Caprella scauroides Mayer, 1903

Peart, Rachael A., Woods, Chris M. C., Sutherland, Judy E. & Cox, Serena L., 2019, Confirmation of Caprella scauroides Mayer, 1903 (Crustacea: Amphipoda) from New Zealand, using integrative techniques, Zootaxa 4686 (3), pp. 361-375 : 362-371

Caprella scauroides Mayer, 1903 has not previously been recorded from New Zealand waters. However, recent collections in New Zealand include samples that appeared to be C. scauroides . In the present study these specimens were morphologically identified and confirmed to be this species. This identification was further verified with DNA sequence data from the mitochondrial cytochrome oxidase I marker (COI) and the nuclear small subunit ribosomal RNA gene (18S). Therefore, this study constitutes the first record of this species from New Zealand waters.

Materials and methods

Sample collection. Two separate sample submissions were received by the Marine Invasives Taxonomic Service (MITS) at the National Institute of Water & Atmospheric Research Ltd (NIWA) during May 2017, both from Ōkahu Bay, Waitematā Harbour ( Fig. 1 ). These sample submissions originated from staff at a vessel haul-out facility (The Landing) at Ōkahu Bay (near the Ōrakei Marina ) reporting large numbers of an unusual caprellid present on several vessels being cleaned of biofouling to the Ministry for Primary Industries (MPI), the lead agency with oversight of national biosecurity in New Zealand. As part of their resulting investigation response, MPI commissioned an additional diver search sample. These were collected from vessel swing moorings through the Summer 2016–17 Marine High Risk Site Surveillance (MHRSS) ( Seaward et al. 2015) survey being conducted in Waitematā Har- bour by NIWA. The sample (MHRSS sample ID AKL24ExtraDive1; MITS ID 73604; 36.85014°S, 174.81415°E; to depth of 5 m; 24 May 2017) consisted of 50+ individual caprellid specimens (males and females, varying life stages, including juveniles and mature life stages). Shortly afterwards, another sample of caprellids was collected by the cleaning staff at The Landing from a recreational vessel on 29 May 2017 and sent to MITS, via MPI. This sample (sample ID AAP14933; MITS 73603, 73683 and 73687) consisted of 65 adult caprellid specimens (very large males, ovigerous females and some juveniles). As a result of the submission of these samples and consequent identification of a new-to-New Zealand species, as well as the extension of records from Mcleod’s Bay, Whāngārei Harbour, detected during a MHRSS survey in November 2017, an older voucher sample of caprellids held by MITS was re-examined. In January 2005, a single sample of over 30 individuals was collected from the hull of the recreational vessel Aoix, in Nelson Marina , as part of a study on biofouling on recreational vessels ( Figure 1 ). This sample consisted of large males, ovigerous and gravid females and many juveniles. These caprellids were identified at the time as C. californica and which was not regarded at that stage as established in New Zealand ( Ahyong & Wilkens 2011).

Morphological methods. All the caprellid specimens were examined, dissected, drawn and photographed for morphological assessment. Takeuchi & Oyamada (2013) was used as a reference point for characters and comparisons as it was the most comprehensive description of the differences between C. scauroides and C. californica . Reference specimens are stored in the NIWA Invertebrate Collection (NIC). Photograph was taken by iphone6S and processed using Adobe Photoshop. Morphological figures were drawn using a camera lucida on a Leica MZ12s microscope, digitally inked using an A3 Wacom Intuos pro tablet and plates constructed using Adobe Photoshop.

Molecular methods. DNA was extracted from three specimens using the Qiagen DNeasy Blood and Tissue kit (Qiagen GmbH, Hilden) according to the manufacturer’s instructions. Varying amounts of tissue were used for the extractions, from a single appendage to a whole animal depending on the size of the specimen. DNA was diluted either 1:10 or 1:100 before amplification by PCR. Each 25 µl PCR reaction contained 5 µL of 5× reaction buffer, 25 pmol of both the forward and reverse primer, dNTPs to a final concentration of 0.2 mM each, 1–3 µl of template DNA and 0.5 U Kapa 2G Robust Hotstart DNA polymerase (Sigma-Aldrich, St Louis, MO). The 18S marker was amplified and sequenced using primers 18S-ai and 18S-bi ( Whiting 2002); the COI marker was amplified and sequenced using the universal primers LCO1490 and HCO2198 ( Folmer et al. 1994). PCR products were purified using ExoSAP-IT (USB, Cleveland, Ohio, USA) according to the manufacturers’ instructions, and were sequenced at Macrogen Inc. (Seoul, South Korea).

