The Effect of Three Gorge Project on the Small Mammals in Yangtze River of China

China has the richest hydro resources on the planet; hence, the development of hydropower is of great importance to alleviate the energy crisis and environmental pollution resulting from this country’s rapid economic growth during the 21^st century [1-2]. The Three Gorge Project (TGP) represents one of the largest hydropower-complex and flood control projects in the world and has been ranked as the key project for the improvement and development of the Yangtze River. The project has economic and social benefits, such as preventing floods and associated disasters, adjusting water quantity, facilitating agricultural irrigation and generating energy. Although, the project has brought tremendous economic and social benefits also posed many adverse impacts on the eco-environment and society. Dams/Reservoirs pose many environmental and ecological challenges and changing the temporal pattern of river discharge to downstream and ultimately the ocean [3].

Increasing numbers of scholars had assessed the effect of dam construction on river ecosystems, including water quality control, water and sediment regulation, biodiversity conservation in downstream areas and reservoirs [4-19]. In parallel, the impact of the TGP on the Yangtze River system had been the subject of much controversy since its launch [18,20]. Many studies had been conducted before and after the construction of the Three Gorges Reservoir (TGR) to determine its actual impact on the surrounding ecosystem [21-32].

There are some species of small mammals (mainly rodents) almost in all habitats. Small mammals, especially rodents, as the herbivores, and also the prey base for many predators, are an important link in food webs and energy flow in ecosystem. They play critical roles in ecosystem functioning, which constitute the primary link between primary producers and secondary consumers. Small mammals are also considered to be ecological indicators, because they are very sensitive to habitat changes. There have been some examples of ecological effects of river regulation on small mammals [33,34]. So the small mammal communities might be affected by the Three Gorge Dam (TGD) both in TGR area and downstream area of TGD along Yangtze River. Specially, small mammals were gotten the attention of the whole world for eruptive population of Yangtze vole (Microtis fortis) in Dongting Lake region in 2007 [35]. Although, there had been limited studies on the small mammal community, there were several study on the area of GTR, Dongting Lake and Poyang Lake area, which could reflect some general phenomena. Here, we conclude small mammal assemblage responses to changes induced by the hydroelectric dam in Yangtze River.

Surveillance surveys in 19 townships of five points (Chongqing, Fengdu, Wanzhou, Fengjie and Yichang) in the TGRA near the Yangtze River from 1997 to 2012 showed that the dominant indoor species were Rattus norvegicus , Mus musculus and Rattus tanezumi, while Insectivora, R. norvegicus and Apodemus agrariuswere the major species captured outdoors [36]. In order to prevent out breaks of rodent-borne diseases, sanitary clearance of the bottom of the reservoir and deratization were performed before every impoundment during the construction of the TGRA. Large-scale deratizations were carried out three times: in 2003 (when the Three Gorges reservoir first began storing water to 135m sea level), 2006 (storing water to 156m sea level) and 2008 (storing water to above 170m sea level). Surveillance data showed that a general decrease in both indoor and outdoor rodent densities [which was showed by relative population abundance, and was indicated by trap success, and calculated as the percentage of success in 100 traps: D = (100N/T) × 100%, where D is the relative population abundance, N is the number of animals caught by all traps and T is all traps collected] and sharp decreased after impoundment compared with before impoundment. After each deratization effort, the rodent density decreased sharply: 3.72% before 2003 to 0.15% in 2003; 1.98% before 2006 to 0.46% in 2006; and 3.41% before 2008 to 0.54% in 2008 [36].

Basing the surveys of Chang, et al. [36], the average indoor rodent density was 2.94%, with the highest and lowest recorded densities in 1997 (6.75%) and 2003 (1.37%), respectively; the average outdoor rodent density was 3.02%, highest in 1997 (7.19%) and lowest in 2006 (1.90%), which showed an overall decrease in the indoor and outdoor rodent density in the TGRA over time [36]. The deratization before impoundment had kept the low density in TGRA, which were reported also by other papers [37-42]. In total, the densities in TGRA did not increase following impoundment, although the small mammal population would migrate from inundated areas to regions of higher altitude due to riding water levels following impoundment.

