Yuwanghe Reservoir is located in the northern part of Liupanshui, Guizhou Province; it belongs to Yuwanghe River, which is a first-class tributary of the Beipan River and is situated approximately 50 km from Shuicheng County (26°08ʹ30–26°09ʹ02N, 104°45ʹ30—104°45ʹ50E; Figure 1 ). The reservoir started to store water and began its operation in 2020 with a maximum depth of 42.3 m and an average depth of 18.8 m. The average capacity is 1.022 × 10⁷ m³. The main function of the reservoir is to provide water and irrigation facilities to the residents of the area. To analyze the horizontal distribution of fish assemblages, the reservoir was divided into three sections —namely, head area, middle area, and tail area —on the basis of its geographical features.

The hydroacoustic survey was conducted during daylight hours in May 2021. A DT-X split-beam echosounder (BioSonics, Seattle, WA, USA) operating at a frequency of 200 kHz was used, with a transducer beam angle of 6.8° × 6.8°. The transducer was mounted on the starboard side of the vessel at a distance from the bow equal to one-third of the hull length; thus, the transducer could reach a depth of 0.5 m below the water. The vertical geographic position and the hydroacoustic route of the beam were recorded in real-time by Garmin GPS 17×HVS. The echosounder was connected to a laptop to display and store the hydroacoustic data in real-time by using Acquisition 6.0 software. Water temperature and salinity data required for the data acquisition program were determined using the YSI6600V2 multiparameter water quality monitor. To reduce the impact of vessel noise on the survey, the hydroacoustic survey was conducted along zigzag transects, with the vessel sailing at the speed of approximately 6–8 km·h^–1 (Lian et al., 2018). The survey started from the head area of the reservoir and ended at the tail area ( Figure 1 ). The route covered the entire reservoir. The investigation was repeated twice back and forth to ensure coverage of more than 6% (Matthias et al., 2012).

Fish catch surveys were conducted to obtain the fish composition and community structure; the aim of these surveys was to determine the consistency between the hydroacoustic survey signals and the distribution of fish body length. Six fish sampling sites were selected, including two sites each at the head, middle, and tail areas ( Figure 1 ). Composite gillnets of multiple meshes (50 m in length and 3 m in height; mesh size of 1.0, 1.6, 2.0, 2.5, 3.1, 4.0, 4.8, 6.0, 7.5, 8.5, 11.0, and 12.5 cm), trammel gillnets (200 m in length and 5 m in height; mesh size of 6.0, 8.0, 10.0, and 12.0 cm), and floor cage net (20 m in length, 0.4 m in width, 0.4 m in height, and mesh size of 5 mm) were used for the field sampling. The floor cage nets were installed in the coastal zone near each sample site at a position as distant as possible. The gillnets were installed at each sample site in the main area. The nets were set for 12 h from 17:00 to 5:00. Fish catches were identified, measured, and counted in fresh condition. Species identification was performed by referring to the guidelines of the Zoology of China (Chen, 1998; Yue, 2000). Spatial and trophic ecological niche classification was performed by referring to the Guizhou Fish Journal (Wu, 1989). Measurements were performed at the accuracy level of 1 mm in length and 0.01 g in weight.

Quantitative samples of plankton and water were collected along with the catch collection. The plankton samples obtained from the mixtures of the bottom, middle, and surface water at each site were collected in 50- ml polyethylene bottles, and 1 ml of formaldehyde solution was added for preservation. The samples were then stored for laboratory tests (Zhang and Huang, 1991). Water samples were collected in 1, 000- ml black polyethylene bottles at a surface water depth of 0.5 m. The samples were brought to the laboratory under low temperature conditions to measure the content of total nitrogen, total phosphorus, ammonia nitrogen, and chlorophyll a. Water depth was measured based on acoustic signals. Water temperature, pH, dissolved oxygen, and conductivity were measured on site by using a YSI 6600V2 multiparameter water quality monitor (Zhang and He, 1991). A transparency disk was used to measure the water transparency.

