Determining the Instream Flow Requirement Monitoring Protocol
1101/1/03

Oct 2003

EXECUTIVE SUMMARY

1. Introduction
Over the last few years groups of different aquatic specialists have been involved in a process developing the Instream flow requirements (IFR's) of rivers across the country. A methodology known as "The building Block Methodology" (BBM) has been developed to assess the Instream flow requirements for any river. It can be used in any river situation and the emphasis is on identifying a complex of different magnitude flows for the maintenance of entire river ecosystems. The Department of Water Affairs (DWAF) uses the information obtained through the BBM workshops, in the planning phases of the proposed water resource developments.

Rivers that have impoundments or related developments planned in near future have been prioritised in the assigning of the recommended flows. According to the new Water Act, there is a need to monitor the effects of water development schemes. To determine the ecological effect of such developments, it is necessary to obtain pre-development data, which will provide baseline information, to which future effects and ecological findings can be compared.

This project is particularly aimed at Luvuvhu and the Sabie/Marite Rivers because both systems already have impoundments planned. Developments have already begun on these two rivers. The Injaka Dam development on the Marite tributary of the Sabie River originally aimed to be completed by the end of 2001, but this has since been postponed to the first quarter of 2002. The Nandoni (formerly known as the Mutoti) Dam on the Luvuvhu aims to be completed by September 2001, but it is also likely that this date will move forward.

IFR sites have been identified and established for both the Marite / Sabie and Luvuvhu Rivers, and the BBM workshops were held in August 1996 and July 1994 respectively. Various refinements to the IFR workshop documents have subsequently been presented, with the most recent for the Sabie being in 1997 and for the Luvuvhu in December 1998. Preliminary monitoring protocol, guidelines and objectives have also been compiled for the Luvuvhu River in December 1998 and the Sabie River in August 1997.

To date there has been considerable amount of monitoring on the Sabie River by different organisations. And some monitoring on the Luvuvhu River. Different sites along the rivers have had varying numbers of surveys completed and to different degrees. For example monitoring sites in the Kruger Park will have had more work done on them, than sites outside the protected area. On the Sabie River there have been quite comprehensive fish studies completed, and more recently macro invertebrate surveys using the SASS:4 method. Most of the monitoring that has taken place to date has been using the rapid assessment, River Health bio-monitoring techniques.

This report concentrates on the geomorphology, riparian vegetation, fish, and aquatic invertebrates.

The project started in September 1999 and was originally expected to be a two-year study. However it has since been extended to December 2001. Most of the fieldwork was completed by May / June 2001. Heavy rains in the 1999/ 2000 season resulted in flood conditions on both the Luvuvhu and Sabie Rivers. While it proved to be a fascinating occurrence, it did hinder the field trips and sampling in the rivers, and to date there are still gauging weirs that have not been re-established. Even by May 2001 the flows were still three to four times the magnitude of those recorded at the start of the project, and at the end of the 1999 dry season the flows were still considered to be above the required daily flow rates.

2.Aims and Objectives:
The main aims of the project and whether they were achieved:

  1. Base line monitoring of the Luvuvhu and Marite / Sabie Rivers. Two sites on the Luvuvhu and four sites on the Sabie / Marite Rivers were monitored. The geomorphology, riparian vegetation, fish and invertebrates were the components monitored. Under the conditions valuable baseline information was obtained.
  2. Develop and test monitoring methods that will check that the recommended ecological flows are being delivered to the different reaches of the Sabie and Luvuvhu rivers. Different monitoring / sampling techniques were tried for the riparian vegetation and fish in particular. Standard methods for the geomorphology and invertebrates were used. A first draft and proposed invertebrate flow scoring system was developed, and an idea for IFR habitat assessment.
  3. Assess whether the recommended flows are achieving the objectives for which they were designed. This objective was not possible to achieve because most of the flows for the project were higher than the recommended flows.
  4. Training / capacity building of 2 technicians. Two technicians were trained in geomorphology, riparian vegetation, fish and invertebrate sampling / monitoring methods.

All of the following components should be considered against the flow rates at the time (obtain nearest gauging weir reading and / or DWAF records), the specific habitats created at that flow rate, and the depths of the different habitats created at that particular flow.
Geomorphology Riparian Vegetation Invertebrates Fish

3. Methods used
The geomorphology was monitored with only one method, the Wadeson, Rowntree geomorphological index. It involves a desktop analysis and a site visit. It is a method designed for any monitoring programme. It is fairly quick to use (the tally of bed material size and distribution taking a little time extra), and requires minimal training. Parts of it are a little subjective.

Four different methods were used for the riparian vegetation. These were a line transect, a general survey, a belt transect and a belt transect with quadrants and indicator species. The belt transect involves recording every woody species, its height, substrate and damage rating (if any). The general survey and belt transect give valuable information for baseline purposes and an idea of the species composition in the riparian community. The line transect is a quick method and gives a summary view of species composition, but many line transects would be required to get an idea of the whole site. The belt and quadrant transect method makes use of riparian indicators, and terrestrial / riparian ecotone. It is less time consuming than a normal belt transect, and doesn't give enough information on the whole of the site, but does provide information on terrestrial encroachment to certain extent as well as information on indicator species. The way to go for long term monitoring is definitely with indicators specific to different sites across the country. Five different techniques were used for sampling the fish. They were electro shocking, a large seine net, a small seine net, a mosquito net and a cast net. The electro shocker was most broadly used and generally gives a high CPUE. It can be tiresome to use (and carry) if the channel is very wide. It is not very effective in very fast deep water, frequently kills fish fry, and if the conductivity of the water is low it is less effective. It can be used in most habitats. The large seine net can also have a large CPUE, but requires more than two people to operate it. It is not effective in medium to fast flows, and has a tendency to snag on rocks and logs in the substratum. The cast net can be very effective for catching unsuspecting adult fish (of the larger species), but it has a low CPUE, and it takes time to master the technique. The small seine is very easy to use with two people, its CPUE is variable, and is fine for shallower pools of slow or no flow. The mosquito net is very useful in a vegetated habitat of slow or no flow and not too deep, its CPUE is variable depending on the circumstances, but is usually high.

