Biology and conservation of the unique and diverse halophilic macroinvertebrates of Australian salt lakes

This study synthesises information on the biology of the unique and diverse halophilic macroinvertebrates of Australian salt lakes, focusing on gastropods and crustaceans. This information is needed to evaluate and manage the threats posed to these invertebrates by increased periods of drought and secondary salinisation. Most of these species are endemic to Australian salt lakes, and some have adapted to extreme conditions (e.g. salinities >100gL–1 and pH <5). This study identifies key general findings regarding the taxonomy, ecology and life histories of these invertebrates, such as that many ‘new’ species have been uncovered in the past 20 years, with more likely to come. The study also identifies critical knowledge gaps, such as the need to elucidate the abiotic and biological drivers of the field distributions of species, including why some species are widespread and common whereas other congeneric species are rare or have narrow distributions. Those species that are either restricted to low salinity environments or survive dry periods as aestivating adults (as opposed to desiccation-resistant eggs) are probably the most vulnerable to increasing salinisation and drought. Future work should prioritise the development of a sound taxonomy for all groups, because this is needed to underpin all other biological research.


Introduction
Salt lakes are defined as enclosed bodies of water with salinity .3 g L À1 , although the salinity is usually much greater than this (Williams 1964). Salt lakes are often classed as either athalassic (inland) or coastal (possessing a current or recent connection with the marine environment, including via groundwater; Bayly and Williams 1966).
Australia has a vast number and variety of salt lakes that support unique faunal communities (De Deckker 1983a;Williams 2002). At least in Australia, coastal and athalassic salt lakes are characterised by different faunas. The former mainly contain species that have current or recent marine or estuarine affinities (Timms 2009a(Timms , 2010aPinder and Quinlan 2015), whereas the latter mainly contain taxa that have evolved in and are restricted to these systems (Bayly 1972). However, some athalassic species occasionally also occur in coastal lakes (Timms 2009a;Pinder and Quinlan 2015). This review focuses on the fauna of Australian athalassic salt lakes but includes data for athalassic species in coastal lakes.
Salt lakes are physically extreme environments, particularly those lakes that are either highly saline or only hold water intermittently (Williams 1985). The physical characteristics of Australian salt lakes are well established and were reviewed recently by Mernagh et al. (2016). Athalassic salt lakes mostly form in locations with semi-arid to arid climates where evaporation exceeds precipitation and drainage is impeded or fully endorheic (Williams 1998a). These lakes vary from expansive playas with long geological records to small, recently formed ponds and wetlands (De Deckker 1983b) that have become saline as a result of the intrusion of saline groundwater. Most Australian salt lakes are alkaline and ionically dominated by NaCl (Bayly and Williams 1966), but some are naturally acidic and can have a pH as low as 3 (Timms 2009b). Some salt lakes hold water permanently but most are either seasonal (usually filling with winter-spring rainfall and drying out over summerautumn) or episodic (filling only after unseasonal rainfall; Williams 1998aWilliams , 1998b.
The terms 'halotolerant', 'halophile' and 'halobiont' are commonly used to classify the aquatic fauna of salt lakes based on their relationship with salinity, although these terms are variously used (Bayly 1972;Williams 1981;Timms 1983;Hammer 1986;Pinder et al. 2002). Herein, 'halotolerant' describes biota that predominately occur in fresh water (salinity ,3 g L À1 ) but occasionally occur in waters with salinity up to ,20 g L À1 , whereas 'halophiles' are biota that occur mostly in athalassic waters (salinity .10 g L À1 ; Williams 1981). We have not attempted to distinguish between 'highsalinity' (sometimes called halobiontic) and 'low-salinity' (sometimes called halophilic) taxa due to varying definitions (e.g. Timms 1983v. Williams 1981v. Bayly 1972 and because many taxa do not neatly fit into one or the other category. It is acknowledged that our definition neglects some species that occur in salt lakes at salinities .3 g L À1 but typically less than 10 g L À1 that have been described as halophilic elsewhere (e.g. Eocyzicus spp. and Branchinella spp. in Timms 2007Timms , 2014. Australian salt lakes support unique communities. Some invertebrate groups, particularly crustaceans, are well represented, with many species, genera and even one family restricted to salt lakes (Halse and McRae 2004;Timms 2014). The gastropod Coxiella (which includes the subgenus Coxielladda) is unique to these environments and the only gastropod genus anywhere in the world to consist entirely of halophilic species (Williams and Mellor 1991). Insects are commonly encountered, but, with few exceptions, these are halotolerant rather than halophilic (Williams and Kokkinn 1988;Timms 1993;Pinder et al. 2005). Fish are rare and are mainly found in low-salinity permanent lakes, but may occur in large episodic lakes during heavy filling events when these lakes become connected to refugia such as mound springs supporting fish (e.g. Craterocephalus eyresii (Steindachner) in Lake Eyre and Lake Torrens; Williams and Kokkinn 1988;Williams et al. 1998). Numerous waterbirds opportunistically use salt lakes for feeding and even breeding (Weston 2007;Pedler et al. 2018) and some, such as banded stilts Cladorhynchus leucocephalus (Vieillot), are largely restricted to these lakes (Pedler et al. 2014). These birds are probably important vectors of dispersal for salt lake invertebrates (Green et al. 2008;Sánchez et al. 2012). Australian salt lakes also have distinctive fringing plant communities dominated by genera such as Tecticornia (samphires) and Frankenia (sea heath; Lyons et al. 2004). Salt-tolerant macroalgae such as Characeae and the vascular plants Ruppia and Lepilaena form thick mats in mildly saline waterbodies (Porter 2007;Casanova 2013), and microalgae such as Dunaliella thrive in hypersaline conditions, where they give lakes a characteristic pink hue (Teller 1987).
Salt lakes and their unique biota are threatened globally (Williams 2002). In Australia, reduced rainfall, associated with anthropogenic climate change, is having a widespread and dramatic effect as it causes lakes to fill less often, with shorter hydroperiods (i.e. periods during which water is present) and higher salinities (Hughes 2003;Nielsen and Brock 2009). These changes can have a profound effect on community composition, especially in previously low-salinity or seasonal lakes (Williams 2002;Pinder et al. 2005). Secondary salinisation, agricultural and mining activities, groundwater extraction, diversion of surface flows and pollution also present significant threats to Australian salt lakes and have already deleteriously affected these environments in some areas (Williams 1995;Timms 2005). For example, increased salinisation, changes to hydroperiod and, in some cases, acidification associated with secondary salinisation have diminished the faunal communities of some salt lake environments on the Eyre Peninsula, South Australia (Williams 1984;Timms 2009a) and in the Wheatbelt region of Western Australia (Cale et al. 2004). In addition to expirations, there is a risk that these hydrological changes may lead to the extinction of entire species, at least in Western Australia, without intervention (Halse et al. 2003). Despite the unique biodiversity and apparent threats, very little has been done to assess the conservation status of salt lake invertebrates ).
Conservation planning should be informed by evidence (Sutherland and Wordley 2017). To effectively assess and manage threats to Australian salt lake environments, it is important to synthesise the available information on these environments and use this information to document general trends and highlight critical knowledge gaps. There is, however, no current systematic evaluation of the state of knowledge of the biota of Australian salt lakes. De Deckker (1983b) reviewed the history, chemistry and biota of Australian salt lakes, but our knowledge of these systems has since improved. There have been significant advances in knowledge of the systematics and biology of a range of taxa (Hebert and Wilson 2000;Halse and McRae 2004;Timms 2014) and expanded coverage of some regions (Pinder et al. 2002;Timms et al. 2006;Timms 2008). The study reviews the current state of knowledge of halophilic macroinvertebrates of Australian salt lakes. It focuses on crustaceans and gastropods because they are an important and conspicuous component of these lakes.
The specific aims of this study were to: (1) synthesise knowledge of halophilic crustaceans and gastropods from Australian salt lakes, focussing on information published since De Deckker (1983b); (2) draw attention to important issues in Australian salt-lake conservation; and (3) identify general trends, gaps in knowledge and directions for future research in this area.