Sequences were trimmed and aligned using Geneious V11.1.5 (http://www.geneious.com, Kearse et al. 2012), and were compared to sequences in GenBank using BLAST ( Altschul et al. 1990). COI and 18S sequences derived from the New Zealand specimens were aligned with sequences from GenBank that showed high homology to those of the New Zealand material. This included specimens identified as Caprella scauroides and Caprella californica , as well as representatives of other species of Caprella . Sequences used in the analysis are listed in Table 1 (including sequences from Best & Stachowicv 2013 and Cabezas et al. 2014).

Sequences were aligned using MUSCLE 3.8.425 ( Edgar 2004) implemented in Geneious and checked by eye. Unalignable regions were removed from the 18S alignment using GBLOCKS, implemented in SeaView V4.7 ( Gouy et al. 2010), at the default settings. We used Partitionfinder 2.1.1 ( Lanfear et al. 2016) to identify appropriate models of sequence evolution and partitioning strategies for the COI and concatenated analyses. For the COI analysis parti- tioning strategies and models selected were: PhyML (no partitioning) HKY+I+G; MrBayes (3 partitions, by codon) Codon1 SYM+I, Codon 2 HKY+I, Codon 3 GTR+G. For the concatenated analysis the following were selected: PhyML (no partitioning) SYM+I+G; MrBayes (4 partitions) Codon1 SYM+I, Codon 2 F81+I, Codon 3 GTR+G, 18S K80+I.

Caprella linearis ( Linnaeus, 1767) and Caprella simia Mayer, 1903 sequences were included as outgroup taxa. We used PhyML ( Guindon et al. 2010) to perform maximum likelihood analyses for both single-gene datasets and for the concatenated alignment of COI and 18S data. Branch support was assessed using the Shimodaira-Hasegawalike approximate likelihood ratio test (SH-like aLRT, Anisimova & Gascuel 2006) and 1000 bootstrap pseudoreplicates for each dataset. Bayesian trees were estimated using MrBayes V3.2.6 ( Ronquist et al. 2012), partitioning the datasets in accordance with the Partitionfinder output. For each dataset, two sets of four MCMC chains were run for 5 million generations, sampling every 1000 generations; burnin was assessed using Tracer 1.7-pre20171127 (http:// tree.bio.ed.ac.uk/software/tracer/) and confirmed by potential scale reduction factor (PSRF) values calculated in MrBayes. Trees were visualised using FigTree 1.4.0 (http://tree.bio.ed.uk/software/figtree).

The sequences are deposited in GenBank and accession numbers are listed in Table 1 . The reference specimens are stored in the NIWA Invertebrate collection.

Results

Morphological. Material examined. NIWA 136876 ( MITS 73603 ), 121 specimens (adults—very large males, ovigerous females), plus over 82 juveniles, Station BNZ (AAP) 14933-AM, The Landing, Ōkahu Bay, Waitematā Harbour , New Zealand, 36.8505271°S, 174.810083°E, 29/05/2017, collectors: Phil Johnstone and Jeff Tyrell.

NIWA 136877 ( MITS 73604 ), over 50 specimens (adults—very large males, ovigerous females + juveniles), Ōrakei Marina, Waitematā Harbour , New Zealand, 36.85014°S, 174.81415°E, mooring line, AKL24ExtraDive 1, 24/05/2017, collectors: NIWA.

NIWA 136878 ( MITS 73858 ), 484 specimens (202 mature male, 132 ovigerous females, 150 juveniles), WRE 26236—AM 7/11/2017, McLeod Bay, Whāngārei Harbour , New Zealand, 35.81502°S, 174.49953°E, 5 m depth from a mooring.

NIWA 136879 ( MITS 24429 ), 10 specimens, NIW193BTI, hull of the vessel Aoix, Nelson Marina , New Zealand, 41.258542°S, 173.281267°E, 12/01/2005, coll.: NIWA.

Remarks. The specimens examined are referable to Caprella scauroides sensu stricto as documented by Takeuchi & Oyamada (2013). The characters that support this include having a rounded antero-distal lobe on the mature male gnathopod 2; a straight head projection; pereonites 3 and 4 the same length as pereonite 5; robust pereopods 6 and 7, longer than pereopod 5; pereopod 7 merus shorter than propodus ( Figs 2–3 ). The only aspect in which our records differ from the account of Takeuchi & Oyamada (2013) is the robustness of pereonite 2 (height 1/4 instead of 1/3 the length) for a large male specimen ( Fig. 3 ). However, the male figured here is a hyper male (22 mm) ( Fig. 3A ), and the specimen noted in Takeuchi and Oyamada (2013) is only 18.36 mm; this may account for this slight difference. When males of a similar length to those documented were compared there was no difference. Caprella californica has different pereonite proportions, an angular gnathopod 2 anterodistal propodal process; a curved anterodistal head projection; and different proportions of the pereopods

