Parvulastra parvivipara (Keough & Dartnall, 1978) and P. vivipara (Dartnall, 1969) are Australian small range endemic seastars occurring on coasts of western Eyre Peninsula (South Australia) and southeast Tasmania, respectively. Both species have atypical life-histories involving a diminutive size, simultaneous hermaphroditism, almost exclusive self-fertilisation, and brooding offspring in the gonads to an advanced juvenile which is achieved by sibling cannibalism (Byrne 1996). Birth occurs through the gonopore of the parent (Fig. 1a). Adult P. parvivipara are among the world’s smallest seastars reaching a diameter of only 11 mm while P. vivipara are slightly larger. Both species have extreme poverty of genetic diversity (Keever et al. 2013).

a

b

c

d

Figure 1.

Photographs from P. parvivipara and P. vivipara habitat, and graph of P. vivipara population trends.

a: Photograph taken during surveys of Liversage (2015) showing P. parvivipara giving birth, with the bright orange juvenile emerging from parent's dorsal side (bar = 1 cm).  
b: Photograph of a boulder underside in P. parvivipara habitat with extensive encrustation of oyster shells that includes invasive Pacific oysters (bar = 5 cm).  
c: Photograph of P. vivipara during 1992 at Pit Water. Populations have become reduced in subsequent years which may be associated with increased siltation and overgrowth from encrusting species (bar = 5 cm).  
d: Trends from the largest P. vivipara population at Pitt Water. Each point is the mean of adult densities from numerous sampling events over each 2 year period. During 1976-83 a fixed 1 m² quadrat was sampled (Prestedge 1998) while a different method was used from 2001-04 involving 25 m transects being sampled across the site (Ecomarine 2014). Other survey types (timed-search) have also found large population declines (see section 2.).  

Both species use habitat underneath intertidal boulders, although P. vivipara is also associated with mussel beds. During night they emerge from under boulders (Prestedge 1998, Roediger and Bolton 2008) to graze on microalgae similar to the much studied P. exigua (Martinez et al. 2016).

The overall distribution of P. parvivipara is limited to seven granitic headlands with the largest population occurring over 3924 m² at the site “Smooth Pool” (Roediger and Bolton 2008). The other populations occur over smaller areas (approx. 1500 m² per site; Roediger and Bolton 2008) raising the question of their vulnerability to extinction, especially since large within-site population fluctuations occur (Roediger and Bolton 2008). For example, the researchers who discovered this species (Keough and Dartnall 1977) stated that a single specimen was found at D’Anville Bay (200 km south of the other populations) but surveys by Roediger and Bolton (2008) failed to find it there recently. Similarly, a small population at a site south of Ceduna (Whittleby; Roediger and Bolton 2008) was not observed by Liversage (2015) despite extensive searching.

P. vivipara is known from six sites in southeast Tasmania (Prestedge 2001), two of which it was artificially introduced to. Population decline is clear from surveys of the most populous site, Pitt Water, from 1976-1983 (Prestedge 1998) and in 2001-2004 (Ecomarine 2014) (Fig. 1d). The site was previously ideal habitat for the species (Fig. 1c) but there has since been widespread habitat deterioration e.g. extensive siltation, encrusting species that cement boulders together, and black anoxic conditions under the boulders. Attempts to find P. vivipara in 30 min timed-search surveys found 68 and 47 individuals in 2016 and 2018, respectively, down from 287 in 2009 (M. Byrne, personal observation).

Approximately 25 % of sites where P. parvivipara occur have been colonised by the non-native Pacific oyster Magallana gigas (Thunberg, 1793). M. gigas co-occurs with the seastar and other native oysters underneath boulders. One aim of this study was to use photographs of boulder undersurfaces in P. parvivipara habitat, taken by Liversage (2015), to analyse potential correlations (linear-regression) between P. parvivipara and oysters. Shells of native oysters could not be differentiated from those of juvenile M. gigas, so shells (alive and dead) of all oyster species were analysed together. The photographs were imported into SketchUp v5 (www.sketchup.com) and the programme was used to calculate % oyster cover.

The result showed that oyster shells reached covers up to 36 % (Fig. 1b), although the mean is 4.65 and 1.12 % at two invaded sites (Cape Vivonne (n = 30 boulders) and Point Brown (n = 22), respectively). Any potential effects of oysters are uncertain at this stage; oyster encrustations clearly change habitat structure but the linear regression between oyster % cover (arcsine transformed) and P. parvivipara density (boulders pooled from abovementioned sites) showed no correlation (R² = 0.002, F = 0.12, P > 0.05). This result suggests that any effect of M. gigas on P. parvivipara may only occur if the invader reaches higher densities than are currently present. M. gigas also co-occurs with P. vivipara in Tasmania; further research in both regions will be required to understand any potential impacts on these rare seastars.

We thank Lana Roediger for discussion on P. parvivipara biology.

This project complies with the Estonian Code of Conduct for Research Integrity (http://www.eetika.ee/en/ethics-estonia/estonian-code-conduct-research-integrity).

The authors declare no conflicts of interest.