Morphometric Variations and Fishery Unit Assessment of Cyclocheilichthys apogon (Actinopterigii: Cyprinidae) from Three-Different Rivers in North-Eastern Thailand

Anan Kenthao and Pornpimol Jearranaiprepame*

Department of Biology, Faculty of Science, Khon Kaen University, Mueang, Khon Kaen 40002, Thailand

ABSTRACT

Beardless barb, Cyclocheilichthys apogon (Valenciennes, 1842) is a freshwater fish of importance as source of low-cost protein in Lower Mekong Region. The present study applied multivariate morphometric technique to identify fishery management units of C. apogon from six populations of three different river drainages: Pong, Chi, and Mun Rivers. Thirty-two truss measures and standard length were obtained using digital calliper from 291 fish individuals, and raw measured data were then subjected to allometric equation to remove size-dependent variation prior further statistical analyses. Multivariate analysis of variance (MANOVA) indicated highly significant differences in morphometric characters between populations (p < 0.01). The first three principal axes of principal component analysis (PCA) explained 49.29% of total variance. The PCA also revealed that morphological variations related to the characters of head depth, body length, body depth, caudal peduncle length and depth. In discriminant function analysis (DFA), the first two discriminant functions accounted for 72.00% of total variation, and discriminated fish samples into three major groups following to their collecting drainages. Furthermore, 96.29 and 90.56% of fish samples were correctly classified into their respect populations with original and cross-validated tests, respectively. The reliable morphometric variations in the present study suggest that management unit of C. apogon should define relied on the isolation of river drainage. Moreover, the study also indicates the involvement of environmental conditions in morphological adaptation, providing useful information for the sustainable conservation of this fish.

Article Information

Received 22 August 2016

Revised 02 March 2017

Accepted 01 April 2017

Available online 01 January 2018

Authors’ Contribution

AK collected samples, performed laboratory analysis, and wrote the article. PJ supervised the work, helped in analysis, and wrote and revised the article.

Key words

Morphological variation, Morphometrics, Ecomorphology, Cyprinids, Fishery management unit.

DOI: http://dx.doi.org/10.17582/journal.pjz/2018.50.1.111.122

* Corresponding author: [email protected]

0030-9923/2018/0001-0111 $ 9.00/0

Copyright 2018 Zoological Society of Pakistan

INTRODUCTION

 

Beardless barb, Cyclocheilichthys apogon (Valenciennes, 1842) is an important food fish particular in lower Mekong region. By having beautiful body colour and patterns, this fish is popular in aquaculture as the ornamental fish (Rainboth, 1996; Vidthayanon, 2012). Thedistribution of this fish species is throughout southeast Asia (Kottelat, 2001; Vidthayanon, 2012). It is usually found in a various habitats from small pond to large lake (Rainboth, 1996) as well as small- throughout large-sized rivers (Kottelat, 2001). The wide range of distribution of this fish probably indicate the formation of stock units based on different ecological conditions of each separated habitats (Akbarzadeh et al., 2009).

Understanding of population structure is an important consideration in developing plans for effective fishery management programmes (Cronin-Fine et al., 2013; Hoggarth, 2006). Each population stocks usually have specific biological attributes that must be taken into account in fishery management (Secor, 2014). The appropriate management plans for each fishery stock will yield high production (Begg et al., 1999) as well as protect population diversity for further fishery use (Turan, 2004). Lacking of knowledge on population biology can also be problematic issues such as the loss of genetic diversity, the decrease of population, and overfishing (Begg et al., 1999; Smith et al., 1991).

Many techniques are applied for understanding population structure in many fish species in order to plan for further beneficial use (Cardin et al., 2005; Hoggarth, 2006). Morphometric analysis especially the truss network technique is frequently used in identification of stock and population structure in various animals including fish (Cardin et al., 2005; Pazhayamadom et al., 2015). The basis of truss network method involved with the measurements of distances between anatomical landmarks forming and reticular network covering the entire fish body (Strauss and Bookstein, 1982; Turan, 1999). The effective in capturing morphological variations regarding shape variability rather than traditional measurement (Cavalcanti et al., 1999) has made the truss network method to become in use more frequently in describing and identifying morphological variability of intraspecific fish groups (Mir et al., 2013; Pazhayamadom et al., 2015).

The objectives of this study were to examine morphological variations of C. apogon habiting in different geographical locations in north-eastern Thailand. A truss network morphometric method was selected and applied on the fish external morphology to provide meaningful information in an identification of fishery units for further use in conservation and management purposes.

 

MATERIALS AND METHODS

Sample collection

Samples of C. apogon were collected in total of 291 individual fish from six populations of three-different rivers in north-eastern Thailand (Table I, Fig. 1). Two populations of 19 and 37 fish from Ubolratana dam (URD1 and URD2) were in reservoir locating along the river Pong. Three populations of 37, 77 and 31 fish from Kaeng Lawa (KLW1 and KLW2) and Kaeng Nam Ton (KNT), and a population of 28 fish from Huai Chorakhe Mak (HCM) were collected from slow-moving water of the river Chi and Mun, respectively.

The fish samples were identified for the right type of fish species based on the identification key of Rainboth (1996) and Kottelat (2001). The identified fish were labelled with a specific code for further traceability and then kept in -20 °C prior for further analysis through truss network method.

Morphometric measurement

The identified fish samples were soaked in running-tap water to make their body soften before placing the right sides posture on a polystyrene board. The landmarks were defined on the basis of homologous points of the external morphology among the specimens with slightly modification from Armbruster (2012). The morphometric valuables were measured on the fish left side to the nearest 0.01 mm with digital callipers based on truss-network system (Strauss and Bookstein, 1982). Thirty-two morphometric variables and standard length (SL) were obtained from 14 anatomical landmarks (Fig. 2). The gender of specimen was identified by dissecting fish body to examine the gonads inside. The specimens were fix in 10% neutral buffered formalin for at least 7 days and changed to preserve in 70% ethanol for voucher-specimen collection.

