In the last few decades, non-native freshwater fishes have been introduced all over the world for economic purposes, including aquaculture and aquarium trade, as well as improvement for wild stocks resulting in adverse environmental and socio-economic effects. The Guangdong province of China is at a high risk of fish invasions owing to its warm and humid climate, abundance of water courses, flourishing aquaculture and ornamental fish trade, and extensive sea ports. A total of 160 non-native freshwater species were introduced in the Guangdong province and 71.9% of them were imported for aquarium purposes. Fourteen species have established self-sustaining populations and 21 species were found in the main river basin of the Guangdong province. Propagule pressure, rapid evolution and abundant resources in the environment were the factors likely to contribute to successful invasion by non-native fishes. The invasion of non-native fishes in the Guangdong province has already resulted in economic losses, decline of native species, as well as negative impacts on the functional diversity of native fish assemblages. To mitigate these effects and prevent future non-native fish invasions, scientists, policy makers and stakeholders should collaborate on the management of non-native fish introductions by developing risk assessments, statutory regulations, public education and scientific research.
Fish invasion poses great threats to aquatic fauna in freshwater ecosystems (Gozlan et al., 2010; Gallardo et al., 2015). The introduction of non-native (NN) freshwater fishes primarily occurs through their transport around the world for commercial purposes such as aquaculture, aquariums, game fishing, fisheries enhancement and biological control (Lin et al., 2015). A total of 439 freshwater fish species have been introduced into China for aquaculture or aquarium purposes (Xiong et al., 2015), however production and cultivation of freshwater fish occur predominantly in the Guangdong province, accounting for 1/3 of introduced fish species in China (Zhong and Power, 1997; Lin et al., 2015).
The Guangdong province is particularly suitable for the aquaculture or aquarium industry because it features a warm and humid climate and abundant water courses. In addition, the Guangzhou city of the Guangdong province has been a historic port city since the Qin (221–206 BC) and Han Dynasty (202 BC–220 AD), allowing easy importation of NN fish species. It also has the largest aquarium markets in China (i.e. Yuehe Flowers, Birds and Fish Wholesale Market). Thousands of NN ornamental species are transported to this market (Wang et al., 2007). These factors are likely to increase the likelihood of accidental or intentional release of NN fishes into native rivers through aquaculture and aquarium activities (Naylor et al., 2001). Consequently, the Guangdong province is at high risk of invasion by NN fish species (Li et al., 2013; Wei et al., 2017). So, this study provides: (1) a summary of the status of NN freshwater species in the Guangdong province; (2) the potential mechanisms driving invasion success; (3) ecological consequences of fish invasion, and (4) a summary of the existing regulations about importation and transportation of NN fish in China. Such information will facilitate the understanding of the status and risks of NN fish in this region, which will be useful for developing management strategies for NN fishes.
The status of non-native freshwater fish in the Guangdong province
A list of NN freshwater fish introduced to the Guangdong province was summarised by consulting peer-reviewed literature and on-line databases, and by a market survey. A non-native species is defined as species which does not occur naturally in a geographical area and is intentionally or unintentionally introduced to the area beyond its natural distribution (synonyms: non-indigenous, exotic and alien) (Richardson et al., 2000; Copp et al., 2005a). Non-native species include those translocated from their natural distribution to other areas within a country, as well as species which are introduced from other countries (Copp et al., 2005a; Xiong et al., 2015). A total of 160 NN freshwater fishes were identified in the import/trade records in the Guangdong province (Table S1?). These species were introduced from South America (35.6%), Asia (27.5%), Africa (17.5%), North and Central America (11.8%), Oceania (5.0%), and Europe (1.9%) for different purposes such as aquarium (71.9%), aquaculture (28.1%), game fishing (3.1%), and biological control (0.6%). This is consistent with previous findings (Lin et al., 2015; Xiong et al., 2015).
The Guangdong province is located in a sub-tropical monsoon climate zone with a low mean temperature of 13.3 °C, a high mean temperature of 28.5 °C, and a mean annual rainfall of 1300–2500 mm (Guangdong Meteorological Service, 2013). Non-native fishes were mainly introduced from regions with similar climatic conditions (Chen, 1994a). The source of ≈74% (119 species) of NN fish introductions were from South America, Africa, South and Southeast Asia, where the climates are similar to the Guangdong province (Kottek et al., 2006). Climate match was considered as an important predictor to assess the invasiveness of NN fish (Bomford et al., 2010). However, this predictor should be combined with other factors (e.g. undesirable (persistent) traits, reproduction etc.) to evaluate the invasiveness of the introduced fish species (Copp et al., 2005b).
