PRIORITISING QUATERNARY CATCHMENTS FOR INVASIVE ALIEN PLANT CONTROL WITHIN THE WORKING FOR WATER MPUMALANGA REGION: PROJECT REPORT March 2011

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga Region

Greg G. Forsyth, Patrick J. O’Farrell and David C. Le Maitre Reviewer: Brian W. van Wilgen

CSIR Natural Resources and the Environment P.O. Box 320 Stellenbosch 7599, South Africa.

Report number: CSIR/NRE/ECO/ER/2011/0031/B March 2011

Prepared for: Andrew Wannenburgh Working for Water Programme Private Bag X4390 Cape Town 8000 Tel: 021 441-2738 E-Mail: [email protected]

Contact person: Gregory Forsyth Tel: 021 888-2406 Fax: 021 888-2684 Email: [email protected]

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Executive Summary Introduction Invasive alien plant control requires the allocation of limited resources to control operations to maximise benefits. The priorities for such allocation are based on a mixture of fact and opinion, interpreted either subjectively or objectively, and often not transparent or repeatable. This project develops an approach that could assist managers and planners in the Working for Water Programme’s Mpumalanga region to prioritise their activities with a degree of transparency. We used the Analytic Hierarchy Process (AHP) to facilitate prioritisation. AHP is a multiple criteria decision-making tool for setting priorities when both qualitative and quantitative aspects of a decision need to be considered, and for achieving group consensus.

Conclusions and recommendations This study has successfully applied, in the Mpumalanga region, the approach developed earlier by van Wilgen et al. (2008). However, a number of follow-up actions will be needed if this approach is to deliver its full potential in terms of assisting the Working for Water Programme to improve its operations and its impact. We recommend the following: •

That the techniques developed at this quaternary catchment scale be adopted by Working for Water’s national and regional planning offices to assist with prioritisation, planning, and the allocation of resources to both existing and new projects on an ongoing basis.

•

The priorities identified here are used to guide the allocation of funds between quaternary catchments and subcatchments of Mpumalanga region.

•

That a spatial database be developed to underpin effective comparisons of areas. This database could contain data relating to most of the criteria identified here, including mean annual runoff, water demand, and water stress, terrestrial and fresh water biodiversity, protected areas, carrying capacity, cultivation potential and harvestable resources, cultural and tourism use features and the abundance and status of invasive alien species.

•

Each Working for Water region should maintain existing datasets and revise them on a regular basis. This period between revisions should not be longer than 3 years so as to coincide with the medium term expenditure framework of government.

Priorities in Mpumalanga A total of 162 quaternary catchments occur within four primary catchments that make up the Mpumalanga region, and were used in this assessment. The five quaternary catchments with the highest relative importance are B42D, X14A, B60B, B60C, and B60A. The top 25 priority catchments are spread across the eastern half of this region. Ten are found just north of Sabie and west of Acornhoek. Two are found just north-west of Waterval Boven. The remainder are found to the south and south-west of Nelspruit, along the edge of Swaziland, south of Ermelo on the Kwazulu-Natal border and to the east of Nelspruit. These catchments have both high water yields and are regarded as having irreplaceable biodiversity.

p. i

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Acknowledgements We thank the Working for Water Programme of the Department of Water and Environment Affairs for funding this work. We thank Mr Ronald Nenungwi and Mr Andrew Wannenburgh of the Working for Water Programme of the Department of Water and Environment Affairs for supporting the project and serving on the reference group. The following managers, implementing agents and researchers are thanked for their informed and enthusiastic participation in the workshop aimed at developing a model for assessing the priority quaternary catchments to clear in Working for Water’s Mpumalanga region: Brian Morris, Willem Bothma, Trent Sinclair, Eric Oosthuizen, Richard Green, Chris Foster, Ronald Nenungwi, Hannes de Lange, Jacobi Bach, Henning Kruger, Johan Brummer, Jeffrey Dibakoane and Andrew Wannenburgh. We also thank Kathleen Saunders of the Working for Water Programme’s Mpumalanga region for commenting on the draft report.

p. ii

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Contents Executive Summary........................................................................................................................................................... i Acknowledgements ........................................................................................................................................................... ii Contents ........................................................................................................................................................................... iii Figures.............................................................................................................................................................................. iv Tables ............................................................................................................................................................................... v 1.

INTRODUCTION

1

2.

SCOPE OF WORK

1

3.

APPROACH

4

4.

3.1

WORKSHOP TO DETERMINE RANKING CRITERIA

4

3.2

SPECIES SELECTION

4

3.3

GOAL AND CRITERIA

4

3.4

SELECTING APPROPRIATE DATA

4

3.4.1 Provision of water 3.4.1.1 Water yield 3.4.2 Biodiversity conservation 3.4.2.1 Terrestrial conservation 3.4.2.2 Fresh water conservation 3.4.3 Land capability 3.4.3.1 Carrying capacity 3.4.3.2 Cultivation potential 3.4.3.3 Harvestable resources 3.4.4 Protected areas 3.4.5 Cultural and tourism sites and routes 3.4.6 Presence of priority invasive alien plants 3.4.6.1 Invader status 3.4.6.2 Abundance

5 5 5 5 6 6 6 6 7 7 7 8 8 8

3.5

SPATIAL DATA SETS USED IN THE PRIORITISATION

8

3.6

CALCULATING THE WEIGHTS USED BY THE EXPERT CHOICE SOFTWARE

9

RESULTS

10

4.1

GOAL AND CRITERIA

10

4.2

MPUMALANGA PRIORITY QUATERNARY CATCHMENTS

12

5.

CONCLUSIONS

19

6.

RECOMMENDATIONS

20

7.

