Bird assemblages as bio-indicators of water regime management and hunting disturbance in natural wet grasslands

Biological Conservation 106 (2002) 115–127 www.elsevier.com/locate/biocon

Bird assemblages as bio-indicators of water regime management and hunting disturbance in natural wet grasslands J.-M. Paillissona,*, S. Reeberb, L. Mariona,b a

UMR 6553 Ecobio, Laboratoire d’Evolution des Syste`mes naturels et modifie´s, Universite´ de Rennes I, Campus de Beaulieu, F-35042 Rennes Cedex, France b ´serve Re naturelle du Lac de Grand-Lieu, Socie´te´ Nationale de Protection de la Nature, 15 rue de la Chaˆtaigneraie, F-44380 Bouaye, France Received 11 April 2001; received in revised form 27 July 2001; accepted 24 October 2001

Abstract Responses of the waterbird community to the management regime of a wet grassland system (primarily small changes in the water regime but also hunting disturbance) in a large floodplain lake, Grand-Lieu, were studied during three consecutive years (winter and spring in 1998–2000). These temporarily flooding grasslands (1100 ha) constituted a very important feeding area by supporting large flocks of waterbird species (52 regular spp., totalling up to 14 250–22 850 birds). A global similar seasonal pattern in the presence of waterbird species was recorded whatever the year. The March–May period was a key phase with the maximum species richness (42–48 spp.) and the maximum abundance (30–40% of the total according to year, essentially Anseriformes). Late in spring Ciconiiformes and to a lesser extent some waders exploit this feeding area. Nevertheless some changes in bird assemblages were exhibited according to the management regime (a decreasing spring water level from 1998 to 2000, and an exceptional hunting ban in January 2000, instead of an usual stop in late February). The hunting disturbance clearly limited diurnal accessibility of ducks to wet grasslands in January and February 1998 and 1999. Also, the duck population in flooding grasslands increased by 55– 65% in January 2000 when hunting was prohibited and in the same time the richness species grew from 14–19 spp. up to 23 spp. in 2000. The overall impact of a high spring water level on waterbirds as in 1998 and to a lesser extent in 1999 (with respectively 54 and 35 cm of mean water level against 25 cm in 2000) was an increase in diving fish-eating birds and larger flocks of resting gulls. Conversely these conditions, notably a reduced period with a spring water level under 25 cm in spring (20 days in 1998–1999 instead of 40 days in 2000) were adverse to numerous ground-feeding waterbirds (Ciconiiformes and waders) that require shallow waters. This study showed that limitation of human disturbance and moderate flooding of wet grasslands can increase their attractiveness for numerous species of high conservation status. # 2002 Elsevier Science Ltd. All rights reserved. Keywords: Biodiversity; Community; Multivariate analysis; Water level; Wetland

1. Introduction Wetlands play an important role in biodiversity because they are attractive to many species due to their large habitat diversity and their great productivity providing nutrients and other resources (Weller, 1988; Elmberg et al., 1994). Birds are among the most conspicuous of wetland animals and various species are extremely sensitive to large hydrological changes (Kushlan, 1986a; Crowder and Bristow, 1988; Pyrovetsi

* Corresponding author. E-mail address: [email protected] (J.-M. Paillisson).

and Papastergiadou, 1992). Creation of reservoirs demonstrated immediate effects on waterbird communities (Hunter et al., 1987; Bildstein et al., 1994) and water conditions are among the main factors affecting the composition and the abundance of waterbird communities directly and indirectly (Dister et al., 1990; Briggs et al., 1998; Osiejuk et al., 1999). Water level fluctuations influence the physical structure of habitats (zonation of vegetation), the availability and accessibility of food (Clausen, 2000) and the presence of safe roosting or breeding sites (Green and Robins, 1993; Guillemain et al., 2000). All these characteristics determine habitat selection by waterbirds (Sanders, 1999) and especially by migrant birds that depend largely on the presence of well functioning wetlands along their

0006-3207/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S0006-3207(01)00239-7

116

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

flyways (Ens et al., 1994; van Eerden, 1997; Madsen, 1998a, b; Ntiamoa-Baidu et al., 1998). Human disturbances, especially hunting, can interfere with these factors and limit the accessibility to resources (e.g. Madsen, 1998a). However, the relative effect of hunting has not been assessed in relation to other parameters influencing the use of wetlands by waterbirds. Moreover, the detailed response of different taxonomic groups of waterbirds to low seasonal changes in water level has rarely been assessed (Elphick and Oring, 1998). Indeed, the studies of natural wetlands that have established relationships between water level and bird population changes concerned essentially responses either of specific groups, Ciconiiformes or ducks (Marion, 1989; DuBowy, 1996; Dimalexis and Pyrovetsi, 1997; Kingsford and Johnson, 1998; Duncan et al., 1999) or of the whole community (Hunter et al., 1987; Crivelli et al., 1995; Briggs et al., 1997; Hertzman and Larsson, 1999) exclusively to large water depth fluctuations or drainage, while small variations in water level on flooding surrounding grasslands have been little studied. Also, most of previous studies have described the responses on a taxonomic basis rather than on an ecological basis. The objectives of the present study are to identify firstly the patterns in the use of a wet grassland system by the whole waterbird community during a 3-year period, and secondly the effect of the management regime of the ecosystem (primarily small changes in the water regime, but also hunting disturbance) on the species assemblages. We compared the response of the different species groups, mainly Ciconiiformes, Anseriformes and Charadriiformes, recorded on flooded wet grasslands over three consecutive years (winter and spring) with partially controlled water regimes and an opportunistic exceptional hunting ban imposed during one of these years. The role of the water level on the breeding populations of the waterbird community in the different habitats of the lake will be addressed in another paper and will complement management recommendations made in the present study. Practical management applications are discussed to define the optimum management regime for preserving or enhancing the value of the wetland as feeding habitat for the whole waterbird community and more specifically for some species of high conservation status.