The 160 NN fishes belong to 15 Orders and 43 Families. Perciformes (36.3%), Siluriformes (24.4%) and Cypriniformes (15.6%) dominated as Orders. Cichlidae (22.9%), Cyprinidae (13.6%) and Loricariidae (15.6%) dominated as Families. Cichlidae (23.4%), Loricariidae (21.7%) and Cyprinidae (9.6%) were popular in aquaria. Cyprinidae (24.4%) and Cichlidae (20.0%) were prevalent in aquaculture. Fourteen species have established self-sustaining populations and 21 species were found in the main river basin of the Guangdong province (Table 1).
There are 27 cichlid species imported through aquarium trade and eight species used in aquaculture in the Guangdong province (Table S). A recent risk assessment using the Aquatic Species Invasiveness Screening Kit (AS-ISK: Copp et al., 2016) suggested that two cichlid species, Cichla ocellaris and Heterotilapia buttikoferi were recognized as invasive in the middle and lower reaches of the Pearl River in the Guangdong province (H. Wei and R. Chaichana, unpublished data). Tilapia species were the most widely cultivated aquaculture species in this region (Chen and Ye, 1994b). About 8 tilapia species were introduced from Africa to improve germplasm through cross breeding in China (Chen and Ye, 1994b). A recent study reported that Oreochromis niloticus, Coptodon zillii, O. aureus, Sarotherodon galilaeus, and hybrids of O. mossambicus × O. niloticus have established self-sustaining populations in the main river basin of the Guangdong province, while O. niloticus and C. zillii dominated in the rivers (Gu et al., 2016). These tilapia species probably escaped from aquarium or aquaculture facilities through channels connected to tributaries and/or the main rivers (Gu et al., 2012b).
Cyprinids were also representative NN species in the Guangdong province, amongst which 11 were introduced for aquaculture, and 11 were introduced for aquarium (Table S). Cirrhinus mrigala and Labeo rohita have established self-sustaining populations in the rivers of the Guangdong province (Gu et al., 2018; F. Chen, unpublished data), whilst Cyprinus carpio (mirror variety) and Tinca tinca were found in the main river basins of Guangdong province (Gu et al., 2012a). Native to central and northern China, C. carpio and Carassius auratus have been frequently introduced for aquaculture and aquarium purposes (Froese and Pauly, 2018). The ornamental variety of C. carpio (koi) and C. auratus were introduced and improved by the Japanese and re-introduced to China (Wang, 2017). Carassius auratus was identified as invasive in Europe, Australia, and America (Chapman, 2010). It also poses a high risk in the middle and lower reaches of the Pearl River (H. Wei and R. Chaichana, unpublished data). Cyprinids accounts for 40.8% of the total number of fish species found in the main river basins of Guangdong province (Li et al., 2013). The pre-adaption hypothesis suggested that NN species could easily invade the communities where native species have close phylogenetic relatedness with NN species, because these species are likely to have the same traits as their native counterparts allowing them to adapt the new environment (Daehler, 2001; Ricciardi and Mottiar, 2006). Therefore, the main rivers of the Guangdong province can be easily invaded by NN cyprinids once accidentally escaped from aquarium and aquaculture facilities. The probability of release from culture facilities is likely to be higher in the rainy season due to flooding connecting aquaculture facilities on the river flood plain to the main river (Schultz et al., 2003, Casimiro et al., 2018). Furthermore, NN cyprinids could potentially compete with native cyprinids for food resources and habitats, resulting in a decline of the latter (Daehler, 2001).
Loricariids which are native to South America, are also prevalent in the aquarium trade. Although 23 species were introduced, only Pterygoplichthys spp. have established self-sustaining populations in the Guangdong province (Wei et al., 2017). Pterygoplichthys spp. were introduced in the 1990s and utilised to clean algae and organic litter in aquarium tanks (Li et al., 2007). Pterygoplichthys spp. were first captured in Meihuajiang and have now spread to the main drainages of Guangdong province (H.Wei, unpublished data). Recent findings suggest that successful invaders are usually of low or no commercial value (Gu et al., 2018). In this case, Pterygoplichthys spp. have little economic value for fishermen and therefore are being thrown back into rivers. Pterygoplichthys spp. have become one of the most abundant NN fish species in the main drainages of Guangdong province and populations have continued to increase in recent years (Wei et al., 2017, H. Wei, personal observation).
