55 Table 3.3: Description of spatial data
Data Description
Shape file of land classification for the eThekwini Municipality (2011)
Attributes such as generic and detailed habitat description, D’MOSS status, ecosystem condition, alien invasive plant infestation and area of open spaces within the eThekwini Municipality
Shape file of land-use for the eThekwini Municipality (2005)
Attributes and descriptions for all land-use within the eThekwini Municipality
Aerial photography of the eThekwini Municipality (2012)
Aerial photography depicting the entire eThekwini Municipality
56 The initial step was to identify and extract all green spaces within the eThekwini Municipality.
Green spaces were selected using existing typologies found in the relevant literature (Bell et al., 2007; Swanwick et al., 2003). It is important to note that as the analysis of all green spaces within the eThekwini Municipality was beyond the scope of the present study, six random green space types were selected for the purposes of the assessment; thus the findings presented in Chapter Four are trends reflective of these six green space types. The green space types selected included: settlement, tree crops, woodland, forest, grassland and thicket. ArcMap v 9.3 was used to overlay all sites corresponding with these green space types on aerial photography of the eThekwini Municipality. Thereafter, the Hawth’s Tools extension was used to select five random sites belonging to each of the six green space types (yielding 30 sites in total). These 30 sites were then subjected to the analysis described below.
The 30 green space sites were first examined using the eThekwini typology and thereafter with the Adapted typology. The two typologies were compared in terms of their utility in classifying/
polarising the selected green spaces into sub-environments on the basis of ecosystem condition (degraded, good, intermediate or transformed). This comparison between the two typologies also allowed for an assessment of the value of the set of criteria employed in the Adapted typology to visually assess the quality of selected green spaces. The criteria described below (Table 3.4) informed the statistics used to assess the quality of the different green space types in the Adapted typology.
57 Table 3.4: Criteria used to assess quality of selected green space types as part of the Adapted typology
Criteria Description
Classification The adapted classification of ecosystem condition (degraded, good, intermediate or transformed) of the green space site
Habitat and vegetation type, and threat status
• The habitat of the green space site.
The habitat type was determined using the Swanwick et al.
(2003) typology which categorises green spaces according to the categories: amenity, functional or semi-natural
• The vegetation found within the green space site
The vegetation type was identified according to the South African National Biodiversity Institute’s (SANBI, 2011a) KwaZulu–Natal vegetation type classification
• Ecosystem threat status of the green space site
The ecosystem threat status was based on the following categories: critically endangered, endangered, vulnerable or least threatened (SANBI, 2011b)
Infringement The level on infringement on the green space site using the likert scale: none, minimal, moderate or considerable
Statistics incorporated in the Adapted typology were calculated for each of the five sites, within the six green space types selected using the criteria listed in Table 3.4. Equations to calculate the indicators used to assess green space site quality in this study are listed below:
1. Quality based land cover condition (area and percentage)
Using the GIS tools described earlier the area and associated percentage of land cover determined to be degraded, good, intermediate or transformed was calculated as follows for each site:
𝑄𝑢𝑎𝑙𝑖𝑡𝑦 𝑏𝑎𝑠𝑒𝑑 𝑙𝑎𝑛𝑑 𝑐𝑜𝑣𝑒𝑟 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛 = (𝐴𝑟𝑒𝑎 𝑑𝑒𝑔𝑟𝑎𝑑𝑒𝑑/𝑔𝑜𝑜𝑑/𝑖𝑛𝑡𝑒𝑟𝑚𝑒𝑑𝑖𝑎𝑡𝑒/𝑡𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚𝑒𝑑
𝑇𝑜𝑡𝑎𝑙 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑠𝑖𝑡𝑒 ) × 100
58 2. Habitat, vegetation and threat status, and infringement (percentage)
For each site the proportions of land associated with the different habitat, vegetation, threat status and infringement categories were calculated as a percentage of the total area assessed for the green space type (i.e. sum of the areas of the five sites sampled) as follows:
𝐻𝑎𝑏𝑖𝑡𝑎𝑡, 𝑣𝑒𝑔𝑒𝑡𝑎𝑡𝑖𝑜𝑛, 𝑡ℎ𝑟𝑒𝑎𝑡 𝑠𝑡𝑎𝑡𝑢𝑠 𝑎𝑛𝑑 𝑖𝑛𝑓𝑟𝑖𝑛𝑔𝑒𝑚𝑒𝑛𝑡 = (𝐴𝑟𝑒𝑎 𝑖𝑛 𝑎𝑔𝑟𝑒𝑒𝑚𝑒𝑛𝑡 𝑤𝑖𝑡ℎ ℎ𝑎𝑏𝑖𝑡𝑎𝑡, 𝑣𝑒𝑔𝑒𝑡𝑎𝑡𝑖𝑜𝑛, 𝑡ℎ𝑟𝑒𝑎𝑡 𝑠𝑡𝑎𝑡𝑢𝑠 𝑎𝑛𝑑 𝑖𝑛𝑓𝑟𝑖𝑛𝑔𝑒𝑚𝑒𝑛𝑡
𝑇𝑜𝑡𝑎𝑙 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑔𝑟𝑒𝑒𝑛 𝑠𝑝𝑎𝑐𝑒 𝑡𝑦𝑝𝑒 ) × 100
3. Deviation index for green space type
For each site a deviation index was scored based on the amount of land that was classified differently to (deviated from) the broad eThekwini typology. The approach adopted for these calculations is based on a statistical technique commonly used to evaluate the level of agreement between two variables, which in this study are eThekwini typology green space condition and Adapted typology green space condition.
More specifically, the deviation index was calculated by first assigning a score to each of the five sites within a particular green space type. These scores were:
0- No deviation from eThekwini typology; land is classified the same by both typologies.
1- Minimal deviation, approximately a third of the land deviated from eThekwini typology (< 33%).
2- Moderate deviation, between a third and two thirds of the land deviated from eThekwini typology (> 33% but < 66%).
3- High deviation, more than two thirds of the land deviated from eThekwini typology (66% >).
The equation below was then used to calculate the percentage deviation for each green space type:
𝑃𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑔𝑟𝑒𝑒𝑛 𝑠𝑝𝑎𝑐𝑒 𝑡𝑦𝑝𝑒 = ( 𝑆𝑢𝑚 𝑜𝑓 𝑠𝑖𝑡𝑒 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑠𝑐𝑜𝑟𝑒𝑠
𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝑝𝑜𝑠𝑠𝑖𝑏𝑙𝑒 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 (3) × 𝑁𝑜. 𝑜𝑓 𝑠𝑖𝑡𝑒𝑠 (5)) × 100
*Based on the criteria used the maximum deviation score for any possible site is equivalent to 3 and in this particular study, 5 sites were sampled within each green space type.
59 The overall goal of the spatial analyses carried out was to assess whether the Adapted typology developed could be used to generate a more refined quantitative assessment of green space quality.