Physico-chemical characteristics of South African estuaries in relation to the zoogeography of the region

Estuarine, Coastal and Shelf Science 61 (2004) 73e87 www.elsevier.com/locate/ECSS

Physico-chemical characteristics of South African estuaries in relation to the zoogeography of the region T.D. Harrison) South African Institute for Aquatic Biodiversity, Private Bag 1015, Grahamstown 6140, South Africa Received 20 October 2003; accepted 5 April 2004

Abstract Selected physico-chemical characteristics of estuaries within three biogeographic regions of South Africa were described and compared using data from an extensive ichthyological survey of the region. The results indicate that estuaries within each zoogeographic region have distinctive physico-chemical characteristics that could be related to regional variations in climate and oceanographic conditions. Estuarine temperatures follow the trend for marine coastal waters, decreasing from the subtropical east coast, along the warm-temperate south coast and up the cool-temperate west coast. Low rainfall and runoff together with high seawater input and evaporative loss results in high salinities and low turbidities in warm-temperate estuaries while high rainfall and runoff results in reduced salinities and high turbidities, particularly in subtropical estuaries. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: estuaries; physico-chemical parameters; biogeography; South Africa

1. Introduction The South African coastline stretches for some 3400 km from the Orange River mouth (28( 38#S; 16( 27#E) on the west (Atlantic Ocean) coast to Kosi Bay (26( 54#S; 32( 48#E) on the east (Indian Ocean) coast (Fig. 1). Over 300 outlets intersect the coastline and these range from relatively large, permanently open estuaries to small coastal streams. Since estuaries are formed where rivers meet the sea, they are affected by variations in both terrestrial and marine conditions (Day, 1981; Cooper, 2001). The coastal region of South Africa spans a number of climatic zones from humid, predominantly summer rainfall conditions in the east and northeast to a Mediterranean climate in the south and southwest, and arid conditions in the northwest (Schulze, 1984; Tinley, 1985). The KwaZulueNatal coast has a warm and humid climate with an average annual rainfall

) Present address: Environment and Heritage Service, Commonwealth House, 35 Castle Street, Belfast BT1 1GU, Northern Ireland. E-mail address: [email protected] 0272-7714/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.ecss.2004.04.005

exceeding 1000 mm, most of which falls in summer (OctobereMarch) (Day, 1981). Toward the Eastern Cape, rainfall decreases to approximately 500 mm per year and occurs almost equally in all seasons although slightly higher rainfall occurs during autumn (March) and spring (October/November) (Heydorn and Tinley, 1980; Schulze, 1984). The southwest Cape, from Cape Agulhas to Cape Columbine, has a Mediterranean climate with cool, wet winters and hot, dry summers (Schulze, 1984; Tinley, 1985). Most rainfall occurs from May to September and is usually between 400 and 700 mm per year (Heydorn and Tinley, 1980; Day, 1981). The west coast of South Africa, north of Cape Columbine, has an arid climate and receives less than 300 mm rainfall per year (Heydorn and Tinley, 1980; Day, 1981), predominantly in winter (Tinley, 1985). The coastal waters of southern Africa are influenced by two major current systems. The east coast, which borders the Indian Ocean, is influenced by the southflowing Agulhas Current. Being tropical in origin, the waters of the Agulhas Current are warm but as it flows south it moves offshore and coastal sea temperatures decline. Average surface sea temperatures off Durban (KwaZulueNatal) exceed 22 (C while off Port Elizabeth

74

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

Fig. 1. Map of South Africa indicating place names mentioned in the text.

(Eastern Cape) they are below 19 (C (Lutjeharms et al., 2000). The west (Atlantic) coast is influenced by the cold, north-flowing Benguela Current (Tinley, 1985; Shannon, 1989). The surface water temperatures of the Benguela system average between 13 and 15 (C; offshore winds result in a pronounced upwelling ‘season’ during the summer (SeptembereMarch) (Shannon, 1989). Features such as climate, geomorphology, and tidal and fluvial patterns play a major role in determining the chemical properties of South African estuaries (Day, 1981; Allanson, 1999). While a number of studies have examined physico-chemical aspects of South African estuaries at the sub-regional scale (e.g. Read, 1983; Begg, 1984a; Emmerson, 1985, 1989; Taljaard et al., 1986; Emmerson and Erasmus, 1987; Cyrus, 1988a, 1988b; Harding, 1994; Slinger and Taljaard, 1994; Scharler et al., 1997) few, if any have considered these systems at a national scale. The aim of this paper is, using data collected during an extensive ichthyofaunal survey, to describe and compare the physico-chemical characteristics of South African estuaries in relation to the zoogeography of the region.

2. Materials and methods 2.1. Field methods An extensive ichthyofaunal survey of South African estuaries was undertaken over the period 1993e1999. Some 250 systems between the Orange River and Kosi Bay were sampled. The sampling regime divided the coastline into arbitrary sections, each containing approximately 40 estuaries; the estuaries in each section of coast were sampled during the spring/summer period and a new section was sampled each year until the entire coastline was covered. During these ichthyofaunal surveys, a number of physico-chemical parameters were measured at various sites within each estuary. The number of sampling sites varied depending on the size of each estuary but generally included the lower, middle and upper reaches of each system. Water depth and transparency was measured using a 20 cm diameter Secchi disc attached to a weighted shot line graduated at 10-cm intervals. Temperature ((C), salinity (Practical Salinity Scale), dissolved oxygen (mg l
1) and turbidity (Nephelometric

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

Turbidity Units (NTU)) was recorded using a Horiba U-10 Water Quality Checker. Where depth permitted, both surface and bottom waters were measured. The mouth condition (open/closed) of each system at the time of sampling was also noted. 2.2. Estuary classification Since the aim of this paper is to describe and compare the physico-chemical characteristics of South African estuaries at a regional scale, intra-regional differences were accounted for by selecting representative estuaries according to a broad agreement between two physical/ morphological classification schemes. Whitfield (1992, 2000) classified South Africa’s estuaries into five broad types based on a combination of physiographic, hydrographic and salinity features while Harrison et al. (2000) classified South Africa’s estuaries into six categories based on the main forms of morphological variability among these systems along the coast. A total of 109 estuaries were selected for the current study; these were divided into two basic types namely ‘open’ estuaries and ‘closed’ estuaries. Forty-two systems were open estuaries and comprised representatives of the permanently open estuaries of Whitfield (2000) and the large open estuaries of Harrison et al. (2000). Sixty-seven estuaries were closed systems representing the temporarily open/closed estuaries of Whitfield (2000) and medium-sized closed estuaries of Harrison et al. (2000). The selected open and closed estuaries were then divided into three biogeographic regions described by Harrison (2002). These included cool-temperate estuaries between the Orange River and Cape Agulhas, warmtemperate estuaries from Cape Agulhas to, and including, the Mdumbi estuary, and subtropical estuaries from the Mdumbi estuary to Kosi Bay (Figs. 2 and 3). 2.3. Data analysis The mean (G SE) depth (m), water temperature ((C), salinity, dissolved oxygen (mg l
1), and turbidity (NTU) were calculated for each estuary within each biogeographic region. Because disturbance of the sediments during field sampling often resulted in high bottom water turbidity values, only surface measurements were considered. Furthermore, in some warm-temperate estuaries, turbidity readings were not taken. In these cases turbidity values were estimated from mean Secchi disc measurements taken at the time of sampling using the formula derived by Cyrus (1988a): SD ¼ 135:0
26:8 ln T where SD is the Secchi disc measurement (cm) and ln T is the natural logarithm of the turbidity value (NTU).

75

The physico-chemical data were also subject to principal component analysis (PCA) using the Plymouth Routines In Multivariate Ecological Research (PRIMER). Principal component analysis is a multivariate ordination method that produces a low-dimensional summary of the inter-relationships between a number of variables and is most suited to environmental data (Clarke and Warwick, 1994). Prior to conducting the PCA, the physico-chemical parameters were first examined for normality. Only turbidity values required log-transformation ðln½1CxÞ. The data were also tested for any inter-correlations (Pearson r). In both closed and open estuaries, dissolved oxygen was significantly negatively correlated with temperature (p ! 0:05); dissolved oxygen was therefore omitted from the analysis. A PCA, based on the normalised Euclidean distance, was then performed on closed and open estuaries separately using the following parameters: depth, temperature, salinity and turbidity. Clarke and Warwick (1994) have suggested that a PCA that accounts for approximately 70e75% of the original variation provides an adequate description of the overall structure of the inter-relationships. For the results of the PCA, the estuaries were labelled according to their zoogeographic position where: C = cool-temperate estuaries; W = warm-temperate estuaries; S = subtropical estuaries. In addition to a PCA, an analysis of similarities (ANOSIM) was also performed to test for any significant differences between estuaries from each biogeographic region. This analysis utilises the (rank) similarity values (based on normalised Euclidean distance) underlying the PCA ordination procedure and tests for differences between and within a priori groupings (Clarke and Warwick, 1994). A test statistic (R) is computed, which reflects the observed differences between groupings, contrasted with differences within groupings. The R statistic usually falls within the range 0 and 1; if R ¼ 1 then all sites within a group are more similar to each other than any sites from different groups and if R ¼ 0 then the similarities between and within groups are the same on average (Clarke and Warwick, 1994).