In summary, although the introduction of NN fish species has brought huge economic benefits and job opportunities (China Fisheries Statistical Yearbook, 2018), the associated risks have been poorly evaluated in the last few decades. Some of these NN fishes (e.g. Pterygoplichthys spp. and Oreochromis spp.) have caused negative impacts on native aquatic ecosystems (Gu et al., 2015; Wei et al., 2017). Further studies should be conducted to assess the risks of the introduced fish species by rapid screening kits (e.g. AS-ISK, Copp et al. 2014) or statistical models (e.g. classification and regression tree, CART and Discriminant Analysis, DA, Kolar and Lodge, 2002), which may help policy makers to manage fish introductions and prevent NN fishes escaping into native rivers.
Potential mechanisms driving fish invasions
Non-native species have to overcome environmental and demographic barriers to establish self-sustaining populations, as well as expand their geographical ranges in the invaded regions (Copp et al., 2005a). The number of individuals released into NN regions (also known as propagule pressure) affects the success of NN species in such regions (Lockwood et al., 2005). Non-native fishes can escape from aquaculture ponds to natural water bodies through connected irrigation channels (Sato et al., 2010). In the Guangdong province, there is a total of 312,081 ha of freshwater water-bodies including 232,031 ha ponds, 3,821 ha lakes, 66,893 ha reservoirs, 1,454 ha rivers, 3,293 ha rice paddies and another 7891 ha for aquaculture (China Fisheries Statistical Yearbook, 2018). Non-native fishes may escape from fish farms located in these water bodies and then spread to the rivers. For example, the production of tilapia species in the Guangdong province accounted for 45.6% of the production in China (China Fisheries Statistical Yearbook, 2018). About 10.5 billion tilapia juveniles accounting for 47.6% of the total amounts in China were reared in fish farms in the Guangdong province (China Fisheries Statistical Yearbook, 2018). In conjunction, populations of tilapia species have grown exponentially in the main drainages of the Guangdong province in recent years, especially in the regions where tilapia species are intensively cultivated (Gu et al., 2015).
Human-mediated release of non-native species (e.g. mercy release and unwanted fish release) is considered an important pathway of biological invasion (Copp et al., 2005c; Shiu and Stokes, 2008). Repeated and mass release by mercy release practitioners can facilitate the establishment of non-native species (e.g. Bullfrog) (Liu et al., 2012). The frequency of occurrence of non-native fishes in aquarium shops is significantly correlated with the likelihood of establishment in the introduced areas (Duggan et al., 2006). Copp et al. (2010) also found that human population density and the number of fish imports are the best predictors of propagule pressure. Britton and Gozlan (2013) further identified the threshold of propagule size required by invasive fish for successful establishment. These findings suggest that direct and indirect human-mediated release can facilitate the spread of NN fishes by increasing the propagule size and numbers in new environments (Lockwood et al., 2005). However, research about propagule pressure of non-native fishes are insufficient relative to other invasive taxa (Duggan et al., 2006). Further studies could be conducted to investigate the factors that affect the propagule size (e.g. intensity of aquaculture and aquarium activities, climate change and hydrological connection), which will be useful for the management of non-native fish.
The plasticity of life-history traits of NN species in new habitats has been well examined to explain their invasion success (Fox and Copp, 2014). Wei et al. (2017) demonstrated that size at maturity and body condition of Pterygoplichthys spp. varied among different populations due to variation in food supply and temperature among drainages of the Guangdong province. Pterygoplichthys spp. appeared to mature at smaller size in the Guangdong province than in other invasive regions (Wei et al., 2017). Pterygoplichthys spp. matured later when total nitrogen increased in the rivers, which suggests that these NN species could adjust their life-history traits to increase fitness in stressful environments (Wei et al., 2018). Cui et al. (2016) found that fecundity of G. affinis was higher in farmlands than in ponds, and diameter of matured eggs was smaller in farmland than in ponds. It is suggested that G. affinis adopts K-strategies in ponds where water level and food supply are relatively stable, and takes r-strategies in farmlands because of frequent human activities (Cui et al., 2016).
