Habitat mapping project of the proposed Iraqi Marshlands National Park area

New ecological surveys in support of the creation of the proposed Iraqi Marshlands National Park were undertaken by Nature Iraq in June 2008 at the Central Marshes in southern Iraq. Surveys that occurred in two focal areas Al Chibaish Marsh (10 sites) and Abu Zirig Marsh (two sites) were supported by a preliminary land cover survey in November 2007. Satellite images from 2007 for the Central Marshes were acquired to support creation of maps. Th e “Iraqi Marshlands Habitat Classifi cation System” based on vegetation types has been developed to inventory habitats in these marshlands and to develop a methodology for application elsewhere in Iraq. Six habitat classes (inland running water, river or canal; inland standing water; marsh vegetation; desert; woodlands; and herbaceous vegetation) are included in this classifi cation system, each of which is divided into several subclasses. Th e dominant habitat subclasses in the Central Marshes study area are: (1) rooted submerged vegetation, (2) helophytic vegetation (reed bed or reed mace bed), (3) free-fl oating vegetation, (4) terrestrial vegetation-shrub, (5) unvegetated river or canal, (6) unvegetated desert, and (7) fl ooded communities. Th is paper constitutes a review of the progress in developing this habitat classifi cation system that remains under development.


Introduction
Th e Government of Iraq is currently considering the establishment of a new National Park in a portion of the Central Marshes of southern Iraq.A "Draft Management Plan for the Central Marsh National Park, Iraq" has been developed (New Eden Group 2008).To assist in this planning process, a baseline habitat survey was deemed to be essential.Th us, in 2007 Nature Iraq with the fi nancial support of the Italian Ministry of Environment, Land and Sea initiated a project to identify, survey and map habitats in the area of the proposed park in collaboration with the Iraq Ministry of Environment.Th is habitat project is also related to other projects that Nature Iraq is carrying out (such as the completion and implementation of a Management Plan for the Hawizeh Marsh, Iraq's fi rst Wetland of International Importance designated under the global Ramsar Convention in 2007; Rubec 2008).Th is paper discusses the national park habitat mapping project.
Th e classifi cation of vegetation types is usually achieved through the grouping of similar types of vegetation according to logical criteria (Sayre et al. 2000).One of the fi rst attempts to classify marsh habitats was Warming (1909) in his book "Oecology of Plants" which identifi ed two major types of wetland depending on plant communities: Saline swamp (with halophytic vegetation) and freshwater swamp.Mader (1991, cited in Tiner 1999: 258) emphasized that classifi cation should be: (1) Flexible, general, and of wide geographic applicability in order to allow for the prediction of distribution patterns over a range of environmental situations; (2) professionally credible, preferably through experimental validation; (3) based on concepts that are understandable by non-technical people; (4) logical, consistent, and objectively quantifi able so as to function within in empirical computer-operated information system; and (5) designed and documented so that regular professional staff can, with nominal training, use the system to identify and map fi eld sites.
Adaptation of the European Nature Information System (EUNIS) habitat classifi cation (Davies et al. 2004) was chosen as the model for classifi cation of habitats in this project.Th e EUNIS habitat types are classifi ed hierarchically (Davies et al. 2004).Other habitat classifi cation systems are also organized hierarchically and contain descriptions of the classifi ed units (FGDC 1996, Grossman et al. 1998).Th us, the Nature Iraq team chose to establish Iraq's classifi cation scheme based on vegetation in addition to other criteria and emulate the experience in the references noted here.Th e Italian partners helped in developing the Iraqi project by assisting with training in the use of the EUNIS system.After fi eld testing, modifi cations were undertaken to make it more applicable in Iraq's marshlands.
Vegetation is the focal issue in this habitat study of a key area of Iraq's southern marshlands due to the importance of plants as food and shelter for people and wildlife.Th is is supported by the economic value of many aquatic plants as food or in manufacturing and their role in cleaning water, and because these plants often are indicators of hydrological and environmental conditions at the sites.
In the last few decades at various times (and again in 2008 due to drought), there has been a signifi cant reduction in the water levels in this area of southern Iraq, which has led to a deterioration in water quality and changes in the distribution and status of the biodiversity of the region.Monitoring of the impact of these variations in water conditions has become critical to the marsh restoration eff orts in Iraq.
Th us, one of the goals of this project is to improve the monitoring of the Iraqi marshlands.Remote sensing and satellite imaging technology, it was hoped, should improve the effi ciency of monitoring fi eld trips and reduce associated cost.Groundtruth fi eld data was essential to developing the classifi cation scheme and to map the habitats.Later, it was also felt that the level of eff ort to deliver the overall project could be reduced if the satellite images indicated that there was signifi cant change in the study sites and the overall proposed national park area.To carry out this project, the work was divided into three steps: S(I) Discussions and planning; (II) land cover survey; and (III) description and defi nition of habitats.
Step I included discussions between Nature Iraq, Italian and other international experts about how to carry out the project and what was needed to achieve it.Step I also included the preparation of a work plan, defi nition of needs and the training of staff .
Step II was supported by an initial fi eld trip in November 2007 designed to identify land cover classes in the Central Marsh of the Al Chibaish Marsh area (CM) and the Abu Zirig Marsh area (AZ).For Step III in June 2008, sites were identifi ed as habitats of specifi c species and described according to water quality, sediments, birds, fi sh, benthic macroinvertebrates, zooplankton and phytoplankton, plants and habitat characteristics and their status.Th e "Iraqi Marshlands Habitat Classifi cation System" is gradually being refi ned, but currently remains provisional.
Additional surveys supporting Step III will cover the environmental parameters that can give Nature Iraq an indication of the environmental or economic values of each habitat subclass.Th is information will help decision-makers to prepare plans for ongoing marsh restoration and conservation of those sites that are important from an environmental point view, such as the National Park in the Central Marshes and Ramsar sites in Iraq.

