Biological Conservation 167 (2013) 57–68

Contents lists available at SciVerse ScienceDirect

Biological Conservation

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

Alteration of wetland hydrology in coastal lagoons: Implicationsfor shorebird conservation and wetland restoration at a Ramsar sitein Sri Lanka

0006-3207/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.biocon.2013.07.013

⇑ Corresponding author. Present address: ILWS, Institute for Land Water andSociety, Charles Sturt University, PO BOX 789, Albury, NSW 2640, Australia. Tel.:+61 02 60519914.

E-mail address: mbellio@csu.edu.au (M. Bellio).

M. Bellio ⇑, R.T. KingsfordAustralian Wetlands, Rivers and Landscapes Centre, School of Biological, Earth and Environmental Sciences, University of New South Wales, NSW 2052, Australia

a r t i c l e i n f o a b s t r a c t

Article history:Received 10 January 2013Received in revised form 1 July 2013Accepted 10 July 2013

Keywords:Wetlands restorationWetlands hydrologyShorebirds conservationRamsarBundalaNational ParkSri Lanka

Many of the world’s lagoons and estuaries, representing the most important habitat for shorebirds, areincreasingly degraded, often associated with dramatic declines in shorebirds, particularly in Asia. Weinvestigated effects of hydrology on shorebird communities of two coastal lagoons of a Ramsar site insouth-east Sri Lanka. One of the lagoons (Bundala) experienced natural water level fluctuations, whilethe other (Embilikala) had high stable water levels (>10 cm), maintained by inflows from irrigated agri-culture. These inflows dramatically affected composition and abundance of shorebird communities andtheir prey. Tactile and visual foraging shorebirds were consistently more abundant on Bundala thanEmbilikala lagoon. Their feeding efficiency was significantly (50%) higher on Bundala, where large(500 lm) benthic prey were more abundant than on Embilikala. Contrastingly, small (60 lm) planktonicprey dominated the hydrologically stable Embilikala lagoon where pelagic foraging shorebird speciesdominated the shorebird community. On both lagoons, visual and tactile species foraged predominantlyin shallow water (1–10 cm). Pelagic shorebirds also foraged in water depths >10 cm but 20% less effi-ciently than in shallower water. Effective shorebird conservation depends on reducing anthropogenicimpacts that detrimentally affect functional processes and habitat value. The detrimental effect of alteredhydrology on Embilikala lagoon could be reversed by diverting drainage water. Re-establishing a naturalhydroperiod, would increase the productivity of the lagoon for foraging shorebirds. Restoration of shore-bird habitat on this Ramsar site could also improve other ecosystem services such as fisheries and tour-ism. Changes to the hydrology of coastal lagoons elsewhere may be similarly affected.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Coastal and marine ecosystems are among the most productive,yet threatened, ecosystems in the world (Halpern et al., 2009), withimpaired natural functions (Agardy and Alder, 2005), because thisis where 60% of global human populations live (UNESCO, 1993).Such coastal areas are particularly affected in Asia with its bur-geoning population growth, unsustainable resource use, escalatingagriculture and urban development. Unfortunately, understandinginteractions with the ecology of wetlands in developing countrieslags well behind well resourced parts of the world (Galbraithet al., 2005).

Many anthropogenic impacts on coastal ecosystems originateupstream (Pringle, 2000; Fabricius, 2005). Agricultural pollutantsdrained into coastal waters cause eutrophication, hypoxia or

anoxia (Rabalais et al., 2002). Damming of rivers, levees, channelsand diversion of water have altered hydrology, increased sedimen-tation rate, salinity and erosion, detrimentally affecting functionsof many coastal estuaries (Dudgeon, 2000; Gillanders and Kings-ford, 2002).

Much of this understanding of coastal impacts relates to effectson hydrodynamic, physical and biological characteristics of estuar-ies, rather than coastal lagoons, common along continental coastsof most continents and occupying about 13% of the world’s coast-line (Boynton et al., 1996). Morphological and functional differ-ences between estuaries and coastal lagoons mean impacts alsomay differ (Kjerfve and Magill, 1989). Nutrient, sediment andfreshwater input from agriculture may be particularly detrimentalto coastal lagoons, with limited tidal exchange, compared to estu-aries (Kjerfve and Magill, 1989; Boynton et al., 1996).

Many coastal lagoon systems provide critical habitat for migra-tory shorebirds (Ferreira et al., 2005; Alfaro and Clara, 2007),whose global populations have declined dramatically with coastaldegradation (Stroud et al., 2006), particularly in Asia (Delany et al.,2010). Some shorebird populations (e.g. Curlew sandpiper (Calidris

58 M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68

ferruginea)) of the Central Asian Flyway have declined by morethan 70% at major wintering grounds along the east coast of India(Point Calimere, Balachandra, 2006) and in Sri Lanka (Li and Mund-kur, 2004), the southernmost point of the flyway. Causes and driv-ers for decline in the flyway remain unclear (Balachandra, 2006)mainly because little is known of shorebird ecology, even thoughit has the greatest number of globally threatened waterbirds(Mundkur, 2006).

Long-term resilience of wetlands and the biota they supportaround the world is seriously compromised by water resourcedevelopment (Kingsford et al., 2006; Vörösmarty et al., 2010).The key international agreement meant to mitigate such impactsis the Ramsar Convention with the listing of wetlands of interna-tional importance around the world (www.ramsar.org). A keyobjective of the Ramsar convention is to avoid change of ecologicalcharacter in a wetland due to anthropogenic impacts (Ramsar,1993; recommendation 5.2 article 3; Finlayson et al., 2011). Eco-logical character is defined as: ‘‘the combination of the ecosystemcomponents, processes and benefits/services that characterise thewetland at a given point in time’’ (Ramsar Bureau 2002, ResolutionX.15) where ecosystem benefits are ‘‘the benefits that people re-ceive from ecosystems’’ (Ramsar Bureau 2002, Resolution IX.1. An-nex A, n 11) including wildlife resources; fisheries; forageresources; agricultural resources and water supply. Maintainingthe ecological character of a wetland should therefore translatein maintaining a healthy wetland that supports a rich biodiversityand people livelihoods and is not affected by anthropogenic im-pacts. Measuring impacts of people on wetlands is often difficult,given paucity of data and natural stochastic variability of theseecosystems. Pressure for water resource development is particu-larly acute in Asia (Dudgeon, 2000; Li and Ounsted, 2007) and safe-guarding ecosystems services for people and biodiversity, throughsustainable development and conservation of wetlands is achallenge.

We investigated the relationship between shorebird ecologyand hydrology on two brackish water (0.05–3% ppt) coastal la-goons that form a Ramsar site in south-east Sri Lanka. The twocoastal lagoons have different flow regimes: altered (from irrigatedagriculture) (Embilikala Lagoon) and natural (Bundala Lagoon)(Smakhtin and Piyankarage, 2003). There has been an alarming de-crease in the abundance of shorebirds and other waterbird speciessuch as the flamingos (Phoenicopterus ruber), an iconic touristicattraction, at the Ramsar site over the last decade (Perera, 2008;Bellio et al., 2009).

We tested for differences in shorebird abundance, communitycomposition, prey abundance and foraging efficiency to cycles offlooding and drying patterns of the two lagoons. This has implica-tions for rehabilitation of the lagoon with altered flow regimes.Habitat restoration and rehabilitation are considered to play a piv-otal role in recovering degraded or destroyed ecosystems (Finlay-son et al., 2001; Roni et al., 2005).

2. Methods

2.1. Study site

Four coastal lagoons (Embilikala 430 ha, Bundala 520 ha, Malala650 ha, Koholankala 390 ha) form part of the wetland system ofBundala National Park, on the south-east coast of Sri Lanka(Fig. 1). Under the Ramsar criteria, wetlands should be selectedfor the Ramsar List on account of their international significancein terms of the biodiversity and uniqueness of their ecology, bot-any, zoology, limnology or hydrology. The Ramsar status wasgranted to Bundala National Park because of the outstanding eco-logical significance of its coastal lagoons (Bambaradeniya, 2001).

The lagoons system supports 324 species of vertebrates which in-clude 32 species of fish, 15 species of amphibians, 48 species ofreptiles, 197 species of birds and 32 species of mammals and 52species of butterflies are among the invertebrates. Shorebird spe-cies include 10 residents and 34 migratory palaearctics (includingseven vagrants) which spend the Northern Hemisphere winter(non-breeding season) on the Bundala wetlands in internationallysignificant numbers (>21,000, Bambaradeniya, 2001; Bellio et al.,2009). In particular curlew sandpipers and black-tailed godwits,regularly winter at the Ramsar site of Bundala, are among the longdistant migrant populations suffering dramatic declines along theCentral Asian flyway.

