Hydrobiologia 464: 227–243, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

227

A comparison of the macrobenthic faunas of permanently open and temporarily open/closed South African estuaries Peter R. Teske & Tris Wooldridge Department of Zoology, University of Port Elizabeth, Private Bag 1600, Port Elizabeth 6000, South Africa E-mail: [email protected] Received 19 October 2000; in revised form 23 August 2001; accepted 17 September 2001

Key words: subtidal macrobenthos, South Africa, estuary, macrobenthic density, species richness, diversity, zonation patterns

Abstract Thirteen estuaries in the Eastern Cape Province, South Africa, were broadly categorised according to size and salinity distribution and were assigned to one of the following categories: permanently open estuaries having a strong salinity gradient between mouth and upper estuary, freshwater-deprived permanently open estuaries, medium-sized temporarily open/closed estuaries, and small, temporarily open/closed estuaries. The macrobenthos collected during surveys was then compared in terms of the following parameters: species composition, salinity, sediment mud content, density of macrobenthic animals, Hill’s N0 (species richness), and Hill’s N1 (diversity). Mud content was found to be the most important environmental variable responsible for biotic patterns found, and sites were consequently assigned to either a sand zone fauna, or a mud zone fauna. Both types of fauna are present in all estuaries sampled, with upper sites of river dominated estuaries having an additional oligohaline fauna, and freshwater-deprived estuaries providing habitat for many marine species. Small, temporarily open/closed estuaries have the highest macrobenthic density, whereas N0 and N1 are highest in freshwater-deprived permanently open systems. River-dominated permanently open estuaries tend to have lower macrobenthic densities, species richness, and diversities compared to estuaries in the other categories. No seasonal differences in these ecological indices were found within any of the estuarine categories.

Introduction Along the South African coastline, there are approximately 250 estuaries (Whitfield, 2000) that comprise less than 600 km2 in total area (Grindley, 1976). A major threat to many of these systems is the impact of freshwater impoundments and irrigation (Whitfield & Bruton, 1989), including a decrease in the frequency and intensity of smaller floods. In 1986, it was estimated that major impoundments had the capacity to retain >50% of the Mean Annual Rainfall (MAR) in the country (Department of Water Affairs, 1986). Many estuaries in South Africa (72%) are not in contact with the sea for part of the year (Whitfield, 2000). These are referred to in this study as ‘temporarily open/closed estuaries’, following the categorisation system proposed by Whitfield (1992). The lagoonal

phase is terminated occasionally or periodically, when systems temporarily behave like true estuaries. Temporarily open/closed estuaries usually are shallow (<2 m average depth) and experience wide variations in physical and chemical attributes (e.g. salinity) that fluctuate between open and closed phases. The average annual rainfall over South Africa is only 493 mm (Wooldridge, 1999) and because of the increasing demand for water by a growing population, many estuaries today are closing more frequently and for longer periods of time than they would under natural conditions (Morant & Quinn, 1999). This can be predicted to have major effects on species composition of such estuaries, including those benthic species considered non-estuarine during part of their life cycle. Wooldridge (1999), for example, found that recruitment to estuarine populations of the mudprawn

228

Figure 1. Map of the sampling area. Estuaries sampled are labelled with acronyms, which are explained in Table 1. Any other estuaries and bodies of water in the sampling area that were not included in the analysis are numbered. Many of the larger rivers are not shown in their entire length.

Upogebia africana ceased during periods of mouth closure, as the planktonic larvae require a marine phase of development. The majority of estuarine studies in South Africa were conducted on permanently open systems, and most of the scientific publications deal with some biological aspect of single estuaries only. This paper presents a comparative study including both permanently open and temporarily open/closed estuaries.

Estuaries studied The 13 estuaries investigated in this study are located in the Eastern Cape Province, South Africa (Fig. 1), from the Kromme (KR) in the West, to the Keiskamma (KE) in the East. Much of the region is characterised by poor farming practices, resulting in large amounts of silt that reach some estuaries after heavy rainfalls. This is particularly evident in the Keiskamma and Great Fish Estuaries. General features of each estuary are given in Table 1, while Figure 2 reflects their relative sizes. South African estuaries are broadly categorized according to the dominant morphological features. Whitfield (1992) suggested five types of estuaries (estuarine bays, permanently open estuaries, river mouths, estuarine lakes, and temporarily open/closed estuaries). This study deals with permanently open and temporarily open/closed estuaries only, but considerable differences exist within each of these types. Consequently, the 13 estuaries have been assigned to five

categories, which are shown in Table 2. Permanently open estuaries in the study area are similar in size (>10 km in length), but they differ with regard to the amount of freshwater they receive. Category 1 consists of estuaries that receive a high freshwater input, whereas estuaries receiving very low amounts of freshwater due to impoundments are assigned to category 1f. Such freshwater deprived open systems are represented by the Kromme and the Kariega. These estuaries remain permanently open to the sea and are marine dominated. Salinity values remain around 35 g kg−1 throughout, and at times these two estuaries have a reversed salinity gradient, with salinity values in the upper reaches exceeding that of seawater. The other permanently open estuaries investigated all have strong freshwater inputs and distinct salinity gradients along their horizontal axes. Temporarily open/closed estuaries were assigned to three categories: Medium sized (<10 km but >1 km) estuaries, represented by the East Kleinemond, Mpekweni and Mtati Estuaries, were assigned to category 2. A fourth medium sized estuary, the Gqutywa Estuary, was hypersaline during the study period (>35 g kg−1 , max. 39.5 g kg−1 ). For that reason it was placed into a category of its own, which was named 2h. Category 3 consists of small estuaries (approximately 1 km or less in length) and included the Kabeljous, Van Stadens and Old Woman’s Estuaries. Taking into consideration the two estuarine types and five estuarine categories, the following hypotheses were tested in this study: 1. Macrobenthic density, species richness and diversity are greater in permanently open estuaries than in temporarily open/closed systems. 2. Within each of the two types of estuaries, different categories can be distinguished on the basis of species composition, macrobenthic density, species richness and diversity. 3. Estuarine categories can be distinguished on the basis of biotic zonation patterns.

