Water temperature and fish distribution in the Sabie River system : towards the development of an adaptive management tool.

35 Figure 2.11 Average annual daily range in water temperatures as a function of dew-nflow distance on the Sabie River. The results for three different scenarios for daily maximum water temperatures are presented 212 Figure 8.11 The relationship between C.

Figure 1.1 The Sabie catchment, showing main rivers, towns and nature reserves. The zone where the transition between cold-water and warm-water fish assemblages on the Sabie River occurs, as identified by Weeks et al

Aims and objectives

Assess how the findings from this model contribute to the tools available to natural resource managers in the Kruger National Park. Hollow: Water temperature is an important determinant of the spatial distribution of fish species in the Sabie River system;

Wider context to this research project

Thus, there is a lag between water temperature data collection and fish data collection. The relationship between water temperature (abiotic process) and fish distribution (biotic pattern) is shown in the fourth chapter.

Figure 1.2 Major goals within Ph.D. thesis , as divided within Part I and 11.

Introduction

Components of water temperature

Despite numerous papers on the relationships between temperature and aquatic organisms, as well as a large amount of research on the physical mechanisms controlling natural water temperature variations, there is limited understanding of the temperature conditions of natural rivers (Smith 1979), especially in southern areas. hemisphere (Ward 1985). This certainly applies to the rivers in the Sabie River basin, where there is limited insight into the intra-annual dynamics of water temperature (Jewitt et al.

Water Temperature Regime

Water temperature and aquatic biota

Figure 2.2 (a) Flow variability (coefficient of variation) within 56 sub-catchments, as determined by Pike and Schulze (2000), of the Sabie catchment. 7 WT8 installed on 6 February 2001, at the confluence of the Lubyelubye and Sabie rivers within the Kruger National Park.

Figure 2.2 (a) Flow variability (coefficient of variation) within 56 subcatchments, as defmed by Pike and Schulze (2000), of the Sabie catchment

Data collection

Hourly relationship curves of water temperatures were calculated for sites on the upper (WT9), “middle” (WT3) and lower (WT8) reaches of the Sabie River. The same seasonal pattern was also evident in the mean annual water temperatures along the longitudinal axis of the Sabie River (Figure 2.15).

Figure 2.3 Fibreglass float with attached thennistors for collection of vertical water temperature profile

Discussion and conclusions

These zones coincide with the zones identified by Weeks et al. 1996), who grouped the fish species composition of the Sabie River into two groups; a cool water group (foot zone) and a warm water group (low field zone). This chapter has illustrated the complex nature of water temperature in the main rivers of the Sabie catchment.

Introduction

Linear regression analysis
Process-based approaches
Dynamic water temperature model
Dynamic water temperature model

Hourly water temperature data were calibrated (cf. section 2.2.1), and used to calculate daily maximum water temperatures. Daily maximum water temperatures were simulated using five different modeling methods at all water temperature monitoring sites (cf. Table 2.1). The second group of models simulated water temperatures using air temperatures unique to each water temperature location.

Autocorrelations on daily maximum water temperatures were performed using the time series analysis routines in Genstat (Genstat 2000). Water temperatures (Tjinal) were a function of the temperature of water entering a reach and mixing with the existing water in the reach (TmLt), and the amount of heat gained or lost over time (t) due to heat- exchange with air (Tdiff ) in that range. Daily maximum water temperatures were approximated with Equation 3.10 (eg Section 3.2.1.3) using the values provided in Table 3.3, for the specific water temperature station used.

Table 3.1 Water temperature loggers and gauging weirs used in flow-dependant multiple linear regression model.

Discussion and conclusions

Physical versus statistical spproaches

Despite an element of non-causality between air and water temperatures, their statistical relationship remains a useful way of predicting water temperatures. However, this should not lead to complacency about the current state of water temperature simulation. These complexities may explain the relatively poor correlations between observed and simulated water temperatures for the upper parts of the Sabie catchment (Table 3.6).

Although the simple linear regression model is the most pragmatic approach to simulating daily maximum water temperatures, this model underpredicted and water temperatures should be adjusted accordingly. This chapter focused on the two main approaches to simulating water temperatures (process-based versus statistical). Simple correlative statistical models were found to be the most pragmatic approach for simulating maximum daily water temperatures.

Introduction

Water temperature and river species patterns
Ordination techniques in explaining species patterns

Webb (1996) provides a comprehensive review of the effect of water temperature on the physical and chemical properties of water. Diversity indices combine the number of species with their relative abundance, making it possible to quantify localities of species, although the resulting reduction in the number of variables is also one of the criticisms of diversity indices. In this respect, canonical correspondence analyzes (CCA) are more suitable for exploring species-environment relationships (ter Braak 1987), and consequently this technique was chosen for fish-environment relationships in the main rivers of the Sabie catchment in this study (c. Section 4.2.2) .

NI is based on Shannon and Weaver's index (H') (Ludwig and Reynolds 1988) and is "a measure of the average degree of 'uncertainty' in predicting what species an individual randomly selected from a is. Note that Pi = ~, and is proportional abundance of the ith species, where n is the number of individuals of the ith species and N is the total number of species at each site.For temperature, values calculated from the average of data from 2001 and 2002 were used.statistics for 2000.