The climate is hot and dry (Cooray, 1984), with a mean annualtemperature of 27.1 �C and an average annual precipitation of1074 mm. There is one dry season (July–September) and two wetseasons: the north east monsoon (October–March) when rainfallis highest in November (IIMI, 1995) and the south west monsoon(April–June). These seasons produce two rice (Oryza sativa) cultiva-tion seasons (called Maha and Yala seasons respectively).

The lagoons’ natural hydrology depended on seasonal rainfall,inflow and seepage of seawater through the sand bars of two nar-row inlets (Fig. 1), connecting Bundala and Malala coastal lagoonsto the Indian Ocean (CEA and Euroconsult, 1993). Flow regimeshave changed in Embilikala since 1989, where freshwater drainageflows in from an upstream irrigation scheme (Kirindi Oya IrrigationSettlement) (drainage canal, Fig. 1) (Jayawardena, 1993; Smakhtinand Piyankarage, 2003). This has decreased salinity (conductivityin Bundala 9732–669 msm�1, Embilikala 973–63.4 msm�1; Bellioet al., 2009) and increased turbidity (CEA and Euroconsult, 1993;Amarasinghe et al., 2001; Bellio et al., 2009). Water levels in theimpacted lagoon have also increased after irrigation development(water level fluctuations range before (0–1.5 m) and after (1–2.2 m)) (Jayawardena, 1993; Smakhtin and Piyankarage, 2003),preventing the lagoon drying out after the monsoon. Contrastingly,Bundala lagoon retains natural drying (post monsoon season) andfilling (monsoon seasons) sequences, reflecting rainfall. We re-corded water level (cm) on both lagoons at least every two weeksfrom 2006 to 2007.

2.2. Shorebird habitat

Most shorebirds feed and roost within the shallows of a wetland(water depth <30 cm), where prey availability and accessibility ismediated by prey behaviour (Zwarts and Wanink, 1993) and spe-cies’ specific bill and neck morphology, leg length and body size(Zwarts and Wanink, 1991). We categorised all habitats used byshorebirds related to water depth: dry margins (D) (0–1 cm), shal-low water (S) (1–10 cm) and high depth (H) (11–30 cm). Thesewere further subdivided on the basis of vegetation/algal coverwhich were generally more developed on Embilikala lagoon thanon Bundala lagoon (Bellio et al., 2009).

We identified five mutually exclusive habitat types on Embili-kala lagoon: two on dry margins (dried grass (DG) and bare soil(DS)); two in shallow (1–10 cm) water (sandy bare soil (SS) andwater with vegetation (SV)) and one in high (11–30 cm) waterdepth with aquatic vegetation (HV), including submerged andemerged plant species. Dominant submerged plant species in-cluded water thyme (Hydrilla spp.), pondweed (Najas sp.), horn-wort (Ceratophyllum demersum), duck lettuce (Ottelia alsimoides)and emergent species included water lettuce (Pistia stratiotes), cat-tails (Typha spp.), the floating fern salvinia (Salvinia molesta),sedges (Cyperus spp.), and water couch (Paspalum spp.).

Bundala lagoon had submerged macroscopic algae (Nitella ssp.)and microscopic blue green algae (Microsystis, Nostoc and Oscillato-ria sp.), creating thick algal mats when water levels were shallow(1–10 cm) but no aquatic vascular plants (CEA and Euroconsult,

M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68 59

1993). There were also areas of bare sandy soil. We identified fivemutually exclusive habitat types on Bundala lagoon: bare soil (DS)along the dry margins of the lagoon; two in shallow (1–10 cm)water (sandy bare soil (SS) and water with vegetation (algae)(SV)) and two in high (11–30 cm) water depth with submergedmacroscopic algae (HV) and sandy bare soil (HS).

2.3. Invertebrates

We sampled macro (>500 lm) and meio (P60 < 500 lm) inver-tebrate fauna, the main prey of shorebirds (del Hoyo et al., 1996),in the two lagoons (February–April, August–September 2007).Most invertebrates, show positive phototropism to light (Heise,1992; Nikolaev, 2007) and so we used underwater light traps, be-cause other methods (e.g. grab-sampling, core sampling) wereinvasive and not permitted. Light traps consisted of 750 ml plasticwater bottles, cut at the top, and reversed (opening of bottle2.5 cm) with a chemiluminiscent stick inside (Schilling et al.,2009). Traps were submerged, set flush to the sediment, in waterat sunset, left overnight and collected the next day at midday, after18 h. We used three different light wavelengths (blue, yellow/green, orange), which could differentially affect behaviour of aqua-tic invertebrate taxa and life cycles (Heise, 1992). Light traps alsocaptured fish and amphibians (Appendix A). We used twelve lighttraps: four replicates for each wavelength on each lagoon. Eachhabitat type had equal numbers of light traps: four for each ofthe three flooded habitat types on Embilikala (12 total) and sixfor the two habitat types on Bundala (12 total). Some samples were

Fig. 1. Study area in south-east Sri Lanka, showing Malala, Embilikala (A–D indicate foulagoons and salt evaporation ponds (Salt ponds) of Bundala National Park (Ramsar siEmbilikala lagoon.

removed by fishermen, but a total of 55 samples on Bundala La-goon and 53 on Embilikala Lagoon remained (Appendix A). Aftercollection, we filtered and washed samples under running water,using two sieves (60 and 500 lm), stacked on top of each other,which separated the fauna into respective size classes. Collectedanimals were retained separately and preserved (70% alcohol,100 ml) for subsequent identification. No published informationexisted on the aquatic invertebrate fauna of the two lagoons andso we counted and identified all individuals retained by the500 lm sieve, categorised into eight broad taxonomic groups: in-sect larvae, adult insects, benthic crustaceans, planktonic crusta-ceans, water mites, molluscs, fish and amphibians (Appendix A).Some of these were further identified to family and genus (Appen-dix A). For 60 lm samples, we collected three subsamples of 1 mlusing a pipette after mixing to ensure animals were evenly distrib-uted in solution (Boros et al., 2006) and counted individuals with aSedgwick rafter cell. We calculated the average of the three subs-amples of 1 ml and then multipled that count (�100 ml total vol-ume in each sample).

2.4. Shorebirds

We surveyed shorebirds in 2006 (February–April, August–Sep-tember, November) and 2007 (February–April, August–September)at Bundala and Embilikala lagoons (11 months over two years).This coincided with arrival of migratory shorebirds in August andtheir departure in April, seasons (wet and dry months) and rice cul-tivation. Rice cultivation started late October to early November in

r different survey sections), Bundala (A–E indicate seven different survey sections)te) and the canal draining freshwater from irrigated agricultural areas (rice) into

Wet season Dry season

Shor

ebird

s ab

unda

nce

Wat

er le

vel

Wet season Dry season

Wat

er le

vel

Shor

ebird

s ab

unda

nce

(a)

(b)

Fig. 2. Total mean monthly shorebird abundance in 2006 (filled) and 2007 (open)and water levels in 2006 (continuous line) and 2007 (dashed line) on (a) Bundala(mean abundance ± SE) and (b) Embilikala lagoons.

60 M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68

2006 but at the end of September in 2007. We also surveyed twicebetween rice cultivation seasons, when irrigation water flowedinto Embilikala lagoon: at the beginning (BMar) and end of March(EMar) each year.

Three shorebird counts (of all thirty species using both lagoons)were done over three consecutive days to estimate temporal vari-ation in shorebird abundance (Weber and Haig, 1996; Masero andPérez-Hurtado, 2001). Shorebirds were identified and countedindividually, using a telescope (60�), with large flocks (>100) esti-mated in groups of 10, 20, 50, 100 birds (Rapold et al., 1985). Totalcounts were possible for Bundala lagoon but not Embilikala lagoonbecause of access restrictions to two sections (C and D; Fig. 1) butpermission was granted to do one monthly total count (A–D). Thetwo frequently counted sections (A and B; Fig. 1) consistently hadmore shorebirds (78 ± 12% S.D. of total), when we compared countsof sections A and B to total counts of Embilikala lagoon.