 

Table I.- Sampling site localities, collection date, population code, and sample sizes of specimens used for this study.

River basin	Sampling sites	Geographical coordinate	Collection date	Population	Sample size
					F	M	total
Pong	Ubolrattana Dam, Ubolratana District, Khon Kaen Province	N 16 43.060 E 102 37.187	November, 2012	URD1	7	12	19
			April, 2013	URD2	22	15	37
Chi	Kaeng Lawa, Ban Phai District, Khon Kaen Province	N 16 09.644 E 102 39.860	April, 2013	KLW1	55	15	37
			November, 2013	KLW2	45	29	77
	Kaeng Nam Ton, Mueang District, Khon Kaen Province	N 16 23.746 E 102 45.903	November, 2013	KNT	19	12	31
Mun	Huai Chorakhe Mak, Mueang District, Burirum Province	N 14 54.293 E 103 01.270	April, 2013	HCM	16	12	28
					164	127	291

 

Statistical analysis

All measured variables was subjected to an allometric transformation equation (Elliott et al., 1995; Reist, 1985) in order to get rid of size-dependent variation from shape information:

Madj = M (Ls/L0)b

Where, Madj is the size-adjusted measurement, M is the original measurement, L0 is the standard length of the fish specimen, Ls is the overall mean of standard length for all fish from all samples in each analysis, and b is an adjusted coefficient estimated from the observed data as the slope of log M against log L0 using all fish in each group. The transformation efficiency was confirmed by testing the correlation significances between the adjusted variables and standard length (Turan, 1999).

Multivariate analysis of variance (MANOVA) was performed in order to evaluate the statistically significant difference between sex and among populations. In addition, a univariate analysis of variance (ANOVA) was performed to examine the statistical differences of each morphometric character for sexual and population effects, respectively. The significant characters (p<0.05) were then subjected to subsequently statistical analyses.

Principal component analysis (PCA) was then implied in order to elucidate patterns of morphological variations between sexes and among populations. The PCA can use for reducing redundancy among the variables and in extracting sets of independent variables that meaningful contributed with morphological differentiation. The univariate t-test and ANOVA were applied to the loading scores of PCA in order to determine significant differences in patterns of morphological variations between sexes and among populations, respectively.

Linear discriminant function analysis (DFA) was executed to predict and classify each specimen to their respective populations based on their morphometric features. Furthermore, the percentages of correct classification were calculated and a cross-validated test was performed to estimate the expected actual error rates of the classification.

All statistical analyses were performed using computer programme R version 3.2.1 (R Core Team, 2016).

 

RESULTS

 

According to the allometric transformation, the correlation analysis revealed no significant correlation

 

 

between transformed truss measurements and standard length (SL) (r <0.03, p>0.05; data not shown). There was an evidence which indicated that size-dependent effects had been efficiently removed from shape information.

Multivariate analysis of variance (MANOVA) on transformed truss variables (Table II) presented significant differences between sexes (p<0.01) and populations (p<0.01), whereas interaction effect of sex and populations was not significantly different (p>0.05).

Patterns of morphological variability among populations

Univariate ANOVA demonstrated that all transformed variables were significantly different between populations (Table III), and the PCA also revealed the first three principal components (PC1 – PC3) accounted for 47.29% of total variation (Table IV). The PC1 accounted for 26.07% of total variation, whereas the PC2 and PC3 explained 11.72 and 9.50% of total variation, respectively.

 

Table II.- Results of multivariate analysis of variance (MANOVA) on transformed truss-network variables for testing effects of sex and population on morphometric variability.

	d.f. 1	d.f. 2	Wilk’s	F-value
Sex	1	32	0.68197	3.6141 **
Populations	5	160	0.00634	13.7077 **
Sex × Population	5	160	0.52947	1.0561

**, highly significantly different (p<0.01).

 

Table III.- Descriptive statistic (mean ± S.D.) of each truss-network measurement of female and male Cyclocheilichthys apogon, and F-statistics of ANOVA for testing morphological variations among populations.

