Responses Of Birds And Their Community To Eco-environment And Bird Habitat Restoration In A Phragmites-dominated Tidal Marsh Ecosystem

The reed(Phragmites australis)-dominated tidal marshes at the Yangtze River Estuary are important buffer zone between the water and the land, as they can not only reduce erosion and protect the nearby dike, but also provide other ecosystem services, such as protecting biodiversity, water conservation, nutrition sources to estuarine and coastal waters. On the one hand, the tidal marsh wetland has suffered decrease in area and degradation in ecosystem functions because of reclamations and disturbance from the near society, and on the other hand, the nearby society needs the tidal marsh to provide critical ecosystem service, such as serving as habitat of wildlife (given that the wetlands of Yangtze River estuary are important staging habitat of shorebird following the Australian-Asia flyway). A solution to this conflict is important for sustainable development of the Yangtze River delta.In this thesis, I have performed field-experiments and observations to reveal the responses of birds and avian community to ecological restoration and bird habitat restoration in Phragmites-dominated tidal marshes of the Chongxi Wetland Research Center. We have performed ecological restoration project first and then have selected and performed two treatments as habitat restoration:construction of forested wetland in reed-dominated tidal marsh by inducing trees that may adapt wetland area (CFW) and reallocation of habitat elements in reed-dominated tidal marsh by eradicating reeds, changing topography, and constructed water area (RHE). I studied the distribution, diets, nest materials and feeding habitat selection of reed habitat specialist-reed parrotbill, Paradoxornis heudei. Through changes of feeding behavior, distribution, and nest site selection of reed parrotbill, use of perch site by the common cuckoo, and micro-habitat use of egret, I studied the responses of birds inhabitating in tidal marsh to the bird habitat restoration. Through comparing the structure, diversity and guilds of bird communities between restored tidal marshes and control ones, I studied the responses of bird community to the ecological restoration and bird habitat restoration conducted at the Phragmites-dominated tidal marsh. Field works have been performed in 2005 to 2007 and main results are below.1. changes of living component to the ecological restoration and bird habitat restorationConstruction of forested wetland in reedbed reduced the illumination in the area, thus casued shading to the reed shoots. In the first year, in area with high density of forest, the density of reed shoots decreased, the height of reed shoots increased, and the number of internodes decreased. The mortality of reed shoots inseased with the density of forest. In the second year after forested, the reed shoots dispeared in area with high density. In patched forested area, the trees and the reed shoots coexisted (Chapter 3). In forested reedbed, the abundance of the insects on reed shoots decreased and dispeared when the reed shoots dispeared (Chapter 5). It seems the introduction of low density forest or patches of trees can reduce the potential negative effects to the reedbed and insects on them.Through reallocation of habitat elements in reed-dominated tidal marsh, we conducted bird habitat restoration. One year after reallocation of habitat elements, the diversity of plant increased. The height of plant community decreased. The plant density and biomass decreased (chapter 4).2. Response of Reed Parrotbill to ecological restorationThrough reviewing papers and field observations, I tried to demonstrate whether the Reed Parrotbill is a reed habitat specialist (Chapter 6). The results showed that in its distribution range in China,83.3% of sites where Reed Parrotbills had been witnessed were reed vegetation or habitat with Common Reeds. In Chongming Island, Reed Parrotbills were observed only in habitats with reeds, such as fish ponds, woodland and tidal marshes. This habitat was either reed-dominated habitat or a habitat with reed patches and near those dominated by reeds. In the local scale, during three-year continuous monitoring, Reed Parrotbills has not been observed in habitats without reeds. The density of Reed Parrotbill was significant higher in reedbeds than in reeds with dense trees and patches of reeds. In the habitat of reeds with dense trees, Reed Parrotbills have not been seen in this habitat since the reeds disappeard in the second year. These results indicate that Reed Parrotbill strongly depend on reed vegetation as its habitat and the Reed Parrotbill is a reed-habitat specialist.I compared the characteristics, material composition and morphology of nests of Oriental Great Reed Warblers (OGW) and Reed Parrotbills (RP), and related these parameters with their