DISCUSSION
Effects of artificial ecological islands on macroinvertebrates
The macroinvertebrate community in the park was mainly composed of Mollusca, Palaemon and Lnsecta, possibly because the plant community was mainly composed of Phragmites australis ,Typha orientalis and Myriophyllum . The substrate was mostly silt that was rich in humus, which meets the requirements of some species for dissolved oxygen and organic debris. The park’s artificially controlled water flow keeps constant, providing excellent conditions for slow-moving species, such as mollusks, to thrive (Zuo et al. 2016; Chen et al. 2014). With the regulation of farmland around the wetland park, it is forbidden to dump waste water into the park, which makes the water quality of the park relatively cleaner, and hydrophyte growth is thick. The abundances of Palaemon (Exopalaemon) modestus , Ephemerptera, Lestidae, Dytiscidae and other aquatic insects such as Nepidae, Belostomatidae and Haliplidae that are suitable for living in aquatic plants also increased. The macroinvertebrate species were abundant and dense, but the biomass was low, which was related to their geographical environment and the short amount of time that had passed since the farmland was converted to wetland. Regarding the increasing trend in the macroinvertebrate diversity, consistent with the works of Du and Lu, the diversity index showed an increasing trend with the extension of time since the construction of conservation engineering projects (Du et al. 2011; Lu et al. 2013). Over time, the ecological effects of these projects become evident.
We found evidence that the islands could affect the composition of macroinvertebrates. First, on the one hand, the islands changed the water depth, causing the islands to have a certain slope. The light and temperature at the bottom of the islands also changed, further influencing the distribution of sediment, primary productivity and hydrophytes and increasing the niche range of the plant species. At the same time, island creation increased the living area, provided diverse habitats for feeding and reproduction and created concealment conditions for macroinvertebrates. The living requirements of different species were met due to the different water levels, water temperatures and light levels (Wen et al. 2016; Carvalho et al. 2012; Freitas et al. 2011). On the other hand, most species live on the sediment surface, which is rich in oxygen and organic matter, and changes in water depth affect the amount of organic carbon, phytoplankton and sedimentary organic matter (Chen et al. 1975), which is conducive to the adjustment of community structure (Zuo et al. 2016; Chen et al. 2014; Du et al. 2011). Therefore, the establishment of islands increased the water depth gradient diversity, which significantly changed the community structure and distribution pattern of the macrobenthos.
Second, it is generally believed that macroinvertebrate species richness differs in different bottom environments (Lu et al. 2007); the community in pebble substrates was richer than that in fine sand substrates (Xie et al. 2007). Therefore, when building the PSI, the park’s designers chose larger stones as the covering, rather than fine sand or small stones, which ensured the stability of the matrix to a certain extent. In this study, the SSIs were surrounded with many hydrophytes, such asPhragmites australis , Typha and floating grass, and the bottom had loose debris that was rich in organic matter; these conditions provided a place for the macroinvertebrates to eat, breed and evade predators and made the substrate stable, reducing the impact of water level changes on the macroinvertebrates (Duan et al. 2007). Therefore, compared with the PSIs, with less vegetation and humus, the SSIs are more suitable for survival.
Third, the biomasses of the dominant species, such as Typha andPhragmites australis , on the island increase each year, and as a result, only single species of hydrophytes exist. Phragmites australis are shallow-rooted scattered plants that have a strong ability to secrete oxygen from their roots, and these conditions can meet the respiratory needs of mollusks, such as Gastropoda, which require high dissolved oxygen levels. However, due to their high growth density, the stems and leaves do not decompose easily (Zuo et al. 2016), and the amount of organic detritus they produce is relatively small; therefore, the abundance and diversity of all macroinvertebrate species were not as high as those on the islands that were constructed relatively later.
Finally, plants are more likely to survive on the soil island (which has a high abundance), but as the soil island ages, the plants tend to become more homogeneous, while the stone island has been scoured by water for a long time, which makes it less stony and more conducive to plant survival. For example, SIMPER analysis showed that the three most contributing species on the island, which was built in 2011, werePhragmites australis (50.63%), Scutellaria scordifolia(16.01%) and Inula japonica (8.38%). However, the top three contributing plants on the soil island built in 2013 were Carex bohemica (20.62%), Calamagrostis epigeios (19.15%) andPhragmites australis (15.93%). Due to mowing a year ago, soil island J has the highest plant biodiversity, species and abundance compared with other soil islands. The number of macroinvertebrate species was negatively correlated with plant abundance (r = -0.689,P = 0.04), as well as, ageing of islands leads to a loss of attractiveness for plant and birds (Scarton et al. 2013). Therefore, it is recommended that wetland parks conduct reed-cutting work regularly to promote the increasing diversification of hydrophytes.