Var.	Descriptive statistic (mean ± S.D.)						ANOVA
	URD1 (n=19)	URD2 (n=37)	KLW1 (n=99)	KLW2 (n=77)	KNT (n=31)	HCM (n=28)	(F-value)
AB	6.08 ± 0.67	5.22 ± 0.49	6.08 ± 0.67	5.24 ± 0.61	4.59 ± 0.59	4.62 ± 0.43	23.496	**
AN	6.03 ± 0.70	6.04 ± 0.65	6.03 ± 0.70	5.38 ± 0.62	5.34 ± 1.16	5.14 ± 0.47	48.049	**
BC	11.24 ± 1.33	11.17 ± 0.73	11.24 ± 1.33	11.21 ± 1.11	9.69 ± 1.16	10.12 ± 0.78	6.9692	**
BL	20.86 ± 2.10	18.52 ± 1.48	20.86 ± 2.10	20.31 ± 2.00	18.10 ± 1.94	17.60 ± 1.80	10.17	**
BM	13.09 ± 1.32	11.84 ± 0.85	13.09 ± 1.32	12.27 ± 1.11	11.25 ± 1.57	11.00 ± 0.92	16.823	**
BN	8.86 ± 0.92	8.22 ± 0.64	8.86 ± 0.92	7.99 ± 0.84	7.17 ± 1.00	7.10 ± 0.71	29.434	**
CD	42.87 ± 4.54	39.31 ± 2.54	42.87 ± 4.54	40.48 ± 3.74	35.43 ± 4.41	34.71 ± 3.10	9.8771	**
CK	42.01 ± 4.65	38.03 ± 2.69	42.01 ± 4.65	39.56 ± 3.72	33.90 ± 4.22	33.29 ± 3.01	14.255	**
CL	20.69 ± 2.03	19.08 ± 1.34	20.69 ± 2.03	19.94 ± 1.71	17.23 ± 2.07	17.53 ± 1.46	13.858	**
CM	18.42 ± 1.84	17.34 ± 1.12	18.42 ± 1.84	17.41 ± 1.45	14.99 ± 1.91	15.89 ± 1.17	19.11	**
CN	17.41 ± 1.71	16.52 ± 0.99	17.41 ± 1.71	16.50 ± 1.34	14.19 ± 1.62	14.88 ± 1.05	12.462	**
DE	17.33 ± 1.67	16.37 ± 0.76	17.33 ± 1.67	16.65 ± 1.58	15.01 ± 1.61	14.85 ± 0.90	2.9346	*
DH	45.47 ± 4.46	42.68 ± 2.18	45.47 ± 4.46	43.93 ± 3.74	38.28 ± 4.83	38.81 ± 2.57	25.934	**
DI	41.26 ± 4.27	39.47 ± 2.28	41.26 ± 4.27	39.52 ± 3.67	34.42 ± 4.42	34.85 ± 2.60	62.991	**
DJ	39.64 ± 4.16	37.73 ± 2.47	39.64 ± 4.16	38.04 ± 3.53	33.27 ± 4.57	33.48 ± 2.73	50.486	**
DK	38.21 ± 4.43	35.36 ± 2.49	38.21 ± 4.43	36.30 ± 3.62	31.63 ± 4.49	31.62 ± 3.04	25.636	**
DL	49.84 ± 5.01	46.51 ± 3.17	49.84 ± 5.01	46.78 ± 4.21	40.89 ± 5.32	40.95 ± 3.12	29.192	**
EF	33.08 ± 3.70	30.55 ± 2.39	33.08 ± 3.70	31.21 ± 2.74	27.98 ± 3.89	28.01 ± 2.95	14.671	**
EG	37.82 ± 4.22	35.09 ± 2.14	37.82 ± 4.22	36.12 ± 2.95	31.80 ± 4.51	31.64 ± 2.76	16.343	**
EH	29.25 ± 3.25	27.40 ± 1.65	29.25 ± 3.25	28.65 ± 2.37	24.30 ± 3.57	24.95 ± 2.15	30.448	**
EI	28.19 ± 3.20	27.08 ± 1.68	28.19 ± 3.20	27.50 ± 2.50	23.30 ± 3.34	23.79 ± 2.11	55.531	**
EJ	28.59 ± 3.20	27.20 ± 1.81	28.59 ± 3.20	27.92 ± 2.54	23.84 ± 3.61	24.21 ± 2.26	50.031	**
EK	38.83 ± 4.49	35.84 ± 2.36	38.83 ± 4.49	37.60 ± 3.57	32.91 ± 4.84	32.77 ± 2.86	40.696	**
FG	13.46 ± 1.58	12.95 ± 0.81	13.46 ± 1.58	13.14 ± 1.22	11.22 ± 1.48	11.23 ± 0.92	44.872	**
FH	19.86 ± 2.25	18.36 ± 1.67	19.86 ± 2.25	19.14 ± 1.76	16.69 ± 2.14	16.29 ± 1.58	10.045	**
GH	12.65 ± 1.72	11.22 ± 1.21	12.65 ± 1.72	11.73 ± 1.25	10.79 ± 1.62	9.84 ± 1.13	2.3604	*
HI	12.87 ± 1.19	11.70 ± 0.96	12.87 ± 1.19	12.50 ± 1.24	10.94 ± 1.41	11.13 ± 0.66	2.7719	*
IJ	5.54 ± 0.84	5.26 ± 0.69	5.54 ± 0.84	5.08 ± 0.80	4.42 ± 0.88	4.77 ± 0.49	21.675	**
JK	26.17 ± 3.43	24.39 ± 2.63	26.17 ± 3.43	25.29 ± 2.74	22.96 ± 4.02	22.61 ± 2.07	35.051	**
KL	30.58 ± 3.45	28.16 ± 2.55	30.58 ± 3.45	27.98 ± 2.95	23.92 ± 3.24	23.57 ± 2.08	16.076	**
LM	10.52 ± 1.36	8.82 ± 1.24	10.52 ± 1.36	10.31 ± 1.46	8.74 ± 1.05	8.89 ± 1.55	9.3031	**
MN	5.85 ± 0.74	4.79 ± 0.70	5.85 ± 0.74	5.72 ± 0.68	5.11 ± 0.81	5.08 ± 0.48	25.95	**

*, significant difference (p<0.05); **, highly significant difference (p<0.01).

AB, snout length; AN, mouth length; BC, forehead length; BL, BM, BN, CL, CM, CN, head depth; CD, pre-dorsal length; CK, diagonal body depth; DE, dorsal fin-base length; DH, DI, DJ, DK, DL, diagonal depth of foretrunk; EF, post-dorsal length; EG, EH, diagonal length of caudal peduncle; EI, EJ, EK, diagonal depth of post-trunk; FG, precaudal depth; FH, diagonal length of caudal peduncle; GH, caudal peduncle length; HI, anal fin-base length; IJ, JK, abdomen length; KL, pectoral length; LM, MN, lower head length.

 

The bivariate plots of PCA also showed some separations of morphological variations among populations (Fig. 3). Populations URD1, URD2 and HCM distributed on the negative side of PC1 and tended to separate from population KLW1, KLW2 and KNT which distributed on the positive PC1 axis. Populations URD1 and URD2 were also separated from population HCM by PC2.

In addition, the ANOVA on loadings scores of the first three PCs indicated significant differences in patterns of morphological variations among populations (Fig. 4). The pairwise-multiple comparisons showed that PC1 (Fig. 4a) grouped the samples into four groups including URD1 and URD2; KLW1; KLW2 and KNT; and HCM. These variations associated with the variations in head depth (CM), forepart body length (DL), hind-part body length (DH, DI, DJ, EJ, EK), body depth (DK, EH), caudal peduncle length (EG, EH) and caudal peduncle depth (FG).