REFERENCES

21

p. iii

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Appendix 1

Priority invasive alien plants in the moist savanna and grassland biomes for the Mpumalanga province

Appendix 2

Agenda for Mpumalanga region

Appendix 3

Participants in expert workshops

Figures Figure 1:

A map showing the regional boundaries of the Working for Water programme in the Mpumalanga region and the two biomes, Savanna and Grassland which dominate this region. ____________________________________________________ 3

Figure 2:

A map showing the regional boundaries of the Working for Water programme in the Mpumalanga region, major towns and the Primary catchments found in this region. _______________________________________________________________ 3

Figure 3

The relative importance and ranking of the 45 top priority quaternary catchments out of the 162 in the Mpumalanga region. These values have been normalised.___________________________________________________________ 13

Figure 4:

The relative importance of quaternary catchments in the Mpumalanga region for each of the eight criteria that were assigned the highest importance in the model. The importance or weight is shown in parentheses. _____________________ 14

Figure 5:

The relative importance of quaternary catchments in the Mpumalanga region th th for the 9 to 16 most important criteria in the model. The importance or weight is shown in parentheses. ________________________________________________ 15

Figure 6:

The relative importance of quaternary catchments in the Mpumalanga region th for the 17 most important criteria in the model. The importance or weight is shown in parentheses. __________________________________________________ 16

Figure 7:

The priority quaternary catchments identified according to priority classes within the Mpumalanga region. Priority weightings reflect the scores for each catchment. ___________________________________________________________ 17

Figure 8:

The expenditure allocated to invasive alien plant clearing projects in the Mpumalanga region in 2010/11, in relation to priorities identified in the study (Current exppenditure). Clearly, most quaternary catchments, including several of high priority, do not receive any funding. The figure also shows the amount of funding that would be allocated to each catchment if the allocations were proportional to priorities (Expenditure according to priority), as well as the funding that would be allocated to each catchment (Top 16 priorities) if the allocation went to the 16 catchments with the highest priority. ___________________ 18

p. iv

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Tables Table 1:

Grazing potential ranges and associated midpoints in large livestock units 2 (LSU) per km (Scholes 1998). ____________________________________________ 6

Table 2:

Spatial datasets used to rank quaternary catchments in relation to selected criteria and sub-criteria. __________________________________________________ 9

Table 3:

Identified criteria and descriptions of the attributes that they incorporate ___________ 10

Table 4:

Nested criteria, together with the relative weightings, identified as significant for the purposes of prioritising quaternary catchments in the Mpumalanga region for the clearing of invasive alien plants. _____________________________________ 11

p. v

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

1. INTRODUCTION Invasive alien plant control requires the allocation of limited resources to control operations to maximise benefits. The priorities for allocating resources typically are based on a mixture of fact and informed opinion and this information may be interpreted either subjectively or objectively. However, the information and the rationale behind the resulting priorities are rarely made explicit so it is difficult to assess the validity of the resulting priorities. The Council for Scientific and Industrial Research (CSIR) has recently completed studies on the prioritisation of primary catchments for the purposes of guiding invasive alien plant control operations. These included and assessment of the terrestrial biomes of South Africa and the established national priorities (van Wilgen et al., 2008), and others that have focused on quaternary catchment scale and established priorities for the Northern Cape and Western Cape (Forsyth et al., 2009, Le Maitre & Forsyth 2010). These studies developed an approach and method that enables managers and planners in the Working for Water Programme to prioritise their activities in a way that is transparent, logical and defensible. A biome-level study has also been undertaken and developed methods for the identification of a priority list of (i) invasive alien plants, and (ii) areas (primary catchments) within the terrestrial biomes of South Africa that should be targeted for control by the Working for Water Programme. Mr Ronald Nenungwi of the Department of Water Affairs and Forestry (DWAF) asked the CSIR to assist in prioritising areas to clear within the Working for Water region in the Mpumalanga province by applying the methods developed for the biomes and the quaternary catchment scale studies. This report presents the results of this study to determine the priority quaternary catchments to clear within the Mpumalanga region. We also make recommendations for further improvements to the prioritisation process and its implementation by the Working for Water Programme.

2. SCOPE OF WORK This project is conducted as part of a collaborative agreement between the DWAF and the CSIR. The work was guided and reviewed by a reference group, appointed by DWAF at the initiation of the project, in terms of the collaborative agreement. Members of the reference group are: •

Mr Ronald Nenungwi (Department of Water and Environment Affairs – Working for Water Programme, Nelspruit)

•

Mr Andrew Wannenburgh (Department of Water and Environment Affairs – Working for Water Programme)

It was agreed at the outset of the study that the planned scope of activities would be as follows: •

The work would focus on the dominant biomes found in the region, the grassland and moist savanna (Figure 1).

Page 1

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

•

The work would entail prioritising areas to clear at a quaternary catchment scale within the portions of the Olifants catchment (B), the Upper Vaal catchment (C), the Usutu catchment (W) and the Inkomati catchment (X) which occur in the Working for Water regional boundaries of the Mpumalanga Province. We selected all the quaternary catchments which fall within this boundary (Figure 2).

•

The Analytic Hierarchy Process (AHP) would be used to facilitate the prioritisation of quaternary catchments using Expert Choice 11.5 decision support software (Anon. 2009).

•

The criteria to use for prioritising the quaternary catchments for the clearing of invasive alien plants would be identified and agreed to at an expert workshop to be held in Mpumalanga.

•

An obvious criterion was whether or not priority alien invasive species are present or likely to spread in a quaternary catchment. In this regard it was agreed that we would start with the list of priority species identified for the Moist Savanna and the Grassland (see Appendix 1) by the recent CSIR study (van Wilgen et al., 2008). In the Mpumalanga study we used data from National Invasive Alien Plant Survey (Kotzé et al. 2010) for prioritising quaternary catchments based on the current extent and density of invasions. The National Invasive Alien Plant Survey (NIAPS) was undertaken by the ARC and used an ecotope classification to extrapolate from the point samples to landscapes.

•

The work of Rouget et al. (2004) and Mgidi et al. (2007) would be used if necessary to identify areas that are likely to become invaded by the species identified in the CSIR study as priority species for clearing.

•

Where applicable and available we would also made use of river (Nel et al. 2007; Nel et al. 2011) and terrestrial (Driver et al. 2005) conservation prioritisation datasets for various spatial scales.

