Ecological and economic impacts of exotic fish species on fisheries in the Pearl River basin

The Pearl River basin has the highest fish diversity in China and has been identified as a global hotspot of fish biodiversity. Twenty-three exotic fish species are distributed in the Pearl River basin, tilapias, North African catfish, roho labeo, and mrigal carp have become important exotic species to the local fisheries economy. We investigated catches in the Pearl River basin using an isometric random sampling method and analyzed the ecological and economic impacts of exotic species. The results show that the biomass percentage of exotic species in the upper reaches of the Pearl River basin was higher than that in other regions, and was higher in the drought season than in the flood season. Pearson correlation analysis indicated that the abundances of exotic species were significantly positively correlated, which indicates that they might have joint impacts. Although exotic species enhance fishing production, they decrease the income of fishermen. A multiple linear model indicated that as the biomass percentage of exotic species increased by one percent, the income decreased by 20.19 RMB/vessel/month (RMB, Ren Min Bi, Chinese currency) at a given site. This inevitably caused the fishermen to catch more fish to increase their income. It is urgent to assess the invasion potential of introduced species in southern China and limit both the biomass and spread of exotic species.


Introduction
The spread of exotic species worldwide has been gradually accelerated by economic globalization, and the effects of exotic species invasions on recipient ecosystems have become global issues (Gurevitch and Padilla 2004).Although the number of exotic species is not sufficient to reflect the degree of ecosystem disturbance (Weaver and Garman 1994), exotic species are one of the major causes of wildlife extinctions by changing habitat and ecosystem processes (García-Berthou and Moreno-Amich 2000; Clavero and García-Berthou 2005).High abundances of exotic species pose a significant threat to native species due to competition, predation, disease transmission, hybridization and erosion the gene pool (Hulme 2007;Hermoso et al. 2011;Avlijas et al. 2018), driving declines in biodiversity (McGeoch et al. 2010).Widespread exotic species could elevate extinction levels and depress speciation in ecosystems, which are closely related to biodiversity loss (Stigall 2011).Exotic species in freshwater ecosystems create greater impacts than those in terrestrial and marine ecosystems (Vilà et al. 2010;Xiong et al. 2017).The number of exotic fish species is higher than the number of aquatic plants and that of other exotic aquatic animals in freshwater systems (Xiong et al. 2015;Hui et al. 2016), and they are difficult to eradicate after populations of exotic fish species have become established (Britton and Brazier 2006).Therefore, invasion by exotic fish species in freshwater systems is an issue of critical management importance in ecology.
Exotic species not only cause ecological harm but also cause enormous economic losses (Xu et al. 2006a).In general, the economic impacts of aquatic exotic species are significantly positively correlated with their ecological impacts (Vilà et al. 2010).Economic losses consist of direct losses, such as costs for the eradication of exotic species or decreases in revenue, and indirect losses are due to losses to the service functions of the landscape (Xu et al. 2006a).It has been reported that exotic fish species cause economic losses in the United States and China of up to almost $5.4 billion and $73.9 million annually, respectively (Pimentel et al. 2005;Xu et al. 2006a).Approximately 439 freshwater fish species have been introduced to China, and almost all of them were intentionally introduced for aquaculture, ornamental purposes or biomass control (Xiong et al. 2015).Great ecological and economic losses caused by the unintentional introduction of aquatic species have been widely reported (Lovell et al. 2006;Vilà et al. 2010;Veer and Nentwig 2015), but the economic impacts of exotic species that were intentionally introduced for aquaculture on natural ecosystems are not understood (Gu et al. 2015).
The Pearl River is the longest river in southern China and has the highest fish diversity in China; it has therefore been identified as a global hotspot of fish diversity (Xing et al. 2016).Historical surveys recorded approximately 659 fish species in the Pearl River basin, and it has the greatest number of endemic and threatened species in China (Wu 2015;Xing et al. 2016).However, this global biodiversity hotspot not only suffers from the threat of drought but also faces exotic invasions (Gu et al. 2012;Shu et al. 2013).Historical records and field surveys have shown that approximately 31 exotic aquatic animals occur in the Pearl River basin, and 23 of these are fish species (Gu et al. 2012;Yang and Xiu 2014).Moreover, some of these species are widely distributed in the Pearl River basin and are fishes that are important to the local economy (Gu et al. 2012;Wei et al. 2017).The present study investigates the spatial and temporal variations of exotic species in the Pearl River basin.In addition, it focuses on the relationships between exotic species and the income of fishermen, fishing production, fishing pressure and the average price of catches.Based on the results, some management implications pertaining to exotic species have been provided for the Pearl River basin.

