Pacific Conservation Biology Pacific Conservation Biology Society
A journal dedicated to conservation and wildlife management in the Pacific region.
RESEARCH ARTICLE

Histological analysis of hatchlings of the Australian lungfish, Neoceratodus forsteri, from water impoundments reveals fundamental flaws in development

Anne Kemp
+ Author Affiliations
- Author Affiliations

School of Environment and the School of Biomolecular and Physical Sciences, Griffith University, 170 Kessels Road, Nathan, Qld 4111, Australia. Email: a.kemp@griffith.edu.au

Pacific Conservation Biology 23(2) 163-179 https://doi.org/10.1071/PC16036
Submitted: 1 September 2016  Accepted: 19 December 2016   Published: 30 January 2017

Abstract

Anomalies in embryos and hatchlings of the Australian lungfish are now found in many of the environments inhabited by lungfish, such as reservoirs (Lakes) created over natural rivers, and affect many tissues and organs in the body, most obviously the epidermis and related sense organs, but extending to blood vessels and skeletal structures as well. Development of muscular tissues is affected, but only by the lack of nutrition in the eggs as laid by the parents, and by failure of the hatchlings to feed. They are found in embryos collected from the Lakes and reared in the laboratory by standard methods, and are also present in embryos and hatchlings collected directly from the Lakes. They are not found in fish collected from unaltered parts of Rivers below the Lakes. Recently, in the Lakes, the anomalies are so widespread, and so serious, that they affect all the young of one season. The most probable cause of abnormal development is a lack of volatile fatty acids in the diet of adults, resulting in the production of poor quality eggs, as has been found in other species of fish. The results of this study have serious implications for survival of the species, since almost every habitat where lungfish are found, and are still spawning, has now been altered by the building of water impoundments. Restoration of freshwater environments in south-east Queensland should be a priority for the State and Federal Governments and for water authorities.

Additional keywords: deficient environments, threats to lungfish survival.

Introduction

Most river catchments in south-east Queensland are now affected by water impoundments, usually designated as Lakes, such as Lake Wivenhoe and Lake Samsonvale (Fig. 1). Populations of the Australian lungfish, Neoceratodus forsteri, originally living in rivers such as the Brisbane River, have been trapped in the Lakes, but have continued to spawn. However, abnormal development of eggs, embryos and hatchlings has recently been identified in specimens from these Lakes, using scanning electron microscopy (Kemp 2011, 2014a).


Fig. 1.  Map of south-east Queensland showing the site at Lowood on the Brisbane River, two sites in Lake Wivenhoe, part of the Brisbane River catchment, and one at the picnic ground at Lake Samsonvale on the Pine River, where eggs were collected. L, Lowood; L I, Logan’s Inlet; P, Paddy’s Gully; P G, picnic ground.
F1

Abnormalities found in the skin and associated sense organs have indicated that survival of the hatchlings in a reservoir environment is unlikely. Ciliated cells on the skin surface, essential for keeping the hatchling clean in the natural environment (Kemp 1996), are either absent or abnormal in structure and unable to function. Electroreceptors on the snout and mandible, and mechanoreceptors in the sensory lines of the head and trunk, lack essential components such as stereocilia, and as a result the hatchling will have difficulty in sensing food or avoiding danger. Even if they can find food, the dentition is poorly formed and this would have affected their ability to chew. Histological analysis reveals that the problems in Lake fish extend much further than deficiencies in skin sense organs and ciliation of the external skin. They affect nerves, blood vessels, lymphatics and complex sensory organs such as the olfactory system. Although a few hatchlings are able to feed and survive for several weeks, they die without developing fully, because of fundamental problems with the skin and sense organs.

The anomalies, which lead to the death of all hatchlings in any season, in the wild and in laboratory reared embryos, may be a result of poor nutrition in the adult lungfish inhabiting water impoundments (Kemp 2011, 2014a). Adult lungfish in the wild feed on snails and clams, their only source of volatile fatty acids. Snails and clams live in shallow water, and are not to be found in water impoundments with fluctuating water levels. The problems in the young fish may result from poor food for the adult lungfish in altered environments, most probably a lack of volatile fatty acids derived from their food (Kemp 2011). The adult lungfish are unable to produce good quality eggs. A similar problem has also been found in other species of fish, in hatchery environments (Peleteiro et al. 1995; Furuita et al. 2003; Fuiman and Falk 2013).

