Correlations Between Morphometric Traits and Body Mass Among Different Geographical Populations of Wild Macrobrachium nipponense in Upper Reaches of the Huaihe River, China

Abstract

The oriental river prawn (Macrobrachium nipponense) is one of the most popular freshwater prawn in China. In order to study the relationships between morphometric traits and body mass across different geographical locations and provide references of phenotypic traits for the breeding of M. nipponense, we collected wild M. nipponense from three locations (main stream, HH; Suyahu Reservoir, SYH; and Wuyue Reservoir, WY) in the upper reaches of the Huaihe River, China, and measured 26 morphometric traits and body masses. We found that the coefficient of variation of body mass varied from 31.88% to 59.27% across the three populations, exceeding that of morphometric traits within each population. All 26 morphometric traits, except for the fourth abdominals somite length in the WY population, were observed to correlate significantly positively with body mass (p < 0.05). A path analysis indicated that propodus length, body length, ischium length, carapace height, and second abdominals somite length in the HH population; body length, propodus length, abdominal height, sixth abdominals somite length, and telson length in the SYH population; and body length, carapace height, fifth abdominals somite length, and abdominal height in the WY population significantly affected body mass directly (p < 0.05). Comparing the pathway analysis with the grey relation analysis, we can conclude that the trait most correlated with body mass was body length across the three geographical populations. These findings provide references of waiting morphological traits for M. nipponense selective breeding in different geographical populations.

1. Introduction

Macrobrachium nipponense, also known as oriental river prawn, originated in mainland China about one million years ago [1] and is now broadly distributed throughout most freshwaters and low-salinity estuarine regions in Japan [2], Korea [3], Myanmar [4], and China [5]. Because of desirable flavor, high nutritive value, and excellent adaptability, it is considered as an important fishery resource and is widely farmed in China, with its farmed production reaching nearly 226 thousand tonnes in 2022 [6]. With the continuous expansion of aquaculture scale, adverse selection and inbreeding resulted in low growth rates, smaller body size, low resistance to diseases, early maturation (neoteny), and dramatic restriction of the development of M. nipponense farming [7]. In recent years, a genetic improvement program for M. nipponense has been initiated to renew the germplasm for production, including screening wild populations for better commercial traits for breeding.

Body mass is the major determinant of abdominal meat weight in prawns [8,9], which can effectively impact the economic value and production. Hence, body mass has been taken as an important target trait for genetic improvement in most farmed aquaculture animals. However, the direct measurement is easily affected by residual water on body surface and intestinal feed residue, and currently, many studies, using path analysis, have consistently demonstrated significant associations between body mass and morphometric traits across various species [10,11,12,13,14]; thus, body mass can be improved by indirect selection on more easily measured morphometric traits.

Morphological traits exhibit significant variation influenced by genetic, physiological, and environmental influences [15,16]. Environmental factors, capable of inducing morphological differentiation through phenotypic plasticity, may play a more decisive role than genetic factors in shaping morphological characters [17,18]. Such plasticity can generate phenotypes better adapted to exploit complex environments. Therefore, the impact of environmental factors on morphology cannot be overlooked.

The Huaihe River is not only the geographical boundary between North and South China, but also the boundary between warm temperate and subtropical zones. Despite its ecological importance, there is a lack of research on the morphological variations and their association with body mass among wild populations of M. nipponense in diverse habitats within the upper reaches of the Huaihe River. This study aims to investigate wild M. nipponense populations in environmentally diverse areas (streams vs. reservoirs, northern vs. southern banks) of the Upper Huaihe River to assess the relationships between morphometric characteristics and body mass across different geographic populations of M. nipponense in varied habitats. We analyzed the correlation between morphometric traits and the body mass of three wild populations from the upper reaches of the Huaihe River using correlation analysis, estimated the contribution of morphometric traits to body mass using path analysis and grey relational analysis, identify effective indicators, and constructed a best-fit linear multiple regression equation using regression analysis to reveal the morphometric variability of wild M. nipponense. This study will provide useful information for better understanding the relationships between morphometric traits and body mass in diverse habitats and enrich the foundation for M. nipponense selective breeding research.