DNA sequence data and phylogenetic analyses. The 18S sequences obtained from all three New Zealand specimens of Caprella scauroides were identical apart from two ambiguous nucleotides over 961 bp in common. The COI sequence from the Whāngārei Harbour specimen was identical to that of the male specimen from Ōkahu Bay. The female specimen from Ōkahu Bay differed from these two by a single substitution over 599 bp. The COI alignment consisted of 50 terminals and 537 nucleotides, while the concatenated COI-18S dataset consisted of 26 terminals and 1441 nucleotides. Figure 5 shows the results of the phylogenetic analysis of the COI data. New Zealand sequences of Caprella scauroides strongly group with five COI sequences from Western Australian specimens previously tentatively referred to C. californica by Cabezas et al. (2014). The Australian and New Zealand sequenc- es differ from one another by 0–4 substitutions (0–0.7%) and are resolved as a separate clade from C. californica sensu stricto from California, from the region of the type locality of the species. This indicates that the Western Australian sequences are referable to C. scauroides sensu stricto. The New Zealand sequences differed from the Californian sequences by 75–84 substitutions over 537 bp total alignment (14–16%). Four sequences from Taiwan which were previously referred to C. scauroides by Cabezas et al. (2014) are resolved as a separate clade, distinct from both the Californian clade ( C. californica sensu stricto) and the Australian/ New Zealand clade ( C. scauroides sensu stricto), and differ from the New Zealand sequences by 71–78 bp (13–14.5%). Phylogenetic analyses of the 18S and COI datasets were consistent, although the less informative 18S dataset was unable to resolve structure within the Caprella scauroides / californica clade, in part because the removal of unalignable regions in the 18S dataset results in removal of the substitutions that differentiate those sequences, which mostly occur within the vari- able V4 region. Before removal of unalignable regions with GBlocks, the New Zealand sequences were identical to one another and to the two 18S sequences from Western Australia, and differed from C. californica sequences from California and Chile by 3 substitutions over the 918 bp alignment. The concatenated COI-18S analysis (Appendix 1), while containing fewer sequences than the COI dataset, strongly supported the same clade structure as the COI dataset, and is presented.

Discussion

Prior to recent taxonomic and phylogenetic review ( Takeuchi & Oyamada 2013; Cabezas et al. 2014), Caprella scauroides was considered a synonym of C. californica . However, recent, detailed appraisals allowing separation of C. californica and some of its synonyms enable clear recognition of C. scauroides as valid. The morphology of the Ōkahu Bay and Nelson Marina specimens are consistent with the taxonomic characteristics described in Takeuchi and Oyamada (2013) of C. scauroides , especially noting the characters mentioned regarding the differences between C. californica sensu stricto and C. scauroides .

Our molecular and morphological analyses indicate the recognition of C. scauroides from New Zealand, which is distinct from C. californica sensu stricto. There were significant morphological similarities to the record- ed Western Australian material and genetically these specimens were very closely related. Takeuchi & Oyamada (2013) noted that the populations identified as C. californica from the Australian region should be recognised as C. scauroides based on morphological criteria. Cabezas et al. (2014) found that COI sequences from the Australian collections were closest to specimens from Taiwan, which they referred to C. scauroides , while 18S sequences were most similar to C. californica sensu stricto and C. scaura f. spinirostris from Chile (not specifically mentioned in Cabezas et al. 2014).

The molecular differences between New Zealand C. scauroides and specimens from Taiwan identified as C. scauroides in Cabezas et al. (2014) suggest they are probably different species. Unfortunately, all the specimens from the Taiwanese collections were destructively sampled for the genetic analysis, and none is available for physi- cal examination of morphological characters. The photographic record of the material, whilst of quite good quality, is not detailed enough to enable confident identification. Takeuchi (pers. com.) noted that there are probably five or six separate but morphologically similar species from the area, and so there is a need for utmost care when identifying to species level. It has not been possible to obtain DNA sequence data from the material morphologically documented in Takeuchi and Oyamada (2013) and attempts to collect fresh material from the descriptive locality (Uwa Sea, Japan) have not been successful.

The samples collected from Ōkahu Bay contained large, sexually mature females (holding many eggs), males and many juveniles, indicating that local conditions were suitable for reproduction and population growth. Mature specimens of C. scauroides were also incidentally observed on biofouling (the non-indigenous spaghetti bryozoan, Amathia verticillata (delle Chiaje, 1822)) on several recreational yachts berthed at Westhaven Marina, Waitematā Harbour , indicating the potential for this non-indigenous species to be more widespread than just Ōkahu Bay in this harbour.