The hydroacoustic data were converted and post-processed with Visual Analyzer version 4.3 software. Because the fish signals from the echosounder were scattered, the blind areas of detection and areas with high interference at 1 m below the level of the transducer and 0.5 m above the bottom of the probe were artificially eliminated; this step minimized the effect of noise (Li et al., 2022; Lian et al., 2022). The echograms of all survey sections were carefully examined and manually edited to regenerate the bottom surface if required. Background noise beyond the target signal, such as bubbles and noise interference from ships, was manually eliminated based on the characteristics of the echogram (Huang et al., 2019). To shield the echo signal from weak scatterers such as plankton, the minimum threshold in the target intensity echo image was set to –70 dB.

The total length of the fish was derived from the empirical formula (TS - TL) of target strength (TS) and total length (TL) of carp as proposed by Frouzova et al. (2005):

where TS is the target strength of the fish body, TL is the total length of the fish, and 103.9 is a constant value.

The conversion between the t otal length and the fish weight is based on the relational expression obtained by applying the power function W = a*TL ^b (W is the fish weight and a and b are the parameters) to fit the total length and weight of the fish catch samples (Ricker, 1975; Ye et al., 2007).

The biomass for each transect was calculated by multiplying the average fish numerical density estimated acoustically with the weight corresponding to the average TS for this transect. The total biomass for the given area was calculated as a weighted mean.

The goodness of fit between fish catch samples and hydroacoustic data was analyzed by Mann–Whitney U-test. The differences between areas for mean fish density and fish biomass were analyzed with two-sample t-test. Pearson correlation coefficient was used to determine the correlation of fish spatial distribution and environmental factors. All analyses were completed with the OriginPro 2021 (OriginLab, 2021) and ArcGIS 10.2 software (Esri, 2013).

In the survey of fish catches, six species of fishes belonging to three orders and four families were found in Yuwanghe Reservoir, and a total of 1,892 fish were sampled. The status and the role of species in the community were classified according to their number and the weighted proportion of catches. Species with an index of relative importance (IRI) > 1,000 are defined as dominant, those with IRI between 100 to 1,000 are important species, those with IRI between 10 to 100 are common species, and those with IRI < 10 are occasional species (Jutagate and Sawusdee, 2022). On the basis of this criterion, topmouth gudgeon (Pseudorasbora parva) was the dominant species in the fish community, Carassius auratus and Culter alburnus were important species, and Micropercops swinhonis and Paramisgurnus dabryanus were common species ( Table 1 ). The curve-fitting result between the total length and the weight of the fish catch samples ( Figure 2 ) by using regression analysis showed that the relation formula was W = 0.00295*TL^3.404 (R ² = 0.9755), and the fitting degree was good.

The total length range of the fish catches was 38 to 236 mm. Small fishes formed the main group, with more than 96% of the fish having a total length of <100 mm. The quantitative proportion of the fish in the catches decreased rapidly with the increase in the total length. The hydroacoustic survey showed that the target strength of fish was –62 to –48 dB, which showed a fluctuation in the value ( Figure 3A ). The target strength was concentrated in the range of –62 to –58 dB, with a peak strength appearing at around –58 dB. The total length of topmouth gudgeon, whose population accounted for 95.74% of the total population, ranged from 38 to 96 mm. Based on the empirical formula (Eq. 1), the target strength of –56 dB was converted into a length of approximately 100 mm. Thus, it can be inferred that the acoustic signal with the target strength of <–56 dB mainly originated from topmouth gudgeon. The results of the Mann–Whitney U-test indicated no significant difference (P > 0.05) in the total length between the fish catches and the hydroacoustic data; this finding indicated a good agreement between the results of the hydroacoustic survey and the field analysis ( Figure 3B ).