2. Materials and Methods

2.1. Sample Collection

Samples were collected from three locations in the upper reaches of the Huaihe River of China: the Huaihe River main stream (HH population), Suyahu Reservoir (SYH population), and Wuyue Reservoir (WY population), in August 2024 (Figure 1). The sampling environmental parameters were as follows: HH: water temperature 32.40 (±0.28) °C, pH value 8.06 (±0.37), dissolved oxygen 7.79 (±0.01) mg/L, ammonia nitrogen 0.41 (±0.25) mg/L, chlorophyll a 2.75 (±0.66) μg/L, and water flow rate 0.2 m/s. WY: water temperature 32.24 (±0.07) °C, pH value 8.78 (±0.57), dissolved oxygen 7.23 (±0.09) mg/L, ammonia nitrogen 0.22 (±0.02) mg/L, chlorophyll a 15.58 (±0.81) μg/L, and water depth 9.5–13.9 m. SYH: water temperature 27.47 (±0.56) °C, pH value 7.94 (±0.31), dissolved oxygen 6.92 (±0.30) mg/L, ammonia nitrogen 0.49 (±0.14) mg/L, chlorophyll a 19.30 (±4.25) μg/L, and water depth 1.3–3.1 m.

All samples were captured by a cage net (4 mm mesh size). After excluding specimens with incomplete appendages and egg-bearing females, 288 samples were measured, of which 122 samples were from HH, 80 were from SYH, and 86 were from WY.

2.2. Measurement of Morphological Traits

Before weighing, absorbent paper was used to dry the body surface to obtain an accurate measurement of body mass. The electronic scale (Meilen: Shenzhen Mobil Electronics Co., Ltd., Shenzhen, China) used for weighing was accurate to 0.01 g. In total, 26 morphometric traits (

Table 1. Description of the morphometric parameters of M. nipponense used in the study.

No.	Morphometric Parameters	Abbreviation	Description
1	Total length	TL	Distance from rostrum tip to the distal tip of the telson with prawn stretched out
2	Body length	BL	Distance from the posterior margin of the right orbit to telson tip
3	Rostrum length	RL	Distance from epigastric tooth basis to rostrum tip
4	Carapace length	CL	Distance from the posterior margin of the right orbit to the midpoint of the posterior margin of the carapace
5	Carapace width	CW	Distance between lateral margin of cephalothoraxes
6	Carapace height	CH	Distance between dorsal and ventral margin of carapace
7	Abdominal length	AL	Distance from the telson tip to the midpoint of anterior margin of the first abdominal somite
8	First abdominals somite length	AL1	Distance between the midpoint of the posterior margin and the anterior margin of first abdominals somite
9	Second abdominals somite length	AL2	Distance between the midpoint of the posterior margin and the anterior margin of second abdominals somite
10	Third abdominals somite length	AL3	Distance between the midpoint of the posterior margin and the anterior margin of third abdominals somite
11	Fourth abdominals somite length	AL4	Distance between the midpoint of the posterior margin and the anterior margin of fourth abdominals somite
12	Fifth abdominals somite length	AL5	Distance between the midpoint of the posterior margin and the anterior margin of fifth abdominals somite
13	Sixth abdominals somite length	AL6	Distance between the midpoint of the posterior margin and the anterior margin of sixth abdominals somite
14	Abdominal width	AW	Distance between lateral margin of first abdominals somite
15	Abdominal height	AH	Distance between dorsal and ventral margin of first abdominals somite
16	Telson length	TeL	Distance from posterior margin of sixth abdominal somite to telson tip
17	Telson width	TeW	Distance between lateral margin of telson taken in his basis
18	Telson height	TeH	Distance between dorsal and ventral margin of telson
19	Second pereiopods length	P2L	Distance from the proximal of ischium to dactylus tip
20	Ischium length	P2L1	Distance between the proximal and the distal margin of ischium
21	Merus length	P2L2	Distance between proximal and distal margin of merus
22	Carpus length	P2L3	Distance between the proximal and the distal margin of carpus
23	Propodus length	P2L4	Distance between proximal and distal margin of palm
24	Dactylus length	P2L5	Distance between proximal and distal margin of dactylus
25	Caudal fan length	FL	Distance between the posterior margin of sixth abdominal somite and the posterior margin of caudal fan
26	Caudal fan width	FW	Maximum distance between lateral margin of the caudal fan

) were measured using IP54 digital display Vernier calipers (Syntek: Deqing Shengtaixin Electronic Technology Co., Ltd., Huzhou, China), accurate to 0.01 mm.