2. Materials and methods 2.1. Study site The Lac de Grand-Lieu is a shallow, turbid, eutrophic natural freshwater ecosystem in Western France (47
050 N, 1
390 W). This lake constitutes one of the most important bird areas in France supporting large numbers of wintering and migrating ducks and Fulica atra

(about 20 000 birds) and breeding populations of ducks and heron species, the latter representing also one of the main European breeding populations and the largest colonies in France for Ardea cinerea, Egretta alba, Platalea leucorodia and Threskiornis aethiopicus (Rocamora and Yeatman-Berthelot, 1999; Marion et al., 2000). The whole wetland is designated a Special Protection Areas under the EC Birds Directive and a Ramsar Site (Deceuninck et al., 2000). The wetland is of value because of its diversity of habitats for waterbirds, with four major complementary functional units (see Marion et al., 1994, for a more detailed description). The lake covers 4000 ha in summer and 6300 ha in winter, by flooding adjacent peaty wet grasslands used mainly as pastures. About 20 km of the lake margin (2000 ha) are covered by a floating peat fen, which becomes progressively exposed in summer. Most of the permanently central flooded area of the lake is covered from April to October by about 1400 ha of floating macrophytes. The central open water region without floating or emergent plants except small patches of submerged macrophytes covers about 600 ha. The wet grasslands of the Lac de Grand-Lieu can be considered as ecotonal habitats, as described by Dobrowolski (1997), and represent the principal feeding area of the wetland for wintering and migrating waterbirds. The grassland consists of level fields intersected by drainage channels. Human activities like mowing and grazing prevent the growth of trees and have formed fields with a heterogeneous vegetation structure dominated by Phalaris arundinacea, Glyceria maxima, Carex spp. or Eleocharis spp. The present study focuses on the use of the southwest unit by the waterbird community, the largest and most important part of the wet grasslands for waterbirds. 2.2. Water regime A drainage channel flows from the lake to the Loire estuary 25 km away and a sluice gate regulates the water level of the lake, essentially during the hydrological storage period (mainly in winter) by discharging the water surplus to the outlet. In the past, pressure from the farming community for land drainage in spring contributed to reductions in water level and duration of flooding. During our study, three different water levels were recorded: a high water level in 1998, a low water level in 2000 and an intermediate one in 1999 (Fig. 1). Such a difference, occurring between three consecutive years in the same site, limits variations due to long term changes in population dynamics of birds related to other factors. The water level is described by using a local depth measure (Buzay scale=0.47 m above French Geographic Level NGF). The mean level of the wet grasslands is 1.89 m Buzay (Marion and Marion, 1975), with

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

117

different fixed-observation points at each 5-day interval to test the overall effect of the two major local descriptors of the management regime (water level and hunting activity) on the waterbird community. 2.4. Data analysis

Fig. 1. Mean water level in each 5-day interval (from January to July) on the wet grassland system of the Lac de Grand-Lieu during the 1998– 2000 period. The horizontal line symbolizes the threshold of 25 cm.

a relatively slow variation (SD=0.10 m). Thus a mean water level of 2.14 m Buzay corresponded to a real mean water level of 25 cm on the grasslands. For each year we calculated the length of time that the level was below this mean threshold of 25 cm commonly defined as the maximum feeding depth for the majority of waterbird species, especially Ciconiiformes and Charadriiformes (Owen and Black, 1990; Ntiamoa-Baidu et al., 1998). 2.3. Bird census All waterbirds (here defined as Podicipediformes, Pelecaniformes, Ciconiiformes, Anseriformes, Gruiformes and Charadriiformes) were censused on the southwest wet grasslands from eight fixed-observation points covering 1100 ha (the other 300 ha, with many trees, not being used by large numbers of birds). For each observation point, waterbird species were counted individually or in units of 10 when flocks were larger than 100. When we encountered very large flocks of resting ducks (> 1000), we first counted them in units of 100 and then we counted each duck species in units of 10 to control both methods (Reeber, 2000). Given the open nature of the wet grassland system studied these counts were likely to estimate absolute abundance accurately for most species. Nevertheless species for which this was unlikely to be true (common snipe Gallinago gallinago and little grebe Tachybaptus ruficollis) are not considered here. Birds seen flying overhead were not included so that they were not counted again at another station. The censuses were conducted at 5-day intervals in three consecutive years (1998–2000) from the fifth period of 5 days in January, noted Jan V, to the second period of 5 days in July, noted Jul II. Hunting activity on the wetland occurred from mid-July to the end of February in all years except 2000, when an exceptional ban was imposed in January (due to the Erika disaster) allowing the effects of disturbance on the waterbird community to be assessed. We combined data for the

We used standard measures of species diversity to characterize the patterns of waterbird community in the wet grassland system: species richness and Shannon diversity (Guillory, 1999). A mean value of the Shannon diversity was calculated for each month from January to July. Annual frequencies of occurrence (% of species present in all the 5-day intervals) and monthly abundance were also defined. The first step of the assemblage analysis was to classify species into groups, achieving this by a posteriori analysis of their abundance and distribution to identify species exhibiting similar patterns (Gawlick et al., 1998, and literature cited therein). We performed a Foucart’s Correspondence Analysis (CA) on a series of three contingency tables (one per year, counts at 5-day intervals for all species). This method (rarely used in ecology) represents a CA based on an average table called a compromise table (see Foucart, 1978, and Blanc et al., 1998, for more details of the method). We used three graphic representations: a graph linking consecutive 5day intervals linked with a line according to years, a dendrogram based on a clustering procedure (Euclidian distances and then UPGMA of factorial coordinates— Unweighted Pair Group Method Algorithm), and plots of axes for each annual data set to identify species assemblages on the compromise plane (convex hull procedure). Analyses and graphics were performed using ADE-4 software (Thioulouse et al., 1997) and species with less than 10 birds recorded for each year were excluded from the matrices. Numbers of waterbirds were log (x+1) transformed prior to analysis. We used Wilcoxon signed rank tests (SYSTAT Software, 1998) to compare the mean abundance of each species between the three years to justify patterns of waterbird species in 5-day intervals and to determine the sensitivity of species groups to water regime. For each species the mean abundance was calculated for the same period of presence for the 3-year period, rather than the complete period of census. The comparisons were conducted according to the different water regime phases (Section 3). Finally some complementary comparisons of mean abundance were performed on the basis of numerical proportions of breeding population size of species feeding on the wet grasslands during the most variable period identified (Reeber, 2000). These latter tests were used to minimize the effect of possible variations in biogeographical population sizes, already reduced by the monitoring during three consecutive years.