flight- and foraging-related morphologies (Chapter 7). The diversity of nest material was higher for OGW than for RP (2006, P<0.05,2007, p=0.07), indicating a narrower resource rage used by RP than OGW.88% of RP's nest material was tissues of the Common Reed and 81% of OGW's nest material was tissues of the Wild Rice. RPs used mainly tissues from living reeds, which need a strong beak to get, while OGWs used mainly dead plant tissues, which are easy to get, but need flying a distance to transport. The RP had stronger beak than OGW, while OGW had strong flight ability than RP, which is adapted well to each other's nest material usage.Through neck ligatures, we collected and compared food composition delivered to nestling by OGWs and RPs (Chapter 8). RPs used mainly insect from the order Hemiptera, while OGWs used mainly insects from the order Diptera. There was a significant difference in food composition between the OGW and the RP. The diversity of food was higher for OGWs than RPs, indicating a relative narrow food resource range of RP.99% of insects delivered to nestling by RP were wingless, while 84% of those by OGWs were winged. The wingless insects mainly hid within reed shoots, which was easier to target but hard to retrieve. A strong beak is needed. The winged insects often flied in the air, which need relative strong flight ability to pursue and get them. The food composition of RPs reflected a strong beak but weak flight ability, while that of OGWs reflected a weak beak but strong flight ability.I studied the foraging habitat selection of RPs by collected reed samples to determine foraging marks, food resource abundance, and characteristics of reed shoots (Chapter 9). The scale insects were the most abundant food resource and the stem-boring caterpillars were the least abundant one. The RPs foraged mainly on scale insects and stem-boring caterpillars and preferred to scale insects more. The RPs foraged on scale insects in areas with short reed shoots, while they foraged on stem-boring caterpillars in areas with dense reed shoots. The amount of insects attacked by RPs was depend on the density of insects, with higher density higher attacked probability. There was a significantly negative relationship between the amount of attacked scale insects and the height of reed shoots. There was a significant positive relationship between the attacked stem-boring caterpillars and the density of newly emerged reed shoots. The predation intensity of each insect significantly positively related to the abundance of each insect. These results indicated that the food resource availability was important factors influenced the foraging habitat selection of RPs. 3. Response of bird species to habitat restorationThrough continuous monitoring the distribution of Reed Parrotbills (RP) and the Vinous-throated Parrotbills (VP) during reed-harvest, I studied the instant responses of RP and VP to reed-harvest (Chapter 10). RPs responded to the reed-harvest fast. RPs reduced their activities in areas where reed-harvest was going on. One week after reed-harvest, RPs came back to the harvested area close to bigger area of reed shoots and used the reed patches as foraging habitat. Four weeks after reed-harvest, RPs redistributed in all reed-harvest area with reed patches. But the group size and density were smaller than those before reed-harvest. The VP did not show obviously signs affected by the reed-harvest during and after harvesting. These results indicated that habitat fragmentation influenced less on a habitat generalist species (VP) than on a habitat specialist (RP), while the RP also had a relative plasticity to adapt to habitat fragmentation, though slower than a habitat generalist.I studied the impacts of reed cutting on foraging of Reed Parrotbill from January to April 2007 (Chapter 11). Reed cutting decreased potential food resource of Reed Parrotbill significantly. Reed cutting changed the foraging habitat of Reed Parrotbill, and shifted their foraging place from dead reed shoots to growing reed shoots before they suited Reed Parrotbill to feed on them. Short time after reed cutting, the residual dead reed shoots have an important role for foraging of Reed Parrotbill. And these indicated that keeping some reed patches uncut could facilitate Reed Parrotbill foraging on growing reed shoots and lessen food shortage impacts of reed cutting.Construction of forested wetland reduced the density of reed shoots and also the abundance of insects on reed shoots (see Chapter 3 and 5). The foraging intensity by Reed Parrotbills of insects on reed shoots in reedbed and forested wetland was compared (Chapter 12). The amount of insects that was preyed by RPs decreased significantly after forested wetland had been constructed shortly. This indicated that construction of forested wetland affected the foraging distribution of RPs.I systematically searched reedbed, reedbed with