Length-Weight Relationships and Other Morphological Traits of Fishes in the Mangrove of Hainan, China

: The length–weight relationships (LWR) and other morphological traits for 6417 specimens in 74 ﬁsh species collected seasonally, from July 2020 to April 2021, in the mangrove of Dongzhaigang Bay, Hainan Province, China, are presented. This involved, for all species, a sample size, and minimum and maximum lengths; in addition, for most species, it involved linear relationships between the standard length and body height and width, the height of the caudal peduncle, the pectoral ﬁn length, and the eye diameter. This extensive coverage of mangrove-resident ﬁsh species (as opposed to species using the mangrove only as a nursery) is a ﬁrst for Hainan. The various morphological traits of the mangrove ﬁsh are, for several species, the ﬁrst to be published since these species’ original descriptions.


Introduction
Mangroves are one of the most important marine ecosystems, supporting small-scale fisheries throughout the tropical and subtropical world [1]; however, their extant has declined dramatically, mainly due to shrimp farming and other coastal developments [2][3][4][5]. The fisheries in mangrove areas are based on two groups of fish: one consisting of species whose ontogeny is fully confined to a mangrove ecosystem; and the other group consisting of fish that use the mangrove only as a nursery, and eventually leave for adjacent ecosystems, such as, e.g., a mudflat or coral reef.
Chiefly because of their small sizes, but also because they are usually exploited by subsistence or small-scale commercial fishers in rural and often remote areas of poorer countries [6], the fishes of mangroves are understudied; i.e., the major features of their life history are not available in the scientific literatures. In the following, we present the first extensive study of morphological traits of mangrove fishes from the Hainan Province; the Province has the highest mangrove cover in all of China ( Figure 1). Among these traits, we emphasize the length-weight relationship (LWR) because, in addition to being an important trait, LWR is also used for various practical purposes; these include turning growth curves in length into growth curves in weight, as required for fish population assessments [7].

Material and Methods
This study is based on data collected during field surveys by Xiamen Univ the Dongzhaigang National Reserve, Hainan Province, China from July 2020 2021. Fish were collected seasonally using 10 m nets with 1 cm meshes. All the c fish specimens were immediately put on ice and transported to a laboratory wh were weighted to the nearest 0.1 g and length measurements (standard, fork, a length; body height and width; height of the caudal peduncle; pectoral fin length diameter) were taken to the nearest millimeter.
The parameters of the length-weight relationships (LWR) of the form W = a estimated by two methods. One consisted of plotting the standard length (L) vs (W) in form of linear regressions of log (W) vs. log (L). These regressions have in (α) which, after elimination of a few obvious outliers, provide estimates of a = addition, they have slopes that provide estimates of b; while their coeffic determination (r 2 ) indicate the faction of the variance in L and W explained by t [8,9].
The other method was used for species for which either few L-W data pa available, or for which specimens were measured that covered only a narrow r sizes. This method consists of assuming b to be equal to 3, as is frequently the calculating values for each available L-W data pair and averaging them; or usin in cases where only one L-W data pair was available [7].
With both methods, the LWR were computed only for species whose m

Material and Methods
This study is based on data collected during field surveys by Xiamen University in the Dongzhaigang National Reserve, Hainan Province, China from July 2020 to April 2021. Fish were collected seasonally using 10 m nets with 1 cm meshes. All the collected fish specimens were immediately put on ice and transported to a laboratory where they were weighted to the nearest 0.1 g and length measurements (standard, fork, and total length; body height and width; height of the caudal peduncle; pectoral fin length and eye diameter) were taken to the nearest millimeter.
The parameters of the length-weight relationships (LWR) of the form W = a·L b were estimated by two methods. One consisted of plotting the standard length (L) vs. weight (W) in form of linear regressions of log (W) vs. log (L). These regressions have intercepts (α) which, after elimination of a few obvious outliers, provide estimates of a = 10 α ; in addition, they have slopes that provide estimates of b; while their coefficients of determination (r 2 ) indicate the faction of the variance in L and W explained by the LWR [8,9].
The other method was used for species for which either few L-W data pairs were available, or for which specimens were measured that covered only a narrow range of sizes. This method consists of assuming b to be equal to 3, as is frequently the case [9]; calculating values for each available L-W data pair and averaging them; or using it as is in cases where only one L-W data pair was available [7].
With both methods, the LWR were computed only for species whose maximum observed length was above 50% of the maximum length provided by FishBase [10] for the species in question. The purpose of this was to ensure that: (i) LWR were computed only for fishes that mostly occur in the mangrove, rather than use them only as nursery grounds; and (ii) the LWR presented here do not refer only to juveniles, but also include at least young adults. Fish species that did not meet the 50% criterion were considered to use the mangrove only as a nursery for their young; and their length-weight relationships were not computed, although their other traits were.
Finally, linear regressions were run for all species for which there were sufficient numbers of measurements, covering a sufficient range of lengths: between their standard length, and their body height and width; the height of the caudal peduncle; and the pectoral fin length and eye diameter.