Effects of artificial ecological islands on waterfowl
In addition to providing habitat and increasing the abundance of macroinvertebrates, artificial islands are also important breeding grounds for waterfowl. Macroinvertebrates provide an important food source for waterfowl, especially during the breeding season when birds lay eggs and broods. This may be due to the particularly high demand for protein during egg development of waterfowl (Joyner 1980; Murkin and Kadlek 1986).
Patra described the interactions occurring among macrophytes, macroinvertebrates and waterfowl in freshwater wetlands as a complex interdependency (Patra. et al., 2010). The low-water areas of the islands may have increased the temperature of the water and/or the light penetrating the water column, thereby promoting the growth of aquatic plants and increasing the food source for birds.(van den Berg et al., 1997;Zimmer et al., 2000).The increased density of aquatic vegetation will provide a carbohydrate-rich food source for waterfowl, which is important for the gathering and migration of waterfowl in the fall (Baschuk 2010; Baldassare and Bolen 2006). At the same time, the higher density of aquatic vegetation will also increase the number of habitats available for invertebrates, which may increase the abundance of invertebrates and further increase the abundance of bird food (van den Berg et al. 1997; Hornung and Foote 2006). The dynamic changes of bird communities were indirectly influenced by macroinvertebrates through the impact of vegetation decomposition on vegetation habitats (Wilson 1990;Schneider et al., 1981;Backwell et al., 1998). Macroinvertebrates indirectly influence the dynamic change of bird community by decomposing vegetation to change vegetation habitat. (Wilson 1990; Schneider et al. 1981; Backwell et al. 1998; Patra. et al., 2010). In the present study, changes in the aquatic substrate conditions directly impacted the macroinvertebrates and aquatic macrophytes and initiated taxonomic changes in the waterfowl assemblages.
The structure and composition of the vegetation within the wetlands also appear to influence the distribution of waterbirds and their use of the wetlands (Desrochers and Ankney 1986; Rehm and Baldassarre 2007).Vegetation species such as Typha spp. and Phragmites australis are preferred by walking marsh birds because they provide dense cover and residual vegetation that allows the birds to move along the water surface (Baschuk 2010). For example, Botaurus commonly forage along the vegetation/water interface, concealing themselves in the vegetation and ambushing passing prey in the open water (Lor 2007; Rehm and Baldassarre 2007). The construction of artificial islands has created more vegetation/water edge that have increased the number of available foraging sites, reducing interspecies competition at these sites and allowing birds to proliferate in the wetlands.
Vegetation species such as Typh and Phragmites australis can reach up to 2.5 m in height in summer, providing deep, over-water nesting habitat for waterfowl species such as some of Anatidae, as well as marsh birds such as Podicipedidae and Fulica atra (Welling et al. 1988; Murkin et al 1997). The interspersion of emergent vegetation provides concealment during foraging. In addition, Cyperaceae on islands provide dry, upland habitat for nesting, which may increase the amount of upland nesting sites (Duebbert and Lokemon 1977, Swanson and Duebbert 1989). A large number of aquatic plants provide runways for birds to fly (Hua et al. 2009) and also play an important role in providing visual isolation between waterfowl breeding pairs during the breeding season. However, Typha was avoided by dabblers and divers, perhaps due to high stem densities and large amounts of residual litter. Dense emergent vegetation can hinder the movement of waterfowl and may also hinder their entry. Therefore, waterfowl may have avoided Typha as it did not provide favourable cover for thermal protection or nesting (Baschuk 2010). To sum up, it is suggested that the park should regularly control and manage the vegetation on the island.