•

The assessment would focus on (a) the criteria and (b) the relative weighting of those criteria that will be used in prioritising the quaternary catchments and not on direct pair-wise catchment comparisons. The primary reason for this is that the AHP approach requires a pairwise ranking and there are too many quaternary catchments in the primary catchments of Mpumalanga to make this feasible. We would therefore apply the procedures which we developed for automating these comparisons for the Western and Northern Cape studies (Forsyth et al., 2009, Le Maitre & Forsyth 2010).

1

The Working for Water Programme’s strategic plan for 2008 – 2012 lists “the reduction of impact of existing priority invasive alien plant problems” as one of three primary goals relating to natural resource management. The other two are related to preventing problems, and building capacity to address problems. This project will assist in the identification of such priorities at a quaternary scale in the Mpumalanga region, which are not clearly defined at present.

1

AHP is a multiple criteria decision-making tool for setting priorities when both qualitative and quantitative aspects of a decision need to be considered. It involves setting a goal, breaking it down into its constituent parts and then assigning relative weights to each of these, thereby progressing from the general to the specific. Scoring is on a relative basis comparing one choice with another. Relative scores for each choice are computed with each level of the hierarchy. Scores are then synthesised through a model contained in Expert Choice. This yields a composite score for each choice at every level as well as an overall score.

Page 2

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Figure 1: A map showing the regional boundaries of the Working for Water programme in the Mpumalanga region and the two biomes, Savanna and Grassland which dominate this region.

Figure 2: A map showing the regional boundaries of the Working for Water programme in the Mpumalanga region, major towns and the Primary catchments found in this region.

Page 3

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

3. APPROACH 3.1 WORKSHOP TO DETERMINE RANKING CRITERIA A one day workshop was held at the Bundu Country Lodge just north of Nelspruit on Thursday 22nd July 2010. The workshop agenda is provided in Appendix 2. A total of 15 people from key sectors, including, Conservation, Agriculture and Water as well as the Working for Water representatives responsible for implementing Working for Water projects in the region, participated in the workshop (see Appendix 3). The topics addressed at the workshop were: •

Findings from previous studies conducted by the CSIR at a regional or provincial level in the Northern Cape and the Western Cape (Forsyth et al. 2009; Le Maitre & Forsyth 2010)

•

An explanation of the Analytic Hierarchy Process

•

Current priority invasive alien plants for the savanna and grassland biomes (See Appendix 1)

•

The goal, criteria (or objectives) and sub-criteria (or sub-objectives) for prioritising quaternary catchments. These were captured on paper and using the Expert Choice software

•

Pair-wise comparisons (ranking) of the agreed criteria and sub-criteria using the Analytical Hierarchy Process approach in the Expert Choice software

•

Identifying the datasets that will allow for the objective comparison of quaternary catchments with regard to particular criteria

3.2 SPECIES SELECTION Workshop participants undertook a species verification exercise. Here local experts examined and commented on the lists of invasive species identified by van Wilgen et al. (2008; 2010) for the Savanna and Grassland biomes. A consolidated list was created which included those verified species as well as additional species which participants felt were problematic. The result was a complete list of the most important alien species for the Mpumalanga region. These are presented according to the regions biomes in Appendix 1.

3.3 GOAL AND CRITERIA The workshop participants established a goal for prioritizing the clearing of catchments in within the Mpumalanga region. Criteria for meeting this goal were also developed and weighted in relation to one another. The AHP was used to compare each individual criterion to each other and to assign weightings to each of these according to their relative importance (Saaty, 1990). The Expert Choice software package, which uses an Analytical Hierarchical Processes, was used to facilitate this process (Anon 2009).

3.4 SELECTING APPROPRIATE DATA A summary of the data sets used, the rationale for using these to address the criteria underlying the hierarchy mode, and methods are provided below. We were limited in our choice to those

Page 4

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

datasets that were readily available (in the public domain) and covered the entire working for water Mpumalanga region. We did not rely on expert opinion in palace of data sets. Where no data was available we used surrogate measures or related variables. Where this was not possible we gave all quaternary catchments the same value.

3.4.1

Provision of water

3.4.1.1 Water yield 3.4.1.1.1

Water yield

Surface water yield information was obtained from the Water Resources 2005 quaternary catchment dataset (Middleton and Bailey 2008). Information relating to the annual run-off was used (coded as “naturalized run-off” in this database). To compare catchments directly we used the mean annual runoff to rank catchments with catchments with higher water yields assigned higher weights in proportion to their area.

3.4.1.1.2

Water demand

Catchment water yield balance or the mean annual yield from its water supply systems, minus the mean annual demand at a quaternary catchment scale, was estimated for the Department of Water Affairs’ Water Situation Assessment Model (WSAM 2003). The WSAM model also accumulates mean annual runoff downstream so cumulative effects are accounted for. Negative values represent a high water demand (i.e. demand exceeds yields) so the positive values were converted to negative and vice versa. The lowest value was added to all the values to make the series positive. The catchments with the greatest demand received the highest proportional weight.

3.4.1.1.3

Water stress

When the model was developed at the workshop we included a sub-criterion for water stress for giving catchments with little or no water availability a higher priority. The focus here was on securing the reserve in highly stressed catchments. We added additional sub-criteria to allow for this and have currently given all the quaternaries an equal weight for this criterion as we have no data sets suitable for this purpose.

3.4.2

Biodiversity conservation

3.4.2.1 Terrestrial conservation The Mpumalanga biodiversity plan (Lötter & Ferrar, 2006) identifies six conservation categories in Mpumalanga, based on the distribution of hundreds of biodiversity and ecological features throughout the province. These six categories are 1) protected areas currently under formal biodiversity protection (14.8% of province), 2) irreplaceable areas in urgent need of protected area status (2.4% of province), 3) highly significant areas, requiring strict land-use controls (12.3% of province), 4), important and necessary areas requiring special care (9.5% of province), 5) areas of least concern, providing sites for development (25.5% of province) 6) areas with no

Page 5

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

natural habitat remaining, providing preferred sites for all forms of development (35.8% with 61.3% - available for development following impact assessment studies). During the stakeholder workshop it was decided that we would include three of these areas as sub-sub-criteria under the terrestrial conservation sub-criteria. These were: irreplaceable areas, highly significant areas, and important and necessary areas. We area apportioned each of the categories according to quaternary catchments, and derived priority weights for each of these.