Materials and methods
This study reviews what is known about halophilic macroinvertebrates from Australian salt lakes, focusing on data published since De Deckker's (1983b) review. For quality control, it mainly relies on peer-reviewed studies, but unpublished theses and reports were used when they contained crucial information that was not otherwise available.
Halophilic macroinvertebrates were identified from published ecological studies of Australian salt lakes (Table 1) and from unpublished data held by the Western Australian Department of Biodiversity, Conservation and Attractions (DBCA). The dataset analysed during this study is available from the corresponding author on reasonable request. In all, 79 described halophilic species in 23 genera were identified. Almost all species were either crustaceans or gastropods. Insects were excluded because most species in Australian salt lakes are halotolerant rather than halophilic (Timms 1993;Pinder et al. 2005). Rotifers have also been excluded on the basis that they are microscopic rather than macroscopic (Blinn et al. 2004).
Various methods were used to find articles on the biology of the identified halophiles, including searches for keywords 'salt', 'saline', 'lake', 'wetland', 'Australia', 'invertebrate', 'fauna', 'halophile', 'halophilic', 'halobiont', 'crustacean' and 'gastropod' on Scopus and Google Scholar databases. Both databases were last accessed on 1 December 2020. This review includes salinity data from multiple studies. These studies typically estimated salinity (the sum of all ion concentrations per unit solution) from measurements of conductivity (which has particular utility in Australian salt lakes because most have a homogenous ionic composition; Bayly and Williams 1966) or as gravimetric total dissolved solids (TDS; which may include small but variable amounts of organic matter that can result in exaggerated salinity values; Williams and Sherwood 1994). Salinity data reported from gravimetric determination (TDS) in the source article have been converted to grams per litre of dissolved solutes in this review to facilitate comparisons across different studies. This was done using a correction factor of 0.91, which is based on the highly correlated relationship between salinity estimated using conductivity and salinity estimated from gravimetric methods (Bayly and Williams 1966).

Parartemia
The biology of Parartemia has been reviewed recently by Timms (2014), so only general points are discussed here.
Taxonomy. Parartemia, the only genus in the family Parartemiidae, is unique to Australia and is composed exclusively of halophilic species (Table 2). With 18 described and one undescribed species (Timms 2014), it is one of the most speciose genera in salt lakes.
The taxonomy of Parartemia is relatively advanced compared with that of most other Australian salt lake invertebrates, although a substantial number of species (nine) have only been described in the past 10 years (Table 2) and another species is yet to be described (Timms 2014). The most recent taxonomic assessments of Parartemia are provided by Timms and Hudson (2009), who described four new species from South Australia, and Timms (2010b), who described six new species from Western Australia. The taxonomic work is based on morphology, but Remigio et al. (2001) used variation in the mitochondrial 16S gene to confirm that the eight species described at that time were genetically distinct and as evidence of undiscovered species in Lake MacLeod (GenBank Accession number AY014794) and another in 'Scadden East' (GenBank Accession number AY014795) in Western Australia. Whether the undescribed species of Remigio et al. (2001) are populations of species subsequently described by Timms (2010b) is not clear.
Ecology. Species diversity in Parartemia decreases from western to eastern Australia, with Western Australia having 13 described species (10 endemics) compared with 6 (3 endemics) in South Australia and 2 (no endemics) in the eastern states ( Table 2). The ubiquity of species varies, ranging from geographically widespread and known from many sites (e.g. Parartemia cylindrifera Linder, Parartemia longicaudata Linder and Parartemia zietziana Sayce) to only known from one or a few sites in a single region (e.g. Parartemia triquetra Timms & Hudson and Parartemia auriciforma Timms & Hudson;). Knowledge of the distributions of some species is inadequate, especially those like P. triquetra and P. auriciforma, which occur in remote areas that have been poorly surveyed (Timms and Hudson 2009;Timms 2010b).
Parartemia species occur mainly in ephemeral and seasonal natural salt lakes, surviving dry periods as drought-resistant cysts (Timms 2012a). They are strong hypoosmotic regulators (Geddes 1981) and the most salt-tolerant of the endemic invertebrates in Australian salt lakes (Timms 2014), but little is known about the physiological basis for this. All 15 species for which relevant data are available have been collected from lakes with salinities .90 g L À1 , and 8 of these were from lakes with salinities in excess of 200 g L À1 (Table 2). Laboratory experiments by Manwell (1978) suggested that one species, P. zietziana, was not able to withstand the highest salinities tolerated by Artemia, possibly due to a lesser capacity to produce sufficient haemoglobin to compensate for low oxygen concentrations at very high salinities. Based on field data, the salinity tolerances of Parartemia species are very broad and overlap (Table 2). Nevertheless, species rarely co-occur in the same waterbody and a combination of physical factors, such as water duration, filling pattern, pH and salinity, are suggested to be important in determining species distributions (Timms 2009c;). In addition, Parartemia eggs sink and become bound up in sediments, reducing dispersal ability ). The majority of species live in alkaline waters, but Parartemia mouritzi Timms, Parartemia contracta Linder and Parartemia acidiphila Timms & Hudson are known to inhabit acidic lakes (Conte and Geddes 1988;Timms 2009bTimms , 2010bTimms and Hudson 2009).
Life history. Sexes are separate for all Parartemia (Timms 2014). Detailed lifecycle and reproductive information is available only for P. zietziana (Marchant and Williams 1977a), which reproduces ovoviviparously, producing multiple cohorts so long as conditions are favourable (Timms 2012b) and uses oviparous reproduction to produce highly resistant cysts when conditions become unfavourable (Geddes 1976;Marchant and Williams 1977b). The cysts of Parartemia laticaudata and Parartemia veronicae have the same stress proteins (p26, artemin and heat shock protein 70) that enable Artemia (cysts) to survive severe desiccation (Clegg and Campagna 2006). Cyst hatching appears to follow a 'bet-hedging' strategy, whereby some eggs hatch within hours of a filling event but not all eggs hatch at once (in case the water does not persist long enough to allow reproduction; Timms 2012b).
Trophic ecology. Dietary information is available only for P. zietziana, which selectively feeds on benthic or suspended particles with a high organic content (Marchant and Williams 1977b). It has been assumed that other species feed similarly (Timms 2014). Timms (2012a) suggested that Parartemia mainly feed on organic matter that is resuspended in the water column, which provides limited opportunity for niche diversification based on food resources. Timms (2012a) hypothesised that this may give a species established in a lake a competitive advantage over a species that recently arrives, and hence explain why lakes typically only contain one Parartemia species. Exactly how this may give a resident species a competitive advantage is not clear.
Parartemia are an important food source for waterbirds, such as the banded stilt, that may fly thousands of kilometres to take advantage of booming Parartemia populations (Pedler et al. 2018).
Conservation status. Parartemia is unique among Australian salt lake genera in that its conservation status has been assessed, at least in Western Australia ). The dominant threats to Parartemia are habitat degradation, including acidification and changes to hydrology (especially prolonged inundation) from secondary salinisation ). Parartemia contracta is currently the only species listed by the International Union for Conservation of Nature (as vulnerable), although it is one of the more common species and may not be as at risk as previously thought ). The species that are regarded as at most risk (Parartemia extracta and Parartemia boomeranga) are only known from the Wheatbelt region, where altered hydrology and secondary salinisation is widespread (Halse et al. 2003).  recommended that P. extracta be considered as vulnerable because it has disappeared from many sites and is secure only in lakes in the Jurien Bay area (Fig. 1;Timms 2014;Pinder and Quinlan 2015). Another species, namely P. boomeranga, may be extinct (Timms 2012b) because it has not been recorded recently and all the lakes this species was known to occur in are affected by secondary salinisation ( Fig. 1; ). Concerns have also been raised for P. mouritzi and Parartemia bicorna, both of which are only known from a few lakes that are currently being affected by either secondary salinisation or mine dewatering discharge (Timms 2014).

