Hongxia LUO,Pimao CHEN,Xiaoguo LI,Huarong YUAN,Lianlian WANG,3,Xue FENG

1.South China Sea Fisheries Research Institute,Chinese Academy of Fishery Science/Scientific Observing and Experimental Station of South China Sea Fishery Resources and Environment,Ministry of Agriculture/Key Laboratory of Marine Ranch Technology,Chinese Academy of Fishery Science,Guangzhou 510300,China;2.College of Marine Sciences,Shanghai Ocean University,Shanghai 201306,China;3.School of Marine Sciences and Environment Engineering,Dalian Ocean University,Dalian 116023,China

Anthocidariscrassispinabelongs to Echinodermata,Echinoidea, Camarodonta and Echinometridae,and is mainly distributed in southeast coast of China[1].A.crassispinahas high nutritional and medicinal value[2-4].Since the 1980s,China’s coastalfishery resources have been severely recessed due to overfishing,environmentalpollution,marine construction,climate change and other factors.A.crassispinaresource has also been severely damaged.Therefore,it is urgent to restore and conserveA.crassispinaresource by propagation and release.In order to protect local germplasm and to avoid invasion of exotic germplasms,genetic structure analysis and local species identification ofA.crassispinaare the premise of propagation and release.

Currently,the studies onA.crassispinamostly focus on nursery and farming[1,3,5].In terms of genetic diversity,Changet al.[6]and Liuet al.[7]in vestigated the genetic relationships a mongA.crassispina,Glyptocidaris crenularis,Hemicentrotuspulcherrimus,Strongylocentrotus nudusandS.intermediusby molecular biology methods.There are rare reports on morphological variations amongA.crassispinagroups.In this study,the sixmorphometric traits ofsixA.crassispinagroups from South China Sea were measured,and the morphological variations among them were analyzed comprehensively using four multivariate analyses,including one-way ANOVA analysis,variation coefficient method,cluster analysis and discriminant analysis.This study will provide certain reference and basic information for future restoration and rational utilization ofA.crassispinaresources.

Materials and Methods

Test materials

The test materials included six wildA.crassispinagroups collected from Daya Bay(DY),Egong Bay(EG),Hailing Island(HL)and Nan’ao Island(NA)in Guangdong Province,Weizhou Island(WZ)in Guangxi Province and Ningde City (ND)in Fujian Province during January to June,2014.The geographical locations of collection sites were shown in Fig.1,and the specific collection sites,times and quantities were shown in Table 1.

Measurement methods

Table 1 Sampling information about A.crassispina groups

The shell diameter(SD),mouth diameter(MD)and shell depth(SDH)ofA.crassispinawere measured accurately using a vernier caliper(accurate to 0.01 mm).To reduce the error caused by individual irregular shape,the shell diameter and mouth diameter ofA.crassispinawere all measured twice,and the final data were expressed as the means of the two replicates.The shell depth ofA.crassispinawas defined as the distance between the highestand lowest points.The mouth weight(MW),total weight(TW)and shell weight(SW)ofA.crassispinawas weighed by an economic scale(accurate to 0.01 g).

Data processing and analysis

For single factors,one-way ANOVA analysis was conducted.The morphological variations among differentA.crassispinagroups were analyzed by LDS,Duncan’s and Tamnaen’s T2 multiple comparisons[8].The variation coefficient was calculated according to the following formula:

Variation coefficient(% )=Standard deviation/Mean×100.

According to the 75% theorem of Maryet al.[10],the trait difference among populations can be calculated by certain trait.Thus the difference between two populations can be compared with the interspecific level.The variation coefficient (CD)was calculated according to the following formula:

In the formula,MiandMjrepresent the means of certain trait in the two populations;SiandSjrepresent the standard deviations of certain trait in the two populations.IfCDvalue is higher than 1.28,the difference between the two populations is judged to reach the interspecific level;but if theCDvalue is lower than 1.28,the difference is judged to reach the inter-population level.

The multivariate analysis among the sixA.crassispinagroups was conducted by cluster analysis method and Euclidean distance method.Based on the calculated means,the dendrogram was drawn.In addition,the Euclidean distancesamongpopulationswere calculated[11].

For each parameter,discriminant function was established.And then,the discriminant analysis was conducted using parameters and established discriminant functions.The typical score plot of discriminant functions was constructed[12].In addition,the discriminant accuracy and comprehensive discriminant accuracy were calculated[13].

