at SciVerse ScienceDirect

Estuarine, Coastal and Shelf Science 125 (2013) 10e19

Contents lists available

Estuarine, Coastal and Shelf Science

journal homepage: www.elsevier .com/locate/ecss

Influence of habitat structure and mouth dynamics on avifauna of intermittently-open estuaries: A study of four small South African estuaries

Anja I. Terörde a,1, Jane K. Turpie b,*

a Percy FitzPatrick Institute, University of Cape Town, Rondebosch 7701, South Africab South African Institute of Aquatic Biodiversity, Grahamstown 6140, South Africa

a r t i c l e i n f o

Article history:Received 26 September 2011Accepted 16 March 2013Available online 6 April 2013

Keywords:aquatic birdsaquatic plantsestuarine dynamicshabitat selectiontemporarily open/closedsubmerged macrophytes

Regional terms:South AfricaEastern Cape ProvinceCoast between Port Alfred (33�3505400S,26�5303900E)Hamburg, (33�1702500S, 27�2704200E)

IOE, intermittently open estuary; ICE, intermitpermanently open estuary.* Corresponding author. Anchor Environmental Con

Silverwood Close, Tokai 7945, South Africa.E-mail addresses: anjater77@gmail.com

anchorenvironmental.co.za, jane.turpie@uct.ac.za (J.K1 Present address: PO Box 390, Port Alfred, 6170, S

0272-7714/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.ecss.2013.03.005

a b s t r a c t

Habitat composition was a major factor in determining waterbird species composition and abundance,particularly the area of floodplain and vegetated channel in four intermittently open estuaries (IOEs) inthe warm-temperate coastal biogeographical region of South Africa. Average bird densities on the fourestuaries varied from 0.5 to 4.2 birds per hectare, and community composition differed significantlybetween estuaries. However, the considerable variation in abundance of macrophytes did not have adetectable effect on waterbirds. Under closed mouth conditions, piscivorous birds dominated theavifauna. Each estuary’s avifauna responded differently in terms of changes in feeding guild compositionwhen the mouth opened. Bird abundance changed immediately after breaching, but not consistently.Diversity was significantly higher under open-mouth conditions for three of the four estuaries, andspecies composition was significantly different from that under closed-mouth conditions at all fourestuaries. Changes in mouth dynamics as a result of climate change, water abstraction and artificialbreaching could lead to significant changes in estuarine fauna.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Estuaries provide important habitats for a number of waterbirdspecies at global and national scales (Hockey and Turpie, 1999;Turpie et al., 2002; Calbrade et al., 2010), but the way in which theyare used is influenced by their type, size and location. While mostwork on estuarine avifauna has taken place on large, permanently-open systems which often attract large aggregations of birds rela-tively little is known about the avifauna of small, intermittently-open or closed estuaries (Teroerde and Turpie, 2012). These estu-aries occur worldwide and are the commonest type of estuary inSouth Africa.

There are 290 estuaries in South Africa, of which 70% are smallsystems that are closed for varying amounts of time (Lamberth and

tently closed estuary; POE,

sultants, 8 Steenberg House,

(A.I. Terörde), jane@. Turpie).outh Africa.

All rights reserved.

Turpie, 2003; Harrison, 2004). These temporarily open/closedsystems are termed either intermittently-open estuaries (IOEs),which are closed most of the time, or intermittently closed estu-aries (ICEs), which are predominantly open to the marine envi-ronment. They are distinctly different from permanently openestuaries (POEs), due to the largely unpredictable temporal varia-tion in habitat and food availability (Whitfield and Bate, 2007).

Small temporarily open/closed estuaries are not only in themajority in some areas, but are also understood to be relativelysensitive to changes in water supply and management in-terventions (Whitfield et al., 2012). Thus, it is important to under-stand the biodiversity in these systems in order to make informedmanagement decisions. The main aim of this study was to inves-tigate the avifauna of a variety of IOEs and identify factors influ-encing community composition and abundance. Specifically, thefollowing key questions were addressed:

� What characterises the avifauna of intermittently openestuaries?

� How does habitat composition influence the bird communitycomposition?

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e19 11

� Howdo bird abundance, community composition and diversitychange with fluctuations in estuary mouth state andvegetation?

2. Materials and methods

2.1. Study site

This study was conducted at four intermittently-open estuariesalong the Eastern Cape coast of South Africa, between the towns ofPort Alfred (33�3505400S, 26�5303900E) and Hamburg (33�1702500S,27�2704200E): the Riet (33�330S, 27�000E), West Kleinemonde(33�320S, 27�020E), East Kleinemonde 33�320S, 27�030E and Bira(33�220S, 27�190E) estuaries (Fig. 1). Rainfall in this area is not asseasonal as in other parts of the country (Stone, 1988). During thestudy period (2005-11), rainfall occurred in all months, with peaksin December 2005 and September 2006 and ranged from 7.6 mm inJuly 2006 to 209.2 mm in August 2006. Mean monthly air tem-peratures ranged from 7.9 �C in July to 26.7 �C in February (SouthAfrican Weather Bureau). The catchments of the estuaries alongthis coast are mainly used for cattle and pineapple farming, and allof the estuaries have small resort developments on their lowerbanks and are used for recreational activities such as fishing,swimming and power boating.

The Riet estuary is largely supratidal and only shows tidal in-fluence above the mouth when a deep channel is formed after veryheavy rainfall. The West Kleinemonde estuary has large areas offloodplains (approximately 58.3 ha) which are inundated whenwater levels in the estuary are high. The mouth is generally morestable than the adjacent East Kleinemonde estuary and often re-mains closed for periods exceeding one year. Salinity in the estuaryunder closed conditions ranges from 5 to 10 in the upper reaches to25e30 in the lower reaches. Water temperatures show a highvariation with a summer range between 22 �C and 29 �C and awinter minimum of 12 �C (Cowley et al., 2003).