Several reasons have been suggested for this phenonomen. Lungfish are known to have low genetic diversity (Frentiu et al. 2001; Lissone et al. 2001; Lissone 2003), but limited genetic diversity does not provide a complete explanation for the high level of anomalies in young lungfish. It is unlikely to account for the observed numbers of dead hatchlings in the laboratory, and the probable lack of recruitment in the wild. The population of young eggs, embryos and hatchlings from the Brisbane River at Lowood, collected between 1986 and 1998, do not show any of the serious anomalies of the skin and sense organs found in the Lake hatchlings, although the genetic diversity of the parent fish from this area was limited as well. Anomalies found in River fish are rare and spontaneous, and usually affect skeletal structures in older hatchlings (Kemp 1994, 2003a).

Pollution is a feature of most reservoirs and rivers in south-east Queensland, but the pollutants vary from one place to another – pesticides in one area, such as Lake Wivenhoe, and herbicides in another, as in Lake Samsonvale (Fig. 1). The anomalies are always the same. Further, lungfish spawned in the Brisbane River (Fig. 1) below Lake Wivenhoe and exposed to herbicides and pesticides from agricultural activities along the shore close to the spawning sites, develop without the anomalies found in Lake fish, and survive in the laboratory. They are also recruited to the adult population. It is possible that the lungfish producing eggs in the reservoirs are too old to produce normal young, having been trapped in the water impoundments for many years. However, analysis of several lungfish from Lake Samsonvale and Lake Wivenhoe that died in a fish kill shows that these fish are not old, but had no food. The dentition of these fish is not worn, and the intestines are empty.

A series of hatchlings from Lake Wivenhoe, collected in 2009, and from Lake Samsonvale, collected in 2010 and 2012–15, have been analysed histologically to assess the extent of anomalies in the developing fish. The analysis was confined to histology and morphology because a preserved series of normal specimens was available from the Brisbane River to compare with specimens from Lake environments. Since spawning has ceased in the Brisbane River, a series of normal live embryos and hatchlings from the River was not available for physiological comparisons with Lake specimens. Structures found in the Lake hatchlings were compared with an equivalent series of hatchlings from Lowood on the Brisbane River, below Lake Wivenhoe, collected between 1986 and 1998.


Materials and methods

Eggs were collected from four sites, the lower Brisbane River at Lowood (1986–98), Lake Wivenhoe at Logan’s Inlet and Paddy’s Gully in 2009 and Lake Samsonvale in south-east Queensland between 2010 and 2015 (Fig. 1). Water quality in all sites was reasonable, and they are used as a water supply for the city of Brisbane. Since lungfish spawn in shallow water, variation in oxygen content is minimal.

Collection sites

Lowood on the Brisbane River

The Brisbane River below the town of Lowood in south-east Queensland (Fig. 1) was used by lungfish for spawning for many years, before the drought of 2001–08 caused alterations to the river. Eggs were laid on both sides of the river, among submerged plants or on the roots of trees. On one side, the river was fringed with established Callistemon viminalis trees, trailing long roots into the water, up to 50 cm deep, with a moderate flow. The site was in shade for most of the day. The other side, in sun most of the day, had some submerged rootlets as well as masses of algae in some years, and, occasionally, Nitella and water hyacinth. The submerged plants and rootlets provided refuges for hatchling lungfish, and small invertebrates, useful as food for young lungfish, could be found there. Basket clams, Corbicula australis, and snails, mostly Thiara balonnensis, were common at Lowood, and these formed the major part of the diet of the adult fish in this part of the river (Kemp 1986). Towards the end of the drought, enormous quantities of Urochloa mutica infested both sides of the river. During the floods of 2011, exacerbated by releases of water from Lake Wivenhoe in the two years after the flood, the spawning site was seriously damaged.

