MATERIALS AND METHODS

21.2.1 STUDY AREA

The study area is the Avon Heathcote Estuary (Ihutai) located at 43.5°S, 172.7°E in the city of Christchurch (Fig. 21.1). The estuary is located between the Waimakariri River and the southern end of a large sandy bay (Pegasus Bay) where it is a prominent local feature (Kirk, 1979). It is a barrier enclosed tidal lagoon type estuary (Hume et ah, 2007) with high ecological and social values including cultural significance for Maori (Jolly and Nga Papatipu Runanga Working Group, 2013; Lang et ah, 2012).

The Avon and Heathcote are the two major rivers of the estuarine system, both of which provide G. maculates spawning habitat. These are spring- fed lowland rivers waterways with average base flows of approx. 2 and 1 cumecs, respectively (White et ah, 2007). They are also among the most well-studied spawning locations in New Zealand with surveys having been conducted periodically since 1988 (Taylor et ah, 1992).

21.2.2 GEOSPATIAL ANALYSES

We analyzed spawning site data from postearthquake studies comprising of seven independent surveys conducted over two years during the peak spawning months using a census-survey methodology designed to detect all spawning in the catchment (Orchard and Hickford, 2018). The areas surveyed were approximately 4 km reaching in each river extending from the saltmarsh vegetation zone (downstream) to 500 m upstream of the inland limit of saltwater (Fig. 21.1). The dataset of 188 records provided details of 121 spawning occurrences in the Avon and 67 in the Heathcote. Each record included upstream and downstream coordinates of the spawning site, mean width of the egg band, and area of occupancy (AOO) of eggs with each site being defined as a continuous or semicontinuous patch of eggs.

Location of postearthquake survey areas for G. maculatus spawning habitat in the Avon and Heathcote River catchments, city of Christchurch, New Zealand

FIGURE 21.1 Location of postearthquake survey areas for G. maculatus spawning habitat in the Avon and Heathcote River catchments, city of Christchurch, New Zealand.

Three spatially explicit protection mechanisms were identified in an analysis of proposed and operative resource management plans (Table 21.2). In this paper, we use the term ‘protected areas’ to denote spatially explicit areas identified in planning methods to address conservation objectives in statutory

Method

Protected area mechanism

Delineation method in plans

Information source

Planning documents

1

Network of small protected areas based on known spawning sites.

20-m diameter areas centered on point data coordinates of known spawning sites, identified in schedule to the plan.

Point data and descriptions from NISD’ and historical reports (Maw and McCallum-Clark, 2015).

Environment Canterbury (2015) Environment Canterbury (2016)

2

Mapped reaches of potential spawning habitat on a catchment basis.

Reaches identified in planning maps and referenced in the plan.

NISD point data and historical reports coupled with field surveys of riparian vegetation to identify potential habitat (Margetts. 2016).

Environment Canterbury (2014)

3

Mapped polygons of predicted spawning habitat coupled with a text description of where in the polygon the protection requirements apply.

Polygons identified in planning maps and GIS layer referenced in the plan.

GIS-based model of predicted spawning habitat (Greer et al., 2015).

Environment Canterbury (2017)

National Inanga Spawning Database.

policies and plans. The areas evaluated in this study are consistent with the IUCN definition of Category IV protected areas being ‘areas to protect particular species or habitats, where management reflects this priority’ (Dudley, 2008). The size of these areas is often relatively small with vaiying management arrangements depending on protection needs (Stolton et al., 2013).

Protected area and spawning site data were visualized in QGIS v2.8.18 (QGIS Development Team, 2016) and reach lengths (RL) calculated in relation to the centerlines of waterway channels digitized from 0.075 m resolution postquake aerial photographs (Land Information New Zealand, 2016). Three comparable RL metrics were calculated to reflect (1) the RL protected under each planning method, (2) extent of occurrence (EOO) of spawning sites, and (3) the total AOO of spawning sites (Table 21.3).

The effectiveness of each protection mechanism was evaluated as the percentage of postearthquake RLA0° located within the PA. Efficiency was considered using two ratios: RLE0° to RLprotec,ed and RLA0° to RLpro,ected. These reflect the size of the area set aside for protection (in terms of reach length) versus the extent of the spawning reach, and the size of the areas actually utilized for spawning, respectively. Each calculation was made on a catchment basis at a yearly temporal scale (i.e., 2015 and 2016) and also using the combined data from both years of postearthquake surveys.

TABLE 21.3 Metrics Calculated to Evaluate the Effectiveness and Efficiency of Protected Area Mechanisms for G. maculatus Spawning Habitat.

Metric

Definition

Calculation method

j protected

Reach length protected areas within a catclunent.

Combined length of waterway channels falling within protected areas, as calculated fr om channel centrelines on a catchment basis.

rleo°

Reach a length of the extent of occurrence (EOO) of spawning within each catclunent during the timeframe under consideration.

