Distributions and Community Composition of Birds in Iraq ’ s Central Marsh

The Central Marsh (CM) in southern Iraq is known to provide important habitats for both resident and migrant birds.The CM has been used extensively by humans, in part due to its high levels of productivity and biodiversity. It was drained in the 1990s by the government and reflooded and restored in 2003. Recent brief surveys of the CM from 2005 to 2010 recorded 94 bird species. Our study combined transects and point counts in detailed monthly surveys from October 2013 to June 2014 in the CM. We found a total of 125 bird species in the CM across all surveys, with 31 species recorded for the first time in the CM and 11 species categorised as red listed by the IUCN. Fourteen species were confirmed breeding in the CM. Cluster analysis using NMDS ordination showed that the study area can be divided into three main clusters of bird assemblages which are presented here. We provide management recommendations based on our findings.


Introduction
The Iraqi Central Marsh (CM) is a globally important open water and freshwater marsh [1] located between three provinces (Missan: 31 ∘ 10  N, 47 ∘ 05E; Thi-Qar: 30 ∘ 50  N to 31 ∘ 30  N; and Basra: 46 ∘ 45  E to 46 ∘ 25  E) in the south of Iraq [2].The CM is almost 300,000 ha in area and is part of a larger marshland complex [3].The CM has always been used extensively by humans, in part due to its high levels of productivity and biodiversity [4].The CM has long been known to provide important permanent habitat for large number of birds and is part of a flyway for thousands more migrating between Siberia and Africa [5,6].Eighty bird species were found in the CM in the last complete census in the 1970s [6] and the area was identified as one of 42 Iraqi Important Bird and Biodiversity Areas (IBAs) [3].Despite its importance for both people and wildlife, the CM was totally drained in the 1990s by the government, which caused huge levels of biodiversity loss and the disappearance of nearly all bird species from the area [7].The motivation for this drainage is not clear, with some sources stating political motivation [1] and others agricultural expansion, which had already begun in the 1970s [8].
Parts of the CM were reflooded and restored in 2003 using the River Euphrates water to feed the CM directly, resulting in a huge reverse migration of both local people and bird species [9,10].Concomitantly, there was a change in governmental attitudes and NGOs toward wildlife and the value of birds, especially regarding the maintenance of healthy ecosystem functioning [11].Reflecting this change, 141,615 ha (47%) of the CM was declared as the country's first national park (NP) in 2013 (Mesopotamian National Park or MNP) [12,13] and the CM was identified as one of 82 Iraqi Key Biodiversity Areas (KBAs), based largely on bird data collected between 2005 and 2008 [2].While the difficult political situation in Iraq made it impossible to conduct bird surveys between 1980 and 2003 [1], the KBA assessment showed that bird species in the CM have begun to recover to predrainage levels (94 species recorded [4]).However, many of these surveys were rapid (visiting the area for one day in the season) (seasonal surveys were conducted between 2005 and 2010 by Nature Iraq (NI) and the Iraqi Ministry of Environment) and therefore need updating to better understand the conservation statuses of CM birds.Furthermore, the KBA assessment did not explicitly examine the distributions or habitat associations of birds within the CM, treating the whole area as one conservation unit [4,10].Treating the whole of the CM in this way could hinder attempts to restore the CM's bird populations, as many of the significant threats that face birds in the CM are not distributed equally in space [7].Here we present data from surveys across a nine-month period in 2013-2014.
Our study had three objectives: (i) to provide an inventory of bird species occurring in the CM; (ii) to investigate the spatial and temporal patterns of bird abundance and diversity across the CM; (iii) to describe bird assemblages within different areas of the CM.We use our findings to suggest management actions in the CM to benefit birds.