Discriminant accuracyP1(% )=(Quantity of accurately identifiedA.crassispinain certain population/Actual quantity ofA.crassispinain certain population)×100;

In the formula, Airepresents quantity ofaccurately identifiedA.crassispinaindividuals in populationi;Birepresents actual quantity of identifiedA.crassispinaindividuals in populationi;k represents population number.

Results and Analysis

One-way ANOVA analysis

The mean,standard deviation,variation coefficient and one-way ANOVA analysis of each trait in each population were shown in Table 2.The means of all the traits,except mouth diameter,ofA.crassispinafrom Hailing Island were smaller than those of other five groups.However,the means of all the traits ofA.crassispinapopulation from Weizhou Island were larger than those of the other five groups.The variation coefficients of all the morphometric traits,except mouth diameter and mouth weight,ofA.crassispinapopulation from Data Bay were larger than those of the five groups.

There were significant differences in the six morphometric traits among the sixA.crassispinagroups.TheA.crassispinapopulation from Weizhou Island showed significant differences in all the six morphometric traits compared with the other five groups.However,between theA.crassispinagroups from Daya Bay and Nan’ao Island,there was only one morphometric trait in which the difference reached the significantlevel(Table 2).

Variation coefficient analysis

The variation coefficients amongA.crassispinagroups from different origins were calculated(Table 3).The variation coefficient of mount weight between theA.crassispinagroups from Daya Bay and Hailing Island was higher than 1.28,and of shell weight between theA.crassispinagroups from Nan’ao Island and Hailing Island was higher than 1.28.The variation coefficients of shell diameter,shelldepth,total weight and shell weight between the groups from Hailing Island and Weizhou Island were all higher than 1.28.Therefore,the differences above all reached the interspecific level.

Table 2 Morphological characters of A.crassispina

Table 3 Variation coefficients of morphometric traits among six A.crassispina groups

Cluster analysis

The cluster analysis (Fig.2)showed that theA.crassispinagroups from Egong Bay,Nan’ao Island and Daya Bay were first clustered into one group,indicating similar morphological characters.The threeA.crassispinagroups above were then clustered into one group with that from Hailing Island.Finally,the fourA.crassispinagroups above were clustered into one group with those from Weizhou Island and Ningde City.TheA.crassispinapopulation from Daya Bay showed the closest genetic distances with those from Nan’ao Island and Egong Bay,followed by that from Hailing Island,and it showed the farest genetic distances with those from Weizhou Island and Ningde City.The genetic relationships between Daya Bay’sA.crassispinapopulation and the other five groups were consistent with the geographical distances between them.

Discriminant analysis

Table 4 showed the discriminant results of the six morphometric traits among the sixA.crassispinagroups from different origins.In terms of discriminant accuracy,the sixA.crassispinagroups ranked as Hailing Island＞Ningde City＞W(wǎng)eizhou Island ＞Egong Bay＞Nan'ao Island ＞Daya Bay.The comprehensive discriminant accuracy was 65.3%.The established discriminant functions were as follows:

F1=6.976X1+1.061X2-13.487X3-1.179X4+7.669X5-1.174X6-150.194;

F2=6.822X1+1.371X2-9.470X3-1.246X4+9.084X5-1.316X6-160.372;

F3=6.986X1+0.934X2-14.655X3-1.209X4+10.141X5-1.266X6-162.252;

F4=6.937X1+0.919X2-12.465X3-1.260X4+9.242X5-1.102X6-158.194;

F5=7.411X1+1.172X2-16.958X3-1.209X4+9.512X5-1.198X6-184.044;

F6=7.640X1+1.053X2-12.448X3-1.332X4+10.410X5-1.380X6-193.934.

F1-F6represent the discriminant functions ofA.crassispinagroups from Daya Bay,Egong Bay,Hailing Island,Nan'ao Island,Weizhou Island and Ningde City,respectively.X1-X6represent shell diameter,shell depth,mouth diameter,total weight,mouth weight and shell weight,respectively.According to the established discriminant functions, the sixA.crassispinagroups from different origins could be classified primarily.

Table 4 Results of discriminant analysis among the six A.crassispina groups

As shown in Fig.3,theA.crassispinagroups from Daya Bay,Egong Bay and Nan’ao Island basically staggered together,and they were distributed over a relatively large area.TheA.crassispinaindividuals from Hailing Island were distributed in the center of the three groups above.TheA.crassispinapopulation from Ningde City was partly overlapped with the four groups above.However,theA.crassispinapopulation from Weizhou Island was distributed in an independent area,and it showed no intersection with the other five groups.