2.3. Path Analysis

The mean value, standard deviation (SD), correlation analysis, and path analysis were calculated using SPSS 20.0. The coefficient of variation (CV) for each of the 26 recorded morphometric traits and body mass were estimated as follows: CV = (SD/mean) × 100%. According to GOMES (1985) [19], the CV is classified as low (CV < 10%), medium (CV between 10% and 20%), high (CV% between 20% and 30%), and very high (CV > 30%).

The correlation coefficient was estimated as follows:

$$r_{xy}={\sum }_{i=1}^n{\displaystyle \frac{(X_i-\overline{X})(Y_i-\overline{Y})}{\sqrt{{\sum }_{i=1}^n{(X_i-\overline{X})}^2{(Y_i-\overline{Y})}^2}}}$$

(1)

where r_xy is the correlation coefficient between traits X and Y; X_i is the independent variable; $\overline{X}$ is the mean value of the independent variable; Y_i is the dependent variable; and $\overline{Y}$ is the mean value of the dependent variable.

The direct path coefficients (path coefficient, P) can be obtained directly, as described by Du and Chen [20]. The determination coefficient was calculated using the formulas:

$$d_i=P_i^2$$

(2)

$$d_{ij}=2r_{ij}{P}_i{P}_j$$

(3)

where d_i is the direct determination of ith trait on the body mass and d_ij is the co-determination of ith trait on the body mass through jth trait (i ≠ j); P_i and P_j are the path coefficients of ith and jth trait on the body mass, respectively; and r_ij is the correlation coefficient between ith and jth trait.

A stepwise multiple regression analysis was used to eliminate nonsignificant morphometric traits, and Student’s t test (α = 0.05) was applied to evaluate significance. The multiple regression equation for body mass (Y) was calculated as follows:

$$Y=a+b_1{X}_1+b_2{X}_2{+b}_3{X}_3+\cdots {+b}_i{X}_i$$

(4)

where Y is the dependent variable, a is the intercept, X_i are the independent variables, and b_i are the partial regression coefficients for X_i on Y.

2.4. Grey Relational Analysis

According to grey system theory [21,22], the body mass and 26 morphometric traits were selected to be a grey system. The body masses were reference sequences (X₀), while the 26 morphometric traits were comparison sequences (X_i, i = 1, 2, 3, …, 26).

Due to different dimensions among different influence factors, the first step was linear normalization of raw data. The data preprocessing was performed by the following equation:

$$X_i^{\text{'}}\left(k\right)={\displaystyle \frac{X_i\left(k\right)-\overline{X_i}}{\mathit{\sigma }}}$$

(5)

where $X_i^{\text{'}}\left(k\right)$ is the value after standardization; k is the population number (k = 1 to 3); $X_i\left(k\right)$ is the value of each morphometric traits; $\overline{X_i}$ is the average value of $X_i\left(k\right)$; σ is the standard deviation of $X_i\left(k\right)$; and i is the morphometric traits number (i = 1 to 26).

Then, the grey relational coefficient was calculated as:

$${\mathit{\xi }}_i\left(k\right)={\displaystyle \frac{\mathit{min}{\Delta }_i\left(k\right)+\mathit{\rho }\mathit{max}{\Delta }_i\left(k\right)}{{\Delta }_i\left(k\right)+\mathit{\rho }\mathit{max}{\Delta }_i\left(k\right)}}$$

(6)

where ${\mathit{\xi }}_i\left(k\right)$ is the grey relational coefficient, which is the relationship between the best and the actual normalized data; ${\Delta }_i\left(k\right)$ are the absolute values between the reference sequence and comparison sequences, ${\Delta }_i\left(k\right)=\left|X_0\left(k\right)-X_i\left(k\right)\right|$; $min{\Delta }_i\left(k\right)$ and $max{\Delta }_i\left(k\right)$ are the minimum and the maximum value of the second level, respectively; and ρ is the distinguishing coefficient (ρ = 0.5).