3.4.2.2 Fresh water conservation We determined areas of conservation significance from the National Freshwater Ecosystem Priority Areas (NFEPA) data (Nel et al. 2011). Areas identified as Fresh Water Ecosystem Priority Areas and Fish Support Areas were selected and their area calculated as a proportion of each quaternary catchment. Catchments with greater proportional areas of these fresh water ecosystem priority areas were assigned higher weights.

3.4.3

Land capability

3.4.3.1 Carrying capacity The relative value of the land for livestock production was estimated by calculating the grazing potential of quaternary catchments. This potential was derived from Scholes’ (1998) estimates of sustainable mean domestic livestock production (Table 1). This approach may underestimate the carrying capacity for browsing antelope but as game farming only occurs in limited areas this would not significantly affect the outcome.

Table 1: Grazing potential ranges and associated midpoints in large livestock units (LSU) per km2 (Scholes 1998).

LSUrange LSUmid-point

0– 1 0.5

1– 2 1.5

2– 3 2.5

3– 4 3.5

4– 6 5

6– 8 7

8 – 10 9

10– 14 12

14– 18 16

18– 22 20

We assumed that only untransformed (natural) vegetation would support livestock, and subtracted the area of transformed vegetation, using the National Land Cover Database 2000, (Van den Berg et al. 2008) from the carrying capacity layer (Scholes 1998) in each catchment before the above calculation was made. We took the midpoint of each class, and multiplied it by the remaining area in that class in each quaternary catchment to get an area weighted mean grazing capacity. The catchments with the highest grazing capacity were given the greatest weights.

3.4.3.2 Cultivation potential We used the Land Capability Data (ARC 2002) to determine the cultivation potential of areas within this region. This database divides the national land surface into eight classes, ranging from 1 = Very high agricultural potential to 8 = Very low agricultural potential. We apportioned quaternary catchments into areas of different land capability using the land capability database once all transformed areas were removed. We multiplied the area under each of the classes by the number of the class (1 to 8), summed these scores for the quaternary catchments and divided by the area of the quaternary catchment. We then subtracted all values from highest category value

Page 6

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

(8), to reverse the order of scores, making the area with the highest score as having the most potential for cultivation and vice versa.

3.4.3.3 Harvestable resources We were unable to obtain spatial data on areas where there is currently harvesting of veld products such as thatching grass and medical plants. However we did find a suitable proxy variable to assist us in determining where the harvesting of indigenous plants would be important for sustaining rural livelihoods. Naudé et al. (2007) described the functioning of settlements in South Africa in terms of a rural – urban typology. They then assigned values corresponding to the dominant settlement type to each of approximately 25 000 irregularly shaped mesozones (approximately 49 km2 or 7 km by 7 km in size) in South Africa. One of the settlement types recognised is “dispersed rural settlements”. Most of these settlements are located in the rural areas of South African that were previously referred to as “homelands”. They typically have high population densities but are distant from formal employment opportunities and social services. This implies a high degree of dependence on the utilization natural veld products to sustain their livelihoods. We calculated the area of mesozones of this type as a proportion of each quaternary catchment. Catchments were allocated weights in proportion to the area occupied by the mesozones.

3.4.4

Protected areas

We used the protected areas, both formal and informal, from the recent update of the national protected areas database prepared for the National Protected Areas Expansion Strategy (Biodiversity GIS 2011). These areas include those managed by South African National Parks, provincial nature conservation authorities and local municipalities, private nature reserves, conservancies and water management areas. We allocated protected areas to each of the following categories: Nature reserves, State protected areas, Protected environments, Conservancies. The sum of the areas of each of these protected areas categories was expressed as a proportion of the total area of the quaternary catchment. The catchment with the greatest proportion of these summed areas received the highest weight for each of these four protected area categories.

3.4.5

Cultural and tourism sites and routes

From the environmental potential atlas (ENPAT 2001), we selected all shape files that related to cultural features and tourism features (these included protected areas, cultural sites, gardens, heritage sites, hiking areas, Ramsar sites, resorts, rock art, routes, cultural interest, historic battle fields). A single data layer of these features was created, the areas of each of these calculated, and proportioned according to quaternary catchments. Quaternary catchments with higher proportional areas of cultural and tourism features received the higher weights.

Page 7

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

3.4.6

Presence of priority invasive alien plants

3.4.6.1 Invader status To determine invader status we used the South African Plant Invaders Atlas (SAPIA) database (Henderson, 1998) which contains records on the presence of invasive alien plants by quarter degree square (QDS) for the country. We used this in place of the National Invasive Alien Plant Survey (Kotzé et al. 2010) as it was more comprehensive. From this database we extracted all records from those QDS that fell within the Working for Water regional boundary, were for species that appeared on the combined list of priority invasive alien plants determined at the project workshop, and where the invader status class was listed as frequent, abundant or very abundant. Priority invasive alien plants species were then assigned an invader status category and score, these being: transformer (score =3), potential transformer (score =2), special effect weed or minor weed (score = 1). We then determined the number of records of each of these species, per invader status class, occurring in each QDS. The invader status category scores were multiplied by the number records for each status class and then summed up producing a value for each QDS (For example, if there were five records of transformer weeds in a QDS and 1 minor weed then the combined value for the QDS would be (5*3) + (1*1) = 16). We then assigned each quaternary catchment a QDS score, selecting the QDS which covered the largest proportion of each quaternary catchment. A proportional weight for each quaternary catchment was then calculated by dividing the designated quaternary score by the sum of all scores for the quaternary catchments in the Mpumalanga region.