Artemia
The genus Artemia comprises seven (Rogers 2013) or possibly nine (Naganawa and Mura 2017) species. Although Artemia is widespread and common in hypersaline lakes in most continents, only Artemia franciscana Kellogg and Artemia parthenogenetica Bowen & Sterling are present in Australia (McMaster et al. 2007;Asem et al. 2018). A. franciscana was introduced to Australia by humans to aid in salt production and is overwhelmingly restricted to constructed evaporative ponds (salt works; Timms and Hudson 2009;Asem et al. 2018). A. parthenogenetica may have also been introduced into salt works by humans (McMaster et al. 2007), but its presence in a range of lakes in south-western Australia could be the result of intercontinental bird-mediated dispersal followed by local dispersal (McMaster et al. 2007).
All Artemia species, including the two species found in Australia, have high dispersal capacity because their cysts float and are effectively transported by animals and wind (Timms and Hudson 2009). Despite this, A. franciscana does not seem to be spreading (Timms and Hudson 2009;Asem et al. 2018), but the distribution of A. parthenogenetica is increasing in southwestern Australia, where it is mostly colonising degraded salt lakes (McMaster et al. 2007). It is unclear how the further spread of A. parthenogenetica may affect Parartemia species, but currently A. parthenogenetica appears limited to lakes not already occupied by Parartemia (McMaster et al. 2007).

Branchinella
Branchinella, with 40 described species, is the most speciose of the Australian anostracan genera (Timms 2015a(Timms , 2015b. Many species have only recently been described (Timms 2015b), with cryptic lineages identified using genetic data (Pinceel et al. 2013b). Only two species, Branchinella buchananensis Geddes and Branchinella simplex Linder, are halophiles; the other species are either strictly freshwater or tolerant of only low levels of salinity (Pinceel et al. 2013a;Timms 2014). Further work is needed to determine whether Branchinella halsei Timms is conspecific with B. simplex, because these taxa are morphologically distinct but molecularly congruent (Pinceel et al. 2013b).
B. buchananensis occurs, usually at salinities of ,15 g L À1 (limit 42 g L À1 ; Timms 2009c), in Queensland and northwestern New South Wales, where it is listed as vulnerable under the Fisheries Management Act 1994 (Timms 2014). B. simplex is known, usually in salinities of ,30 g L À1 (limit 62 g L À1 ; Timms 2015b), from lakes in central Western Australia, northern South Australia and the southern Northern Territory.
The results of a molecular phylogeographic study suggest that the two halophilic species have evolved from different lineages (Pinceel et al. 2013b). B. buchananensis and four halotolerant species comprise a clade that is estimated to have evolved from a freshwater ancestor somewhere between c. 62 and 23 million years ago, which coincides with increasing availability of temporary saline aquatic habitats in Australia (van de Graaff 1977;Pinceel et al. 2013a).

Notostraca
Triops Both living notostracan genera, Triops and Lepidurus, occur in Australia (Pinder et al. 2005), although only Triops is known from salt lakes (Timms 2009c  thought to be represented in Australia by a single, widespread and morphologically variable species (Triops australiensis Spencer & Hall; Timms 2012a). However, a recent molecular appraisal has revealed the presence of a range of putative new species and predicted that further diversity would be found once lakes in central and western Australia are adequately surveyed (Murugan et al. 2009;Meusel and Schwentner 2017). Little ecological information is available for halophilic Triops except that undescribed species have been reported from Lake Carey (salinity 11.6-84.3 g L À1 , mean 30.3 g L À1 from 10 records; Timms et al. 2006), the Esperance region in Western Australia (salinity 27 and 31 g L À1 , 2 records; Timms 2009b), the Paroo wetlands in New South Wales (salinity 0.3-19.3 g L À1 ; Timms 1993) and Lake Torrens in South Australia (salinity ,16 g L À1 ; Williams et al. 1998).

Cladocera
The bulk of Cladocera that occur in inland waters in Australia are restricted to fresh water, but three groups have representatives in salt lakes. These groups comprise: (1) six species of Daphniopsis (or Daphnia; see below); (2) two species of Daphnia (Daphnia salinifera Hebert and Daphnia neosalinifera Hebert) from the Daphnia carinata (King) subgenus; and (3) three species of chydorid: Moina baylyi Forró, Moina mongolica Daday and Extremalona timmsi Sinev & Shiel. All species are endemic to Australia except for M. mongolica, which is also known from northern Africa, Europe, the middle-east, Russia and China (He et al. 2001), notwithstanding that molecular data suggest that there are many unrecognised species of Moina (Bekker et al. 2016). Colbourne et al. (2006) collected D. salinifera from Lake Wyora in Queensland and D. neosalinifera from Colac in Victoria and suggested that these species are usually found at lower salinities than the Australian species of Daphniopsis. However, apart from this, there are virtually no published data on the former two species, and they are not considered any further herein.