We also investigated habitat associations of shorebirds (for allthirty species) in the two lagoons. In section A of Embilikala lagoon(Fig. 1), we marked 18 static sampling areas (50 � 120 m) coveringall habitat types, parallel to the shoreline in two rows of 9 samplingareas each. These were small enough to include most of the eightmore abundant species and large enough to avoid sampling thesame individual twice. We recorded the number and species ofshorebirds in sampling areas each survey, in the five habitat typesfor which we estimated percentage cover. We divided the entireperimeter of Bundala lagoon into seven sampling areas, incorporat-ing the five habitat types (A–G, Fig. 1): bare sandy soil (B1, A, F) andareas with submerged macroscopic algae or algal mats (B, C, D, E,G) (Fig. 1). We tracked changes in habitats in these sections withwater level and drying patterns. We derived the shoreline lengthof each survey section, relative to a full (no shoreline exposed)perimeter margin, from a remotely sensed image (March 2006 In-dian Remote Sensing Image). We recorded which section had ex-posed shallow shoreline (1–10 cm) during each survey andderived the percentage of total lagoon exposed shoreline. Werecorded number and species of shorebirds within each surveysection at each water depth shallow (1–10 cm) (S) and high (11–30 cm) (H)), at every survey. Water depth was estimated usingleg lengths of shorebird species (Higgins and Davies, 1996).

We also recorded diurnal foraging behaviour of the eight mostcommon (present during all surveys) shorebirds species, on thetwo lagoons. Data were collected early (sunrise to 11 am) or late(4 pm to sunset), avoiding the hottest part of the day when birdspredominantly rested. We collected data from different parts ofthe same lagoon to avoid pseudo-replication. Shorebirds specieswere separated into three foraging guilds (Piersma and Ntiamoa-Baidu, 1995) on the basis of sensory food detection and feedingstyle: visual surface foraging shorebirds, plovers (Charadriidae);tactile surface foraging shorebirds, including sandpipers (Scolopac-idae); pelagic foraging shorebirds, including marsh sandpipers(Scolopacidae) and stilts (Recurvirostridae). Tactile feeders(black-tailed godwit (Limosa limosa), curlew sandpiper (Calidris fer-ruginea), little stint (Calidris minuta)) inserted their bills into thesubstrate (sand or algal mats) while visual feeders (Kentish plover(Charadrius alexandrinus), lesser sand plover (Charadrius mongolus))moved quickly or slowly (common redshank (Tringa totanus))pecking prey from the water surface, wet algal mats or areas ofdry soil. Pelagic feeders black-winged stilt (Himantopus himanto-pus), marsh sandpiper (Tringa stagnatilis) moved their bills fromside to side or ran with their bill submerged chasing prey or rapidlypecking prey on and below the water surface.

Birds clumped in multispecies flocks along the margins of thelagoons within sampling areas. We used focal sampling (observingone individual for a specified amount of time) over 1-min intervals(using a 60� telescope) to record foraging behaviour of the eightmost common species (Martin and Bateson, 1986). The eight most

common species were: black-tailed godwit, curlew sandpiper, littlestint, black-winged stilt, marsh sandpiper, lesser sand plover, Ken-tish plover, common greenshank. If a bird flew before the end of aminute, the datum was discarded. We selected individuals withinthe flock by constantly shifting the telescope position in front ofthe flock and focusing on only one individual to avoid dependenceamong individuals. We recorded each peck (converted to a peckrate, number of pecks in a minute) when the bird’s bill enteredthe water or substrate or when multiple probes were made, beforethe bill exited (e.g. ‘‘stitching’’ by black-tailed godwit and curlewsandpiper). Pecks were defined as successful (‘‘success rate’’) ifone of four criteria was met: captured prey item was swallowed(usually black-tailed godwits); throat (gular region) movement oc-curred, consistent with swallowing (plovers); ‘‘head-throwing’’indicated swallowing (inertial feeding in curlew sandpiper andblack-winged stilt, see Gans, 1961) or; mandibular spreading (i.e.,surface tension feeding by black-winged stilts and marsh sand-piper; see Rubega, 1997). We could not detect swallowing in littlestints and so only recorded ‘‘peck rate’’. For each focal individual,we also recorded habitat type and feeding depth using leg lengths(Higgins and Davies, 1996).

We also compared behavioural activities of flocks at the two la-goons. To determine flock behaviour, we instantaneously scannedbehaviour of each individual (Altmann, 1974). Each scan was sep-arated in time (30–60 min), to avoid dependency (Martin and Bate-son, 1986). We recorded four mutually exclusive behaviours:feeding, preening, resting and walking. Feeding birds actively for-aged by gleaning, pecking or manipulating food. During preening,

Table 1Percentage of shallow shoreline relative to total lagoon perimeter recorded onBundala and Embilikala lagoons, during wet and dry seasons in 2006 and 2007.

Season Month Bundala Embilikala

2006 2007 2006 2007

Wet February 29 73 29 75March (early) 29 73 66 51March (late) 63 100 56 57April 73 73 68 38November 14 – 18 –

Dry August 100 100 83 75September 100 100 54 75

M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68 61

individuals re-arranged feathers with their bills or bathed. Duringresting, the head or bill was tucked in their feathers (godwits, com-mon redshanks, large sandpipers). Walking birds moved within theflock but did not feed. Birds within the flock doing each activitywere counted and recorded.

2.5. Data analyses

We tested for differences in monthly water levels between 2006and 2007. We also compared the variability of seasonal (wet anddry) water levels fluctuations between the lagoons, using the coef-ficient of variation (CV). We used simple regression to separatelyinvestigate relationships between shorebird and invertebrateabundances and water levels for each lagoon.

Shorebirds abundance data (total counts) were first trans-formed into densities (number of birds ha�1) to allow comparisonbetween the two lagoons of different size. Invertebrate abundancewas calculated as number of individual caught per trap within18 h. We compared differences in shorebird densities and inverte-brate abundance between the lagoons. All comparisons were doneusing the Kruskal Wallis test as data were not normal andvariances could not be stabilised by transformation (SPSS, 2009).We used the single monthly total shorebird counts for Embilikalalagoon. For Bundala lagoon, there were no significant differencesin shorebird abundance among daily counts (H = 0.14 p = 0.93),

Water level (cm)

Num

ber o

f sho

rebi

rds

Num

ber o

f sho

rebi

rds

Water level (cm)

(b)

(a)

Fig. 3. Significant negative relationships (R2, p value) between water level and shor(individuals h�1) over five months on (a) Bundala lagoon and (b) Embilikala lagoon.

but significant differences among months (H = 48.01 p < 0.001)and so we averaged daily counts for each month of survey forBundala lagoon.

We compared (Kruskal Wallis test) pecks and ‘‘prey capturerate’’ (success rate) and feeding efficiency (pecks rate/success rate)(Shepherd and Boates, 1999) for the eight most common shore-birds species and compared the percentage of time that individualsdevoted to different activities (feeding, preening, resting, walking)between the lagoons. Data on behavioural activities of flocks (feed-ing, preening, resting, walking were analysed separately for eachseason (wet and dry). For each period (wet and dry), we calculatedthe average number of birds engaged in this activity, among allscans. This average was the separate percentage of foraging andresting birds for each scan (Appendix B).

For the invertebrates samples, we first determined our samplesize adequacy (108 samples) using a conservative power analysisof 80% to detect reductions of 25% in prey abundance (Mathotet al., 2007). We also tested (Kruskal–Wallis test) whether therewere differences in invertebrate abundances among different lightcoloured traps for each taxon. There were no differences (p > 0.05)in invertebrate abundance among different light coloured traps foreach taxon on Embilikala lagoon while on Bundala lagoon, exceptfor blue light traps which caught more molluscs than yellow or or-ange light traps (Hmolluscs = 9.34 p = 0.01, post hoc p < 0.02). There-fore, invertebrate abundance (expressed as number of preycaptured per trap per hour) was averaged across different colourlight traps for each taxon except molluscs, which were separatelytested for each light colour (blue–yellow and orange).

We investigated the relationships between shorebirds andinvertebrate abundance and habitat type using principal compo-nent (PCA) ordination techniques, PRIMER 6.1 (Clarke and War-wick, 2001). Data on invertebrate abundance were square roottransformed to meet PCA assumptions and standardised (Clarkeand Warwick, 2001), removing differences in total abundancewithin samples when differences could be attributable to uncon-trollable sampling factors (Clarke and Warwick, 2001). We sepa-rately calculated similarities for the invertebrate and shorebirdcommunities of the two lagoons using the Bray–Curtis association

Water level (cm)

Num

ber o

f inv

erte

brat

es

Num

ber o

f inv

erte

brat

es

Water level (cm)

ebird abundance (log transformed) over 13 months and invertebrate abundance

Fig. 4. (a) Cluster analysis of similarities (Bray–Curtis) of invertebrate communities,amphibians and fish (individuals h�1), standardised and square-root transformed)between Bundala (BUN) and Embilikala (EMB) lagoons in wet . and dry seasons s

and (b) principal component ordination of seven taxa, discriminating the seasonal(wet . and dry season s) differences in invertebrate, amphibians and fishcommunities captured using light traps, between Bundala and Embilikala lagoons.