Conclusions and Discussion

Multivariate analysis is widely applied in analysis and studies of morphological differences among aquatic biological communities.Wanget al.[12,14]studied the morphological differences among threeMonodonta labiogroups and sixCoilia nasusgroups;Hanet al.[15]investigated the morphological differences among 4Procambarus clarkiigroups;Gaoet al.[16]investigated the morphological differences among 4Portunus trituberculatusgroups.In this study,the morphological differences among sixA.crassispinagroups from different origins were analyzed by one-way ANOVA analysis,variation coefficient method,cluster analysis and discriminant analysis,thereby providing certain reference for future relevant studies and analysis onA.crassispina.

One-way ANOVA analysis is a relatively basic difference analysis method.The results of this study showed that no significant differences were found in five out of the six morphometric traits between theA.crassispinagroups from Daya Bay and Nan’ao Island;theA.crassispinapopulation from Weizhou Island showed significant differences in all the morphometric traits compared with the other five groups.It suggests that the morphology ofA.crassispinagroups from Daya Bay and Nan’ao Island is the most similar,but theA.crassispinapopulation from Weizhou Island is extremely different compared with the other fiveA.crassispinagroups.According to Mayr’s 75% theorem,variation coefficient of certain trait between two groups is calculated,and it is used to judge whether the variation between two groups reaches the interspecific level.This study found that the variation coefficients of four morphometric traits between theA.crassispinagroups from Weizhou Island and Hailing Island were higher than 1.28,so the morphological differences between the two groups is judged to reach the interspecific level.Cluster analysis collects data and classifies them by utilizing the similarities among data sets.Based on the calculated-Manhattan distances or Euclidean distances of certain trait among several populations,dendrogram isdrawn.From the dendrogram,the genetic relationships among the groups can be observed directly[9].The smaller the distance is,the more similar the morphology is.In this study,the genetic distances among theA.crassispinagroups from Daya Bay,Egong Bay and Nan'ao Island were smallest,so they were clustered into one group.Subsequently,they were clustered into one larger group with that from Hailing Island.Finally,the fourA.crassispinagroups were clustered into one group withthosefrom Ningde Cityand Weizhou Island.The genetic relationships among the sixA.crassispinagroups showed certain correlation with their geographicaldistances. The smaller the geographical distance is,the closer the genetic relationship is.According to certain criterions,discriminant analysis establishes discriminant functions by utilizing certain traits.And then,the discriminant values are calculated.Based on the calculated discriminant values,the genetic relationships among certain groups area analyzed.The results of this study showed that the discriminant accuracies ofA.crassispinagroups ranged from 47.8% to 86%,and the comprehensive discriminant accuracy was 65.3%.The discriminant accuracies ofA.crassispinagroups from Daya Bay,Egong Bay and Nan’ao Island were lower.On one hand,the morphological differences among the threeA.crassispinagroups were smaller,and the discriminant was interspecific identification.On the other hand,the closer geographical distances might lead to stagger among the groups.In the discriminantfunction,mouth diameter and mouth weight were of great importance forA.crassispina.The six morphometric traits not only provide important reference for population discriminant ofA.crassispina,but also can be used as important indicators in future breeding process.

