The sea nettle jellyfish of the genus Chrysaora are classified under the family Pelagidae (order Semaeostomeae) and are typically recognized by having 32–48 lappets, with eight marginal sense organs, with three or more tentacles per octant, with 16 gastrovascular pouches, and with numerous warts on the exumbrella (Kramp, 1961). Species of sea nettle have a worldwide distribution and have been reported to occur in the South China Sea, North, Central and South America, Africa, Europe and Australia (Morandini and Marques, 2010; Yap and Ong, 2012). According to the World Register of Marine Species (http://www.marinespecies.org) there are possibly 15–18 species, whereby 5 species have been verified: Chrysaora achlyos, Chrysaora hysoscella, Chrysaora pacifica, Chrysaora quinquecirrha and Chrysaora chinensis.

Although some regions of Southeast Asia have been reported to harbour Chrysaora species such as Chrysaora quinquecirrha and Chrysaora melanaster (Kramp, 1961; Yap and Ong, 2012), there is the possibility they were misidentified base on colour variations and this possibility warrants verification. In the past, jellyfish species were notoriously difficult to identify as there were no reliable taxonomic keys, specimens were badly preserved, and confounded by the existence of cryptic species complexes that could be detected only by the application of molecular genetic techniques. For example, C. chinensis in Malaysia has been previously identified as C. hysoscella or C. quinquecirrha based on reports of envenomation and toxicology studies of jellyfish stings (Azila and Othman, 1993) and those in the Singapore Straits as C. melanaster (Yap and Ong, 2012)— possibly misidentified due to the colour variations. Some features, such as nematocysts have been used to aid in the species identification, hence sea nettles from the South China Sea were believed to be C. chinensis (Morandini and Marques, 2010; Yap and Ong, 2012). More recently, applications of DNA sequence analysis and phylogenetics have helped identify and barcode some of the Malaysian jellyfish species, including C. chinensis (Rizman-Idid et al., 2016), which concurred the notion that specimens with different colour morphs are often genetically similar and regarded as the same species. However, it is important to realize that the efficiency of such DNA barcoding method depends on the availability of reference sequences of correctly identified voucher specimens in the GenBank database. Although the study provided 16S and ITS1 sequences, it did not have the required COI sequences to definitively barcode C. chinensis. Furthermore, the morphological description of the species in the study was relatively simple and preliminary. Hence a more detailed morphological evaluation of the Malaysian sea nettle (C. chinensis) is required.

Chrysaora chinensis in Malaysia has tendencies to bloom, sometimes causing blockage of cooling systems of coastal power plants, contamination of fishing nets and a nuisance to fishing activities (pers. obs.). Beach tourism is also affected as beach goers are often warned about its stings that are intense with painful burning sensation. In general, jellyfish blooms have been linked to eutrophication and nutrients added to the diet of the jellyfish (Purcell et al., 2007; Richardson et al., 2009). Moreover, morphological adaptations to different localities with various ecological and environmental conditions are well documented for jellyfish (Dawson, 2005). Therefore, this present study aims to distinguish morphologically the populations of Malaysian C. chinensis by using geometric morphometrics, especially between populations of the Straits of Malacca that are heavily impacted by anthropogenic activities from those found in the South China Sea where the water quality is somewhat better.

Geometric morphometric analysis is the quantitative study of the biological shape, shape variation and covariation, which until recently have been applied to study the shape of organisms such as oak leaves (Viscosi and Cardini, 2011), dogs (Drake and Klingenberg, 2010) and cichlids (Maderbacher et al., 2008). Traditional multivariate morphometrics usually analyse multiple measurements together by using canonical variates analysis (CVA), principal components analysis (PCA), whereas geometric morphometrics combine multivariate analysis and methods for the direct visualization of biological form, making it appealing to researchers since the results are no longer presented only as a series of statistical data but also as graphical representations of the actual organism being studied (Rohlf and Marcus, 1993; Adams et al., 2004). Furthermore, these visualizations provide information on morphological changes in their immediate anatomical context (Klingenberg, 2013; Bookstein, 2015).

It is noteworthy that the revival of C. chinensis was made recently by Morandini and Marques (2010) based on specimens from Philippines and China and assumed the species is distributed throughout the South China Sea. Other than the work by Yap and Ong (2012) on specimens found in the Singapore Straits, there has been no in-depth morphological study of the C. chinensis from Malaysia waters, other than the preliminary taxonomic survey and molecular documentation undertaken by Rizman-Idid et al. (2016). Thus, this study is the first of its kind to provide a more detailed morphological description of C. chinensis of Peninsular Malaysia and to employ geometric morphometric analysis to distinguish its populations.

The detailed morphological description of C. chinensis in the present study represents the first of its kind in Malaysia. Chrysaora chinensis was originally described by Vanhöffen (1888), and known to occur in South China Sea (Vanhöffen, 1911). The species is commonly found in all beaches in Malaysia. Based on personal observations and interviews with fishermen, they occur more abundantly at more polluted areas, particularly during the dry season (April to October). Prior to the resurrection of C. chinensis by Morandini and Marques (2010), many Chrysaora specimens found in the South China Sea and Malay Archipelago have been identified as either C. melanaster (Yap and Ong, 2012) or C. helvola (Kramp, 1961). However, Morandini and Marques (2010) emphasized that C. chinensis is very different from C. helvola (valid name C. fuscescens) and C. melanaster due to the differences in tentacle and lappet numbers. Both of the latter species also are larger than their observed specimens of C. chinensis. Recent studies by Yap and Ong (2012) ascribed the specimens from St. John’s Island in Singapore as C. chinensis. Likewise, findings from this present study, which focused on specimens obtained from the east and west coasts of Peninsular Malaysia confirms that the sea nettles found in these waters were morphologically identified as C. chinensis. Furthermore, DNA sequences from some of the specimens obtained in the present study (unpublished data) revealed that they are C. chinensis and genetically similar and phylogenetically clustered to those reported by Rizman-Idid et al. (2016). Although there is no correction to the species description by Morandini and Marques (2010), the morphological description of C. chinensis from this study is more detailed than that by Rizman-Idid et al. (2016). It is noteworthy that the colourations of C. chinensis can vary greatly. Some specimens are transparent whereas some are reddish with radiating stripes on the exumbrella with pigmentations on oral arms and reddish-brown lappets. Nonetheless, this colouration is corroborated by the colouration diagnosis of C. chinensis by Morandini and Marques (2010) and Yap and Ong (2012).