118

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

3. Results 3.1. Water regime The water level in wet grasslands reached a maximum value in 1998 in January (Jan V: 117 cm, Fig. 1). In February, the highest water level was recorded in 2000 (83 cm in Feb IV) while the other years were similar. The smallest between-year difference occurred in March (54.5  2.5 cm) despite variable patterns (a stable water level in 1998, a decreasing water level in 1999 and 2000). Major prolonged differences were noted from April to mid-June (phase 2) with a high flooding in 1998 (maximum value of 83.5 cm in May I) and a more moderate one in 1999 (a peak of 56.0 cm in May I) before grasslands were on average exposed in the second part of June in all the three years. In spring the mean water level was 53.5  16.5 cm in 1998, 34.5  10.0 cm in 1999 and 24.5  8.0 cm in 2000. Duration of water levels of less than 25 cm was thus more limited in 1998 and 1999 (20 days, from May V to June III) and twice as long in 2000 (40 days, from May II to June III). 3.2. Composition and biodiversity indices of the waterbird community A total of 52 regular different waterbird species was recorded in the wet grasslands each year, except in 1999 with no Tringa glareola (Table 1). The most abundant waterbird species were Charadriiformes (27 species), Anseriformes (13) and Ciconiiformes (8). The different taxonomic groups included species that spend part of their life at the study site, some being always present, as well as migrant species making more temporary use of the wet grassland system, especially waders (10 species with a frequency of occurrence of less than 50%). There were great differences in the frequency of occurrence of waterbirds between the three years, notably among Charadriiforme species, but also among some Anseriformes. The monthly species richness patterns were generally similar between the 3 years (Fig. 2), with a maximum number of waterbird species from March to May (from 39 to 50) and fewer species in January and in July, especially in the two first years. The species richness pattern in 2000 was slightly different because of a high value in January (23 spp.) during the exceptional hunting ban followed by a resumption of hunting in February. This change in human disturbance induced a marked difference in curve pattern this year comparatively to the usual curves of the other years (see also mean abundance). The attractiveness of the wet grasslands was prolonged in 2000 since a larger number of species was recorded in July (30) than for 1998 (19) and 1999 (14). However the species richness was higher in February 1998 (35) than during the other years (26).

The mean abundance patterns showed three major results on the use of the wet grassland system (Fig. 2). The maximum mean abundance of waterbirds was observed in March (from 11200 to 14 800 respectively in 1999 and 2000, essentially Anseriformes and Gruiformes, Fig. 3) and represented from 30 (2000) to 40% (1998 and 1999) of the total annual abundance. The mean abundance was relatively lower in the other months, especially in the May–July period whereas the species richness was still high. At this period the wet grasslands were however largely used by Charadriiformes (41–46%) and Ciconiiformes (13–31%) which previously used this habitat little or not at all when the water level was higher (12–21% together). Finally the grasslands were more heavily used by the waterbird community in 2000, with more birds counted in almost all months (Fig. 2). The Shannon diversity was relatively constant between months in 1998 with several similar values (Fig. 4), whereas this index increased in 1999 and 2000 from low values in the first 2 months (1.06–1.35) to the highest diversity in the May–July period (2.85–3.31). During the last 2 years reduced values were linked to dominance of Fulica atra in January and February, when it accounted for 65–70% of the total abundance (Fig. 3). The intermediate values were noted when the waterbird community was large but also composed of large flocks of Anseriformes (March and April). Finally the Shannon diversity reached a peak when the community was again diversified and more equilibrated. 3.3. Definition of species assemblages and responses to the water regime The multivariate analysis was performed on contingency tables composed of 40 species, excluding species with less than 10 birds recorded for each year. The first two axes of the Foucart’s CA accounted for 67% of the total variation in sample-waterbird composition in the 3 years. The graph linking subsequent 5-day intervals and the dendrogram of 5-day intervals confirm the relative similarity of the use of grasslands by the waterbird community from January to July in all years (Fig. 5A, B). However, Jan V (with the influence of the break in hunting in 2000) and to a lesser extent Feb V in 1998 (with a large species richness, see dendrogram) and the 5-day intervals from May I to Jul II showed major differences in waterbird patterns between years. The first axis accounting for 49% of the total variation arranged mainly species assemblages along a seasonal gradient in the use of grasslands (Fig. 5C), from wintering species confined in the negative part of the X-axis to species recorded in spring with positive values. Two groups were very similar between years (groups 1 and 3). Group 1 clumped together some Anseriforme species (Anas crecca, Anas acuta, Anas penelope and Aythya

119

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

Table 1 Species of waterbirds recorded during census of the wet grassland system of the Lac de Grand-Lieu from 1998 to 2000 (scientific and common names)a Scientific name

Common name

Species code

Frequency of occurrence (%)

Mean abundance 1998

1999

2000

n

Podicipediforme species Podiceps cristatus Podiceps nigricollis*

Great crested grebe Black-necked grebe

POCR PONI

91.20–97.05 8.80–47.05

9.202.55 2.851.95

3.851.10 0.450.45

5.25 1.80 0.65 0.85

34 23

Pelecaniforme species Phalacrocorax carbo*

Great cormorant

PHCA

76.45–100

8.853.10

4.951.75

6.45 1.65

34

Ciconiiforme species Ardeola ralloides** Bubulcus ibis Egretta garzetta* Egretta alba** Ardea cinerea Ardea purpurea** Threskiornis aethiopicus Platalea leucorodia**

Squacco heron Cattle egret Little egret Great white egret Grey heron Purple heron Sacred ibis Spoonbill

ARRA BUIB EGGA EGAL ARCI ARPU THAE PLLE

38.25–47.05 79.40–88.25 75.55–88.25 100 100 55.88–61.75 73.55–79.40 5.90–61.75

– 23.056.25 37.7013.40 4.600.65 141.2537.05 7.852.45 41.7512.65 1.200.95

– 27.2011.95 24.108.20 6.050.90 102.1524.95 6.852.85 51.4021.00 0.250.40

– 92.1036.20 48.9517.20 15.403.10 117.9530.00 12.104.80 58.5020.35 3.85 1.35

– 34 32 34 34 21 31 24

Anseriforme species Cygnus olor* Cygnus atratus Anser anser Tadorna tadorna Anas penelope Anas strepera* Anas platyrhynchos Anas acuta Anas clypeata* Anas querquedula* Anas crecca* Aythya ferina* Aythya fuligula*

Mute swan Black Swan Grey lag goose Shelduck Wigeon Gadwall Mallard Pintail Shoveler Garganey Teal Pochard Tufted duck

CYOL CYAL ANAN TATA ANPE ANST ANPL ANAC ANCL ANQU ANCR AYFE AYFU

76.45–100 8.80–64.70 5.90–20.60 11.75–44.10 55.90–61.75 73.55–79.40 94.12–100 44.10–50.00 82.35–88.25 67.35–79.40 20.60–32.35 35.30–52.95 8.80–29.40

11.352.15 – – – 343.00219.00 109.0082.00 64.5013.50 132.1592.00 970.00820.50 27.6512.50 5.856.55 7.055.05 1.701.50

8.352.55 – – – 357.00249.50 120.0083.50 31.559.00 83.5073.50 1006.50730.70 21.158.85 0.850.65 4.353.90 1.001.65

15.254.30 – – – 431.00204.00 151.0096.00 33.007.50 202.50135.00 2077.001403.00 24.9511.00 3.30 2.00 8.35 11.50 5.50 8.95