The second index (Fig. 4b), PC2 clustered samples into three groups including URD1 and KLW1, URD2, and KLW2, KNT and HCM. Such groupings have morphological differentiations in head characters (AN, BN, CM, CN), body depth (EH) and caudal peduncle length (EG).

 

For abbreviations, see Table III

 

The PC3 yielded population separations similar to PC2 (Fig. 4c) in correlation with the variations in head features (BC, CL, LM), thoracic length (KL) and forepart body length (DL).

 

Table IV.- Results of principal component analysis (PCA) and analysis of variance (ANOVA) of factors scores for investigating morphological variations among populations.

Variables	Factor loadings
	PC1	PC2	PC3
AB	0.230	-0.377	0.022
AN	-0.244	-0.598	0.028
BC	-0.325	-0.268	0.086
BL	-0.172	-0.223	-0.896
BM	-0.311	-0.428	-0.173
BN	-0.172	-0.746	0.051
CD	-0.400	-0.005	-0.062
CK	-0.412	-0.379	0.060
CL	-0.516	-0.411	-0.612
CM	-0.479	-0.570	0.072
CN	-0.310	-0.615	0.160
DE	-0.092	-0.209	-0.021
DH	-0.768	0.239	-0.019
DI	-0.875	-0.049	0.105
DJ	-0.892	0.095	0.004
DK	-0.797	-0.086	-0.121
DL	-0.615	-0.212	0.414
EF	-0.435	0.422	0.135
EG	-0.583	0.477	0.101
EH	-0.718	0.467	-0.010
EI	-0.898	0.209	0.082
EJ	-0.871	0.241	0.010
EK	-0.713	0.087	-0.208
FG	-0.662	0.007	0.108
FH	-0.295	0.085	0.163
GH	0.088	0.088	0.118
HI	0.157	-0.007	-0.109
IJ	-0.123	-0.360	0.179
JK	-0.411	0.220	-0.189
KL	-0.025	-0.308	0.796
LM	0.011	-0.118	-0.844
MN	0.116	0.417	-0.168
EigenValue	8.3421	3.7506	3.3090
% variation	26.07%	11.72%	9.50%
F-value	81.03**	32.00**	15.09**

**, highly significantly different (p<0.01).

For abbreviations, see Table III

 

Discrimination of population using morphometric characteristics

The discriminant function analysis (DFA) revealed five discriminant functions which could be used as morphological descriptors for classifying the samples into their own groups (Table V). The first two discriminant functions which were meaningful for DFA (Eigenvalue > 1) accounted for 72% of total variation among population. The first discriminant function (DF1) accounted for 49.33% of total variation. The measurements from head (AN, BN, LM) and hind-part body length (EL, HI) highly

 

Table V.- Structure matrix of discriminant functions obtained from discriminant function analysis (DFA) on truss-network variables.

Variables	Discriminant function (DF)
	DF1	DF2	DF3	DF4	DF5
AB	-0.092	-0.166	0.292	-0.345	-0.356
AN	0.616	-0.129	-0.491	-0.149	-0.025
BC	0.156	0.036	0.164	-0.025	0.329
BL	0.111	0.251	-0.117	0.407	-0.288
BM	0.331	0.206	-0.387	0.056	-0.309
BN	0.561	-0.255	-0.107	-0.119	-0.265
CK	0.344	0.219	0.077	-0.030	-0.101
CL	0.429	0.091	0.276	0.095	-0.159
CM	0.304	0.300	0.228	0.182	-0.097
CN	0.375	0.198	0.365	-0.179	-0.059
DE	0.317	-0.095	0.351	-0.160	0.037
DJ	0.133	-0.089	-0.080	-0.153	0.185
DK	0.233	0.441	0.192	-0.064	0.296
DL	0.517	0.375	0.197	-0.264	0.328
EF	0.220	0.584	0.195	-0.339	0.374
EG	0.164	0.581	0.149	-0.163	0.090
EH	0.424	0.224	0.198	-0.393	0.096
EI	0.146	0.490	-0.002	-0.112	-0.050
EJ	0.148	0.461	-0.004	-0.038	0.161
EK	0.071	0.526	0.277	0.107	0.249
FG	0.345	0.471	0.297	-0.044	0.393
FH	0.156	0.610	0.258	-0.051	0.363
GH	0.072	0.681	-0.005	0.051	0.185
HI	0.508	0.176	0.119	0.082	0.419
IJ	0.337	0.154	-0.015	0.194	0.047
JK	0.087	-0.057	-0.258	0.063	-0.055
KL	-0.118	-0.172	0.011	0.099	-0.052
LM	0.592	-0.124	0.011	0.136	-0.303
MN	-0.020	0.639	-0.216	-0.126	0.234
Eigenvalue	3.8967	1.8510	0.9886	0.6947	0.4681
% variation	49.33%	23.43%	12.52%	8.79%	5.93%

For abbreviations, see Table III

 

Table VI.- Percentage of specimens classified in each group from original and cross-validation tests of discriminant function analysis (DFA) on truss-network data.

Predicted populations	Original populations						Global accuracy
	URD1	URD2	KLW1	KLW2	KNT	HCM
Original test
URD1	100	0	0	0	0	0
URD2	0	97.30	0	2.70	0	0
KLW1	0	0	95.96	4.04	0	0	96.29
KLW2	0	1.30	6.49	90.91	1.30	0
KNT	0	0	6.45	0	93.55	0
HCM	0	0	0	0	0	100
Cross-validation test
URD1	100.00	0	0	0	0	0
URD2	2.70	91.90	0	2.70	0	2.70
KLW1	0	2.02	88.89	6.06	2.02	1.01	90.59
KLW2	0	1.30	7.79	89.61	1.30	0
KNT	0	0	9.68	6.45	83.87	0
HCM	0	3.57	0	3.57	3.57	89.29

For abbreviations, see Table III

 

 

contributed to DF1. The second discriminant function (DF2) explained 23.43% of total variation, and this function highly correlated with the measurements from head (MN), body depth (EJ), hind-part body length (EI, EK), and caudal peduncle traits (EF, EG, GH, FG, GH). It indicated that all of those variables were important in discrimination of morphological variation in C. apogon populations.

