Australatya hawkei, Choy & Page & Mos, 2019

Choy, Satish, Page, Timothy J. & Mos, Benjamin, 2019, Taxonomic revision of the Australian species of Australatya Chace 1983 (Crustacea, Decapoda, Atyidae), and the description of a new species, Zootaxa 4711 (2), pp. 366-378 : 368-375

Australatya hawkei sp. nov.

( Figs. 1 , 2 , 3 , 4 , 5 , Table 1 )

Atyoida striolata Smith & Williams 1982 (in part), all specimens from Paluma (Mt Spec), Smoko Ck, and 25 km N. of Cooktown as listed on p. 345 of Smith & Williams (1982), and those designated as Far Northern Queensland in Table 3 of Smith & Williams (1982).

Australatya striolata Smith 1994 (in part), all specimens from Site Numbers 7 to 15 as listed in Table 1 of Smith (1994) and corresponding sites in Table 2 of Smith (1994).

Holotype. QM W19563 ovigerous female, 14.0 mm CL, 3.0 mm RL, 9.3 mm CD, 4.5 mm A1P, 55.1 mm SL, 61.5 mm TL, Adeline Ck , tributary of Daintree River, NE Queensland, Australia, 16°13.1’S, 145°5.1’E, ca. 1000 m altitude, 16.03.1993, B. Herbert, P. Graham, J. Peeters.

Allotype. QM W19563 male, same data as holotype.

Paratypes. ‘ Southern’ morphotype: QM W195637 ovigerous females, medium to large sized, same data as holotype; QM W19263 Mt Windsor Tableland , tributary of Daintree River, NE Queensland, Australia , 16°13.9’S, 145°3.8’E, 13.11.1993, J. Short, P. Davie, A. Humphreys, 4 large specimens. stn. m, freshwater, lotic, coarse sand, moderate flow, leaf litter, fallen timber, water clarity high, fringing rainforest; QM W22018 Annan River nr Shipston Flat, FN Queensland, Australia , 15°45.6’S, 145°13.3’E, ca. 150 m elevation, 13.11.1993, J. Short, P. Davie, A. Humphreys, 2 medium sized specimens, electrofished in freshwater, riverine upper reach, riffle, rocks, moderate flow, silt, sand, gravel, rocks, water clarity fair, some macrophytes, fallen timber, leaf litter, fringing monsoonal forest, water temp. 20.7 °C, pH 6.0, DO 8.6 ppm, photographed by J. Short; QM W15002 Bells Ck , east of Sarina, Queensland, Australia , 21°26’S, 149°13’E, 19.12.1978, R.J. McKay , J. Johnson , freshwater. All deposited in the Queensland Museum, Brisbane, Australia .

‘Northern’ morphotype: QM W13216–W13218, W13220 Claudie River, Iron Range, about 20 km NW of Lockhart River, Queensland, Australia, 12 o 45’S, 143 o 12’E, “way up”, 24– 25.08.1986, Ross Smith , 30 small to medium sized specimens; QM W13219 Leo Creek, McIlwraith Range, Queensland, Australia , 13 o 45’S, 143 o 23’E, 500 m altitude, 28.08.1985, Ross Smith , 31 small to medium sized specimens; QM W13221 Rocky River, McIlwraith Range, Queensland, Australia , 13 o 48’S, 143 o 27’E, 50–120 m altitude, 27– 28.08.1985, Ross Smith , 12 small specimens. All specimens deposited in the Museum of Tropical Queensland, Townsville, Australia .