33 – – – 21 30 34 24 33 27 20 29 25

Gruiforme species Fulica atra

Coot

FUAT

100

1647.50412.80

1809.40456.00

3104.55769.00

34

Charadriiforme species Himantopus himantopus* Charadrius dubius Charadrius hiaticula Pluvialis apricaria* Pluvialis squatarola Vanellus vanellus* Calidris alpina Philomachus pugnax* Limosa limosa* Numenius arquata Numenius phaeopus Tringa erythropus Tringa totanus* Tringa nebularia Tringa ochropus Tringa glareola Actitis hypoleucos* Larus melanocephalus* Larus ridibundus Larus minutus* Larus canus

Black-winged stilt Little ringed plover Ringed plover Golden plover Grey plover Lapwing Dunlin Ruff Black-tailed godwit Curlew Whimbrel Spotted redshank Redshank Greenshank Green sandpiper Wood sandpiper Common sandpiper Mediterranean gull Black-headed gull Little gull Common gull

HIHI CHDU CHHI PLAP PLSQ VAVA CAAL PHPU LILI NUAR NUPH TRER TRTO TRNE TROC TRGL ACHY LAME LARI LAMI LACA

26.45–67.65 29.40–61.75 23.55–50.00 8.80–26.45 14.70–26.45 97.05–100 38.25–52.95 41.20–64.70 41.20–61.75 8.80–17.65 11.75–23.55 20.60–26.45 70.60–76.45 29.40–50.00 5.90–35.30 0–14.70 14.70–32.35 41.20–58.80 100 17.65–47.05 26.45–55.90

3.401.60 5.307.05 – 5.704.70 – 230.80139.30 1.851.15 32.4536.10 39.1522.60 – 6.909.15 – 17.854.40 12.259.00 – – 0.850.80 3.451.10 1122.00203.50 5.758.90 –

1.651.35 4.603.25 – 0.250.25 – 88.8038.70 2.751.55 50.3539.75 14.7515.70 – 3.404.75 – 17.054.15 9.307.05 – – 0.300.25 4.051.20 752.00150.00 0.250.15 –

60.5019.55 18.8013.00 – 0.15 0.15 – 122.1537.20 6.85 2.90 18.7020.20 13.8511.20 – 11.809.35 – 43.7519.60 19.6015.20 – – 2.80 1.95 4.00 2.55 734.00127.50 12.9016.15 –

24 22 – 22 – 34 20 24 29 – 10 – 27 19 – – 17 20 34 27 –

120

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

Table 1 (continued) Scientific name

Common name

Species code

Frequency of occurrence (%)

Mean abundance 1998

Larus argentatus Larus cachinnans Larus fuscus Sterna hirundo* Chlidonias niger** Chlidonias hybridus*

Herring gull Yellow-legged herring gull Lesser black-backed gull Common tern Black tern Whiskered tern

LAAR LACA LAFU STHI CHNI CHHY

64.70–94.10 55.90–67.75 64.70–67.65 5.90–20.60 47.05–52.95 52.95–58.80

3.401.65 – 5.753.65 – 12.454.00 258.8595.40

1999 3.251.95 – 3.801.85 – 13.153.85 81.3068.80

2000 2.25 1.35 – 2.05 0.95 – 14.804.10 417.25129.15

n 34 – 31 – 18 20

a

Four letter acronyms correspond to species codes. Species of high (French list, see Rocamora and Yeatman-Berthelot, 1999; or European ‘Birds’ Directive) or very high (national and European criteria) conservation value are mentioned by one or two asterisks respectively, after the scientific name. Frequency of occurrence (minimum and maximum values for the 3 years) and mean abundance (number of individuals) are presented for the shared period of occurrence for each species (n=number of 5-day intervals). The mean abundance was not calculated for the rare species (– : n<10 individuals for each year). A natural breeding population of sacred ibis Threskiornis aethiopicus expands in the present site from escaped birds since 1993 (Marion and Marion, 1994).

fuligula) only observed in the early months (peaks of abundance near Mar V) and insensitive to the water levels tested (Table 2). Late migrant Charadriiformes species (Numenius phaeopus, Tringa nebularia and Actitis hypoleucos whose peaks of abundance were reached roughly in May I) constituted the second markedly stable group (group 3) and were relatively unaffected by fluctuations in water level. On the other hand Fig. 5C also revealed annual differences related to water conditions for species having intermediate coordinates. The different patterns of water regime between years resulted in the formation of group 6 in 1999 and 2000 from the set of waterbird species forming group 5 in 1998. It combined the majority of non-sedentary breeding Ciconiiforme species foraging in wet grasslands (Bubulcus ibis, Egretta garzetta, Ardea purpurea and Threskiornis aethiopicus) but nesting in the surrounding habitats (floating peat fen) and breeding Charadriiformes in grasslands (Tringa totanus and Chlidonias niger). All these species seemed negatively affected by the spring flooding in 1998 (significant differences in comparisons of mean abundance during phase 2, Table 2) except Chlidonias niger. Platalea leucorodia, Himantopus himantopus and Chlidonias hybridus were added to this group only in 2000 because of an increased number of birds feeding in grasslands this year. In contrast to this latter group, some sedentary species were always noted in group 5 whatever the year (Podiceps cristatus, Phalacrocorax carbo, Egretta alba, Ardea cinerea, Cygnus olor, Anas platyrhynchos, Fulica atra, Vanellus vanellus, Larus ridibundus, Larus cachinnans and Larus argentatus). Most of these species showed a positive relationship with the high water level in phase 2 in 1998 when no real trend was noted in phase 1 (Table 2). Finally the other not previously mentioned waterbird species were not included in the same groups between years first because of some pairwise differences in mean abundance (Anas strepera, Anas clypeata, Anas quequedula, Larus fuscus) and second because of delayed staging for some

migrant species with either no effect of water fluctuations in the use of wet grasslands (Philomachus pugnax, Limosa limosa and Larus melanocephalus) or no trend (Charadrius dubius and Calidris alpina) resulting in group 7 formed in 1999. In the same time comparisons of mean abundance, based on numerical proportions (Table 3), confirmed results on the effect of the water level variations summarized in Table 4. Tests were performed only for waterbird species nesting in the other habitats of the lake but foraging onto the wet grasslands in order to assess the attractiveness of the wet grasslands as feeding areas. Comparisons showed trends similar to those obtained from Table 2: an increasing proportion of the population size affected by the high water level of 1998 (Podiceps cristatus, Phalacrocorax carbo, Ardea cinerea) or by the low water level of 2000 (Egretta alba, Threskiornis aethiopicus and Platalea leucorodia). In the same time no trend was obtained for the other breeding species whose population size has increased during the 3 years (Bubulcus ibis, Egretta garzetta and Ardea purpurea). The second axis of the factorial planes reflected more the distribution pattern of presence duration in the waterbird community. The larger the number of 5-day intervals in which species were counted the more negative the species ordinates were in each assemblage.