3.4.6.2 Abundance Condensed invaded area was extracted for each priority species from the NIAPS data (Kotzé 2010) and expressed as a proportion of the area of the respective quaternary catchment. Whilst we note that not all invasive alien species have been captured in this survey the dominant species which make up most of this density are captured. Quaternary catchments with a greater proportion of condensed hectares of invasive alien plants received the higher weights.

3.5 SPATIAL DATA SETS USED IN THE PRIORITISATION The prioritisation of the catchments relies on matching of criteria data sets that either provide a direct measure of the criterion, or a surrogate measure, where direct data are not available (Table 2).

Page 8

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Table 2: Spatial datasets used to rank quaternary catchments in relation to selected criteria and sub-criteria.

Criterion

Water provision

Biodiversity conservation

Land resources

Sub criteria

Spatial data1

Water yield

Water Resources 2005 (Middleton & Bailey 2008)

Water demand

Water Situation Assessment Model at quaternary catchment scale (WSAM 2003)

Water stress

Water Situation Assessment Model at quaternary catchment scale (WSAM 2003)

Terrestrial ecosystems

Mpumalanga conservation plan (Lötter & Ferrar, 2006).

Fresh water (rivers, free flowing rivers & wetlands)

National freshwater ecosystem priority areas (Nel et al.2011)

Forage production

Areas of homogenous grazing potential (Scholes, 1998)

Utilisable species

Scattered rural population surrogate – CSIR GAP data (Naudé 2007); National Land Cover Database 2000 (Van den Berg et al. 2008)

Cultivation potential

National land capability (ARC 2002); National Land Cover Database 2000 (Van den Berg et al . 2008)

Nature reserves

National Protected Areas Database (SANBI)(Biodiversity GIS 2010)

State protected areas

National Protected Areas Database (SANBI)(Biodiversity GIS 2010)

Protected environments

National Protected Areas Database (SANBI)(Biodiversity GIS 2010)

Conservancies

National Protected Areas Database (SANBI)(Biodiversity GIS 2010)

Invader status

South African Plant Invaders Atlas (Henderson 1998 and revisions) - copy of database March 2010

Abundance of invasive alien plants

National Invasive Alien Plant Survey (Kotzé 2010)

Protected areas

Presence of priority invasive alien species (current extent)

Cultural and tourism sites and routes 1

Environmental Potential Atlas (ENPAT 2001)

See reference section for complete references

3.6 CALCULATING THE WEIGHTS USED BY THE EXPERT CHOICE SOFTWARE The Expert Choice software (Anon 2009) calculates the weights of alternatives (quaternary catchments in this case) as proportions that sum to one. For each of the criteria and sub-criteria used by the AHP model we calculated the sum of the values for each variable for each quaternary catchment. Each quaternary catchment’s value for that variable was then divided by the corresponding total to give the final weight.

Page 9

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

4. RESULTS 4.1 GOAL AND CRITERIA The goal that that was agreed on at the Nelspruit workshop was “To prioritise the areas to manage invasive alien plant species within Mpumalanga so as to secure sustainable ecosystem benefits”. Goal setting was followed by the identification of six criteria for prioritising the clearing of quaternary catchments within the Mpumalanga region. The criteria identified were: Provision of water, biodiversity conservation, land capability, protected areas, presence of priority invasive alien species, tourism and cultural features. Descriptions of these criteria are provided in Table 3. Many of the criteria were further divided into sub-criteria, for example, biodiversity conservation was divided into the sub-criteria of terrestrial conservation and fresh water conservation. These sub-criteria were further divided into sub-sub-criteria. The final ranking for prioritisation, considering all the criteria, sub-criteria, and sub-sub-criteria is provided Table 4.

Table 3: Identified criteria and descriptions of the attributes that they incorporate Criterion

Description of criteria attributes

Provision of water

The degree to which alien species affect our limited water resources. This includes water quantity and quality (of both surface and ground water), high yielding areas and ground water recharge areas. The objective behind this criterion was to secure the sustainable provision of water.

Biodiversity conservation

The degree to which the alien species are able to displace indigenous species, in particular their impact on protected areas, terrestrial, wetland, and riparian systems, and the biodiversity they contain. The objective behind this criterion was to protect biodiversity priority areas.

Land capability

The degree to which the alien species are able to displace indigenous species which are important for grazing and harvestable resources. The objective behind this criterion was to retain land capability.

Protected areas

The degree to which the alien species is able to displace indigenous species, within protected areas. The objective behind this criterion was to maintain protected areas.

Presence of priority invasive alien species

The actual presence of invaders, both density and number (i.e. whether and invasive species is an ecosystem transformer). The objective behind this criterion was to identify areas with a large scale presence of invasive alien species capable of transforming ecosystems.

Cultural and tourism sites and routes

The degree to which the alien species is able to displace indigenous species, negatively affecting cultural and tourist features and sites, e.g. restricting access, impacting views and restricting recreation opportunities and cultural practices. The objective behind this criterion was to maintain these important cultural and tourism sites and routes.

Page 10

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Table 4: Nested criteria, together with the relative weightings, identified as significant for the purposes of prioritising quaternary catchments in the Mpumalanga region for the clearing of invasive alien plants.

Criterion

Provision of water

Weighting assigned (%)

Sub-criterion

Weighting assigned (%)

Water yield

0.269

Water demand

0.045

Water stress

0.09

Sub-subcriterion

Weighting assigned (%)

0.404

Irreplaceability 0.288 Biodiversity conservation

0.216

Terrestrial conservation

Fresh water conservation

0.072

Carrying capacity

0.094

Cultivation potential

0.013

Utilisable species

0.022

Highly significant Important and necessary

0.158 0.041 0.017

0.129 Land capability

0.09 Protected areas

Presence of priority invasive alien species Key tourism and cultural features

0.055

Nature reserves

0.035

Protected environments

0.009

Conservancies

0.005

State protected areas

0.041

Invader status

0.041

Abundances

0.014

0.033

0.033

The most important sub-criterion identified overall is water yield. This carries a weighting of 26.9% (Table 4). Next in order of importance is are irreplaceable terrestrial areas for biodiversity conservation (15.8%), water stress (9.0%) and carrying capacity (9.4%). State protected areas and nature reserves are seen as less important sub-criteria and assigned weightings of 4.1% and 3.5% respectively (Table 4). Socio-economic factors were discussed, in particular the location of poor people with respect to prioritising areas. It was decided that, given the pervasive nature of poverty within South African society, all areas of this region contained many more unemployed poor people than the Working for Water programme would be able to employ. Such a criterion would therefore be unhelpful in deciding between priorities. Nevertheless, we believe that social and economic criteria are relevant and should be revisited when revising these priorities.