3. Results 3.1. Physico-chemical characteristics 3.1.1. Closed estuaries Four closed estuaries were represented in the cooltemperate region (Fig. 2). One system, the Diep estuary was open at the time of this survey. Average water depths ranged between 0.5 and 1.2 m with most systems not exceeding 1.0 m. Mean water temperatures varied between 16.9 (C in the Diep estuary and 20.8 (C in the Sand system. Salinities measured between 2.3 in the Wildevoe¨l system and 21.5 in the Diep estuary; generally,

76

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

Fig. 2. Map of closed estuaries in the cool-temperate, warm-temperate and subtropical biogeographic regions included in this study. The relative positions of the closed estuaries included in the study are indicated with arrows.

mean salinities did not exceed 15. Mean dissolved oxygen levels ranged between 4.9 mg l
1 in the Krom estuary and 8.7 mg l
1 in the Wildevoe¨l system. Mean surface water turbidities varied between 6 and 20 NTU in the Krom and Sand estuaries, respectively; most estuaries had turbidities below 15 NTU (Table 1). Forty-one closed estuaries were included in the warmtemperate region (Fig. 2). Seven systems (Tsitsikamma, Mcantsi, Kwenxura, Nyara, Haga-Haga, Morgan, and Qolora) were open at the time of this study. Mean water depths ranged from 0.4 m in the Haga-Haga estuary to 3.4 m in the Qolora system; most estuaries were between 1.0 and 2.0 m deep. Average water temperatures

varied between 16.3 (C in the Groot (Wes) system and 26.1 (C in the Kasuka estuary, with most values between 18.0 and 24.0 (C. Mean salinities ranged between 0.6 in the Tsitsikamma estuary and 49.3 in the Gqutywa system; overall, salinities generally ranged between 15 and 30. Mean dissolved oxygen levels measured between 3.5 mg l
1 in the Nyara estuary and 11.0 mg l
1 in the Kiwane system. Dissolved oxygen concentrations were mostly in the range 5e8 mg l
1. Mean surface water turbidities varied between 0 and 100 NTU recorded in the Groot (Wes) and Morgan estuaries, respectively. Most estuaries had turbidities below 5 NTU (Table 2).

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

77

Fig. 3. Map of open estuaries in the cool-temperate, warm-temperate and subtropical biogeographic regions included in this study. The relative positions of the closed estuaries included in the study are indicated with arrows.

Twenty-two closed estuaries were represented in the subtropical region (Fig. 2). Six systems (Kandandlovu, Mpenjati, Little Manzimtoti, Manzimtoti, Mhlanga, and Siyai) were open at the time of this study. The mean water depth ranged between 0.5 m in the Mhlanga estuary and 2.9 m in the Mdlotane system with most estuaries having water depths of 1e2 m. Mean water temperatures between 20.7 and 28.2 (C were recorded in the Mhlanga and Fafa estuaries, respectively. Most estuaries had water temperatures exceeding 24 (C. Salinities ranged between 0.1 in the Mdloti estuary and 16.1 in the Siyai system. Overall, mean salinities did not exceed 15; a high proportion of systems had values

below 5. Mean dissolved oxygen levels measured between 1.7 mg l
1 in the Mdlotane estuary and 7.4 mg l
1 in the Mtentwana system. Approximately 40% of the estuaries had dissolved oxygen levels below 5.0 mg l
1. Surface water turbidities ranged between 4 and 58 NTU in the Mtentweni and Zinkwasi estuaries, respectively. Overall, turbidity values were mainly within the range 5e10 NTU (Table 3).

3.1.2. Open estuaries Four open estuaries were represented in the cooltemperate region (Fig. 3). Average water depths ranged

78

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

Table 1 Mean physico-chemical parameters recorded in closed cool-temperate estuaries (number of samples; G SE) System

Depth (m)

Diep Wildevoe¨l Krom Sand

0.77 0.45 0.53 1.23

(4; (3; (2; (3;

0.18) 0.12) 0.21) 0.47)

Temperature ((C)

Salinity

16.86 19.75 19.25 20.78

21.53 2.28 9.08 11.32

(8; (6; (4; (6;

0.87) 0.78) 0.13) 0.03)

between 0.8 m in the Uilkraals system and 2.7 m in the Berg estuary. Mean water temperatures did not exceed 18.0 (C and varied between 14.0 (C in the Olifants estuary and 17.3 (C in the Uilkraals system. Mean salinities between 15.4 and 20.0 were recorded in the Uilkraals and Berg estuaries, respectively. Dissolved

(8; (6; (4; (6;

4.51) 1.03) 0.20) 0.78)

Dissolved oxygen (mg l
1)

Turbidity (NTU)

5.69 8.68 4.92 7.90

14.25 6.67 6.00 20.00

(8; (6; (4; (6;

0.69) 0.30) 0.18) 0.54)

(4; (3; (2; (3;

8.67) 3.71) 4.00) 4.04)

oxygen levels ranged between 6.8 mg l
1 in the Berg estuary and 8.7 mg l
1 in the Olifants system with most estuaries having values exceeding 8 mg l
1. Average surface water turbidities ranged between 4 NTU in the Palmiet estuary and 31 NTU in the Olifants estuary; most values were below 15 NTU (Table 4).

Table 2 Average physico-chemical parameters recorded in closed warm-temperate estuaries (number of samples; G SE) System

Depth (m)

Blinde Hartenbos Groot (Wes) Tsitsikamma Seekoei Kabeljous Van Stadens Boknes Kasuka Riet Wes-Kleinemond Oos Kleinemond Old Woman’s Mpekweni Mtati Mgwalana Bira Gqutywa Mtana Ngqinisa Kiwane Ross’ Creek Ncera Mlele Mcantsi Gxulu Goda Hickmans Qinira Cintsa Cefane Kwenxura Nyara Haga-Haga Morgan Gxara Ngogwane Qolora Cebe Zalu Ngqwara

1.77 1.35 1.97 0.65 1.08 1.00 2.20 1.80 1.53 1.67 1.92 1.23 2.30 1.74 2.20 1.04 1.70 0.98 0.90 0.37 1.95 0.57 1.22 1.07 1.03 1.37 1.97 1.60 1.70 1.38 0.93 0.80 0.55 0.35 0.57 1.30 1.47 3.40 1.20 1.10 0.90

(3; (4; (3; (2; (4; (3; (4; (3; (4; (3; (6; (4; (3; (5; (5; (5; (7; (4; (4; (3; (4; (3; (5; (3; (3; (6; (3; (3; (4; (4; (4; (3; (2; (2; (3; (3; (3; (3; (3; (3; (3;

0.30) 0.25) 0.13) 0.25) 0.23) 0.06) 0.46) 0.12) 0.25) 0.28) 0.13) 0.21) 0.57) 0.20) 0.22) 0.11) 0.21) 0.24) 0.15) 0.13) 0.52) 0.23) 0.16) 0.09) 0.03) 0.09) 0.23) 0.21) 0.11) 0.13) 0.18) 0.38) 0.25) 0.15) 0.07) 0.40) 0.20) 0.44) 0.35) 0.36) 0.15)

Temperature ((C)

Salinity

Dissolved oxygen (mg l
1)

Turbidity (NTU)

20.85 20.19 16.28 17.73 19.26 17.65 20.69 21.12 26.11 23.98 23.20 23.00 17.42 18.14 17.52 19.42 19.99 20.60 19.28 17.20 18.45 18.94 20.46 19.77 23.03 22.06 21.00 21.98 21.50 23.78 23.50 19.70 19.03 21.33 21.98 21.37 22.62 21.28 24.22 20.58 21.90

12.33 17.64 13.93 0.63 6.13 16.45 14.55 16.67 24.84 20.02 18.58 15.13 25.87 19.57 19.47 27.99 29.76 49.34 29.40 31.53 20.16 6.58 33.66 15.30 14.10 29.64 32.55 18.83 28.01 31.13 28.90 29.10 21.77 25.80 8.63 18.97 20.58 16.52 25.67 20.62 25.58

5.46 7.05 4.29 8.45 10.59 5.45 6.07 7.93 4.41 4.51 4.92 6.37 6.68 8.65 7.56 9.50 8.14 5.82 7.60 8.81 11.01 6.62 7.70 7.51 6.96 6.41 6.09 6.42 6.40 5.53 5.00 6.98 3.50 8.01 8.37 5.71 5.64 4.52 6.74 6.25 6.41

7.67 (3; 3.00 (4; 0.00 (3; 34.6* 9.4* 2.5* 0.0* 1.8* 0.8* 1.2* 1.1* 3.7* 2.00 (3; 8.40 (5; 8.40 (5; 28.20 (5; 14.14 (7; 15.75 (4; 9.50 (4; 15.00 (3; 6.25 (4; 20.00 (3; 2.40 (5; 16.00 (3; 6.67 (3; 3.17 (6; 5.33 (3; 3.67 (3; 4.50 (4; 3.50 (4; 6.00 (4; 35.67 (3; 94.50 (2; 23.00 (2; 100.33 (3; 16.50 (2; 0.33 (3; 4.33 (3; 11.33 (3; 4.33 (3; 1.00 (3;