4. Discussion 4.1. The avian conservation value of the wet grasslands The biological integrity and ecological value of a wetland can be judged by the occurrence and the number of waterbird species that represent a high nature conservation value (Owen and Black, 1990; van Eerden, 1997). The presence of 26 species of high conservation value out of a total of 52 species (Table 1) showed that the temporarily flooding grasslands of the Lac de

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

Fig. 2. Mean abundance (histograms) and species richness (lines) of waterbirds recorded in each month of the three years. Values (with a 95% confidence interval) in 1998, 1999 and 2000 are indicated respectively by light grey, white and dark grey histograms for the mean abundance and by a solid line with circles, a solid line with squares and a dashed line with diamonds for the species richness.

Fig. 3. Proportion of each taxonomic group in the total abundance counted each month during the 1998–2000 period. Podicipediforme and Pelecaniforme species are excluded since they only accounted for 0.02–0.35% of the total number of waterbirds.

Grand-Lieu are really an important ornithological wetland by supporting numerous wintering, migrating and breeding waterbird species. These wet grasslands constituted a very important feeding area for wintering ducks, breeding Ciconiiformes, and a stop over migrating site for waders, while few waterbirds nest on these marsh grasslands. The March–May period was generally a key phase with the maximum biodiversity (42– 48 spp. from a total of 52 spp.) and the maximum abundance recorded (30–40% of the total bird sightings, peaks of 14 250–22 850 birds according to years). However some differences were noted between years, and 2000 was characterized by many waterbirds foraging in the wet grasslands in most of the months and by a rich and equilibrated community until summer. 4.2. Relationships between patterns of the waterbird community and the management regime The multivariate analysis exhibited a relatively similar seasonal pattern of use of the flooded grasslands by

121

Fig. 4. Mean Shannon diversity (with a 95% confidence interval) of each month monitored from 1998 to 2000.

waterbirds whatever the year. Both migrating behaviour (early vs late species) and breeding behaviour (species with long breeding season vs short late breeding season or regular territorial feeding inducing small dispersed birds vs opportunist social feeding with large flocks of birds) appear to be the main factors determining this pattern of use of the habitats. Nevertheless some changes in bird assemblages, notably the formation of group 6 in 1999 and 2000 from a large set of species, associated with comparisons of mean abundance within and between years, revealed the effect of the management regime (water level and hunting disturbance) of the wet grasslands. Although regional and/or continental-wide factors can influence the fluctuations in abundance recorded among waterbird populations which is particularly true for migratory species whose population sizes depend on survival and reproduction in different geographical areas, sometimes very far apart (Kushlan, 1986b; Duncan et al., 1999), most of the changes in abundance of these species could probably be attributed to hunting and water regime factors. The monitoring during three consecutive years reduces the effect of possible variations in biogeographical population sizes. In any case tests performed on numerical proportions of the breeding population on the lake that fed on the wet grasslands avoid this eventual biogeographical device (Table 3). Moreover habitat selection is known to be strictly related to the availability and accessibility of food (Clausen, 2000) and the presence of safe roosting or breeding sites (Green and Robins, 1993; Guillemain et al., 2000) which are mainly governed by the hydrological and disturbance conditions in wetlands (Ferns et al., 1989). The changes in location of 5-day intervals in January 2000 (Fig. 5B,C with an unusual stop of hunting), corresponding mainly to the presence of large flocks of Fulica atra, showed the weight of the hunting disturbance. An optimal use of feeding resources by ducks was prevented when wet grasslands were insecure during the hunting season (January–February).

122

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

Fig. 5. Results of the Foucart’s Correspondence Analysis on the three-year contingency tables (5-day intervals-species abundance). (A) Plots of samples (5-day intervals) on the first two axes of the compromise table identifying trajectories of the 3 years. (B) Dendrogram of 5-day intervals exhibiting the different levels of clustering. 1998 in dotted line, 1999 and 2000 respectively in black solid plain and bold lines. (C) Coordinates and grouping of the waterbird species on the compromise plot of the two axes (see Table 1 for species codes).

123

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

Table 2 Level of significance of the difference between the abundance of species during the three years and for the two different phases of water regime (respectively Jan V–March VI and Apr I–Jul II) using Wilcoxon’s signed rank testa Comparison of mean abundance Phase 1

Phase 2

1998–1999

1998–2000

1999–2000

1998–1999

1998–2000

1999–2000

Podicipediforme species Podiceps cristatus Podiceps nigricollis

** **

nsb *

ns ns

**

**

ns

Pelecaniforme species Phalacrocorax carbo

ns

ns

ns

**

**

+**

Ciconiiforme species Bubulcus ibis Egretta garzetta Egretta alba Ardea cinerea Ardea purpurea Threskiornis aethiopicus Platalea leucorodia

* * +** ns ns ns ns

+** ns +** ns ns ns ns

+** +* +* ns ns ns +*

+** +* ns * ns +* **

+** +** +** * +** +** +**

+** +** +** ns +** +* +**

Anseriforme species Cygnus olor Anas penelope Anas strepera Anas platyrhynchos Anas acuta Anas clypeata Anas querquedula Anas crecca Aythya ferina Aythya fuligula

ns ns ns ** ns ns * ns ns ns

+** ns ns ** ns +** ns ns ns ns

+** ns ns ns * ns +** +* ns ns

**

**

ns

ns **

ns *

* ns

** ns

* ns

ns ns

ns

+*

+*

Gruiforme species Fulica atra

ns

+**

+**

**

+**

+*

Charadriiforme species Himantopus himantopus Charadrius dubius Pluvialis apricaria Vanellus vanellus Calidris alpina Philomachus pugnax Limosa limosa numenius phaeopus Tringa totanus Tringa nebularia Actitis hypoleucos Larus melanocephalus Larus ridibundus Larus minutus Larus argentatus Larus fuscus Chlidonias niger Chlidonias hybridus

ns +** * ** ns ns ** ns ns ns ns ns * +** ns ns ns ns

+** +* * ** +* ns ** ns ns ns ns ns * ns ns ** ns +*

+** ns ns ns ns ns ns +* +** +** +* ns ns +** ns ns ns +*

ns ns

+** ns

+** +*

ns ns ns ns ns ns ns ns ns ** ns ** ns ns **

+* ns ns ns ns +** ns +* ns * ns * ns ns +*

+** +** ns ns ns +** +* ns ns +* ns ns ns ns +**

a The ‘‘+’’ and ‘‘’’ signs indicate the direction of the difference between two years. No test was performed for some species not recorded during phase 2. b ns, Not significant. * 0.05< P<0.01. ** P40.01.