Materials and methods Sampling procedure Field procedure The 13 estuaries were sampled in June 1998 and in December/January 1998/1999. In each case, benthic macrofauna was sampled at sites that varied in number between two and nine, depending on the length

229 Table 1. Characteristics of the 13 estuaries studied. Data for salinity and Secchi depth ranges for each estuary may originate from two different sampling excursions. Tidal level was not taken into consideration when measuring maximum water depth. Salinity measurements were taken near the bottom. Minimum sediment mud content is not shown, as it approached zero in the mouth area of most estuaries Name

Kromme Kabeljous Van Stadens Swartkops Sundays Kariega East Kleinemond Great Fish Old Woman Mpekweni Mtati Gqutywa Keiskamma

Acronym

Max. depth (m)

Salinity range (g kg−1 )

Secchi depth range (m)

KR KJ VS SW SU KA EK GF OW MP MT GQ KE

4.7 1.5 2.0 5.0 5.5 4.5 2.0 1.5 1.2 2.2 2.7 1.6 3.0

31.7–35.9 34.4–35.0 13.4–18.9 23.7–35.5 2.0–33.7 25.7–35.7 10.0–24.6 0.0–35.3 31.5–32.2 20.0–28.3 18.0–20.5 35.5–39.6 0.0–34.8

1.0–4.5 1.0–>1.5 >2.0 0.1–>1.5 0.2–1.0 0.8–2.0 0.6–1.6 0.01–0.05 0.8–>1.2 0.6–1.6 0.3–1.5 0.01–0.5

Max. sediment Special features mud content (%) 88 30 1 91 90 94 41 89 6 41 91 90 99

Marine dominated

Industrial and urban pollution Agricultural run-off in catchment Marine dominated River dominated, very turbid Very small size

Hypersaline River dominated, soil erosion in catchment

Figure 2. Maps of all estuaries investigated in this study, showing sampling sites. Sites sampled in winter 1998 only are in round brackets, sites sampled in summer 1998/99 only in square brackets. The figure also reflects the relative sizes of the different estuaries.

of individual estuaries. Sites were spaced at variable distances to each other and spanned the entire length of each estuary in order to incorporate possible salin-

ity variations, sediment changes, depth and turbidity changes. Before sampling commenced at each particular site, near bottom salinity was measured using a

230 Valeport CTD meter, and a sediment sample was taken using a van Veen grab (200 cm2 in area, sampled to 10 cm sediment depth). The grab sampler was operated from a small boat. At each site, three sets of macrobenthic samples were taken, each consisting of three grab samples (i.e. n=3, 9 grab samples in total), using the van Veen grab. Each set of samples was collected in an area of about 10 m2 , taking care not to sample repeatedly in the same spot. All samples were taken in, or close to the main channel, to ensure that all sites were subtidal. All grab samples were immediately transferred to large plastic bags and preserved in 10% formaldehyde. Each replicate represented a sampling area of 3 × 200 cm2 , but data were then corrected to number of ind m−2 . Laboratory procedure Macrobenthos was extracted from sand samples by repeated decantation. This was done at least three times, or until no more animals were visually detected. Before discarding the remaining sand, samples were carefully examined for the presence of macrobenthic organisms too heavy to be effectively extracted by decantation (large bivalves and gastropods). Muddy samples were first thinly spread over a fine sieve (1 m2 surface area, 0.26 mm aperture mesh size) and gently sprayed with a garden hose, until all mud was removed. Samples containing varying amounts of plant material were thoroughly checked for any animals that clung to plant debris. The remaining portion of the samples was subjected to the decanting procedure outlined above. All samples were then gently sieved through a mesh of 0.5 mm aperture size, and all animals were preserved in 10% formaldehyde, containing rose bengal dye. The two-step fixation procedure employed may not be ideal for dye uptake and preservation of fragile individuals, but as the extraction procedure took several weeks, the main purpose of using formaldehyde was to prevent rotting of the samples. Macrofauna from these samples was identified as far as possible, and counted using a binocular microscope. Sediment samples were dried in an oven at 100 ◦ C for 24 h. Approximately half of every sample was weighed, gently ground, and the mud (i.e. clay and silt) portion of each sample was removed by washing it through a sieve with 63 µm aperture size. The remaining sediment was dried at 100 ◦ C for 24 h and weighed. The difference in weight was expressed as the percentage of the initial weight, and it represen-