Objectives
Th is project has three objectives: -To survey and obtain specifi c data that can support Nature Iraq projects; -To use standard criteria for describing the status of the marshes in terms of vegetation cover, water quality and biodiversity; and -To facilitate conservation of these sites.

Study area
Th e ground-truth fi eld surveys focused on describing the ecological characteristics and habitat structure at representative sites.All selected sites were within the proposed Central Marshes National Park area and were distributed between the Al Chibaish Marsh area and Abu Zirig Marsh area.Table 1 includes the names, codes and GPS coordinates of each habitat survey site.Th e exploratory fi eld trip conducted in November 2007 was the starting point of the project.By using satellite images from 2006, an initial land cover survey and water quality study was conducted for nine candidate survey sites in these marshes.Th ere is also data from previous surveys in August 2007 for all of Iraq's southern marshlands and their adjacent areas (Abdulhassan 2007).Th e most recent survey was in June 2008 (by using another satellite image from 2007) and some of the results of this work are presented in this paper.Twelve sites were surveyed from the 14 to 18 June 2008, ten of which were in Al Chibaish Marsh area and two were in Abu Zirig Marsh area.

Satellite image processing
Remote sensing has long been identifi ed as a technology capable of supporting the development of habitat maps over large areas.Satellite images contain a information regarding land and water characteristics and the application of digital image process- ing allows for the extracting of data from a digital image very eff ectively.In the work, remote sensing activities allowed for survey of the extension and the distribution of the land cover classes of marshes and to ability to analyze the development of wetland vegetation.
A fi rst map of the Central Marshes was created on the basis of SPOT satellite images acquired in July 2006.Th e pre-processing of SPOT satellite data has included radiometric calibration and atmospheric eff ect correction (dark object subtraction).Image interpretation and analysis of vegetation indices allowed for the spectral analyses of surfaces and the characterization of the diff erent land cover classes.Th en, supervised image classifi cation allowed the creation of detailed land cover maps at the scale of 1:50,000.
In a second phase of the work, ASTER satellite images acquired in July 2007 were processed to obtain updated maps of the study area.Th e same techniques of image pre-processing and supervised classifi cation used for the SPOT images were applied.Th e monitoring survey ground-truthing gave parameters necessary to refi ne and validate the land cover classifi cation obtained from the remote sensing analysis.Th e fi nal products, based on the "Iraqi Marshlands Habitat Classifi cation System", are landcover maps at the scales of 1:50,000 and 1:100,000 (see Fig. 1).It is expected that the project can eventually permit eff ective, low-cost monitoring of these Iraqi marshlands by applying remote sensing and satellite imaging technology.