The East Kleinemonde estuary is mostly shallow with a mainchannel depth of 1e2 m (Cowley and Whitfield, 2002). Tidal in-fluence can be strong after breaching events caused by large riverfloods which scour a deep channel in the mouth region. Between

Fig. 1. Map of the study area and sampled e

1993 and 2006 the estuary was closed to the sea 74% of days,experienced overwash on 11% of days and was completely open on15% of days (van Niekerk et al., 2008). The longest recorded periodthat the estuary was connected to the marine environment was 32days in 2004. River base flow and tidal exchange in this system areinsufficient to maintain open mouth conditions and the mouthgenerally closes a few days after river flooding has subsided (vanNiekerk et al., 2008). Mean monthly salinity ranges from 0 to 25,depending on mouth condition and freshwater inflows (Cowleyet al., 2003).

The Bira estuary was the largest of the four systems studied. Itcan stay open for extended periods of time (several weeks tomonths), but once closed, it usually remains so for several months.

2.2. Bird counts

2.2.1. Seasonal monitoringAll four estuaries were sampled from a motorised boat or on foot

along their navigable length. Sampling was carried out fromDecember 2005 until November 2006. Each estuary was dividedinto four regions and all water-associated birds encountered in eachregion were recorded. Terrestrial-feeding birds such as Black-headed Heron, Cattle Egret and Brown-hooded Kingfisher wereexcluded from the study. The status of each estuary mouth (open/closed) was recorded daily during the entire study period. Eachestuary was counted one to three times per month on randomly-selected days over the period of one year, apart from during theperiod immediately after breaching, when counts took place everyday for a week. A total of 110 waterbird counts were conducted ofwhich 52 and 58 were carried out under open and closed estuarymouth conditions, respectively (Table 1). Summer was defined asNovember to April and winter as May to October, according to thepresence or absence of themajority of Palearctic-breedingmigrants.

2.2.2. Long term monitoringFollowing the intensive first year of monitoring, summer and

winter bird counts were conducted at the East Kleinemonde andWest Kleinemonde estuaries until February 2011, over a periodduring which macrophyte cover was recorded in detail by Riddin(2011).

stuaries (modified from Walton, 1984).

0

5

10

15

20

25

0123456789

10Riet

00

100

200

300

400

500

50100150200250300350400450 West Kleinemonde

0

20

40

60

80

100

120

0

10

20

30

40

50

60

70East Kleinemonde

9060

01020304050607080

0

10

20

30

40

50

Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov

Bira

Waterfowl Invertebrate feedersDiving piscivores Wading piscivoresSwimming piscivores Total

no data

Fig. 2. Variation in the mean monthly numbers of birds of five groupings and total birdnumbers over the period of one year. The arrow denotes main breaching event whenthe estuary changed from closed to open.

Table 1Summary of estuary characteristics (Badenhorst, 1988;Vorwerk, 2000; Cowley andWhitfield, 2002; Whitfield and Bate, 2007; Hughes, 2008) and the number of birdcounts conducted between December 2005 and November 2006.

Riet WestKleinemonde

EastKleinemonde

Bira Total

Length (km) 2 7 3.5 9Max width (m) 50 330 120 300Survey area (ha) 8.9 101.2 34 116.3 260.4MAR (m3$yr�1) 2 � 106 4 � 106 2 � 106 13 � 106

Catchment(km2)

48 94 46 255

Bird countsMouth closed 9 21 19 9 58Mouth open 13 12 15 12 52Total 22 33 34 21 110

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e1912

2.3. Habitat characteristics

Seven habitat types were identified: overhanging trees/rocks,exposed banks, reedbeds, inundated floodplains, shallow margins,unvegetated channel and vegetated channel (with submergedmacrophytes). The area of each habitat type in four sections of eachestuarywas estimated using percentage cover estimates from aerialphotographs.

2.4. Data analysis

Count data were summarised by taxonomic group, species,migratory status and feeding guild over the whole year and bymouth condition. Species were grouped according to dominantdietary items into piscivores, herbivores (ducks, geese, coots,moorhen) and invertebrate-feeding waders. Piscivores weregrouped further according to feeding method into pursuit-swimming piscivores (cormorants, darters, grebes), aerial-divingpiscivores (kingfishers, raptors, terns, gulls) and wading pisci-vores (herons, egrets).

Count data were analysed using PRIMER v6 statistical software.BrayeCurtis similarity matrices were generated using untrans-formed count data for all estuaries combined and for each indi-vidual estuary. Non-metric multi-dimensional scaling (MDS) plotswere generated to examine the influence of estuary, mouth stateand season on avifauna. The ANOSIM and SIMPER routines (one-way and two-way cross-designs) were used to test for significantdifferences between groups in each analysis and to determine therelative contribution of key species to the differences betweengroups. The ShannoneWiener Index was used to measure diversity.c2 tests of independence were used to compare feeding guildcomposition of the bird community under open and closed estuarymouth conditions.

Multiple regression analyses were used to investigate re-lationships between feeding guild abundance and habitat areas in16 estuary sections. As habitat size was calculated for closed con-ditions, only counts made under closed and full conditions wereused. All 16 estuary sections were treated as independent sampleson the basis of the assumptions that (1) the pool of birds that usethese systems was not limited to the studied estuaries and (2)birds moved freely between and within these systems and othersystems.

Long term monitoring data were related to available data onsubmerged macrophyte cover at the East Kleinemonde estuary(Riddin, 2011) and West Kleinemonde. Spearman rank coefficientswere calculated to analyse correlations between bird numbers andsubmerged macrophyte cover.