dense trees, reed patches and protection forest for nests of the RP to study the effects of habitat restoration on nest site selection of RPs (Chapter 13).117 nests were found from 2006 to 2008.99% of these nests were found in reed vegetation and 86% was in reedbed. Only one nest was found in reedbed with dense trees. The nest density was significant higher in reedbeds than in reed patches. There were no significant difference in reed- and food-related characteristics between nest site and site without nest. These results indicated that, first, RPs were heavily depend on reed vegetation as their nesting habitat, second, construction of forested wetland and reallocation of habitat elements in reed-dominated tidal marsh could reduce the quality of these areas as RPs'nesting habitat.The distribution and group size of RPs in 2006 to 2008 among reedbed, reed patches, reedbed with dense trees and reedbed with sparse trees was compared (Chapter 14). The density of RPs was significant higher in reedbed than in all other habitats. Since the reed shoots in reedbed with dense trees disappeared in 2007, RPs has not been seen in this habitat. There was RPs in reedbed with sparse trees in all three years, and the density in reedbed with sparse trees was not significantly different with that in reed patches. The frequencies of different group size (1-2 inds./group, 3-7inds./group and>7inds./group) in reedbed, reedbed with dense trees and reedbed with sparse trees were similar. These results indicated that RPs responded to the reed vegetation fragmentation induced by construction of forested wetland and reallocation of habitat elements in their distribution and this response was consistent with distribution in nest site and foraging intensity.Construction of forested wetland can increase perch site for some birds. I studied the response of Common Cuckoo to a simulation of perch site increase by erecting bamboo pole in reedbeds (Chapter 15). The abundance and emerging frequency of Common Cuckoo increased with increased perch sites. Common Cuckoo showed more frequency in perching behavior in area with increased perch sites than in area without and more flying frequency in areas without perch sites. The nest parasitism rates of OGW by Common Cuckoos increased in area with perch sites. These results demonstrated that construction of forested wetland could increase perch sites, thus increase abundance of the Common Cuckoo and also change the behavior of the Common Cuckoo.Observations were conducted to determine the use of microhabitat (waterline, waters, naked area and perch site simulated by bunch of reed shoots) in area with reallocation of habitat elements (RHE) by Litter Egret (Chapter 16). Little Egrets showed more frequency of rest behavior in RHE area and waterline was more frequently used than other microhabitats. 45% of foraging happened near waterline and 40% of resting in bunch of reed shoots. The abundance of Litter Egret in RHE areas was higher when there was more in tidal flat nearby. While the tide covered the tidal flat nearby, there was also no egret in RHE area. This indicated that usage of RHE area as habitats by egrets depended on the abundance of egrets in a larger scale.4. Responses of avian community to ecological restoration and habitat restorationWith BACI (before-after-control-impact), the avian community in winter was compared between reedbed and forested wetland (Chapter 17). After construction of forested wetland (with dense trees), during one year, the density and species richness of the whole area decreased and the diversity increased. The density, species richness and diversity index of wintering birds depending on reedbed decreased. Within the forested wetland, the density decreased and the diversity increased. These results indicated that construction of forested wetland could show negative effects on avian community depending on habitat before habitat restoration; also provide opportunity for new avian community.In CFW area, the avian community had the highest species richness while in RHE area, the species richness of whole community and wetland birds were similar with those in reedbeds and tidal flat. The area with CFW and RHE had higher species richness in whole community and wetland bird community than vicinal area, while the abundance was similar to the vicinal area. The abundances of migrants, residents and wintering birds belonging to wetland birds were higher than vicinal area (Chapter 18).There was the highest species richness and number of response guilds in CFW area. In RHE area, there was the highest diversity index in species and guilds. The species and abundance of obligate wetland species were higher in RHE, reedbed and tidal flat than in CFW, While there almost the same proportion of obligate wetland species and upland species, indicating the RHE area was both good habitat for wetland birds and land birds (chapter 19).