Th e Habitat Hectare Approach (HHA) for assessing habitat
In order to characterize vegetation classes and subclasses of high conservation priority and to gather quantitative data on species richness, plot studies were used (as suggested by Sayre et al. 2000).Th e number of plots at the site was determined by the range of distinct habitats defi ned in preliminary classifi cations of the sites (usually between one and four habitat types).Th e HHA involves assigning a habitat score to a habitat zone that indicates the quality of the vegetation relative to established benchmarks.Th is habitat score can then be multiplied by the area of the habitat zone (in hectares) to determine the quality and quantity of vegetation (thus calculating "habitat hectares").Th e components are divided into two groups refl ecting an assessment of both "site condition" and "landscape context".Th is is useful for habitat assessment and ground-truthing (DSE 2004).Th e HHA method was applied in the ground-truthing exercises as a methodology to check the classifi cation of the land-cover classes resulting from remote sensing application.Due to the broader complexity of the HHA method, only the determination of vegetation cover from this method was applied within each hectare and without the scoring (Fig. 2).

Vegetation
Plant genera and species were identifi ed using botanical keys (Townsend and Guest 1966, 1968, 1974, 1980a, 1980b, 1985).Th e descriptions of aquatic plants were checked using other Iraq-specifi c references (e.g.Al-Sa'ady and Al-Mayah 1983).Internet botanical resources were also used to confi rm the identifi cation of some plant species (Google Image Search 2008).Species were identifi ed in the fi eld where possible or collected in nylon bags, pressed and transferred to the lab for identifi cation with appropriate botanical keys.Th ese specimens are now preserved in Nature Iraq's herbarium in Sulaimani, Iraq.

Coordinates
Photographic records of fresh specimens were also used to aid identifi cation.Percent vegetation cover for each plant species found at each site was estimated (using the HAA method) and used to calculate the total percentage vegetation cover for the whole site.

Vegetation
Th e dominant vegetation type at each site was used as the basis for identifying the habitat types.Some plant species were common and found at most of the sites while others were restricted to one or two sites, but most of the identifi ed plants are common in these Iraqi marshes.Some species are native to Iraq such as Aeluropus lagopoides that exists near the margins of the marshlands (Townsend and Guest 1968).Hydrilla verticillata is known to be an invasive species in some parts of the world and may be a new exotic species in Iraq as it was not mentioned in (Townsend and Guest 1985).Th ose authors listed only three plant species belonging to Hydrocharitacea of which Hydrilla verticillata was not included.It is possible that it was introduced during the period of great ecological change that occurred with the drainage and later refl ooding of the marshlands of southern Iraq in the 1990s and after 2003.A noteworthy point about these plant communities is that the reed growth probably expanded in the last few years because of the decrease in water level (those species need a water depth of more than 2 m in the open water areas to avoid reed expansion).Th is has led to the decrease of total open water area and the closing off of many water passages due to the expansion of reeds.Table 2 provides a listing of the percentage vegetation cover at each of the 12 survey sites examined in June 2008.

Habitat classifi cation system for the southern Marshlands of Iraq
Classifi cation systems have been developed in order to divide habitats into groups with similar features or functions.Th is is important in Iraq for identifying and describing habitats in order to assess their biodiversity status and habitat functions and then establish conservation plans for Iraq's ecologically important habitats.As in many classifi cation systems, including the EUNIS (Davies et al. 2004), the classifi cation developed for Iraq's habitats is organized hierarchically.It includes a description of the classes and subclasses of habitats.Th is provisional Iraqi system is modeled only partially on the EUNIS system as some of the classes are chosen while others are not because they are not applicable for Iraq's marshlands.However, even the applicable parts were subjected to some modifi cations to make them fi t more readily with the uniqueness of Iraq's marshlands.For example, the class "permanent lake ice" that is used in the EUNIS system, is excluded from Iraqi marshlands classifi cation system because there is no such habitat in the marshes of Iraq.Also, the class "permanent inland saline and brackish lakes, ponds, and pools" was retained but under the subclass "salt water".
It is important to underscore that this is a provisional classifi cation system that will be modifi ed in the future as more knowledge of Iraqi habitats is acquired.At this point, it is divided into three major habitat categories: (1) Water, (2) marsh and (3) terrestrial habitat.Th ese three categories include six classes and each one of them is divided into subclasses and, in some cases, secondary subclasses.Table 3 outlines the provisional "Iraqi Marshlands Habitat Classifi cation System" for the marshes of southern Iraq and their associated surrounding terrestrial habitats.