Daphniopsis
Taxonomy. Historically, the daphniids in Australian salt lakes were classified in the genus Daphniopsis (Sars) and just one species, Daphniopsis pusilla (Serventy), was recognised. Six species, all supported by genetic data (Hebert et al. 2002), have now been described (Sergeev and Williams 1985;Sergeev 1990aSergeev , 1990bHebert and Wilson 2000). Furthermore, one of six ephippia morphotypes identified by Kokkinn and Williams (1987), namely 'morphotype six' from Lake Eyre South, has not been accounted for by any described species (Hebert and Wilson 2000), and there are reports of undescribed species from salt lakes in the Wheatbelt region in Western Australia (Pinder et al. 2005) and the Eyre Peninsula in South Australia (Timms 2009a). The diversity of Daphniopsis in Australia is high, given that Daphniopsis is represented by a single species in either fresh or saline water in other regions where it occurs (Bayly 1995;Hebert and Wilson 2000;Gibson and Bayly 2007). Because all the described Australian species form a monophyletic group relative to other daphnids (Colbourne et al. 2006), it appears that this group has undergone a significant radiation in Australia.
The taxonomic status of Daphniopsis, specifically its relationship to the genus Daphnia, has long been debated (Schwartz and Hebert 1984). Benzie (2005) concluded that Daphniopsis as a whole is not morphologically or genetically distinct enough to warrant genus status and suggested that this taxon should be subsumed into the Daphnia subgenus Ctenodaphnia. Data from three mitochondrial genes support the view that Daphniopsis is a component of the genus Daphnia (Colbourne et al. 2006). Nevertheless, all the Daphniopsis species that were included in the above genetic study formed a monophyletic group, except for Daphniopsis ephemeralis (Schwartz & Hebert), which is the only representative of this taxon from the Northern Hemisphere. Most recent studies retain the name Daphniopsis (Säwström et al. 2009;Ismail et al. 2010aIsmail et al. , 2011aMcCloud et al. 2018;Wang et al. 2019) and this practice has been adopted herein.
Ecology. Given that several species have only been recently described, it is difficult to gauge the reliability of historical accounts of species distributions, although, based on data from recent field records, it appears that south-western Australia has the greatest diversity (Hebert and Wilson 2000). All six described species have been found from at least two sites in this region and two, Daphniopsis wardi (Hebert & Wilson) and Daphniopsis pusilla, are endemic. Historical reports of the latter species from other parts of southern Australia predate the current taxonomy and are due to misidentification (Hebert and Wilson 2000). Another two species, Daphniopsis quadrangulus (Sergeev) and Daphniopsis australis (Sergeev & Williams), predominately occur in lakes in south-eastern Australia, but each has been recorded twice in south-western Australia (Fig. 2a). Daphniopsis truncata (Hebert & Wilson) and Daphniopsis queenslandensis (Sergeev) are common in Western Australia and South Australia, and D. queenslandensis also occurs in New South Wales and Queensland (Fig. 2b).
Osmoregulation has been studied in two species of Daphniopsis (D. pusilla and D. australis). Both were determined to be osmoregulators, where the haemolymph is hyperosmotic to the environment at salinities under 8 g L À1 , isosmotic in waters with salinity 8-20 g L À1 and hypoosmotic in waters with salinity .20 g L À1 (Aladin and Potts 1995).
All Australian species of Daphniopsis are halophilic. Field salinity data suggest that most species are common in 17-30 g L À1 , but nevertheless the salinity ranges of most species are very broad ( Table 3). The salinity distribution of one species, D. quadrangulus, is mostly unknown because there are only two reported occurrences of this species (Table 3). Some reports of species from very high salinities should be viewed with caution because they may relate to preserved dead or dying specimens. For example, experimental data indicate that D. australis experiences significant mortality at salinities below 5 and above 33 g L À1 (Ismail et al. 2010a), suggesting that reports of live D. australis at a salinity of 154.1 g L À1 by Sergeev and Williams (1985) are unlikely.
seasonally in the winter-spring (Campbell 1994). Similarly, D. queenslandensis and D. truncata have been reported to be more abundant in winter-spring than in summer-autumn (Timms 2008(Timms , 2009b(Timms , 2018, despite minimal differences in salinity between seasons. Life history. Like most daphniids, Australian Daphniopsis are cyclical parthenogens, although there are interspecific hybrids that reproduce by obligate parthenogenesis (Hebert and Wilson 2000). The switch from parthenogenesis to sexual reproduction and the production of desiccation-resistant ephippial eggs is trigged by unfavourable conditions, possibly one (or a combination) of low food availability, high population density, high salinity and high temperature (Sergeev and Williams 1983;Williams 1986;Ismail et al. 2010b). Detailed lifecycle data are available only for D. australis under laboratory conditions, as described by Ismail et al. (2010b). These data suggest that the parthenogenetic females usually live for 20-30 days, take 6-7 days to reach maturity and produce ,10-12 clutches during their lifetime. The ephippial females have a shorter lifespan and usually produce a maximum of two diapausing eggs. These females are produced under conditions of high population density and can switch from sexual to parthenogenetic reproduction, particularly at low and high food densities in the presence of males whose persistent attempts at mating may stress the females.
Trophic ecology. Little is known about the trophic ecology of Daphniopsis in Australian salt lakes. Based on information for other species of Daphniopsis (Säwström et al. 2009) and Daphnia (Taipale et al. 2012), these species likely consume microalgae and bacteria and are important herbivores in the lake food webs. Ismail et al. (2011a) fed different combinations of three microalgae species commonly used in aquaculture to D. australis and found that the composition of algae in the diet influenced the growth and reproductive output of individuals. Based on detailed observations of swimming behaviour under laboratory conditions, McCloud et al. (2018) developed the following hypotheses for D. australis: (1) males mainly occur in the water column, where they filter feed; (2) parthenogenetic females spend most of their time close to the bottom and mainly rely on benthic resources; and (3) the habits of the ephippial females are intermediate between those of the males and parthenogenetic females. The ostracod Australocypris insularis is known to predate on Daphniopsis (Campbell 1995).

Chydorids
The endemic halophile M. baylyi is widespread in central Australian salt lakes (Timms 2007) and the predominantly Northern Hemisphere M. mongolica has been encountered once in a coastal saline lake near Carnarvon in Western Australia ( Fig. 3; Halse et al. 2000a). Very little information is available for M. baylyi other than it is commonly reported at salinities between 2.9 and 60 g L À1 and up to 86.7 g L À1 , and is more common during the summer rather than winter filling of lakes (Timms 1987;Williams and Kokkinn 1988;Timms 1993Timms , 2001Williams et al. 1998). The monotypic E. timmsi is the most recently described halophilic cladoceran from Australia (Sinev and Shiel 2012) and is known only from two acidic salt lakes near Esperance, Western Australia (as Alona spp. in Timms 2009b).

Ostracoda
A diverse and heterogeneous range of halophilic ostracods occurs in Australian salt lakes. These ostracods can be divided into two main groups: (1) giant ostracods from the Australian endemic subfamily Mytilocypridinae (herein called giant ostracods); and (2) non-Mytilocypridinae ostracods (herein called small ostracods).