ted the percentage mud content (by dry weight) of the particular sample. Data analysis In order to test the suitability of the categorisation system for grouping Eastern Cape estuaries shown in the previous section, all the systems studied were compared with regard to taxon presence, macrobenthic density, species richness and diversity. Hill’s numbers N0 and N1 (Hill, 1973) were used to represent species richness and diversity, respectively. Macrobenthic density is defined here as the number of individuals per square meter that are retained by a sieve of 0.5 mm aperture size. Because of the skewness of the data, only non-parametric tests were considered appropriate for this study, even though in a few cases assumptions for parametric tests were satisifed (i.e. normal distribution and equality of variances). The use of Kruskal–Wallis tests to compare all estuarine categories simultaneously proved unsuitable, as statistically significant differences were found in a very small number of cases only. Whenever a significant difference was found, Dunn’s test (a pairwise multiple comparison procedure) was employed to identify subgroupings. However, the results of Dunn’s test were ambiguous in each case, which indicates that the amount of data was too low to employ this test (Zar, 1974). Consequently, macrobenthic densities, N0 and N1 of different estuarine categories were analysed by pooling the data from all sites of the estuaries in one particular category, and comparing them with the other categories in a pairwise fashion using a two-tailed Mann–Whitney U -test. The Mann–Whitney U -test is only 5% less powerful than a t-test, and in cases where the assumptions for a parametric test are violated, it is considerably more powerful (Zar, 1974). Whenever the null-hypothesis was rejected, a onetailed Mann–Whitney test was run to determine which of the two data sets had the greater values. The results of this method were only used to support general trends observed on the plots regarding the relationship of two categories of interest. They were not considered suitable to make statements regarding all categories simultaneously. In two cases, values measured at sites of a single estuary were compared with a pooled data set of the values found at the sites of several other estuaries using Mann–Whitney U -tests. Additionally, Kruskal– Wallis ANOVA on ranks was used to test for in-

231 tergroup differences between a number of estuaries consisting of the one under investigation and several others. Only when the null hypothesis was rejected by both tests, the difference was accepted as real. Summer and winter data from individual sites within each category were compared using a Wilcoxon signed rank test. Non-metric multidimensional scaling (NMDS) was chosen as an ordination method to identify faunistic zonation patterns and relate these to trends found between the different estuarine categories. NMDS can also be used to interpret environmental variables responsible for biotic patterns observed. A geometric shape representing the magnitude of the variable is superimposed on to a species or site score, and the significance of the variable determined by visually assessing how clustering of sites or species correspond to geometrical shapes of a certain size range. Given the volume of data from the 83 sites on 13 estuaries, differences between faunistic zones were not sufficiently distinct for site clusters to form. In order to identify biotic zones, the method of superimposing objects onto site scores was expanded in the following way. A list of the most common species was compiled that included 14 taxa, from the polychaete Capitella capitata complex and the cumacean Iphinoe truncata (found in all 13 estuaries) to the isopod Anopsilana ?fluviatilis, the tanaid Apseudes digitalis, the amphipod Corophium triaenonyx and the insect Chironomid larvae (found in 9 of the estuaries studied). At each site, code names of the two most abundant taxa from the list were superimposed onto site scores. Thus, by grouping abundant taxa represented in a specific area on the ordination plot, biotic grouping patterns were identified. These were then compared with patterns shown by the environmental data. Consensus between biotic and environmental patterns resulted in the designation of individual sites to different biotic zones. Geometric shapes representing different ranges of magnitude of the following variables were then superimposed onto their particular site score: salinity, sediment mud content, macrobenthic density, species richness, and diversity. Results for the two excursions were similar and only data collected during austral summer (December 1998 and January 1999) are presented here. Results from the second sampling trip (undertaken in June 1998 during the austral winter) are contrasted if seasonal differences were evident.

Table 2. Detailed categorisation of the estuaries investigated. Estuaries are listed from west to east. 1 = permanently open, large estuaries (>10 km in length); 2 = medium-sized (<10 km but >1 km in length) temporarily open/closed estuaries; 3 = small (approximately 1 km or less in length) temporarily open/closed estuaries; f = freshwater deprived estuary; h = estuary was hypersaline during the study period Estuary

Description

Kromme Kabeljous Van Stadens Swartkops Sundays Kariega East Kleinemond Great Fish Old Woman Mpekweni Mtati Gqutywa Keiskamma

Large, permanently open Small, temporarily open/closed Small, temporarily open/closed Large, permanently open Large, permanently open Large, permanently open Medium, temporarily open/closed Large, permanently open Small, temporarily open/closed Medium, temporarily open/closed Medium, temporarily open/closed Medium, temporarily open/closed Large, permanently open

Category 1f 3 3 1 1 1f 2 1 3 2 2 2h 1

Results Taxa identified A total of 72 taxa were identified in the 13 estuaries studied, which are listed in Table 3. Results from a previous expedition to the East Kleinemond, Gqutywa and Keiskamma Estuaries are included in this table. A maximum of eight taxa are new to science and are yet undescribed. A large number of taxa was recorded in permanently open estuaries only, whereas only three taxa were found exclusively in temporarily open/closed estuaries: Grandidierella chelata, Pontogeloides latipes and Thaumastoplax spiralis. G. chelata is known to inhabit permanently open estuaries (Wooldridge, unpubl. data). The records of the other two species are based on single individuals only, and are thus not representative. Comparison of macrobenthic community structure Comparison of species composition An arbitrarily placed cut-off point on the presence/absence dendrogram (Fig. 3) at approximately 140 yields five major groups, which only partly resemble the grouping used in the categorisation system shown in Table 2. Permanently open estuaries that receive a high freshwater input (category 1) showed

Table 3. List of all taxa found on three sampling expeditions to Eastern Cape estuaries (A = February 1998, B = June 1998, C = Dec/Jan 1998/99)

232

Table 3. Continued

233

234

Figure 3. Hierarchical clustering using presence/absence data of species from all three sampling expeditions.

little similarity in species composition, with the Great Fish and Sundays Estuaries being grouped together, the Swartkops Estuary being grouped with freshwater deprived estuaries (category 1f), and the Keiskamma Estuary being in a group of its own. Medium sized temporarily open/closed estuaries were all grouped together (including categories 2 and 2h). Species composition in the small Old Woman’s Estuary was very similar to that of the medium-sized estuaries (category 2), which geographically are located in closer proximity than other estuaries of its size (Kabeljous & Van Stadens, category 3). Comparison of indices Prior to detailed statistical investigations including all categories, two estuaries were focused on in greater detail, because upon inspection of Figures 3 and 4, the need for minor modifications of the categorisation system shown in Table 2 became evident.