Habitat Types
Th e proposed "Iraqi Marshlands Habitats Classifi cation System" presented in Table 3 is based on vegetation due to the ecological importance of vegetation communities and because vegetation is a result of the ecological conditions.Table 4 describes the specifi c habitats seen within the study areas along with a basic site description of example study sites.Similar to Site CM-2 (thus an open water area with randomly distributed reed groups).Th ere is a road adjacent to the site from the east and there are small Typha groups (on the east side of the area).Th ere are small areas beside the road where submerged vegetation is absent and the water is deeper than the rest of area.Th e submerged plants are more dense than Site CM-2 but similarly decayed.HAB-CM-10 2.4.2Rooted submerged vegetation 3.1.1.1Helophytic vegetation (reed beds) Th is is Lake is also known as a "Bargah".It is a large open water area with submerged vegetation and surrounded from all directions by reeds (reed beds).Th ere are small groups of reed (known as "Tahala") in the middle of the Bargah.Fishing occurs in the area by nets and electroshock.HAB-CM-11 2.4.2Rooted submerged vegetation 3.1.1.1Helophytic vegetation (reed beds) 2.4.1 Free-fl oating vegetation Th is is considered as an extension to Abu Sobatt canal, which is an inlet to Al Baghdadia Lake (Bargah).Th is canal divides the area into two sides (east and west) but the habitats are the same on both sides of the canal.Th ey have small open water areas with a high density of submerged plants and are surrounded by reeds and Typha from all directions.All submerged plants are decayed on the surface of water.Th e canal is bordered by a line of Typha followed by a line of reeds on both sides.Th is is a water buff alo grazing area.Th ere is extensive fi shing with nets in the moving water of the canal.HAB-CM-12 2.4.2Rooted submerged vegetation 3.1.1.1Helophytic vegetation (reed beds)

Free-fl oating vegetation
Th is open water area (known locally as "Bargah") has submerged vegetation in diff erent densities.It is surrounded on all sides by reeds beds and there are groups of reeds inside the area of the Bargah.Most of the submerged plants are decayed on the water surface.Th is area had been burned before and the ground was brownish and included spots with a low density of submerged plants.HAB-CM-13 2.3 Amphibious communities 2.4.1 Free-fl oating vegetation 3.1.1.1Helophytic vegetation (reed beds) 4.2 Unvegetated desert 6.3 Sparse vegetation A paved road divides this area into two sides: Th e west side is an aquatic habitat with reed beds and a water passage close to the road.Th ere is also an area of high ground to the southwest with terrestrial plants (Tamarix sp.) and aquatic plants (dry Phragmites australis).Th e soil is wet indicating that this is a seasonal marsh.Th e east side includes three types of habitat, (a) t in the northern portion are reed beds and reed mace beds; (b)in the middle area is dry land without plants that is use by the local people; and (c) in the southern portion is terrestrial vegetation.Th is area includes high usage by water buff alo, including breeding activity.Th is is a dry area with a mix of terrestrial plants (to the southwest) and aquatic plants (to the northwest and northeast).Th ere is a small area that still contains some shallow water (5-20 cm depth).Th e area is considered a seasonal marsh.Th ere is a paved road adjacent to the area on the west.Th ere are many people who live along the road and breed water buff alo.Note: Th e siting of this area was determined by two coordinates to the west and the description places it about 1 km toward the east.HAB-CM-26 5.1 Unvegetated desert 3.1.1Helophytic vegetation (reed beds) 4.1 desert shrubs Th is is a dry area (a seasonal marsh) with Phragmites that was dry.It also has terrestrial plants (Tamarix sp. and Suaeda sp.).Th ere is a paved road to the east of area and as one moves northward, the plant cover decreases and the area becomes more desert-like.HAB-CM-27 3.1.1Helophytic vegetation (reed beds) 4.1 desert shrubs Th is is a dry site (seasonally wet) that is located to the west of a soil embankment that extends beside the area from north to south.Th e entire area is covered by dry reeds with low density intermixed with terrestrial vegetation (Tamarix sp. and Suaeda sp.).HAB-CM-28 2.4.2Rooted submerged vegetation 3.1.1.2Helophytic vegetation (reed mace beds) 4.1 desert shrubs 2.4.1 Free-fl oating vegetation Th e main part of this area is open water with rooted submerged vegetation and surrounded by reed mace beds (Typha domingensis) from the east and west.From the north there are reed beds.From the south, there is a small canal and road.Th ere are small soil embankments to the southeast of the area.Th e area is used for water buff alo grazing.

