Conchology Variations in Species Identification of Pachychilidae (Mollusca, Gastropoda, Cerithiodea) through Multivariate Analysis

Pachychilidae is one of the freshwater gastropod family which was previously known under the Potamididae and Thiaridae families. Studies on freshwater gastropods especially on conchcology examinantions are still inadequate compared to marine gastropods. Morphological and morphometric studies of gastropods are practically used to identify and differentiate between species and necessary to complement molecular studies due to its low cost and tolerable resolving power of discrimination. The aim of the current study is to provide information on morphological and morphometric characteristics of Pachychilidae in Bintulu, Sarawak stream. A total of 20 individuals from each species of Sulcospira testudinaria, Sulcospira schmidti, Brotia siamensis, and Tylomelania sp. from Pachychilidae familiy were collected at three different sites from a small stream within the Bintulu area. Fourteen measurement of shell morphometrics were converted into proportioned ratios and analysed for univariate and multivariate analysis. Three shell morphometric (Aperture width, AW; Whorl width, WW2; and, Interior anterior length, AINL) of Pachychilidae indicated significant differences (P < 0.05) between species. However, multivariate analysis revealed that these shell morphometrics are pre-eminent factors to discriminate genus Sulcospira, Brotia and Tylomelania, as well as between Sulcospira species. This current study also suggests that these three characteristics are unique to Sulcospira species due to strong distinction among species. Findings on these three characteristics are significant for Sulcospira spp. as this study is the first shell morphometric report on the Pachychilidae species in Sarawak.

• The study suggests that three shell characteristics are unique to Sulcospira species due to strong distinction among the species.

INTRODUCTION
Freshwater gastropods can be classified into two groups which are Prosobranchia or gilled gastropods and Pulmonata or lunged gastropods. There are approximately more than 500 families of freshwater Prosobranchia and 150 families of freshwater Pulmonata found all over the world (Leveque et al. 2005). Family Pachychilidae is under the subclass Prosobranchia along with five other families within Sorbeconcha clade (Strong et al. 2008). Pachychilidae was previously known as Potamididae or Thiaridae (Poppe & Tagaro, 2006). Nevertheless, study on freshwater gastropods taxonomy is inadequate compared to marine gastropod. Species discrimination can be done through shell morphology, reproduction characteristic (Köhler et al. 2004), molecular phylogenies (Köhler & Glaubrecht, 2003) and morphometric variation (Pramithasari et al. 2017). Shell morphology is commonly designed for evolutionary studies (Seilacher 1984;Palmer 1985). However, morphological studies on snails are still practically used to identify and differentiate between species. Morphological and morphometric studies of gastropod shells are important, even though present-day molecular technology is widely used to identify organisms precisely (Prioli et al. 2002). Variations between gastropod shells are not only caused by genetic changes, but also by environmental characteristics and existence of predators (Preston & Roberts 2007). In addition, genetic analysis is very costly compared to morphometric analysis that requires analysis of data using available software and hardware such as thermal cycler and DNA sequencer. The usage of morphometric analysis has obtained wide acceptance in the current biological scene as it is increasingly used as a necessary complement to molecular studies due to its reasonable cost and tolerable resolving power of discrimination (Doyle et al. 2018). Therefore, morphometric and morphology studies are faster compared to genetic analysis that usually takes more time to produce results. Species identification can be done in-situ based on morphological observation and this helps to avoid the misidentification of species.
Gastropods form significant roles in the ecology and promote worldwide economy. In ecology, gastropods are reliable water quality indicators and abundance of certain species may indicate or alarm of any water impoundment (Savic & Randjelovic 2016;Galan et al. 2015). Economically, many molluscs are edible thus became important due to human consumption and as the alternative source of protein (Hamli et al. 2012;Robinson 1999). Despite their ecological and economic importance, the world's population of freshwater gastropods battles with loss and decline in number. Freshwater gastropods especially may encounter threats from habitat loss and degradation (Strong et al. 2008). In fact, a total number of 279 species had extinct whilst 1,794 threatened Gastropoda species were listed in the Red List of Threatened Species in 2012.
There are scarce number of studies about Pachychilidae in Malaysia, particularly on morphometric and morphological studies. While in Sarawak, there are handfuls of published studies about Pachychilidae and one of them was conducted by Hamli et al. (2013). Hamli et al. (2013) mentioned that there were 21 species of edible gastropods found in markets around Sarawak (Sibu, Mukah, Bintulu, Kuching, Miri, Limbang and Lawas) and two species are from Pachychilidae. Malaysia is one of the regions that is rich in biodiversity, however there is limited information can be found recorded on malacofauna communities especially freshwater gastropods, in Sarawak. Therefore, this study aims to determine on morphological and morphometric characteristics of Pachychilidae that essentially contribute to information in freshwater ecosystem.

MORPHOLOGICAL STUDY
Samples were brought back to the laboratory for cleaning and visceral mass removal. Gastropod species was identified based on Köhler and Glaubrecht (2002), Köhler and Glaubrecht (2005), and Köhler and Simonis (2012). The gastropod shell was then photographed using camera (Cybershot DSCW810, Sony) from the apertural and abapertural view.
A total of 15 shell morphometrics was measured using Mitutoyo digital callipers at 0.00 ± mm following Sherely's (1996) (Fig. 2, Table 1). All shell morphometric were divided to shell length as proportion ratio to ensure no ontogeny bias.  Sherely (1996). Detail shell morphometrics and characteristics as shown in Table 1.