Table 3Eigenvalues of the first two principal components (PC1–PC2) and contribution of fiveinvertebrates (insect, insect larvae, plankton, molluscs, benthic crustaceans) and twovertebrate taxa (fish and amphibians) to the ordination.

Variable PC1 PC2

Insect �0.436 �0.819Insect larvae �0.232 0.380Benthic Crustaceans 0.677 �0.255Molluscs 0.376 �0.050Fish �0.052 �0.249Amphibians �0.009 0.045Plankton �0.393 0.229Eigenvalue 5.44 0.849% Variation 84.6 13.2Cumulative % variation 84.6 97.8

62 M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68

measure to form similarity matrices. We then examined differ-ences between the communities on the lagoons, using hierarchicalagglomerative clustering with group-average linking and non-metric multi-dimensional scaling (MDS) (Clarke and Warwick,2001). We accepted stress values 60.1, corresponding to a goodordination with no prospect of misleading interpretation (Peppinget al., 1999).

3. Results

Annual patterns of flooding and drying were similar within la-goons, contrasting differences between lagoons (Fig. 2a and b).

Table 2Mean ± SE abundance of aquatic invertebrates, amphibians and fish processed (individuals/Embilikala lagoons, during dry and wet seasons in 2007.

Mesh size Taxa Bundala

Wet (31)

500 lm Insect 426 ± 172Insect larvaea 29 ± 23Benthic Crustaceansa 125 ± 39Planktonic Crustaceansa 38 ± 21Molluscsa 188 ± 82Fish 19 ± 5Amphibians 0

60 lm Insect larvaea 0.5 ± 0.4Crustaceansa 22 ± 8Planktonic crustaceansa 254 ± 62

a Taxa for which significant difference (p < 0.05) existed in invertebrate abundance be

Bundala lagoon’s water levels progressively declined after thewet season in both years, following a drying pattern (Fig. 2a), withsome variation. Wet and dry season levels were respectively 50 cm(v0.05,1 = 3.84; Hwet = 14.73 p < 0.001) and 10 cm (Hdry = 12.75p < 0.001) lower in 2007 than 2006 (Fig. 2a). Contrastingly, Embili-kala lagoon’s water levels were reasonably stable over the year(CVEmbilikala = 9.9%, CVBundala19.2%, Fig. 2a and b) and not differentbetween years (H = 0.82 p = 0.36, Fig. 2b).

These water level fluctuations were reflected in habitat avail-ability. As Bundala lagoon dried, shallow shoreline became avail-able, increasing as water levels decreased (y = �54.50x + 6515,r = �0.96, p < 0.001, n = 11), (Table 1). Bundala lagoon’s inlet wasopened in January 2007 after heavy rains made the Bundala road(Fig. 1) impassable. Thus, drier conditions in January 2007 pro-duced more (30%) shallow shoreline than in 2006, during the wetseason months (H = 4.05 p = 0.04, Table 1). In both years, all sevensectors had shallow water at the end of the dry season (Table 1).

Similarly, the amount of shallow shoreline increased as waterlevels decreased on Embilikala lagoon (y = �89x + 18197,r = �0.76, p < 0.01, n = 11) although not as much as on Bundala la-goon (Table 1). There was relatively less flooded shallow shorelineon Embilikala lagoon compared to Bundala lagoon (Table 1), aswater draining from irrigated areas maintained high water levelson Embilikala (Fig. 2b). Even though water levels were artificiallylowered in January 2007 on Embilikala lagoon by opening the Mal-ala inlet, water levels increased by March, flooding shallow shore-line(Fig. 2b) as water drained from the agricultural area.

3.1. Invertebrates

Water levels also influenced the abundance and spatialdistribution of invertebrates. As water levels increased, abundanceof aquatic invertebrates decreased on both lagoons

trap, sample sizes in parentheses) for two mesh sizes (500 lm, 60 lm) on Bundala and

Embilikala

Dry (24) Wet (25) Dry (28)

10 ± 3 273 ± 128 9 ± 20.2 ± 0.1 1 ± 0.3 10 ± 2548 ± 125 3 ± 2 46 ± 377 ± 4 12 ± 3 31 ± 12773 ± 163 11 ± 3 12 ± 34 ± 1 2 ± 0.8 1 ± 0.60 0 0.08 ± 0.050.1 ± 0.1 4 ± 1 3 ± 1658 ± 257 0 0133 ± 44 122 ± 25 386 ± 108

tween lagoons.

M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68 63

(yBundala = �42.81x + 1996.5, r = �0.91 n = 5 p < 0.05; yEmbilikala =�16.47x + 2822.9, r = �0.88 n = 5 p < 0.05, Fig. 3a and b). Therewas also a clear seasonal pattern in invertebrate abundance andcommunity composition between the two lagoons, related towater depth (Fig. 4a and b). During the wet season, insects (pre-dominantly Corixidae) were the most dominant taxon, occurringin similar abundance (H = 0.17 p = 0.89, Table 2, Fig. 4a and b, Ta-ble 3), making up 31% and 16% of all invertebrates respectivelyon Embilikala and Bundala lagoons (Appendix A) (Table 2). Duringthe dry season, crustaceans and molluscs were the most abundanttaxon, but the type and size of invertebrates differed between thelagoons (Fig. 4a and b, first axis PCA 85% variance explained, Ta-ble 3). Benthic crustaceans (mainly amphipods) and molluscs(>500 lm) were in higher numbers on Bundala lagoon (AppendixA; 67% of total), compared to Embilikala lagoon (Hcrustaceans = 73.99p < 0.001, Hmolluscs = 46.81 p < 0.001, Fig. 4b, Tables 2 and 3) whileplanktonic crustaceans (mainly copepods, ostracods and cladocer-ans) (>60 < 500 lm) were more abundant on Embilikala lagoon(Appendix A; 64% of total), compared to Bundala lagoon(H = 48.51 p < 0.001, Fig. 4b, Tables 2 and 3). These seasonal differ-ences reflected habitat availability on the two lagoons (Fig. 5a andb, Table 4).

Fig. 5. Principal component ordination of five invertebrates and two vertebrate taxa (fishforaging guilds (right), with discrimination of habitat type relative to water depth on (adepth (1–10 cm) areas with sandy soil and no vegetation (SS) h or with vegetation/ algae(HS) d or with vegetation/ algae (HV) j. Embilikala lagoon did not have areas of mediu

On Bundala lagoon, water depth (first axis 48.6% variance ex-plained) and vegetation cover (second axis 28.5% variance ex-plained) explained most variation in invertebrate communities(Fig. 5a, Table 4). Molluscs and benthic crustaceans were abundantin shallow water habitats while insects and plankton were numer-ous in medium water depths (Fig. 5a and b, Table 4). Molluscs asso-ciated with shallow areas with no vegetation cover while benthiccrustaceans occurred in shallow sandy areas or habitats with algalmats. Contrastingly, planktonic crustaceans occurred where therewere well developed submerged aquatic plants and algal mats,while insects associated with vegetated areas (Fig. 5a, Table 4).Similarly on Embilikala lagoon, planktonic crustaceans were moreabundant in medium water depths (first axis 54% of total vari-ance); insects predominantly occurred in shallow vegetated areas(Fig. 5b, Table 4) and; molluscs and benthic crustaceans were inshallow sandy areas (second axis 24% of total variance, Fig. 5b,Table 4).

3.2. Shorebirds

Shorebird communities partitioned in response to hydrologicaldifferences, habitat availability and invertebrate communities

and amphibians) communities (left) and abundance (birds ha�1) of three shorebird) Bundala lagoon and (b) Embilikala lagoon. Habitat types included: shallow water

SV s; and medium water depth (11–30 cm) areas of sandy soil with no vegetationm water depth with no vegetation.

Table 4Eigenvalues of the first two principal components (PC1–PC2) and contribution of fiveinvertebrates (insect, insect larvae, plankton, molluscs, benthic crustaceans) and onevertebrate taxa (fish) and three shorebird foraging guilds (tactile, visual, pelagic) tothe ordination and discrimination of habitat type and water depth on Bundala andEmbilikala.