Giant ostracods
The giant ostracods are .3 mm in length and are a conspicuous component of the biota of Australian salt lakes.
Taxonomy. The morphotaxonomy of giant ostracods is relatively well established from the early work of De Deckker (1978, 1981a) and more recent work by Halse and McRae (2004). Halse and McRae (2004) added 2 new genera and 6 new species, bringing the total to 21 described species in 6 genera, although there are uncertainties regarding the status of a few described species (e.g. whether Mytilocypris tasmanica (McKenzie) and Mytilocypris praenuncia (Chapman) should be synonymised; Finston 2000). There are also reports of an undescribed species of Mytilocypris (Pinder et al. 2010;Quinlan et al. 2016) and of Lacrimicypris (L. Bourke, K. Brown, and G. Paczkowska, unpubl. data, 2018), both from Western Australia. Further species are likely to be found, particularly in remote areas (Halse and McRae 2004). Other than for some allozyme data that have helped resolve species questions in Mytilocypris Table 3. Salinity records for six species of Australian Daphniopsis based on field measurements Sources are as follows: 1, Cale et al. (2004); 2, Edward (1983); 3, Halse et al. (2000a);4, Pinder et al. (2005); 5, Pinder and Quinlan (2015); 6, Sergeev and Williams (1985); 7, Timms (1993); 8, Timms (1998);9, Timms (2001); 10, Timms (2009a) (Finston 2000;Halse and McRae 2004), the current taxonomy of giant ostracods is based on morphological characters, mainly the structure of the hemipenis. Although some giant ostracod species are known only from fresh or low-salinity waterbodies, most are halophiles (Table 4). These halophiles occur in four of six genera: Australocypris (all seven species), Repandocypris (all two species), Mytilocypris (five of seven species, assuming M. praenuncia and M. tasmanica are distinct) and Trigonocypris (one of two species; Table 4). The broad taxonomic distribution of halophilic, low-salinity and freshwater forms suggests that halophily has probably evolved more than once in the Mytilocypridinae.
Ecology. Other than for some taxonomic and distribution information, very little is known about the ecology of giant ostracods, even for Australocypris, which is a very diverse, widespread and abundant component of the salt lake ecosystems.
The highest diversity of halophilic giant ostracods is in Western Australia (Table 4), particularly for Australocypris, which shows a marked decrease in species richness and endemism from west to east (Table 4). Most halophilic Mytilocypris species occur broadly across Australia, although M. praenuncia-tasmanica (see above) has only been reported from southeastern Australia (Table 4). Repandocypris has one species only known from Western Australia (Repandocypris austinensis Halse & McRae) and another only known from South Australia (Repandocypris gleneagles Halse & McRae; Table 4). The halophilic species of Trigonocypris (Trigonocypris globulosa (De Deckker)) is found in Western Australia, South Australia, New South Wales and Queensland (Table 4). Most halophilic species of giant ostracod are broadly distributed, although a few seem to have a very narrow range (Halse and McRae 2004). Australocypris mongerensis Halse & McRae is only known from a small mildly saline claypan next to Lake Monger in Western Australia (Halse and McRae 2004). However, knowledge of species distributions, particularly for those species first described by Halse and McRae (2004) like A. mongerensis, is likely to be incomplete.
With the exception of A. mongerensis, the halophilic species of giant ostracod have been collected from a broad and overlapping range of salinities (Table 4). Eight species have been recorded from salinities .100 g L À1 (Table 4). On average, Mytilocypris species tend to occur in a lower and narrower range of salinities than Australocypris species (Table 4). However, the salinity distribution of Mytilocypris mytiloides (Brady) is very broad and extends into highly saline water (Table 4). Both species of Repandocypris have only ever been found at salinities .40 g L À1 (Table 4), although they are only known from a few sites (Halse and McRae 2004). The salinity range of T. globulosa is broad and exceeds 100 g L À1 . Based on data for M. praenuncia, halophilic giant ostracods are osmoregulators (Aladin and Potts 1996).
All giant ostracods occur in neutral or slightly alkaline waters, but the distributions of A. insularis (Chapman),

Australocypris bennetti
Trophic ecology. Giant ostracods tend to actively swim in the water column (De Deckker 1983a). This, and their large size, is likely related to the general absence of fish predators in the salt lakes or temporary freshwater waterbodies they inhabit (De Deckker 1983a). The diet of giant ostracods is poorly understood, although at least A. insularis consumes detritus and is also planktivorous and capable of significantly reducing the abundance of Calamoecia, Daphniopsis and Diacypris in microcosm experiments (Campbell 1995).

Small ostracods
Taxonomy. Most of the small ostracods in Australian salt lakes belong to the genera Diacypris and Reticypris. These genera contain seven and five described species respectively (Table 5), all of which are halophilic (De Deckker 1981b). They comprise the subfamily Diacypridinae, which is endemic to Australia, and have no known recent freshwater or marine ancestors (De Deckker 1981b). The taxonomy of Diacypridinae has not been addressed since De Deckker (1981aDeckker ( , 1981c, and the extent to which it captures the biodiversity of this group is not clear. There are reports of multiple undescribed species of Diacypris and Reticypris, predominately from Western Australia (Williams 1984;Williams and Kokkinn 1988;Timms 1993;Pinder et al. 2005;Quinlan et al. 2016). Species diagnoses are based on the structure of the hemipenis (see fig. 13 in De Deckker 1981c for Diacypris). Some characters, such as variation in the carapace shape and size, are ecophenotypic at least in several Diacypris and Reticypris species (De Deckker 1981c, 1981d. Other small halophilic ostracods from Australian salt lakes are the endemic Platycypris baueri (Herbst), Patcypris outback Halse & Martens and Trilocypris horwitzi Halse & Martens, which are all in the cosmopolitan subfamily Cyprinotinae (De Deckker and Geddes 1980;De Deckker 1981b;Karanovic 2012;Halse and Martens 2019). In addition, Billcypris davisae Halse & Martens has been collected from Lake Lefroy and Lake Cowan in the Goldfields region of Western Australia (Halse and Martens 2019). These are both large hypersaline lakes and this species is likely halophilic, although no information on the salinity at the time of its collection is available (Halse and Martens 2019). The endemic species Cyprideis australiensis Hartmann and Leptocythere lacustris De Deckker have also been reported from Australian salt lakes but, because these species have marine ancestries, they lack desiccation-resistant eggs and so are restricted to lakes with permanent water (De Deckker 1983a;Pinder et al. 2002;Schön et al. 2017). A few species of Cyprinotus, most notably Cyprinotus cingalensis Brady, which is widely distributed in  (Karanovic 2008), but usually in low-salinity waters, and are more halotolerant than halophilic (Timms 1993(Timms , 2009bPinder et al. 2002).
Ecology. Knowledge of the biology of the small halophilic ostracods from Australian salt lakes is rudimentary. Two of the Cyprinotinae species (P. outback and T. horwitzi) have only recently been described (Halse and Martens 2019) and thus far reported from only one or a small number of sites in Western Australia. These species are not considered in the following discussion. Understanding the ecology of Diacypris and Reticypris is complicated by the fact that these taxa are not identified to species level in many studies (Timms 1998(Timms , 2018Timms et al. 2006).
All described Diacypris and Reticypris species, as well as P. baueri, have broad and overlapping geographic distributions (Table 5). All these species are known from both Western Australia and South Australia, except for Reticypris herbsti McKenzie (known from South Australia but not Western Australia) and Reticypris pinguis De Deckker (known from Western Australia but not South Australia; Table 5). Some of these species also occur in other states, the exact details of which vary from species to species (Table 5). Many species, such as Diacypris phoxe De Deckker, R. herbsti, Reticypris kurdimurka De Deckker, R. pinguis and Reticypris walbu De Deckker, are patchily distributed (i.e. reported from a small number of sites that are typically separated by hundreds or even thousands of kilometres; Fig. 4).
It is often said that dispersal of inland ostracods, including halophilic small and giant species, is mediated by birds (De Deckker 1977;Halse 2002). This is supported by the fact that ostracod eggs have been hatched from the faeces of waterbirds that frequent the temporary wetlands in arid Australia, although no information on the species hatched is available (Green et al. 2008).
Ecology. Both C. clitellata and C. salina occur throughout mainland southern Australia; the former is also found in Tasmania (Fig. 6). By contrast, C. trilobata is endemic to Western Australia (Fig. 6a).
Despite being osmoconformers (Bayly 1972), these calanoids are often reported from high salinities; for example, C. clitellata and C. salina have been reported from salinities up to 132 and 195 g L À1 respectively (De Deckker and Geddes 1980;Timms 2009b). Both C. salina and C. clitellata are often found in the same lake, with C. salina succeeding C. clitellata at higher salinities, which is attributable to its higher hatching salinity (61.4-82.9 g L À1 ) than that of C. clitellata (20-71 g L À1 ; Geddes 1976). Less information is available for C. trilobata, which has only been reported from acidic salt lakes (pH 2.96-6.3) and primarily occurs at salinities of ,31-63 g L À1 (Pinder et al. 2005;Timms 2009b). Some records of C. trilobata at salinities of 160 and 240 g L À1 are probably of dead, brine-preserved individuals (Halse and McRae 2001) and therefore overestimate the upper salinity tolerance of this species (Halse and McRae 2001).
Life history. Calamoecia, like most salt lake crustaceans, produces two types of eggs: (1) subitaneous eggs that maintain populations during favourable conditions; and (2) resistant resting eggs that allow populations to persist through unfavourable environmental conditions (Whitehead 2005). These resting eggs are likely important for dispersal. Waterbirds are considered the most likely vector for dispersal (Whitehead 2005), although water or human movement may also be important (Maly et al. 1997). Population genetic data for C. clitellata suggest that gene flow is restricted between regions and drainage basins, and sometimes even on fine spatial scales, and therefore populations are typically isolated and self-sustaining (Whitehead 2005). However, occasional dispersal may have significant effects on evolutionary timescales, resulting in secondary contact between highly divergent lineages (Whitehead 2005).
Trophic ecology. Apart from reports that Calamoecia feed on microalgae (Whitehead 2005) and are preyed upon by A. insularis and potentially other giant ostracods (Campbell 1995), little is known of the trophic ecology of the salt lake calanoids.