(6; 0.27) (8; 0.43) (6; 0.55) (3; 0.37) (8; 0.44) (6; 0.18) (8; 0.42) (6; 0.22) (8; 0.25) (6; 0.22) (12; 0.20) (8; 0.51) (6; 0.17) (10; 0.21) (10; 0.14) (10; 0.25) (14; 0.15) (8; 0.19) (8; 0.19) (3; 0.45) (8; 0.16) (5; 0.18) (10; 0.13) (6; 0.04) (6; 0.16) (12; 0.29) (6; 0.19) (6; 0.13) (8; 0.14) (8; 0.40) (8; 0.32) (5; 1.11) (3; 0.80) (3; 0.55) (6; 0.45) (6; 0.31) (6; 0.42) (6; 0.48) (6; 0.14) (6; 0.16) (6; 0.54)

(6; 0.40) (8; 0.71) (6; 3.47) (3; 0.23) (8; 0.79) (6; 0.27) (8; 0.58) (6; 0.13) (8; 0.39) (6; 1.88) (12; 0.44) (8; 0.52) (6; 0.07) (10; 0.14) (10; 0.42) (10; 0.35) (14; 0.37) (8; 0.08) (8; 0.28) (3; 0.03) (8; 0.06) (5; 0.09) (10; 0.24) (6; 0.04) (6; 0.29) (12; 0.20) (6; 0.04) (6; 0.80) (8; 0.20) (8; 0.30) (8; 0.17) (5; 3.10) (3; 7.49) (3; 0.42) (6; 5.32) (6; 1.26) (6; 0.32) (6; 4.33) (6; 0.08) (6; 0.11) (6; 0.26)

Turbidity values estimated from Secchi disc measurements are denoted by an asterisk (*).

(6; 0.67) (8; 0.44) (6; 1.00) (3; 0.24) (8; 0.69) (6; 0.45) (8; 0.51) (6; 0.36) (8; 0.30) (6; 0.85) (12; 0.40) (8; 0.19) (6; 0.12) (10; 0.56) (10; 1.03) (10; 0.44) (14; 0.29) (8; 0.17) (8; 0.11) (3; 0.11) (8; 0.08) (5; 0.55) (10; 0.65) (6; 0.09) (6; 0.51) (12; 0.21) (6; 0.21) (6; 0.77) (8; 0.25) (8; 0.21) (8; 0.08) (5; 0.38) (2; 0.20) (3; 0.04) (6; 0.16) (6; 0.61) (6; 0.44) (6; 1.50) (6; 0.07) (6; 0.31) (6; 0.53)

2.19) 0.41) 0.00)

1.00) 1.86) 1.94) 3.60) 3.83) 2.02) 2.33) 6.43) 1.65) 0.00) 1.17) 0.00) 1.86) 1.05) 2.33) 0.33) 0.96) 0.50) 2.12) 18.52) 44.50) 9.00) 6.49) 6.50) 0.33) 2.19) 4.48) 2.33) 0.58)

79

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87 Table 3 Average physico-chemical parameters recorded in closed subtropical estuaries (number of samples; G SE) System

Depth (m)

Mtentwana Kandandlovu Mpenjati Umhlangankulu Kaba Mbizana Bilanhlolo Mhlangeni Mtentweni Mhlangamkulu Intshambili Fafa Sezela Mpambanyoni Mahlongwa Little Manzimtoti Manzimtoti Mhlanga Mdloti Mdlotane Zinkwasi Siyai

1.80 1.00 1.77 2.00 0.85 2.13 1.37 1.07 1.50 1.68 1.15 0.73 1.80 0.68 1.35 1.23 1.05 0.45 2.33 2.93 1.30 1.18

(3; (2; (3; (3; (2; (4; (3; (3; (3; (2; (3; (3; (3; (3; (3; (3; (3; (3; (3; (3; (4; (2;

0.51) 0.30) 0.09) 0.29) 0.25) 0.22) 0.27) 0.15) 0.17) 0.18) 0.22) 0.34) 0.29) 0.28) 0.36) 0.09) 0.08) 0.05) 0.47) 0.18) 0.11) 0.73)

Temperature ((C)

Salinity

28.05 24.48 22.22 23.37 26.73 24.70 26.37 27.03 25.57 22.90 25.40 28.16 25.70 24.14 26.77 20.83 21.83 20.67 21.83 28.07 27.73 26.10

14.20 11.38 10.10 4.05 12.08 3.49 10.53 13.60 10.02 0.50 3.17 2.34 4.27 4.46 2.30 1.45 1.70 9.57 0.10 0.23 12.25 16.13

(6; (4; (6; (6; (4; (8; (6; (6; (6; (4; (6; (5; (6; (5; (6; (6; (6; (3; (6; (6; (8; (4;

0.47) 0.59) 0.55) 0.31) 0.81) 0.43) 0.38) 0.43) 0.92) 0.39) 0.26) 0.36) 0.12) 1.09) 0.10) 0.34) 0.15) 0.20) 0.16) 0.78) 0.44) 0.30)

Twenty-eight open estuaries were represented in the warm-temperate region (Fig. 3). Average water depths ranged between 1.0 m in the Qora and Shixini estuaries to 3.4 m in the Buffalo estuary; the water depth of the vast majority of estuaries exceeded 1.5 m. Mean water temperatures ranged between 14.9 (C in the Heuningnes estuary and 23.0 (C in the Kariega system and were mostly within the range 18e22 (C. Average salinities varied between 4.4 in the Great Kei estuary and 33.6 in the Gourits estuary; the salinity of most estuaries fell within the range 25e35. Mean dissolved oxygen values ranged between 5.2 mg l
1 in the Keurbooms estuary and 8.5 mg l
1 in the Nahoon estuary; most estuaries had dissolved oxygen concentrations between 6.0 and 8.0 mg l
1. Average surface water turbidities fell between 1 and 1300 NTU recorded in the Keurbooms and Great Kei estuaries, respectively; most estuaries had turbidities of less than 10 NTU (Table 5). Ten open estuaries were represented in the subtropical region (Fig. 3). The average water depth ranged from 0.9 m in the Mkomazi estuary to 4.1 m in the

(6; (4; (6; (6; (4; (8; (6; (6; (6; (4; (6; (5; (6; (5; (6; (6; (6; (3; (6; (6; (8; (4;

1.63) 4.35) 2.66) 0.03) 0.57) 0.45) 3.79) 2.44) 2.55) 0.00) 0.06) 0.49) 0.09) 0.99) 0.29) 0.65) 0.51) 7.57) 0.00) 0.02) 0.76) 7.72)

Dissolved oxygen (mg l
1)

Turbidity (NTU)

7.43 4.30 5.62 5.94 7.36 6.15 6.05 6.94 5.39 4.70 5.02 6.83 3.11 6.91 6.10 2.31 5.24 4.22 2.93 1.74 4.40 4.25

11.00 8.00 8.00 9.00 9.50 5.00 5.00 13.00 4.33 9.50 9.33 8.67 18.67 18.33 10.00 26.33 49.33 36.67 8.33 10.67 58.25 19.50

(6; (4; (6; (6; (4; (8; (6; (6; (6; (4; (6; (5; (6; (5; (6; (6; (6; (3; (6; (6; (8; (4;

0.49) 0.63) 1.23) 0.26) 0.32) 0.26) 0.67) 1.09) 0.79) 0.97) 0.44) 0.28) 0.35) 1.58) 0.15) 0.48) 1.41) 1.00) 1.04) 0.82) 0.65) 1.41)

(3; (2; (3; (3; (2; (4; (3; (3; (3; (2; (3; (3; (3; (3; (3; (3; (3; (3; (3; (3; (4; (2;

2.08) 5.00) 0.58) 1.00) 1.50) 0.41) 0.00) 1.53) 0.67) 0.50) 0.33) 1.33) 4.18) 4.48) 0.58) 6.98) 7.45) 14.53) 1.45) 1.20) 2.10) 9.50)

Msikaba system; most estuaries had water depths exceeding 2.0 m. Mean water temperatures varied between 23.5 (C in the Mngazana, Mzimkulu and Matigulu/ Nyoni estuaries and 27.0 (C in the Mlalazi system. The water temperature of most estuaries fell within the range 22e28 (C. Average salinities ranged from 3.0 in the Matigulu/Nyoni estuary to 28.5 in the Mngazana system with most estuaries having salinities of below 20. Mean dissolved oxygen levels ranged between 5.2 mg l
1 in the Matigulu/Nyoni estuary and 7.4 mg l
1 in the Mtentu system. Average surface water turbidities varied between 5 NTU in the Mtentu estuary and 591 NTU in the Mkomazi system; most estuaries had turbidities within the range 10e20 NTU (Table 6). 3.2. Multivariate analysis The results of the PCA of closed estuaries revealed that the first two PC axes accounted for approximately 71% of the variation between the samples. The first PC axis was strongly related to turbidity and depth, while