Study Site and Sampling
Protocol.Three longitudinal water transects (each 30 km in length) were chosen inside the CM to identify and count birds in the marsh from both sides of transects [14].These three transects ran approximately north-south and followed existing water courses, thus causing only minor disturbance to habitats whilst surveying (Figure 1).The area has historically been supplied with water from the north by the Tigress River.The Euphrates River crosses Chibayish City from the West to the East towards Modina City in Basra province [11].The Euphrates River has been closed using soil embankments between Chibayish City and Modina City due to the scarcity of water in the river.Therefore, all water coming from the West of Iraq in the Euphrates River now goes directly to the CM.
To aid management, we subdivided the study area into three zones.We did this based on the dominant type of vegetation and similarities in the type of human activity that occurred in each zone (e.g., fishing, reed cutting, and the intensity of water buffalo used for distributing water buffalo milk).This classification was made by visual inspection of the CM and was descriptive only (based on qualitative impressions made during the survey work).Zone one started from the south in the Euphrates River and crossed Chibayish City, with zones two and three extending north inside the national park (Figure 1).Zone one had the most human activity, grazing by water buffalo, and dominance of the plant species Typha domingensis.Zone two had intermediary levels of human activity and water buffalo grazing with Typha domingensis and Phragmites australis the dominant plant species, and zone three had the least amount of human activity and grazing with Phragmites australis the dominant plant species.
Nine surveys were carried out in the CM from October 2013 to June 2014 (for exact dates see Appendix).A motor canoe was for transport, with surveys starting from the riverside in the south of the national park and finishing in the north.Three days were spent in the area for each survey (one day/transect), and six-eight hours per day were spent moving along each transect.All field observations were conducted by the same observer and were started from the river in zone one in the morning and finished in the afternoon at the end of zone three (05:30-12:30/13:30 in the summer, and 07:00-14:00/15:00 in the winter).The stop points (point counts) are shown in Figure 1; time spent at each stop was 30-40 minutes (the survey included water and sediment sampling; therefore, almost 15-20 minutes were spent for bird records in each point).Our chosen survey time for bird counts was designed to maximise our chances of detecting as many cryptic species as possible, while minimising our chances of doublecounting.Additionally, our sampling design was constrained by safety considerations and logistical difficulties, which, although making it impossible to conduct sampling in zone three in the morning, reduced our risk of double-counting further.A Canon 7D camera with Sigma lens 135 × 400 and Canon lens 100 × 400 and 8 × 42 binoculars were used to observe and count birds in the CM.No observations and counts were undertaken on the way back, apart from new species that were not recorded in the main survey.A Garmin GPS device was used to map the three transects digitally.We used official data from the Iraqi Ministry of Water Resources, Chibayish Branch, to record the monthly water level in the Euphrates River during the survey period.

Analysis of Species Richness and Abundance and Community Composition.
To investigate the role of management zone, month (coded as Julian day), and transect in species richness and abundance, we used general linear models with a Gaussian error structure as our models with a Poisson, quasi-Poisson, and negative binomial error structure were severely overdispersed.To produce acceptable model residual plots, species abundance was square-root transformed but species richness was left untransformed.For each dependent variable, we constructed a full model and used a multimodel inference approach to define the relative importance of each independent variable, as recommended by Burnham and Anderson (2002) [15] and Grueber et al. (2011) [16].We used the package "MuMIn" [17] to produce all possible candidate models, which were ranked by AICc.We then used model averaging across the full set of candidate models to produce parameter estimates and measures of the relative importance of each parameter [15].
Nonmetric Multidimensional Scaling (NMDS) ordination was conducted using the FactoMine R package [15] in R to identify whether there were differences in community composition between different management zones.To do this, nine different sites were identified based on the point where each of the three transects intersected each of the three management zones (see Figure 1).The rationale behind splitting up the zones further for this analysis was in response to preliminary analysis of our field observations, which suggested that there were distinct bird communities at finer scales than the three management zones that we originally identified.Then, a Bray-Curtis dissimilarity matrix and dendrogram were created to identify clusters of sites in the CM that were most similar in their bird species' communities (based on both species' identities and abundance).These clusters were then overlaid on the results of the ordination to help identify parts of the CM with similar bird communities.We chose to use both a dendrogram and NMDS ordination to identify clusters to ensure that our results were robust.All data manipulation and statistical analysis were undertaken in R version 3.1.3(R Development Core Team 2015).

Which Bird Species Occur and Breed in the CM and Which
Species Are of Conservation Concern?  1 for detailed information about breeding status and more details in the Appendix).

Important
Bird Species for Conservation.The survey found 11 species that are important conservation priorities in the CM and are red listed by the IUCN (Table 2), and two endemic species (Basra Reed Warbler and Iraq Babbler), and three near endemic species (Black Francolin, Little Grebe, and Mesopotamian Crow) according to [18].In addition, there are two species that are regionally threatened (Pygmy Cormorant and Sacred Ibis) and four species that are regionally near threatened (Grey Heron, Purple Heron, Western Marsh Harrier, and Common Kingfisher) in the Arabian Peninsula [19].

Spatial Distribution of Bird Abundance and Diversity
across the CM.According to the raw data from the study survey, the greatest bird species abundance was in winter (December, January, and February).The areas of the CM with the greatest abundance of birds was along transect two and in zone two.The highest species richness was in January and April along transect one, and in zone two (Figures   the highest overall species richness ( = 9.37, SE = 1.62); however, species abundance did not appear to significantly decline from January to December ( = −0.01,SE = 0.01).