Cyclopoidea
In Australian salt lakes, cyclopoids are represented by three endemic species from two genera, as well as the cosmopolitan halophiles Apocyclops dengizicus (Lepeshkin) and Pescecyclops arnaudi (Sars) (Anufriieva 2015). However, the taxonomy of these copepods is confused, due, in part, to a series of name changes (Table 6) and to reports of a significant number of undescribed halophilic species from throughout Australia (De Deckker and Geddes 1980;Brock and Shiel 1983;Williams and Kokkinn 1988;Williams et al. 1990;Timms 1993;Timms and Boulton 2001;Pinder et al. 2005). It is impossible to meaningfully review the biology of these species until the taxonomy is resolved.

Harpacticoidea
The harpacticoid genera Mesochra and Schizopera occur in Australian salt lakes, but only one described species, M. baylyi, is known to be halophilic (Hamond 1971;De Deckker and Geddes 1980;Brock and Shiel 1983). Reports of other harpacticoid species from Australian salt lakes are either of undescribed species or of marine and estuarine species that have colonised permanent or semi-permanent coastal salt lakes (Bayly 1970;Hamond 1971;Timms 1993Timms , 2001Timms , 2009bPinder et al. 2005). M. baylyi is widely distributed in southern Australia (south-western Western Australia, south-eastern South Australia, Victoria and Tasmania) and has been reported in waters with salinities up to 129 g L À1 , but is typically reported in waters with salinities below 25 g L À1 (De Deckker and Geddes 1980;De Deckker and Williams 1982;Pinder et al. 2005). It is one of only a few halophilic species suggested to have colonised salt lakes directly from the marine environment (Williams 1981).

Isopoda
The oniscoid Haloniscus searlei Chilton is the only described halophilic isopod that occurs in Australia, although recent molecular phylogenetic analyses have suggested that the diversity of this genus is underappreciated ). The putative new species (revealed by molecular data) are mostly from groundwater, freshwater or semiterrestrial environments, but at least one is from salt lakes (Cooper et al. 2008;Stringer et al. 2019).

Meridiecyclops platypus Fiers
Microcyclops platypus in Geddes et al. (1981) and Williams and Kokkinn (1988) Metacyclops platypus in Timms (1993) and Timms (2001) Meridiecyclops platypus in Timms et al. (2006) H. searlei sensu lato is geographically widespread, occurring in seasonal and permanent salt lakes in Western Australia, South Australia, Victoria and Tasmania (Williams 1983;Guzik et al. 2019). It is a powerful osmoregulator commonly found in salinities from 3.6 to 161 g L À1 (Williams 1983). It primarily grazes on diatoms, but has been noted to occasionally consume chironomid larvae (Blinn et al. 1989). This species is unusual among halophilic crustaceans because it has colonised the salt lakes via a terrestrial rather than freshwater ancestor (Ellis and Williams 1970). Related to this is the fact that it survives dry lake periods as an adult rather than with resistant eggs (Williams 1983) and appears unable to colonise lakes that only fill episodically (Williams and Kokkinn 1988;Timms 2008).

Coxiella
All halophilic gastropods in Australian salt lakes belong to the endemic genus Coxiella, the only gastropod genus in the world to consist exclusively of halophilic species. Coxiella is currently placed within the family Pomatiopsidae, which mainly contains freshwater species (Wilke 2019). However, molecular phylogenetic analysis suggests that Coxiella and Tomichia, a morphologically similar genus from Africa that includes freshwater, brackish and halophilic species, may merit recognition as a separate family (Tomichiidae; Wilke et al. 2013). Despite the ubiquity and abundance of Coxiella in Australian salt lakes, the genus is poorly studied (Williams and Mellor 1991;Pinder et al. 2002).

Taxonomy
The taxonomy of Coxiella has effectively remained unchanged since it was reviewed by Macpherson (1957), who listed 10 species. Kendrick (1978) added one further species (Coxiella roeae), known only from fossil material from Western Australia. Coxiella badgerensis (Johnston) has since been synonymised with Coxiella striata (Reeve) (Smith 1979), and Coxiella molesta Iredale and Coxiella minima Macpherson have not been collected since they were initially described (for a discussion about C. molesta, see Bayly and Williams 1966). Currently, nine extant species in the two subgenera Coxiella (eight species) and Coxielladda (one species) are recognised (Davis 1979;Wilke 2019), with seven species having been collected in recent decades. The discussion below focuses on these seven species.
Current species descriptions are based on external shell and operculum characters. Subsequent studies have found single specimens that exhibit the diagnostic morphological characteristics of more than one species (De Deckker and Geddes 1980;Williams and Mellor 1991) or have suggested the existence of undescribed species (Pinder et al. 2005;Timms 2009b). In addition, operculum characters have been found to vary with salinity (De Deckker and Geddes 1980) and size (Williams and Mellor 1991). Currently, researchers are unable to confidently identify Coxiella material to species level (Williams and Mellor 1991;Pinder et al. 2002;Timms 2009b). Owing to this inadequate taxonomy, all species-level information in this group, included that presented herein, must be considered tentative.

Ecology
Coxiella species are mostly found in southern Australia, including Tasmania, but also occur in some parts of central and southern Queensland (Williams and Mellor 1991;Timms 1998). Western Australia is the most diverse area for Coxiella, with six described species, including four that are endemic to this region: Coxiella striatula (Menke), Coxiella exposita (Iredale), Coxiella glabra Macpherson and Coxiella pyrrhostoma (Cox) (Table 7). Coxiella glauerti Macpherson occurs in both Western Australia and South Australia, whereas C. striata occurs in South Australia, Victoria and Tasmania, but not Western Australia (Table 7). Coxielladda gilesi (Angas) has been reported from remote inland lakes in Western Australia, South Australia and Queensland (Macpherson 1957;Timms 1998).
The geographical ranges of some Coxiella species overlap, but different species rarely appear to occur in the same waterbody, although there are reports of C. striata and C. glauerti cooccurring in lakes on the Eyre Peninsula (Timms 2009a). Weston (2007) identified four Coxiella species from bird faecal samples collected from a single lake in Western Australia, but the species identifications are questionable.

