CAI Xing-Wei ， LI Wei LI Zhong-Jie ZHANG Tang-Lin YE Shao-Wen FAN Hou-Rui and LIU Jia-Shou

(1. State Key Laboratory of Freshwater Ecology and Biotechnology， Institute of Hydrobiology， Chinese Academy of Sciences，Wuhan 430072， China; 2. University of Chinese Academy of Sciences， Beijing 100049， China)

Abstract: This study is based on monthly sampling (from April 2013 to April 2015) of 18 fish species representing three families. Samples were obtained using benthic fyke-nets， block nets， and multi-mesh gillnets in Lake Kuilei， a shallow lake of the Yangtze River basin， China (N 31°24′ ， E 120°51′). Using ordinary leastsquares linear regression， length-weight relationships of these fish species were estimated， with equations for 3 species reported for the first time. The maximum total length records of the 15 species exceeded those reported in FishBase. This paper provides new LWR equations for these species to be included in FishBase.

Key words: Growth; Length-weight relationships; Shallow lake; Lake Kuilei

Length-weight relationships (LWRs) are extremely important aspects of fisheries biology[1，2].These relationships are often reported from the parameters from regression equations or mathematical power functions. These equations are used for many miscellaneous purposes including stock assessment，routine monitoring， and as components of fisheries management programs[3]. Equations also provide necessary data for comparing life-history characteristics of fishes from different areas or habitats[4—6].

The Yangtze River is the third largest river in the world. There are many freshwater lakes distributed along the Yangtze River basin， which cumulatively account for about 60% of the total surface area of all of the freshwater lakes in China[7]. Historically， these lakes were connected directly or indirectly with the Yangtze River main stem， forming a series of potamo-lacustrine ecosystems that were highly unique in China[8]. These lakes are typically shallow， do not thermally stratify， have abundant vegetation， and exhibit relatively high biological productivity and fish diversity[9，10]. Because of the important ecological functions that these potamo-lacustrine ecosystems serve and their high levels of biodiversity， the World-Wide Fund for Nature (WWF) has designated the Yangtze River and its adjacent lakes as one of the Global 200: Priority Ecoregions for Global Conservation[11].

Knowledge about the biological characteristics of fishes resident to these shallow Yangtze River lakes is still very limited[12，13]. Previous reports on LWR equations have been largely focused on largebody fish species of commercial interest or from reaches in the upper basin[14—17]. On the other hand，LWR information of smaller-bodied fishes is relatively scarce， with little literature available. To address this information gap， intensive fish sampling was conducted in the lower reaches of the Yangtze River basin where by 18 fish species were collected and analyzed to estimate length-weight relationships.These analyses also included 3 species that had not previous published LWR equations.

1 Materials and Methods

The study was conducted at Lake Kuilei (N 31°24′， E 120°51′) in Jiangsu Province of Eastern China. This lake has a surface area of 670 ha and an average depth of 2.74 m. This lake is characteristic of shallow lakes in the lower Yangtze River basin， and also is a drinking water source for Kunshan City. The physico-chemical characteristics of the lake were monitored from April 2013 to April 2015 and are presented in Tab. 1.

Tab. 1 The mean， standard deviation (±SD) and range of the physico-chemical parameters monitored in Lake Kuilei

Fishes were sampled monthly from April 2013 to April 2015 using benthic fyke-nets (long 15 m，stretched mesh size 0.4 cm)， multi-mesh gillnets (long 30 m， high 1.5 m， stretched mesh size ranging from 1.0—12.5 cm)， and block nets (stretched mesh size 4 cm). The benthic fyke-nets and block nets were deployed between AM 8:30 and 10:30 hours， and the catches from the end pockets were collected after 24 hours; The multi-mesh gillnets were deployed at PM 18:00， and were collected at AM 6:00 next day.All the fishes collected were placed into plastic bags，labeled， refrigerated， and later identified and analyzed. In the laboratory， sampled fish specimens were identified to species level following Chen (1998) and Ni (2005) and enumerated[18，19]. Total length to the nearest 0.1 cm somatic weight to the nearest 0.01 g were measured. All scientific names of fishes were checked against FishBase[6]. The relationship between body weight and total length was calculated using the equation:W=aLb， whereWis the body weight (g)，Lis the total length (cm)，ais the regression intercept parameter andbis the regression slope parameter.Outliers were identified and excluded according to the method of Froese (2006)， with 95% confidence limits ofaandbestimated using a bootstrapping procedure based on 1000 iterations.

2 Results

The length and weight measures of 15864 specimens representing 18 species were used to estimate length-weight relationships. All statistical descriptions of LWR parameters for each species are provided in Tab. 2. Records in FishBase[6]indicates this study provides new LWR records for 3 species denoted with*in Tab. 2. Sample sizes ranged from 33(Taenioides cirratus) to 1821 (Toxabramis swinhonis).Coefficients of determinationr2were all high， ranging from 0.910 (Hemiculter leucisclus) to 0.995(Sarcocheilichthys sinensis). Estimates ofbranged from 1.6678 (T. cirratus) to 3.3780 (T. swinhonis)，and averaged 3.0183 (Tab. 2). Maximum lengths of 15 species exceeded the FishBase records and are shown in bold type in Tab. 2.

3 Discussion

Compared to estimates contained in FishBase[6]，no LWR information was available forMicrophysogobio microstomus，S. sinensisandT. cirratus. In addition， revised (i.e.， greater) maximum total lengths are presented for 15 species， includingParacheilognathus imberbis，Acheilognathus macropterus，Sarcocheilichthys nigripinnis，Hemibarbus maculates，Odontobutis obscura，Culter mongolicus，Pseudorasbora parva，Pseudobrama simoni，T. swinhonis，Paracanthobrama guichenoti，Acheilognathus chankaensis，Hemiculter leucisclus，Cultrichthys erythropterus，Rhinogobius giurinusandMicropercops swinhonis.The present study indicated thatbestimates for all species exceptT. cirratusfell within the range of 2.5—3.5[3]， which is normal for the vast majority of freshwater fishes. AlthoughT. cirratushad an estimatedbof 1.6678， the aberrant value was probably due its small， elongated shape. Additionally， there are some differences inbestimates calculated from the present study compared to other authors[12，16，17，20—23]for 10 other fishes. These species includedP. imberbis，H. swinhonis，H. leucisclus，P. simoni，S. nigripinnis，T. swinhonis，P. Parva，P. guichenoti，C.erythropterusandA. chankaensis. However， thesebestimates for these 10 species are in general agreement with other studies， which suggest that thebparameter may be affected by factors such as habitat，length ranges used， sex， stomach fullness， sample size， diet， preservation techniques， and annual differences in environmental conditions[3，16，24，25]. In the present study， we were not able to account for these factors in our estimations.

In conclusion， this study successfully provided new and revised information on LWRs for 18 fish species from the Yangtze River basin， which should be useful for sustainable utilization， management， and protection of fishery resources in lower Yangtze River lakes of China.

Tab. 2 Length-weight relationships for 18 fish species collected Lake Kuilei， China during 2013—2015

Acknowledgements

We would like to acknowledge Chuansong Liao，Geng Huang， Jixin Yu， Xianghong Dong and all other Colleagues for their assistance in the field samping.