Description. Rostrum short ( Fig. 4A ), without dorsal or lateral teeth, holotype 0.21 CL (0.19–0.31 in paratypes), reaching to or just beyond first segment of antennular peduncle but never reaching end of second segment of antennular peduncle; lateral carina well developed on posterior quarter, continuing to tip which can be straight or bent downwards, armed with 7 ventral teeth (5–8 in paratypes). Ocular cornea well developed, 0.14 CL, well pigmented. CL/CD 1.5 (1.39–1.8 in paratypes), antennal spine well developed. Basal segment of antennular pe- duncle with sharply pointed stylocerite, stylocerite not reaching to tip of basal segment of antennular peduncle. Scaphocerite about twice length of antennal peduncle. Third maxilliped relatively stout, ultimate segment beyond antennal peduncle; ultimate segment 0.95 times as long as penultimate (0.93–1.06 in paratypes), 0.86 times as long as antepenultimate (0.84–0.90 in paratypes), clear sexual dimorphism with a terminal claw-like spine at the tip in males and none in females; exopod to about half of penultimate segment (ischiomerus).

First pereiopod ( Fig. 4B ) reaching to tip of antennular peduncle, shorter and smaller than second pereiopod but generally of similar shape; dactylus and propodus of equal length and width, distal setae dense and longer than propodus; propodus length 5.31 times width (4.33–5.65 in paratypes); carpus length 0.43 of propodus length (0.43–0.55 in paratypes); carpus length 1.06 times width (1.06–1.56 in paratypes), relative excavation depth of carpus 0.81 (0.70–0.95 in paratypes); carpus length 0.50 of propodus length (0.43–55 in paratypes); merus length 2.09 times width (2.01–2.15 in paratypes), 0.89 times propodus length (0.77–0.98 in paratypes); ischium length 3.14 times width (3.10–3.25 in paratypes), 0.73 times propdus length (0.68–0.79 in paratypes).

Second pereiopod ( Fig. 4C ) reaching to beyond antennular peduncle and first pereiopod; dactylus and propodus of equal length and width, distal setae dense and longer than propodus; propodus length 5.14 times width (4.35–5.51 in paratypes); carpus length 0.40 of propodus length (0.40–0.52 in paratypes); carpus length 1.03 times width (1.00– 1.50 in paratypes), relative excavation depth of carpus 0.70 (0.66–0.80 in paratypes); carpus length 0.40 of propodus length (0.40–50 in paratypes); merus length 2.36 times width (2.30–2.40 in paratypes), 0.85 times propodus length (0.84–1.10 in paratypes); ischium length 3.78 times width (3.70–3.85 in paratypes), 1.0 times propdus length (0.95–1.10 in paratypes).

Third pereiopod ( Fig. 4D ) reaching tip of antennular peduncle; dactylus short, ending in robust claw and 6 stout posterior medial spines, dactylus length 0.28 times propodus length (0.25–0.32 in paratypes); propodus length 5.50 times longer than wide (5.05–6.70 in paratypes); carpus length 3.95 times width, carpus length 0.84 times propodus length (0.80–0.90 in paratypes); merus length 5.61 times width, merus length 1.78 times propodus length (1.60–1.85 in paratypes), ischium length 0.56 times propodus length (0.53–0.62 in paratypes).

Fourth pereiopod reaching by one third dactylus beyond scaphocerite to just reaching tip of dactylus at distal end of scaphocerite; generally similar to third pereiopod but slightly longer and more slender; dactylus short, with robust terminal claw and 5 stout posterior medial spines, dactylus length 0.22 times propodus length (0.22–0.30 in paratypes); propodus length 5.90 times width (5.80–8.02 in paratypes); carpus length 4.33 times width (3.50–4.50 in paratypes), carpus length 0.67 propodus length (0.62–0.69 in paratypes), merus length 5.89 times width (4.12–5.95 in paratype), merus length 1.32 times propodus length (1.28–1.42 in paratypes), ischium length 2.13 times width (2.11–2.25 in paratypes), length 0.44 times propodus length (0.41–0.49 in paratypes).