Habitat selection by birds is strongly related to the food distribution, water depth and food availability. Water depth is the most important factor that limits the use of waterfowl habitat and affects the composition of nesting and thermal cover vegetation. Water depth limits the feeding behavior and energy consumption of waterbirds, which affects the availability and availability of food and determines whether the habitat can be used (Ma et al. 2010, Murkin et al. 1997). For example, the length of the waders’ beaks and legs limits the gate’s range in the shallows. (Nolet et al. 2002). Waterbirds with the same water depth and the same feeding habits need to reduce spatial niche overlap by utilizing habitats with different water depths, while the construction of artificial islands increases the types of available habitats (Zhang et al. 2014). Most waterfowl used the shallow water habitats most; Ardeidae, Charadriidae and Scolopacidae, due to their morphological characteristics and foraging strategies, were limited to feeding in the shallow water habitats, such as shallow water areas at the edges of islands that were less than 20 cm deep (Shao et al. 2016). It may also be related to the higher frequency of food (such as hydrophytes, zooplankton, fish and other invertebrates) in shallow waters (Xia et al. 2010). Anatidae, Podicipedidae and Rallidae generally feed on seeds, fish, and other foods, mostly use deep water areas. The preference for deeper water by the Aythya ferina andPodiceps cristatus was expected as these species require deep water to allow mobility during foraging and escape (Baschuk et al. 2012). During the whole observation period, the open water area changed within a small range, and the water depth was maintained at approximately 2-3.5 m. Because Fulica atra individuals that did not breed during the year also clustered on the water surface, the observed numbers remained high. The density of wading birds was negatively correlated with the water depth, while the density of swimming birds was positively correlated with the water depth (Baschuk et al. 2012). The change in water depth is an important factor for the formation of niche differentiation and the stable coexistence of waterbird communities (Shao et al. 2016). However, in this study area, the deepest water depth was only approximately 3.5 m. Due to topographic characteristics, the water depth did not change significantly, and the difference in the distribution of the bird habitats was not obvious.
The habitat selection by the waterbirds showed a ”nesting pattern” in the spatial gradient. Wader’s quantity of foraging was strongly related to water depth: if there was not enough water or the substrate was too dry, it was not easy to dig for food; furthermore, if the water level was too shallow, there were fewer large fish and more small fish, and the deep water area had more fish, but they were difficult to obtain (Zhang et al. 2014). For example, universal species (Fulica atraand Anatidae) usually choose deep water habitats, while obligate species (Charadriidae and Scolopacidae) choose shallow water habitats. The shallow water areas of artificial islands can solve this problem, providing more diverse water depth gradients for different birds to forage and inhabit, making them more competitive. Shallow water at the edge of the island with water depths of 10-20 cm can provide a series of water gradients that attract relatively more birds and promote bird diversity and richness (Colwell et al. 2000). An increase in surface area can increase the diversity of swimming birds, and water depth variation affects the diversity of waders (Hua et al. 2009). Island construction changed the original single topography of the wide open water and then changed the water depth, that is, the bird diversity increased by increasing the water depth gradient and increasing the available habitats for obligate species. By increasing the water depth heterogeneity of the habitat, the micro-habitat diversity increased to accommodate waterbirds with different ecological niches.
Whether a habitat is suitable for bird migration and reproduction determines the distribution level of birds. During the construction of water level management, the park provides a mosaic of deep and shallow wetlands, staggers the water level of the wetland complex, and creates a diverse range of habitats available in relatively close proximity. Artificial islands could be used as shallow wetland habitat to promote the use by dabblers, whereas the open-water area could be used as deep water habitat to promote the use by divers and marsh birds, thus creating a diverse wetland habitat. Wetland diversity will provide wide range of available habitats to waterfowl and help and promote avian biodiversity. The wetland park can play a positive role in protecting the important passage of bird migration in northeast Asia. Unfortunately, the number of visitors to the wetland park has been increasing for three years, which has disturbed the normal habitat of waterbirds. Therefore, the number and species of waterbirds in this study showed a downward trend. We also recommend that wetland parks separate tourist areas from bird breeding areas to reduce disturbance to birds. This study was conducted only in wetlands in China, and we did not study whether the increase in bird species was due to individual migration from adjacent sites or other factors. We also did not investigate whether the increase in birds came from other adjacent wetlands or was due to increased breeding rates. These need further study in the future. But the results clearly show that the island is useful in increasing the biodiversity of waterfowl and macroinvertebrates. It is not difficult to predict that over time; the ecological benefits of artificial wetland islands will become more and more prominent, which can provide reference for wetland restoration work around the world.