Species Description Distribution
Coxiella exposita Iredale (1943) Endemic to south-western WA glabra Macpherson (1957) Endemic to south-western WA striatula Menke (1843) Endemic to south-western WA A pyrrhostoma Cox (1868) Endemic to south-western WA glauerti Macpherson (1957) South-western WA and Eyre Peninsula striata Reeve ( Reported from the Wheatbelt, Western Australia, but the validity of the identification is unclear (see Pinder et al. 2002).
Coxiella are osmoconformers (Williams and Mellor 1991). Most field records of active individuals are from a salinity range between 10 and 70 g L À1 ; the lower and upper recorded ranges are 0.3 and 130 g L À1 Timms 1983;Williams et al. 1990;Williams and Mellor 1991;Pinder et al. 2005). These field data cannot always be accurately ascribed to particular species due to taxonomic uncertainties. Experimental data indicate that the adults of one species (possibly C. striata) from Lake Tallinga in South Australia have a broad salinity tolerance, with LD 50 limits of 2 and 95.5 g L À1 for gradual acclimation, and 6 and 83 g L À1 for a direct transfer (Williams and Mellor 1991). The extent to which Coxiella species vary in salinity tolerance is not clear, but some (e.g. C. glauerti) seem to occur at higher salinities than others (e.g. C. striata ;Timms 2009a;Timms et al. 2014).

Life history
The life histories of Coxiella species are poorly known. Species are dioecious and reproduction is sexual and probably iteroparous (Williams and Mellor 1991). Individuals are longlived (compared with most halophilic invertebrates), possibly for as long as 2 years (Williams and Mellor 1991). Related to this, Coxiella, like H. searlei, aestivate as adults during dry lake periods in areas of high humidity (e.g. under plant or algal mats) or in plates of densely packed individuals to reduce desiccation (Williams 1985;Timms et al. 2014). Laboratory experiments suggest that the Coxiella from Lake Tallinga can aestivate for at least 2 months with little mortality (Williams and Mellor 1991), whereas aestivating specimens of C. striata have been reanimated from lakes known to have been dry for at least 3 months (De Deckker and Geddes 1980). The total length of time that individuals can survive aestivated, and whether this varies among species, is not known. However, it must be limited because Coxiella are found in permanent and seasonal lakes, but not in episodic ones (Williams 1998a).

Trophic ecology
The diets of Coxiella species have not been investigated directly, but it has been suggested that they feed on benthic algae or detritus (De Deckker 1982;Bayly 1993). Coxiella are known to be consumed by native fish (e.g. Galaxias maculatus (Jenyns), Atherinosoma microstoma (Günther), Pseudogobius olorum (Sauvage) and Philypnodon grandiceps (Krefft); Chessman and Williams 1987;Becker and Laurenson 2007) and introduced fish (e.g. Cyprinus carpio Linnaeus; Khan 2003) in some waterbodies. They are also an important food source for some waterbirds, such as the hooded plover Thinornis rubricollis in south-western Australia, where Coxiella has been found to make up to 90% of the diet (Weston 2007).

Discussion
This study synthesises available information on the halophilic macroinvertebrates of Australian salt lakes, focusing on crustaceans and gastropods. The results show that our understanding of the taxonomy and diversity of some groups has improved markedly since the main previous review of these organisms (De Deckker 1983b), as has our knowledge of aspects of the ecology and life histories of a small number of species, although many knowledge gaps remain. This progress and these gaps, and their conservation implications, are discussed below.

Unique biodiversity
This review highlights the evolutionary distinctiveness of the halophilic macroinvertebrates in Australian salt lakes. The fauna is dominated by crustaceans and molluscs, of which 1 family (Parartemiidae), 2 subfamilies (Diacypridinae and Mytilocypridinae), 11 genera (Parartemia, Extremalona, Australocypris, Repandocypris, Diacypris, Reticypris, Platycypris, Patcypris, Trilocypris, Meridiecyclops and Coxiella) and 74 of the 79 described species are endemic to these lakes. These high levels of endemism are likely due to a long association between the fauna and Australian salt lake environments or their precursors (De Deckker 1981b;Williams 1981). The number of macroscopic halophilic invertebrate species in Australian salt lakes is also higher than in lakes elsewhere in the world (Alonso 1990;Bayly 1993;Williams et al. 1995;De Los Rios-Escalante and Amarouayache 2016). In addition, many components of the Australian fauna are capable of tolerating extreme levels of salinity, and some have other unusual characteristics. For example, although ostracod species are usually absent from environments with pH ,5 (because a lower pH inhibits formation of the calcite carapace; Ruiz et al. 2013), several species of ostracod occur in highly acidic salt lakes (Timms 2009b). By virtue of its distinctiveness, the invertebrate biodiversity of Australian salt lakes is of high conservation value.

Taxonomy
What we do not know, we cannot protect [Martens and Savatenalinton 2011].
In the past 20 years, considerable progress has been made regarding the discovery and description of species in a few groups of Australian salt lake invertebrates, notably Parartemia and giant (Mytilocypridinae) ostracods (Halse and McRae 2004;Timms and Hudson 2009;Timms 2010b). However, the taxonomy of other groups is much less advanced and in some cases (e.g. Coxiella gastropods and cyclopoid copepods) the current taxonomy does not allow consistent and accurate species identifications (Williams and Mellor 1991;Pinder et al. 2002). Many new species, and sometimes even new genera, have been discovered in those groups that have been subject to recent taxonomic revision, and more biodiversity likely remains to be discovered, especially in poorly studied groups and regions.
Molecular data have substantially changed our understanding of the diversity of many groups, such as Triops (Meusel and Schwentner 2017), Daphniopsis (Hebert and Wilson 2000), Parartemia (Remigio et al. 2001), Mytilocypris (Finston 2000(Finston , 2007, Branchinella (Pinceel et al. 2013a) and Haloniscus . Nevertheless, the discovery and identification of Australian salt lake invertebrates is still overwhelmingly reliant on morphotaxonomy. Using molecular and morphological data to document the full extent of the diversity of halophilic invertebrates of Australian salt lakes is essential to progress our understanding and conservation of these ecosystems.