On a larger scale, the compositions of the dominant species did not alter after impoundment, although the indoor and outdoor densities of different key rodent species changed differently over time [36]. Whereas, in some place of TGRA, such as Yichang, both the species and the predominant species of small mammal changed before and after impoundment [38,39]. As for the small mammal species composition, the indoor and outdoor dominant species before impoundment were R. norvegicus and those of the genus Sorex, accounting for 71.07% and 42.11%, respectively. The dominant species after impoundment, both indoor and outdoor, were R. tanezumi , accounting for 41.67% and 42.86%, respectively [39]. The result of the survey of rodent population by he in the upper of the dam also showed that the dominant species was R. tanezumi (28.24%) and the following were A. agrarius (26.47%), M. musculus (20.00%) and Rattus losea(14.71%) [43]. These observation might due to the tolerance of R. tanezumi and A. agrarius to drugs used for deratization and habitat change [44,45], for example the newly built houses for local immigration in the reservoir region after relocation, with harder floors are not suitable for ground-burrowing rodents such as R. norvegicus .

With full operation of hydroelectric facilities, as a result, some mountains had already become islands in the TGRA. A number of permanent and seasonal islands had been formed. After habitat insularization, the ecosystem functioning might be highly disrupted on small islands [46]. The investigation of the species composition and distribution of rodent populations at two geographic locations in the TGRA showed that the species richness of island and mainland rodent communities was not significantly different at either site, but the dominant species differed significantly between island and mainland habitats: the total rodent density on the islands was significantly higher than that of mainland areas [29]. Stable isotope analysis revealed that the dietary composition of rodents was more diverse on islands than at nearby mainland sites. Moreover, the island populations had greater overlap in food sources than did the mainland populations, suggesting more intense competition for food in the newly insularized habitats. Thus, habitat fragmentation due to the construction of the TGD might substantially increase intra- and inter specific competition among local rodent populations, leading to further changes in species composition and biodiversity [29].

Because Anourosorex squamipes utilized aquatic food sources, it had become the most abundant species on some island in TGRA. As water levels continued to rise and island area decreased, the population of A. squamipes might have an even greater advantage of food sources [29]. Some studies have already revealed that aquatic resources could serve as a complementary food source to fuel high reproduction among small mammal populations on islands [47].

In hydro-fluctuation belt of TGR, small mammals might move into the drawdown area of the reservoir [48]. The average capture rate was low and there was various species of rodent in hydro?fluctuation belt. The density and species composition of small mammals varied with different regions, altitude and monitoring time [42]. According surveillance by Liu, et al [48] in June 2010?A. agrarius was the predominant species(80.65%), R. tanezumi , A. squamipes and R. norvegicus were also found; in September, R. norvegicus and A. squamipes were the predominant species?During 2010-2014, the captured small mammals were classified into six species: A. agrarius, A. squamipes, R. norvegicus , R. tanezumi , M. musculus and R. edwardsi. A. agrarius was the predominant species accounting for 46.55% and A. squamipes was next (25.86%) [42].

Poyang Lake and Dongting Lake are the two largest freshwater lakes of the Yangtze River floodplain in China. Both lakes directly connect with the Yangtze River and represent the most frequently flooded areas in China. The region has asymmetrical annual precipitation, with 70~90% of total annual precipitation occurring during the rainy season, from May to October (with the greatest concentration of rain falling between May and July). For example, the annual water level of the Dongting Lake changes by as much as 15m, rising in summer and falling in winter. Both lakes extensively exchange water with the river, with the water level substantially elevating following the monsoon season, which is termed as a “winter-land, summer-water” ecohydrological condition that only favors a few rodents (such as M. fortis in Dongting Lake) that immigrate to the seasonal lake beaches.

Following the onset of TGR operation in 2003, the flow of the middle and lower reaches of the Yangtze River was altered, which in turn changed the exposure time (i.e., period not covered by water) of the beaches. During the water storage periods of the TGR, the water level of Dongting Lake decreased by 2.03m in 2006 and 2.11m in 2009 at the outlet of the lake, with extreme decreases of up to 3.30m and 3.02m, respectively [49]. Based on the regular pattern of beach emergence periods during the dry season, in parallel to simulated water level data for the lake after TGD construction, Zou, et al., predicted that the period of low and medium level beach emergence in Dongting Lake would be significantly lengthened once the TGP was operational [50]. Furthermore, for the mid- and long-term periods (30 and 50 years) after the completion of the project, the emergence period of the entire beach is expected to gradually lengthen.

These changes in water level inevitably induced alterations to the inundation patterns of the wetlands of lakes, which, in turn, disturbed the ecological function of the lake wetlands as habitats for plant and animal communities. Therefore, the manipulation of flow regulation by the TGR might also influence the small mammal communities inhabiting the beaches of rivers and lakes in the middle reaches of the Yangtze River, downstream of the TGR. To date, a few small mammal studies have been conducted, with a focus on the small mammals inhabiting the beach areas in the Dongting Lake region and the Poyang Lake region [35, 51-55].