Study area
The Pearl River is the largest river in southern China and is located within a tropical and subtropical monsoon climate.The river originates in Yunnan province and drains the majority of Guangdong and Guangxi provinces, with a total length, catchment area and annual discharge of 2,320 km, 450,000 km 2 and 10,000 m 3 /s, respectively.This study focused on Guangdong and Guangxi provinces in southern China, which is a region with abundant non-native fish species.The wild-caught freshwater fisheries in these two provinces were approximately 267,506 t in 2016, accounting for approximately 11.5% of the total yield in China (FDMA 2017).

Data collection and analysis
A total of 10 sampling sites were selected from present studies conducted in various areas of Guangdong and Guangxi provinces.Among them, five sites were distributed in the mainstream of the Pearl River, while others were either distributed in tributaries or in other rivers that independently flow into the South China Sea (Figure 1).The sampling sites BMX and LJS are located in reservoirs.The information on fishing boats in each county was obtained from the local fisheries management department, and each boat has a registration number that can be used to select random samples.Ten boats were sampled at each site using the isometric random sampling method (Madow and Madow 1944).The tonnage and power of the fishing boat are in the 0.2-30 tonnage and 0.3-40 kW range, respectively.The fishing gear mainly include gillnets, benthic fyke nets, long-lines, seines, trawls and others (eg.cast nets) (Table 1).The mesh size of three-layer gillnets are with a central mesh size of 4-8 cm in central and outer mesh sizes of 10-20 cm.Mesh size of benthic fyke nets, seines, trawls and cast nets are 1-2 cm, 2-4 cm, 8-10 cm and 4-8 cm, respectively.For each boat sampled, fishermen recorded the production of each species (mostly economic species) and the type and number of fishing nets, costs, and income each day.The fishermen of every boat sampled were learned for data records before research.The researchers collected records from fishermen before the 10th of every month, and the study lasted for the whole year (from February 2016 to January 2017).The temperature (Tem) data were collected from the weather report (http://www.tianqihoubao.com/lishi/).The distances from the sampling sites to the estuary (Dist) were determined using Google Earth.Number of fishing boats represents the fishing pressure at the sampling sites and was obtained from the local fisheries management department.Biomass percentage of each exotic species: Biomass percentage of the four studied exotic species: P f = P i1 + P i2 + P i3 + P i4 , Average price of catches: P a = total income from catches/total yield The abundance of exotic species was represented by biomass percentage in the present study.W i is the weight of exotic species i, W t is the total weight of all catches.
Catch per unit effort (CPUE, kg/vessel/day) data were obtained from the recorded data.Sampling confidence was assessed using the standard deviation (SD), coefficient of variation (CV) and relative error (RE) of the CPUE value.Linear regression analysis was used to determine whether the exotic species affected the income of fishermen and the average price of catches.All statistical analyses were performed using R 3.3.1 (R development Core Team 2013) and Excel.

Distribution of economic exotic species
The total catch in the present study collected from the 10 sampling sites was 2.41 × 10 5 kg.The biomass percentage and price of each economic species in the catches at the different sites are shown in Supplementary material Table S1.The biomass percentages of tilapia species, the North African catfish Clarias gariepinus (Burchell, 1822), the roho labeo Labeo rohita (Hamilton, 1822), the mrigal carp Cirrhinus mrigala (Hamilton, 1822) and native species in the catches were 9.38%, 0.04%, 0.14%, 0.28% and 90.15%, respectively (Table S1).The biomass percentage of exotic species at each site ranged from 0 to 33.86%, and the percentages in the upper reaches of the Pearl River basin (BMX, XBQ, FSX, HX) were higher than those in other regions (Figure 2A).The biomass percentage of exotic species in each month ranged from 6.38% to 13.75% (Figure 2B), and the percentages in the drought seasons (Jan-Feb, Nov-Dec) were higher than that in the flood season (Mar-Oct).The tilapia species, North African catfish, roho labeo, mrigal carp and native species contributed 6.64%, 0.07%, 0.13%, 0.24 and 92.92% of the total fishing income, respectively.