Plasticity in life history traits appears to be driven by abiotic factors, which can help NN species to survive and spread in new environments (Stearns and Koella, 1986). Clinal variations of life-history traits were observed in fish species which can be explained by temperature, precipitation, prey occurrence and exploitation (Wilson et al. 2019). Also, genetic admixture of divergent intraspecific linages and hybridization of congeneric species could explain plasticity of these quantitative traits (Rius and Darling, 2014; Hodgins et al., 2018). Admixture through multiple introduction of a species or inter-and intraspecific hybridization increases genetic variation, which might be favorable for invasive species to overcome detrimental mutation through superior alleles or a more robust mix of alleles, resulting in colonization success and population expansion (Rius and Darling, 2014; Hodgins et al., 2018). Kahilainen et al. (2011) reported that the hybrid and backcross of native whitefish (Coregonus lavaretus) and introduced vendace (C. albula) had higher growth rates and fecundities. Muhlfeld et al. (2014) also demonstrated that NN rainbow trout (Oncorhynchus mykiss) admixture spread rapidly within the Flathead River system, which can be explained by decreased precipitation and increased temperature in the streams. Facon et al. (2005) found that hybrid lines of invasive freshwater snail Melanoides tuberculate have significant variation in life-history traits (e.g. growth, fecundity and juvenile size), which increases its invasiveness. These findings suggest that both environmental and genetic admixture can affect trait variation of fish species. Further studies could be conducted to test whether and how genetic admixture could affect invasive fishes to adapt to rapid environmental change under climate warming and intensive urbanization scenarios.
Biotic/abiotic resistance or facilitation
Invasion success of NN species is affected by biotic or abiotic factors in NN ranges (Copp et al., 2005a; Davies et al., 2005). Gu et al. (2018) indicated that the biomass of invasive fish decreased with increasing richness of native fish species. Fish communities were dramatically affected by extreme climate or anthropogenic disturbances, which provided “windows” for invasive species’ colonization (Diez et al., 2012; Mouillot et al., 2013). These factors may lead to biodiversity loss, in turn resulting in fish invasions. Studies have also argued that biotic resistance has little impact on fish invasion (Moyle and Light, 1996). Ricciardi and Mottiar (2006) further demonstrated that invasion of NN fishes was not always successful in new communities where there were close phylogenetic relatives (not consistent with pre-adaption hypothesis) due to less intensive competition in fish communities.
Resource availability in the environment plays an important role in the successful establishment of NN species (Moyle and Light, 1996; Davis et al., 2000). A recent study suggested that female Pterygoplichthys spp. matured at larger sizes with an increase in total phosphorus in rivers, whilst males matured at larger sizes with an increase of total nitrogen (Wei et al., 2018). This result suggested that Pterygoplichthys spp. might benefit from nutrient enrichment in rivers that provide abundant food supply for these NN and hypoxia tolerant fishes. Additionally, Pterygoplichthys spp. matured at larger sizes in its NN range than its native range. This might be ascribed to the variation of resource availability in the wet season and presence of predators in the native range, resulting in higher growth rates at juvenile stage (de Mérona and Rankin-de-Mérona, 2004; Wei et al., 2017). This was consistent with a global life-history traits analysis also demonstrating that invasive fish exhibit larger body size and delayed maturation (Liu et al., 2017). It seems that invasive fish might benefit from escaping from the inhibition of natural enemies or detrimental abiotic factors (e.g. low-water level in dry season) in their native range, resulting in high growth rate and fecundity in the invasive range (Keane and Crawley, 2002).
Ecological consequences of non-native fish introductions
Numerous studies have demonstrated that fish invasion can result in adverse impacts on local ecosystems and/or socio-economic activities (Gozlan et al., 2010; Gallardo et al. 2015). By way of example, the invasion of tilapia species in the Guangdong province reduced the catch-per-unit- effort of native fish (Gu et al., 2015). Furthermore, fishermen’s incomes decreased when the capture of tilapia species increased (Gu et al., 2015). Local fishermen also observed a decline of economically important taxa (e.g. shrimp) due to invasion of Pterygoplichthys spp. (Wei et al., 2017).
Invasive fish could compete with native species for resources or prey upon native species, resulting in biodiversity loss. Invasive fish G. affinis preyed on the early stages of native protected fish Tanichthys albonubes (Chen, 2010). Gambusia affinis also disturbed the reproduction of T. albonubes by chasing and biting the adults (Chen, 2010). A recent study has also suggested that NN fish species could significantly decrease functional diversity of native fish species, thereby de-stabilising aquatic ecosystem function and structure (Shuai et al., 2018).
The invasion of NN fishes can also affect water quality and nutrient dynamics in freshwater ecosystems due to their unique stoichiometric traits (Figueredo and Giani, 2005; Capps and Flecker, 2013). Studies indicated that phosphorus excretion of Pterygoplichthys spp. were comparatively lower than native species and thus could act as phosphorus sinks, thereby increasing the severity of phosphorus limitation of periphyton (Capps and Flecker, 2013; Datri et al., 2015; H. Wei unpublished data). The presence of O. niloticus increased nitrogen and phosphorus availability via excretion, resulting in proliferation of phytoplankton populations (Figueredo and Giani, 2005), whilst, water quality was found to benefit from removing a population of tilapia in a hypereutrophic reservoir (Starling et al., 2002).
