIS data layer.
Data age: Obtained from Department of Conservation in September 2004. Original data ranges from 1980s to 2000s.
Coverage: Extensive coverage throughout the Catchment.
Caveats: Unknown data age.
Related: Overlap with numerous ecology layers.
Values: All areas of natural significance shown are of high ecological value.
Threatened birds: Birds_Pt
Description: Locations of threatened bird sightings from Department of Conservation and Ornithological Society staff.
Classes: Species name, threat status, Department of Conservation source of data.
Data source: Department of Conservation "Biotheme" GIS data layer.
Data age: Obtained from Department of Conservation in September 2004; variable original data age
Coverage: Extensive coverage throughout the Catchment.
Caveats: Due to the mobility of birds, this point data layer is not as useful as the Birds_Poly layer.
Mohua sightings and some other sightings were excluded, as they lacked point data.
Related: Birds_Poly layer.
Values: All records are for threatened species, indicating high ecological value.
Wetland bird habitat: Birds_Poly
Description: Polygons showing the locations and significance of indigenous bird habitats of rivers and open waters.
Classes: Location name, water body type, wetland bird guilds and their key habitats, threatened species and their key habitats, habitat ranking and significance, threats (actual and potential), information sources, comments, data quality.
Data source: Environment Canterbury Report U00/37 (O'Donnell, 2000) and associated Environment Canterbury GIS data.
Data age: Variable age of original sources, although Department of Conservation still consider this reference and GIS data to have good coverage of information for the area.
Coverage: Excellent coverage of lakes and mainstem of major rivers. Poor coverage of tributaries.
Caveats: None.
Related: Birds_Pt layer.
Values: All locations denote areas of high ecological value, mainly due to the presence of habitat for threatened bird species.
Wetland vegetation: NatVeg_Poly
Description: Polygons showing the locations of significant indigenous vegetation associated with water bodies.
Classes: Name of water body and ecological district, water type, altitude, presence/absence of 10 plant community classes, presence/absence of 508 listed plant species, presence/absence of threatened and uncommon plants, criteria for conservation significance, threats, reserve status, information sources, weed presence/absence.
Data source: Environment Canterbury Report U01/45 (Allen, 2001) and associated Environment Canterbury GIS data.
Data age: 1972 - 2000.
Coverage: Restricted to Waitaki mainstem (from Lake Waitaki downstream), Ahuriri River and delta, and the Tekapo River.
Caveats: None.
Related: Threat_Plants layer.
Values: All records are for threatened species, indicating high ecological value.
Salmonid habitat: Salmonid_Poly
Description: Polygons with data fields describing salmonid habitat values and threats.
Classes: Name of water body, salmonid habitats present (for each of six salmonid species), salmonid habitat value (low, medium, high), threats (actual and potential), data source and age, site comments.
Data source: Environment Canterbury Report U00/31 (Langlands and Elley, 2000) and associated Environment Canterbury GIS data.
Data age: 1980s - 2000.
Coverage: Excellent coverage of lakes and mainstem of major rivers. Poor coverage of tributaries.
Caveats: None.
Related: Fish layer.
Values: Ecological values are determined by selecting appropriate fields, the presence of the layer in an area does not automatically infer high values.
Freshwater fish: Fish
Description: New Zealand Freshwater Fish Database (NZFFD) records, with additional information added.
Classes: Fish species collected, NZFFD card number, catchment and locality name, native (vs introduced or marine wanderer), migratory status (diadromous, non-diadromous, marine wanderer), salmonid (vs non-salmonid), threat status and category, plus numerous NZFFD attributes.
Data source: NZFFD search of Waitaki Catchment records in October 2004.
Data courtesy of the National Institute of Water and Atmospheric Research (NIWA).
Summary fields (eg, threat status) added by Kingett Mitchell.
Data age: Based on a search of the database in October 2004.
Individual records range from 1952 to 2004.
Coverage: Excellent coverage throughout catchment.
Probably best coverage of all the freshwater ecological data available.
Caveats: NZFFD is provided by NIWA on the assumption that it is for non-commercial use.
Related: REC, REC_Fish and Salmonid layers.
Values: Not all fish species necessarily have high ecological value. However, ecological values can be determined by looking at fields such as threat status.
Caddisflies (Trichoptera): Caddisflies
Description: New Zealand Caddisfly (freshwater invertebrates in the order Trichoptera) Database records.
Classes: Species collected, altitude, collector, location of type specimens, ecoregion, locality, number of adults, larvae, and pupae collected.
Data source: Database held by John Ward (Canterbury Museum).
Data age: Last updated in May 2004.
Coverage: Sparse catchment-wide coverage.
Caveats: None.
Related: SoE_Inverts layer.
Values: Many larval caddisflies (which are aquatic) are sensitive to pollution, plus they are important food for many freshwater fish species.
Freshwater invertebrates: SoE_Inverts
Description: Freshwater invertebrate values data.
Classes: Number of taxa collected, percentage abundance comprised of pollution-sensitive Ephemeroptera, Plecoptera, and Trichoptera (%EPT), percent dominance of the most abundant taxa (a diversity measure), site name, comments, land use.
Data source: Raw invertebrate data provided by Environment Canterbury, collected as part of their regional State of the Environment (SoE) monitoring.
Invertebrate summary data calculated by Kingett Mitchell.
