Spatiotemporal variations in length-weight relationship, growth pattern and condition factor of Periophthalmus variabilis Eggert, 1935 in Vietnamese Mekong Delta

Length-weight relationship (LWR), growth pattern and condition factor (CF) play a vital role in fish resource evaluation and management but data on this is limited for Periophthalmus variabilis. This is an amphibious fish that lives in the mudflats of the mangrove forests in the Western Pacific regions, including the Vietnamese Mekong Delta (VMD). This paper contributes to the understanding of the LWR, growth pattern and CF of P. variabilis. Fish specimens were collected by hand at four sites in the VMD from April 2020 to March 2021. The analysis of 495 individual fish (259 females and 236 males) showed that fish weight could be estimated from a given fish length due to high determination values (r2 = 0.70 − 0.97). Dusky-gilled mudskippers of the VMD exhibited positive allometry, as indicated by their larger than 3 b value (b = 3.094 ± 0.045, p = 0.04). However, the growth pattern of the mudskipper between the sex and maturation stage (immature versus mature) showed isometry. These fish displayed positive allometry in the dry season (b = 3.138 ± 0.065) (>3, p = 0.04) but isometry in the wet season (b = 3.058 ± 0.061) (≈3, p = 0.34). Fish growth ranged from isometry to positive allometry and varied by site (b = 2.850 ± 0.103–3.370 ± 0.114) and month (b = 2.668 ± 0.184–3.588 ± 0.299) based on the b value. The CF of P. variabilis was not affected by sex (p = 0.29), body size (p = 0.64) or season (p = 0.43), but was affected by site (p = 0.01) and month (p = 0.01). The CF of this species (1.05 ± 0.02) was higher than 1 (p < 0.001), indicating that the fish adapted well to their habitat.

The mudflats and mangroves are habitats for a number of animal species including fish (Sanders et al., 2010;Sasmito et al., 2020). The mudskipper is a unique fish group that lives mainly in these habitats (Murdy, 1989) and can obtain oxygen directly from the air using their skin and gills (Jaafar, Perrig & Ming Chou, 2009). Periophthalmus is one of the largest genera of mudskippers with 19 described species found throughout the world (Murdy & Jaafar, 2017). Periophthalmus of the VMD is comprised of three species: P. chrysospilos, P. gracilis and P. variabilis (Tran et al., 2013). Of these three species, P. variabilis is found quite frequently in mudflat and mangrove regions (Murdy, 1989;Kottelat et al., 1993;Jaafar & Hou, 2012;Tran et al., 2013;Tran et al., 2020;Tran & Dinh, 2021;Tran et al., 2021a) and can move flexibly in and out of the water to catch prey (Wicaksono et al., 2020). In the VMD, our observations show that the number of individual P. variabilis has decreased over time. However, there is no data on its biology and ecology. This study was conducted to document fish LWR, growth pattern and CF to further understand the ecological adaptations of the P. variabilis population.

Study site and fish analysis
The present research was carried out in four locations along the estuarine and coastal regions in the VMD, including Duyen Hai-Tra Vinh (DHTV,9 • 40 29.5 (Fig. 1). There are two seasons at these sites: the dry season from January to May and the wet season from June to December (Le et al., 2006). The dominant plants in DHTV, TDST and DHBL are Sonneratia caseolaris, Avicennia marina and Bruguiera gymnorrhiza, respectively. In DDCM, A. marna and B. gymnorrhiza are equally dominant (Dinh et al., 2021a). The pH ranged from 7.6-8.0, and the salinity varied widely from 12.3 to 23.5%. The pH varied with the site but not with the season, whereas the salinity varied with the season but not the site (Dinh et al., 2021a).
Fish samples were collected monthly by hand for 4-hours at a time in an area of 120 m 2 (6-m width × 20-m length) from April 2020 to March 2021. Fish specimens were easily distinguished from congeners using their external features (Murdy & Jaafar, 2017). Periophthalmus variabilis are covered by many irregular blackish dots, whereas P. chrysospilos and P. gracilis are covered by small orange spots and are greyish brown with several irregular narrow white bars, respectively. The roughly triangular first dorsal fin is found on P. variabilis and P. chrysospilos, but not on P. gracilis. Filamentous spines are found in the first dorsal fin of P. chrysospilos but are not found in P. variabilis. MS222 was used to anaesthetize fish specimens before being preserved in a 5% formalin buffer and shipped to the laboratory. In the laboratory, fish sex was differentiated using genital papilla, e.g., a triangle shape in males and an oval shape in females. Fish total length (TL) was measured to the nearest 0.1 cm using a ruler, and fish weight (W ) was recorded to the nearest 0.01 g using an electric scale.