As caprellids can readily colonize artificial structures such as vessels, ropes, fish cages, netting, buoys and pontoons (e.g., Buschbaum & Gutow 2005; Greene & Grizzle 2007; Ros & Guerra-García 2012), movement of any ‘infested’ immersed equipment/structures outside of detected localities provides a potential pathway for transport of caprellids. Previous records for C. scauroides have noted it as being found extensively at pearl and fish aquaculture facilities ( Takeuchi & Oyamada 2013 and on vessel hulls ( Montelli & Lewis 2008). Most caprellid species are noted as tolerant of broad environmental conditions (Mittman & Müller 1998; Takeuchi et al. 2003). Caprella scauroides has been observed amongst biofouling on naval vessels in Australia (where it is considered non-indigenous) ( Montelli & Lewis 2008). The Ōkahu Bay samples (both submissions) were collected from a recreational vessel swing mooring line and from the hull of a recreational vessel being cleaned, and the Nelson samples also from a recreational vessel hull. As caprellids are early biofouling colonisers on vessels and can move around to find low-drag niches during vessel movement, caprellids can be dispersed widely and easily by vessels, including on relatively clean vessel surfaces. Thus, one of the most likely vectors for the introduction of C. scauroides to New Zealand is vessel biofouling (importation of used commercial aquaculture equipment is not permitted under any current Im- port Health Standards in New Zealand). Natural dispersal of C. scauroides following establishment is possible via rafting on floating substrata or passive transport in currents for dislodged caprellids, or via gradual benthic ‘creep’ across natural substratum.

It is not currently known what effect, if any, C. scauroides might have on native New Zealand ecosystems. There are no documented impacts of this species in the scientific literature, although it may act as a potential controller of seaweed communities (to the benefit of the algae by consuming algal epiphytes) (e.g. Caprella penantis Leach, 1814 in Duffy 1990 ). Caprellids are typically detritivores, although some species (mainly those that are widely distributed) are opportunistic or even predatory ( Guerra-García & de Figueroa 2009; Alarcón-Ortega et al. 2012).

There is potential for displacement of ecologically similar species by C. scauroides . Shucksmith et al. (2009) demonstrated that the non-indigenous C. mutica , could displace native C. linearis ( Linnaeus, 1767) and Pseudoprotella phasma ( Montagu, 1804) when resource space was limited. As C. scauroides is morphologically very similar to C. equilibra Say, 1818 and appears to inhabit similar environments and substratum, it may compete with C. equilibra . Caprella scauroides may have a stronger competitive effect on the much smaller Caprellina longicollis Nicolet, 1849 which also inhabits similar environments and substrata, but typically occurs at lower densities (CW, pers. obs.). However, C. scauroides is not expected to affect the less common native caprellid species, which are restricted to deeper water or to New Zealand’s sub-Antarctic islands.

As a first record for a region it is important to establish the status of species such as these, which have the potential to influence fisheries and ecosystems. It is also important to have the most accurate identification as possible, which has been achieved to the extent possible by both morphological and molecular means. As this study highlights, cryptic morphology can lead to confusion in distribution and influence. This study does not rule out the possibility of further undetected species of this group also occurring in New Zealand waters. Ideally, sequences from the type locality and pathways would secure the placement of the course of these species. However, as recorded above, this extra material has not yet been able to be obtained.

Figures

FIGURE 1. Map of New Zealand with incidence localities for Caprella scauroides plotted.

FIGURE 2. Caprella scauroides Mayer, 1903: A, male 21 mm length. (Photograph: R. A. Peart, NIWA). Scale bar represents 1 mm

FIGURE 3. Caprella scauroides Mayer, 1903: A, male 22 mm length (NIWA 136876). B, female 11 mm body length (NIWA 136876).

FIGURE 5. Maximum-likelihood phylogram estimated from the COI dataset. Support values are shown on each branch: approximate Likelihood Ratio Test (aLRT) values and ML bootstrap (%) above, and Bayesian PP values below. Support values for some very short internal branches are omitted for clarity. Only values greater than 0.8 (aLRT), 70% (bootstrap) and 0.9 (PP) are shown. All support values are shown if two support methods reach the cut-off value. Voucher numbers or isolate designations (for GenBank sequences) or collection locations (NZ collections) are appended to names when two or more representatives of a single taxon are included. Scale bar represents substitutions per site. Vertical lines indicate regions of collections for the Caprella scauroides/californica lineages.