STATISTICAL ANALYSIS
The proportion ratio of all measured morphometrics were analysed using analysis of variance (ANOVA) with Tukey's mean comparison test using statistical analysis computer software (SAS) version 9.4. Significant differences of all shell morphometrics between species were further tested with Principal Component Analysis (PCA) and Cluster analysis using Paleontological Statistics (PAST) Version 3.14.

<0.0001
Note: Different superscripts indicate a significantly difference (P < 0.05) using Tukey mean comparison test.

Multivariate Analysis
Four and two clusters had been formed at 0.04 and 0.1 distances, respectively, based on the AW, WW2, and AINL characteristics (Fig. 4). Most samples of S. schmidti belong to Cluster I except one individual sample. However, one sample of B. siamensis and S. testudinaria were clustered along within Cluster I. Cluster II was representative of S. testudinaria due to 17 out of 20 samples belonging to this species. However, four samples within this cluster were associated to B. siamensis and S. schmidti. A total of 7 from 20 Tylomelania sp. samples were grouped under Cluster III and the rest of Tylomelania sp. clustered within Cluster IV together with B. siamensis. Therefore, Cluster IV was a combination of Tylomelania sp. and B. siamensis. Both S. testudinaria and S. schmidti were grouped in A which corresponds to the genus Sulcospira. However, B. siamensis and Tylomelania sp. were grouped under Cluster B due to overlapping shell morphometric. The hierarchical clustering of Cluster A, B, I, II, III, and IV was similarly interpreted through scattered plot of PCA which indicated prominent separation between S. testudinaria and S. schmidti compared to B. siamensis and Tylomelania sp. (Fig. 5). The principal components and shell morphometric showed a strong correlation for the first principal component with AW and AINL compared to WW2 (Table 3). The first principal component is primarily a measure of shell AINL (0.74215) which can be used as a tool to separate between S. testudinaria, S. schmidti, B. siamensis, and Tylomelania sp. Moreover, the second and third principal components mostly relate to shell AW and WW2, respectively for species distinction.
A proportion of AINL is considered a unique characteristic that can differentiate S. schmidti from the rest of Pachychilidae species found. Shell AW and WW2 can also be used to support the differentiation of species of Pachychilidae in the present study.

DISCUSSION
Univariate analysis showed eight measurement of shell morphometric can be used to differentiate among species. However, only three measurement of shell morphometrics were proportionate to shell length (AL, WW2, AINL) and able to verify variation between four species of Pachychilidae. Köhler and Glaubrecht (2006) used five shell characteristics to discriminate between the two species of Pachychilidae, which are the minimum measurement of shell morphometric study. However, Köhler and Glaubrecht (2006) study was further supported by molecular work, thus make distinction among species more concrete. Most of the previous studies on gastropod shell morphometrics used at least four to thirteen shell characteristics (Dillon Jr. & Jacquemin 2015;Ismail & Elkarmi 2006;Köhler & Glaubrecht 2006;Chiu et al. 2002). Common shell characteristics used in morphometric studies are shell length, shell width, body whorl length, penultimate whorl width, aperture length, and aperture width. Nevertheless, current findings only indicated that aperture width was able to differentiate between species compared to other common shell features used. Application of proportion ratio to this study is crucial to eliminate biases of sample age or ontogeny. Ontogeny is bias that affect morphology characteristic as organisms grow and develop (Herrel et al. 2016). The application has also been adapted for Meretrix spp. inner shell differentiation (Hamli et al. 2016). Therefore, it is important for shell measurements to be proportionate with shell length before further statistical analysis been done. Three shell characteristics AW, WW2 and AINL had indicated species distinction of Sulcospira genus compared to Brotia and Tylomelania genus. Furthermore, clear distinction between S. schmidti and S. testudinari indicates that AW, WW2, and AINL are strong differentiation characteristics for Sulcospira spp. However, differentiation between Brotia and Tylomelania was not apparently in this study. According to Köhler and Glaubrecht (2002), Brotia and Tylomtelania genus display similar characteristics which sometimes lead to both species being identified as Brotia. Additionally, both species have similar habitats and are distributed in the Southeast Asian region (Köhler et al. 2004).
Brotia and Tylomelania are able to be differentiated through reproductive biology characteristics, specifically a subhaemocoelic brood pouch and brood pouch formed by the pallial oviduct respectively (Köhler et al. 2004). Therefore, additional methods or other shell characteristics are needed to differentiate between Brotia and Tylomelania. Most morphometric analysis involves another method to increase accuracy of species discrimination. Generally, shell morphometric analysis is combined with molecular analysis using mitochondrial DNA (Potkamp et al. 2017;Köhler & Dames 2009). Other delineation studies are a combination of shell morphometrics with reproductive morphology (Köhler & Glaubrecht 2003), nervous system, radula, or alimentary system (Köhler & Glaubrecht 2006). Dillon and Jacquemin (2015) categorised shell morphometrics into two groups, traditional morphometrics and geometric morphometrics. The current study is categorised under traditional morphometrics which is based on linear measurement. Geometric morphometrics is a recent morphometric analysis method that is based on digitised landmarks of shells which was first introduced by Johnston et al. (1991). The use of geometric morphometrics has increased as an alternative for shell morphometric analysis that is widely accepted (Doyle et al. 2018;Dillon & Jacquemin 2015;Farman & Almukhtar 2015;Cruz et al. 2012).

CONCLUSION
Based on the findings, application of more tools in morphometric studies must be emphasised to ensure strong differentiation between species and less dependence on just one practice. More discoveries on shell characteristics and additional techniques are needed to differentiate between Brotia sp. and Tylomelania sp. due to weak differentiation methods between these genera.