The discriminant plot of fish individuals on DF1-DF2 axes (Fig. 5) distinguished samples into three groups regarding river drainage systems of the collecting sites including URD (Pong River basin), KLW and KNT (Chi River basin), and HCM group (Mun River basin).

The global accuracy of classification was 96.29% for original classification and 90.56% for cross-validated test (Table VI). The correct classification rate was highest in population KLW1 (100% for both original and cross-validated tests). For the original test, the corrected classification rates ranged from 90.91 – 100%, and the highest misclassification was the misclassification of population KLW2 into KLW1 (6.49%). For the cross-validated test, the corrected classification ranged from 83.87% - 100%, and the highest misclassification is the classification of KNT into KLW1 (9.68%). These misclassifications may indicate the morphological similarities of the C. apogon collected from same river drainage.

 

DISCUSSION

 

The morphological variation of C. apogon in the present study occurred in both between different sexes and among population. These findings are consistent to many previous studies in fish such as the rohu labeo Labeo rohita from Ganga basin in India (Mir et al., 2013), silver perch Leiopotheron plumbeus from three lakes in the Philippines (Quilang et al., 2007), Günther’s Mouthbrooder Chromidotilapia guntheri from three coastal rivers of Africa (Boussou et al., 2010), the three populations of orange-fin labeo Labeo cabasu from two isolate rivers in Bangladesh (Hossain et al., 2010) and spotted snakehead Channa punctatus from three Indian rivers (Khan et al., 2013).

The results of the present work were also indicative of geographical isolation according river drainage systems of the collecting sites: Pong River (URD populations), Chi River (KLW and KNT populations), and Mun River (HCM population). Most of the fish samples (96.29 and 90.56% with original and cross-validation tests, respectively) were correctly classified into their respect locations by DFA (Table VI) and the discrimination plots of each individual along discriminant axes showed quite separation regarding river isolation (Fig. 5).

Interestingly, morphological similarity was observed in KLW2 and KNT which are geographically isolated populations. This observation may be due to the morphological plasticity to the similar ecological impacts (Mir et al., 2013), or due to local migration of the fish between connected locations (Hossain et al., 2010; Khan et al., 2013). On the other hand, this finding suggest that an insufficient degree of geographical isolation might not be involved the formation of different stock especially if the ecological conditions of the habitats are quite similar.

Regarding the large degree of morphological variation between populations obtaining from the same locality at different times (URD1-URD2 and KLW1-KLW2), significant morphometric differences between those two pairs of populations were observed and resulted in a high correct reclassification rate of each populations (Table VI, Fig. 5). These findings provide further evidence for the complexity of the stock structure within that locations (Zhang et al., 2016). The separation of URD1 and URD2 samples may possibly due to isolation of portion of populations within large local habitat area (Mir et al., 2013; Turan et al., 2004) that may be sufficient to enforce populations to adapt and involve as independent biological entities with specific phenotypes in different ecological conditions (Turan et al., 2004). In addition, the differences may also be attributed to spatial variation in environment factors varying in different season during the year. Hydrological regime is considered the key factor driving ecological functioning in river floodplain system (Bunn and Arthington, 2002; Thomaz et al., 2007). The water level and water current affecting by differential flooding cycles are causally related to ecosystem attributes in the habitat especially food availability (Cochran-Biederman and Winemiller, 2010; Thomaz et al., 2007) which will be affected biological parameters of populations, leading to the differentiation in morphology among that populations.

The plots of the first three principal component axes of PCA (Fig. 4) also confirmed a high degree of morphological variations. The subsequently observed differences in morphology were significantly in overall body shape from head to tail. The variation of size and shape was usually occurred in fish more than other vertebrates and were considered as the involvement of environmental influences (Cadrin, 2000; Wimberger, 1992). Such variations in body depth and caudal peduncle characteristics could possibly be related to environmental conditions in relation to water depth and current flow (Pazhayamadom et al., 2015). The adaptation in body depth and caudal peduncle traits may be associated with swimming performance (Boily and Magnan, 2002; Peres-Neto and Magnan, 2004; Webb, 1984), which could also be related to foraging efficiency (Boily and Magnan, 2002; Swain et al., 2005) and predator evasiveness (Chipps et al., 2004; Swain et al., 2005). Adaptation with deep robust body is required for attain faster burst velocity with transient propulsion in the less turbulent water, while the shallow body depth is optimal for periodic propulsion against fast-following water currents (Blake et al., 2005; Webb, 1984). The variability in the head parts which reflected for a differential habitat use (Boily and Magnan, 2002; Robinson and Wilson, 1994; Wainwright, 1996; Webster et al., 2011), especially regarding the feeding regimes with variable diets (Berchtold et al., 2015; Hyndes et al., 1997; Wainwright and Richard, 1995). In addition, variation in the head morphology will also attributed to water parameters and current velocity (Langerhans et al., 2007). It is well known that the phenotypic plasticity allows fish to adaptively react to environmental changes for fitness by modifications in their physiology and behaviour, which lead to changes in morphology, reproduction and survival (Turan et al., 2004). Variations of environmental factors such as water current, flooding patterns, water turbidity, and food availability could also be involved as particular factors in morphological variations during the early development stages when the individual’s trait is more susceptible to environment influences (Wimberger, 1992).