Ecology
The species diversity of halophilic invertebrates is highest in Western Australian semi-arid and arid regions. Therefore, this part of Australia is an important region for the conservation of these invertebrates, especially because the salt lakes here are experiencing exceptionally high levels of disturbance due to climate change, secondary salinisation and mining (Halse et al. 2003;Timms 2005;Nielsen and Brock 2009).
Conservation planning and assessment requires a sound knowledge of species' distributions and how these are changing through time (Cardoso et al. 2011). This point is illustrated by attempts to assess the conservation status of Parartemia species, where concerns about P. contracta have decreased through time as more intensive sampling has revealed a range of additional populations while concerns about P. extracta have increased because temporal sampling suggests that the number of populations and the geographic range of this species are limited and shrinking Pinder and Quinlan 2015).
Our knowledge of the geographic distribution of most halophilic invertebrates is incomplete. This problem is particularly acute for those species that are difficult to identify because either the existing taxonomy is incomplete or because they cannot be reliably identified using morphological characters (e.g. Coxiella). It is also problematic for species that are rare or occur in remote locations. In addition, some older distributional data may be misleading because they are based on outdated taxonomy (e.g. D. pusilla). Obtaining a comprehensive inventory of the distribution of halophilic invertebrates in Australian salt lakes is a challenging prospect, partly because there are many lakes that either occur in remote locations or hold water infrequently. Surveying some sites may benefit from a flexible sampling approach, such as raising individuals from egg banks (Timms 2012b). Analysis of environmental DNA in dry sediments could also become an important survey tool but requires a database of reference sequences and a well-developed taxonomy to be effective (Cristescu and Hebert 2018).
The available evidence indicates that most of the halophilic invertebrates in Australia are osmoregulators (e.g. Parartemia, giant and small ostracods, Daphniopsis), but some are osmoconformers (e.g. calanoid copepods and Coxiella). Regardless, most species appear to have broad and overlapping salinity tolerances and many are able to tolerate extremely high salinities, as described above. However, the bulk of information on salinity tolerances in these invertebrates is derived from scattered field data. Such data do not necessarily indicate the optimal conditions for reproduction or completing the lifecycle, which may be much more restrictive than the conditions adults can survive (Hammer 1986). Those species that can only reproduce or complete their lifecycle at low salinity, or within a narrow salinity range, are seemingly more vulnerable to increases or changes in the salinity of the lakes. Lake Corangamite, a large permanent lake in western Victoria, provides a clear example of how increasing salinity can have a big effect on resident organisms. Due to the diversion of the main inflow creek dating back to the 1960s, water levels in this lake have fallen and salinity has increased, resulting in a change in the biota from one characteristic of moderate salinity (35-.50 g L À1 ) to one of higher salinity (50-100 g L À1 ; Williams 1995). Associated with this, populations of halophilic species have been lost from the lake (e.g. C. striata). Some salt lakes in central Western Australia and elsewhere are being subjected to the discharge of highly saline waste water from mining activities (Timms 2005;Timms et al. 2006). How these lakes are being affected by increased salt loads is unclear (Timms 2005) but, based on the Lake Corangamite example, community changes and local extinctions are possible.
Two interesting features of the distribution of the halophilic invertebrates are that: (1) some species are much more common and broadly distributed than others, even congeners; and (2) congeneric species rarely occur in the same lake, even though many have overlapping geographic and salinity distributions (see above). It is currently unknown whether the common and widespread species have relatively broader ecological tolerances, greater dispersal capacity or both (Williams 1984). The factors controlling the distribution of halophilic invertebrates are likely to be complex. The effects of salinity are correlated with other chemical and physical changes; for example, the availability of dissolved oxygen decreases as salinity (and temperature) increases (Williams 1998a). Thus, although the ecological tolerances of species are usually measured in terms of salinity, this is an oversimplification. Abiotic factors alone are unlikely to explain why congeneric species are rarely found in the same lake, although more intensive sampling and better taxonomy may lead to more examples of co-occurrence. Understanding the relative importance of various abiotic and biotic factors in controlling the distribution of the halophilic invertebrates is critical to predicting how species and communities will respond to perturbations and for developing effective interventions (Nielsen and Brock 2009).

Life history
Detailed life history information is typically available for one species for each of the main types (e.g. ostracod, cladoceran, gastropod) of halophilic macroinvertebrate in Australia. This information indicates that these invertebrates' life histories are fundamentally similar to those of their relatives from other environments (Ellis and Williams 1970;Finston 2002;Ismail et al. 2011b). Most species have a rapid life cycle, with shortlived individuals. Multiple generations may be produced while environmental conditions remain favourable and desiccationresistant eggs are used to survive dry or unfavourable conditions (Timms 2014). By contrast, H. searlei and Coxiella species survive dry or unfavourable periods as adults (Williams 1983;Williams and Mellor 1991). The total length of time that the aestivating adults can remain viable is not known, but those of C. striata can survive for at least 3 months (De Deckker and Geddes 1980) and those of H. searlei for at least 1 month depending on humidity (Williams 1983). This time is probably minimal compared with the duration that desiccation-resistant eggs of other taxa may remain viable (Williams and Kokkinn 1988). Consequently, Coxiella species and, particularly, H. searlei are excluded from waterbodies that dry for extended periods (Williams and Kokkinn 1988), and therefore seem especially vulnerable to a drying climate. Understanding the long-term viability and hatching cues for desiccation-resistant eggs, or the equivalent for the aestivating adults, and how these vary among species is crucial for assessing the capacity of species to survive as the climate dries.
It is generally believed that halophilic invertebrates in Australia mainly disperse via desiccation-resistant eggs (when present) and that aquatic birds and, to a lesser extent, wind and water flow are important dispersal vectors (Finston 2002;Green et al. 2008). Nevertheless, other than for population genetic studies of the copepod C. clitellata (Whitehead 2005) and several species of Mytilocypris ostracod (Finston 2002), the amount or pattern of dispersal in these invertebrates is unknown. Although it is unlikely that the amount of dispersal is enough to directly affect the population dynamics of a species, if the rate of local extinction is high, a species will need to have a 'reasonable' chance of dispersing into a suitable habitat in order to persist in the landscape.

Biological interactions
Information on biological interactions, such as predation and competition, in Australian salt lake communities is rudimentary (Timms 2021). This makes it difficult to assess the resilience of these systems to the loss of a species. Some waterbirds are known to prey on the invertebrates (Weston 2007;Pedler et al. 2018), although comprehensive data on the relationships between these predators and prey are lacking. It is possible that declines in the abundance and availability of the invertebrates (e.g. if salt lakes are dry for longer periods of time and are more saline) will affect the population dynamics of these birds (Senner et al. 2018). Fish predation may also be important at certain times in some waterbodies (Chessman and Williams 1987;Becker and Laurenson 2007), although fish are absent from most of the Australian salt lakes, which are typically either ephemeral or highly saline (De Deckker 1983b).
Competitive exclusion (Timms 2012a) may explain why congeneric species in a range of taxa (e.g. Parartemia, Australocypris and Coxiella) rarely co-occur in a single lake, although there is currently no evidence to support that competitive exclusion is more important than niche differentiation or dispersal limitation. The effect of the recent spread of A. parthenogenetica, and the potential (although not yet observed) spread of A. franciscana, on the native halophilic invertebrates is difficult to predict without detailed information on the ecological requirements of the native species (McMaster et al. 2007). However, it is worth noting that A. franciscana has outcompeted and replaced native Artemia species on other continents (Ruebhart et al. 2008).

Conclusion
A very diverse and unique range of halophilic invertebrates inhabit Australian salt lakes. These invertebrates are threatened by a range of processes, the most significant of which are increasing periods of extensive drought and secondary salinisation over large parts of the Australian landscape. Populations, and potentially species, have already become extinct (Williams 1995;Timms 2012b) and, without further study, more may be lost without notice (Halse et al. 2003). It is currently difficult to properly evaluate the consequences of the threats due to critical gaps in our knowledge of the biology of the invertebrates. Future studies of Australian halophilic invertebrates should prioritise the following: documenting the full extent of biodiversity in these invertebrates; this is the most critical gap to be filled because a sound taxonomy is needed to underpin all other biological research developing a better understanding of species' distributions and community composition, particularly in remote regions ascertaining species' tolerances to salinity and other physicochemical parameters across all stages of their lifecycle elucidating the long-term viability and hatching cues for desiccation-resistant eggs, and the equivalent for aestivating adults, and how these vary among species elucidating the abiotic and biotic factors that control species' distributions, including why congeneric species rarely occur in the same lake despite overlapping geographic and salinity distributions determining the patterns and mechanisms of dispersal and how these may influence the capacity of species to persist in the landscape elucidating the biological relationships between species, including how the spread of Artemia in Australian salt lakes will affect native species on a more holistic level, documenting the levels of anthropogenic stressors, including pollution and mining activities, on salt lake environments and identifying and preserving any key habitats that may serve as refugia for Australian halophilic invertebrates, particularly in the face of a drying climate.

Conflicts of interest
The authors declare that they have no conflicts of interest.

Declaration of funding
Angus Lawrie was supported by a Commonwealth Supported Research Training Program while conducting this research.