3. Results

3.1. Overall abundance

Waterbird abundance varied greatly between the four sampledestuaries (Fig. 2). Count data at all four estuaries were normallydistributed (KolmogoroveSmirnov). The Riet estuary supportedvery few birds. The difference between the mean number of birdsduring the closed and open phase was not significant (15.2 vs 16.7;t ¼ 2.086; df ¼ 20; p > 0.05). The West Kleinemonde estuary hadthe highest number of birds (Fig. 2), and overall bird abundancewas significantly higher when the estuary mouth was closed (441.0vs 263.0; t ¼ 2.040; df ¼ 31; p < 0.001). There was no significantdifference in mean bird numbers between open and closed estuarymouth conditions at the East Kleinemonde (76.5 vs 86.9; t ¼ 2.037;df¼ 32; p> 0.05), but at the Bira, numbers were significantly lowerwhen the estuary mouth was closed (35.3 vs 66.0; t ¼ 2.093,df ¼ 19; p < 0.001).

Bird abundance was not correlated with estuary size. The WestKleinemonde estuary had the highest density of waterbirds with amean number of 3.6 birds per hectare, followed by the East

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e19 13

Kleinemonde estuary with 2.3 birds/ha. The small Riet estuary hada higher bird density (1.8 birds/ha) than the large Bira estuarywhich only had a mean number of 0.5 birds/ha.

3.2. Community composition

The Riet estuary had the lowest diversity of waterbirds with 25recorded species from 13 Families in four Orders (Table 2) and aShannon’s diversity index of 1.85. East Kleinemonde had the secondlowest Shannon’s diversity index (2.26).West Kleinemonde had thehighest diversity of waterbirds and the highest Margalef’s speciesrichness score (4.13) with a total of 54 recorded species from 18Families in 5 Orders. Bira estuary’s Shannon diversity score wasslightly higher (2.33) than that of the West Kleinemonde estuary(2.32).

While theWest Kleinemonde estuary had a higher species countand Margalef’s species richness score than Bira estuary, the popu-lation there was dominated by a few species, namely Yellow-billedDuck and Reed Cormorant. Charadriiformes formed the largestOrder at the East Kleinemonde (48.6% of recorded species), WestKleinemonde (51.9%) and Bira (50%) estuaries. Ciconiiformes wasthe largest Order at the Riet estuary (45.8%; Table 2). Piscivorousbird species comprised the largest group in terms of species di-versity on all four estuaries followed by invertebrate-feeding spe-cies and herbivorous species. Piscivorous birds also dominated theavifauna in terms of abundance at the Riet, West and East Kleine-monde estuaries, whereas invertebrate-feeding waders were thedominant group at the Bira (Fig. 2).

3.3. Factors influencing community composition

3.3.1. Influence of estuary, mouth state and season on birdcommunities

Estuary was the overriding factor determining the avifaunalcommunity composition (ANOSIM, R¼ 0.780, p< 0.001; Fig. 3), andestuaries were significantly different from one another (ANSOM,p < 0.001 for all combinations). Season and mouth state also had asignificant influence on avifauna (ANSOM, season: R ¼ 0.176,p< 0.001; mouth state R¼ 0.051, p< 0.01). For each estuary, mouthstate and season have a noticeable influence (Fig. 4), though thesewere only significant for the East Kleinemonde (mouth state:R ¼ 0.469, p < 0.05, season: R ¼ 0.415, p ¼ 0.05) and West Klei-nemonde estuaries (mouth state: R ¼ 0.695, p < 0.001, season:

Table 2Taxonomic composition of the avifauna at the four sampled estuaries.

Order Family

Anseriformes Anatidae Ducks, geeseCoraciiformes Alcedinidae Alcedinid kingfishers

Dacelonidae Dacelonid kingfishersGruiformes Rallidae Moorhens, cootsCharadriiformes Scolopacidae Greenshanks, whimbrels, stints,

sandpipers, ruff, turnstonesRostratulidae Painted-snipesBurhinidae Thick-kneesHaematopodidae OystercatchersRecurvirostridae StiltsCharadriidae Plovers, lapwingsLaridae Gulls, terns

Ciconiiformes Accipitridae Typical raptors, ospreysPodicipedidae GrebesAnhingidae DartersPhalacrocoracidae CormorantsArdeidae Egrets, heronsScopidae HamerkopThreskiornitidae Ibises, spoonbillsCiconiidae Storks

Total number of species

R ¼ 0.472, p < 0.001). However, the effects of season and mouthstate are confounded, because in most cases most of the closedestuary observations were in summer and vice versa, apart fromthe East Kleinemonde.

Feeding guild composition was significantly different underopen and closed estuary mouth conditions on all four estuaries (c2

values: Riet 57.1, West Kleinemonde 495.6, East Kleinemonde 36.3Bira 67.6; all p < 0.001). Overall, herbivorous waterfowl were moreabundant in winter than summer and in closed than open condi-tions (Fig. 5). Invertebrate feeding waders were slightly moreabundant in summer, diving piscivores were slightly more abun-dant in closed conditions, and swimming piscivores were moreabundant in winter. However, the overall pattern was not general,and patterns for all the groups differed markedly among the estu-aries (Fig. 6).

3.3.2. Influence of habitat areasA multiple regression analysis of bird numbers under closed

conditions with area of habitat types was significant for all feedingguilds (Table 3). Numbers of wading piscivores were significantlyinfluenced by the area of all habitats apart from vegetated channelwhereas the other groups were significantly influenced by one ortwo habitats. Waterfowl and benthic invertebrate feeders weremainly influenced by floodplain area. Arial diving piscivores werepositively influenced by area of shallow water habitat and vege-tated channel, while pursuit-swimming piscivores were positivelyinfluence by the area of unvegetated channel.

3.3.3. Influence of water levels and submerged macrophyte cover onavifauna

At the East Kleinemonde estuary, both water levels andmacrophyte abundance varied considerably over the study period,and the latter increased greatly during a sustained period of mouthclosure and stable water levels for about 12 months betweenAugust 2007 and August 2008 (Riddin, 2011). However, no re-lationships were found between bird numbers in any guild andeither water levels or macrophyte cover at this estuary.