Page 11

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

4.2 MPUMALANGA PRIORITY QUATERNARY CATCHMENTS Within the Mpumalanga region the ten catchments with the highest relative importance rankings are: B42D, X14A, B60B, B60C, B60A, W52A, B60D, X31C, X21H, X31A (Figure 3). All criteria have been mapped according to quaternary catchments indicating priorities for each of these (Figures 4-6). Of the 162 quaternary catchments investigated, the first 25 accounted for more than 30% of the total weight of the priorities in the model. The top 25 priority catchments are spread across the eastern half of the region. Ten are found just north of Sabie and west of Acornhoek. Two high priority catchments are found just north-west of Waterval Boven, and four to the south-west of Nelspruit, and four along the north eastern edge of the Swaziland border. A further two high order priority catchments are found south of Ermelo on the Kwazulu Natal border, and two to the immediate east of Nelspruit. Water yield was the most important criterion with 26% of the total weight of the prioritisation model (Table 4). Many of these catchments are in high lying areas and areas that descend from the escarpment to the Lowveld and therefore have high runoff and water yield (Figure 4). The majority of these catchments contain highly important or irreplaceable biodiversity areas (Figure 4-6). Following on from this they typically contain protected areas of various statuses (Figure 4-6). A further criterion that appear to contribute to the selection of these catchments is the generally high carrying capacity found within these catchments (Figure 4).

Page 12

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Figure 3 The relative importance and ranking of the 45 top priority quaternary catchments out of the 162 in the Mpumalanga region. These values have been normalised.

Page 13

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Figure 4: The relative importance of quaternary catchments in the Mpumalanga region for each of the eight criteria that were assigned the highest importance in the model. The importance or weight is shown in parentheses.

Page 14

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Figure 5: The relative importance of quaternary catchments in the Mpumalanga region for the 9th to 16th most important criteria in the model. The importance or weight is shown in parentheses.

Page 15

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Figure 6: The relative importance of quaternary catchments in the Mpumalanga region for the 17th most important criteria in the model. The importance or weight is shown in parentheses.

The highest priority catchments are mainly those which have high water yields, contain areas that are considered to be irreplaceable for a biodiversity perspective, have relatively high carrying capacities and contain fresh water biodiversity priority areas (Figure 7). Of the 16 quaternary catchments which are currently have clearing projects in operation, only one is in the top ten priority quaternary catchments, and six in the top 40 priority catchments identified here (Figure 8). Most of the high-priority catchments receive no funding, but there are many funded projects in low-priority catchments.

Page 16

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Figure 7: The priority quaternary catchments identified according to priority classes within the Mpumalanga region. Priority weightings reflect the scores for each catchment.

Page 17

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Current expenditure

Budget (Millions of Rands)

Exp. According to priority Top 16 priorities

Quaternary catchment priority ranking (1= most important to 162 =least important) Figure 8: The expenditure allocated to invasive alien plant clearing projects in the Mpumalanga region in 2010/11, in relation to priorities identified in the study (Current expenditure). Clearly, most quaternary catchments, including several of high priority, do not receive any funding. The figure also shows the amount of funding that would be allocated to each catchment if the allocations were proportional to priorities (Expenditure according to priority), as well as the funding that would be allocated to each catchment (Top 16 priorities) if the allocation went to the 16 catchments with the highest priority.

Page 18

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

5. CONCLUSIONS This study has identified the highest priority quaternary catchments for managing invasive alien plants within the Olifants, the Upper Vaal, Usutu and Inkomati primary catchments which are managed by the Working for Water Mpumalanga region and compared them with the current budget allocations. There is little alignment between identified priority areas and the location of existing projects and most quaternary catchments with high priorities do not have projects at present. The regional Working for Water management team needs to assess how best to improve the current alignment between budgets and priority catchments over time. The techniques we have developed to determine the priority areas for clearing invasive alien plants at a quaternary catchment scale are workable but it is not really clear whether the results correspond with what the managers and experts would intuitively expect. The difference that this new approach makes is that the managers can now evaluate the roles of the individual data elements which contribute to each score assigned by the Expert Choice (AHP) software and adjust them where necessary. An advantage of using AHP is that it can handle a large number of alternatives enabling comparisons to be made on any number of quaternary catchments. This study has identified a number of shortcomings regarding the available spatial data and, in other instances, the lack of appropriate spatial data to represent the criteria and sub-criteria that were considered important by the experts. We are however still limited in our ability to determine impacts on: cultural features and natural resource use, and are unable to identify water stressed catchments. Ms Kathleen Saunders suggested in a comment that the Working for Water Programme should give a low priority to properties already under consideration or earmarked for mining. This is something that could be considered for future assessments but because a number of other factors may be involved (e.g. ability of the species to disperse beyond the property, opportunities for conservation offsets or for the mining to clear the invasions under is environmental management plan), this issue may be more appropriately addressed at the property level than as a general rule.

Page 19

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

6. RECOMMENDATIONS This study has been successful in applying the approach developed by van Wilgen et al. (2008), Forsyth et al. (2009) and Le Maitre and Forsyth (2010), at a quaternary catchment scale in the Mpumalanga region. However, a number of follow-up actions will be needed if this approach is to deliver its full potential in terms of assisting the Working for Water Programme to improve its operations and its impact. With this in mind, we recommend the following: •

That the techniques developed at the primary and quaternary catchment scale be adopted by Working for Water’s national and regional planning offices to assist with prioritisation, planning, and the allocation of resources to both existing and new projects on an ongoing basis. This would assist in establishing a uniform approach to prioritisation across the organization and allow for regular reassessments as needed and when new or improved datasets become available.