No significant differences in macrobenthic density, species richness and diversity in either season were found when comparing the Gqutywa Estuary with the three estuaries in category 2 using a two-tailed Mann– Whitney test. In winter, p-values of density, species richness and diversity were 0.239, 1.000 and 0.957, respectively. In summer, the three corresponding pvalues were 0.328, 0.217 and 0.766. A Kruskal–Wallis ANOVA on ranks also did not find any statistically significant differences between the four medium-sized estuaries, with p-values for macrobenthic density, species richness and diversity being 0.135, 0.499 and 0.170 in winter, and 0.241, 0.428 and 0.625 in summer. Due to its similarity to the other three mediumsized estuaries (including similar species composition, Fig. 3), and because there was no other hypersaline estuary in the study area to use as a replicate for category 2h, the Gqutywa Estuary was grouped with the other

235

Figure 4.1. Mean macrobenthic density at each sampling site; (a) winter data; (b) summer data. Individual columns shown for each estuary represent a specific station sampled and are based on the mean of three samples. Standard errors are shown as whiskers. Stations are sequenced away from the mouth, with upper sites shown on the right side within each estuary.

Figure 4.2. Magnitude of Hill’s N0 (species richness) at each sampling site; (a) winter data; (b) summer data.

Figure 4.3. Magnitude of Hill’s N1 (diversity) at each sampling site; (a) winter data; (b) summer data.

medium-sized estuaries in all subsequent statistical comparisons. The Swartkops Estuary was also focused on in greater detail, because in several cases values measured at its sites appeared greater than in other estuaries of category 1. The null hypothesis was rejected using both Mann–Whitney U -tests and Kruskal–Wallis ANOVA on ranks for summer data only. P -values for the Mann–Whitney U -test for macrobenthic density, species richness and diversity were 0.002, 0.000 and 0.094, respectively, and corresponding p-values for the Kruskal–Wallis ANOVA on ranks were 0.005, 0.001 and 0.020. Consequently, in subsequent statistical comparisons of summer data, this estuary was excluded from category 1. The following results are based on combined information from Figure 4 and Table 4. Macrobenthic densities (Fig. 4.1) were generally low at stations sampled in category 1 estuaries, with the exception of the two upper sites in the Swartkops Estuary during summer (Fig. 4.1b). Category 1f estuaries had higher densities compared to category 1. Temporarily open/closed estuaries (categories 2 and 3) had greater macrobenthic densities when compared to open systems (categories 1 and 1f). Densities in small temporarily open/closed estuaries (category 3) were greater than those in medium sized temporar-

Table 4. One-tailed Mann–Whitney tests of winter and summer data. Only those comparisons are shown where a significant difference was found between the two data sets using a two-tailed Mann–Whitney. Null hypothesis tested: The magnitude of the variables in category B is equal or smaller than the magnitude of variables in category A (α = 0.05). The conclusion column shows the results of each test as a relationship between categories. Data from all sites of a particular category were pooled. Category 2h has been grouped with category 2. The Swartkops Estuary has been excluded from category 1 for the summer data

236

237 ily open/closed estuaries (category 2). In most cases, highest macrobenthic densities in estuaries in categories 2 were found at sandy sites close to the mouth region. There was no consistent trend of such a spatial pattern in any of the other categories, and for that reason, sandy and muddy sites were not compared separately for different estuarine categories. Similar trends were found for species richness (N0) (Fig. 4.2). This ecological index was lower in category 1 estuaries than in any of the other categories, and it was greater in small temporarily open/closed estuaries than in medium sized temporarily open/closed estuaries (categories 3 and 2, respectively). Again, higher values were determined for those sites of category 2 located closest to the mouth. Estuaries of category 1f had N0 values equal to those determined for small estuaries. This clearly set this group apart from other open systems, with the exception of the Swartkops Estuary. However, the total number of species found in small estuaries was lower than in estuaries of category 1f (Table 4). Species richness at each site is thus the same at individual sites within both estuarine categories, but overall, marine dominated estuaries provide habitat to significantly more species, which further sets this group apart from the other categories. The results of Hill’s N1 (Fig. 4.3) determined for the different estuaries closely resembled those of Hill’s N0, with category 1 having the lowest values, categories 1f and 3 being equal, and category 3 having greater values than category 2. A Wilcoxon signed rank test comparing summer and winter data of all estuaries combined showed no differences for macrobenthic density and diversity (Table 5). A significant difference in species richness between summer and winter samples was found in the Swartkops Estuary only. Ordination The stress value for the 2-dimensional NMDS ordination plot (Fig. 5) was high (0.24). However, the number of sites was greater than 50 (see Clarke & Warwick, 1994), and it was more important to consider the grouping of environmental variables and dominant taxa rather than the proximity of site scores to each other. Site arrangement and magnitude of salinity do not show any relationship (Fig. 5a), suggesting that salinity plays only a minor role in determining community patters. Fig. 5b indicates that the nature of the sediment, rather than salinity, is the major determinant of observed faunistic zonation patterns.