Variable PC1 PC2

Bundala habitatInsect 0.480 0.452Insect larvae 0.102 0.057Benthic Crustaceans �0.257 �0.719Molluscs �0.602 0.373Fish �0.019 0.199Plankton 0.574 �0.311Eigenvalue 18.7 11% Variation 48.6 28.5Cumulative % variation 48.6 77.1

Bundala water depthVisual �0.318 0.751Tactile �0.861 0.028Pelagic �0.398 �0.660Eigenvalue 8.91 0.334% Variation 96.1 3.6Cumulative % variation 96.1 99.7

Embilikala habitatInsect �0.656 0.483Insect larvae 0.079 �0.309Benthic Crustaceans 0.025 �0.032Molluscs �0.187 �0.766Fish �0.099 0.047Amphibians �0.007 �0.031Plankton 0.719 0.283Eigenvalue 17.4 7.93% Variation 54.3 24.8Cumulative % variation 54.3 79.2

Embilikala water depthVisual 0.184 0.484Tactile 0.471 0.720Pelagic 0.863 �0.497Eigenvalue 2.97 0.919% Variation 76.4 23.6Cumulative % variation 76.4 100

Fig. 6. (a) Multidimensional scale (MDS) ordination of shorebird communities(monthly average abundance) on Bundala (BUN) . and Embilikala (EMB) D lagoons,and (b) percentage composition of three foraging guilds (pelagic, tactile, visual)during wet . and dry D seasons in 2006 and 2007 on Bundala and Embilikalalagoons.

64 M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68

(Fig. 3a and b, Fig. 5a and b, Table 4) with consistent changes inshorebird abundance between the lagoons. As water levels in-creased, shorebird abundance decreased on both lagoons (yBundala =�0.011x + 3.38, r = �0.73, n = 11 p < 0.01; yEmbilikala = �0.014x + 4.45,r = �0.60 n = 11 p < 0.05, Fig. 3a and b) but there were differences inshorebird communities between the two lagoons that reflected theirhydrological state and habitat availability (Fig. 6a and b, Table 1,Appendix B). Tactile foraging shorebirds dominated Bundala lagoon(Fig. 6b), while pelagic and visual foraging shorebirds always com-prised >50% of total shorebird density on Embilikala lagoon(Fig. 6b). There was only similarity between shorebird communitieson the two lagoons early in the 2007 wet season, when water levelswere artificially lowered on Embilikala lagoon, more shoreline shal-low habitat became available (Table 1), and abundance of tactile spe-cies was highest (Fig. 6a and b). Tactile feeders however spent moretime resting than feeding on Embilikala than Bundala lagoon, even ifshallow water levels were available (dry season 2006 and 2007;Appendix B).

Tactile foraging shorebirds always occurred in shallow water(first axis PCA explained 96% variance for Bundala lagoon and73% variance for Embilikala lagoon) (Fig. 5a and b, Table 4) on bothlagoons but feeding efficiency differed between lagoons.Black-tailed godwit and curlew sandpiper had higher pecksrates (Hblack-tailed godwit = 100.6 p < 0.001; Hcurlew sandpiper = 15.26p < 0.001) and feeding success (Hblack-tailed godwit = 126.7 p < 0.001;Hcurlew sandpiper = 21.85 p < 0.001) on Bundala lagoon compared to

Embilikala lagoon (Table 5). There was no difference in peck ratesof little stint between the lagoons (HLittle stint = 3.48 p = 0.06).

Similarly, visual foraging shorebirds also associated with shal-low areas (Fig. 5a and b), their densities were higher (Hvisual = 17.98p < 0.001) in 2007 compared to 2006 on Bundala lagoon reflectingthe availability of shallow water habitat (Tables 1). Abundance ofvisual foragers also increased with drying on Embilikala lagoon(Table 1, Fig. 6b, Appendix B), with availability of shallow water(Fig. 5b, first axis PCA 76% variance explained) but similarly to tac-tile feeders, visual foraging shorebirds spent more time resting thanfeeding on Embilikala than Bundala lagoons, even if shallow waterhabitat was available (Table 1, Appendix B). Feeding efficiency of vi-sual foragers, however, differed between the two lagoons. All visualforaging species (Kentish plover, common redshank and lesser sandplover) had higher peck rates (HKentish plover = 7.28 p = 0.007;Hcommon redshank = 10.87 p = 0.001; Hlesser sand plover = 9.49 p = 0.002)and peck success (HKentish plover = 9.17 p = 0.002; Hcommon redshank =27.15 p < 0.001; Hlesser sand plover = 7.86 p = 0.005) on Bundalalagoon, compared to Embilikala lagoon (Table 5).

Contrastingly, pelagic foraging shorebirds were not responsiveto availability of shallow areas and declining water levels (Fig. 5aand b, Table 4).They foraged for most of the time (>71% feeding)at all depths (second axis PCA 24% variance explained for Embili-kala lagoon; 3% variance explained for Bundala lagoon) (Fig. 5aand b, Appendix B). There were differences in the foragingefficiency of pelagic foragers between the lagoons. Black-winged

Table 5Mean (±SE, min�1) pecks, success (number of prey swallowed) and efficiency rate (success/pecks) of the eight most abundant shorebird species in relation to foraging guild.Sample size indicated for each species (n).

Foraging guild Species Bundala lagoon Embilikala lagoon

n Pecks Success Efficiency n Pecks Success Efficiency

Tactile Black-tailed godwit 249 48(1.5) 41 ± 1.6 86% 75 17 ± 0.7 7 ± 0.5 42%Curlew sandpiper 99 47(2.0) 44 ± 2.2 94% 29 32 ± 1.7 24 ± 2.3 74%Little stint 223 50 ± 1.6 – – 64 44 ± 2.4 – –

Pelagic Black-winged stilt 89 28 ± 1.9 26 ± 2.1 90% 93 16 ± 1.0 12 ± 1.1 74%a 74 17(1.2) 13(1.2) 79%b 19 15(1.8) 8(1.4) 56%Marsh sandpiper 67 33 ± 1.9 30 ± 2.0 93% 20 39 ± 3.7 34 ± 4.0 87%

Visual Lesser sand plover 103 23 ± 0.2 22 ± 0.2 96% 51 15 ± 1.2 11 ± 0.9 71%Kentish plover 56 29.5 ± 2.3 29.2 ± 2.3 99% 19 17 ± 1.8 16 ± 1.6 95%Common redshank 61 38 ± 2.3 35 ± 2.6 92% 28 26 ± 1.8 15 ± 1.7 59%

aFeeding in water depth <12 cm.bFeeding in water depth >12 cm.

M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68 65

stilts had higher (H Black-winged stilt = 24.6 p < 0.001) feedingefficiency on Bundala lagoon than on Embilikala lagoon while peckrates of marsh sandpipers were higher on Embilikala lagoon com-pared to Bundala lagoon (H Marsh sandpiper = 6.11 p = 0.013) (Ta-ble 5) but there was no difference in feeding efficiency betweenlagoons (H Marsh sandpiper = 0.32 p = 0.06). On Embilikala feedingefficiency of black-winged stilts was also 79% when feeding in<12 cm water depth but considerably less, 54% when feeding indeeper water (Table 5).

4. Discussion

We showed that one of two lagoons in a Ramsar-listed wetlandwas degraded by upstream water resource development. Our com-parative approach of two adjacent functionally similar coastal la-goons, originally with similar flow regimes, provided powerfulevidence that flow altered shorebird habitat and food resources.This anthropogenic impact could be ameliorated through habitatrehabilitation.

There is increasing evidence that alteration of flow regimes sig-nificantly impacts downstream wetlands and their dependent bio-ta. Most understanding relates to reductions in flow impactingbiota (Dudgeon, 2000; Bunn and Arthington, 2002; Dudgeonet al., 2006) but there is also evidence that increasing flows impacton aquatic biota of wetlands (Sierzen et al., 2006; Nilsson andRenöfäit, 2008). Increased inflows from agricultural areas to oneof our two coastal lagoons in Sri Lanka altered its hydrology,changing its value for shorebirds. This change reduced temporalhydrological variability (months, years) for the altered lagoon,compared to the highly variable unaltered lagoon.

Shorebird guilds responded predictably to these changes asavailability (extent, timing and duration) of exposed foraging hab-itat affects shorebird foraging decisions (Hamer et al., 2006). Tac-tile and visual foraging shorebirds were abundant on the Bundalalagoon where slow natural drawdown (>7 months) allowed contin-uous feeding in shallow water habitats (1–10 cm). Contrastingly,hydrological stability on Embilikala lagoon impacted on the tem-poral availability of the foraging niche for tactile and visual forag-ers. Drawdown events on Embilikala lagoon lasted less than twomonths, exposing shallow foraging areas for a short period of timebefore areas were re-flooded by freshwater inputs from agricul-ture. It was not simply the size of habitat that was important butrather the accessible habitat (mediated by water level fluctuations)consistent with other published studies (Weber and Haig, 1996;Kelly, 2000; Colazzo et al., 2002). Further, on Embilikala even whenconditions were favourable (shallow water habitat available), tac-tile and visual foraging species spent more time resting than feed-

ing pointing to the lower habitat quality available on Embilikala. Ingeneral, shorebirds around the world spend most of their day for-aging, often at least two-thirds, except when affected by tides(Zwarts et al., 1990; Piersma and Ntiamoa-Baidu, 1995; Zwartset al., 1996; Ntiamoa-Baidu et al., 1998; Masero et al., 2000; vande Kam et al., 2004). In particular, small visual foragers (sandpip-ers, stints, plovers),the most abundant shorebirds at Bundala, feed-ing on small prey, may spend more than 75% of their time feeding(van de Kam et al., 2004). Pelagic foraging shorebirds were notresponsive to availability of shallow areas and declining water lev-els as they could forage also in medium water depth (11–30 cm)despite their foraging preferred depth was 1–10 cm and spentmost of the time foraging on both lagoons.