Fifth pereiopod ( Fig. 4E ) is the longest and most slender of all the pereiopods, generally similar to fourth pereiopod with dactylus having a short robust terminal claw but behind it are a row of about 50 comb-like spines on the medial margin; dactylus length 0.19 times propodus length (0.16–0.27 in paratypes); propodus length 11.00 times width (7.5–11.5 in paratypes); carpus length 4.50 times width (3.5–5.00 in paratypes), carpus length 0.56 times propodus length (0.48–0.58 in paratypes); merus length 5.55 times width (4.6–6.10 in paratypes); merus length 0.78 times propodus length (0.74–0.81 in paratypes); ischium length 0.36 of propodus length (0.33–0.41 in paratypes).

First female pleopodal endopod is of similar length to exopod, narrowly ovate and tapers distally; setae on lateral borders. First male pleopodal endopod about 0.67 times exopodal length, distal third with a narrow appendix interna. Second female pleopodal endopod slightly shorter than exopod, both broadly ovate; appendix interna short, narrow, and with dense cluster of retinacula at distal end. Second male pleopodal endopod slightly shorter than exopod; appendix masculina prominent and a shorter appendix interna.

Molecular results. GenBank accession numbers for the new sequences are MN244315 – MN244320 . In the 16S analyses, A. striolata and A. hawkei sp. nov. formed a clade relative to the other taxa, which was strongly supported in the parsimony (PA) analysis (91%; tree score 260 steps; not displayed) but only moderately supported in the maximum likelihood (ML) analysis (65%; tree score—1830.20; Fig. 1A ). The genetic distance (K2P) between A. striolata and A. hawkei sp. nov. for 16S was 1.7%. The genus Australatya was strongly supported (ML 97%, PA 91%), with Atyoida spp. as its sister (ML 81%, PA 88%), with a net K2P genetic distance between these genera of 5.8%.

In the COI analyses, A. hawkei sp. nov. and A. striolata formed geographically structured clades relative to each other (ML: tree score—1321.03, Fig. 1B ; PA tree score: 120 steps, not displayed). All specimens of A. hawkei sp. nov. from north–eastern Australia (<22°S) were recovered as a strongly supported clade (ML 96%, PA 99%), and all specimens of A. striolata from south–eastern Australia (> 26°S) in a less strongly supported clade (ML 79%, PA 96%). The net COI K2P genetic distance between the species was 6.1%, and 18.6% between them both and A. obscura . The 6.1% COI difference is at the lower end of the range of average divergences between species within decapod genera ( Costa et al. 2007), but it is well within the range. In Costa et al. (2007), the lowest divergence was 4.92%. However, we found there was no clear geographic pattern when each species was considered in isolation. Within A. hawkei sp. nov., the most common haplotype was found at all seven sites which stretch about 1100 km along the north–eastern Australian coast ( Fig. 1B ).

Specimens from Claudie River and Leo Creek (CL and LE in Fig. 1B ) that morphologically equate to the ‘northern’ morphotype of A. hawkei sp. nov. were not differentiated at this mitochondrial locus from its ‘southern’ morphotype, and shared haplotypes with it. Similarly, within A. striolata , there was no clear geographically structured pattern, with haplotypes from all sites spread throughout the tree, and the two most common haplotypes were found at nine of the ten sites ( Fig. 1B ).

Etymology. The species is named in honour of the late Robert (Bob) James Lee Hawke, Prime Minister of Australia from 1983–1991. His government led many socially important initiatives such as Medicare and Superannuation, and it also established continuing funding for the Landcare movement that brought communities together to work to tackle within-catchment land degradation and water quality problems. The common name of this species will be “Hawke’s shrimp.