This study showed that the results of the four kinds of analysis methods were highly consistent.The morphology ofA.crassispinagroups from Daya Bay,Egong Bay and Nan’ao Island was most similar.The similarity of morphologybetweentheA.crassispinapopulation from Hailing Island and the other five groups ranked second.TheA.crassispinapopulation from WeizhouIsland showed the greatestdifference in morphology compared with the other five groups.The variation degree amongA.cras-sispinagroups might be correlated with their geographical distribution to some extent.Weizhou Island is far apart from the other waters due to the existence of Leizhou Peninsula,so theA.crassispinapopulation from Weizhou Island was extremely different from those from the other origins.Ningde is also far from the other five origins,so theA.crassispinapopulation from Ningde City also showed significant differences in morphology compared with the other five groups.The geographical distances among the remaining groups are closer,so the morphological differences among them were also smaller.A.crassispinais a kind of benthic organism,and it usually perches between rocks and stones.TheA.crassispinaindividual migration is weak,and it can only move a few centimeters every five minutes[1].Therefore,genetic exchange betweenA.crassispinagroups is likely to be limited by geographical distance[17].Since the existence of geographical barriers,one population is isolated with other populations of the same species,and it loses the chance to mate with other groups.In longterm geographical isolation,different population accumulates certain variation,leading to differences in morphology,ecology,physiology and genetic molecule among different populations[18].In addition to geographical isolation,the morphological difference among different populations may also caused by the difference in living environment[19-20].Ningde City has higher altitude,lower water temperature and shorter sunshine time.In the bay where Ningde City is located,there is river emptying into the sea,so the water salinity is relatively low.In addition,the bay greatly extends into the sea.Due to the blocking by Leizhou Peninsula,water velocity in Weizhou Island is relatively low.These environmental differences may affect the growth ofA.crassispina,leading to morphologicaldifferences among differentA.crassispinapopulations.Nan’ao Island is far from Daya Bay,so frequency of genetic exchange between twoA.crassispinagroups is low in theory.However,the cluster analysis showed that there was small difference between the twoA.crassispinagroups from Nan’ao Island and Daya Bay.The one-way ANOVA analysis also showed that there were no significant differences in five out of the six morphometric traits between the two groups.On one hand,the living environments of the twoA.crassispinagroups are similar.There are no rivers diluting the salinity of the sea,so their salinities are all relatively high.On the other hand,in recent years,a lot of propagating and releasing activities ofA.crassispinahave been carried out in coastal area of Guangdong Province,while the parents ofA.crassispinamay come from other areas.The propagation and release of non-nativeA.crassispinaincreases the frequency of genetic exchange among different groups.This further illustrates the importance of researches on genetic relationships among differentA.crassispinapopulations and establishment of methods to identify nativeA.crassispinapopulation.

In researches on comparison and classification of aquatic populations,morphological method is a traditional,intuitive,simple and efficient research method.The analysis on morphological differences amongA.crassispinapopulations can provide certain basic data for identification ofA.crassispinapopulations.However,the morphological differences amongA.crassispinapopulations may also be affected by external environmentto some extent,and they can not accurately reflect the true genetic variations inA.crassispina.Therefore,to comprehensively and accurately identifyA.crassispinapopulations,genetic researches are further needed to be carried out.

[1]CHRN C(陳錘).Biology and culture ofAnthocidaris crassispina(紫海膽的生物學(xué)與養(yǎng)殖)[J].Oceanic and Fisheries(海洋與漁業(yè)),2007(7):32.

[2]LIU H(劉恒).Review on the world sea urchin fishery(世界的海膽漁業(yè))[J].Marine Sciences(海洋科學(xué)),2001,25(3):38-41.

[3]PENG JS(彭景書(shū)),ZHOU WC(周文川),HUANG MH(黃敏紅),et al.Preliminary study on artificial breeding ofAnthocidaris crassispina(紫海膽人工養(yǎng)殖技術(shù)初步研究)[J].Oceanic and Fisheries(海洋與漁業(yè)),2007(9):21.

[4]Jiangsu New Medical College(江蘇新醫(yī)學(xué)院).Traditional Chinese medicine dictionary(中藥大辭典)[M].Shanghai:ShanghaiScience and Technology Press(上海:上?？茖W(xué)技術(shù)出版社),1992.

[5]CHEN YF(陳顏鋒),YOU XT(游小艇).Farming test ofAnthocidaris crassispina(紫海膽?zhàn)B殖試驗(yàn))[J].Scientific Fish Farming(科學(xué)養(yǎng)魚(yú)),2003(10):29.

[6]CHANG YQ(常亞青),DING J(丁君),XING RL(邢榮蓮),et al.Genomic DNA polymorphism in five species of sea urchin by RAPD(用RAPD技術(shù)對(duì)5種經(jīng)濟(jì)海膽基因組DNA多態(tài)性的研究)[J].Journal of Fishery Sciences of China(中國(guó)水產(chǎn)科學(xué)),2004,11(2):129-134.

[7]LIU XH(劉曉慧),HUANG JQ(黃佳琪),ZHOU ZC(周遵春),et al.Sequence analysis of mtDNA of 16S rRNA gene fragments among five 5 species of sea urchins(種經(jīng)濟(jì)海膽線粒體16SrRNA基因片段的序列分析)[J].Fisheries Science(水產(chǎn)科學(xué)),2007,26(6):331-334.