Table 4 Average physico-chemical parameters recorded in open cool-temperate estuaries (number of samples; G SE) System

Depth (m)

Olifants Berg Palmiet Uilkraals

1.48 2.73 1.83 0.83

(5; (5; (3; (3;

0.33) 0.60) 0.20) 0.07)

Temperature ((C)

Salinity

Dissolved oxygen (mg l
1)

Turbidity (NTU)

13.95 16.42 14.97 17.33

17.85 20.00 17.77 15.43

8.73 6.76 8.43 8.49

30.80 13.60 4.00 5.33

(10; 0.79) (10; 0.55) (6; 0.18) (6; 0.27)

(10; 4.71) (10; 3.56) (6; 6.25) (6; 2.11)

(10; 0.29) (10; 0.16) (6; 0.58) (6; 0.11)

(5; (5; (3; (3;

7.75) 3.57) 0.00) 0.88)

80

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

Table 5 Average physico-chemical parameters recorded in open warm-temperate estuaries (number of samples; G SE) System

Depth (m)

Heuningnes Bree¨ Duiwenhoks Goukou Gourits Keurbooms Kromme Gamtoos Swartkops Sundays Bushmans Kariega Kowie Great Fish Keiskamma Buffalo Nahoon Gqunube Kwelera Great Kei Kobonqaba Ngqusi/Inxaxo Qora Shixini Mbashe Xora Mtata Mdumbi

1.63 2.52 2.68 1.48 2.10 1.63 2.39 1.59 2.22 2.41 2.44 2.39 2.75 1.92 2.01 3.43 2.32 1.68 1.56 1.78 1.53 2.14 1.03 1.03 2.75 2.10 3.08 2.08

(3; (5; (4; (4; (4; (8; (8; (8; (6; (7; (7; (8; (6; (5; (8; (3; (5; (4; (5; (6; (4; (7; (3; (3; (4; (4; (5; (4;

0.44) 0.48) 0.69) 0.13) 0.35) 0.31) 0.37) 0.24) 0.38) 0.32) 0.38) 0.18) 0.36) 0.12) 0.26) 1.13) 0.61) 0.30) 0.33) 0.36) 0.60) 0.40) 0.12) 0.38) 0.90) 0.46) 0.82) 0.69)

Temperature ((C)

Salinity

Dissolved oxygen (mg l
1)

Turbidity (NTU)

14.85 19.06 15.88 17.26 17.41 21.36 18.84 19.16 18.58 20.45 21.05 22.95 21.48 21.55 18.41 18.17 19.41 19.98 21.01 22.17 20.38 18.06 17.02 18.84 20.24 22.43 21.40 21.54

18.17 10.71 31.63 23.33 33.64 27.19 30.23 19.80 30.21 18.11 32.28 31.59 30.04 12.60 22.09 31.17 32.61 32.81 32.12 4.43 28.36 28.03 23.38 30.86 14.64 27.84 12.59 27.76

7.49 7.36 6.58 6.17 6.28 5.23 6.53 8.08 6.25 7.61 6.91 6.59 7.02 7.88 6.77 7.85 8.45 6.78 6.82 7.10 7.05 6.55 7.94 7.83 7.36 6.70 7.62 7.32

25.00 (3; 7.40 (5; 5.75 (4; 2.75 (4; 10.75 (4; 0.63 (8; 1.3* 11.4* 1.5* 23.9* 14.5* 4.7* 6.9* 73.0* 43.38 (8; 10.00 (3; 5.80 (5; 15.75 (4; 15.80 (5; 1300.00 (6; 5.50 (4; 7.14 (7; 83.33 (3; 14.33 (3; 163.00 (4; 17.67 (3; 100.20 (5; 8.00 (4;

(6; 0.10) (10; 0.22) (8; 0.13) (7; 0.13) (8; 0.20) (16; 0.28) (18; 0.21) (16; 0.15) (12; 0.09) (14; 0.30) (13; 0.40) (16; 0.51) (13; 0.38) (10; 0.84) (16; 0.08) (6; 0.59) (10; 0.29) (8; 0.37) (10; 0.54) (11; 0.23) (8; 0.97) (14; 0.26) (6; 0.82) (5; 0.88) (7; 0.65) (7; 0.73) (10; 1.06) (8; 0.74)

(6; 5.16) (10; 1.59) (8; 1.51) (7; 3.36) (8; 0.68) (16; 1.56) (18; 0.47) (16; 2.58) (12; 0.77) (14; 2.66) (13; 0.34) (16; 0.47) (13; 0.84) (10; 4.15) (16; 1.94) (6; 0.64) (10; 0.21) (8; 0.04) (10; 0.16) (11; 3.05) (8; 2.50) (14; 2.99) (6; 4.29) (5; 0.53) (7; 5.25) (7; 2.66) (10; 3.62) (8; 2.38)

(6; 0.06) (10; 0.06) (8; 0.07) (7; 0.11) (8; 0.07) (16; 0.27) (18; 0.22) (16; 0.49) (12; 0.27) (14; 0.44) (13; 0.16) (16; 0.20) (13; 0.11) (10; 0.19) (16; 0.31) (6; 0.57) (10; 0.38) (8; 0.22) (10; 0.18) (11; 0.13) (8; 0.09) (14; 0.28) (6; 0.14) (5; 0.07) (7; 0.58) (7; 0.48) (10; 0.10) (8; 0.30)

10.50) 0.75) 0.25) 0.25) 4.19) 0.18)

12.66) 1.73) 1.36) 5.85) 3.2) 0.00) 0.87) 1.58) 26.12) 2.96) 120.41) 5.61) 31.23) 1.41)

Turbidity values estimated from Secchi disc measurements are denoted by an asterisk (*).

the second axis was strongly related to salinity and temperature (Table 7). The pattern produced by the ordination showed a broad gradation from subtropical estuaries in the upper half of the plot toward cool- and warm-temperate systems in the lower half of the ordination (Fig. 4). The first two axes of the PCA of open estuaries accounted for approximately 76% of the variation between the sites; the first PC axis was related to salinity and turbidity, while the second axis was correlated with depth and temperature (Table 8). The ordination

produced by the PCA showed warm-temperate systems situated toward the left of the plot with subtropical systems forming a broad scatter along the top and right of the ordination. Cool-temperate estuaries were mostly situated toward the bottom of the ordination (Fig. 5). The ANOSIM test revealed that both closed and open estuaries in the subtropical region were slightly but significantly different (p ! 0:01) to cool-temperate and warm-temperate systems (Table 9). Warm-temperate and cool-temperate systems, however, did not exhibit any significant difference.

Table 6 Average physico-chemical parameters recorded in open subtropical estuaries (number of samples; G SE) System

Depth (m)

Mngazana Mngazi Mntafufu Msikaba Mtentu Mzamba Mzimkulu Mkomazi Matigulu/Nyoni Mlalazi

2.33 2.08 3.18 4.08 3.85 2.30 1.06 0.85 1.12 1.72

(6; (5; (4; (4; (4; (3; (5; (4; (7; (5;

0.30) 0.55) 1.16) 0.80) 1.23) 0.61) 0.17) 0.31) 0.16) 0.23)

Temperature ((C)

Salinity

Dissolved oxygen (mg l
1)

Turbidity (NTU)

23.48 24.10 24.76 25.84 25.60 26.75 23.53 24.43 23.54 26.98

28.48 19.25 20.30 17.00 18.90 23.93 14.08 3.44 3.00 9.20

5.90 5.90 6.86 7.33 7.43 7.01 7.06 6.50 5.21 6.66

12.50 86.20 19.50 10.25 4.50 13.33 12.00 591.25 55.71 19.60

(12; 0.40) (10; 0.21) (8; 0.46) (8; 0.68) (8; 0.54) (6; 0.66) (10; 0.50) (7; 0.19) (14; 0.20) (10; 0.27)

(12; 2.22) (10; 4.34) (8; 4.45) (8; 4.67) (8; 4.32) (6; 5.19) (10; 4.14) (7; 3.39) (14; 1.84) (10; 2.98)

(12; 0.48) (10; 0.44) (8; 0.70) (8; 0.56) (8; 0.41) (6; 0.58) (10; 0.35) (7; 0.35) (14; 0.48) (10; 0.19)

(6; (5; (4; (4; (4; (3; (5; (4; (7; (5;

2.64) 15.73) 1.19) 0.85) 1.04) 2.03) 0.89) 85.46) 5.88) 1.63)

81

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87 Table 7 Coefficients in the linear combinations of the physico-chemical variables making up the principal components for closed estuaries; the percentage variation explained by the principal components are also shown

Table 8 Coefficients in the linear combinations of the physico-chemical variables making up the principal components for open estuaries; the percentage variation explained by the principal components are also shown