Data analysis
The LWR was determined using the formula W = a ×TL b (W : fish weight, a: intercept parameter, TL: fish total length, and b: slope parameter) (Ricker, 1973). The condition factor (CF ) was calculated as suggested by Le Cren (1951) using the formula CF = W /(a The quality of LWR s was determined by using the determination coefficient (r 2 ) (Metin et al., 2011). The t -test was performed to verify if the b value obtained from the LWR s was close to a cubic value. Species displayed positive allometry (b > 3), negative allometry (b < 3) and isometry (b =3) (Martin, 1949). The t -test was applied to confirm whether the CF varied by sex, size and season, while one-way ANOVA was used to assess any variations in th CF between months and sites (Mahmood et al., 2012). The t -test was used to verify if the CF was significantly different from the threshold of 1. A General Linear Model was used to confirm if the interactions of sex × season, sex × site and season × site affected the CF (Dinh, 2016a). The length at first maturity (L m ) was used to divide fish into the immature group if TL < L m and into the mature group if TL ≥ L m (e.g., L m values of females and males were 5.0 cm and 5.7 cm in DHTV; 4.6 cm and 5.8 cm in TDST; 4.9 cm and 5.2 cm in DHBL; and 6.2 cm and 5.9 cm in DDCM, unpublished data). SPSS v.21 was used for data analysis and all tests were set at p < 0.05. The Benjamini-Hochberg procedure was used to decrease the Type I error of all tests (Benjamini & Hochberg, 1995;McDonald, 2014).
The growth pattern of P. variabilis did not change based on sex as both males and females showed isometry, suggesting that the different ovarian and testicular weights did not regulate fish growth type. A similar growth pattern in the two sexes was found in P. barbarus in Nigeria (King & Udo, 1998) but not in P. modestus in the VRD (Tran, Nguyen & Ha, 2021b). The growth pattern of P. variabilis was not impacted by size as evidenced by immature and mature groups both exhibiting isometry. This was also observed in P. barbarus in Nigeria (King & Udo, 1998) but not in P. modestus in the VRD (Tran, Nguyen & Ha, 2021b). Like its congener, P. modestus in the VRD (Tran, Nguyen & Ha, 2021b), the growth pattern of P. variabilis in the dry season was different from in the wet season, indicating that the difference in precipitation between these two seasons affected fish growth type. The various vegetation among the four sites could have an impact on P. variabilis' growth patterns as well. The impact of environmental conditions on fish growth patterns was also found in P. waltoni in Nigeria (Sarafraz et al., 2012) and P. modestus in the VRD (Tran, Nguyen & Ha, 2021b). Like P. variabilis, some gobies living in and out of the VMD, e.g., P. barbarus (King & Udo, 1998), P. waltoni (Sarafraz et al., 2012), G. giuris (Dinh & Ly, 2014;Phan et al., 2021), B. boddarti (Dinh, 2014b), P. serperaster (Dinh et al., 2016), T. vagina (Dinh, 2016c), P. schlosseri (Dinh, 2016b) and P. modestus (Tran, Nguyen & Ha, 2021b) showed variations in their growth patterns that fluctuated based on the month.
Like B. koilomatodon in the VMD (Lam & Dinh, 2021), the CF of P. variabilis was not affected by its developmental stage as its CF did not change with sex and size. Likewise, the CF of P. variabilis's congener, P. barbarus, in Nigeria did not indicate changes in the CF based on sex differences (King & Udo, 1998;Chukwu & Deekae, 2011). By contrast, P. modestus, another congener in the VRD, showed a change in its CF due to sex as this value was high in females towards the end of gonadal maturation. Like P. variabilis, the CF of some gobies in the VMD, e.g., Parapocryptes serperaster (Dinh et al., 2016), P. schlosseri (Dinh, 2016b), T. vagina (Dinh, 2016c) and G. giuris (Phan et al., 2021) did not vary with size. P. variabilis could adapt well to environmental conditions as its CF showed a similar pattern in the dry and wet seasons. In contrast, the dry season was preferable for P. modestus in the VRD as its CF value in the dry season was higher than it was in the wet season (Tran, Nguyen & Ha, 2021b). Like P. variabilis, a similar CF value was found between the dry and wet seasons in co-occurring gobiid species such as P. elongatus (Tran, 2008), P. serperaster (Dinh et al., 2016), T. vagina (Dinh, 2016c), G. giuris (Phan et al., 2021) and B. koilomatodon (Lam & Dinh, 2021). The difference in abiotic and biotic factors among sites (Dinh et al., 2021a) could be impacting the CF of P. variabilis as it showed a spatial variation in its CF. This same spatial variation in the CF was also found in the co-occurring goby species B. koilomatodon (Lam & Dinh, 2021) but was not found in the P. modestus species in the VRD (Tran, Nguyen & Ha, 2021b). Although P. variabilis showed spatiotemporal variations in its CF, overall its CF was higher than the threshold of 1 which could indicate that the study site had favourable environmental conditions. Congeners living outside of the VMD e.g., P. barbarus (King & Udo, 1998), P. chrysospilos (Abdullah & Zain, 2019) and P. modestus (Tran, Nguyen & Ha, 2021b) were also well adapted to their habitats as evidenced by higher CFs. This was also observed in some other fish species in the VMD, e.g., P. elongatus (Tran, 2008), P. serperaster (Dinh et al., 2016), T. vagina (Dinh, 2016c), P. schlosseri (Dinh, 2016b), G. aureus (Dinh, 2019), P. chrysospilos (Dinh et al., in press) and G. giuris (Phan et al., 2021).

CONCLUSIONS
Due to high determination values fish weight can be estimated using a given fish length. As the slope value obtained from its LWRs was higher than the cubic value, P. variabilis displayed positive allometry. The growth pattern of this species was not affected by sexual and intraspecific factors since male, female, immature and mature fish displayed isometric growth as indicated by a b value ≈3. The CF was not affected by sex, size or season, but did vary by site and month. The CF of this species was higher than 1, showing it adapted well to the environment. These results can play a vital role in the assessment and management of the P. variabilis population in the VMD.