Abu Zirig (AZ)
HAB-AZ-1 1.1 Unvegetated river and canal 2.4.2Rooted submerged vegetation 3.1.1.1Helophytic vegetation (reed beds) 2.4.1 Free-fl oating vegetation Th e major habitat here is reed beds and there are small open water areas inside the reed beds.Th is area is adjacent to the road on the south and to a soil embankment of the river that is adjacent to the area and has openings that feed the marsh with water from the river on the west side of the marsh.Th ere are date palm trees on the soil embankment.Th is area is considered a water buff alo grazing area; local people cut and collect the reeds for water buff alo feeding and manufacturing of goods.Th e area is close to Al Fuhood City.HAB-AZ-2 1.1 Unvegetated river and canal 2.4.2Rooted submerged vegetation 3.1.1.1Helophytic vegetation (reed beds) 2.4.1 Free-fl oating vegetation Th is is a water passage (canal) with a depth of about 2 meters and width of about 25-30 meters, bordered on both sides by reeds that achieve heights of about 2-3 meters above the water surface.Th e canal extends from north to south.Th ere are areas close to the reeds with dense and decayed submerged plants, and there is a narrow area in the middle of an open, moving water area devoid of plants and deeper than the rest of canal.Th is area is used for breeding by some birds on the submerged plants (the tops of these plants have emerged above the water surface due to the decreasing water level).

Conclusions and recommendations
Th e study area and its 12 survey sites in the Central Marsh of Iraq, were set up to assess the application in a real life situation of the Habitat Hectare Assessment (HHA) methodology and to develop a practical classifi cation system for Iraqi habitats based on anticipated habitat classes that had been previously observed.A provisional hierarchical classifi cation system, the "Iraqi Marshlands Habitat Classifi cation System", was created to facilitate mapping these habitats into distinct units.Six habitat classes have been identifi ed, each of which is divided into several subclasses.However, the people of the local communities in these marshes use specifi c terms to describe habitat types.Two examples are "bargah" which means a pond or lake with unvegetated standing water habitats, and "sibil" which means inland running water/river or canal.Eff orts are being made to associate these local names with the scientifi c categories that have been identifi ed so the classifi cation system and maps becomes practical for local use.Th is habitat classifi cation system is still undergoing development and will be subjected to further rigorous review.
Th e author urges consideration of the following recommendations that directly consider plant diversity and the health of the habitats of the southern marshes of Iraq: It is recommended that additional habitat survey work in other areas in Iraq be initiated in order to verify the applicability of the provisional classifi cation system and its methodology.
In 2008, the Nature Iraq team observed that, in this area, water levels are still decreasing in addition to having on-going poor water quality conditions.To return to a better state comparable to pre-1990 conditions (e.g.before drying of these marshes) for both nature and the local people of the southern marshes of Iraq, it is recommended that Iraq should maximize the restoration of as much of the ecological character of this area as possible.
Application of a standardized habitat classifi cation is essential to monitoring of the progress of this restoration eff ort.It is recommended that the provisional habitat classifi cation system discussed in this paper be completed and put into operational use as soon as feasible.
Th e habitat types of greatest importance in restoration eff orts will be those that bring stability and economic value to the local communities.Th e area strongly needs a permanent increase in the quantity of useable water to help restore habitats that are critical to water buff alo for example, a mainstay of the local marsh people.Th us, it is recommended that Iraq make all eff orts to foster the restoration of its freshwater sources that once fl owed from the Euphrates River and the Tigris River.
Restoration of healthy habitats also requires improvement in water quality.It is further recommended that the Government of Iraq should establish wastewater treatment plants at the points of discharge in the cities that are located on the inlets of these marshes and help to restore the hydrological regime.Increases of water quality and water levels can likely assist in limiting the growth of reeds (such as Phragmites australis) and help restore the ecological character of these marshes.
Finally, it is also recommended that stakeholders initiate training programs to reduce human disturbance and engage the local community in these restoration eff orts through education and awareness building about ecosystem health and the importance of particular habitats.

Figure 1 .
Figure 1.Satellite-based land-cover classifi cation of the Central Marsh (Al Chibaish and Abu Zirig) showing the selected survey sites (circled areas).

Figure 2 .
Figure 2. Application of the Habitat Hectare Approach (HHA) method; fi eld data sheet for describing habitats (codes A-F indicate habitat classes).

Table 2 .
Vegetation cover (%) of each of 12 survey sites.

Table 4 .
Th e habitat classes identifi ed at each site with site description.