4. Discussion

Very little attention has been paid to the many intermittently-open estuaries in South Africa and elsewhere, probably becauseof the relatively small size of both the estuaries and their

Riet East Kleine-monde West Kleine-monde Bira

3 3 7 51 2 2 22 2 2 20 0 2 02 8 10 5

0 0 1 01 1 1 11 1 1 00 1 1 03 5 6 61 2 7 41 1 2 11 1 1 11 1 1 13 3 3 25 5 4 40 0 0 10 2 2 00 0 1 0

25 38 54 35

Season

SummerWinter

2D Stress: 0.15

Estuary

RietEKMWestBira

2D Stress: 0.15

Mouth state

ClosedOpen

2D Stress: 0.15

Fig. 3. Two-dimensional MDS plot of bird count data for four estuaries, showing theinfluence of estuary, mouth condition and season.

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e1914

catchments. However, these estuaries, which cumulativelycontribute a significant portion to the South African estuarinecomplement, are increasingly threatened by development andwater use. Indeed because of their small size they are particularlysensitive to these changes (Reddering and Rust, 1990; Whitfield,1992; Cooper et al., 1999; Whitfield et al., 2012). It is thus impor-tant to improve our understanding of the biodiversity of thesesystems and how they relate to habitat structure and dynamics.

4.1. Birds of intermittently open estuaries

This study found that while the numbers and density of birds onsmall intermittently-open estuaries is low, it can reach notable levelson larger systems that have a relatively high habitat diversity. The

avifauna was dominated by piscivorous birds and waterfowl, whichis markedly different from large, open estuaries that tend to have amuch larger invertebrate-feeding wader component. The generalpredominance of piscivorous birds indicates that while estuariesprovide critical sheltered habitat for many fish species, smallintermittently-open systems do not provide a refuge frompredation.

4.2. How does habitat structure influence community composition?

Habitat diversity within and betweenwetland systems is knownto influence waterbird diversity (Brock et al., 1999), as differentspecies can occupy more niches in a diverse mosaic habitat(Perrings et al., 1992; Gray, 1997). In this study, each estuary had adistinct bird community composition, and habitat structure was amajor factor in determining community composition, through ef-fects on the abundance of different feeding guilds.

Floodplain habitat had the strongest influence on bird abun-dance. The West Kleinemonde estuary, which had the highestabundance and diversity of birds also had the largest area offloodplain. In addition, the channel was vegetated with submergedmacrophytes, which had a positive influence on the numbers ofpursuit-swimming piscivores. By contrast, the Bira estuary whichhad the lowest abundance and density of birds, had no visiblesubmerged macrophyte growth in its channel and also only a smallarea of floodplain habitat. The East Kleinemonde estuary had thesecond most diverse and dense bird community and also had arelatively large floodplain area. While submerged macrophytegrowth in the channel was low, wading and aerial-diving piscivoreswere attracted by the shallow margin habitat, which covered alarge portion of the estuary.

The abundance of pursuit-swimming piscivores was stronglycorrelated with vegetated channel habitat. The most abundantpursuit-swimming piscivorous species was Reed Cormorant, whichprefers to fish in water 30e200 cm deep (Whitfield and Blaber,1979). Possibly the concentration of prey fish is higher in waterswith submerged macrophytes as they provide shelter and refuge.Aerial-diving piscivores showed the weakest overall relationship tohabitat (adjusted R2 ¼ 0.21). Many of these species (for exampleterns and gulls) can feed over the entire water surface area and aretherefore less reliant on the availability of specific habitat types.Prey availability possibly plays a greater role in this feeding guild.Wading piscivore numbers showed a positive correlationwithmosthabitats, particularly with floodplain habitat, suggesting that theirpresence is a correlation of overall estuary size.

Invertebrate-feeding waders were most strongly correlatedwith floodplain habitat. Water levels in the floodplains on thestudied estuaries varied greatly, so that very shallow water andmargins provided a suitable feeding ground for these species. It isinteresting to note that they were not correlated with the presenceof exposed banks, which are relatively unproductive in theselargely non-tidal systems.

Herbivorous waterfowl correlated strongly with floodplainhabitat, and were deterred by reedbeds. Since waterfowl are relianton macrophyte and other plants for feeding and shelter this was anunexpected result. However it is likely that areas dominated byreedbeds, such as at the Riet estuary, may be less attractive towaterfowl. Indeed, the floodplain areas were well vegetatedincluding moderate reedmarsh fringes interspersed with suitableloafing and feeding areas.

4.3. How do breaching events and mouth condition affect theavifauna of IOEs?

As the estuary mouth breaches and water levels drop, com-munities of plants and animals colonising the substratum can be

Bira

East Kleinemonde

West Kleinemonde

RietMouth state

ClosedOpen

2D Stress: 0.14Season

SummerWinter

2D Stress: 0.14

Mouth state

ClosedOpen

2D Stress: 0.1Season

SummerWinter

2D Stress: 0.1

Mouth state

ClosedOpen

2D Stress: 0.17Season

SummerWinter

2D Stress: 0.17

Mouth state

ClosedOpen

2D Stress: 0.14

Season

SummerWinter

2D Stress: 0.14

Fig. 4. Two-dimensional MDS plot of bird count data for four estuaries, showing the influence of mouth condition and season for each estuary.

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e19 15

exposed (Perissinotto et al., 2000). Fish and invertebrates can bestranded on floodplains and immigration of fish and invertebratesinto remnant pools occurs, providing highly concentrated feedingopportunities for waterbirds (Kushlan, 1976). Bird numbersincreased immediately following breaching events on the Riet, EastKleinemonde and Bira estuaries suggesting that they were initiallyattracted to intermittently open estuaries following mouthbreaching and used these highly unpredictable events as feedingopportunities.