Collecting at Lowood began in 1982, and continued until spawning ceased in 2002 during the drought. The large number of normal fish reared from eggs collected in Lowood between 1986 and 1998 were used for this paper. Apart from rare cases of inability to break free of the egg case, anomalies in River fish are unusual (Kemp 1994, 2003a).

Logan’s Inlet in Lake Wivenhoe

The spawning areas at Logan’s Inlet (Fig. 1), within Lake Wivenhoe, provided a deficient environment, with clumps of dead and rotting grass, masses of filamentous algae, dead leaves and cattle faeces (Kemp 2011). The substrate consisted of coarse sand or gravel. Snails and basket clams were absent, but large numbers of carnivorous beetle larvae infested the shallows. Eggs were laid loose in shallow water, up to 50 cm deep, among strands of filamentous algae, tufts of dead grass, detritus, dead leaves and cattle faeces. These provided poor refuges for the eggs and hatchlings. Many eggs were subject to wind drift, which often carried them into water along the margin of the lake that was only a few centimetres deep, or caused them to aggregate in patches of submerged paddock grass and weeds. In 2009, spawning in this site was prolific, but in 2010 only a few dead eggs were collected. Most of the eggs that were laid in Logan’s Inlet had unusually fragile shells, and many embryos hatched too soon, before they had developed sufficient sensory systems or muscular activity to protect them from predators in the environment, or assist them to find places to hide (Kemp 2011).

Paddy’s Gully in Lake Wivenhoe

The spawning site at Paddy’s Gully (Fig. 1), close to the site of entry of the Brisbane River into Lake Wivenhoe, was cleaner and less contaminated than the site in Logan’s Inlet (Kemp 2014a). The substrate consisted of soft soil covered in dead leaves and cattle faeces. Some water hyacinth plants with short rootlets provided a little shade along the water edge, but eggs and hatchlings were otherwise unprotected. Water beetle larvae did not infest the area and filamentous algae were scarce. Eggs were not attached to the hyacinth rootlets, but were shed loose into the water. Eggs were collected from this site in 2009, but the spawning event was not repeated in 2010.

Picnic ground at Lake Samsonvale on the Pine River

Like other water impoundments in south-east Queensland, Lake Samsonvale on the Pine River near the Brisbane River catchment, has been opened for recreation, and the shores where lungfish spawn are adjacent to a popular picnic and fishing spot. The spawning site in Lake Samsonvale (Fig. 1) consists of a shallow beach below the picnic area. The substrate consists of coarse gravel or small stones, and short leaved plants of Vallisneria grow in patches, along the shore, at depths of 30–80 cm depending on the water levels in the reservoir. In some seasons, Nitella grew among the Vallisneria. Water in the Lake Samsonvale sites was clean, but weed cover was spasmodic at best, and the leaves of the submerged plants were always covered in short filaments of algae and fine sediment. The cyanobacterial species Microcystis aeruginosa, of a non-toxic strain (Orr, pers. comm.), was present along the shore line, usually dispersed in the water, but occasionally as a scum on the surface. Small fish, shrimps, a few snails and occasionally a basket clam lived among the submerged water plants, and predatory insect larvae were rare. Eggs were laid loose among the Vallisneria plants or on the substrate (Kemp 2014a). The exposed nature of the spawning site, and the lack of refuges for young hatchlings, resulted in heavy predation. The site can also be affected by strong winds, and this causes eggs to drift to the edge of the reservoir, where they are exposed to the sun, and die in large numbers.

The Lake Samsonvale site was used in the spring of 2010, but not in the following year after the disastrous flooding of 2011. In 2010 and in 2012, large numbers of eggs were laid, and some late stage embryos and three hatchlings were found in 2012. Spawning was less prolific in 2013 and 2014, late stage embryos were rare, and hatchlings not present. In 2015, numbers of eggs laid improved, and the embryos were healthier, although only three survived to a late hatchling stage in the laboratory, when they died.