Finally, the grey relational grade was calculated as follows:

$$r_i={\displaystyle \frac{1}{n}}\sum _k^n{\mathit{\xi }}_i\left(k\right)$$

(7)

where $r_i$ is the grey relational grade and n is the number of performance characteristics.

3. Results

3.1. Descriptive Statistics of Parameters and Correlation Coefficients

The mean, SD, and CV for the 26 morphometric traits and body mass of wild M. nipponense are presented in

Table 2. Descriptive statistics of the morphometric traits and body mass among different geographical populations of wild M. nipponense.

Trait	HH Population (N = 122)			SYH Population (N = 80)			WY Population (N = 86)
	Mean	SD	CV	Mean	SD	CV	Mean	SD	CV
BW/g	2.92	1.73	59.27%	1.92	0.67	35.01%	1.18	0.38	31.88%
TL/mm	62.05	10.71	17.26%	56.91	5.43	9.54%	49.63	4.64	9.36%
BL/mm	50.03	8.27	16.53%	45.11	4.57	10.13%	39.33	3.61	9.18%
RL/mm	12.02	2.95	24.53%	11.80	1.50	12.70%	10.29	1.81	17.63%
CL/mm	16.00	2.90	18.14%	13.59	1.66	12.22%	13.04	1.57	12.07%
CW/mm	9.08	1.94	21.41%	8.08	1.15	14.27%	6.88	0.74	10.83%
CH/mm	9.93	1.76	17.70%	8.82	1.02	11.60%	7.96	0.86	10.83%
AL/mm	34.03	5.65	16.60%	31.52	3.27	10.39%	26.29	2.89	10.98%
AL1/mm	2.44	0.41	16.98%	2.15	0.29	13.39%	1.97	0.37	18.62%
AL2/mm	4.68	0.87	18.63%	3.62	0.55	15.32%	3.44	0.58	17.01%
AL3/mm	5.67	0.90	15.89%	4.70	0.52	11.18%	4.29	0.54	12.59%
AL4/mm	4.22	0.84	20.02%	3.72	0.66	17.84%	3.55	0.79	22.35%
AL5/mm	3.22	0.55	17.21%	2.60	0.36	13.81%	2.42	0.33	13.45%
AL6/mm	5.17	0.93	17.96%	4.47	0.56	12.59%	4.12	0.43	10.36%
AW/mm	7.22	1.46	20.27%	6.73	0.93	13.84%	6.01	0.65	10.88%
AH/mm	8.00	1.38	17.23%	7.57	0.85	11.27%	6.78	0.77	11.43%
TeL/mm	8.67	1.40	16.18%	7.58	0.81	10.72%	6.91	0.74	10.71%
TeW/mm	2.40	0.53	21.94%	2.12	0.34	15.90%	1.83	0.33	17.84%
TeH/mm	1.67	0.39	23.04%	1.30	0.24	18.68%	1.24	0.20	15.97%
FL/mm	11.01	1.84	16.68%	9.86	0.93	9.43%	9.14	0.74	8.06%
FW/mm	15.83	3.30	20.82%	14.21	1.63	11.44%	12.57	1.30	10.32%
P2L/mm	50.65	20.65	40.77%	52.23	12.82	24.54%	41.72	11.41	27.35%
P2L1/mm	8.67	2.38	27.45%	8.32	1.64	19.71%	6.95	1.63	23.49%
P2L2/mm	10.10	4.04	40.01%	10.52	2.61	24.79%	8.53	2.23	26.09%
P2L3/mm	14.03	6.20	44.15%	14.69	3.95	26.88%	11.34	3.40	29.99%
P2L4/mm	10.37	5.16	49.73%	10.29	3.03	29.49%	8.02	2.54	31.60%
P2L5/mm	7.48	3.41	45.53%	8.41	2.09	24.79%	6.87	2.09	30.38%

. In the case of the morphometric traits, the CV of P2L4 in HH, SYH, and WY populations were the highest, which were 49.73%, 29.49%, and 31.60%, respectively, whereas the CV for AL3 in HH (15.89%) and FL in SYH (9.43%) and WY (8.06%) were the lowest. The CV on the body masses of the HH, SYH, and WY populations were 59.27%, 35.01%, and 31.88%, respectively, which were higher than any of the morphometric traits in the same population. The CV of body mass of M. nipponense in the HH population was larger than that of the SYH and WY populations.