Apparently, there is possibility that morphological variabilities among geographically isolated populations observed in the recent work may be correlated to genetic differentiation. Since the fragmentation of habitat localities that prevent the genetic exchange among populations designates the enrichment of the established genetic differences resulting a heighten degree of inter-population differences (Cadrin, 2000; Poulet et al., 2004). The relationship of morphological variation and genetic difference was explained in several fish including the pikeperch Sander lucioperca from a fragmented delta (Poulet et al., 2004) and Swedish postglacial stickleback Pungitius pungitius from coastal and inland lakes (Mobley et al., 2011). The correlation between genetic and morphological variations also supported the existence of distinct population groups of Moenkhausia oligolepis from different tributaries (Domingos et al., 2014).

However, morphological variations observed in isolated populations in the recent study may not be involved with genetic differentiation as in Coilia ectenes populations from three-isolated lakes that found no obviously genetic variation in relation to geographical differentiation (Xie, 2012). In contrary with the study of Eurasian perch, Perca fuuviatilis that showed the difference of morphology between literal and pelagic populations more closely related to the environmental adaptation than genetic variation (Svanbäck and Eklöv, 2006).

The present study suggests that the extent of morphological divergence probably related to ecological differences which also related to the distances of geographical isolation. The variation of C. apogon will be resulted from phenotypic plasticity that allows the fish to consequently suitable to the environmental conditions of the habitats (Wimberger, 1992). A sufficient degree of isolation, which can arise because of geographical distant or flood cycle of the river, will result in notable morphological differentiation as well as genetic variability between stock of C. apogon (Cronin-Fine et al., 2013; Turan, 2004). An effective planning for fishery management should establish separately based on isolate stock function as basic unit (Begg et al., 1999; Cardin et al., 2005; Hoggarth, 2006). A failure to account for fishery stock can lead to erosion of biological attributes of population, which would be subsequently accelerated a loss of genetic diversity and a potential decrease of fishery productivity of the species resource (Begg et al., 1999; Sterner, 2007; Zhang et al., 2016).

 

CONCLUSION

 

The output of the study will provide useful baseline information of C. apogon for appropriate management and conservation of the species. The results indicated that fisheries management of C. apogon should be considered strategic planning independently along each of the river drainages. However, further study in genetic information is necessary to investigate correlation between morphological variation and genetic attributes of this species, resulting to sufficient information for sustainable utilisation of fishery resources.

 

Acknowledgements

 

The present work was financially supported by Development and Promotion of Science and Technology Talents Project (DPST). The authors would also like to thank Department of Biology, Faculty of Science, Khon Kaen University for providing necessary facilities to carry out this work and sincere thanks to anonymous reviewers for great comments to help improving this manuscript.

 

Statement of conflict of interest

Authors have declared no conflict of interest.

 

References

 

Akbarzadeh, A., Farahmand, H., Shabani, A.A., Karami, M., Kaboli, M., Abbasi, K. and Rafiee, G.R., 2009. Morphological variation of the pikeperch Sander lucioperca (L.) in the southern Caspian Sea, using a truss system. J. appl. Ichthyol., 25: 576–582. https://doi.org/10.1111/j.1439-0426.2009.01308.x

Armbruster, J.W., 2012. Standardized measurements, landmarks, and meristic counts for cypriniform fishes. Zootaxa, 3586: 8–12.

Begg, G.A., Friedland, K.D. and Pearce, J.B., 1999. Stock identification and its role in stock assessment and fisheries management: an overview. Fish. Res., 43: 1–8. https://doi.org/10.1016/S0165-7836(99)00062-4

Berchtold, A.E., Colborne, S.F., Longstaffe, F.J. and Neff, B.D., 2015. Ecomorphological patterns linking morphology and diet across three populations of pumpkinseed sunfish (Lepomis gibbosus). Can. J. Zool., 93: 289–297. https://doi.org/10.1139/cjz-2014-0236

Blake, R.W., Law, T.C., Chan, K.H.S. and Li, J.F.Z., 2005. Comparison of the prolonged swimming performances of closely related, morphologically distinct three-spined sticklebacks Gasterosteus spp. J. Fish. Biol., 67: 834-848. https://doi.org/10.1111/j.0022-1112.2005.00788.x

Boily, P. and Magnan, P., 2002. Relationship between individual variation in morphological characters and swimming costs in brook charr (Salvelinus fontinalis) and yellow perch (Perca flavescens). J. exp. Biol., 205: 1031–1036.

Boussou, C.K., Konan, F.K., Edia, E.O., Quattara, M., Bony, Y.K., Quattara, A. and Gourène, G., 2010. Morphometric analysis of populations of Chromidotilapia guntheri (Sauvage, 1882) (Cichlidae, perciformes) in four coastal rivers of Côte d´Ivoire (West Africa). Pan-Am. J. aquat. Sci., 5: 387–400.

Bunn, S.E. and Arthington, A.H., 2002. Basic principles and ecological consequences of altered flow regimes for aquatic biodiversity. Environ. Manage., 30: 492–507. https://doi.org/10.1007/s00267-002-2737-0

Cadrin, S., 2000. Advances in morphometric identification of fishery stocks. Rev. Fish Biol. Fish., 10: 91–112. https://doi.org/10.1023/A:1008939104413

Cardin, S.X., Friedland, K.D. and Waldman, J.R., 2005. Chapter 1 - Stock identification methods: An overview. In: Stock identification methods (eds. S.X. Cadrin, K.D. Friedland and J.R. Waldman), Academic Press, Burlington, pp. 3–6. https://doi.org/10.1016/B978-012154351-8/50002-2

Cavalcanti, M.J., Monteiro, L.R. and Lopes, P.R.D., 1999. Landmark-based morphometric analysis in selected species of serranid fishes (Perciformes: Teleostei). Zool. Stud., 38: 287–294.