Consequently, sites were grouped into two major zones both based on the mud content and on the presence of typical estuarine species: a sand zone, whose sites had a mud content no greater than 5% and a mud zone, whose sediment contained more than 5% mud (Fig. 5c). The list from which representative taxa were chosen included the following: Capitella capitata complex (Ccap) and Iphinoe truncata (I) (both present in 13 estuaries), Cyathura estuaria (C) and Leech 1 (L) (both present in 12 estuaries), Urothoe serrulidactylus (U) and Assimenia sp. 1 (Ass) (both present in 11 estuaries), Prionospio sp. (Prio), Amphipod 1 (A1), Xenathura sp. (X), and Macoma litoralis (M) (all present in 10 estuaries), and Anopsilana ?fluviatilis (Aflu), Apseudes digitalis (A), Chironomid larvae (Chi) and Corophium triaenonyx (Ctri) (all present in 9 estuaries). The sand zone was dominated by the cumacean Iphinoe truncata (I) and the amphipod Urothoe serrulidactylus (U), while Amphipod 1 (A1), Leech 1 (L), and ?Assimenia sp. 1 (Ass) were also common. The mud zone was dominated by four estuarine endemics, namely the tanaid Apseudes digitalis (A), the isopod Cyathura estuaria (C), the polychaete Dendronereis arborifera (D), and the bivalve Macoma litoralis (M). Representatives of the stenohaline marine sandy fauna were not found in temporarily open/closed systems, and they were present only in a few of the six larger permanently open estuaries (mostly Kromme, Swartkops and Kariega) (Teske & Wooldridge, unpubl. data). Because of this, they are not represented in Figure 5c, and confirmation of the borders of the sand zone is based almost exclusively on estuarine endemics, which were abundant both in temporarily open/closed and permanently open estuaries. A third zone was identified in summer only in addition to the two major zones described above. At most sites within this zone, the sediment contained a large amount of mud (Fig. 5b), and water salinity was close to zero (Fig. 5a). This oligohaline zone was dominated by chironomid larvae (Chi). Oligochaetes were also common, and were particularly characteristic as none were found in the other zones. Even though the more typical sand zone species (U. serrulidactylus and I. truncata) were absent, species composition in the oligohaline zone was more similar to that of the sand zone than to that of the mud zone, which explains the overlap between sand zone and oligohaline zone. All sites in the oligohaline zone were located upstream of

238

Figure 5a. NMDS plot of summer data showing sites and superimposed magnitude of salinity; acronyms refer to the stations in the estuaries studied. Borders of the biotic zones on this and subsequent figures are based on consensus between Figures 5b and 5

Figure 5b. Mud content.

typical mud zone sites. In this case, low salinity played a greater role than sediment characteristics. Little correlation was found between biotic zones and estuarine categories (Fig. 5d). All 13 estuaries had sites located in the sand zone, as well as sites located in the mud zone. Sites with near freshwater salinities were encountered exclusively in the upper reaches of estuaries of category 1, namely in the Sundays, Great Fish and Keiskamma Estuaries. However, it can not be ruled out that several of the other estuaries may have an oligohaline zone beyond the navigable section.

Macrobenthic density increases from top to bottom in Figure 5e. As found previously, highest densities were measured in small estuaries (category 3) and at individual sites of marine dominated estuaries (category 1f). Most of the sites of estuaries receiving a high freshwater input are located in the top region of the plot, whereas sites of medium-sized estuaries (category 2) are located towards the bottom. Highest counts for both species richness (N0) and diversity (N1) were found at sites that were difficult to assign to either of the biotic zones (Fig. 5f, g).

239

Figure 5c. Most common 14 taxa (taxon codes of most abundant and second most abundant taxon shown at every site).

Figure 5d. Network of estuarine categories.

At these sites, fauna associated with sand and fauna associated with mud were more likely co-occur than at other sites. Site scores representing mud fauna in its purest form are present towards the left of the plot, and they have lower species richness and diversity values. There was little indication of such a trend in the sand zone. Many sandy sites from which mud zone species were absent were not only inhabited by estuarine sand zone species, but also by marine species, particularly in the freshwater deprived systems.

Discussion No obvious differences in macrobenthos density and diversity were recorded between summer and winter samples (Table 5). Species richness was significantly greater in the Swartkops Estuary during summer, when its catchment area was the only one to experience a dry period (temperature in ◦ C > 2× precipitation in mm, Walter & Lieth, 1960) (SA weather bureau). The relatively low freshwater input resulted in an increased marine dominance, which may have favoured marine species living in the estuary. In all other estuaries, sea-

240

Figure 5e. Macrobenthic density.