Table 4.1 Site information relating to May electrofishing surveys

Results

Species patterns in the Sabie River using diversity indices
Quantification of niche dimensions using the niche hypervolume concept The first approach was based on the niche hypervolume concept. These were represented

For three consecutive years, similar trends were found in the downstream gradient of the Sabie River. The selection of appropriate indicators of changes in annual water temperatures in the Sabie River is discussed in this chapter through analyzes of niche dimensions of five fish species occurring in the lower Sabie River. 1996) proposed that species of the genus Chiloglanis are good indicators of changes in cumulative heat units in the Sabie River.

Populations showed slow recovery to adverse flow conditions, based on the sampling conducted in the early 1990s in the rivers of the Sabie catchment (Weeks et al. 1996). In the context of this study, the most important abiotic determinant of fish community patterns in the rivers of the Sabie catchment was water temperature. The results of the analyzes of niche dimensions of the five fish species considered from the Sabie River are discussed below.

Table 4.7b Diversity (N) and evenness (E) indices for May 2001 electrofishing survey. The di versity indices No, NI and N] are measures of the total, abundant and ve abundant s ecies respectively, at each site.

Discussion and conclusions

6 TPC FOR CUMULATIVE ANNUAL WATER TEMPERATURE IN THE SABLE RIVER AND ITS ESTIMATION USING BIOLOGICAL INDICATORS. Furthermore, the fish species Chiloglanis anoterus and C.paratus (Mochokidae) were shown to be sensitive to annual changes in thermal units in the Sabie River due to their small size and niche dimensions. Water temperature has been shown to be one of the primary variables explaining the distribution of fish in the Sabie River (cf. Chapter 4), and consequently an important driver of change.

In this chapter, a spatially explicit TPC is proposed for changing the annual water temperature range in the Sabie River (cf. section 6.4.3), where exceedance is measured using biological indices (cf. section 6.1.3.1-2). These biological indices are based on the response of two species of Chiloglanid fish to annual changes in water temperatures and can be used to indicate a significant change in the thermal regime of the Sabie River by measuring the response of the species to annual water temperatures over time at a specific spatial point. An appropriate monitoring method to detect changes in the annual thermal regime of the Sabie River.

Figure 6.1 The adaptive management process for the Kruger National Park. Arrow numbers define the sequence in which steps are taken (Rogers and Biggs 1999).

Methods for evaluating Chiloglanid biological indices

C. anoterus condition factor index

Logistic regressions were calculated using the presence (1) or absence (0) of C.anoterus at 16 locations, to calculate the probability of C occurrence. Two approaches were taken to determine the degree of species affinity that may exist between C.anoterus and C. A second qualitative approach was used to test the degree of significance of species covariation, and in particular to determine whether there was a significant negative was related. between C.anoterus and C.paratus.

The next step in using the condition factor as an index of thermal change was to determine whether there was a trend between the condition factor and downstream distance (km) along the longitudinal axis of the Sabia River. Box-and-whisker plots were used to illustrate the relationship between the variability of the condition factor and downstream distance. The effect of sampling date and gender on the fitness factor was tested using one-way analysis of variance (without blocking).

Table 6.1 Chiloelanis electro IS mz site names an eo-or ma e m orma I

Results

Annual frequency of MWAT exceedance was plotted as a function of downstream distance (Figure 6.4); this relationship could be approximated by the appropriate logistic equation (Equation 6.5). The slope of the regression lines using Equation 6.3a was more similar (greater constancy) (Figure 6.6b) than the regression slopes derived using alternative ratio calculations;. A box-and-whisker plot (Figure 6.9) of the condition factor (as a pooled data set from all four electrofishing surveys; n = 629) versus distance downstream along the longitudinal axis of the Sabie River showed that while the range in condition was greatest in upstream areas , median condition followed a decreasing trend, where condition was highest at the top of the watershed and decreased with distance downstream.

Based on the trend in length and mass with distance downstream (Figure 6.8), this is more likely to be a function of a change in mass than a change in length. Mean, median, and mode factor values, calculated using combined data from all sampling occasions, were each regressed against the lower distance to determine which statistic provided the best prediction (or correlation) of the index. of the condition factor (Figure 6.10). The relationship between relative abundance and mean condition for the May 2001 data was approximated by a third-order polynomial (Figure 6.12), although this relationship was not significant, suggesting that condition and abundance are not correlated.

Figure 6.3a Relati ve abundance curves of C. anoterus and C. paratus for May 2001 electrofishing survey

Discussion and conclusions .1 General considerations

Modelling approaches for natural systems
Modelling and adaptive management
Objectives and conceptual approach

The MWAT exceedance of 25 °C for 200 days coincided with site WT4 (Chapter 2), located 106 kilometers downstream of the head of the Sabie catchment. The importance of the relationship between water temperature and fish diversity in the Sabie River has been emphasized (cf. Chapter 4). Ecological processes can be simulated at different scales within the same model, depending on the purpose of the model (Matsinos et al. 1994; Mooij and Boersma 1996).

Ecological processes can be modeled at different scales, depending on the purpose of the model. One of the management objectives of the Kruger National Park is the conservation of biodiversity (Rogers and Bestbier 1997). Biological indices, such as the condition of Chiloglanis anoterus, and the ratio of the relative abundances of C.