Chipps, S.R., Dunbar, J.A. and Wahl, D.H., 2004. Phenotypic variation and vulnerability to predation in juvenile bluegill sunfish (Lepomis macrochirus). Oecologia, 138: 32–38. https://doi.org/10.1007/s00442-003-1396-z

Cochran-Biederman, J.L. and Winemiller, K.O., 2010. Relationships among habitat, ecomorphology and diets of cichlids in the Bladen River, Belize. Environ. Biol. Fishes, 88: 143–152. https://doi.org/10.1007/s10641-010-9624-y

Cronin-Fine, L., Stockwell, J.D., Whitener, Z.T., Labbe, E.M., Willis, T.V. and Wilson, K.A., 2013. Application of morphometric analysis to identify Alewife stock structure in the Gulf of Maine. Mar. Coast. Fish., 5: 11–20. https://doi.org/10.1080/19425120.2012.741558

Domingos, T.J., Moraes, L.N., Moresco, R.M., Margarido, V.P. and Venere, P.C., 2014. Genetic and morphological diversity of Moenkhausia oligolepis (Characiformes: Characidae) populations in the tributaries of the Araguaia River, Brazil: implications for taxonomy and conservation. Genet. mol. Res., 13: 7979–7991. https://doi.org/10.4238/2014.September.29.11

Elliott, N.G., Haskard, K. and Koslow, J.A., 1995. Morphometric analysis of orange roughy (Hoplostethus atlanticus) off the continental slope of southern Australia. J. Fish Biol., 46: 202–220. https://doi.org/10.1111/j.1095-8649.1995.tb05962.x

Fairbairn, D.J., 1997. Allometry for sexual size dimorphism: Pattern and Process in the coevolution of body size in males and females. Annu. Rev. Ecol. Syst., 28: 659–687. https://doi.org/10.1146/annurev.ecolsys.28.1.659

Fairbairn, D.J. and Preziosi, R.F., 1994. Sexual selection and the evolution of allometry for sexual size dimorphism in the water strider, Aquarius remigis. Am. Natural., 144: 101–118. https://doi.org/10.1086/285663

Hedrick, A.V. and Temeles, E.J., 1989. The evolution of sexual dimorphism in animals: Hypotheses and tests. Trends Ecol. Evol., 4: 136–138. https://doi.org/10.1016/0169-5347(89)90212-7

Herler, J., Kerschbaumer, M., Mitteroecker, P., Postl, L. and Sturmbauer, C., 2010. Sexual dimorphism and population divergence in the Lake Tanganyika cichlid fish genus Tropheus. Front. Zool., 7: 4. https://doi.org/10.1186/1742-9994-7-4

Hoggarth, D.D. (Ed.), 2006. Stock assessment for fishery management: a framework guide to the stock assessment tools of the Fisheries Management Science Programme. FAO Fisheries Technical Paper, Food and Agriculture Organization of the United Nations, Rome.

Hossain, M.A.R., Nahiduzzaman, M., Saha, D., Khanam, M.U.H. and Alam, M.S., 2010. Landmark-based morphometric and meristic variations of the endangered carp, Kalibaus Labeo calbasu, from stocks of two isolated rivers, the Jamuna and Halda, and a hatchery. Zool. Stud., 49: 556–563.

Hughes, A.L. and Hughes, M.K., 1986. Paternal investment and sexual size dimorphism in North American Passerines. Oikos, 46: 171–175. https://doi.org/10.2307/3565464

Hyndes, G.A., Platell, M.E. and Potter, I.C., 1997. Relationships between diet and body size, mouth morphology, habitat and movements of six sillaginid species in coastal waters: implications for resource partitioning. Mar. Biol., 128: 585–598. https://doi.org/10.1007/s002270050125

Khan, M.A., Miyan, K. and Khan, S., 2013. Morphometric variation of snakehead fish, Channa punctatus, populations from three Indian rivers. J. appl. Ichthyol., 29: 637–642. https://doi.org/10.1111/j.1439-0426.2012.02058.x

Kottelat, M., 2001. Fishes of Laos. WHT Publications, Colombo, Sri Lanka.

Langerhans, R.B., Chapman, L.J. and Dewitt, T.J., 2007. Complex phenotype–environment associations revealed in an East African cyprinid. J. Evol. Biol., 20: 1171-1181. https://doi.org/10.1111/j.1420-9101.2007.01282.x

Mir, J.I., Mir, F.A., Chandra, S. and Patiyal, R.S., 2013. Pattern of morphological variations in Alghad snowtrout, Schizopyge niger (Heckel 1838) from Kashmir Himalaya using truss network analysis. Ichthyol. Res., 60: 256–262. https://doi.org/10.1007/s10228-013-0347-x

Mir, J.I., Sarkar, U.K., Dwivedi, A.K., Gusain, O.P. and Jena, J.K., 2013. Stock structure analysis of Labeo rohita (Hamilton, 1822) across the Ganga basin (India) using a truss network system. J. appl. Ichthyol., 29: 1097–1103. https://doi.org/10.1111/jai.12141

Mobley, K.B., Lussetti, D., Johansson, F., Englund, G. and Bokma, F., 2011. Morphological and genetic divergence in Swedish postglacial stickleback (Pungitius pungitius) populations. BMC Evol. Biol., 11: 287. https://doi.org/10.1186/1471-2148-11-287

Pazhayamadom, D.G., Chakraborty, S.K., Jaiswar, A.K., Sudheesan, D., Sajina, A.M. and Jahageerdar, S., 2015. Stock structure analysis of “Bombay duck” (Harpadon nehereus Hamilton, 1822) along the Indian coast using truss network morphometrics. J. appl. Ichthyol., 31: 37–44. https://doi.org/10.1111/jai.12629

Peres-Neto, P.R. and Magnan, P., 2004. The influence of swimming demand on phenotypic plasticity and morphological integration: a comparison of two polymorphic charr species. Oecologia, 140: 36–45. https://doi.org/10.1007/s00442-004-1562-y