Results
We collected 74 fish species belonging to 12 orders and 68 families during the sampling period; however, the above criteria allowed the estimation of the LWR for only 36 species, 22 using linear regression (Table 1) and 14 by assuming b = 3 ( Table 2). Figure 2A    ) are provided to document: (1) that no species is included that did not have at least one L-W data pair above 50% of the maximum length in FishBase; this ensures that species that use the mangrove only as nurseries are not included; and (2) that the estimate of 'a' is based on adult fish, and not juveniles.  Table 2 for more cases similar to (B-D), and the text for comments on the assumption that b = 3.  Table 2 for more cases similar to (B-D), and the text for comments on the assumption that b = 3.
The linear relationships between the standard length and various morphological traits of fish species with enough data are illustrated, for example, by the greenback mullet (Liza subviridis) in Figure 3; and they are documented in tabular and/or graphic form for the other 64 species we investigated. The Supplementary Materials include species which grow well beyond double the maximum length that was sampled, and which therefore can be assumed to use the mangrove only as a nursery. For these species, the traits reported here should be assumed to apply only to their juveniles.

PEER REVIEW 5 of 7
The linear relationships between the standard length and various morphological traits of fish species with enough data are illustrated, for example, by the greenback mullet (Liza subviridis) in Figure 3; and they are documented in tabular and/or graphic form for the other 64 species we investigated. The Supplementary Materials include species which grow well beyond double the maximum length that was sampled, and which therefore can be assumed to use the mangrove only as a nursery. For these species, the traits reported here should be assumed to apply only to their juveniles.

Discussion
This study estimated the LWR in numerous species of mangrove fish for which important information was lacking in the literature, and thus in FishBase [10] as well. Hence, the data in Tables 1 and 2 will fill numerous gaps; even though FishBase compensates for the gaps by generating the LWR via a Bayesian routine, from

Discussion
This study estimated the LWR in numerous species of mangrove fish for which important information was lacking in the literature, and thus in FishBase [10] as well. Hence, the data in Tables 1 and 2 will fill numerous gaps; even though FishBase compensates for the gaps by generating the LWR via a Bayesian routine, from taxonomically related species of similar shape [11]. The point here is that the LWR of Table 2, even if based on only a few L-W data pairs, or even on one, will improve the predicted LWR.
The computed LWR (Table 1) had a mean exponent b = 3.03, which corroborated the hypothesis that 3 should be the default assumption in LWR studies [9]; and which implies that, with b ≈ 3, the fish in question maintain their overall shape as they grow from juveniles to adults, i.e., their growth is isometric.
One species that did not fit well in our study was the goldfish Carassius auratus, of which one specimen of 19.3 cm SL was caught in the mangrove. C. auratus can briefly tolerate brackish water [12]; however, it can reproduce only in freshwater. Thus, we assume that this specimen escaped from a nearby freshwater pond or was discharged from an aquarium.
Current biodiversity assessments tend to focus on taxonomic diversity, such as species richness or abundance; these do not account for the different biological characteristics or traits of species; i.e., for their functional diversity [13]. Yet, species with different morphological traits perform different roles in ecosystems [14]. Thus, considering functional diversity can contribute to better understand how ecosystems function and provide services [15].
Therefore, the data on the morphological diversity of the fishes of the mangroves of Hainan presented therein, jointly with related data on fish biodiversity in FishBase [10], will help in future work aiming to characterize these ecosystems.