Remarks. Australatya hawkei sp. nov. is genetically and morphologically different from A. striolata (Mc- Culloch & McNeill 1923). Specifically with regards to morphology, the carapace length of A. hawkei sp. nov. is generally shorter in relation to its depth and the ratio of rostral length to carapace length is intermediate in A. striolata when compared to the two morphotypes of A. hawkei sp. nov.. The ratio of rostral length to scaphocerite length is smaller in A. hawkei sp. nov. and the excavation of the carpus on the first and second pereiopds is much deeper in A. hawkei sp. nov.. The propodus to dactylus length on the third, fourth and fifth pereiopds and the scaphognathite length to carapace length in A. hawkei sp. nov. are greater than in A. striolata ( Table 1 ). It should be noted, however, the two species have some very similar morphological features (e.g. mouthparts, uropods, telson, and pre-anal carina) and there is overlap in other characters ( Table 1 ), and this should be taken into consideration when identifying specimens. Given that individual morphometric and meristic characters can be highly variable, identification of closely related species should be based on multiple characters and, where possible, verified with genetic data ( Page et al. 2005; de Mazancourt et al. 2019; Choy et al. 2019).

There is also a clear allopatry between the two species, with A. hawkei sp. nov. occurring north of Mackay and A. striolata only known to occur much further south, to at least Genoa River in eastern Victoria ( Fig. 3 ). Specimens examined by Smith & Williams (1982, p. 345, Table 3 and Fig. 5 ) comprised both species but can be clearly separated based on locality. All their specimens belong to A. striolata except for those from Paluma, Mt Spec, Smoko Creek, and 25 km north of Cooktown which, being from far northern Queensland, belong to Australatya hawkei sp. nov. ( Table 1 , Fig. 3 ). This locality-based separation is also supported by the corresponding values for proportional meristic characters provided in Table 3 of Smith & Williams 1982. Similarly, the specimens examined and analysed by Smith (1994, Tables 1 –2 and Fig. 1 ) can be separated on the basis of locality ( Table 1 ).

Within the genetically homogeneous Australatya hawkei sp. nov., there are two distinct morphotypes that are allopatric; a ‘southern’ one (currently known to occur from near Mackay to about Cooktown) and a ‘northern’ one (occurring in the McIlwraith and Iron Ranges, north of Coen) ( Fig. 3 ). The data for the characters used to examine locality-related morphological variation in Australatya , provided in Table 2 of Smith (1994), clearly demonstrate this. As shown in Table 1 , the ‘northern’ morphotype has a relatively long carapace and rostrum, short propodus length to width ratio of the first pereiopod, longer scaphognathite and shorter telsonic length to width ratio. When these data were further analysed using ANOVA, principal component, and hierarchical clustering (Systat Software, San Jose, CA), the results show much clearer separation ( Fig. 2 ). The cluster analysis suggests that the ‘northern’ A. hawkei sp. nov. (e.g. from Leo Creek, McIlwraith Range) are morphologically very distinct from the ‘southern’ A. hawkei sp. nov. (e.g. from Yuccabine Creek, Herbert River) for these characters. It is also interesting that the ‘southern’ morphotype of A. hawkei sp. nov. is morphologically more similar to A. striolata for these characters than to its ‘northern’ morphotype ( Fig. 2 ). This paradox can be explained by contemporary environmentally driven factors such as stream gradient and current speed, leading to clinal variation in morphology ( Smith 1994), which do not reflect deep phylogenetic relationships. Hence, the morphological characters provided in Table 1 and Fig. 2 can be used to separate not only A. hawkei sp. nov. and A. striolata , but also the two geographically separated A. hawkei sp. nov. morphotypes. It could be argued that if the species were based on morphology alone, the ‘northern’ A. hawkei sp. nov. could be designated as a species or sub-species. However, given that both are genetically identical (at least, based on the molecular markers used here), it is more appropriate to treat them identically and, if necessary, refer to them as geographically distinct.

Ecology and biogeography. All species of Australatya are found in riparian forest covered, moderate to fast flowing streams that have high water clarity ( Smith 1994; Han & Klotz 2015, Fig. 5 A–D); hence the common name, “riffle” or “filter-feeding” shrimps. A. striolata and A. hawkei sp. nov. are generally found under large stones, boul- ders, and plant litter in riffles or cascades, although some may be found amongst tree roots, hanging from aquatic vegetation, and under plant litter in pools ( Fig. 5 A–D). Both sexes are often active at night, with large females (> 4 cm TL) frequently sighted in the open. Being primarily filter feeders, they have a clear preference for flowing waters but may be found in stagnant conditions where water levels are low.