Variable

PC1

PC2

PC3

PC4

Variable

PC1

PC2

PC3

PC4

Depth Temperature Salinity Turbidity % Variation

0.674 0.050 0.171
0.717 36.7

0.242 0.677
0.686 0.112 34.0


0.305 0.733 0.600
0.093 17.3

0.627 0.042 0.373 0.682 12.0

Depth Temperature Salinity Turbidity % Variation


0.197 0.332
0.659 0.646 44.9

0.750 0.656 0.029
0.080 30.8


0.630 0.669 0.179
0.353 14.9


0.053 0.110 0.730 0.673 9.3

4.1.1. Closed estuaries Closed estuaries, also sometimes referred to as blind estuaries (Day, 1981) or lagoons (Begg, 1984a), have small catchments (!500 km2) and river flow is insufficient during most months to prevent closure of the mouth by a bar built up from longshore and/or onshore movement of sand (Whitfield, 1992). Although many systems on the west coast, north of Cape Columbine have relatively large catchments, due to the arid climate, most comprise dry riverbeds and only carry water at times of exceptional rainfall (Heydorn, 1991). All closed estuaries in the cool-temperate region were situated on the southwest coast between Cape Columbine and Cape Agulhas (Fig. 2). Estuaries in this region frequently open during the winter rainfall

period but are closed by a sandbar in summer (Millard and Scott, 1954; Morant, 1991; Quick and Harding, 1994). In the warm-temperate region, where rainfall (and runoff) is relatively low, estuaries usually only breach during periods of high fluvial discharge (Perissinotto et al., 2000; Cowley and Whitfield, 2001). Seven systems were open during this study and most of these had breached as a result of heavy rains preceding sampling. Subtropical estuaries are normally closed during the dry winter season but frequently open following increased river discharge during the summer (Whitfield, 1980; Begg, 1984b; Blaber et al., 1984; Harrison and Whitfield, 1995; Cooper et al., 1999; van der Elst et al., 1999). Six systems were open at the time of this study and these had breached following recent rains in the catchment. Although rainfall and river flow is usually responsible for breaching the mouths of closed estuaries, breaching can also occur as a result of high seas overtopping and lowering the sand bar to a point

Fig. 4. PCA ordination of physico-chemical variables for closed South African estuaries (C Z cool-temperate estuaries; W Z warm-temperate estuaries; S Z subtropical estuaries).

Fig. 5. PCA ordination of physico-chemical variables for open South African estuaries (C Z cool-temperate estuaries; W Z warm-temperate estuaries; S Z subtropical estuaries).

4. Discussion 4.1. Physico-chemical characteristics

82

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

Table 9 Test statistic (R) (and significance) of the ANOSIM test applied to closed and open estuaries within each biogeographic region

Global R Cool-temperate versus warm-temperate Cool-temperate versus subtropical Warm-temperate versus subtropical

Closed

Open

0.365 (p ¼ 0:001) 0.080 (p ¼ 0:284)

0.440 (p ¼ 0:001) 0.235 (p ¼ 0:075)

0.511 (p ¼ 0:004)

0.586 (p ¼ 0:002)

0.411 (p ¼ 0:001)

0.487 (p ¼ 0:001)

which enables rising water levels to form an outlet (Begg, 1984a). Closed estuaries in the cool-temperate region are generally shallow systems with water depths of !1.0 m being reported (Millard and Scott, 1954; Day, 1981). Similar conditions were recorded during this study where water depths usually did not exceed 1.0 m (Table 1). Warm-temperate estuaries were generally deeper than cool-temperate systems with average water depths mostly 1e2 m (Table 2). Similar water depths of between 1.0 and 2.0 m have also been reported in many warmtemperate estuaries (Day, 1981; Cowley and Whitfield, 2001; Vorwerk et al., 2001). When closed estuaries breach there is usually a strong outflow to sea followed by a fall in water level. Closed estuaries in the warmtemperate region, however, develop behind low-elevation barriers fronted by wide, gently sloping beaches; these systems are impounded at or close to high tide level, and as a result, breaching does not result in a dramatic reduction in water level (Cooper, 2001). In contrast, closed estuaries that occur in the subtropical region develop behind steep beaches and have high berms that maintain a water level above high tide level (Cooper, 2001). Due to their bed levels being elevated above mean sea level, these systems tend to drain when they open (Begg, 1984b; Cooper, 1989; Cooper and Harrison, 1992; Cooper et al., 1999). During periods of mouth closure, however, water levels increase and the adjacent floodplain often becomes inundated (Begg, 1984a). The vast majority of subtropical estuaries were closed at the time of sampling and this probably accounts for the relatively deep conditions (mostly 1e2 m) recorded during this study (Table 3). Mean water depth was positively correlated with the first principal component of the PCA (Table 7). Many warm-temperate systems were situated toward the right of the PCA plot with relatively shallow cool-temperate estuaries located slightly toward the left of the ordination (Fig. 4). Although subtropical systems were also situated mostly to the left of the PCA plot, water depths were similar to those recorded in warm-temperate systems. Their position in the PCA ordination is probably a reflection of turbidity, which was also correlated with the first PC axis.

Estuarine water temperatures generally decreased from the subtropical region toward the cool-temperate zone; this corresponds to the trend for marine coastal waters. Mean water temperatures of estuaries in the cool-temperate region usually did not exceed 20 (C (Table 1). During open mouth phases, marine waters can exert a cooling influence in the lower reaches of these estuaries (Millard and Scott, 1954). The cooler temperatures recorded in the Diep estuary during this study (17 (C) are probably a reflection of the open mouth condition at the time of sampling. Average water temperatures in warm-temperate estuaries were mostly in the range 18e24 (C (Table 2) and subtropical estuaries were generally above 24 (C (Table 3). This pattern was also evident in the PCA ordination. Temperature was positively related to the second principal component of the PCA (Table 7) and the ordination produced a gradation from subtropical systems at the top of the plot toward cool- and warmtemperate systems toward the bottom of the ordination (Fig. 4). Although tidal exchange and salinity gradients may be present when closed systems open, during the normally closed phase, salinities in these systems depend on the ratio between losses through evaporation and seepage through the sand bar, and gains through river discharge plus direct precipitation (Day, 1981). If the gains exceed the losses, the salinity decreases; if the losses by evaporation exceed the gains of fresh water (e.g. during droughts) then the salinity of the estuary may become hypersaline (O40) (Whitfield and Bruton, 1989). In cool-temperate estuaries, high winter runoff reduces salinities in these systems while elevated temperatures and high evaporation rates increases salinities during the summer (Millard and Scott, 1954; Day, 1981). During closed phases, seawater may occasionally enter these systems by overtopping the sand bar at the mouth (Millard and Scott, 1954). Mean salinities in closed cool-temperate estuaries during this study were generally low (!15) (Table 1); the relatively high salinity recorded in the Diep estuary (O20) is probably a result of marine influence due to the open mouth condition. Salinities in warm-temperate estuaries are typically high; this is due to a combination of low freshwater input, high evaporation rates as well as seawater introduction via barrier overwash (Cooper et al., 1999; Cowley and Whitfield, 2001; Vorwerk et al., 2001). Mean salinities during this study were generally between 15 and 30; one system (Gqutya) was hypersaline (Table 2). Occasional heavy rains may reduce salinities in warm-temperate estuaries, especially during the outflow phase after mouth breaching (Perissinotto et al., 2000; Cowley and Whitfield, 2001). High river runoff was also responsible for the relatively low salinities (!10) recorded in the Tsitsikamma and Morgan estuaries (Table 2). High rainfall and runoff together with limited seawater input

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

results in perched subtropical estuaries having typically low salinities (Whitfield, 1980; Begg, 1984a, 1984b; Harrison and Whitfield, 1995; van der Elst et al., 1999). During this study, mean salinities generally did not exceed 15 with a large proportion of systems having values below 5 (Table 3). Although salinities in subtropical estuaries are generally low, seawater penetration often occurs when these systems open (Begg, 1984b; Blaber et al., 1984; Whitfield, 1990; van der Elst et al., 1999); seawater can also occasionally enter these estuaries via barrier overwash (Begg, 1984a). Salinity was negatively related to the second principal component of the PCA (Table 7). The ordination also showed a pattern of relatively saline warm-temperate estuaries occurring toward the bottom of the plot and less saline subtropical systems toward the top of the ordination; most cool-temperate estuaries were situated in the centre of the PCA ordination (Fig. 4). Dissolved oxygen concentrations during this study was significantly negatively correlated with temperature and generally declined from cool-temperate to subtropical estuaries. Closed cool-temperate estuaries during this study appeared to be well oxygenated with mean values generally exceeding 5.0 mg l
1 (Table 1). Dissolved oxygen values in closed warm-temperate estuaries were mostly within the range 5e8 mg l
1 (Table 2). A high proportion of subtropical estuaries had mean dissolved oxygen levels of below 5.0 mg l
1 (Table 3). Begg (1984b) found that some closed KwaZulueNatal estuaries were well mixed and well oxygenated throughout the year and this was attributed to their shallow condition. In other estuaries, however, reduced oxygen levels (generally !5.0 mg l
1) were ascribed to poor water circulation, protection from the wind and the decomposition of leaf litter (Begg, 1984a, 1984b). A number of factors influence the turbidity of estuarine waters and these include river flow, substratum type, wind and tides (Cyrus, 1988b). Turbidities in cooltemperate estuaries generally increase following winter rains but are reduced when river flow ceases in summer (Millard and Scott, 1954). Turbidities in closed cooltemperate estuaries during this study were generally below 15 NTU (Table 1). Warm-temperate estuaries are predominantly clear systems (Vorwerk et al., 2001). Low turbidities were also recoded in warm-temperate estuaries during this study with values mostly below 5 NTU (Table 2). High turbidities (O80 NTU), however, were recorded in systems that had recently experienced heavy rains in their catchments (e.g. Nyara, Morgan). Water transparency in subtropical estuaries tend to decline during the rainy season while the rivers are flowing but are much clearer in winter, particularly during the closed phase (Whitfield, 1980; Begg, 1984a, 1984b; Blaber et al., 1984; Cyrus, 1988b; Harrison and Whitfield, 1995). The low turbidities (mostly 5e10 NTU) recorded in closed subtropical estuaries during this study (Table 3)