124

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

Table 3 Level of significance of the difference between the abundance of breeding species nesting outside the wet grasslands but feeding in these habitats (expressed in proportion of the breeding population) during the 3 years and for the second phase of the water regime (Apr I–Jul II) using Wilcoxon’s signed rank test Comparison of mean numerical proportion of breeding species (Phase 2)a

Podiceps cristatus Phalacrocorax carbo Bubulcus ibis Egretta garzetta Egretta alba Ardea purpurea Threskiornis aethiopicus Platalea leucorodia Ardea cinerea

1998–1999

1998–2000

1999–2000

** ** ns ns ns ns ns * **

** ** ns ns +** ns +** +** *

nsb ns ns ns +** ns +** +** ns

a The ‘‘+’’ and ‘‘’’ signs indicate the direction of the difference between two years. b ns, Not significant. * 0.05< P<0.01. ** P40.01.

In January 1998 and 1999, ducks feeding on the hunted wet grasslands represented only 0.21% (25) and 0.16% (27) of the wintering population on the whole wetland, against 10.56% (1941) in January 2000 when hunting was prohibited (unpublished data). In February when hunting occurred, there were only 136, 46 and 227 ducks on the wet grasslands, respectively, in 1998, 1999 and 2000. The largest flocks were observed only and suddenly in March after the end of the hunting season (1302, 2385 and 922 birds, respectively, in 1998, 1999 and 2000 in the first days of the month and then up to 13 736, 10 326 and 14 278 birds in the second half of March), resulting from movements of wintering birds from the other parts of the lake preserved from disturbance (Reserve Nature status). Dehorter and Tamisier (1997) reported similar results in the Camargue, where ducks tended to exploit protected brackish habitats even though they are poorer in food resources than the freshwater habitats which are subject to hunting. The water regime (period, duration and amplitude) induced various responses for numerous species (cf. results summarized in Table 4). Fulica atra (dominating in January and February) and ducks (dominating in March and April) were less sensitive to water level fluctuations than Ciconiiformes and Charadriiformes. During the spring period the water regimes were very contrasted, the mean water level being 30 cm higher from April to mid-June in 1998 than in 2000 and 10 cm higher than in 1999, with a maximum difference in May II (38.5 cm for 1998/1999, 54 cm for 1998/2000 and 15.5 cm for 1999/2000). The overall impact of this rise in water level on waterbirds was an increase in diving fisheating birds such as Podiceps cristatus and Phalacro-

Table 4 Synthesis of significant effects of changes in the management regime (water level) from 1998 to 2000 on the use of the wet grasslands by waterbirds during the phase 2 (April–July)a Waterbird species Late breeding species Bubulcus ibis Egretta garzetta* Egretta alba** Ardea purpurea** Threskiornis aethiopicus Platalea leucorodia** Himantopus himantopus* Tringa tetanus* Chlidonias hybridus*

Effect of changes in the water regime (1998–2000)b + + + + + + + + +

Sedentary, migrant and precocious breeders Podiceps cristatus  Phalacrocorax carbo  Ardea cinerea  Cygnus olor*  Anas platyrhynchos  Anas clypeata*  Aythya farina* + Fulica atra + Vanellus vanellus* + Larus ridibundus  Larus argentatus  a

Species of high (French list, see Rocamora and Yeatman-Berthelot, 1999; or European ‘Birds’ Directive) or very high (national and European criteria) conservation value are mentioned by one or two asterisks respectively, after the scientific name b The ‘‘+’’ and ‘‘’’ signs indicate the direction of the effect.

corax carbo and larger flocks of resting gulls. Conversely these conditions were adverse to numerous ground-feeding waterbirds that require very shallow waters. The pattern of use of flooding grasslands did not always correspond to taxonomy but to ecological responses of well-defined species assemblages. Morphological and behavioural adaptations like the tarsus length for Ciconiiformes or Charadriiformes (DuBowy, 1996; Ntiamoa-Baidu et al., 1998) or the neck length for dabbling ducks (Po¨ysa¨, 1983) and life history strategies like the diet composition and the foraging behaviour for instance for herons (Marion et al., 2000) are the likely mechanisms explaining variations in the use of the wet grasslands by birds according to suitability of feeding habitats. The increased attractiveness of the wet grasslands as feeding habitats in 2000 for numerous Ciconiiformes is all the more true since the population size of these colonial species nesting in trees in the neighbouring floating peat fen has increased between years, notably for Egretta alba and Ardea cinerea (unpublished data). These long-legged herons are known to be favoured by high water levels in wetlands (Marion et al., 2000) and

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

field observations indicated that particularly Egretta alba was often present in deep waters (about 40 cm). In the present study, Tringa totanus and Himantopus himantopus were grouped with the summer migrant breeding Ciconiiformes. At low water levels, wet grasslands provide suitable habitats (Sanders, 1999). These two species and Vanellus vanellus nest in wet grasslands at the Lac de Grand-Lieu contrary to the majority of Charadriiformes recorded and they are more vulnerable than migrant species to the water regime since extensive spring flooding during the breeding season can delay or prevent waders from breeding as it was recorded elsewhere (Self et al., 1994). 4.3. Recommendations for management regime of wet grasslands for the waterbird community Numerous studies have shown large changes in the use of wetlands by bird populations according to severe trends in water conditions, from total drought (Kushlan, 1986a; Catsadorakis et al., 1996; Sanders, 1999) to major flooding such as 2–5 m at Kerkini reservoir (Pyrovetsi and Papastergiadou, 1992; Crivelli et al., 1995). But when modest water level fluctuations occurred, the presence and pattern of many species at a site becomes less predictable (Elphick and Oring, 1998). The present study shows that, in relation to the semi-natural functioning of the ecosystem, it is essential to use appropriate management strategies to ensure and perhaps to enhance the value of the wet grassland system of the Lac de Grand-Lieu as feeding habitat for waterbirds. Any change in lake water regime needs to be monitored carefully in view of the rapid response of many species to the hydrological conditions. The maintenance of a water regime close to that recorded in 2000 seems more favourable to the species of high conservation value (Table 4). A water level around 50 cm during winter until April ensured suitable conditions for large flocks of wintering and migrating Anseriformes. During the following months, the increased duration of a water level of less than 25 cm provided feeding areas for long-legged Ciconiiformes and also for short-legged waders. In this geographical area, the maintain of grasslands flooded until June is very important since numerous inland wet grasslands are exposed and the attractiveness of the wet grasslands of the Lac de Grand-Lieu is likely increased. In the same way, Fasola and Ruiz (1996 and literature cited therein) have shown the importance of rice fields in the Mediterranean Region as valuable replacements for lost natural habitats for a variety of waterbirds because of the long duration of flooding through winter and to a lesser extent as breeding sites. Finally the water regime applied in 2000 for the Lac de Grand-Lieu favours early accessibility for livestock which is essential in maintaining the diversity of habi-