Delayed development of Lake hatchlings

Young eggs collected from Lake and River environments and raised under identical circumstances pass through similar stages of cleavage, formation of a blastula and a neurula and development into an embryo and eventually a hatchling. However, rates of development differ, with Lake hatchlings lagging behind River hatchlings and taking up to three weeks longer to hatch and even longer to reach feeding stages. Anomalous development is evident early in development (Kemp 2011, 2014a) and increases in severity as the embryo approaches hatching.

River eggs take 30–32 days to develop to the stage of hatching. In Lake fish this process takes 55–60 days. River fish begin to feed six weeks after hatching, when most of the yolk in the endodermal cells is digested. For the first few weeks after hatching, the young fish hide amongst water plants, and move only when stimulated. This changes as they grow and become active. In the laboratory, they feed voraciously, quickly consuming all of the food provided, and swimming actively, particularly at night. Lake fish do not feed before they are eight weeks old, if they start to feed at all. They are also completely inactive, and move very little, even when stimulated.

Processing of material

Eggs, embryos and hatchlings were raised in the laboratory as described in Kemp (1981). Stages of development follow Kemp (1982, 1999).

Specimens were preserved at regular intervals throughout the period of the experiment, before and after hatching. Embryos and hatchlings were anaesthetised with clove oil before fixation in formalin. Specimens for scanning electron microscopy and for histology were fixed in 10% neutral buffered formalin and kept under refrigeration until they were required for analysis. Hatchlings intended for the Alcian blue/alizarin red technique (Kelly and Bryden 1983) were fixed in Carnoy’s fixative.

For scanning electron microscopy, specimens were removed from the egg membranes where required, sonicated in water for five minutes, and rinsed in clean water. Some of the specimens used for scanning electron microscopy were sectioned lengthwise with a razor blade to reveal internal details such as the structure of the developing olfactory openings and the gills. Specimens were then dehydrated in a graded series of alcohols before drying in a critical point drier. They were mounted on aluminium stubs and coated with gold before being examined and photographed in a scanning electron microscope (Jeol 6300).

To assess the histology of the developing skin and skin sense organs, two specimens from each Lake site were preserved in formalin, washed, dehydrated in a graded series of alcohols and embedded in Technovit resin, which was allowed to set at room temperature. Blocks were sectioned at 3 µm and sections expanded and dried onto clean glass slides before staining with 1% toluidine blue in phosphate buffered saline. Histology of the Lake hatchlings was compared with the histology of hatchlings of similar stages from the Brisbane River.

A series of 20 hatchlings from Lake Samsonvale, old enough for skeletal elements to have developed, were washed and treated with 0.5% potassium hydroxide for three days with the addition of 2% hydrogen peroxide to bleach pigment from the skin. Ten of the hatchlings were washed again in water, and dehydrated in alcohol. These specimens were stained in 0.01% Alcian blue in 96% alcohol acidified with acetic acid, following the methods of Kelly and Bryden (1983). Excess stain was removed by washing in fresh 96% alcohol and the hatchlings were rehydrated in distilled water and cleared in a solution of trypsin in saturated sodium tetraborate before being washed in water and preserved in glycerol. The other 10 hatchlings of similar stages were washed in water, treated with 0.5% potassium hydroxide and stained with Alizarin Red to delineate the bones and teeth. After staining they were washed, cleared in glycerol, and photographed.


Results

Numbers of affected specimens

The numbers of affected specimens from the Brisbane River and from the Lake sites are recorded in Table 1. Altered characteristics of Lake hatchlings are listed in Table 2. Numbers of specimens assessed using scanning electron microscopy are not always the same. The trait or sense organ may not be developed in young embryos, and not every specimen could be assessed for the trait, depending on the preparation.


Table 1.  Percentages of anomalies assessed during scanning electron microscopy
Numbers assessed vary. The trait or sense organ may not be developed in young embryos, and not every specimen could be assessed for the trait, depending on the preparation
Click to zoom


Table 2.  Characteristics of hatchling River and Lake fish
Click to zoom

No hatchlings from the River showed the characteristic of chronic ill thrift (Table 1). All fed well and developed appropriately. Nearly half of the specimens from Lake Wivenhoe that were old enough to assess, and over a quarter of the fish from Lake Samsonvale, showed ill thrift (Table 1). Trapping of an embryo, or inability to hatch, is uncommon in River fish, but often affects fish from Lake sites (Table 1). Oral anomalies, deficient skin cells and cilia, and malformed mechanoreceptors and electroreceptors were present in every single Lake fish examined, as were poorly developed gills, and blocked or absent vents (Table 1). None of these anomalies were to be found in any of the River fish examined. A few hatchlings from Lake Wivenhoe had pustular cells, and this trait was found in a quarter of the fish from Lake Samsonvale (Table 1). No fish from the Lowood site on the Brisbane River had any sign of pustular cells.