83

are probably a reflection of the closed mouth condition of the majority of systems. Although mean turbidities in subtropical estuaries were low, they were generally higher than those recorded in warm-temperate systems. This is also supported by the PCA. Turbidity was negatively correlated with the first principal component of the PCA (Table 7). Warmtemperate estuaries were mainly situated to the right of the ordination while subtropical estuaries were mostly situated to the left of the plot. Cool-temperate estuaries were also situated toward the left of the PCA ordination. Overall, the PCA for closed estuaries showed a broad gradation from the warm, turbid, relatively low salinity subtropical estuaries in the upper half of the plot toward the cooler, clearer, more saline systems of the warmtemperate region in the lower right half of the ordination (Fig. 4). 4.1.2. Open estuaries Open estuaries are relatively large systems with catchments usually exceeding 500 km2 and a perennial river flow (Whitfield, 1992, 1998). In the cool-temperate region, two systems (Olifants, Berg) were situated on the west coast, and two (Palmiet, Uilkraals) were located on the southwest coast (Fig. 3). The estuaries on the west coast have large catchments that drain areas beyond the arid coastal zone and thus river flow is generally sufficient to maintain a near permanent outlet to the sea (Cooper et al., 1999). The estuaries on the southwest coast have smaller catchments (and MAR) than those on the west coast. The position of the mouth of the Palmiet estuary is situated against a rocky promontory and this, together with an almost continuous runoff throughout the year helps maintain an open mouth condition. The system may, however, close briefly during the dry summer months (Branch and Day, 1984; Bennett, 1989). Little is known about the Uilkraals estuary; this system opens to the sea over a beach with a relatively flat profile and does not appear to close. Strong tidal currents probably contribute toward maintaining an open mouth condition in this estuary. Tidal currents play a major role in maintaining a connection with the sea in warm-temperate estuaries while in subtropical estuaries, river flow is the major factor responsible for maintaining an open mouth condition (Cooper et al., 1999; Cooper, 2001). Unlike tidedominated estuaries, river-dominated systems may close under low flow or drought conditions (Begg, 1984b). Water depth in open estuaries is strongly influenced by the state of the tide; however, this was not accounted for in the present study. Given this variability, however, the measurements recorded during this study compare with those reported by other workers. Water depths in cool-temperate estuaries were found to be mainly between 2.0 and 3.0 m (Day, 1981; Taljaard et al., 1986;

84

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

Slinger and Taljaard, 1994). The mean water depth recorded in cool-temperate estuaries during this study generally exceeded 1.5 m (Table 4). Average water depths recorded in warm-temperate estuaries generally exceeded 2.0 m (Table 5). Water depths greater than 2.0 m have also been reported in a number of open warm-temperate systems (Day, 1981; Marais, 1984; Reddering and Esterhuysen, 1987; Plumstead et al., 1989a, 1989b). The average water depths of subtropical estuaries were largely over 2.0 m (Table 6). Similar water depths (O2 m) have also been reported in many open subtropical estuaries (Connell, 1974; Branch and Grindley, 1979; Begg, 1984b; Plumstead et al., 1991). Water depth was positively correlated with the second principal component of the PCA (Table 8). Given the wide overlap of estuarine water depths recorded within each region, the pattern produced by PCA ordination is more easily explained as a function of temperature rather than depth. Temperature was also positively correlated with the second principal component of the PCA (Table 8). The ordination showed subtropical estuaries to be situated toward the top of the plot with a gradation to warmtemperate systems in the centre of the ordination and cool-temperate systems toward the bottom of the plot (Fig. 5). Overall, estuarine water temperatures declined from subtropical estuaries toward cool-temperate systems, following the pattern for oceanic waters. Mean water temperatures in cool-temperate estuaries did not exceed 18 (C, water temperatures in warm-temperate estuaries were mostly between 18 and 22 (C and water temperatures in subtropical estuaries were above 22 (C (Tables 4e6). Salinities in open estuaries are governed primarily by the mixing of freshwater inflow from the catchment and seawater inputs driven by tidal currents. High salinity conditions often characterise cool-temperate estuaries during the summer while in winter, when the rivers are flowing very strongly, seawater penetration is often limited and low salinity conditions (!3) can occur (Slinger and Taljaard, 1994). Mean salinities in cooltemperate estuaries during this study were within the range 15e20 (Table 4). Tide-dominated estuaries in the warm-temperate region have large tidal prisms (Cooper et al., 1999; Cooper, 2001) and as a result, salinities in these systems are high, typically exceeding 30 (Marais, 1984; Emmerson, 1985; Emmerson and Erasmus, 1987; Hecht and van der Lingen, 1992; Scharler et al., 1997; Colloty et al., 2001). Mean salinities in warm-temperate estuaries during this study were mostly in the range 25e35 (Table 5). Drought or low flow conditions can lead to hypersaline conditions and even a reversed salinity gradient (Marais and Baird, 1980; Ter Morshuizen and Whitfield, 1994) while increased runoff following heavy rainfall or floods can result in lowered salinities (Marais and Baird, 1980; Plumstead et al., 1985, 1989a,

1989b; Whitfield et al., 1994). Low salinities (!15) were reported in the Great Kei, Mbashe and Mtata estuaries (Table 5) and this was primarily due to increased runoff following rains in the catchment. The river-dominated estuaries in the subtropical region have small tidal prisms (Cooper et al., 1999; Cooper, 2001) and salinities are typically reduced (Begg, 1984b). Salinities recorded in subtropical estuaries during this study were mostly below 20 (Table 6). Heavy rains in the catchment accounts for the relatively low salinities (!15) recorded in the Mzimkulu, Mkomazi, Matigulu/Nyoni, and Mlalazi estuaries during this study. Salinity was negatively correlated with the first principal component of the PCA (Table 8). Estuaries in which low salinities were recorded (mostly subtropical systems) were situated toward the right of the ordination while estuaries with high salinities (mostly warmtemperate estuaries) were situated toward the left of the plot (Fig. 5). Cool-temperate systems occupied an intermediate position between subtropical and warmtemperate estuaries along the first PC axis of the ordination. Dissolved oxygen concentrations, which were negatively correlated with temperature, generally increased from subtropical estuaries toward cool-temperate systems. Dissolved oxygen levels in cool-temperate estuaries mostly exceeded 8.0 mg l
1 (Table 4). Relatively high dissolved oxygen levels have also been reported in a number of cool-temperate estuaries (Branch and Day, 1984; Slinger and Taljaard, 1994). Mean dissolved oxygen values recorded in warm-temperate estuaries exceeded 6.0 mg l
1 (Table 5). Other studies have also found that warm-temperate estuaries are well oxygenated with dissolved oxygen values mostly above 5.0 mg l
1 (Emmerson, 1985, 1989; Plumstead et al., 1985, 1989a, 1989b; Emmerson and Erasmus, 1987; Hecht and van den Lingen, 1992; Scharler et al., 1997). Subtropical estuaries were also well oxygenated with average dissolved oxygen concentrations exceeding 5.0 mg l
1 (Table 6). Good water column mixing was attributed to the well-oxygenated waters reported in many subtropical estuaries (Connell, 1974; Branch and Grindley, 1979; Begg, 1984b; Plumstead et al., 1991). Turbidities in cool-temperate estuaries during this study did not exceed 30 NTU and were mostly below 15 NTU (Table 4). Turbidity in cool-temperate estuaries is strongly linked to rainfall and runoff; heavy rain in the catchment often results in turbid conditions, particularly in winter (Day, 1981). Being strongly marine influenced, open warm-temperate estuaries are relatively clear systems (Marais, 1984; Hecht and van der Lingen, 1992; Whitfield, 1994; Scharler et al., 1997). Turbidities in warm-temperate estuaries during this study were generally below 10 NTU (Table 5). High turbidities (O80 NTU), however, were recorded in a number of