Historical data for Dongting Lake showed that, before the TGP, extensive flooding in summer made the lake beaches unstable habitats for small mammal populations, other than M. fortis [54], with studies suggesting that the lake beach represented suitable living and breeding habitat for the vole living [56,57]. It was predicted that the regulation of water flow by the TGR might reduce water levels in fall and increase the amount of time that M. fortis had access to its preferred beach habitats, which would, consequently, result in larger vole populations and greater damage to surrounding crops [51,52]. The outbreak of the vole population in the Dongting Lake region during 2007 seemed to reasonably, support this forecast [35]. However, the percentage represented by the vole in species compositions has declined compared to that before dam operation, due to the immigration of other species, particularly the striped field mouse (A. agrarius) and the Norway rat (R. norvegicus ), to beaches that were previously uninhabitable because of seasonally high water levels in the Dongting Lake region [54]. Out of the eight species captured on the beaches after the onset of TGR operation, only three (M. fortis, R. norvegicus and M. minutus) were recorded inhabiting the beaches during the 1990s (Table 1). Before the TGP, M. fortiswas the only confirmed dominant species that inhabited the beaches of Dongting Lake. In comparison, once water was impounded by the TGR, A. agrarius and R. norvegicus became abundant (with marked population growth) [54]. In other words, the increase of A. agrarius and R. norvegicus in the species composition of beach habitats has caused the ratio of M. fortis to decline accordingly. Therefore, small mammals that are strongly affected by flooding would now benefit from the reduced extent and duration of water inundation because of water regulation management by TGR operations. The regulation of water discharged by the TGR had caused an increase in the species diversity (Diversity index included the dominant concentration, the Shannon-Weiner diversity index, the evenness index and so on) of the small mammal community on the beaches, which is becoming increasingly similar (Similarity coefficients included the Sorenon index and the Whittaker index) to that on farmlands adjacent to the lake [54].

Period	Species composition (%)*
	Microtus fortis	Apodemus agrarius	Rattus norvegicus	Micromys minutus	Niviventer fulvescens	Rattus nitidus	Suncus murinus	Erina ceuseuropaeus
1990s	99.68 (622)	0.00	0.16 (1)	0.16 (1)	0.00	0.00	0.00	0.00
2003–2006	81.77  (314)	15.89 (61)	0.52  (2)	0.00	0.00	0.00	1.82 (7)	0.00
2007–2008	79.36 (919)	17.70  (205)	1.47 (17)	0.17  (2)	0.09 (1)	0.00	1.12  (13)	0.09  (1)
2009–2012	51.00  (559)	42.61 (467)	4.29  (47)	0.46 (5)	0.09 (1)	0.09 (1)	1.46 (16)	0.00
Total	74.00  (2414)	22.47 (733)	2.05 (67)	0.25 (8)	0.06 (2)	0.03 (1)	1.10 (36)	0.03 (1)

Although no studies of small mammal had not been published on the beaches surrounding Poyang Lake, based on the changes to the structure of the small mammal community described for Dongting Lake, Dai, et al., compared the communities of small mammal in beaches and farmland areas [55]. A similar situation has been documented around Poyang Lake following the TGP (Table 2). The numbers of small mammal species captured both on farmland and beaches was low (five species on farmland and four species on beaches). A. agrarius and R. losea were the dominant species on lakeside farmlands in the Poyang Lake region. Investigation of Poyang Lake showed that the species abundance and diversity indexes of the small mammal community on the beaches were similar to those on farm lands [55]. A. agrarius, R. losea and R. norvegicus were also found universally on the beach, suggesting that the farmland community has expanded to the beach habitat. Based on the similar structuring of the small mammal communities on the farmland and beach habitats of the Dongting Lake and Poyang Lake regions, the potential impact of the TGP on downstream lakes should be incorporated into future ecological impact assessments.

Habitats	Species composition (%) #
	A. agrarius	R. losea	R. norvegicus	M. musculus	M. minutus	S. murinus
Beach	66.01 (101)	28.76 (44)	3.92 (6)	0.00	0.00	1.31 (2)
Paddy field	46.73 (157)	50.30 (169)	2.08 (7)	0.60 (2)	0.30 (1)	0.00
Dry field	80.65 (25)	19.35 (6)	0.00	0.00	0.00	0.00
Total farmland	49.59 (182)	47.68 (175)	1.91 (7)	0.54 (2)	0.27 (1)	0.00
Total	54.42 (283)	42.12 (219)	2.50 (13)	0.38 (2)	0.19 (1)	2 (0.38)

Studies from the Dongting Lake and Poyang Lake showed more species of the small mammal immigrated to the beaches of lakes during the dry season. When flooding would normally occur, small mammals occupying beach habitats would be forced to aggregate on the dikes and migrate to farmland areas [58], with high population numbers potentially causing severe damage/losses to crops, as well as presenting a health hazard to humans. A. agrarius and R. norvegicus in Dongting Lake or A. agrarius and R. losea in Poyang Lake were found on all beaches after the onset of TGR water impoundment. Therefore, it is important to monitor fluctuations in the populations of small mammal to identify possible population outbreaks. In addition, shifts in the succession of the small mammal community structure on the beaches should be monitored, which may indicate potential trends leading to the outbreak of particular species.