After the Kolmogorov–Smirnov normality test, the body masses of wild M. nipponense in the three geographical populations all obey normal distribution (p > 0.05). The result suggested that the data satisfied the condition of regression analysis. The correlation coefficients of various traits of different geographical populations of wild M. nipponense are shown in Figure 2. In the HH population, the results showed that there was a striking association between the phenotypic traits of 351 pairs (Table S1, p < 0.05); the highest correlation coefficient was found in the correlation between P2L and P2L3, with a value of 0.992, followed by the correlations of P2L-P2L4 (0.989) and P2L-P2L2 (0.986). All 26 morphometric traits were significantly correlated with body mass (p < 0.01), with the correlation coefficient ranging from 0.348 (between body mass and AL4) to 0.904 (between body mass and CW). In the SYH population, there was a significant association between the 340 pairs of phenotypic traits (Table S2, p < 0.05): the highest correlation coefficient was found in the correlation between P2L and P2L3 (0.985), followed by the correlations of P2L-P2L2 (0.972) and BL-TL (0.970). All 26 morphometric traits were remarkably correlated with body mass (p < 0.05), with the correlation coefficient ranging from 0.250 (between body mass and AL4) to 0.929 (between body mass and BL). In the WY population, there was a remarkable correlation between the phenotypic traits of 300 pairs (Table S3, p < 0.05), and the strongest correlation coefficient was found in the correlation of P2L-P2L3 (0.982), followed by P2L-P2L2 (0.970) and P2L-P2L4 (0.957). Except for AL4, which had no significant correlation with body mass (p > 0.05), with a correlation coefficient of 0.152, the other 25 morphometric traits were remarkably correlated with body mass (p < 0.05), with the correlation coefficient ranging from 0.210 (between body mass and AL2) to 0.890 (between body mass and BL).

3.2. Path Analysis of Morphometric Traits on Body Mass

The direct and indirect effects of the significant morphometric traits on the body mass of wild M. nipponense among different geographical populations are shown in

Table 3. Path analysis of significant morphometric traits on body mass among different geographical populations of wild M. nipponense.

Population	Trait	Correlation Coefficient	Direct Effect	Indirect Effect
HH				P2L4	BL	P2L1	CH	AL2	∑
	P2L4	0.890	0.539 **		0.383	−0.220	0.232	−0.043	0.351
	BL	0.899	0.472 **	0.438		−0.213	0.247	−0.045	0.427
	P2L1	0.826	−0.246 **	0.483	0.408		0.233	−0.052	1.072
	CH	0.897	0.270 **	0.462	0.432	−0.213		−0.055	0.627
	AL2	0.410	−0.093 *	0.251	0.231	−0.138	0.158		0.503
SYH				BL	P2L4	AH	AL6	TeL	∑
	BL	0.929	0.332 **		0.139	0.228	0.110	0.121	0.597
	P2L4	0.733	0.229 **	0.201		0.134	0.079	0.091	0.504
	AH	0.900	0.253 **	0.299	0.121		0.111	0.119	0.651
	AL6	0.796	0.151 **	0.241	0.119	0.186		0.098	0.645
	TeL	0.909	0.137 *	0.292	0.152	0.220	0.108		0.772
WY				BL	CH	AL5	AH		∑
	BL	0.890	0.583 **		0.291	−0.061	0.076		0.307
	CH	0.824	0.384 **	0.442		−0.067	0.065		0.440
	AL5	0.314	−0.147 **	0.240	0.175		0.045		0.461
	AH	0.656	0.116 *	0.382	0.214	−0.057			0.540

** represents significant difference (p < 0.01); * represents significant difference (p < 0.05).

. In the HH population, P2L4, BL, and CH exerted a greater direct effect on body mass (p < 0.01), while P2L1 and AL2 had a negative direct effect on body mass (p < 0.05). The direct effect of P2L4 (0.539) and BL (0.472) were most important, and they were higher than the indirect effects on body mass. In the SYH population, BL, P2L4, AH, AL6, and TeL showed significant direct effects on body mass (p < 0.05), ranging from 0.137 (TeL) to 0.332 (BL). However, the direct effects of these traits were lower than the corresponding indirect effects on body mass. In the WY population, BL, CH, AL5, and AH exhibited a significant direct effect on body mass (p < 0.05). Only the direct effect of BL (0.583) was greater than the indirect effect (0.307) on body mass.