However, bird numbers decreased immediately following thefirst breaching event at the West Kleinemonde estuary, andcontinued to decrease thereafter under open-mouth conditions.

The West Kleinemonde estuary has a large floodplain area whichwas inundated when the estuary mouth was closed and waterlevels were high. Macrophyte cover was visibly high during theclosed phase of the study as the estuary had been closed to the seafor an extended period of time before the study. Herbivorouswaterfowl such as Yellow-billed Duck and Red-billed Duck favourfeeding in the littoral zone where aquatic vegetation is closest tothe surface and thus easily accessible (Whitfield and Cyrus, 1978),while wading piscivores such as Grey Heron and Great Egret huntfor prey in shallow waters of a maximum depth (Kushlan, 1976).These feeding guilds were attracted to the floodplains of the WestKleinemonde estuary in high numbers during the closed phase.

0

20

40

60

80

100

120

140

Herbivorous & omnivorous waterfowl

Benthic invertebrate

feeders

Diving piscivores

Wading piscivores

Swimming piscivores

All estuaries Summer Closed

Summer Open

Winter Closed

Winter Open

Fig. 5. Mean numbers of birds counted in five feeding categories under open andclosed conditions during summer and winter.

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e1916

As the estuary mouth opened, these floodplains drained within avery short period of time and a large area of wading and feedinghabitat was lost. Herbivorous waterfowl and pursuit-swimmingand wading piscivores departed. The tidal influence was low atthe West Kleinemonde estuary and water levels were extremelylow, even at low tide, so that the water surface area of the systemwas greatly reduced. Thus, the initial influx of birds which wasrecorded at the other three studied estuaries was outweighed bythe loss of available feeding habitat on this estuary. The relativelyhigh abundance and density of birds at the West Kleinemondeestuary can therefore be attributed to the extensive floodplainsfound in this system. As no other water bodies in the area areknown of by local bird watchers in which waterfowl occur insimilar numbers (A. Williams, pers. comm.), theWest Kleinemondeestuary could potentially be of regional importance for herbivorouswaterfowl.

At the Bira estuary a strong initial influx of birds was recordedwhen themouth opened, and bird numbers then dropped again butsettled at a higher level than when the mouth was closed. The Biraestuary showed a high tidal influence reaching approximately10 km up the estuary, and water levels remained relatively high athigh tide. The drop in water levels and tidal influence at Bira could

0

5

10

15

20

25

05

1015202530354045

Herbivorous & omnivorous waterfowl

Benthic invertebrate

feeders

Diving piscivores

Wading piscivores

Swimming piscivores

East Kleinemonde

Summer Closed

Summer Open

Winter Closed

Winter Open

0123456789

Herbivorous & omnivorous waterfowl

Benthic invertebrate

feeders

Diving piscivores

Wading piscivores

Swimming piscivores

Riet Summer Closed

Summer Open

Winter Closed

Winter Open

5

10

15

20

25

30

35

Fig. 6. Average numbers of birds counted in five feeding categories under open

have led to an increase in available wading habitat for birds with amaximumwading depth, such as egrets and herons, as deeper areasin the channel became shallower. The strong tidal influence is alsolikely to be the reason that pursuit-swimming piscivores wereattracted to the estuary after the mouth opened, as a narrowerchannel leads to a higher concentration of fish, making it easier tohunt underwater.

While there was a sharp initial influx of birds at the East Klei-nemonde estuary, bird abundance decreased thereafter to a similarlevel as when the estuary mouth was closed. While there was a lossof floodplain wading habitat after the opening, this did not affectoverall abundance. The floodplain, which lacks significant pans, didnot attract the large numbers recorded at the adjacent West Klei-nemonde and so this loss of wading habitat did not seem to haveaffected overall numbers. In addition a heronry of Great Egret andGrey Heronwas in use near the road during the opening events andthis is likely to have kept wading piscivores from leaving the area.They were observed flying in and out to feed, but remained at theestuary to breed as they had the previous summer.

Waterbird abundance did not increase significantly when themouth opened at the Riet estuary. This differed from the othersystems in that when it opened to the sea, no deep channel wasformed at the mouth and tidal influence was minimal, because thesystem was more perched. The water level dropped only slightly,with water trickling out at the mouth at a rate probably similar towater running in from the top. Most of the exposed sandbanksbecame supra-tidal rather than inter-tidal, which causes the fooditems of waders (small shallow-burrowing invertebrates) todesiccate (Perissinotto et al., 2004). Only a limited intertidal areawas therefore available for invertebrate-feeding waders and wad-ing piscivores to use as a feeding ground. Since the water level didnot change substantially after opening, there were no large-scalechanges in habitat available to waterbirds.

This study demonstrates that responses to mouth openingevents are highly variable and differ from system to system. The keyfactor which influences the abundance of birds in these systems ishow habitat availability changes with changes in mouth condition.Birds are highly mobile organisms and are attracted to estuariesfor a variety of reasons and seasonal migration patterns, food

Herbivorous & omnivorous waterfowl

Benthic invertebrate

feeders

Diving piscivores

Wading piscivores

Swimming piscivores

Bira Summer Closed

Summer Open

Winter Closed

Winter Open

0

0

0

0

0

0

0

0

Herbivorous & omnivorous waterfowl

Benthic invertebrate

feeders

Diving piscivores

Wading piscivores

Swimming piscivores

West Kleinemonde Summer Closed

Summer Open

Winter Closed

Winter Open

and closed conditions during summer and winter for each of four estuaries.

Table 3Multiple regression results of mean bird numbers and habitat sizes under closed mouth conditions.