Along the eastern seaboard of Australia, the mountains (the Great Dividing Range) get closer to the coastline as one travels northwards and so northern streams are generally shorter and steeper, reaching a much higher altitude closer to the coast ( Fig. 3 ). Australatya striolata is generally found in the lower gradient streams of southern Queensland, New South Wales, and Victoria, with juveniles and adults both occurring at lower altitudes ( Smith 1994). The maximum recorded altitude of occurrence is less than 200 m ( Smith 1994). In contrast, A. hawkei sp. nov. is found in the higher gradient streams of northern Queensland. In general, only their juveniles are found at lower altitudes while adults tend to occur mainly at higher altitudes (above 100 m). The maximum recorded altitude of occurrence is about 1000 m for both of the morphotypes of A. hawkei sp. nov..

Both Australian Australatya are relatively larger than the insular A. obscura , with females of the former growing to about 13 mm CL and males of the former to about 10.4 mm CL ( Han & Klotz 2015). Samples suggest that A. striolata is slightly smaller and less robust than A. hawkei sp. nov. (females about 13 mm CL vs. 16 mm CL). Both species are protandrous ( Smith & Williams 1982), changing from male to female at about 5 mm CL and 8 mm CL, respectively. Live colouration is very similar among all Australian Australatya , typically solid to translucent black or grey with a prominent white or yellow dorsal stripe ( Fig. 5E,F ), but can vary to shades of red, blue, brown, and green with stress, mood, and/or time of day (e.g. Fig. 5G,H ).

Both A. striolata and A. hawkei sp. nov. appear sporadically in the aquarium trade, being robust survivors in aquaria, and are occasionally bred in captivity. In the wild, the reproductive cycle of A. hawkei sp. nov. is likely similar to A. striolata where embryos carried by females are triggered to hatch during flood events, pelagic larvae then develop in estuaries or the ocean where they feed on phytoplankton and/or zooplankton, before returning to freshwater as juveniles (i.e. amphidromy) ( Smith 1994). It is interesting to note that while the number of eggs carried by female A. hawkei sp. nov. are quite numerous (1000–2600) and the size of eggs is small (0.30–0.35 × 0.50–0.55 mm), the lifecycle can be completed in freshwater if sufficient food is made available, albeit in low numbers in specialist aquaria, suggesting the possibility of facultative amphidromy. It is most likely reports in aquarium literature of direct development as “miniature adults” stem from cases where juvenile shrimp appear in aquaria where there was sufficient microorganisms or suspended organic material to provide nutrition to early pelagic life stages that are small (<5 mm TL), somewhat transparent, and easily overlooked.

Species with amphidromous life histories tend to have widespread distributions ( McDowall 2007, Bauer 2013, Page et al. 2013). A. striolata and the ‘southern’ morphotype of A. hawkei sp. nov. have a similar but disjunct distribution along Australia’s east coast, while the ‘northern’ morphotype of A. hawkei sp. nov. has a much smaller, but also disjunct, distribution ( Fig. 3 ). A. obscura , which also has numerous, small eggs (about 1000, 0.33–0.34 × 0.55–0.56 mm), has been reported to have a prolonged larval development and the possibility of a wider distribution range in South-East Asia ( Han & Klotz 2015). Indeed, A. obscura was recently recorded from the Ryukyu Islands, Japan, and it was suggested that the species was transported northwards by the Kuroshio Current ( Inui et al. 2019).