The hydroacoustic survey showed an uneven horizontal distribution of fish in Yuwanghe Reservoir ( Figure 4 ). Fish density and fish biomass were significantly varied in the different areas of the reservoir (P < 0.05). The mean fish density of the entire reservoir was 318.7 ± 256.1 individuals/1,000 m², and the fish density tended to increase from the head area and the middle area to the tail area of the reservoir, with values of 168.5 ± 60.1 individuals/1,000 m², 306.8 ± 124.7 individuals/1,000 m², and 696.4 ± 288.9 individuals/1,000 m², respectively ( Figure 5A ). On the basis of the W - TL relation formula and the weighted average calculation, the mean fish biomass was 984.8 g/1,000 m², and the fish biomass from the head area and the middle area to the tail area of the reservoir was 532.5 ± 230.4 g/1,000 m², 1,264.2 ± 400.6 g/1,000, m² and 2,482.6 ± 600.8 g/1,000 m², respectively ( Figure 5B ). Based on fish density and biomass, the highest distribution of fish was in the tail area of the reservoir, and the lowest was in the head area ( Figure 4 ).

Based on 3 m as the water layer according to the water depth of the reservoir, the proportion of fish density in different water layers in different areas of the reservoir to the sum of fish density in the entire water column was calculated as shown in Figure 5 . The majority of the fish population was found in the water layer within the depth range of 0 to 12 m, where more than 97.3% of fish were distributed. In the water layer at depths of more than 12 m, the fish density tended to decrease gradually with the increase in water depth. However, differences were noted in the fish distribution pattern in the different areas of the reservoir along the vertical direction. Fish at the tail area preferred to move to the surface, and the proportion of fish in the depth range of 0 to 6 m was 81.5%; this proportion was higher than that at the head area (69.9%) and at the middle area (61.3%) ( Figure 6 ). The reason for this difference may be the fact that the tail area of the reservoir has a shallow depth on average, and the fish were closer to the water surface. Consequently, the proportion of fish density in the surface waters was higher in the entire water column.

According to Pearson correlation coefficient analysis, the main factors that influenced the fish distribution in Yuwanghe Reservoir were water depth (WD), water temperature (WT), dissolved oxygen (DO), total nitrogen (TN), and plankton (including zooplankton and phytoplankton) density ( Figure 7 ). WT, DO, and plankton density were positively correlated with fish density, whereas WD was negatively correlated with fish density on the contrary. The pH, conductivity (Cond), total phosphorus (TP), ammonium nitrogen (NH₄ ⁺–N), and chlorophyll a (Chl a) were not observed to have a significant connection with fish density.

From the survey data, only six species of fish were found in Yuwanghe Reservoir ( Table 1 ). Regarding ecosystem stability, Yuwanghe Reservoir had fewer fish species, and the community structure was more based on a single species. Furthermore, the variety of trophic levels was poor, particularly for secondary consumers. A community with a very simple structure is extremely unstable and prone to retrograde succession under external interferences (Balazs et al., 2022; Wu et al., 2022 ). In terms of biological manipulation, the proportion of small fish in Yuwanghe Reservoir was too high. The hydroacoustic survey results showed the proportion of fish with a target strength of <–56 dB; accordingly, the proportion of fish with a total length of <100 mm was 92.9%. The proportion of fish with a total length of <100 mm in the collected fish catches was 96.3%, and the predominant species was topmouth gudgeon (P. parva). Zooplankton is an important prey population for small fish (Zhang, 2005; Liu et al., 2007; Lian et al., 2018). However, zooplankton feeds on phytoplankton, and the presence of a large number of small fish will result in a large decrease in zooplankton biomass in the water column, which is not favorable for controlling the phytoplankton biomass. Secondly, the absence of filter-feeding fish that feed on phytoplankton can easily lead to a massive growth of some algae under suitable conditions, which can trigger the occurrence of water blooms (Liu and Xie, 1999; Liu and Xie, 2003; Xie and Liu, 2001). Thirdly, because of the short generation cycle, early sexual maturity, and high reproductive capacity (Pollux and Korosi, 2006), topmouth gudgeon has been widely distributed in Central Asia, Europe, and North Africa in less than 50 years (Gozlan et al., 2010), and this fish species is listed as one of the most severe aquatic invasive species in Europe (Sala et al., 2000; Andreou et al., 2012) Following invasion into new areas, topmouth gudgeon poses a serious threat to the survival and reproduction of indigenous fish of the water bodies and can impair the stability of aquatic ecosystems by competing for food and habitat, carrying new diseases, and changing the characteristics of the habitat (Fletcher et al., 2016)— for example, topmouth gudgeon invasion in the Yunnan Plateau lakes led to the reduction or even extinction of many native fish species and a series of knock-on ecological effects (Chen et al., 1998; Yan and Chen, 2009). In the British Lake District, the whelk has become the most abundant fish in the region after 4 years of invasion, thereby threatening the survival of other native fish species (Britton et al., 2007). Unfortunately, there were too few carnivorous fish in the reservoir, and topmouth culter (Culter alburnus) was the only carnivorous fish species; its proportion in the catches was also extremely low, with a population share of only 1.01%. Moreover, topmouth culter prefers to inhabit the upper layers of the water column, while topmouth gudgeon inhabits the lower layers; this makes the topmouth culter population insufficient in gaining effective control over topmouth gudgeon in terms of energy conversion efficiency and habits.