Regulations regarding to non-native freshwater fish in the Guangdong province
In China, legislation regarding NN species is included in a number of Acts comprising the Quarantine Act, Environmental Protection Act, Agriculture Act and Wildlife Protection Act. Some statutory documents about environmental protection have stipulated that quarantine or scientific assessment shall be conducted before importing NN species. The Fishery Act and Regulation on Aquaculture Seedlings have stipulated that trans-provincial and trans-continental importation of aquaculture seedlings that are not approved by the department of fishery administration is prohibited. In practice, each of 31 provinces and municipalities in mainland China have developed their own regulations for the introduction or release of NN aquaculture propagules. These regulations are supplementary to the national regulations (e.g. Fishery Act) and are tailored to the local situation. Despite local adaptations, legislations are relatively consistent amongst different Chinese regions.
In the Guangdong province, risk assessments shall be conducted before the introduction of NN freshwater fishes, and high risk species shall not be imported according to the Fishery Management Regulation of Guangdong Province. This regulation also stipulates that a person or corporate body who plans to introduce NN aquatic species from abroad shall be approved by the Department of Fishery Administration, otherwise, the seedling shall be confiscated and the person or corporate body shall be fined up to 30,000 CNY (approx. 4361 USD). The regulation on Aquaculture Seedlings of Guangdong province also stipulates that a person or corporate body must provide reports on the biology of the introduced species, and environmental conditions in its native range, as well as set-up of quarantined areas in the introduced site before its introduction. However, similar to EU, the enforcement of these regulations in China is difficult due to the limits of political or economic considerations (Genovesi et al. 2015). Further efforts should make to define the regulatory responsibilities of different agencies to improve the efficacy of enforcement.
A total of 160 NN freshwater fishes were recorded in the import/trade records in the Guangdong province and 71.9% of them were introduced for aquarium purposes. Fourteen species have established self-sustaining populations and 21 species were recorded in the main river basin of Guangdong province. Although it doesn’t represent a comprehensive dataset due to the lack of a comprehensive species list relating to introduction activities, trade or monitoring programmes to detect NN species, the current dataset indicates that the Guangdong province will be at a high risk of fish invasions in the future owing to its benign climate which allows these NN species to survive or colonize in this region (Kottek et al., 2006). In this regard, risk assessment, statutory regulations, public education and scientific research should be developed to prevent future fish invasions.
Balancing the conflicts between socio-economic benefits and ecological risks becomes a challenging task for policy makers (Gozlan and Newton 2009; Britton and Orsi 2012). Risk assessment or horizon scanning should be conducted for NN species by a group of multi-disciplinary experts including fish biologists, policy makers, ecologists and pathologists to provide comprehensive background information to the policy makers (Copp et al., 2014). However, this approach can be costly, which could discourage government agencies to organize such panels, especially if the benefits are not defined and the negative impacts not well-studied. Therefore, rapid and user-friendly screening toolkits could be developed to provide preliminary results for policy makers (Genovesi et al., 2015; Roy et al., 2017).
Furthermore, studies should also be conducted to investigate the mechanisms driving invasions, and to assess their ecological consequences. Such information could be helpful for the policy makers to weigh the costs and benefits of intentional fish introductions, so that actions for management can be implemented.
Global proactive and reactive abilities to manage introduced species are different and unbalanced (Early et al., 2016). Therefore, it is recommended that countries should take collaborative actions, especially in neighboring regions, to save on the costs of fighting invasions (Charles et al., 2010; Singh and Lakra, 2011; Ricciardi et al., 2017).
This study was conceived during the international workshop “Marine and Freshwater Invasive Species: Solutions for water security (MFIS-China)”, held in Beijing, China on August 27-29, 2018. We appreciate the help from the organizers of this workshop, especially Dr Mohiuddin Munawar and Dr Aibin Zhan. Thanks also to Dr Gordon Copp, Dr Darren Yeo and Dr Kit Magellan whose presentations in the workshop sparked the idea for the manuscript. We appreciate Dr Amanda Arnold of Queen Mary University of London for her efforts in improving the language of the manuscript. We also appreciate two anonymous reviewers for their edits and comments to improve the manuscript.
Supplementary data for this article can be accessed on the publisher's website.
This work was funded by the Natural Science Foundation of Guangdong Province (Grant No. 2017A030310669 and 2016A030313145), Central Public-interest Scientific Institution Basal Research Fund, CAFS (NO. 2018GH11) and National Natural Science Foundation of China (Grant No. 31700473).
(Please see additional references listed as supplementary material on the publisher's website)