Skull bones in hatchlings from the Brisbane River are well ossified and in articulation (Table 2). Bones of the jaws are normal, with a correctly formed dentition. Cells on the external surface of the epidermis are all of similar size, with the appropriate surface structure. Ciliated cells are spaced regularly, with cilia of equal length. Up to four sections of lymphatic vessels are present in sections of the snout of young hatchlings, and the vessels are filled with lymphocytes. At least 58 electroreceptors are present in the skin of the snout and up to 10 in the mandible (Table 2).

Skull bones in hatchlings from Lake environments are poorly ossified and several bones are missing (Table 2). Bones of the jaws are present but affected by the dentition, usually lacking one or more ridges or with cusps of an abnormal shape. Cells on the skin surface vary in size, and many lack normal surface structure and are pustular. Ciliated cells, when present, carry abnormal and deformed cilia. Lymphatic vessels are few, and lymphocytes are sparse. Electroreceptors on both the snout and the mandible are significantly reduced in numbers (Table 2).

Chronic ill thrift

In River hatchlings, structures such as developing eyes, muscles and the endodermal mass are covered by an epidermis and a thick dermis, and the organs below the skin are not obvious (Fig. 2a). In Lake hatchlings of similar age, the epidermis is shrunken and the dermis is withered (Fig. 2b). The young Lake fish are emaciated in comparison with hatchlings from the River. Specimens collected directly from water impoundments are all thin, even the hatchlings that were old enough to find and ingest food.


Fig. 2.  Scanning electron micrographs of River and Lake specimens to compare morphology. (a) Head of a normal hatchling of Stage 50, from the Brisbane River at Lowood. (b) Head of a hatchling of Stage 50 from Paddy’s Gully, Lake Wivenhoe, showing wrinkled skin covering the underlying structures. Scale bar = 1 mm.
F2

The process of hatching and subsequent development

Inability to break out of the egg case is rare in River fish, and failures in the hatching process are usually the result of damage when the inner membranes of the egg break into large pieces and injure the embryo. In River fish, embryos begin to move while they are still enclosed in the egg membranes, and this causes the inner membranes of the egg to expand, allowing further movements (Kemp 1994). As the embryo grows, the inner membranes break up completely, and gaps appear in the outer albumen coat. Eventually the embryo can move in and out of the egg case, and finally shift away. In a small number of River fish this process does not occur, and the embryo is trapped, unable to hatch. Trapping is rare in River fish before the time of the long drought (2001–08), and is found in only 2.7% of River embryos (Kemp 1994).

Failure to hatch is common in Lake fish, because the embryos are too weak to move inside the egg case and the normal process of enlargement of the space within the shell does not occur. The inner membranes do not break down when they should and the embryo is unable to hatch (Fig. 3a). Some embryos die without being able to hatch (Fig. 3b), and the few that can emerge from the egg case after being enclosed for too long may be unable to move. Continued development is compromised.


Fig. 3.  Lake embryos affected by trapping in the egg case. (a) Embryo of Stage 35, trapped in the egg case, with fragments of the egg case adhering to the embryo. (b) Embryo of Stage 42 removed after being trapped in the egg case. The embryo had died before being released. Scale bars = 500 µm.
Click to zoom

Most Lake embryos that are able to pass through the hatching process normally are able to swim, but are less active than usual. Food animals such as black worms or small crustacea placed in the dishes and tanks are ignored by most of the hatchlings, and these eventually die. A few hatchlings begin to feed, and appear to be developing normally, even passing through the normal process of posture change. They ingest food and produce faeces, but they are sluggish. After a few weeks, they also die, despite reaching late hatchling stages, and looking apparently normal.

