Microphis retzii (Bleeker, 1856)

Haÿ, Vincent, Mennesson, Marion I., Carpentier, Camille, Dahruddin, Hadi, Sauri, Sopian, Limmon, Gino, Wowor, Daisy, Hubert, Nicolas, Keith, Philippe & Lord, Clara, 2025, Phylogeography of Microphis retzii (Bleeker, 1856) and Microphis brachyurus (Bleeker, 1854) in the Pacific, Journal of Fish Biology 106 (2), pp. 602-620 : 613-614

4.2 | M. retzii , a species complex?

The haplotype network obtained for M. retzii revealed three distinct haplogroups: one in EI (Ceram/ Ambon / Papua / Sulawesi, haplogroup 1), one in WI (Bali / Java /Lombok, haplogroup 2), and one in the NPO ( China and Taiwan, haplogroup 3) (Figure 4), and no haplotypes are shared between these three haplogroups. This phylogeographic pattern suggests at least distinct mitochondrial divergences among these sets of populations as a result of limited connectivity between these regions. It is currently unknown if M. retzii is amphidromous. Haÿ et al. (2023b) have validated an amphidromous life cycle for Microphis nicoleae , a closely related species to M. retzii , with a relatively short marine phase of 19.7 ± 5.8 days. However, it is important to note that the life cycle of a taxon is not fixed and can vary. The loss of amphidromy is quite common in fish species or populations ( Liao et al., 2020; Murase & Iguchi, 2019) and has already been observed in freshwater pipefish ( Lord et al., 2024). The marked genetic structuring of these three lineages could be partly explained by biotic factors, such as life-cycle variations (i.e., facultative amphidromy or short marine duration), which limit dispersal and enhance geographic isolation.

These three lineages are found in different areas in Southeast Asia, which is divided into several biogeographic subregions (or hotspots), of which three are represented here: the Sunda Shelf (represented by haplogroup 2), Wallacea (represented by haplogroup 1), and Philippines (represented by haplogroup 3) ( Woodruff, 2010) (Figure 4). These subregions present complex biogeographical histories, leading to major vicariance events ( Hutama et al., 2016; Lohman et al., 2011). Indeed, genetic structuring can be significantly impacted by biogeography and environmental conditions; rapid changes during the geological history of a region can create barriers to dispersal, which in turn limits gene flow ( Lohman et al., 2011; Sholihah et al., 2021a; Sholihah et al., 2021b; Wibowo et al., 2023). Several events on the scale of geological time have caused successive interruptions of connectivity in Southeast Asia. For example, sea-level fluctuations during the glacial cycles of the Pleistocene (2.7 MAY– 11,700 years) have led to the establishment of geographical barrier by connecting Borneo, Sumatra, and Java to the mainland, a process that happened repeatedly during the late Pleistocene ( Sholihah et al., 2021a; Sholihah et al., 2021b; Woodruff, 2010). Moreover, the Makassar Strait (known as Wallace's line), between the Sunda Shelf and Wallacea, although known to serve as a marine barrier to the dispersal of land animals to Borneo and Sulawesi, could be involved as a dispersal barrier to marine organisms and therefore lead to the genetic isolation of amphidromous species. For instance, sharp genetic breaks were described for populations of the mantis shrimp Haptosquilla pulchella among these oceanographic regions, suggesting that Wallace's line has a role in shaping species distribution and population structure ( Barber et al., 2000). Murphy and Austin (2005) also suggested a possible effect of Wallace's line on Macrobrachium rosenbergii , a freshwater prawn, for which strong genetic divergences are observed between the Australian and Thai populations. Therefore, oceanic currents in Southeast Asia could be involved in the isolation of lineages on each side of Wallace's lines like the Indonesian Throughflow current passing in the Makassar Strait ( Godfrey, 1996), thus reducing the connectivity between these two subregions. Isolation of the NPO population ( Taiwan and China) can also be influenced by a combination of currents present in this area. The presence of the Kuroshio Current (Figure 6), on the western side of the NPO basin, could act as a dispersal barrier and promote lineage diversification or population differentiation, as it has been observed in some marine organisms and the gobioid Periophtalmus modestus ( He et al., 2015) . Iida et al. (2010) have also shown the important role of the Kuroshio Current to maintain the population structure in the amphidromous goby S. japonicus from Taiwan to northern Japan, thus limiting the range of this species to the islands of the North Pacific. The isolation of the different lineages of M. retzii may have been influenced by these past and current barriers.

1.8

1.6

1.4

1.2

1

0.8

0.6

0.4

0.2

0

Time

(

MYA

)

Considering our results, what is the taxonomic status of these different lineages? Compared to M. brachyurus , mitochondrial MRCA for M. retzii lineages is relatively more ancient and was dated around 1.8 MYA (Figure 5). However, these divergence age estimates should be considered with caution, as the existence of discontinuous gene flow and divergent mitochondrial lineages within the complex may violate the assumptions of the phylogenetic reconstructions of haplotypes presented here (i.e., the lack of genetic structuring). The percentages of divergence between individuals from the three geographical areas are relatively high, ranging from 3.3% between NPO ( China and Taiwan) and WI to 5.1% between WI and EI (Table 6), with high and significant Φst values (Table 5). These results suggest that the three mitochondrial haplogroups of M. retzii represent closely related species as follow: (i) M. retzii (Bleeker, 1856) (type locality: Manado, Sulawesi, Indonesia) present in Sulawesi, Ceram, Ambon, and in Papua (haplogroup 1) (ii) M. cf. 1 retzii present in WI on the islands of Java, Bali, and Lombok (haplogroup 2), and (iii) M. cf. 2 retzii present in the North Pacific, China, and Taiwan (haplogroup 3). Each of these mitochondrial lineages is therefore restricted to limited geographical areas. These results were expected as endemism is high in this region (Parenti, 2011). Endemism between various close islands of the Indo-Pacific and Indonesia has already been observed ( De Mazancourt et al., 2020; Dwiyanto et al., 2021; Haÿ et al., 2021; Jamonneau et al., 2024; Keith et al., 2015; Lord et al., 2012; Wibowo et al., 2023). Indonesian ichthyofauna hosts several radiations of morphologically similar species, and the use of molecular approaches allows us to uncover hidden diversity, including either cryptic or unnamed taxa ( Hubert et al., 2015; Kottelat & Lim, 2021; Sholihah et al., 2021a; Sholihah, Delrieu-Trottin, Sukmono, et al., 2021b; Utami et al., 2022). The species of this complex are morphologically very similar, but their differentiation, based on genetics and geography, constitutes strong argument in favor of elevating them to the species level. These results warrant a taxonomic revision of M. retzii in Indonesia based on the analysis of nuclear markers, including more specimens and more localities (especially in Borneo, Sulawesi, and the Philippines), and a detailed examination of their morphological characters to revalidate or describe these two potentially new taxa from Java/Bali and from China / Taiwan. This work is in progress.