Regression coefficient

Overall model Intercept Over-hangingtrees/rocks

Exposedbank

Reed-beds

Flood-plain

Shallowmargins

Unvegetatedchannel

Vegetatedchannel

Pursuit-swimming piscivores Adjusted R2 ¼ 0.47F(7,432) ¼ 56.5;p < 0.001

0.07 3.33 �3.07 �2.76 0.15 0.66 0.70*** 4.48

Aerial-diving piscivores Adjusted R2 ¼ 0.219F(7,432) ¼ 18.08p < 0.001

3.96 �0.12 �0.72 �11.76 �0.01 0.97** �0.06 0.95***

Wading piscivores Adjusted R2 ¼ 0.39F(7,432) ¼ 40.95p < 0.001

�2.79** 2.77** �2.69** 12.08** 0.92*** 2.35* 0.57 �0.29

Inverte- brate- feeding waders Adjusted R2 ¼ 0.27F(7,432) ¼ 24.74p < 0.001

1.709 �0.781 1.410 0.830 0.741*** 0.526 �0.010 �0.087

Waterfowl Adjusted R2 ¼ 0.46F(7,432) ¼ 54.0p < 0.001

1.67 �0.13 0.92 �13.68 1.95*** 0.19 �0.30 �0.64

***: p < 0.001; **: p < 0.01; *: p < 0.05.

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e19 17

availability, lack of disturbance, as well as habitat availability allplay a role (Ntiamoa-Baidu et al., 2000). The occurrence of birds ona system is likely to be influenced by conditions elsewhere as wellas conditions on the estuary itself. Piscivorous birds, waterfowl andwader numbers increased abruptly and significantly on the Oos-terschelde estuary after the artificial closure of the adjacent Gre-velingen estuary (Doornbos, 1987; Lambeck et al., 1989; Slob, 1989).

The results indicate that each estuary has a distinct avifaunalcommunity, which is related to the unique morphology and habitatcomposition of each estuary, which was therefore investigatedfurther. Also, species compositions in one particular estuary underopen and closed conditions resemble one another more closelythan species compositions of the samemouth condition at differentestuaries. Species composition and feeding guild compositionchanged significantly from the closed to the open mouth state, andtwo distinct communities were found at each estuary. There was achange from a piscivore-dominated community to a more diversecommunity with an increased proportion of invertebrate feedingwaders at the Riet, West Kleinemonde and East Kleinemonde es-tuaries. Only at Bira were more piscivores attracted to the estuaryunder open conditions, which can be attributed to the prevailinggeomorphological conditions and greater fish recruitment arisingfrom a prolonged marine connection. While a significant change inspecies composition still took place when the mouth opened, di-versity was not increased significantly. In general, however, theintermittently-open estuarine (IOE) avifauna is a piscivore-dominated community, which changes towards the type of com-munity typical of permanently open estuaries when the mouthopens. If an increase in the open-mouth phase duration takes placeit can be expected that the typical IOE avifaunawill be replaced. It isnot known to what degree waterbirds, particularly piscivores relyon intermittently open estuaries as feeding and roosting groundsand where they move if conditions become unfavourable.

4.4. How does submerged macrophyte cover influence theabundance of certain feeding guilds?

Submerged macrophytes form an important food source forherbivorous waterfowl found in South African estuaries (Heÿl andCurrie, 1985; Russell et al., 2009) and submerged macrophytecover is correlated to the abundance of certain feeding guilds(Stewart and Bally, 1985; Russell et al., 2009). A weak positiverelationship between herbivore numbers andmacrophyte densitiessuggested that any influence may have been difficult to detect

because of the low numbers of birds at the system and the fact thatthe birds appeared tomove regularly between this and the adjacentWest Kleinemonde estuary. Indeed, at the West Kleinemonde es-tuary, bird abundance increased greatly from 150 to 250 individualsper count during the open-mouth phase to over 1100 individualsper count one year after mouth closure, after which abundancedecreased. While no detailed data on submergedmacrophyte coverare available for this period, macrophyte data from 2007 suggestthat submerged macrophyte cover increased during this period ofmouth closure, as would be expected (Riddin and Adams, 2008).Thereafter, the estuary became very saline causing the submergedmacrophytes to die-off, which is probably the cause of the ensuingdecline in bird numbers. In the case of invertebrate foragers, anegative relationship might be expected due to reduced access tobare substrata, but would have been confounded by seasonalvariation in bird numbers, as well as factors such as water level.There were too few data to determine the trends involving multiplefactors.

4.5. How do densities of birds on intermittently open estuariescompare with other types of estuaries?

Overall waterbird densities ranged from 0.46 birds/ha at the Biraestuary to 3.64 birds/ha at the West Kleinemonde estuary. Fewanalyses of total waterbird densities on South African estuariesexist (Hockey and Turpie, 1999), with many studies focussing onlyon invertebrate-feeding waders (Martin and Baird, 1987; Hockeyet al., 1992). Large permanently-open systems generally havehigher waterbird densities than the intermittently-open systemsstudied here. Martin and Baird (1987) recorded 11.5 birds/ha onSwartkops estuary, while Smit (1981) found densities of 8.5 birds/ha on the Wadden Sea in Europe. At Langebaan Lagoon high den-sities (15.9 birds/ha) were recorded in summer of which the largemajority (14.7 birds/ha) were waders, with densities dropping to4.5 birds/ha in winter (Underhill, 1987), due to the departure ofmigrating waders.

Martin et al. (2000) recorded 2.7 birds/ha at Knysna estuary, alarge estuarine bay, but attribute this relatively low abundance tohigh levels of recreational disturbance and low availability ofmacrobenthic invertebrates. The high bird densities in permanentlyopen systems are largely due to high numbers of invertebrate-feeding waders that use these systems, especially in the summermonths (Martin and Baird, 1987; Underhill, 1987; Martin et al.,2000). Invertebrate-feeding wader numbers did not increase

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e1918

significantly during open mouth phases at the intermittently openestuaries in this study. However, the estuarymouths were all closedduring most of the summer months and were only open in Octoberand November and therefore unlikely to attract many summermigrants. Moreover, resident intertidal-feeding waders were notattracted by open mouth conditions and seem to prefer the morestable and predictable environment of permanently open estuariesand shoreline habitats. Small, intermittently open estuaries there-fore probably do not play an important role for intertidal-feedingwaders.