Relationships between environmental variables and exotic species
The linear regression showed that the CPUE was significantly correlated with the temperature, and that the CPUE was higher in areas with higher temperatures (R 2 = 0.776, P < 0.001, Figure 3A).The number of economic exotic species in the catches was positively correlated with the distance from the estuary (R 2 = 0.778, P < 0.001, Figure 3B), and there were more exotic species farther away from the estuary.The abundance of exotic species was significantly and positively correlated with the distance from the estuary (Figure 3C): exotic species increased along the longitudinal profile of the Pearl River basin.Pearson correlation analysis showed that the biomass percentage of exotic species was not significantly related to the fishing pressure (P = 0.431, Figure 3D).
The pearson correlation analysis indicated significant positive correlations among the abundances of the exotic tilapia, North African catfish, roho labeo and mrigal carp (Figure 4).The abundance of tilapia was negatively significant related to the native species (P < 0.01, Figure 4).

Economic impact of exotic species
The multiple linear regression analysis showed that the abundance of exotic species was significantly and negatively correlated with the local fishing income; that is, the fishing income was lower in regions with abundant exotic species than in other regions (Table 2, R 2 = 0.871, P < 0.01,).The yield (kg/vessel/month) and sampling site also was significantly related to the income (P < 0.01), while the month has no significant impact on the income (P = 0.960).The results indicated that when the proportion of exotic species increased by one percent, the income decreased by 20.19 RMB at a given site (Figure 5).Exchange rate RMB 6.29 = US $ 1.00 (April 2018).
The average price of catches was significantly correlated with the sampling sites (R 2 = 0.896, P < 0.01, Table 3).The abundance of exotic species and yield have negative effect on the price, but the effect wasn't statistically significant (P > 0.05).The multiple linear regression model indicated that the yield was significantly correlate with sampling site and month (R 2 = 0.760, P < 0.01), but has no significant correlation with abundance of exotic species (P = 0.08).And the biomass percentage of exotic species was significantly correlated with sampling site (R 2 = 0.862, P < 0.01), not significantly related to number of native species (P = 0.112) and month (P = 0.08).

Ecological impacts of exotic species
Tilapia species, the North African catfish, the roho labeo and the mrigal carp were intentionally introduced for cultivation in the last century and are commonly cultured in southern China due to their rapid growth, high disease resistance and high adaptability (Chen et al. 2007;Lin et al. 2015).These exotic species escaped from ponds and then established thriving populations in the natural ecosystem (Xu et al. 2006b).Historical records and field surveys indicate that these four exotic species are widely distributed in southern China and have become local economic species (Figure 2A) (Gu et al. 2012;Gu et al. 2018).The intentionally introduced species enhance fishing production (P < 0.05) because of their high growth rate.However, the increasing abundance of exotic species will inevitably threaten native species through competition, reducing the relative abundance of native species (Figure 2).It has been reported that exotic species cause great ecological impacts after introduction by decreasing the abundance of native species, depressing native species growth, and simplifying the food web in the ecosystem (Britton et al. 2010).The presence of tilapia has decreased the growth of the native mud carp and negatively affected the CPUE in the Pearl River basin (Gu et al. 2015).It has been reported that exotic species Nile tilapia (Oreochromis niloticus Linnaeus, 1758) declined catches of native carp but positively contributing to fishing production in Ganga River (Singh et al. 2013).In the present study, exotic species increased fishing production, reduced native species and depressed the income of fishermen.
The number and biomass of exotic species in the upper reaches of the Pearl River were higher than in the lower reaches (Figure 3B, C).And the number of native species was also significant positively linear correlated with the distance from the estuary (P < 0.05).The upper reaches of the Pearl River occur in karst regions, which support high fish diversity (Shu et al. 2013) and harbor many dams and reservoirs (Chen et al. 2017).The sampling site HX is located in a tributary, the Yujiang River, and 9 hydroelectric dams are distributed throughout the main stream of this river.The sampling sites BMX and LJS are also located in reservoir areas and show a high abundance of exotic species (Figure 2A).The present study indicated that these sites and areas close to reservoirs have a higher abundance of exotic species than other areas.Previous studies have shown that reservoir areas have more exotic species than other regions in Iberian streams, which suggests that habitat stability plays an essential role in the colonization and proliferation of exotic species (Hermoso et al. 2011).After Three Gorges Dam impoundment, the number of exotic species also significantly increased (Xiong et al. 2018).The four economic exotic species considered in this study are pond culture species and prefer lentic habitats; therefore, reservoir habitats could promote the proliferation of these species.
Native species have a higher ability to take advantage of competitionfree conditions than exotic species, but tend to be more negatively response to competition (Kempel et al. 2013).Many studies indicated that exotic species have negative effects on native species through trophic competition, predation, habitat degradation and disease transmission (Mills et al. 2004;Hermoso et al. 2011;Avlijas et al. 2018).However, interactions between exotic species are poorly studied.Exotic species can interact with one another, which likely helps increase populations (Simberloff and Von Holle 1999).Interaction networks of the receipt ecosystem could be altered after one exotic species invaded (Tylianakis et al. 2010), that could contribute to the establishment of afterward exotic species.The abundances of four exotic species in the Pearl River basin are positively correlated.The research on interactions between exotic species is urgently conducted in the future.We suspected that the joint impacts of these species on the Pearl River would be more severe than those when the species are acting alone.