[8]HAN Z(韓真),XIAO YS(肖永雙),GAP TX(高天翔).Comparison of morphological characteristics of 9Larimichthys polyactispopulations in China(中國(guó)近海9個(gè)小黃魚(yú)群體的形態(tài)學(xué)比較研究)[J].South China Fisheries Science(南方水產(chǎn)科學(xué)),2012,8(3):25-33.

[9]DING JQ(丁金強(qiáng)),LIU P(劉萍),LI J(李健),et al.Analysis of morphological variation among four wild populations ofCharybdis japonicafrom coastal waters of China(中國(guó)沿海日本蟳4個(gè)地理群體的形態(tài)差異比較分析)[J].Journal of Fishery Sciences of China(中國(guó)水產(chǎn)科學(xué)),2012,19(4):604-610.

[10]MAYRRE,LINSLEYEG,USINGERRL.Methods and principles of systematic zoology[M].New York and London:Mc Craw-Hill,1953:125-154.

[11]SHUI BN(水柏年),SUN XF(孫希福),HAN ZQ(韓志強(qiáng)),et al.Morphological variation analysis among populations of Japanese spanish mackerelScomberomorus niphoniusin the Yellow Sea and East China Sea(黃海、東海藍(lán)點(diǎn)馬鮫群體的形態(tài)特征分析)[J].Journal of Fisheries of China(水產(chǎn)學(xué)報(bào)),2009,33(3):91-97.

[12]WANG DT(王丹婷),YANG J(楊健),JIANG T(姜濤),et al.A comparative study of the morphology of different geographical populations ofCoilia nasus(不同水域刀鱭形態(tài)的分析比較)[J].Journal of Fisheries of China(水產(chǎn)學(xué)報(bào)),2012,36(1):78-90.

[13]WU YP(吳楊平),YAO GX(姚國(guó)興),CHEN AH(陳愛(ài)華),et al.Multivariate morphometric analysis of two species of Meretrix,wit notes on Japanese Meretrix lusoria’s validity(文蛤?qū)?2 種貝類多變量形態(tài)分析及日本文蛤的物種有效性)[J].Journal of Fisheries of China(水產(chǎn)學(xué)報(bào)),2011,35(9):1410-1418.

[14]WANG X(王旭),XU H(徐衡),ZOU L(鄒莉),et al.Morphological variation analysis of three different geographic populations of Monodonta labio(3個(gè)單齒螺地理群體形態(tài)差異分析)[J].South China Fisheries Science(南方水產(chǎn)科學(xué)),2013,9(4):22-27.

[15]HAN XL(韓曉磊),MA Q(馬強(qiáng)),LI XR(李小蕊),et al.Morphological variations analysis of different geographic populations ofProcambarus clarkii(不同地區(qū)克氏原螯蝦群體的形態(tài)差異分析)[J].Journal of Hydroecology(水生態(tài)學(xué)雜志),2011,32(3):82-87.

[16]GAO BQ(高保全),LIU P(劉萍),LI J(李健),et al.Analysis of morphological variations among four wild populations ofPortunus trituberculatus(三疣梭子蟹4個(gè)野生群體形態(tài)差異分析)[J].Journal of Fishery Sciences of China(中國(guó)水產(chǎn)科學(xué)),2007,14(2):223-228.

[17]SONG FF(宋菲菲),WANG QX(王其翔),LIU M(劉名),et al.Morphological variation analysis ofScapharca subcrenatapopulations in China(不同地區(qū)毛蚶群體的形態(tài)差異的比較研究)[J].Journal of Ocean University of China(中國(guó)海洋大學(xué)學(xué)報(bào) (自然科學(xué)版)),2012,42(11):40-45.

[18]LI Y(李勇),LI SF(李思發(fā)),WANG CH(王成輝),et al.Establishment and application of morphological discrimination modelforjuveniles Eriocheir sinensis from Liaohe,Yangtze and Qujiang rivers(三水系中華絨螯蟹幼蟹形態(tài)判別程序的建立和使用)[J].Journal of Fisheries of China(水產(chǎn)學(xué)報(bào)),2001,25(2):120-126.

[19]KONGLF,LIQ,QIUZX.Genetic and morphologicaldifferentiation in the clamCoelomactraantiquate(Bivalvia:Veneroida)along the coast of China[J].Exper Mar Biol Ecol,2007(343):110-117.

[20]PLEJDRUPJK,SIMONSENV,PERTOLDIG,et al.Genetic and morphological diversity in populations ofNucella lapillusL.(neogastropoda)in response to tributyltin contamination[J].Ecotoxicol Environm Saf,2006(64):146-154.