Figure 5f. Hill’s N0.

sonal effects were of little importance. However, high variances and low power of the test employed renders conclusions from this study tentative. A separate study including a smaller number of estuaries and a greater number of samples from each estuary should be more suitable to address this issue. Even though species composition differed considerably in estuaries of categories 1 and 3 (Fig. 3), the grouping of estuaries shown in Table 2 was generally replicated when comparing the different indices used. Data from the Old Woman’s Estuary show that even in a case where species composition was very different

from that of the other estuaries in its category, ecological indices, particularly species richness, were of equal magnitude. The Gqutywa and the Swartkops Estuaries did not fall into the original format. In case of the Gqutywa Estuary, statistical tests suggested that there was no justification for grouping this estuary separately from the other estuaries of similar size. Also, a species composition similar to that found in other medium-sized estuaries suggests that hypersalinity had little effect on the macrobenthic fauna of this estuary. Figure 5a demonstrates that salinity is of minor importance in

241

Figure 5g. Hill’s N1.

determining zonation patterns in Eastern Cape estuaries; true estuarine species are able to tolerate a wide range of salinities. Many estuarine species in South Africa survive salinities of up to 55 g kg−1 (de Villiers & Hodgson, 1999), which explains why salinity values of up to 39.5 g kg−1 measured in the Gqutywa Estuary did not result in the exclusion of taxa present in medium-sized estuaries. In the Swartkops Estuary, macrobenthic density, species richness (N0) and diversity (N1) were higher than in other estuaries in category 1 during summer (Fig. 4.1b, 4.2b and 4.3b). It is unlikely that the effects of urbanisation and industrialisation are responsible for this difference, as one would expect the values to be lower than in the other three estuaries in category 1. The Swartkops receives a comparatively low input of freshwater. The oligohaline zone found in the upper reaches of the other three estuaries in category 1 during summer was absent from this estuary (Fig. 5). Instead, salinity values in the upper reaches were approaching 30 g kg−1 . The estuary therefore tends towards marine dominance, which in turn favoured marine associated species present in the estuary. This explains why its species composition was similar to that of the freshwater deprived Kromme and Kariega Estuaries (Fig. 3). Hence, due to the elevated values determined, this estuary tends towards the pattern shown for category 1f systems. Apart from these two cases, results supported the hypothesis that the categories could be distinguished

on the basis of macrobenthic density, species richness and diversity. In each case, all indices of category 1 systems were smaller than those measured for category 1f, 2 and 3, whereas those of category 2 estuaries were always smaller than the ones of category 3 (Table 4). The two estuaries in category 1f (marine dominated systems) had comparatively low density values; by contrast, N0 and N1 values were high. This is probably due to the fauna resident in these estuaries having a large number of eury- and stenohaline marine species in addition to the estuarine endemic species. Many polychaete taxa found in the Kromme, Kariega, and also the Swartkops, are absent from other estuaries (Table 3). In summary, low freshwater input correlated with a greater number of marine associated species present in the estuaries and, consequently, gave rise to a negative correlation between freshwater input and species richness observed in open estuaries. The total number of species found in temporarily open/closed estuaries was low compared to estuaries of category 1f (Table 3). The relatively high species richness at individual sites of temporarily open/closed estuaries can be best explained by focusing on small estuaries (category 3). In these systems, species found in the mouth region in larger estuaries (e.g. Urothoe serrulidactylus, Iphinoe truncata, Grandidierella lutosa and Prionospio sexoculata) were frequently found at the same sites as species which in larger estuaries are characteristic of the middle and upper reaches (e.g. Cyathura estuaria, Macoma litoralis

242 Table 5. Wilcoxon paired sample testing of the null hypothesis of no difference in macrobenthic density (ind m−2 ), Hill’s N0 (species richness) and Hill’s N1 (diversity) between summer and winter samples for each estuarine category (α = 0.05). The Gqutywa Estuary has been included in category 2, whereas the Swartkops Estuary (SW) has been excluded from category 1 Variable

Category

N

W

T+

T−

Density

1 SW 1f 2 3

27 6 19 27 9

−86 19 32 −59 −22

119.5 20.0 101.5 97.0 3.0

−205.5 −1.0 −69.5 −156.0 −25.0

0.253 0.063 0.495 0.346 0.078

N0

1 SW 1f 2 3

27 6 19 27 9

−34 21 47 −62 2

109.5 21.0 109.0 74.0 11.5

−143.5 0.0 −62.0 −136.0 −9.5

0.589 0.031 0.325 0.261 0.844

N1

1 SW 1f 2 3

27 6 19 27 9

52 7 29 −9 −4

176.0 14.0 100.0 122.0 12.0

−124.0 −7.0 −71.0 −131.0 −16.0

0.466 0.563 0.551 0.900 0.813

and Chironomid sp.) (Teske & Wooldridge, unpubl. data). In open estuaries, spatial separation between these two groups is thus more clearly developed than in temporarily open/closed systems. Diversity (N1) was lowest in estuaries of category 1. These estuaries receive relatively high inputs of freshwater, and fluctuations in salinity are more extreme because of variations in freshwater inflow. Only hardy estuarine species that are able to tolerate such challenging conditions were sufficiently abundant to be included in the N1 value. Fluctuations in freshwater inflow may also be responsible for the consistently low density values measured in the estuaries of this category, but anthropogenic factors are also likely to play a role. The Sundays Estuary is strongly affected by agricultural run-off, particularly by pesticides from the citrus industry which abuts a long portion of the river. The water in both the Great Fish and Keiskamma Estuaries is turbid (Secchi depths as low as 1 cm during summer), which is the result of sediment loading due to extensive soil erosion in the rivers’ catchments. Particularly in the Keiskamma Estuary, large amounts of soft silt accumulating in the lower reaches (Fig. 5b) are likely exclude colonisation by the macrobenthos.