•

Future funding should be channelled into establishing invasive alien clearing projects in catchments that have been identified as having high priorities. In addition projects in low priority catchments should be brought to a conclusion.

•

Each Working for Water region should maintain existing datasets and revise them on a regular basis. This should not be longer than 3 years so as to coincide with the medium term expenditure framework of government.

•

That a spatial database be developed to underpin effective comparisons of areas. This database could contain data relating to most of the criteria identified here, including water provision, biodiversity conservation, land resources, protected areas, invasive alien abundance and status, and cultural and tourism factors. The Working for Water Information Management System should be used to store the necessary data.

•

The results are as good as the underlying spatial datasets but as new or revised datasets become available they should be incorporated into the hierarchical model and used to generate a revised set of rankings (catchment scores). In addition as understanding improves, the weightings assigned to the criteria and sub-criteria in the hierarchical model can be adjusted, and criteria and sub-criteria added or removed.

Page 20

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

7. REFERENCES Anonymous (2009) Expert Choice 11.5. Expert Choice Inc., Pittsburgh, PA, United States of America. Agricultural Research Council (2002) A System for Soil and Land Capability Classification for Agriculture in South Africa. Department of Agriculture and Environmental Affairs. Biodiversity GIS (2010) South African National Biodiversity Institute's spatial biodiversity planning information system. http://bgis.sanbi.org. Accessed: November 2010. Driver, A., Maze, K., Rouget, M., Lombard, A.T., Nel, J., Turpie, J.K., Cowling, R.M., Desmet, P., Goodman, P., Harris, J., Jonas Z., Reyers, B., Sink, K. and Strauss, T. (2005) National Spatial Biodiversity Assessment 2004: Priorities for biodiversity conservation in South Africa. Strelitzia 17, South African National Biodiversity Institute, Pretoria. ENPAT (2001) Environmental Potential Atlas. Department of Environmental Affairs and Tourism, South Africa. Forsyth, G.G., Le Maitre D.C. and van Wilgen, B.W. (2009) Prioritising quaternary catchments for invasive alien plant control within the Fynbos and Karoo biomes of the Western Cape Province. CSIR Report CSIR/NRE/ECO/ER/2009/0094/B. Natural Resources and the Environment, CSIR, Stellenbosch. Henderson, L. (1998) Southern African Plant Invaders Atlas (SAPIA). Applied Plant Science 12, 3132. Kotzé, I., Beukes, H., van den Berg, E. and Newby, T. (2010) National Invasive Alien Plant Survey. Report No. GW/A/2010/21, Agricultural Research Council – Institute for Soil, Climate and Water, Pretoria. Le Maitre, D.C. and Forsyth, G.G. (2010) Prioritising quaternary catchments for invasive alien plant control within the Savanna, Nama and Succulent Karoo biomes of the Northern Cape province. CSIR Report CSIR/NRE/ECO/ER/2010/0015/B. Natural Resources and the Environment, CSIR, Stellenbosch. Lötter, M.C. and Ferrar, A.A. (2006) Mpumalanga Biodiversity Conservation Plan Map. Mpumalanga Parks Board, Nelspruit. Mgidi, T.N., Le Maitre, D.C., Schonegevel, L., Nel, J.L., Rouget, M. and Richardson, D.M. (2007) Alien plant invasions – incorporating emerging invaders in regional prioritization: a pragmatic approach for southern Africa. Journal of Environmental Management 84, 173-187. Middleton, B.J. and Bailey, A.K. (eds) (2008) Water Resources Of South Africa, 2005 Study (WR2005). Report TT 380/08, Water Research Commission, Pretoria.

Page 21

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Naudé, A.H., Badenhorst, W., Zietsman, H.L., Van Huyssteen, E., Maritz, J. (2007) Technical overview of the mesoframe methodology and South African Geospatial Analysis Platform. CSIR, Pretoria. (Report number CSIR/BE/PSS/IR/2007/0104/B). Nel, J.L., Roux, D.J., Maree, G., Kleynhans, C.J., Moolman, J., Reyers, B., Rouget, M. and Cowling, R.M. (2007) Rivers in peril inside and outside protected areas: a systematic approach to conservation assessment of river ecosystems. Diversity and Distributions 13, 341–352. Nel, J.L., Driver, A., Strydom, W., Maherry, A., Petersen, C., Roux, D.J., Nienaber, S., van Deventer, H., Smith-Adao, L.B. and Hill, L. (2011) Atlas of Freshwater Ecosystem Priority Areas in South Africa: Maps to support sustainable development of water resources. Atlas and accompanying data available from CSIR or WRC. Rouget, M., Richardson, D.M., Nel, J.L., Le Maitre, D.C., Egoh, B. and Mgidi, T. (2004) Mapping the potential ranges of major plant invaders in South Africa, Lesotho and Swaziland using climatic suitability. Diversity and Distributions 10, 475 – 484. Saaty, T.L. (1990) How to make a decision: The analytic hierarchy process. European Journal of Operational Research 48, 9-26. Scholes, R.J. (1998) The South African 1:250 000 maps of areas of homogenous grazing potential. Report ENV-P-C 98190, CSIR, Pretoria. Schulze, R.E. and Horan, M.J.C. (2007) Soils: Hydrological Attributes. In: Schulze, R.E. (Ed). 2007. South African Atlas of Climatology and Agrohydrology. Water Research Commission, Pretoria, RSA, WRC Report 1489/1/06, Section 4.2. Van den Berg, E.C., Plarre, C., Van den Berg, H.M. and Thompson, M.W. (2008) The South African National Land Cover 2000. Agricultural Research Council (ARC) and Council for Scientific and Industrial Research (CSIR), Pretoria. Report No. GW/A/2008/86. van Wilgen, B.W., Forsyth, G.G. and Le Maitre, D.C. (2008) The prioritization of species and primary catchments for the purposes of guiding invasive alien plant control operations in the terrestrial biomes of South Africa. CSIR Report CSIR/NRE/ECO/ER/2008/0070/C. Natural Resources and the Environment, CSIR, Stellenbosch. van Wilgen, B.W., Le Maitre, D.C., Forsyth, G.G. and O’Farrell, P.J. (2010) The prioritization of terrestrial biomes for invasive alien plant control in South Africa. CSIR Report CSIR/NRE/ECO/ER/2010/0004/C. Natural Resources and the Environment, CSIR, Stellenbosch. WSAM (2003) Water situation assessment model, Version 3.002. Department of Water Affairs and Forestry (DWAF), Republic of South Africa.