Development of the chondrocranium

In fish from the lower Brisbane River, and in specimens from both Lake Wivenhoe and Lake Samsonvale, the chondrocranium develops in the same way, and similar structures appear, up to the time the hatchlings from the water impoundments survive, unless the hatchlings are totally abnormal in appearance. In early stages, chondral structures are initiated as the parachordal and trabecular cartilages below the brain and in the snout, and as Meckel’s cartilage and the ceratohyals in the mandible. The quadrate has begun to form on either side of the parachondral cartilage, and is in articulation with Meckel’s cartilage. Traces of the otic capsule have appeared behind the quadrate. The otic quadrate process develops to link the quadrate and the otic capsule.

In slightly older River fish, the form of the juvenile chondrocranium is apparent, with rostral and lateral extensions from the parachondral cartilages, forming the pila antotica, and the development of the basal quadrate process (Fig. 4a). Dorsal and lateral trabecular horns grow from the trabeculae cranii and begin to encircle the olfactory organ, linked to the olfactory bulbs by the olfactory tract and by the olfactory nerve, which passes through the opening in the anterior chondrocranium (Fig. 4c). The otic capsules behind the basal quadrate process increase in size and in density. The lateral sphenotic foramen forms anterior to the otic capsule, and the optic foramen appears anterior to this large opening. Sclerotic cartilages develop a little later, to protect the eye, and the primordium of the ectethmoidal process extends from the ventral chondrocranium to support the tissues of the upper lip (Fig. 4c). Meckel’s cartilage in the mandible is united anteriorly. The hyoid arch is completed by the appearance of the paired hypohyals and the medial basihyal, and the gill arches develop (Fig. 4c).


Fig. 4.  Chondrocrania of hatchlings from the Brisbane River and from Lake Samsonvale. (a) River hatchling at Stage 50, lateral view. (b) Lake hatchling at Stage 50, lateral view. (c) River hatchling at Stage 51, dorsal view. (d) Lake hatchling at Stage 52, dorsal view. Structure of the chondrocranium appears similar in River and Lake specimens. Scale bars = 1 mm.
Click to zoom

Lake hatchlings follow a similar developmental process in formation of the chondrocranium (Fig. 4b), although it can be affected by the shape of the head in young fish that are abnormal. The ceratohyal and Meckel’s cartilage may be twisted and deformed, and this also happens in hatchlings from the River (Kemp 2003a). The gill arches may be reduced in length in Lake hatchlings and are poorly developed (Fig. 4b, d).

In older hatchlings from the River (Fig. 4c), and in the few surviving Lake fish (Fig. 4d), development of the lateral walls of the chondrocranium is advanced, with a large lateral sphenotic foramen from which nerves V and VII emerge. A small notch behind the quadrate marks the position of the jugular foramen for nerves IX and X. Development of the trabecular horns into capsules surrounding the olfactory organ and fusion with the pila antotica posteriorly completes the lateral chondrocranium. The olfactory nerve, and the minute nervus terminalis, leave the chondrocranium through the wide anterior foramen behind the developing olfactory capsules. The membranous labyrinth is now completely enclosed in cartilage. The components of the hyoid arch are present, but the Lake fish have poorly developed gill arches (Fig. 4d) and the basihyal may be missing. Meckel’s cartilage is a large element, with a shovel shaped anterior process supporting the lower lip. This element articulates with the quadrate distally. Otherwise, the chondrocranium has a normal appearance, until late in hatchling life (Fig. 4c, d), when the oldest of the fish from the Lakes died for reasons not associated with skeletal development.

Deformities and deficiencies in the chondrocranium of River fish occur in older hatchlings (Kemp 2003a). The anomalies usually involve the jaws and jaw articulations, and are not found in specimens from the Lakes because they do not survive for long enough. Up to 10% of River fish may show a specific problem, such as failure of the trabeculae cranii to fuse, affecting development of the olfactory cartilage, or malformation of the ceratohyal, which affects the suctorial function of the jaws, but the fish continue to feed and grow in captivity despite the anomalies.