Total length of waterway channels between the upstream and downstream limits of spawning, as measured along channel centrelines on a catchment basis.

rla0°

Reach a length of the area of occupation (AOO) of all spawning sites within each catclunent during the timeframe under consideration.

Total length of all individual spawning sites, as measured along channel centrelines on a catclunent basis.

RESULTS

The three protected area mechanisms provided considerably different RLpro- 'ected vaiues reflecting their spatial basis (Table 21.4). However, for each mechanism, the RLprot was comparable between catchments. An overlay of each protection mechanism on combined postquake spawning site data is provided for each of the study catchments in Figure 21.2.

TABLE 21.4 Reach Length (RL) Protected by each of the Three Protected Area Mechanisms Evaluated in the Two Study Catchments.

Method Description of protected area mechanism

Reach length protected (in)

Avon river

Heathcote river

1

The network of small protected areas based on known spawning sites.

120

80

2

Mapped reaches of potential spawning habitat on a catchment basis.

3230

3098

3

Mapped polygons of predicted spawning habitat coupled with a text description of where in the polygon the protection requirements apply.

19.100

16.600

Method 3 was highly effective at protecting spawning habitat, achieving 92.7% protection in the Avon and 100% in the Heathcote using the combined postquake data (Table 21.5). The anomaly in the Avon relates to a few spawning sites that occurred outside of the mapped polygon in the vicinity of a small tributary and this occurred in both years. In the Avon, the effectiveness of method 2 was similar with close to 100% achieved (Table 21.4). However, in the Heathcote, only 69.9% of spawning habitat fell within the protected area and 45.6% in 2016. This reflected the occurrence, in both years, of spawning downstream of the protected area (Fig. 21.2d). In comparison, the effectiveness of method 1 was low. The percentage of habitat protected ranged from 3.9-14.2% (Table 21.4). This reflected the extent to which spawning occurred at previously known sites which formed the basis for delineation of the PAs (Fig. 21.2a,b).

In the efficiency evaluation, all of the protection mechanisms were relatively inefficient in terms of land use allocation when the evaluation metric was RLA0°. For all methods, more than half of the RLprotec,ed was allocated to areas that were not utilized for spawning habitat over the study period, even when the areas allocated were very small and targetted at previously known spawning sites. The highest percentage overlap with RLA0° was 47.5% achieved by method 1 in the Avon in 2016. However, when the evaluation metric was RLE0° the percentage overlap results changed considerably. Method 1 achieved a 100% overlap in the Avon in both years but in the Heathcote only 12.5%. Method 2 achieved 67.6% overlap in the Avon

Overlay of the spatial extent of three protection mechanisms found in conservation plans on the footprint of post-earthquake G

FIGURE 21.2 Overlay of the spatial extent of three protection mechanisms found in conservation plans on the footprint of post-earthquake G. maculatus spawning sites recorded in 2015 (n = 85) and 2016 (n = 103). (a) Method 1. Avon river, (b) Method 1. Heatlicote river, (c) Method 2, Avon river, (d) Method 2. Heatlicote river, (e) Method 3, Avon river, (f) Method 3, Heatlicote river.

(2016) and 48.7% in the Heathcote (2016), whilst method 3 achieved 11.5% in the Avon (2016) and 17.6% in the Heathcote (2016).

TABLE 21.5 Effectiveness of Three Protected Area Mechanisms for G. maculatus Spawning Habitat Following Earthquake-induced Landscape Change.

Protection mechanism

Time period

Percentage of habitat protected (% RLA0°)

Avon river

Heathcote river

Method 1

2015

5.4

7.5

2016

14.2

6.3

2015+2016

9.3

3.9

Method 2

2015

96.9

69.9

2016

99.0

45.6

2015+2016

98.0

52.5

Method 3

2015

96.9

100

2016

96.5

100

2015+2016

97.2

100

Comparing these results, method 3 was the least efficient in terms of land use allocation for the purposes of protection in all comparisons in the Avon. However, in the Heathcote method 1 was even less efficient in terms of RLE0°. This reflected that the protected areas identified were not well located in relation to the areas utilized for spawning (Fig. 21.2). In the Avon, the PAs under method 1 was much better located with all PAs overlapping the RLE0°. In terms of RLA0° method 1 also performed better in the Avon versus the Heathcote as a result of the PAs coinciding several of the areas actually utilized. However, even here the efficiency of the PA mechanism was rather variable with 47.5% of the RLprotected overlapping with spawning sites in 2016 but only 17.5% in 2015. This variability is associated with the repeated use of some, but not all, previously used spawning sites between years (Fig. 21.2).

Overall, method 2 produced relatively consistent results in the efficiency comparisons between years. This reflects that the RLE0° was similar in both catchments between years and also located in a similar position in the catchment versus the reaches mapped for protection. Within the RLE0° the total RLAO° was also very similar between years (Avon 386 m2 and 410 m2, Heathcote 133 m2and 158 nr for 2015 and 2016, respectively) despite considerable variation in the location of the sites used each year (Fig. 21.2).

 
Source
< Prev   CONTENTS   Source   Next >