125

tats in wet grasslands (e.g. Colwell and Dodd, 1995), contrary to either high water levels as in 1998 or unflooding grasslands from May in the past (Marion and Marion, 1975). Although the water conditions in 2000 resulted in providing suitable areas for exploitation by the majority of waterbirds, they did not meet the requirements of all species, notably diving fish-eating species (Table 4). But most of these species have less conservation value and above all the wetland is composed of several complementary units and the permanently central flooded area is the most favourable feeding area for these species. The water regime is not the only factor influencing the use of wetlands by waterbirds. The hunting activity, and more generally all types of disturbance likely predominate. The hunting disturbance was specifically studied in numerous studies (e.g. Bell and Fox, 1991; Madsen and Fox, 1995) and a ‘reserve’ effect was often demonstrated (e.g. Hockin et al., 1992). In the present study, the impact of the hunting activity depleted the use of the wet grasslands in January. The total implementation of the European ‘Birds’ and ‘Habitats’ Directives by means of notably the new French law prohibiting hunting in February from 2001 will probably induce in the future important changes in the use of wet grasslands by ducks. This last example emphasizes the need to also define management guidelines at large spatial scales including multi-wetland sites (Haig et al., 1998). Long-term and multi-species monitoring of the use of the wet grasslands of the Lac de Grand-Lieu might elucidate the response of some waterbird species (essentially Charadriiformes) for which there might be a delay before possible effect can be detected.

Acknowledgements This study was supported by grants from the Life Grand-Lieu Program (European Community, Ministe`re de l’Environnement, Agence de l’Eau Loire-Bretagne, Re´gion Pays de la Loire, De´partement de Loire Atlantique), directed by L. Marion for the Socie´te´ Nationale de Protection de la Nature and the Centre National de la Recherche Scientifique. We thank two anonymous referees for comments on a previous draft.

References Bildstein, K.L., Gawlik, D.E., Ferral, D.P., Brisbin, I.L., Wein, G.R., 1994. Wading bird use of established and newly created reactor cooling reservoirs at the Savannah River Site, near Aiken, South Carolina, USA. Hydrobiologia 279/280, 71–82. Bell, D.V., Fox, P.J.A., 1991. Shooting disturbance: a review. In: Matthews, G.V.T. (Ed.), Managing Waterfowl Population. I.W.R.B. Publication 12, pp. 159–171.

126

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127

Blanc, L., Chessel, D., Dole´dec, S., 1998. Etude de la stabilite´ temporelle des structures spatiales par analyses d’une se´rie de tableaux de releve´s faunistiques totalement apparie´s. Bulletin Franc¸ais de la Peˆche et de la Pisciculture 348, 1–21. Briggs, S.V., Thornton, S.A., Lawler, W.G., 1997. Relationships between hydrological control of River Red Gum wetlands and waterbird breeding. Emu 97, 31–42. Briggs, S.V., Lawler, W.G., Thornton, S.A., 1998. Relationships between control of water regimes in River Red Gum wetlands and abundance of waterbirds. Corella 22, 47–55. Catsadorakis, G., Malakou, M., Crivelli, A.J., 1996. The effects of the 1989/1990 drought on the colonial waterbirds nesting at Lake Mikri Prespa, Greece, with special emphasis on Pelicans. Colonial Waterbirds 19, 207–218. Clausen, P., 2000. Modelling water level influence on habitat choice and food availability for Zostera feeding brent geese Branta bernicla in non-tidal areas. Wildlife Biology 6, 75–87. Colwell, M.A., Dodd, S.L., 1995. Waterbird communities and habitat relationships in coastal pastures of Northern California. Conservation Biology 9, 827–834. Crivelli, A.J., Grillas, P., Jerrenthrup, H., Nazirides, T., 1995. Effects on fisheries and waterbirds of raising water levels at Kerkini Reservoir, a Ramsar Site in Northern Greece. Environment Management 19, 431–443. Crowder, A.A., Bristow, J.M., 1988. The future of waterfowl habitats in the Canadian lower Great Lakes wetlands. Journal of Great Lakes Research 14, 115–127. Deceuninck, B., Rufray, X., Jourde, P., 2000. France. In: Heath, M.F., Evans, M.I. (Eds.), Important Bird Areas in Europe: priority sites for conservation. 2: Southern Europe. Birdlife International (BirdLife Conservation Series 8), Cambridge, pp. 157–245. Dehorter, O., Tamisier, A., 1997. Typology of wetlands in the Camargue (France). A necessary framework for waterfowl studies. In: Farago, S., Kerekes, J. (Eds.), Limnology and Waterfowl. Monitoring, Modelling and Management. Wetlands International Publications 43, Slimbridge, pp. 9–17. Dimalexis, A., Pyrovetsi, M., 1997. Effect of water level fluctuations on wading bird foraging habitat use at an irrigation reservoir, Lake Kerkini, Greece. Colonial Waterbirds 20, 244–252. Dister, E., Gomer, D., Obrdlik, P., Petermann, P., Schneider, E., 1990. Water management and ecological perspectives of the Upper Rhine’s floodplains. Regulated Rivers: Research & Management 5, 1–15. Dobrowolski, K.A., 1997. Bird diversity in ecotonal habitats. In: Lachavanne, J-B., Juge, R. (Eds.), Biodiversity in Land-Inland Water Ecotones. UNESCO & The Parthenon Publishing Group, Geneva, pp. 205–221. DuBowy, P.J., 1996. Effects of water levels and weather on wintering herons and egrets. The Southwestern Naturalist 41, 341–347. Duncan, P., Hewison, A.J.M., Houte, S., Rosoux, R., Tournebize, T., Dubs, F., Burel, F., Bretagnolle, V., 1999. Long-term changes in agricultural practices and wildfowling in an internationally important wetland, and their effects on the guild of wintering ducks. Journal of Applied Ecology 36, 11–23. Elmberg, J., Nummi, P., Po¨ysa¨, H., Sjo¨berg, K., 1994. Relationships between species number, lake size and resource diversity in assemblages of breeding waterfowl. Journal of Biogeography 21, 75–84. Elphick, C.S., Oring, L.W., 1998. Winter management of Californian rice fields for waterbirds. Journal of Applied Ecology 35, 95–108. Ens, B.J., Piersma, T., Drent, R.H., 1994. The dependence of waders and waterfowl migrating along the east Atlantic flyway on their coastal food supplies: what is the most profitable research programme? Ophelia 6, 127–151. Fasola, M., Ruiz, X., 1996. The value of rice fields as substitutes for natural wetlands for waterbirds in the Mediterranean Region. Colonial Waterbirds 19, 122–128. Ferns, P.N., Hinsley, S.A., Read, M., 1989. Wintering wildfowl community of a gravel pit lake complex in Southern England. In: Sharitz,