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

systems (e.g. Great Kei, Mbashe and Mtata) and this is linked to heavy rains in the catchments prior to sampling. Increased turbidities have also been reported in warm-temperate estuaries during high river flow periods (Marais, 1983; Plumstead et al., 1985, 1989a, 1989b; Hecht and van der Lingen, 1992; Whitfield, 1994; Whitfield et al., 1994; Ter Morshuizen et al., 1996; Vorwerk et al., 2001). Turbidity in subtropical estuaries is also linked to rainfall and runoff with values often increasing following rainfall in the catchment and heavy runoff (Hill, 1966; Begg, 1984b; Cyrus, 1988b; Plumstead et al., 1991). Turbidities recorded in subtropical estuaries during this study were mostly in the range 10e20 NTU (Table 6). High turbidities were recorded in those systems that were sampled following heavy rains in the catchment (e.g. Mngazi, Mkomazi, Matigulu/ Nyoni). Turbidity was negatively correlated with the first principal component of the PCA. Predominantly clear warm-temperate estuaries were situated toward the left of the ordination while turbid (subtropical) estuaries were situated to the right of the plot (Fig. 5). Cooltemperate estuaries occupied an intermediate position between subtropical and warm-temperate estuaries along the first PC axis of the ordination. The PCA ordination for open systems showed a broad gradation from cool, clear, saline warm-temperate systems in the bottom left of the plot toward the warmer, less saline, turbid estuaries of the subtropical region in right hand of the ordination (Fig. 5). 4.2. General Although based on a single, once-off survey, the physico-chemical parameters recorded during this study compare with results from other studies in South African estuaries. The study is also unique in that physico-chemical conditions of estuaries within each zoogeographic region were compared using data that were collected in a systematic and comparable manner. The results have showed that estuaries within each zoogeographic region have somewhat distinctive physicochemical characteristics. This is also supported by the PCA. The ANOSIM test also revealed that in both closed and open estuaries, subtropical systems were significantly different to cool-temperate and warm-temperate systems (Table 9). Although the test indicated that there was no difference between warm-temperate and cooltemperate systems, this may be due to the small number of systems represented in the cool-temperate region. Day (1981) grouped southern African estuaries into three main provinces based mainly on water temperature, rainfall and river flow. The estuaries of southern Mozambique to the Great Kei were classified as subtropical and were characterised by warm waters

85

(O16 (C), a predominantly summer rainfall pattern and high river discharge during this season. Warm-temperate estuaries from the Great Kei to Cape Point have minimum winter temperatures of between 12 and 14 (C and experience variable rainfall. The estuaries on the west coast, between Cape Point and the Orange River are characterised by very low summer rainfall and high evaporation during this period. Cooper (2001) also found that the geomorphological variability of estuaries in South Africa followed a broad pattern around the coast that reflected regional variability in climate, topography and sediment availability. The subtropical east coast was characterised by predominantly perched closed systems and river-dominated open estuaries; warm-temperate estuaries were mainly non-perched closed estuaries and tide-dominated open systems. Closed estuaries in the cool-temperate region were predominantly perched systems while open estuaries comprised both tide- and river-dominated systems (Cooper, 2001). Variations in climate, river and oceanographic regime have also been found to be responsible for regional differences in estuary physical/environmental attributes in Australia and North America. Based on a PCA of eight physical/environmental attributes Pease (1999) classified the estuaries of New South Wales (Australia) into three regions that also broadly corresponded to the biogeographic provinces of that region. Key parameters identified in the analysis included mouth depth and width, and latitude. Mouth depth and width were related to estuary size, geomorphology, runoff, and the degree of marine influence while latitude was related to temperature, rainfall and wind patterns (Pease, 1999). Kench (1999) also described broad differences in Australian estuarine morphology that correspond to the regional differences in climate, river, and oceanographic regime. The interaction of these factors results in a range of estuarine types around the Australian continent from wave-dominated, microtidal, bar-built estuaries in the south to low energy macrotidal estuaries in the north (Kench, 1999). In North America, Dame et al. (2000) found that the estuaries of the southeastern Atlantic coast could be divided into three general areas. The North Carolina coast is dominated by extensive and poorly flushed sounds; the central region of South Carolina and Georgia is dominated by well-flushed bar-built and riverine estuaries, and the Florida coast is dominated by moderate to poorly flushed bar-built estuaries with little freshwater input (Dame et al., 2000). On the west coast, Emmett et al. (2000) found that southern Californian estuaries were relatively small systems with little freshwater inflow; many of these systems are closed to the sea for much of the year but during the winter rainfall period, these systems often breach (Emmett et al., 2000). Further north, the estuaries are typically

86

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87

drowned-river valley systems that are morphologically larger and have higher freshwater inflows (Emmett et al., 2000). This study has demonstrated that there is a strong relationship between estuarine zoogeography and the physico-chemical conditions around the South African coast. This could be related to regional variations in climate and oceanic conditions. Estuarine temperatures follow the trend for marine coastal waters, decreasing from the subtropical east coast, along the warmtemperate south coast and up the cool-temperate west coast. Low rainfall and runoff, together with higher seawater input and evaporative loss, results in elevated salinities and low turbidities in warm-temperate estuaries. In both cool-temperate and subtropical estuaries, high rainfall and runoff often results in low salinities and high turbidities, the duration and intensity of which varies according to river flow regimes.

Acknowledgements I would like to express my gratitude to the South African Department of Environmental Affairs and Tourism for funding the project that generated the data for this research. I would also like to thank the CSIR, Division of Water Environment and Forestry Technology. I am also grateful to the various national and regional conservation authorities for permission to sample the estuaries and for logistical assistance. Thanks is also extended to all those who helped with field sampling and laboratory analyses. I am also grateful to Arjoon Singh, CSIR, for generating the maps and to Dr Alan Whitfield, South African Institute for Aquatic Biodiversity, for his helpful comments on an earlier draft of this manuscript. Finally, I would like to thank the referees for their useful comments.

References Allanson, B., 1999. Chemistry. The primary chemical structure. In: Allanson, B.R., Baird, D. (Eds.), Estuaries of South Africa. Cambridge University Press, Cambridge, pp. 53e70. Begg, G.W., 1984a. The comparative ecology of Natal’s smaller estuaries. Natal Town and Regional Planning Report 62, 1e182. Begg, G.W., 1984b. The estuaries of Natal. Part 2. Natal Town and Regional Planning Report 55, 1e631. Bennett, B.A., 1989. A comparison of the fish communities in nearby permanently open, seasonally open and normally closed estuaries in the south-western Cape, South Africa. South African Journal of Marine Science 8, 43e55. Blaber, S.J.M., Hay, D.G., Cyrus, D.P., Martin, T.J., 1984. The ecology of two degraded estuaries on the north coast of Natal, South Africa. South African Journal of Zoology 19, 224e240. Branch, G.M., Day, J.A., 1984. Ecology of southern African estuaries. Part XIII. The Palmiet River estuary in the south-western Cape. South African Journal of Zoology 19, 63e77.

Branch, G.M., Grindley, J.R., 1979. Ecology of southern African estuaries. Part XI. Mngazana: a mangrove estuary in Transkei. South African Journal of Zoology 14, 149e170. Clarke, K.R., Warwick, R.M., 1994. Change in marine communities: an approach to statistical analysis and interpretation. Plymouth Marine Laboratory, Plymouth, 144 pp. Colloty, B.M., Adams, J.B., Bate, G.C., 2001. The botanical importance rating of estuaries in the former Ciskei/Transkei. Water Research Commission Report TT 160/01, pp. 1e119. Connell, A.D., 1974. Mysidacea of the Mtentu River estuary, Transkei, South Africa. Zoologica Africana 9, 147e159. Cooper, J.A.G., 1989. Fairweather versus flood sedimentation in Mhlanga lagoon, Natal: implications for environmental management. South African Journal of Geology 92, 279e294. Cooper, J.A.G., 2001. Geomorphological variability among microtidal estuaries from the wave-dominated South African coast. Geomorphology 40, 99e122. Cooper, J.A.G., Harrison, T.D., 1992. Effects of nesting activities of Oreochromis mossambicus (Pisces: Cichlidae) on bank stability in small lagoons. South African Journal of Science 88, 398e401. Cooper, A., Wright, I., Mason, T., 1999. Geomorphology and sedimentology. In: Allanson, B.R., Baird, D. (Eds.), Estuaries of South Africa. Cambridge University Press, Cambridge, pp. 5e25. Cowley, P.D., Whitfield, A.K., 2001. Ichthyofaunal characteristics of a typical temporarily open/closed estuary on the southeast coast of South Africa. Ichthyological Bulletin of the J.L.B. Smith Institute of Ichthyology 71, 1e19. Cyrus, D.P., 1988a. Turbidity and other physical factors in Natal estuarine systems. Part 2: Estuarine lakes. Journal of the Limnological Society of Southern Africa 14, 72e81. Cyrus, D.P., 1988b. Turbidity and other physical factors in Natal estuarine systems. Part 1: Selected estuaries. Journal of the Limnological Society of Southern Africa 14, 60e71. Dame, R., Alber, M., Allen, D., Mallin, M., Montague, C., Lewitus, A., Chalmers, A., Gardner, R., Gilman, C., Kjerfve, B., Pinckney, J., Smith, N., 2000. Estuaries of the South Atlantic coast of North America: their geographical signatures. Estuaries 23, 793e819. Day, J.H., 1981. Estuarine Ecology with Particular Reference to Southern Africa. A.A. Balkema, Cape Town, 411 pp. Emmerson, W.D., 1985. The nutrient status of the Swartkops River estuary, Eastern Cape. Water SA 4, 189e198. Emmerson, W.D., 1989. The nutrient status of the Sundays River estuary South Africa. Water Research 23, 1059e1067. Emmerson, W.D., Erasmus, T., 1987. The nutrient status of the Kromme River estuary, South Africa. Hydrobiologia 148, 87e96. Emmett, R., Llanso, R., Newton, J., Thom, R., Hornberger, M., Morgan, C., Levings, C., Copping, A., Fishman, P., 2000. Geographic signatures of North American west coast estuaries. Estuaries 23, 765e792. Harding, W.R., 1994. Water quality trends and the influence of salinity in a highly regulated estuary near Cape Town, South Africa. South African Journal of Science 90, 240e246. Harrison, T.D., 2002. Preliminary assessment of the biogeography of fishes in South African estuaries. Marine and Freshwater Research 53, 479e490. Harrison, T.D., Whitfield, A.K., 1995. Fish community structure in three temporarily open/closed estuaries on the Natal coast. Ichthyological Bulletin of the J.L.B. Smith Institute of Ichthyology 64, 1e80. Harrison, T.D., Cooper, J.A.G., Ramm, A.E.L., 2000. State of South African estuaries. Geomorphology, Ichthyofauna, water quality and aesthetics. Department of Environmental Affairs and Tourism State of the Environment Series Report 2, 1e127. Hecht, T., Van der Lingen, C.D., 1992. Turbidity-induced changes in feeding strategies of fish in estuaries. South African Journal of Zoology 27, 95e107.