The determination coefficients of the morphometric traits on body mass are listed in

Table 4. Determination coefficients of morphometric traits on body mass among different geographical populations of wild M. nipponense.

Population	Trait	Determination Coefficients
HH		P2L4	BL	P2L1	CH	AL2	∑
	P2L4	0.291	0.413	−0.238	0.250	−0.047	0.905
	BL		0.223	−0.201	0.233	−0.043
	P2L1			0.061	−0.115	0.026
	CH				0.073	−0.029
	AL2					0.009
SYH		BL	P2L4	AH	AL6	TeL
	BL	0.110	0.092	0.151	0.073	0.080	0.950
	P2L4		0.052	0.061	0.036	0.042
	AH			0.064	0.056	0.060
	AL6				0.023	0.030
	TeL					0.019
WY		BL	CH	AL	AW
	BL	0.340	0.340	−0.071	0.089		0.865
	CH		0.147	−0.052	0.050
	AL5			0.022	−0.013
	AH				0.013

The co-determinant and determination coefficients are shown on the off-diagonal and the diagonal (highlighted in bold), respectively.

. The sum of the determination coefficients on body mass in the HH, SYH, and WY populations was 0.905, 0.950, and 0.865, respectively. In the HH population, the determination coefficient of P2L4 was the largest (0.291), followed by BL (0.223), whereas that of AL2 was the lowest (0.009). The co-determinant coefficient of P2L4 and BL on body mass was the highest at 0.413. In the SYH population, the determination coefficient of BL was most important (0.110). The co-determinant coefficient of BL and AH on body mass was the highest, which was 0.151. In the WY population, the BL had the most remarkable decisive effect on body mass (0.340). The co-determinant coefficient of BL and CH on body mass was the largest, with a value of 0.340.

3.3. Construction of Multiple Regression Equations

The regression coefficient test of the morphometric traits on body mass are presented in

Table 5. Regression coefficient test among different geographical populations of wild M. nipponense.

Population	Model	Partial Regression Coefficient	Standard Error	t	p	VIF
HH	Constant	−4.116	0.439	−9.372	0.000
	P2L4	0.181	0.024	7.454	0.000	6.353
	BL	0.099	0.017	5.966	0.000	7.611
	P2L1	−0.179	0.058	−3.084	0.003	7.727
	CH	0.265	0.087	3.044	0.003	9.536
	AL2	−0.184	0.075	−2.466	0.015	1.726
SYH	Constant	−3.973	0.191	−20.766	0.000
	BL	0.049	0.01	4.833	0.000	6.960
	P2L4	0.051	0.008	6.393	0.000	1.895
	AH	0.199	0.053	3.737	0.000	6.792
	AL6	0.181	0.048	3.772	0.000	2.378
	TeL	0.113	0.053	2.129	0.037	6.121
WY	Constant	−2.507	0.176	−14.234	0.000
	BL	0.061	0.007	8.405	0.000	2.901
	CH	0.167	0.028	5.94	0.000	2.525
	AL5	−0.169	0.054	−3.161	0.002	1.308
	AH	0.056	0.027	2.117	0.037	1.820

. It can be found that all variance inflation factor (VIF) values were below 10, indicating that there is no multicollinearity among the retained traits. According to the test of the significance of partial regression coefficients of these significant traits, the regression equations of the body masses of different geographical M. nipponense populations were constructed as follows:

HH population:

Y = −4.116 + 0.181 P2L4 + 0.099BL − 0.179 P2L1 + 0.265CH − 0.184AL2 (R² = 0.905)

(8)

SYH population:

Y = −3.973 + 0.049BL + 0.051P2L4 + 0.199AH + 0.181AL6 + 0.113TeL (R² = 0.950)

(9)

WY population:

Y = −2.507 + 0.061BL + 0.167CH − 0.169 AL5 + 0.056AH (R² = 0.866)

(10)

The results of the analysis of variance on the multiple regression equations are shown in

Table 6. Analysis of variance of multiple regression equations among different geographical populations of wild M. nipponense.