R.R., Gibbons, J.W. (Eds.), Freshwater Wetlands and Wildlife. DOE Symposium Series 61, USDOE Office of Scientific and Technical Information, Oak Ridge, pp. 647–658. Foucart, T., 1978. Sur les suites de tableaux de contingence indexe´s par le temps. Statistique et Analyse des Donne´es 2, 67–84. Gawlick, D.E., Douglas Slack, R., Thomas, J.A., Harpole, D.N., 1998. Long-term trends in population and community measures of colonial-nesting waterbirds in Galveston Bay Estuary. Colonial Watrebirds 21, 143–151. Green, R.E., Robins, M., 1993. The decline of the ornithological importance of the Somerset levels and Moors, England and changes in the management of water levels. Biological Conservation 66, 95–106. Guillemain, M., Houte, S., Fritz, H., 2000. Activities and food resources of wintering teal (Anas crecca) in a diurnal feeding site: a case study in Western France. Revue d’Ecologie (Terre Vie) 55, 171–181. Guillory, V., 1999. Ecolostat software, version B. University of Rennes, France. Haig, S.M., Mehlman, D.W., Oring, L.W., 1998. Avian movements and wetland connectivity in landscape conservation. Conservation Biology 12, 749–758. Hertzman, T., Larsson, T., 1999. Lake Hornborga, Sweden—the return of a bird lake. Wetlands International Publication 50, Wageningen. Hockin, D., Ounsted, M., Gorman, M., Hill, D., Keller, V., Marker, M.A., 1992. Examination of the effects of disturbance on birds with reference to its importance in ecological assessments. Journal of Environmental Management 36, 253–286. Hunter, W.C., Anderson, B.W., Ohmart, R.D., 1987. Avian community structure changes in a mature floodplain forest after extensive flooding. Journal of Wildlife Management 51 495-502. Kingsford, R.T., Johnson, W., 1998. Impact of water diversions on colonially-nesting waterbirds in the Macquarie Marshes of Arid Australia. Colonial Waterbirds 21, 159–170. Kushlan, J.A., 1986a. Responses of wading birds to seasonally fluctuating water levels: strategies and their limits. Colonial Waterbirds 9, 155–162. Kushlan, J.A., 1986b. The management of wetlands for aquatic birds. Colonial Waterbirds 9, 246–248. Madsen, J., 1998a. Experimental refuges for migratory waterfowl in Danish wetlands. I. Baseline assessment of the disturbance effects of recreational activities. Journal of Applied Ecology 35, 386–397. Madsen, J., 1998b. Experimental refuges for migratory waterfowl in Danish wetlands. II. Tests of hunting disturbance effects. Journal of Applied Ecology 35, 398–417. Madsen, J., Fox, A.D., 1995. Impacts of hunting disturbance on waterbirds: a review. Wildlife Biology 1, 193–207. Marion, L., 1989. Territorial feeding and colonial breeding are not mutually exclusive: the case of the Grey Heron (Ardea cinerea). Journal of Animal Ecology 58, 693–710. Marion, L., Marion, P., 1975. Contribution a´ l’e´tude e´cologique du Lac de Grand-Lieu. Socie´te´ des Sciences Naturelles de l’Ouest de la France, Nantes. Marion, L., Marion, P., 1994. Premie`re installation spontane´e d’une colonie d’Ibis sacre´ Threskiornis aethiopicus, au lac de Grand-Lieu. Donne´es pre´liminaires sur la production en jeunes et sur le re´gime alimentaire. Alauda 62 (4), 275–280. Marion, L., Uleaners, P., van Vessen, J., 2000. Herons in Europe. In: Kushlan, J.A., Hafner, H. (Eds.), Heron Conservation. Academic Press, London, pp. 1–31. Marion, L., Clergeau, P., Brient, L., Bertru, G., 1994. The importance of avian-contributed nitrogen (N) and phosphorus (P) to Lac de Grand-Lieu, France. Hydrobiologia 279/280, 133–147. 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 coastal lagoons in Ghana. Ibis 140, 89–103.

J.-M. Paillisson et al. / Biological Conservation 106 (2002) 115–127 Osiejuk, T.S., Kuczynski, L., Jermaczek, A., 1999. The effect of water conditions on breeding communities of pastures, meadows and shrub habitats in the Slonsk reserve, NW Poland. Biologia 54, 207– 214. Owen, M., Black, J.M., 1990. Waterfowl Ecology. Blackie & Chapmann & Hall, New York. Po¨ysa¨, H., 1983. Morphology-mediated niche organization in guild of dabbling ducks. Ornis Scandinavia 14, 317–326. Pyrovetsi, M., Papastergiadou, A., 1992. Biological conservation implications of water level fluctuations in a wetland of international importance: Lake Kerkini, Macedonia, Greece. Environmental Conservation 19, 235–243. Reeber, S., 2000. Impact du rele`vement du niveau d’eau du Lac de Grand-Lieu sur l’avifaune. Re´serve naturelle du Lac de Grand-Lieu. SNPN, Bouaye. Rocamora, G., Yeatman-Berthelot, D., 1999. Oiseaux menace´s et a` surveiller en France. Listes rouges et recherche de priorite´s. Populations. Tendances. Menaces. Conservation. Socie´te´ d’Etudes Ornithologiques de France/Ligue pour la Protection des Oiseaux, Paris.

127

Sanders, M.D., 1999. Effect of changes in water level on numbers of black stilts (Himantopus novaezelandiae) using deltas of Lake Benmore. New Zealand Journal of Zoology 26, 155–163. Self, M., O’Brien, M., Hirons, G., 1994. Hydrological management for waterfowl on RSPB lowland wet grassland reserves. RSPB Conservation Review 8, 45–56. SYSTAT, 1998. Systat 8.0 for Windows: Statistics. SPSS Inc, Chicago. Thioulouse, J., Chessel, D., Dole´dec, S., Olivier, J.-M., 1997. ADE-4: a multivariate analysis and graphical display software. Statistics and Computing 7, 75–83. van Eerden, M.R., 1997. Introduction: patch use, habitat exploitation and carrying capacity for water birds in Dutch freshwater wetlands. In: van Eerden, M.R. (Ed.), Patchwork: Patch Use, Habitat Exploitation and Carrying Capacity for Water Birds in Dutch Freshwater Wetlands. Ministerie van Verkeer en Waterstaat, Lelystad, pp. 15–30. Weller, M.W., 1988. Issues and approaches in assessing cumulative impacts on waterbird habitat in wetlands. Environmental Management 12, 695–701.