P

In summary, the hypothesis that permanently open estuaries had greater macrobenthic densities, species richness and diversity, was partly rejected, as values determined in category 1 estuaries were significantly lower in the majority of cases. Species composition not only depends on the type of estuary, but possibly also on geographical position. Excluding freshwater deprived estuaries, the length of estuaries is negatively correlated with the magnitude of their ecological indices. A possible explanation is that smaller estuaries are fed by rivers that originate close to the coast and consequently are neither affected by anthropogenic disturbances, nor by extensive fluctuations in freshwater inflow. In contrast, plankton counts have shown to be greatest in river dominated permanently open systems (particularly Keiskamma and Great Fish), and were low in all temporarily open/closed systems (Teske, pers. obs.; Wooldridge, unpubl. data). The lack of variation experienced in smaller estuaries may thus favour the macrobenthos as compared to other groups of estuarine consumers. All estuaries studied have a sand zone in the lower reaches and a mud zone in the upper reaches. The estuarine endemic faunas of both zones are similar in

243 all estuaries. An additional marine associated fauna is common in the sand zone of freshwater deprived estuaries, whereas oligohaline species are found primarily in the upper reaches of river dominated permanently open estuaries. Temporarily open/closed estuaries are inhabited almost exclusively by estuarine endemic species, which are present in all Eastern Cape estuaries.

Acknowledgements Thanks to Eberhard Teske, Nadine Strydom, Ané Oosthuizen, Ntsikelelo Sambokwe, Quentin van Staden and Larraine Becker for assistance with fieldand laboratory work. Niel Bruce, Wim Vader, Mary Bursey, Andrew Mackie, Elín Sigvaldadóttir, Steven Weerts and Mary Elizabeth Petersen are thanked for advice and suggestions on identification, taxonomy and distribution of many of the macrobenthic taxa. We also thank two anonymous referees for their suggestions. This study was made possible by a Prestige Bursary from the National Research Foundation (NRF) awarded to PT.

References Branch, G. M., M. L. Branch, C. L. Griffiths & L. E. Beckley, 1994. Two Oceans – a guide to the marine life of Southern Africa. David Philips Publishers (Pty) Ltd, Cape Town & Johannesburg. Clarke, K. R. & R. M. Warwick, 1994. Change in marine communities: an approach to statstical analysis and interpretation. Natural Environment Research Council, U.K. Hitchings & Mason Ltd., Plymouth, U.K. Day, J. H., 1967. A monograph on the Polychaeta of southern Africa. Trustees of the British Museum (Natural History), London. Day, J. H., 1974. A guide to marine life on South African shores. A. A. Balkema, Cape Town & Rotterdam.

De Villiers, C. & A. Hodgson, 1999. Studies on estuarine macroinvertebrates – The macrobenthos. In Allanson, B. R. & D. Baird (eds), Estuaries. Cambridge University Press: 167–191. Department of Water Affairs, 1986. Management of Water Resources of Southern Africa. Pretoria: University Printers. Griffiths, C. L., 1976. Guide to the benthic marine amphipods of Southern Africa. The Rustica Press (pty.) ltd, Wynberg, Cape. Grindley, J. R., 1976. Report on ecology of Knysna estuary and proposed Braamkraal Marina. Rondebosch, UCT School of Environmental Studies. Hill, M. O., 1973. Diversity and evenness: A unifying notion and its consequences. J. Ecol. 54: 427–432. Jackson, L. F. & S. Lipschitz, 1984. Coastal sensitivity atlas of Southern Africa/ Kussensiwiteitsatlas van Suidelike Afrika, National Book Printers, Goodwood. Kensley, B., 1978. Guide to Marine Isopods of Southern Africa. Trustees of the South African Museum, Cape Town. Morant, P. & N. Quinn, 1999. Influence of man and management of South African estuaries. In Allanson, B. K. & D. Baird (eds), Estuaries of South Africa. Cambridge University Press: 289–320. Ruppert, E. E. & R. D. Barnes, 1994. Invertebrate Zoology. Saunders College Publishing, Harcourt Brace College Publishing, Fort Worth, Philadelphia, San Diego, New York, Orlando, San Antonio, Toronto, Montreal, London, Sydney, Tokyo. Steyn, D. G. & M. Lussi, 1998. Marine shells of South Africa – an illustrated collector’s guide to beached shells. Ekogilde Publishers, P.O. Box 178, Hartebeespoort 0216, Rep. of South Africa. Walter, H. & H. Lieth, 1960. Klimadiagramm-Weltatlas, VEB Gustav Fischer Verlag, Jena. Whitfield, A. K., 1992. A characterization of southern African estuarine systems. Sth. Afr. J. aquat. Sci. 16: 89–103. Whitfield, A. K., 2000. Estuarine databases in South Africa: available scientific information on individual South African estuarine systems. Available from the Internet: http://www.ru.ac.za/cerm/datab.html. Whitfield, A. K. & M. N. Bruton, 1989. Some biological implications of reduced freshwater inflow into eastern Cape estuaries: a preliminary assessment. Sth. Afr. J.Sci. 85: 691. Wooldridge, T. H., 1999. Estuarine zooplankton community structure and dynamics. In Allanson, B. K. & D. Baird (eds), Estuaries of South Africa. Cambridge University Press: 141–166. Zar, J. H., 1974. Biostatistical Analysis. Prentice-Hall, Inc., Englewood Cliffs, N.J.

A comparison of the macrobenthic faunas of ...

ders College Publishing, Harcourt Brace College Publishing,. Fort Worth, Philadelphia, San Diego, New York, Orlando, San. Antonio, Toronto, Montreal, London, ...