The avifauna of all four systems was dominated by piscivorousresident (non-migrating) species. Despite the low densities of birdsrecorded in this study, compared to permanently open estuariesmentioned above, IOEs collectively provide a large area of feeding,roosting and breeding habitat for piscivorous birds in South Africa.The degree to which they rely on these habitats requires furtherinvestigation.

5. Conclusions

Habitat structure plays the most important role in determiningbird community composition in intermittently-open estuaries andchanges in mouth state influence avian community compositionand diversity. Breaching events attract waterbirds and providefeeding opportunities for a short period.

While diversity increased when the mouth of the studied es-tuaries opened, species composition also changed. An increase inmouth-opening events could therefore lead to the distinct inter-mittently open estuarine avifauna being replaced by communitiesresembling those of permanently open estuaries.

Submerged macrophyte cover can influence the abundance ofcertain feeding guilds. However due to the large variety of factorsthat can influence bird abundance in a small system with low birdnumbers it is difficult to make predictions based on one factor suchas submerged macrophyte cover. An intense sampling effort over along period of time would be needed to make more accurate pre-dictions needed for management decisions.

Acknowledgements

This study was funded by the Water Research Commission, theSouth African Institute for Aquatic Biodiversity, and the NRF Centreof Excellence Percy FitzPatrick Institute, University of Cape Town.We are grateful to Tarryn Riddin for provision of macrophyte data,and to Barry Clark, Alan Whitfield and two anonymous referees fortheir inputs.

References

Badenhorst, P., 1988. Report on the Dynamics of the Kleinemonde West and EastEstuaries (CSE 13 and 14). CSIR Report EMA-T 8805.

Brock, M.A., Smith, R.G.B., Jarman, P.J., 1999. Drain it, dam it: alteration of waterregime in shallow wetlands on the New England Tableland of New South Wales,Australia. Wetlands Ecology and Management 7, 37e46.

Calbrade, N., Holt, C., Austin, G., Mellan, H., Hearn, R., Stroud, D., Wotton, S.,Musgrove, A., 2010. Waterbirds in the UK 2008/09-The Wetland Bird Survey.BTO/RSPB/JNCC in Association with WWT, Thetford.

Cooper, A., Wright, I., Mason, T., 1999. Geomorphology and sedimentology. In:Allanson, B.R., Baird, D. (Eds.), Estuaries of South Africa. University Press,Cambridge, pp. 5e25.

Cowley, P.D., Whitfield, A.K., 2002. Biomass and production estimates of a fish-community in a small South African estuary. Journal of Fish Biology 61,74e89.

Cowley, P.D., Wood, A.D., Corroyer, B., Nsubuga, Y., Chalmers, R., 2003. A Survey ofFishery Resource Utilisation on Four Eastern Cape Estuaries (Great Fish, WestKleinemonde, East Kleinemonde and Kowie). Eastern Cape Estuaries Manage-ment Programme e Sustainable Use Sub-Programme Report on behalf of theInstitute for Natural Resources, p. 32.

Doornbos, G., 1987. The Fish Fauna of Lake Grevelingen (SW Netherlands). Ph.D.Thesis. University Amsterdam, Delta Instituut voor Hydrobiologisch Onderzoek,Yerseke.

Gray, J.S., 1997. Marine biodiversity: patterns, threats and conservation needs.Biodiversity and Conservation 6, 153e175.

Harrison, T.D., 2004. Physico-chemical characteristics of South African estuaries inrelation to the zoogeography of the region. Estuarine. Coastal and Shelf Science61, 73e87.

Heÿl, C.W., Currie, M.H., 1985. Variations in the use of the Bot River estuary bywater-birds. Transactions of the Royal Society of South Africa 45, 397e417.

Hockey, P.A.R., Turpie, J.K., 1999. Estuarine birds in South Africa. In: Allanson, B.R.,Baird, D. (Eds.), Estuaries of South Africa. Cambridge University Press, Cam-bridge, pp. 235e268.

Hockey, P.A.R., Navarro, R.A., Kalejta, B., Velasquez, C.R., 1992. The riddle of thesands: why are shorebird densities so high in southern estuaries? The AmericanNaturalist 140, 969e979.

Hughes, D.A., 2008. Appendix B. Specialist report:simulated catchment hydrology.In: van Niekerk, L., Bate, G.C., Whitfield, A.K. (Eds.), East Kleinemonde EstuaryReserve Determination Study: Technical Report. Department of Water Affairsand Forestry, Pretoria.

Kushlan, J.A., 1976. Wading bird predation in a seasonally fluctuating pond. The Auk93, 464e476.

Lambeck, R.H.D., Sandee, A.J.J., de Wolf, L., 1989. Long-term patterns in the waderusage of an intertidal flat in the Oosterschelde (SW Netherlands) and theimpact of the closure of an adjacent estuary. Journal of Applied Ecology 26,419e431.

Lamberth, S.J., Turpie, J.K., 2003. The role of estuaries in South African fisheries:economic importance and management implications. African Journal of MarineScience 25, 131e157.

Martin, A.P., Baird, D., 1987. Seasonal abundance of birds on the Swartkops estuary,Port Elizabeth. Ostrich 58, 122e134.

Martin, A.P., Von Korff, J., Watt, L., 2000. Abundance and distribution of waterbirdson the Knysna estuary. Transactions of the Royal Society of South Africa 55,215e222.