Poulet, N., Berrebi, P., Crivelli, A.J., Lek, S. and Argillier, C., 2004. Genetic and morphometric variations in the pikeperch (Sander lucioperca L.) of a fragmented delta. Arch. Hydrobiol., 159: 531–554. https://doi.org/10.1127/0003-9136/2004/0159-0531

Quilang, J.P., Basiao, Z.U., Pagulayan, R.C., Roderos, R.R. and Barrios, E.B., 2007. Meristic and morphometric variation in the silver perch, Leiopotherapon plumbeus (Kner, 1864), from three lakes in the Philippines. J. appl. Ichthyol., 23: 561–567. https://doi.org/10.1111/j.1439-0426.2007.00862.x

R Core Team, 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

Rainboth, W.J., 1996. Fishes of the Cambodian Mekong, FAO species identification field guide for fishery purposes. Food and Agriculture Organization of the United Nations, Rome.

Reist, J.D., 1985. An empirical evaluation of several univariate methods that adjust for size variation in morphometric data. Can. J. Zool., 63: 1429–1439. https://doi.org/10.1139/z85-213

Robinson, B.W. and Wilson, D.S., 1994. Character release and displacement in fishes: A neglected literature. Am. Natural., 144: 596–627. https://doi.org/10.1086/285696

Secor, D.H., 2014. Chapter two - The unit stock concept: Bounded fish and fisheries. In: Stock Identification Methods. (Eds. Cadrin, S.X., Kerr, L.A. and Mariani S.) 2nd Edition) Academic Press, San Diego. pp.7–28.

Smith, P.J., Francis, R.I.C.C. and McVeagh, M., 1991. Loss of genetic diversity due to fishing pressure. Fish. Res., 10: 309–316. https://doi.org/10.1016/0165-7836(91)90082-Q

Sterner, T., 2007. Unobserved diversity, depletion and irreversibility. The importance of subpopulations for management of cod stocks. Ecol. Econ., 61: 566–574. https://doi.org/10.1016/j.ecolecon.2006.05.015

Strauss, R.E. and Bookstein, F.L., 1982. The Truss: Body form reconstructions in morphometrics. Syst. Zool., 31: 113–135. https://doi.org/10.2307/2413032

Svanbäck, R. and Eklöv, P., 2006. Genetic variation and phenotypic plasticity: causes of morphological and dietary variation in Eurasian perch. Evol. Ecol. Res., 8: 37–49.

Swain, D.P., Hutchings, J.A., Foote, C.J., 2005. Chapter 4 - Environmental and genetic influences on stock identification characters. In: Stock identification methods (eds. S.X. Cadrin, K.D. Friedland and J.R. Waldman), Academic Press, Burlington, pp. 45–85. https://doi.org/10.1016/B978-012154351-8/50005-8

Thomaz, S.M., Bini, L.M. and Bozelli, R.L., 2007. Floods increase similarity among aquatic habitats in river-floodplain systems. Hydrobiologia, 579: 1–13. https://doi.org/10.1007/s10750-006-0285-y

Turan, C., 2004. Stock identification of Mediterranean horse mackerel (Trachurus mediterraneus) using morphometric and meristic characters. ICES J. mar. Sci., 61: 774–781. https://doi.org/10.1016/j.icesjms.2004.05.001

Turan, C., 1999. A note on the examination of morphometric differentiation among fish populations: The Truss system. Turk. J. Zool., 23: 259–263.

Turan, C., Ergüden, D., Gürlek, M., Başusta, N. and Turan, T., 2004. Morphometric structuring of the anchovy (Engraulis encrasicolus L.) in the Black, Aegean and Northeastern Mediterranean Seas. Turk. J. Anim. Sci., 28: 865–871.

Vidthayanon, C., 2012. Cyclocheilichthys apogon (Beardless barb) [WWW Document]. Cyclocheilichthys Apogon. Available at: http://www.iucnredlist.org (accessed 7.14.15).

Wainwright, P.C., 1996. Ecological explanation through functional morphology: the feeding biology of sunfishes. Ecology, 77: 1336-1343. https://doi.org/10.2307/2265531

Wainwright, P.C. and Richard, B.A., 1995. Predicting patterns of prey use from morphology of fishes. Environ. Biol. Fishes, 44: 97–113. https://doi.org/10.1007/BF00005909

Webb, P.W., 1984. Form and function in fish swimming. Scient. Am., 251: 72–82. https://doi.org/10.1038/scientificamerican0784-72

Webster, M.M., Atton, N., Hart, P.J.B. and Ward, A.J.W., 2011. Habitat-specific morphological variation among threespine sticklebacks (Gasterosteus aculeatus) within a drainage basin. PLoS One, 6: e21060. https://doi.org/10.1371/journal.pone.0021060

Wimberger, P.H., 1992. Plasticity of fish body shape. The effects of diet, development, family and age in two species of Geophagus (Pisces: Cichlidae). Biol. J. Linn. Soc., 45: 197–218. https://doi.org/10.1111/j.1095-8312.1992.tb00640.x

Wyman, M.J., Stinchcombe, J.R. and Rowe, L., 2013. A multivariate view of the evolution of sexual dimorphism. J. Evol. Biol., 26: 2070–2080. https://doi.org/10.1111/jeb.12188

Xie, J.Y., 2012. Genetic and morphological variation of three freshwater lake populations of Coilia ectenes (Engraulidae). Genetika, 48: 1200–1206. https://doi.org/10.1134/s1022795412070150

Zhang, C., Jiang, Y.Q., Ye, Z.J., Li, Z.G. and Dou, S.Z., 2016. A morphometric investigation of the small yellow croaker (Larimichthys polyactis Bleeker, 1877): Evidence for subpopulations on the Chinese coast. J. appl. Ichthyol., 32: 67–74. https://doi.org/10.1111/jai.12923