1MB Sizes 0 Downloads 204 Views

Recommend Documents

A Probabilistic Comparison of the Strength of Split, Triangle, and ...
Feb 4, 2011 - Abstract. We consider mixed integer linear sets defined by two equations involving two integer variables and any number of non- negative continuous variables. The non-trivial valid inequalities of such sets can be classified into split,

Comparison of Square Comparison of Square-Pixel and ... - IJRIT
Square pixels became the norm because there needed to be an industry standard to avoid compatibility issues over .... Euclidean Spaces'. Information and ...

A Comparison of Medium-Chain
From the Department of Surgery, Peking Union Medical College Hospital, Beijing, China, and the ..... An infusion of fat-free nutrition solution was started at. 8:00 A.M. and .... using standard software StatView SE.25 Results were ex- pressed as ...

A Comparison Study of Urban Redevelopment Strategies_final.pdf ...
Whoops! There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. A Comparison Study of Urban Redevelopment Strategies_final.pdf. A Comparison Study of Urban Redevelo

On the contact domain method: A comparison of ...
This work focuses on the assessment of the relative performance of the so-called contact domain method, using either the Lagrange multiplier or the penalty ...

15 A Structured Comparison of the Goodman ...
In this chapter, we present an extensive and structured comparison of basic forms of .... values taken by NBi and NWi , or by Bi and Wi , once the aggregate data ...

economics of the internet: a comparison between
using the number of Internet hosts for Western Europe as dependent ... (HOST), telephone lines per 1000 people (TELLINE), rural population as a ..... countries, SMS through mobile phones has become affordable and a popular mode of ... Hargittai, E.,

a comparison of methods for determining the molecular ...
2009). The spatial and mass resolution of the cosmological simulation is matched exactly to the spatial and mass resolution of the simulation with the fixed ISM. Due to computational expense, the fully self- consistent cosmological simulation was not

A comparison of ground geoelectric activity between three regions of ...
A comparison of ground geoelectric activity between three regions of different level of ..... To go further inside in the comparison of our data sets, we constructed ...

A comparison of communication models of traditional ...
and accordingly, how it impacts health care providers' communication of instructions ... ing health care executives currently use or plan on .... lines, the Internet, an Integrated Services Dig- ... col used for the study of virtual visits in home ca

1 Institutional Accountability: A Comparison of the ...
Jan 18, 2017 - Bank of New York 2016), up from $390 billion in nominal dollars ten years prior (Federal. Reserve Bank of ...... Business is up in keeping default rates down. The Chronicle ... 28 April 2016. Six recent trends in student debt.

A Total Cost of Ownership Comparison of MongoDB & Oracle - Media16
6. Maintenance and Support. 9. Summary. 9. Topline Implications of Using MongoDB. 9 ... Because the choice of MongoDB versus a relational database is not the major driver of these .... Standard Edition) plus Oracle Real Application Clusters.

A comparison of ground geoelectric activity between three regions of ...
ing exponents for short and large lags arisen from crossover points in the geoelectric ... we introduce the method of data processing; in Sect. 4 the re- sults of the ...

A Comparison of Engine Performance and Emissions of Fusel Oil ...
A Comparison of Engine Performance and Emissions of F ... d Gasoline Mixtures at Different Ignition Timings.pdf. A Comparison of Engine Performance and ...

A comparison of numerical methods for solving the ...
Jun 12, 2007 - solution. We may conclude that the FDS scheme is second-order accurate, but .... −a2 − cos bt + 2 arctan[γ(x, t)] + bsin bt + 2 arctan[γ(x, t)]. − ln.

Performance comparison of a novel configuration of beta-type ...
Performance comparison of a novel configuration of beta-type Stirling engines with rhombic drive engine.pdf. Performance comparison of a novel configuration ...

comparison of techniques
Zircon. Zr [SiO4]. 1 to >10,000. < 2 most. Titanite. CaTi[SiO3](O,OH,F). 4 to 500. 5 to 40 k,c,a,m,ig,mp, gp,hv, gn,sk. Monazite. (Ce,La,Th)PO4. 282 to >50,000. < 2 mp,sg, hv,gp. Xenotime. YPO4. 5,000 to 29,000. < 5 gp,sg. Thorite. Th[SiO4]. > 50,000

Comparison of MINQUE and Simple Estimate of the ... - Springer Link
1,2Department of Applied Mathematics, Beijing Polytechnic University, Beijing ... and a Project of Science and Technology of Beijing Education Committee.

Comparison of electrochemical techniques during the corrosion of X52 ...
J. Genesca, R. Galvan-Martinez, ... G. Garcia-Caloca, R. Duran-Romero, J. Mendoza-Flores, .... In order to analyze the measured electrochemical noise data.

Comparison of Results
Education Programs Office. The authors would also like to ... M.S. Thesis, Virginia Polytechnic Institute and State. University, Blacksburg, Virginia, 2000.

Comparison of electrochemical techniques during the corrosion of X52 ...
2 shows the best fitting parameters obtained in the nu- merical analyses. In this table ... ing, at each analysed frequency, the power spectral density. (PSD) of the ...

Comparison of MINQUE and Simple Estimate of the ... - Springer Link
Vol.19, No.1 (2003) 13–18. Comparison of MINQUE and Simple Estimate of the. Error Variance in the General Linear Models. Song-gui Wang. 1. , Mi-xia Wu. 2.