Based on the index of diversity, the small mammal community on the beaches had already undergone major changes with a tendency toward diversification. TGD water regulation had created a transitional small mammal community, characterized by the encroachment of species from areas surrounding the lake into beach habitat. However, it remains unclear as to whether the observed changes will facilitate the control of rodent damage or whether the introduction of more species will generate greater damage to agricultural crops and human communities. Therefore, it is important to conduct studies on interspecies interactions.

Although most of the world’s river systems are regulated by humans to some extent, information about the effects of river regulation on small mammals remains limited. In fact, speculation about how river regulation affects mammal communities is more common than actual empirical data [33]. This paper presents some studies of how a dam, the TGP, affects a small mammal community. The TGP represents one of the world’s largest hydraulic projects and, as such, has been the subject of much controversy. After decades of planning and 17 years of construction, the project has demonstrated comprehensive benefits with respect to various issues, including flood control, power generation and navigation. In parallel, various environmental and ecological issues have begun to emerge following the onset of TGR operation, particularly when operating at full capacity [10,20].

According to the results presented in reported study, in upstream area of the dam, impoundment had caused some changes in species compound of small mammal. The densities were keeping low as a whole duo to the deratization in the TGRA. Most investigation showed that measures that reduce rodent population densities could be effective in controlling rodent-borne diseases during the large-scale hydraulic engineering construction. Meanwhile, habitat fragmentation due to the construction of the TGD had led to further changes in species composition and biodiversity, nevertheless, the species richness of island and mainland rodent communities was not significantly different [29]. Along-term observations in Central Brazilian Amazonia showed small mammal assemblages further exhibited a nested structure in insular habitat, smaller and more isolated islands exhibited depauperate small mammal assemblages, in which some species were typically either missing or persisted at very low abundances [46]. So, it may need to have a long period of monitoring in TGRA.

Studies from the Dongting Lake and Poyang Lake showed that the hydrological and meteorological changes implemented through TGD operation might favor the small mammal community, resulting in higher species immigration to the beaches of lakes during the dry season. Water regulation might potentially affect small mammal communities by altering habitat, in addition to changing species movement and survival patterns [34, 59,60]. The possible underlying mechanism for such a shift in species distributions is the longer period of beach habitat exposure, which increases the opportunity for other species to establish home ranges and reproduce. It might be concluded that lake beach habitats have become suitable for harboring a broad variety of small mammal species, because of changes in habitat structure and composition (e.g., vegetation structure, habitat type, landscape composition, connectivity, substrate, moisture and size), in parallel to changes in flood regimes.

Those studies just presented some changes that have occurred in the small mammal community after the onset of TGR operation. It was not determined that the exact factors were had led to the observed changes in the small mammal community. Understanding how biodiversity is affected by hydropower projects is then of critical importance for management actions [46]. Therefore, future research should focus on determining to what extent and in what way the TGP has contributed toward changing the structure of the small mammal community inhabiting the up and downstream of TGD. Further research about succession mechanisms and the factors that influence succession, is required, through long-term surveys of the small mammal communities inhabiting the various habitats.

Meanwhile, it was said that leptospirosis and Hemorrhagic Fever with Renal Syndrome (HFRS) were endemic in Yangtze River. Rodents and Insectivora are the most important vectors of these diseases. Although effective deratization before impoundment during construction of the TGR likely helped to decrease the rodent density and prevent outbreaks of rodent-borne diseases, the development of some species of rodents and Insectivora in beaches of lakes downstream should be given more attention. They had immigrated to the beaches of lakes during the dry season, for example A. agrarius, R. losea and R. norvegicus . They are very important reservoir hosts of the causative pathogens of HFRS and leptospirosis. Specially, A. agrariusis the primary host of HFRS and leptospirosis, which also are remarkably resistant to many disease-causing pathogens, and have a strong potential to transmit diseases. Up to 17 diseases at least are known to be transmit by A. agrarius. Thus, long-term and effective surveillance of small mammal for monitoring future demographic changes as well as its impact on public health needs to be established.