Data age: Field sampling from 1999 to 2003.
Coverage: 24 records from a good range of localities throughout the Catchment. Good coverage in areas where most water permit records occur.
Caveats: None.
Related: Caddisfly layer.
Values: A combination of high %EPT and high taxa richness often indicates high ecological values. Freshwater invertebrates are also important fish food.
Modeled fish locations: REC _ Fish genus and species
Description: REC streamlines classified according to the likelihood of threatened native fish (four species layers) and introduced sports fish (three species layers) occurring there.
Classes: Likelihood of selected threatened and sports fish occurring, expressed as a percentage (and rank) of all NZFFD records falling in REC classes.
Data source: Created by Kingett Mitchell in October 2004, by combining NZFFD data (see Fish layer above) and REC data. Details of this process are described below.
Data age: Recent - 2004.
Coverage: Coverage depends on the species (see Figs. 11 and 12 for examples). Purpose of the layer was to extrapolate fish values throughout the Catchment.
Caveats: Excluded fish records from lakes and where it was difficult to assign a fish record to an REC reach (eg, records >100m from an REC reach that fell between differing REC classes).
This layer was a trial and has not been field-validated. The values are intended to provide an indication of the likelihood of fish occurring - numerous factors other than the coarse REC classes used could influence fish distribution.
Related: REC and REC_highorder layers
Values: High percentage likelihood of species occurrence indicates there is potential for high fish values in the reach.
4.3 Modelling freshwater fish values using the REC
4.3.1 Background
The REC contains a great deal of environmental data of relevance to freshwater biological communities. For example, the REC can be used to determine the proportion of the catchment with pasture, forest, and urban land cover for each stream reach, and it is well understood that land cover and river habitat quality are closely related (Quinn et al, 1997). Most recently, Joy and Death (2004) combined REC data with additional habitat data and freshwater records to predict fish species occurrence in the Wellington region. Freshwater fish are good species to model, because of the wealth of location and habitat data stored in the NZFFD. The value of ecological modeling is that information can be extrapolated from areas where data is present into areas where data is sparse.
In an attempt to improve the geographic spread of freshwater ecological values in the Waitaki Catchment, Kingett Mitchell created new REC layers with freshwater fish attributes. The REC fish layers were limited to five threatened species and the three most common salmonid sports fish in the catchment. These species were chosen because of their conservation and sports fishery values, respectively. Lamprey (Geotria australis) and Canterbury mudfish (Neochanna burrowsius) have a threatened status but were excluded from the REC analysis, as they had too few records in the Catchment to be able to reliably extrapolate data from. It is worth noting that sampling by Department of Conservation may increase the number of Canterbury mudfish records in the near future (Leanne O'Brien, University of Canterbury, pers comm.).
What follows below is a description of how the REC_Fish layers were created and a brief discussion of the results of the modeling and use of the data.
4.3.2 Methods
- REC reach attributes were joined to the nearest NZFFD record, using a 'spatial join' GIS grouping command.
- All lake records were removed, as the REC is only for rivers. This required some checking of original NZFFD cards to confirm whether fish records were from lakes or rivers.
- For all NZFFD records >100 m from an REC reach, points were excluded if it was unclear which REC reach to assign them to (eg, if the record fell between two REC reaches with different classes). For clarity, we refer to these excluded records as 'ambiguous' points.
- The remaining NZFFD records with REC attributes were used for the next stage of extrapolating throughout the Catchment. The total number of NZFFD records (cards) for the Catchment dropped from 712 to 597, and individual species records dropped from 1851 to 1544 after this data editing stage. Thus, 83% of original NZFFD data remained, which is a reasonable proportion. Between 78 and 100% of records remained for the seven 'target' fish species being used for subsequent modeling.
- For each of the seven target fish species, record counts were tabulated and plotted against all permutations of the climate, source-of-flow, geology, and land-cover REC categories (see example figures attached). The percent frequency and rank occurrence of each target fish species in each of the REC classes was then calculated.
- An REC fish layer was then created for each target fish species by adding a field to the REC layer stating the percent occurrence of the fish species within the REC category. REC categories sharing the same percent fish occurrence value were given the same classification in the percent occurrence field. Where the number of percent occurrence classifications exceeded 10, they were truncated to 10.
4.3.3 Results
Fig. 11 and Fig. 12 are sample GIS plots showing the REC fish data for chinook salmon and longfin eels, respectively, overlain with actual NZFFD records for each species. These layers illustrate how chinook salmon records, and by inference their likelihood of occurrence (displayed using the REC_Oncorhynchus_tshawytscha layer), are mostly restricted to the lower catchment (Fig. 11). This reflects the effect of the hydroelectric dams on upstream salmonid migration.
The REC fish layers will be most reliable for those fish species with the greatest number of records, due to the greater REC environmental data and spatial data available. Thus, while REC predictions for rainbow trout (Oncorhynchus mykiss) are considered reasonable, records for less common threatened galaxiid species should be treated with caution and regarded as being indicative only.
Initial examination of the data indicates that the REC fish layers may help fill gaps where ecological data is sparse, thus fulfilling the objective of the exercise. However, there has been no field validation of the data, and therefore the absence of a high percentage REC fish ranking in a given area will not necessarily mean the species will not be found there. Field data would still need to be collected, or local experts consulted, to fill ecological data gaps in areas where water demands are high.