These hydrological differences also affected the distribution,abundance and composition of shorebird prey. Water depth andhydroperiod profoundly influence the structure of wetlandmacro-invertebrate communities (Nielsen et al., 2002; Tarr et al.,2005; Whiles and Goldwitz, 2005). Extended inundation decreasesabundance of invertebrates adapted to a cycle of wetting and dry-ing (i.e. amphipods, cladocerans, ostracods) (Neckles et al., 1990;Nielsen et al., 2002) and promotes the prevalence of planktonicspecies (Sierzen et al., 2006). There were similar patterns reflectedin the composition and abundance of invertebrates on Embilikalaand Bundala lagoons. Abundance and size of benthic prey, particu-larly amphipods and molluscs, were significantly higher on thehydrological dynamic Bundala lagoon than on the hydrologicallystable Embilikala. As a result, tactile foragers, feeding predomi-nantly on benthic prey (Higgins and Davies, 1996; del Hoyoet al., 1996), fed 20–40% more efficiently on Bundala lagoon thanon Embilikala lagoon. Prolonged inundation and high water levelson the altered Embilikala lagoon were also reflected in a higher ra-tio, and abundance of pelagic prey (plankton, small shrimp larvae),relative to benthic prey (molluscs, crustaceans, midges larvae).Even though pelagic shorebirds (e.g. black-winged stilt and marshsandpiper), were most common and abundant on Embilikala, theirforaging success was not as good as on Bundala lagoon. For exam-ple, black-winged stilt were 20% less efficient at foraging on thehydrologically stable Embilikala lagoon where water depths were>10 cm, compared to Bundala lagoon. Thus, the three shorebirdforaging guilds exhibited different distribution and foraging behav-ioural patterns on the two lagoons reflecting the patterns of habitatavailability and invertebrate prey composition, indicating the sub-optimal foraging habitat provided by Embilikala lagoon, comparedto Bundala lagoon. High quality and dependable food resources ofwintering grounds and stop over sites are critical for survival andreproduction of migratory shorebirds (Myers, 1983; Baker et al.,2004) and declining of quality of foraging habitat at stopover sitesis correlated to decline of many shorebirds populations around the

66 M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68

world (Baker et al., 2004; Smith et al., 2012). The reduced quality offoraging habitat of the impacted lagoon is a real concern for theconservation of the long distant migrants, such as curlew sandpip-ers and godwits, of the Ramsar site of Bundala.

Other factors associated with altered hydrologic regimes alsoprobably contributed to differences between the lagoons. Turbid-ity, caused by run-off from agriculture, was on average 1.5 timeshigher (Bellio et al., 2009) in Embilikala lagoon than in Bundala la-goon (Bellio et al., 2009). This could have contributed to reducedabundances of benthic amphipods, negatively affected by turbidity(Anteau and Aftond, 2008). It would have also decreased efficiencyof visual foraging shorebirds, reflected in reduced (<20–40%) feed-ing efficiency in Embilkala lagoon, compared to Bundala lagoon,even though insects and larvae were similarly abundant on thetwo lagoons. The abundant macrophytes on Embilkala lagoon com-pared to Bundala lagoon (Bellio et al., 2009) may also have pro-vided opportunity for abundant insect prey, such as waterboatmen, Corixidae, to escape visual foraging shorebirds (Brendocket al., 2003).

5. Conclusion

Habitat restoration and rehabilitation will play in the future apivotal role in recovering degraded or destroyed ecosystems (Er-win, 2009). The detrimental changes to flow regime and turbidityon Embiliklala lagoon could be altered to significantly rehabilitatehabitats for shorebird conservation and the Ramsar site more gen-erally. This could be done by reinstating the timing, amplitude andduration of natural flooding and drying cycles, similar to Bundalalagoon. It would require diversion of the drainage water to paddyfields, rather than flowing into Embilikala lagoon (Fig. 1: Weligattaarea). Such management could benefit rice growers, potentiallyimproving availability of water for rice production. We predict thatthis change would increase the productivity of the Embilikala la-goon for foraging shorebirds, rehabilitating this Ramsar site. Drain-age water from the paddy fields could be then released directly tothe sea. Restoration of shorebird habitat on Embilikala could alsoimprove fisheries and tourism. Freshwater inflows from agricul-ture have caused a decline in the shrimp industry on Embilikala la-goon, lowering water salinity of the lagoon (CEA and Euroconsult,1993). Shrimps are also the main food source for wintering flamin-gos (Phoenicopterus ruber) (Matsuno et al., 1998), the iconic touris-tic attraction of the Ramsar site; which have declined (Matsunoet al., 1998). Such rehabilitation could be easily measured usingshorebirds and changes in their composition and foraging efficien-cies. Diversion of inflows of freshwater from irrigation will allow toincrease the water salinity in Embilikala, restoring the pre-irriga-tion development characteristics of this brackish water coastal la-goon. Similarly, decrease of sediments-rich run-off fromagriculture will also potentially improved water turbidity whichlevel have been known to have increased after irrigation develop-ment (CEA and Euroconsult, 1993; Amarasinghe et al., 2001; Bellioet al., 2009).

Often environmental degradation of wetlands is the product ofidiosyncratic local changes to flooding regime, significantly affect-ing ecological character of wetlands. If the cause can be identified,as with Embilkala lagoon, it offers opportunity for local rehabilita-tion, if there is sufficient will and resources. Effecting such changeis probably difficult and would require Sri Lankan authorities torecognise the problem and embrace the solution. Given the areais a Ramsar site, support for rehabilitation could be garneredthrough argument that a change in ecological character, a funda-mental premise of the Ramsar Convention, has occurred to Embili-kala lagoon. Finally, the ecological character of this Ramsar site isalso threatened by other impacts. Invasive plants such as Opuntia

sp. (cactus) and Prosopis juliflora (mesquite) are a serious threatto Ramsar site of Bundala National Park (Sunandal, 2008). Fresh-water inflows from agriculture have decreased the salinity onEmbilikala and increased the duration of high water levels, possi-bly favouring the spread and development of invasive aquaticplants such as Salvinia sp., and cattail (Typha angustifolia). Theseinvasive species have caused major changes to the ecosystemand affected areas used by animals for grazing as well as habitatsfor shorebirds and fish (Sunandal, 2008). It is highly recommendedthat management of these invasive species be considered in anyhabitat rehabilitation.

Shorebird conservation ultimately depends on maintenance ofkey functional processes (flooding and drying), often affected byanthropogenic impacts. There are opportunities to rehabilitatewetland sites, such as Embilikala lagoon within the Bundala Ram-sar site, to improve opportunities for long-term conservation ofshorebird populations. Such approaches may similarly be applica-ble to identification of causes and consequences of altered floodingregimes of other important wetland or Ramsar sites around theworld.

Acknowledgments

This study was financially supported by the International WaterManagement Institute (IWMI) Colombo, Sri Lanka, and by the Aus-tralian Wetlands and Rivers Centre of the University of New SouthWales, scholarship APA (Australian Post Graduate ResearchAward). We thank Prof. S. W. Kotagama and the Department ofWildlife Conservation, Sri Lanka, for permits for surveys in BundalaNational Park and onsite support by park rangers Nuwan Chamaraand Hapugoda P.A. Prasantha.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.biocon.2013.07.013.

References

Amarasinghe, F., Matsuno, Y., de Silva, R.I., Piyankarage, S.C., Bambaradeniya, C.N.B.,Mallawatanthiri, A., 2001. Impact of irrigation on water quality, fish andavifauna of three lagoons in southern Sri Lanka. Conference Proceedings of theAsian Wetland Symposium, Penang Malaysia, p. 183.

Anteau, M.J., Aftond, A.D., 2008. Amphipods densities and indices of wetland qualityacross the upper Midwest, USA. Wetlands 28 (1), 184–196.