Economic impacts of exotic species on fisheries
In general, the introduction of species for aquaculture promotes fisheries production due to their high growth rate, high disease resistance and high adaptability (Xu and Ming 2018).Exotic species with low commercial value are more likely to successfully establish populations in natural environments than others due to arbitrary disposal and poor management (Gu et al. 2018).In the present study, the exotic species were shown to enhance fishing production, but the income decreased because of the low price of exotic species (Figure 5, Table 2).This likely caused the fishermen to catch more fish to offset their decreased income.

Management implications
In general, fishing catches decrease in the winter because most native species, with the exception of some migratory fish, hide in deep water to escape the low temperatures, making it difficult to catch these species.However, the relative biomass of exotic species was high in the winter due to their biological traits, including tolerance to wide ranges temperature (Das et al. 2005;Russell et al. 2012).This feature makes it convenient to remove exotic species in the winter.
It has been reported that the introduced North African catfish has a high risk of becoming invasive, as assessed by the Fish Invasiveness Screening Kit (FISK) (Piria et al. 2016).In the present study, the North African catfish was found in the upper reaches at BMX and in the tributary section at FSX, and this species accounted for 0.4% and 0.02% of the catches in these areas, respectively.There is no doubt that this species will spread to other areas of the Pearl River basin due to its high adaptability.There is an urgent need to assess the invasion potential of introduced species in southern China and establish strict management policies regarding exotic aquatic animals to protect the biodiversity before this biodiversity hotspot is lost.

Figure 1 .
Figure 1.Locations of the 10 sampling sites in the Pearl River basin, China.

Figure 2 .
Figure 2. Biomass percentages (mean ± SE) of exotic and native species at different sites (A) and in different months (B).

Figure 3 .
Figure 3. Correlation analysis between the CPUE, number of economic exotic species, biomass percentage of exotic species and environmental variables.

Figure 4 .
Figure 4. Multipanel display of the pairwise relationships between the biomass percentages of different species in catches, with bivariate scatter plots in the lower panels, histograms on the diagonal and Pearson's r correlations in the upper panels (statistical significance: P < 0.001 ***, P < 0.01 **, P < 0.05 *).

Figure 5 .
Figure5.The multiple linear regression analysis of income with yield and biomass percentage of exotic species at sampling site DQX.The lines shows that the relationships between income and yield while the biomass percentage of exotic species (invper) are 0, 10% and 20%, respectively.Exchange rate RMB 6.29 = US $ 1.00(April 2018)

Table 1 .
Basic information of fishing gear at each sampling site.The total number of gillnets represents the number of days with gillnet fishing and other specific nets as same.

Table 2 .
Results of multiple linear regression model of income with yield (kg/vessel/month), biomass percentage of exotic species (invper, %), sampling sites and month.The sampling sites that have significant effects on the model are shown in the table.

Table 3 .
Results of multiple linear regression model of average price of catches with yield (kg/vessel/month), biomass percentage of exotic species (invper, %), sampling sites and month.The sampling sites that have significant effects on the model are shown in the table.