T.D. Harrison / Estuarine, Coastal and Shelf Science 61 (2004) 73e87 Heydorn, A.E.F., 1991. The conservation status of southern African estuaries. In: Huntley, B.J. (Ed.), Biotic Diversity in Southern Africa. Concepts and Conservation. Oxford University Press, Cape Town, pp. 290e297. Heydorn, A.E.F., Tinley, K.L., 1980. Estuaries of the Cape. Part 1. Synopsis of the Cape coastdnatural features, dynamics and utilization. CSIR Research Report 380, 1e97. Hill, B.J., 1966. A contribution to the ecology of the Umlalazi estuary. Zoologica Africana 2, 1e24. Kench, P.S., 1999. Geomorphology of Australian estuaries: review and prospect. Australian Journal of Ecology 24, 367e380. Lutjeharms, J.R.E., Cooper, J., Roberts, M., 2000. Upwelling at the inshore edge of the Agulhas Current. Continental Shelf Research 20, 737e761. Marais, J.F.K., 1983. Fish abundance and distribution in the Gamtoos estuary with notes on the effect of floods. South African Journal of Zoology 18, 103e109. Marais, J.F.K., 1984. Feeding ecology of major carnivorous fish from four eastern Cape estuaries. South African Journal of Zoology 19, 210e223. Marais, J.F.K., Baird, D., 1980. Seasonal abundance, distribution and catch per unit effort of fishes in the Swartkops estuary. South African Journal of Zoology 15, 66e71. Millard, N.A.H., Scott, K.M.F., 1954. The ecology of South African estuaries, Part VI: Milnerton estuary and the Diep River, Cape. Transactions of the Royal Society of South Africa 34, 279e324. Morant, P.D., 1991. The estuaries of False Bay. Transactions of the Royal Society of South Africa 47, 629e640. Pease, B.C., 1999. A spatially orientated analysis of estuaries and their associated commercial fisheries in New South Wales, Australia. Fisheries Research 42, 67e86. Perissinotto, R., Walker, D.R., Webb, P., Wooldridge, T.H., Bally, R., 2000. Relationships between zoo- and phytoplankton in a warmtemperate, semi-permanently closed estuary, South Africa. Estuarine, Coastal and Shelf Science 51, 1e11. Plumstead, E.E., Prinsloo, J.F., Schoonbee, H.J., 1985. A survey of the fish fauna of Transkei estuaries. Part 1. The Kei River estuary. South African Journal of Zoology 20, 213e220. Plumstead, E.E., Prinsloo, J.F., Schoonbee, H.J., 1989a. A survey of the fish fauna of Transkei estuaries. Part 2. The Mbashe estuary. South African Journal of Zoology 24, 273e281. Plumstead, E.E., Prinsloo, J.F., Schoonbee, H.J., 1989b. A survey of the fish fauna of Transkei estuaries. Part 3. The Mtata River estuary. South African Journal of Zoology 24, 282e289. Plumstead, E.E., Prinsloo, J.F., Schoonbee, H.J., 1991. A survey of the fish fauna of Transkei estuaries. Part 4. The Mntafufu and Mzamba River estuaries. South African Journal of Zoology 26, 153e163. Quick, A.J.R., Harding, W.R., 1994. Management of a shallow estuarine lake for recreation and as a fish nursery: Zandvlei, Cape Town, South Africa. Water SA 20, 289e297. Read, G.H., 1983. The effect of a dry and wet summer on the thermal and salinity structure of the middle and upper reaches of the Keiskamma estuary, Ciskei. Transactions of the Royal Society of South Africa 45, 45e62. Reddering, J.S.V., Esterhuysen, K., 1987. The effects of river floods on sediment dispersal in small estuaries: a case study from East London. South African Journal of Geology 90, 458e470.

87

Scharler, U.M., Baird, D., Winter, P.E.D., 1997. Diversity and productivity of biotic communities in relation to freshwater inputs in three Eastern Cape estuaries. Water Research Commission Report 463/1/98, pp. 1e197. Schulze, B.R., 1984. Climate of South Africa. Part 8. General survey. Weather Bureau, Department of Transport, Republic of South Africa, 300 pp. Shannon, L.V., 1989. The physical environment. In: Payne, A.I.L., Crawford, R.J.M. (Eds.), Oceans of Life off Southern Africa. Vlaeberg Publishers, Cape Town, pp. 12e27. Slinger, J.H., Taljaard, S., 1994. Preliminary investigation of seasonality in the Great Berg estuary. Water SA 20, 279e288. Taljaard, S., Eagle, G.A., Hennig, H.F.-K.O., 1986. The Palmiet estuary: A model for water circulation using salinity and temperature measurements over a tidal cycle. Water SA 12, 119e126. Ter Morshuizen, L.D., Whitfield, A.K., 1994. The distribution of littoral fish associated with eelgrass Zostera capensis beds in the Kariega estuary, a southern African system with a reversed salinity gradient. South African Journal of Marine Science 14, 95e105. Ter Morshuizen, L.D., Whitfield, A.K., Paterson, A.W., 1996. Distribution patterns of fishes in an eastern Cape estuary and river with particular emphasis on the ebb and flow region. Transactions of the Royal Society of South Africa 51, 257e280. Tinley, K.L., 1985. Coastal dunes of South Africa. South African National Scientific Programmes Report 109, 1e300. Van der Elst, R.P., Birnie, S.L., Everett, B.I., 1999. Siyaya catchment demonstration project: an experiment in estuarine rehabilitation. Oceanographic Research Institute Special Publication 6, 1e99. Vorwerk, P.D., Whitfield, A.K., Cowley, P.D., Paterson, A.W., 2001. A survey of selected Eastern Cape estuaries with particular reference to the ichthyofauna. Ichthyological Bulletin of the J.L.B. Smith Institute of Ichthyology 72, 1e52. Whitfield, A.K., 1980. Distribution of fishes in the Mhlanga estuary in relation to food resources. South African Journal of Zoology 15, 159e165. Whitfield, A.K., 1990. Life-history styles of fishes in South African estuaries. Environmental Biology of Fishes 28, 295e308. Whitfield, A.K., 1992. A characterization of southern African estuarine systems. Southern African Journal of Aquatic Sciences 18, 89e103. Whitfield, A.K., 1994. Abundance of larval and 0C juvenile marine fishes in the lower reaches of three southern African estuaries with differing freshwater inputs. Marine Ecology Progress Series 105, 257e267. Whitfield, A.K., 1998. Biology and ecology of fishes in southern African estuaries. Ichthyological Monographs of the J.L.B. Smith Institute of Ichthyology 2, 1e223. Whitfield, A.K., 2000. Available scientific information on individual southern African estuarine systems. Water Research Commission Report 577/3/00, pp. 1e217. Whitfield, A.K., Bruton, M.N., 1989. Some biological implications of reduced fresh water inflow into eastern Cape estuaries: a preliminary assessment. South African Journal of Science 85, 691e694. Whitfield, A.K., Paterson, A.W., Bok, A.H., Kok, H.M., 1994. A comparison of the ichthyofaunas in two permanently open eastern Cape estuaries. South African Journal of Zoology 29, 175e185.