Population	Index	Sum of Squares	df	Mean Square	F	p
HH	Regression analysis	326.92	5	65.384	219.883	0.000
	Residual	34.494	116	0.297
	Total	361.414	121
SYH	Regression analysis	33.795	5	6.759	280.84	0.000
	Residual	1.781	74	0.024
	Total	35.576	79
WY	Regression analysis	10.348	4	2.587	130.532	0.000
	Residual	1.605	81	0.020
	Total	11.953	85

. The p value of three equations were < 0.01, indicating that the prediction of body mass, in relation to these significant morphometric traits, was reliable.

3.4. Grey Relational Analysis of Morphometric Traits on Body Mass

Figure 3 illustrates the grey relational analysis of 26 morphometric traits on the body mass of wild M. nipponense among different geographical populations. In the HH population, the relational grade of the factors on body mass ranged between 0.755 and 0.922. The order of the top five grey relational grades was CW > CL > BL > TL > P2L. In the SYH population, the grey relational grade of the morphometric traits ranged from 0.770 to 0.901. The highest grey relational grade was found in BL, followed by CW > CL > TL > AH. In the WY population, the range of 0.775–0.891 was the relational grade of the factors on body mass. The main factors affecting the body mass were BL > CW > TL > CH > AL.

4. Discussion

We analyzed 26 morphometric traits and body mass from three geographical populations of M. nipponense, and there was a relatively high CV ranging from 8.06% for the FL of WY population to 59.27% for the body mass of HH population. The abundant morphological variability across all sites suggested substantial internal variation within the population, potentially indicating high genetic diversity of M. nipponense in the upper reaches of the Huaihe River. Comparative descriptive statistics indicated a higher CV for body mass than other morphological traits across three populations: HH (59.27%), SYH (35.01%), and WY (31.88%). This result is consistent with Wang et al. (2022), who found higher CVs for body mass (25.25% for females and 29.26% for males) relative to other morphometric traits in a domestic M. nipponense population [8]. Elevated CVs for body mass have also been documented in other aquatic species, such as Parabramis pekinensis [11], Pampus argenteus [23], and Scylla paramamosain [24]. The second pereiopods (chelicerae) of M. nipponense were well developed and play a crucial role in predation and defense. Among the three populations, the CV for traits associated with second pereiopods were 19.71~49.73%, after that for body mass. It suggests that the second pereiopods traits were highly susceptible to environmental influences, which was in agreement with the studies conducted by Li et al. [14] on crayfish in different culture environments. Furthermore, the CV of body mass and the second pereiopods traits of M. nipponense from the main stream were larger than those of M. nipponense from the SYH and WY reservoirs, which is probably related to environmental factors, such as stream flow rate, water depth, concealment, and bait abundance.

In aquaculture breeding programs, the value of CV is closely related to genetic potential. Traits with greater CV often exhibit higher levels of variability and respond better to selective breeding, which are typically prioritized as target breeding traits. Body mass had a larger CV than other morphological traits; thus, it is a major selection trait for breeding. However, body mass is influenced by genetics, physiological, and environmental factors, and it has proven difficult to achieve satisfactory results in the selection programs when only taking body mass into account [23,25]. Body mass is closely correlated with morphometric traits in fish, spawn, and molluscs. Imanpour Namin et al. [26] found that the body mass of M. nipponense had a highly significant correlation with TL in both sexes. Wang et al. [8] found that the body mass of M. nipponense was mainly affected by BL, CL, AL, CW, AW, AH, and abdominal meat weight. In the present study, all morphometric traits, except for AL4 in the WY population, were observed to correlate significantly positively with body mass (p < 0.05). This means that it is feasible to select body mass through the selection of one or more other morphometric traits in the M. nipponense breeding program. Nevertheless, correlation analysis can only measure the degree of relationship between morphometric traits and body mass and cannot clarify the specific scale of their role or degree of influence on body mass.