The major ocean current on the eastern coast of Australia, the East Australian Current (EAC) splits in two around Mackay, continuing south as the EAC, and north as the South Equatorial Current (SEC) ( Fig. 3 ). If larval A. hawkei sp. nov. rarely cross the north-south split in the EAC, this might explain why A. hawkei sp. nov. is restricted to northern latitudes. This same area is also a biogeographic boundary for a number of marine taxa ( Cook et al. 2012). Similarly, the shape of the coastline near Coen and the dominance of the northerly flowing ocean currents in this area ( Fig. 3 ) might explain the persistence of the ‘southern’ and ‘northern’ morphotypes of A. hawkei sp. nov.. A lack of adult habitat might also be acting as a barrier between populations.

The gap between the distributions of the ‘southern’ and ‘northern’ morphotypes of A. hawkei sp. nov. (from Cooktown to Coen) and between the distributions of A. hawkei sp. nov. and A. striolata (near Mackay) coincide with areas where there are few or no mountains and high gradient streams near the coast ( Fig. 3 ). It is possible, however, that the distribution of A. hawkei sp. nov. is larger than reported in this study. The prevailing ocean currents along Australia’s north–eastern coast flow north towards Papua New Guinea and the Solomon Islands ( Fig. 3 ). Other amphidromous atyids from the eastern coast of far northern Queensland, Australia are found in the Solomon Islands (e.g. Caridina sp. 3 Solomon, de Mazancourt et al. 2019).

Figures

FIGURE 1. (A) Maximum likelihood topology of 16S rDNA dataset, featuring Australatya spp. and related genera, showing GenBank accession numbers and geographic sampling area. (B) Maximum likelihood topology of COI dataset, including only Australian Australatya. Individuals of the ‘northern’ morphotype of A. hawkei sp. nov. are bold and underlined. Nodes with probability values <50% have been collapsed. Site codes from north to south: A. hawkei sp. nov.: CL = Claudie R, LE = Leo Ck, MT = Mt Windsor, HU = Hutchinson Ck, DA = Damper Ck, OL = Ollera Ck, CE = Cedar Ck; A. striolata: MA = Maroochy R, MO = Mooloolah R, LO = Logan R, CO = Coomera R, TW = Tweed R, RI = Richmond R, BE = Bellinger R, WI = Wilson R, MN = Manning R, SH = Shoalhaven R.

FIGURE 2. Cluster tree showing Australatya from different locales (north to south) clustered according to morphometric parameters (rostrum length/carapace length, rostrum length/depth, telson length/width, antennular peduncle length/antennule segment 1 length, peraeopod 1 propodus length/width, peraeopod 3 propodus length/width, peraeopod 5 propodus length/width, peraeopod 5 propodus length/dactylus length, peraeopod 5 propodus merus length/width; data from Table 2 of Smith (1994)).

FIGURE 3. Map showing the distribution of Australian Australatya, mountain ranges, and the primary direction of ocean currents along the east coast of Australia.

FIGURE 4. Australatya hawkei sp. nov. southern’ morphotype (A) cephalothorax; (B) 1st pereiopod; (C) 2nd pereiopod; (D) 3rd pereiopod; (E) 5th pereiopod.

FIGURE 5. Habitats and live colouration of Australian Australatya. Riffle habitats of Australatya striolata: (A) Never Nev- er River, Bellinger River catchment (30°21’39.4”S, 152°54’15.4”E). (B) Pine Brush Creek, flows direct to Pacific Ocean (30°15’05.4”S, 153°07’56.9”E). Riffle habitats of Australatya hawkei sp. nov.: (C) Gordon Creek, Chinaman Creek catchment (16°57’57.7”S, 145°43’50.0”E). (D) Little Mulgrave River, Mulgrave River catchment (17°08’00.7”S, 145°42’42.5”E). Live colouration: (E) A. striolata, typical colouration in aquaria. (F) A. striolata, female, live colouration on capture, Never Never River. (G) A. striolata, female, colouration at night, Pine Brush Creek. (H) Australatya hawkei sp. nov., ‘stress’ colouration after capture. Note: typical colouration is very similar to A. striolata in panels E and F. Photos: A–G by B. Mos. H by Bob Kroll.