Ntiamoa-Baidu, Y., Nyame, S.K., Nuoh, A.A., 2000. Trends in the use of a smallcoastal lagoon by waterbirds: Muni Lagoon (Ghana). Biodiversity and Conser-vation 9, 527e539.

Perissinotto, R., Walker, D.R., Webb, P., Wooldridge, T.H., Bally, R., 2000. Re-lationships between zoo- and phytoplankton in a warm-temperate, semi-permanently closed estuary, South Africa. Estuarine, Coastal and Shelf Science51, 1e11.

Perissinotto, R., Blair, A., Connell, A., Demetriades, N.T., Forbes, A.T., Harrison, T.D.,Iyer, K., Joubert, M., Kibirige, I., Mundree, S., Simpson, H., Stretch, D., Thomas, C.,Thwala, X., Zietsman, I., 2004. Responses of the biological communities to flowvariation and mouth state in two temporarily open/closed KwaZulu-Natal es-tuaries. In: Adams, J.B. (Ed.), Contributions to Information Requirements for theImplementation of Resource Directed Measures for Estuaries, pp. 56e162. WRCReport No. 1247/2/04.

Perrings, C., Folke, C., Mäler, K.G., 1992. The ecology and economics of biodiversityloss: the research agenda. Ambio 21, 28e33.

Reddering, J.S.V., Rust, I.C., 1990. Historical changes and sedimentary characteristicsof southern African estuaries. South African Journal of Science 86, 425e428.

Riddin, T., 2011. Dynamics of Macrophytes in the East Kleinemonde, a SmallTemporarily Open/closed South African Estuary. Nelson Mandela MetropolitanUniversity, South Africa. PhD thesis.

Riddin, T., Adams, J.B., 2008. Influence of mouth status and water level on themacrophytes in a small temporarily open/closed estuary. Estuarine, Coastal andShelf Science 79, 86e92.

Russell, I.A., Randall, R.M., Randall, B.M., Hanekom, N., 2009. Relationships betweenthe biomass of waterfowl and submerged macrophytes in a South Africanestuarine lake system. Ostrich 80, 35e41.

Slob, G.J., 1989. 15 aar Vogelontwikkelingen in het afgesloten Grevelingenbekken.Staatsbosbeheer, Goes, p. 50.

Smit, C.J., 1981. The importance of the Wadden Sea for estuarine birds. In:Smit, C.J., Wolff, W.J. (Eds.), Birds of the Wadden Sea. A.A. Balkema, Rotterdam,pp. 280e289.

Stewart, B.A., Bally, R., 1985. The ecological role of the Red-knobbed Coot Fulicacristata GMELIN at the Bot River estuary, South Africa: a preliminary investi-gation. Transactions of the Royal Society of South Africa 45, 419e426.

Stone, A.W., 1988. Climate and Weather. In: Lubke, R.A., Gess, F.W., Bruton, M.N.(Eds.), A Field Guide to the Eastern Cape Coast. The Grahamstown Centre of theWildlife Society of Southern Africa, Grahamstown, pp. 19e30.

Teroerde, A., Turpie, J.K., 2012. Use of a small, intermittently-open estuary by wa-terbirds: a case study of the East Kleinemonde estuary, Eastern Cape, SouthAfrica. African Journal of Aquatic Science 37, 183e190.

Turpie, J.K., Adams, J.B., Joubert, A., Harrison, T.D., Colloty, B.M., Maree, R.C.,Whitfield, A.K., Wooldridge, T.H., Lamberth, S.J., Taljaard, S., Van Niekerk, L.,2002. Assessment of the conservation priority status of South African estuariesfor use in management and water allocation. Water SA 28, 191e205.

Underhill, L.G., 1987. Waders (Charadrii) and other waterbirds at Langebaan Lagoon,South Africa, 1975e1986. Ostrich 58, 145e155.

van Niekerk, L., Cowley, P.D., Bornman, T.G., 2008. Appendix C. Specialist report:physical processes. In: van Niekerk, L., Bate, G.C., Whitfield, A.K. (Eds.), EastKleinemonde Estuary Reserve Determination Study: Technical Report. Depart-ment of Water Affairs and Forestry, Pretoria.

A.I. Terörde, J.K. Turpie / Estuarine, Coastal and Shelf Science 125 (2013) 10e19 19

Vorwerk, P.D., 2000. Ichthyofaunal Community Structures in Different Types ofEastern Cape Estuaries. MSc Thesis. Rhodes University, Grahamstown.

Walton, C., 1984. Readers Digest Atlas of Southern Africa. Readers Digest Associationof South Africa, Cape Town.

Whitfield, A.K., 1992. A characterization of southern African estuarine systems.South African Journal of Aquatic Science 18, 89e103.

WRC Report No. 1581/1/07. In: Whitfield, A.K., Bate, G.C. (Eds.), A Review of Infor-mation on Temporarily Open/closed Estuaries in the Warm and Cool TemperateBiogeographic Regions of South Africa, with Particular Emphasis on the Influ-ence of River Flow on These Systems.

Whitfield, A.K., Bate, G.C., Adams, J.B., Cowley, P.D., Froneman, P.W., Gama, P.T.,Strydom, N.A., Taljaard, S., Theron, A.K., Turpie, J.K., Van Niekerk, L.,Wooldridge, T.H., 2012. A review of the ecology and management of tempo-rarily open/closed estuaries in South Africa, with particular emphasis on riverflow and mouth state as primary drivers of these systems. African Journal ofMarine Science 34 (2).

Whitfield, A.K., Blaber, S.J.M., 1979. Feeding ecology of piscivorous birds at Lake StLucia, Part 3: swimming birds. Ostrich 50, 10e20.

Whitfield, A.K., Cyrus, D.P., 1978. Feeding succession and zonation of aquatic birds atFalse Bay, Lake St. Lucia. Ostrich 49, 8e15.