The invertebrates were sampled using the kicking method. Should any different invertebrate be caught during a fish sampling method for example shocking, then it was recorded as well. A more in depth SASS version 4 survey was used.

Sometimes it was necessary to make use of a two man inflatable boat, especially with the flows higher than average.

4. Results
Under 'normal' conditions geomorphology would not have changed much in a two year survey, but when an extreme event like the February floods takes place there are definitely geomorphological changes. Some of the (observed) sites changed less than others some changed only in terms of vegetation cover. The Luvuvhu IFR 1 site changed quite considerably in that the large downstream island changed from being mainly sand to a smaller cobble / boulder island. The Sabie IFR 4 site had the left hand channel opened up. IFR 2 on the Sabie became quite sedimented, large diagonal bedrock bars that were formerly visible are no longer exposed. The channel 1 at the Sabie IFR 5 site changed from a fairly wide slow to medium flow, and fairly deep channel to a narrow channel that has different flow velocities, but is not likely to remain perennial.

With the riparian vegetation monitoring, each site had at least one method used at the site, but it was generally two methods, and some sites had three different methods applied. There were fourty six species recorded at the Luvuvhu IFR 1 site. Luvuvhu IFR 2 site had 20 species recorded, 2 were exotic (woody) species and 2 were definite terrestrial species, the rest a combination of unknown and riparian species. Marite IFR 1 site had 41 species recorded at it, of which 6 were exotic species and 5 terrestrial species. The Sabie IFR 2 site had 31 species, of which 3 were exotics and 12 were terrestrial species. The Sabie IFR 4 site had 32 species, 7 were terrestrial and 2 exotics.

During the high flow conditions a total of eighteen fish were added to all the historic species lists, some being the same species at different sites, and some being debatable as to whether they are actually additional species, and have been recorded at the site or in the vicinity, but are not on the list. Whatever the case it is definite that additional species were recorded at some sites and it can be attributed to the different habitat conditions that have been created as a result of the floods. But then there are species that were not caught at any stage in the survey, that were caught fairly regularly in the past and are on the historic species list. This can also be attributed to a few reasons (sampling method, chance of capture, capture avoidance, and loss of habitat), but at least some of the species were not caught because of the flood conditions. Whether the species composition will stabilise or gradually change to the former, remains to be seen, much of which depends on whether previous habitats re-establish themselves. This refers particularly to the vegetation component of habitats.

The monitoring of invertebrates involved separating biotopes more specifically than a normal SASS survey, for example the stones in current biotope was separated from the stones out of current. Flows were also more refined, and were catogorised into no flow, slow flow, medium, fast and very fast flows. By distinguishing the biotopes and flows more specifically some interesting information was obtained on the biotopes that different taxa prefer, and as a result a preliminary flow scoring system was put together. SASS also applies in terms of water quality, and generally the scores were acceptable and good considering that the invertebrates were re-establishing themselves after the floods.

5. IFR Monitoring Protocol Workshop
A workshop held in August 2001, proved to be successful in most respects. Participants were experts from different aquatic fields. The main aims of the workshop were: to assist Department of Water Affairs in defining a monitoring policy for the Ecological Reserve; to design a monitoring system which will feedback into management, to review IFR monitoring methods and suggest further priorities, and to design a monitoring system which can be implemented and effective within available resources. Alison Howman of DWAF presented the Monitoring and Assessment Information System (MAIS) that DWAF are in the process of implementing. Dirk Roux presented the model of Strategic Adaptive Management (SAM), and assessment criteria, which contribute to the success of the design and implementation of an environmental monitoring programme, using the River Health Programme as a case study. It was agreed that TPC's (Thresholds of probable concern) were appropriate, but the TPC's for each component of the monitoring programme need to be defined and refined. A process diagram was devised to represent the processes involved in IFR monitoring and the proposed steps to be taken if a TPC is exceeded. The workshop participants then divided into different speciality groups: geomorphology; riparian vegetation; invertebrates and fish. Each group produced a proposed monitoring programme with different levels of detail.

6. Conclusion
While the February 2000 floods may have burdened the project to a certain degree (loss of cross sections, loss of gauging weirs and continued high flow conditions for most of the project), there has certainly been interesting and valuable information to be gained as well. With both the Injaka and Nandoni dams to be closed soon, there has now been some additional baseline monitoring at some sites. While it was not the objective of the project to monitor the responses of biota to high flow conditions, nor were sufficient sites monitored to gain a broader spectrum of information, the information collected is valuable under the circumstances. Initially it was thought that perhaps a fish scoring system needs to be developed, something similar to SASS for example, but flow related. But after the IFR Monitoring Protocol Workshop it was evident from discussions that the current RHP method FAII (Fish assemblage integrity index) is more than sufficient. A great deal of previous research on macro-invertebrate communities in the Sabie River in particular, identified the organisms to species level. At the out start of the project it was considered an option to do the same. However it was decided that for the present, that to the existing taxon level (mainly families) was sufficient. Perhaps certain indicator species could be identified at a site rating for a river and specific site, that could be used in the future long term IFR monitoring.

7. Recommendations