Page 22

Page 23

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

APPENDIX 1: PRIORITY INVASIVE ALIEN PLANTS IN THE MOIST SAVANNA AND GRASSLAND BIOMES FOR THE MPUMALANGA PROVINCE The invasive alien plant taxa selected for prioritisation in the Savanna and Grassland biome. A # indicates those species identified by van Wilgen, Forsyth and Le Maitre, (2008) as a threat but not regarded by the Mpumalanga experts as a threat. An * denotes species regarded as problematic and added to the list during the workshop by the local experts. Savanna species Acacia mearnsii (Black wattle) Acacia melanoxylon (Blackwood) Arundo donax (Giant reed)# Caesalpinia decapetala (Mauritius thorn) Cereus jamacaru (Queen of the night) Cestrum laevigatum (Inkberry) Chromolaena odorata (Triffid weed) Eucalyptus camaldulensis (Red river gum)* Jacaranda mimosifolia (Jacaranda) Lantana camara (Lantana) Melia azedarach (Persian lilac) Opuntia sp. (Cactus pear)* Parthenium hysterophorus (Parthenium) Pereskia aculeata (Barbados gooseberry)# Pinus spp. (elliottii and patula) Psidium guajava (Guava) Rubus cuneifolius (American bramble) Schinus terebinthifolius (Brazilian pepper tree)# Senna spp. (Peanut butter cassia and others) Solanum mauritianum (Bugweed) Tecoma stans (Yellow bells)* Tithonia diversifolia (Mexican sunflower)* Grassland species Acacia mearnsii, dealbata and decurrens (Black, silver and green wattle) Acacia melanoxylon (Blackwood)* Arundo donax (Giant reed)# Campuloclinium macrocephalum (Pom-pom weed) Chromolaena odorata (Triffid weed) Cotoneaster franchetii and pannosus (Orange and silver-leaf cotoneasters) Datura stramoniu (Common thorn apple)* Eucalyptus camaldulensis (Red river gum) Eucalyptus cinerea (Florist gum)# Hedychium coronarium (White ginger lily)* Ipomoea indica (Morning glory)# Opuntia sp. (Cactus pear)* Pinus elliottii (Slash pine)# Pinus patula (Patula pine)#

Page 24

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Populus x canescens and alba (Grey and white poplars) Prunus persica (peach)* Pyracantha angustifolia (Yellow fire thorn) Ricinus communis (Castor-oil-plant)* Robinia pseudoacacia (Black locust)# Rubus cuneifolius (American bramble) Salix babylonica (Weeping willow)# Salix fragilis (Crack willow)# Solanum mauritianum (Bugweed)

Page 25

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

Appendix 2: AGENDA for MPUMALANGA REGION AGENDA for MPUMALANGA REGION RANKING THE IMPORTANCE OF CRITERIA TO USE IN PRIORITISING QUATERNARY CATCHMENTS TO CLEAR OF INVASIVE ALIEN PLANT SPECIES IN PRIMARY CATCHMENTS IN MPUMALANGA

Bundu Lodge Near Rocky Drift, White River (013) 758-1221

Thursday 22nd July 2010 1. 2. 3. 4. 5. 6.

Welcome (08h30) Introduction: Aims of workshop and AHP methodology (08h45) Results of previous prioritisation studies (09h00) Discuss current rankings of important invasive alien species in Mpumalanga (09h30) Tea (10h00 – 10h20) Agree on goal, criteria (objectives) and sub-criteria (sub-objectives) for prioritising quaternaries catchments in Mpumalanga (10h20) 7. Pairwise comparisons of criteria and sub-criteria (11h30) 8. Lunch (13h00 – 13h45) 9. Continuation of pairwise comparisons of criteria and sub-criteria (13h45) 10. Tea (15h00 – 15h20) 11. Discussion of relevant and available datasets for Mpumalanga (15h20) 12. Workshop ends at approximately 16h00

Page 26

Prioritising quaternary catchments for invasive alien plant control within the Working for Water Mpumalanga region

APPENDIX 3: PARTICIPANTS IN THE EXPERT WORKSHOP PARTICIPANTS IN EXPERT WORKSHOPS Participants in the workshop held at the Bundu Country Lodge, Nelspruit on 22nd July 2010 to determine criteria and rank these to use in prioritising the clearing of invasive alien plants from quaternary catchments within Mpumalanga #

Name

Organisation

Telephone

1

Greg Forsyth

CSIR

021 888-2400 / 2609

2

Patrick O’Farrell

CSIR

021 888-2400 / 2476

3

Andrew Wannenburgh

Working for Water

021 441-2738

4

Brian Morris

5

Willem Bothma

6

Trent Sinclair

7

Mpumalanga Tourism and Park Agency

084 5797979

De Kaap Valley Conservancy

084 999 8019

Mount Anderson Catchment Nature Reserve

084 5504286

Eric Oosthuizen

Mountainlands Nature Reserve

0825757412

8

Richard Green

National: Agriculture, Forestry and Fisheries

013 7597308 083 6770839

9

Chris Foster

Komatiland Forests

10

Ronald Nenungwi

Working for Water

11

Hannes de Lange

Working for Water

082 727 2015

12

Jacobi Bach

Crocriver Mountain Conservancy

082 900 2681

13

Henning Kruger

Schoemankloof Conservatries

013 733 5333

14

Johan Brummer

Schoemankloof Conservatries

013 733 4901

15

Jeffrey Dibakoane

Working for Water

013 759 7379

Page 27