Path analysis employs a model based on multiple linear regression that decomposes the correlation coefficient into a direct path coefficient (the direct effect of an independent variable on the dependent variable) and an indirect path coefficient (the indirect effect of the independent variable on the dependent variable through other in dependent variables) to directly compare the relative importance of each factor [14,24]. Path analysis has been widely used in the genetic selection of aquatic animals [24,27,28,29]. Our results revealed that, out of 26 morphometric traits, P2L4, BL, P2L1, CH, and AL2 in the HH population; and BL, P2L4, AH, AL6, and TeL in the SYH population showed significant direct effects on body mass (p < 0.05), whereas in the WY population, BL, CH, AL5, and AH had significant direct effects on body mass (p < 0.05). Not all morphometric traits significantly correlated with body mass were retained, and the traits retained in each population were different; this finding further indicates that correlation analysis fails to comprehensively consider the interactions among various morphological traits and also confirms the necessity of conducting path analysis. The effects of morphometric traits on the body mass of M. nipponense were dissimilar at different populations; this is similar to the findings of Huang et al. [27] in M. nipponense, Li et al. [28] in Haliotis discus hannai, and Zhang et al. [29] in Larimichthys crocea. In this study, the sum of the determination coefficients and the multiple correlation indices of the three populations were both greater than 0.850, indicating that the traits retained by path analysis were the main morphometric traits affecting the body mass of each population [11,24]. Among these, the co-determination coefficient of P2L4 and BL (0.413) to body mass for the HH population; BL and AH (0.151) for the SYH population; and the determination coefficient of BL (0.340) for the WY population were the largest. Liu et al. [30] found that the three top effective factors on the body mass of M. nipponense in both sexes from lower reaches of the Yangtze River were TL, BL and P2L. Huang et al. [27] reported that the traits most correlated with body mass in five provinces of China were BL and carapace traits. Due to wide distribution of M. nipponense and its high adaptability, the morphological characteristics exhibit variations across distinct habitats, necessitating consideration of phenotypic alterations influenced by environmental factors in breeding programs.

The grey relational analysis, developed by Deng [21,22] based on grey system theory, describes the correlation between various factors by assessing their degree of geometric similarity. It has been proven to be useful for dealing with poor, incomplete, and uncertain information [31], and it has begun to be used in morphological traits [14] and physiological traits [32] for aquatic animals. This study employed a grey relational analysis to investigate the relationships between morphometric traits and body mass in M. nipponense and found that the key traits influencing body mass differed at different populations. Specifically, the top five traits for body mass were CW, CL BL, TL, and P2L for the HH population; BL, CW, CL, TL, and AH for the SYH population; and BL, CW, TL, CH, and AL for the WY population.

By comparing the morphological traits retained by the pathway analysis with the top five traits ranked by the grey relational analysis in terms of grade, it was concluded that the trait primarily correlated with body mass in the HH population was BL (grade rank three), the most relevant traits for body mass in the SYH population were BL (grade rank one) and AH (grade rank five), and in the WY population, the traits most correlated with body mass were BL (grade rank one) and CH (grade rank four). The results show that these two methods were in agreement with each other, but the discrepancies were observed in the order of influence of morphometric traits on the body mass obtained by the two methods. Path analysis is suitable for high-capacity sample data, but grey relational analysis is suitable for low-capacity sample data. Furthermore, path analysis can analyze the relationships among independent variables, whereas grey relational analysis evaluates independent variables individually, without regard of their significance or collinearity [33]. Various analytical methods are founded on distinct principles with specific data prerequisites, so it is important to choose suitable statistical methods or conduct comparisons across multiple analytical methods to ascertain the key factors influencing the target traits.

5. Conclusions

In the present study, body mass had the highest CV, followed by the second pereiopods traits of M. nipponense. All morphometric traits, except for AL4 in the WY population, were observed to correlate significantly positively with body mass (p < 0.05). According to the path analysis and grey relational analysis, the traits primary correlated with body mass were BL in the HH population, BL and AH in the SYH population, and BL and CH in the WY population. The main morphometric traits affecting body mass in different populations were diverse, but BL always maintained a high correlation with body mass. Considering the influence of environmental factors on the morphological characteristics of wild populations, future research could be strengthened by incorporating analyses of key environmental parameters (e.g., water quality, temperature, habitat type) alongside morphometric data. This study will be useful for further research about the relationships among morphological traits and for the selective breeding of M. nipponense and other crustacean species.