Both PCA and CVA were used for data analysis of geometric morphometrics, but the result of CVA better illustrates the shape variation between the different geographical. CVA is a type of analysis that efficiently summarizes the maximal difference between more than two pre-defined groups (relative to within-group variance), unlike PCA which is a quick way to explore overall variation across the dataset (such as outliers and clusters of specimens) and may not be able to detect differences between groups (Webster and Sheets, 2010). In this study, the result of CVA shows that the shape difference is mainly between the east and west coast, where the shape of the gastrovascular pouch of specimens from the east coast is slightly elongated compared to those from the west coast. Unlike Rhizostomeae jellyfish, C. chinensis do not have well developed canal. The gastrovascular pouches of C. chinensis are used to store nutrients from the central stomach that permeates to various parts of the jellyfish. The east coast of Peninsular Malaysia is facing the South China Sea, whereas the west coast is facing the Straits of Malacca, one of the busiest sea routes in the world, hosting annually about 65 000 vessels for international navigation, and up to another additional 15 000 fishing vessels (Ibrahim and Khalid, 2007). The low water quality of the Straits of Malacca is due in a large part to pollution from the shipping industry, aquaculture, a densely populated coastal areas resulting in nutrients and effluent discharged into the straits, increasing the total suspended solids and decreases the dissolved oxygen, thus causing eutrophication (Rezai et al., 2003; Bong and Lee, 2008; Praveena et al., 2011). Human activities that lead to eutrophication have been associated with the bloom of the jellyfish (Arai, 2001; Parsons and Lalli, 2002; Malej et al., 2007; Purcell et al., 2007; Richardson et al., 2009). Jellyfish appear to be capable of adapting to fluctuating levels of water quality and other environmental parameters, such as salinity, temperature, food source (Purcell et al., 2012). According to Brotz (2011), there were different trends of jellyfish populations in the coastal ecosystems and oceans worldwide. Using data that were aggregated and analysed at the scale of Large Marine Ecosystem (LME), he showed there is an increase in native and invasive jellyfish populations in the Bay of Bengal, which included the Straits of Malacca in the LME’s definition. In contrast, the abundance of such jellyfish populations in the South China Sea was stable or variable.

Since the main function of the gastrovascular system is for the circulation of nutrients (Arai, 1997), these differences in gastrovascular pouch shapes observed in the present study may be a form of morphological adaptation of C. chinensis populations to the different conditions of the coastal areas of Peninsular Malaysia. The landmass of Peninsular Malaysia acts as the main physical barrier which separates the east from the west populations. The EC population also appears distinct from the EN population. Although the distance between the sampling areas of both populations is only approximately 350 km; a distance that may not be of significant magnitude to explain the isolation by distance for these populations; these populations may have been kept separated due to occurrences of eddies and differing water circulations at 5°N, just off the east coast of Peninsular Malaysia (Daryabor et al., 2016). According to Daryabor et al. (2016), such water circulations may influence the distribution of the nutrient balance in regulating primary productivity and the changes in the marine ecosystem.

The present study has successfully demonstrated that internal structure such as the gastrovascular pouch is suitable, as the result appears to be robust with low measurement errors. Although geometric morphometric analysis has been proven useful in analysing and detecting even very minute shape variations in various organisms (Klingenberg, 2013), its application as demonstrated in the present study on gelatinous organism such as jellyfish, may still be informative albeit with a few practical challenges. Geometric morphometric analysis seems more suited for organisms with rigid structures, as landmarks are more easily identified and configured, thus detection of shape variation would be attributed due to real morphological differences rather than inconsistencies in measurements due to mishandling of specimens. To date, there has been no shape variation study using geometric morphometric analysis on soft or gelatinous organisms. Jellyfish, being gelatinous, can be quite easily distorted, either from bad sampling methods that can damage their delicate structures or even from prolonged preservation of specimens. In fact, formalin can contribute to the change of shape of jellyfish, especially after a long period of immersion (Kapiris et al., 1997). Therefore, using geometric morphometric analysis to detect shape variation in jellyfish poses a higher degree of difficulty compared to rigid organisms.

The practical usage may be improved by choosing a different structure for landmark configurations, depending on the species. As for other jellyfish species, such as Rhopilema esculentum of the Rhizostomeae family are bigger in size and generally more robust, such that their external or internal rigid structures may be chosen as a better target for geometric morphometric study.

Although the study of jellyfish is still new in Malaysia, the present study contributes much needed baseline data for the country and region. The present study has successfully ascribed the sea nettles found in Peninsular Malaysia as C. chinensis and demonstrated that its populations are morphologically varied between the different coastal areas base on gastrovascular pouch shape. Application of geometric morphometric analysis may also facilitate future studies of jellyfish to either delineate species or distinguish populations.

We sincerely thank Tek Ho, the staff of Institute of Biological Sciences (ISB) and Institute of Ocean and Earth Sciences (IOES), University of Malaya for their assistance in the field. We are also grateful to Siti Aisyah Hj. Alias from IOES for helping to proofread the manuscript.