Agardy, T., Alder, J., 2005. Coastal systems. In: Hassan, R., Scholes, R., Ash, N. (Eds.),Millenium Ecosystem Assessment. Ecosystems and Human Well-being. IslandPress, London, pp. 15–39.

Alfaro, M., Clara, M., 2007. Assemblage of shorebirds and seabirds on Rocha Lagoonsandbar, Uruguay. Ornithol. Neot. 18, 421–434.

Altmann, J., 1974. Observational study of behaviour: sampling methods. Behav 49,227–267.

Baker, A.J., González, P.M., Piersma, T., Niles, L.J., Nascimento, I.d.L.S.d.P., Atkinson,W., Clark, N.A., Minton, C.D.T., Peck, M.K., Aarts, G., 2004. Rapid populationdecline in red knots: fitness consequences of decreased refuelling rates and latearrival in Delaware Bay. Proceedings of the Royal Society of London: Series B271, 875–882.

Balachandra, S., 2006. The decline in wader populations along the east coast of Indiawith special reference to Point Calimere, south-east India. In: Boere, G.C.,Galbraith, C.A., Stroud, D.A. (Eds.), Waterbirds Around the World. The StationeryOffice.

Bambaradeniya, C.N.B., 2001. Guide to Bundala: a Guide to the Biodiversity ofBundala National Park – a Ramsar Wetland in Sri Lanka. IUCN (ed), Sri Lanka, p.54.

Bellio, M.G., Kingsford, R.T., Kotagama, S.W., 2009. Natural versus artificial –wetlands and their waterbirds in Sri Lanka. Biol. Conserv. 142 (12), 3076–3085.

Boynton, W.R., Murray, L., Hagy, J.D., Stokes, C., 1996. A comparative analysis ofeutrophication patterns in a temperate coastal lagoon. Estuaries 19 (28), 406–421.

Boros, E., Szabolcs, B., Forro, L., 2006. Anostracans and microcrustaceans as potentialfood sources of waterbirds on sodic pans of the Hungarian plain. Hydrobiologia5676, 341–349.

M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68 67

Brendock, L., Maes, J., Rommens, W., Dekeza, N., Nhiwatina, T., Barson, M., Callebaut,V., Phiri, C., Moreau, K., Gratwicke, B., 2003. The impact of water hyacinth(Eichornia crassipes) in a eutrophic subtropical impoundment (Lake Chivero,Zimbabwe). II Species diversity. Arch. Hydr. 158 (3), 369–405.

Bunn, S.T., Arthington, A.H., 2002. Basic principles and ecological consequences ofaltered flow regimes in aquatic biodiversity. Environ. Manage. 30 (4), 492–503.

CEA (Central Environmental Authority), Euroconsult, 1993. Wetland site report andconservation management plan. Bundala National Park. Central EnvironmentAuthority – Sri Lanka and Euroconsult (ed), The Netherland.

Clarke, K.R., Warwick, R.M., 2001. Change in marine communities: an approach tostatistical analysis and interpretation, 2nd ed. Primer-E Ltd. Plymouth MarineLaboratory, UK.

Colazzo, J.A., O’Harra, D.A., Kelly, C.A., 2002. Accessible habitat for shorebirds:factors influencing its availability and conservation implications. Waterbirds 25(special issue), 13–24.

Cooray, P.G., 1984. The Geology of Sri Lanka (Ceylon). National Museum Publication,Department of Government Printing (ed), Colombo, Sri Lanka.

Delany, S., Szabolcs, N., Davidson, N., 2010. State of the World’s Waterbirds 2010.Wetland International (ed), The Netherlands.

del Hoyo, J., Elliott, A., Sargatal, J., 1996. Order Charadriiformes. In: del Hoyo, J.,Elliott, A., Sargatal, J. (Eds.), Handbook of the birds of the World, Hoatzin toAuks, vol. 3. Lynx Edicions, Barcelona, Spain, pp. 276–537.

Dudgeon, D., 2000. Large-scale hydrological changes in tropical Asia: prospects forriverine biodiversity. Bioscience 50 (9), 793–806.

Dudgeon, D., Arthington, A.H., Gessner, M.O., Kawabata, Z., Knowler, D.J., et al., 2006.Freshwater biodiversity: importance, threats, status and conservationchallenges. Biol. Rev. 81, 163–182.

Erwin, K.L., 2009. Wetlands and global climate change: the role of wetlandsrestoration in a changing world. Wetlands Ecol. Manage. 17, 71–84.

Fabricius, K.E., 2005. Effects of terrestrial runoff on the ecology of corals and coralreefs: review and synthesis. Mar. Pollut. Bull. 50, 125–146.

Ferreira, W.L.S., Bemvenuti, C.E., Rosa, L.C., 2005. Effects of the shorebirds predationon the estuarine macrofauna of the Patos Lagoon, South Brazil. Thal. 21 (2), 77–82.

Finlayson, C.M., Davidson, N., Pritchard, D., Milton, G.R., MacKay, H., 2011. TheRamsar convention and ecosystem-based approaches to the wise use andsustainable development of wetlands. J. Int. Wil. Law Pollut. 14, 176–198.

Finlayson, C.M., Davidson, N.C., Stevenson, N.J. (Eds.), 2001. Wetland inventory,assessment and monitoring: Practical techniques and identification of majorissues. Proceedings of Workshop 4, 2nd International Conference on Wetlandsand Development, Dakar, Senegal, 8.14 November 1998, Supervising ScientistReport 161, Supervising Scientist, Darwin.

Galbraith, H., Amerasinghe, P., Huber-Lee, A., 2005. The Effects of AgriculturalIrrigation on Wetland Ecosystems in Developing Countries: a Literature Review.International Water Management Institute, Colombo, Sri Lanka.

Gans, C., 1961. The feeding efficiency of snakes and its possible evolution. Am. Zool.1, 217–227.

Gillanders, B.M., Kingsford, M.J., 2002. Impact of changes in flow of freshwater onestuarine and open coastal habitats and associated organisms. Oceanogr. Mar.Biol.: Ann. Rev. 40, 233–309.

Halpern, B.S., Ebert, C.M., Kappel, C.V., Madin, E.M.P., Micheli, F., Perry, E., Selkoe,K.A., Walbridge, S., 2009. Global priority areas for incorporating land–seaconnections in marine conservation. Conserv. Lett. 2 (4), 186–196.

Hamer, G.L., Heske, E.J., Brawn, J.D., Brown, P.W., 2006. Migrant shorebird predationon benthic invertebrates along The Illinois river, Illinois. Wild. J. Ornithol. 118(2), 152–163.

Heise, B.A., 1992. Sensitivity of mayfly nymphs to red light: implications forbehavioural ecology. Freshwater Biol. 28, 331–336.

Higgins, P.J., Davies, S.J.J.F. (Eds.), 1996. Handbook of Australian, New Zealand andAntarctic Birds. Volume Three – Snipe to Pigeons. Oxford University Press,Melbourne, Victoria.

IIMI (International Irrigation Management Institute), 1995. Irrigation managementand crop diversification. Volume II, final report on the Technical AssistanceStudy, Phase II. IIMI Colombo, Sri Lanka.

Jayawardena, H.M.K.S., 1993. Hydrological aspects of the Bundala SanctuaryWetland Conservation Program Report. Central Environmental Authority (ed),Colombo, Sri Lanka.

Kelly, C. 2000. Shorebirds populations and habitat quality at Merrit Island.Unpublished M.S. Thesis. North Carolina State University, Raleigh.

Kingsford, R.T., Lemly, A.D., Thompson, J.R., 2006. Impacts of dams, rivermanagement and diversions on desert rivers. In: Kingsford, R.T. (Ed.), Ecologyof Desert Rivers. Cambridge University Press, Cambridge, UK, pp. 203–247.

Kjerfve, B., Magill, K.E., 1989. Geographic and hydrodynamic characteristics ofshallow coastal lagoons. Mar. Geol. 88, 187–199.

Li, Z.W.D., Mundkur, T., 2004. Numbers and distribution of waterbirds and wetlandsin the Asia Pacific Region. Results of the Asian Waterbird Census 1997-2001.Wetland International (ed), The Netherlands.

Li, Z.W.D., Ounsted, R., 2007. The status of coastal waterbirds and wetlands inSoutheast Asia: results of waterbird surveys in Malaysia (2004–2006) andThailand and Myanmar (2006). Wetlands International (ed), Malaysia.

Martin, P., Bateson, P., 1986. Measuring Behaviour. An Introductory Guide.Cambrideg University Press, UK.