What Bird Assemblages Exist within Different Zones in the CM?
The dendrogram based on the Bray-Curtis dissimilarity matrix showed that there were at least two main clusters of bird species, with potentially a third cluster in site nine.These results were aligned with the site cluster analysis undertaken using NMDS ordination (the linear fit from a stress plot was  2 = 0.98), which shows that the study area can be categorised into three main clusters of bird species (cluster one = T1-Z1, T1-Z2, T2-Z1, and T3-Z1; cluster 2 = T1-Z3, T2-Z2, T2-Z3, T3-Z2, and T3-Z3; Figures 5 and 6

Discussion
We recorded 125 bird species in the CM over nine months of survey work, which increased the species list for the area by 24% and we confirmed the breeding of 14 species and coded 20 species as species observed in breeding season in suitable nesting habitat (H according to British Trust for Ornithology breeding status codes).Earlier work over a five-year period recorded 94 species using less intensive survey efforts [4].Fifty-one breeding bird species were reported by Salim et al. [20] and 77 species were recorded by Salim and Porter [21] in the Iraqi southern marshland complex where the CM is found.In comparison to the wider region at large, 357 bird species were recorded in wetland conservation areas found in the Arabian Peninsula, Iraq, Syria, and Lebanon, covering an area of 3,000,000 km 2 [22].
Our survey recorded 31 more bird species versus the most recent, rapid assessment which is potentially due to the nature of the previous surveys, which were rapid assessments of bird species undertaken across the entirety of the Mesopotamian marshlands (Hawiza Marsh to the east of the Tigris River, Hammar Marsh to the south of the Euphrates River, and the Central Marsh between the Tigris and Euphrates Rivers) [4].By focusing on one of the marshes, we were able to conduct more intensive surveys, which detected more bird species.For this reason, there is a clear need to undertake further, intensive surveys in the other two Mesopotamian marshes to provide accurate, up-to-date information.Recent national water and biodiversity strategies have indicated that it is no longer appropriate to consider all three marshes as one conservation unit, as the connections between the water bodies have severely declined due to extreme water scarcity [10].As such, there is an urgent need to develop management plans for each of the three Mesopotamian marshes.
The greatest bird species abundance in the CM was in winter (December, January, and February).The area of the International Journal of Biodiversity     CM with the greatest abundance of birds was along transect two and in zone two.The highest species richness was in January and April along transect one and in zone two.However, there is potential of seasonal and observer bias.In winter, larger birds often form flocks in open areas and are therefore easy to detect.In summer many breeding birds may be less easy to detect (i.e., hidden from view in the reeds).
Similarly in spring and autumn there will be huge numbers of migrant passerines using the marshes as a "stop-over" site to rest and feed.Our analysis is useful for comparing abundance and richness between areas within the same season but is perhaps less useful for comparing between seasons.
There has been recent interest in the potential population size of one of the most threatened species we recorded, the globally endangered Basra Reed Warbler, with major discrepancies of opinion regarding breeding population size (e.g., between [18,19] and [23]).We recorded 11 nests in our study area (10 of them were independent nests; one in April, four in May, and five in June 2014).All recorded nests of Basra Reed Warbler in our surveys were built on reed stems only.Basra Reed Warbler shares its preferred reed bed habitat with the similar Great Reed Warbler.Our survey highlighted the dominance of Great Reed Warbler in April (the number of individuals recorded of Basra Reed Warbler in the area of study was 14, 42, and 66 and for Great Reed Warbler 38, 57, and 42 in April, May, and June 2014, resp.).It is unsafe to extrapolate from the numbers of nesting Basra Reed Warblers that we detected in our survey to the whole site, but clearly the Iraqi Marshlands holds a substantial population of this globally endangered species.
Although the CM is the first national park to be designated in Iraq [11] there is still a lack of detailed information on bird distribution across the site or in current management plans [4].The CM area was considered as a Key Biodiversity Area (KBA) and an Important Birds Areas site and was divided into two main areas: core and buffer zones [4,7,13].
Our study provides more detailed information about bird abundance, richness, and assemblages.Three bird clusters were suggested for the first time in the area with zone two identified as containing large numbers of breeding birds.In contrast zone one had fewer species and a lower overall abundance; due to the proximity of human settlements zone one is also likely to receive more human disturbance (fishing, reed cutting, and buffalo grazing) than the other zones.More analysis is needed for the data to highlight differences between the 3 clusters.
Water levels in the Euphrates River, the main source of the CM's water, varied between autumn, winter, and summer (1.29 m, 1.79 m, and 1.66 m in October 2013 and January and June 2014, resp.; see Table 3), which is the main source of control of water levels and water quality inside the CM [11].Rising water levels in the winter expand the flooded area away from Chibayish City and provide more suitable habitats for water birds in zone three.In addition, rising water levels in the Euphrates enable easier access for local people in zone one.Thus, keeping the minimum water level at 1.29 m in summer and winter could help and support wildlife in the CM, especially in zone two and zone three.
Our results clearly show that the CM provides habitat for many bird species and that more intensive survey methods are needed for other two marshes found within the Mesopotamian marshland complex.Owing to national water scarcity, there is a real danger that the Mesopotamian marshlands will shrink in size, reducing the effective area of suitable habitat for many wetland species.We provide detailed information on the bird species found within the International Journal of Biodiversity