In the horizontal direction of Yuwanghe Reservoir, the fish densities and biomass showed a trend of gradual increase from the head area to the tail area of the reservoir; this trend was related to WT, DO, TN, and plankton feed from the correlation analysis between fish density and environmental factors ( Figure 7 ). Although there was no significant difference in water temperature between the different areas of Yuwanghe Reservoir, the overall trend of a gradual increase in temperature from the head area to the middle area to the tail area of the reservoir was observed. Fish do not like to move in areas with low water temperatures; hence, the fish density tends to increase as they move further away from the dam. A similar result was reported by Prchalova et al. in their study on Rimov Reservoir (Prchalová, 2009). The DO and TN levels were another major cause of differences in fish horizontal densities. The inflowing river is located at the tail area of the reservoir, and the highest DO levels are found at the tail area of the reservoir where the water flows in. The DO concentration slowly decreased as the exogenous nutrient salts diffused away from the inlet river. As the predominant species in Yuwanghe Reservoir, topmouth gudgeon prefers to inhabit shallow waters (Asaeda and Manatunge, 2005; Asaeda and Manatunge, 2007). The shallow water area of Yuwanghe Reservoir provides the necessary habitat for P. parva, and it also has little risk of fish predation because of the low percentage of carnivorous fish. Consequently, fish density is higher in shallow waters than in deep water. Furthermore, the sampling survey of plankton in Yuwanghe Reservoir revealed that the horizontal distribution of plankton in the reservoir area showed a gradual increase from the head area to the tail area of the reservoir; thus, the difference in bait organisms was also a primary factor that led to a difference in the horizontal distribution pattern of fish in Yuwanghe Reservoir in the present study.

Vertically, most of the fish in Yuwanghe Reservoir were distributed in the water layer within a depth of 0 to 12 m, and the fish density tended to decrease gradually with the increase in water depth. Some studies have been conducted on the distribution of fish density or biomass in different water layers. Mou et al. (2012) surveyed Hongfeng Lake, a large deep lake, by using the hydroacoustic method, and they found significant differences in fish density in different water layers, and the fish density gradually decreased with an increase in water depth. Lian et al. (2018) reported that more than 97.6% of fish were distributed in the water layer within the depth range of 0 to 10 m in a hydroacoustic study of deep-water reservoirs in Yunnan Plateau. Zhou et al. (2021) showed a negative correlation between fish density distribution and water depth in Hongchaojiang Reservoir. The results of the present study agreed with those of former studies. Fish lack the ability to regulate body temperature, and most warm-water fish prefer a suitable water temperature of 20 to 30°C; consequently, the either too high or too low water temperature will affect the growth and movement of fish (Yin, 1995). Hence, most fish tend to stay away from the water layer below the thermocline, while zooplanktons—which is the food of most fish —are mainly distributed in the water layer above the thermocline (Lian et al., 2018). DO is also an important factor that affects the vertical distribution of fish. In general, DO concentration in the deep-water reservoir gradually decreases as the water depth increases; thus, fish will actively migrate to avoid low-oxygen or anoxic water (He and Cai, 1998; Lian et al., 2018). It was also the reason why most fish species were distributed in the upper water layer in the present study.