Masero, J.A., Pérez-Hurtado, A., 2001. Importance of the supratidal habitats formaintaining overwintering shorebird populations: how redshanks use tidalmudflats and adjacent saltworks in Southern Europe. The Condor 103 (1), 21–30.

Masero, J.A., Perez-Hurtado, A., Castro, M., Arroyo, G.M., 2000. Complementary useof intertidal mudflats and adjacent salinas by foraging waders. Ardea 88 (2),177–191.

Mathot, K.J., Smith, B.D., Elner, R.W., 2007. Latitudinal clines in food distributioncorrelate with differential migration in Western Sandpiper. Ecology 87, 781–791.

Matsuno, Y., van der Hoek, W., Ranawake, M., 1998. Irrigation water managementand the Bundala National Park. In: Proceedings of the Workshop on WaterQuality of the Bundala Lagoons., Colombo. International Water ManagementInstitute (IWMI), Sri Lanka, pp 48.

Mundkur, T., 2006. East Asia-pacific flyway. Workshop introduction. In: Boere, G.C.,Galbraith, C.A., Stroud, D.A. (Eds.), Waterbirds Around the World. The StationeryOffice, Edinburgh, UK, pp. 315–318.

Myers, J.P., 1983. Conservation of migrating shorebirds: staging areas, geographicbottlenecks and regional movements. American Birds 37, 23–25.

Neckles, H.A., Murkin, H.R., Cooper, J.A., 1990. Influences of seasonal flooding onmacroinvertebrate abundance in wetland habitats. Freshwater Biol. 23, 311–322.

Nikolaev, N.E., 2007. Study on phototropism of freshwater invertebrates usingSvetolovushek underwater camera. Herald TSU Serie. Biol. Ecol. 6, 115–117.

Nielsen, D., Hillman, T., Smith, F., Shiel, R., 2002. The influence of seasonality andduration of flooding on zooplankton in experimental billabongs. Riv. Res. Appl.18 (3), 227–237.

Nilsson, C., Renöfäit, B.M., 2008. Linking flow regime and water quality in rivers: achallenge to adaptive catchment management. Ecol. Soc. 13 (2) (special feature,art.18).

Ntiamoa-Baidu, Y., Piersma, T., Wiersma, P., Poot, M., Battley, P., Gordon, C., 1998.Water depth selection, daily feeding routines and diets of waterbirds in coastallagoons in Ghana. Ibis 140, 89–103.

Pepping, M., Piersma, T., Pearson, G., Lavaleye, M., 1999. Interdital sediments andbenthic animals of roebuck Bay Wetsern Australia. Netherland Institute for SeaResearch (NIOZ). Report of the Roebuck Bay Intertidal benthic Mappingprogramme, June 1997 (ROEBIM-97).

Perera, N., 2008. An exceptional wetland facing multitude of problems. Siyoth 2 (1),4–8.

Piersma, T., Ntiamoa-Baidu, Y., 1995. Waterbird ecology and the management ofcoastal wetlands in Ghana. NIOZ (Netherland Institute for Sea Research). Texel,The Netherland. Wildlife Department and Ghana Wildlife Society, Accra, Ghana.pp. 105.

Pringle, C.M., 2000. Integration of aquaculture into crop-animal systems in Asia.Agr. Syst. 71, 127–146.

Rabalais, N.N., Turner, R.E., Diaz, R.J., Justic, D., 2002. Global change andeutrophication of coastal waters. ICES J. Mar. Sci. 66, 1528–1537.

Ramsar, 1993. Recommendation 5.2: Guidelines for interpretation of Article 3(‘‘ecological character’’ and ‘‘change in ecological character’’). In: 5th Meeting ofthe Conference of the Contracting Parties Kushiro, Japan 9–16 June, 1993.

Rapold, C., Kersten, M., Smit, C., 1985. Errors in large scale shorebirds counts. Ardea73, 13–24.

Roni, P., Hanson, K., Beechie, T., Pess, G., Pollock, M., Bartley, D.M., 2005. Habitatrehabilitation for inland fisheries. Global review of effectiveness and guidancefor rehabilitation of freshwater ecosystems. FAO Fisheries Technical Paper. No.484. Rome, FAO. 2005. 116pp.

Rubega, M.A., 1997. Surface tension prey transport in shorebirds: how widespreadis it? Ibis 139, 488–493.

Schilling, E.G., Loftin, C.S., Huryn, A.D., 2009. Effects of introduced fish onmacroinvertebrate communities in historicallly fishless headwater and kettlelakes. Biol. Conserv. 142 (12), 3030–3038.

Shepherd, P.C.F., Boates, J.S., 1999. Effects of commercial baitworm harvest onsemipalmated sandpipers and their prey in the bay of Fundy HemisphericShorebird Reserve. Conserv. Biol. 13 (2), 347–356.

Sierzen, M.E., Peterson, G.S., Trebitz, A.S., Brazner, J.C., West, C.W., 2006. Hydrologyand nutrient effects on food-web structure in ten Lake Superior coastalwetlands. Wetlands 26 (4), 951–964.

Smakhtin, V.U., Piyankarage, S.C., 2003. Simulating hydrologic reference conditionof coastal lagoons affected by irrigation flows in southern Sri Lanka. Wetlands23 (4), 827–834.

SPSS, 2009. SPSS grad-pack for students version 17.0.2, SPSS Inc. 2009.Smith, P.A., Gratto-Trevor, C.L., Collins, B.T., Fellows, S.D., Lanctot, R.B., Liebezeit, J.,

McCaffery, B.J., Tracy, D., Rausch, J., Kendall, S., Zack, S., Gates, H.R., 2012. Trendsin abundance of Semipalmated Sandpipers: evidence from the Arctic.Waterbirds 35 (1), 106–119.

Stroud, D.A., Baker, A., Blanco, D.E., Davidson, N.C., Delany, S., Ganter, B., Gill, R.,Gonzáles, P., Haanstra, L., Morrison, R.I.G., Piersma, T., Scott, D.A., Thorup, O.,West, R., Wilson, J., Zöckler, C., 2006. The conservation and population status ofthe world’s waders at the turn of the millenium. In: Boere, G.C., Galbraith, C.A.,Stroud, D.A. (Eds.), Waterbirds Around the World. The Stationery Office,Edinburgh, UK, pp. 643–648.

Sunandal, K.D., 2008. Control of invasive Alien plants in Bundala National Park. Bookof abstracts. National Symposium on Invasive Alien Species. Sri LankaFoundation Institute, 11 November, Colombo, pp. 25.

Tarr, T.L., Baber, M.J., Babbitt, K.J., 2005. Macroinvertebrate community structureacross a wetland hydroperiod gradient in southern New Hampshire, USA.Wetlands Ecol. Manage. 13, 321–334.

UNESCO, 1993. Coasts – Managing Complex Systems. <http://www.unesco.org/csi/intro/coastse.pdf>.

van de Kam, J., Bruno, E., Piersma, T., Zwarts, L., 2004. Shorebirds. An illustratedbehavioural ecology. KNNV Publisher, The Netherlands, pp. 157–160.

68 M. Bellio, R.T. Kingsford / Biological Conservation 167 (2013) 57–68

Vörösmarty, C.J., McIntyre, P.B., Gessner, M.O., Dudgeon, D., Prusevich, A., Green, P.,Glidden, E., Bunn, S., Sullivan, C.A., Reidy Liermann, C., Davies, P.M., 2010. Globalthreats to human water security and river biodiversity. Nature 467, 555–561.

Whiles, M.R., Goldwitz, B.S., 2005. Macroinvertebrate communities in Central PlateRiver wetlands: patterns across a hydrologic gradient. Wetlands 25 (2), 462–472.

Weber, I., Haig, S.M., 1996. Shorebirds use of South Carolina managed and naturalcoastal wetlands. J. Wild. Manage. 60, 73–82.

Zwarts, L., Blomert, A.-M., Hupkes, R., 1990. Increase of feeding time in waderspreparing for spring migration from the Banc d’Arguin, Mauritania. Ardea 78,237–256.

Zwarts, L., Ens, B.J., Goss-Custard, J.D., Hulscher, J.B., Kersten, M., 1996. WhyOystercatchers Haematopus ostralegus cannot meet their daily energyrequirements in a single low water period. Ardea 84A, 269–290.

Zwarts, L., Wanink, J.H., 1991. The macrobenthos fraction accessible to waders mayrepresent marginal prey. Oecologia 87, 581–587.

Zwarts, L., Wanink, J.H., 1993. How the food supply harvestable by waders in theWadden Sea depends on the variation in energy density, body weight, biomass,burying depth and behaviour of tidal-flat invertebrates. Netherland J. Sea Res.31, 441–476.