2 InternationalFigure 1 :
Figure 1: Area of study in the CM showing locations of transect surveys (T1: transect 1, T2: transect 2, and T3: transect 3) and the location of the CM within the Middle East within the overview map (as shown in orange).

Figure 2 :
Figure 2: Bird species abundance (all species observations summed) in the Iraqi CM (monthly surveys from October 2013 to June 2014).(a) shows abundance for the whole CM, (b) shows abundance by transect, and (c) shows abundance by zone.

Figure 3 :
Figure 3: Bird species richness in the Iraqi CM (monthly surveys from October 2013 to June 2014).(a) shows species richness for the whole CM, (b) shows species richness by transect, and (c) shows species richness by zone.

Figure 4 :
Figure 4: Bird species richness in the Iraqi CM (monthly surveys from October 2013 to June 2014).(a) and (b) show species abundance for the three transects and the three zones in the CM, and (c) and (d) show species richness by the three transects and the three zones.

Figure 5 :Figure 6 :
Figure 5: NMDS cluster analysis for birds in the CM.(a) shows a dendrogram based on a Bray-Curtis dissimilarity matrix.This dendrogram was then cut to give three groupings based on similarities in bird community composition as shown both by the dendrogram and by the NMDS ordination plot.(b)shows the results of the NMDS ordination, with each zone/transect complement (e.g., zone one transect one is site one and zone one transect two is site two) coloured according to grouping and surrounded by a convex hull.The ordination shows three different groupings of sites (note: the inclusion of site three within cluster one does not represent a failing with our method but a consequence of the way NMDS plots are visualized).

B a s r aR e e dW a r b l e
o u s -T a i l e dS c r u b -R o b i

Table 1 :
Bird breeding statuses in the CM for April-June 2014.Key: T = transect and Z = zone.

Table 2 :
Threatened bird species recorded in the CM by this study (details of numbers recorded are presented in the Appendix).

Table 3 :
Maximum and minimum monthly water levels in the Euphrates River (Chibayish City station) from October 2013 to June 2014 (official data of the Iraqi Ministry of Water Resources-Chibayish Branch).

Table 4 :
Species observations and counts (October, 2013-June, 2014) in the Iraqi Central Marsh.Scientific names are provided in Appendix Table7.

Table 5 :
Field notes and breeding evidence of birds from Oct, 2013, to Jun, 2014, in the Iraqi Central Marsh.Wide distribution of Pied Kingfisher, Cattle Egret (especially on the back of the water buffalos), Crested Lark, House Sparrow (ii) Recording Greylag Goose as first arrival in this month to the CM (iii) Recording Red-Backed Shrike (we just recorded this species in Oct and May) (iv) Recording Water Rail just in this month Wide distribution of Pied Kingfisher, Cattle Egret (especially on the back of the water buffalos), Crested Lark, House Sparrow (ii) Recording Starling, Coot, Yellow Wagtail, Robin, Arminian Gull, Ferruginous Duck, Bluethroat as first arrival in this month to the CM (iii) Large number of Dead Sea Sparrow, Winchat comparing with the observations of same species in the 9 surveys (iv) Recording Little and Spotted Crake (iv) First arrivals of huge number of water birds and ducks to the area of study (e.g., Teal, Northern Pintail, Mallard, Shoveler).(v) Increasing of Common Kingfisher, White-Breasted Kingfisher, White Wagtail, Marsh Harrier, Kentish Plover, Starling, Slender Bill and Black-Headed Gulls, Coot, Chiffchaff (vi) First arrival of White Stork and Pelican White (vii) First observation of Spanish Sparrow, Marbled Teal, Common Kestrel, Glossy Ibis of the 9 surveys (viii) Increasing of birds hunting

Table 6 :
List of survey dates in the Central Marsh.

Table 7 :
The status of birds in the Central Marshes, southern Iraq.Survey from Oct, 2013-Jun, 2014 Key: IUCN Red List category (ver.3.1.);Status in Iraq (Salim et al. 2012); Months recorded; highest count; Status in the CM based on the study survey; Breeding status in CM with BTO breeding code.