Development of skull bones

A further series of whole hatchlings was also stained to reveal bone development in the head, and compared with an equivalent series of stained specimens from the Brisbane River (Fig. 5a). Although the skeletal and dental elements are not completely developed in the Lake fish, which all die in the laboratory at late hatchling stages or before, assessment of the formation of skeletal and dental structures up to that point is still possible (Fig. 5b).


Fig. 5.  Bones visible in hatchlings from the Brisbane River and from Lake Samsonvale. (a) River hatchling in ventral view at Stage 51. (b) Lake hatchling in ventral view at Stage 51. Arrows indicate calcified otoliths. The bones of the Lake hatchling lack ossification, except for the otoliths, which are heavily mineralised. Scale bars = 1 mm.
F5

Normal development of the bones of the lungfish skull is described in Kemp (1999). The first bones to appear in the head of River hatchlings are the parasphenoid in the palate, and the bones that support the tooth plates. The ascending process of the pterygopalatine bone appears slightly later, along with the dermal rostral bone, and the opercular and subopercular bones. The ceratohyal begins to ossify perichondrally as the dentition and bones of the jaws form. The hypohyal, basihyal and gill arches do not ossify, nor do the otoliths in the membranous labyrinth at early stages of development. Posterior bones of the skull roof, as well as the angular bone in the mandible, follow later. Other bones of the skull appear later in older hatchlings (Kemp 1999).

Development of some ossified elements follows a similar pattern in Lake hatchlings, while they survive (Table 2), but it lags behind bone development in River fish, and there is little calvarial development in late hatchlings (Fig. 5b). The bones that support tooth plates, including the transient dentary, with associated dental elements, are present. The parasphenoid, ceratohyal, the opercular, subopercular, and the ossified elements of the shoulder girdle are present as well but poorly calcified, and the calvarial bones are represented only by a trace of the rostral bone. In addition, there is malformation, enlargement and heavy ossification of the minute otoliths inside the membranous labyrinth, not found in River fish. The bones show no grossly obvious anomalies, except where the bone below the developing tooth plate is affected by absent or oversized cusps (Kemp 2014a, fig. 8d, f).

Olfactory openings in the oral cavity

The openings of the olfactory organ in lungfish are to be found within the oral cavity. In River fish, the anterior olfactory opening is wide, and placed just inside the upper lip. Numerous ciliated cells are arranged within the opening to direct a current of water into the olfactory organ (Fig. 6a, b). The posterior opening, situated to one side of the second ridge of the upper tooth plate, is a little narrower than the anterior opening, and also has many ciliated cells (Fig. 6a, b). Ciliated cells are still present inside the olfactory openings when they have disappeared from the lining of the oral cavity (Fig. 6b). Oral and olfactory deformities are not found in River fish (Table 1). In Lake fish, although the anterior opening is usually present, and in the correct place, it is often deformed and lacks ciliated cells (Fig. 6c, d). The posterior opening may be absent (Fig. 6c) or partially occluded (Fig. 6d) and has no ciliated cells. Oral and olfactory deformities are present in many of the Lake fish that are old enough to examine (Table 1).


Fig. 6.  The olfactory openings of hatchlings from the Brisbane River and from Lake Samsonvale. (a) One side of an upper jaw with normal olfactory openings, River hatchling, Stage 47. (b) One side of an upper jaw with normal olfactory openings, River hatchling, Stage 51. (c) Lake Samsonvale upper jaw, Stage 45, with absent posterior olfactory openings on both sides. The anterior openings (arrowed) are slits, and have no ciliated cells. (d) Upper jaw from a Lake Samsonvale hatchling, Stage 50, anterior opening with no ciliated cells, posterior olfactory opening (arrowed) slit-like and partially occluded. Scale bars = 100 µm.
Click to zoom

Gustatory organs are present in the oral endothelium, in both upper and lower jaws, and on the tongue, and are arranged in a regular manner, in hatchlings from the River. The endothelial cover of the cusps, before they erupt, is complete, with squamous cells of even sizes, as in the endothelium of the oral cavity. Gustatory organs in the oral cavity of Lake hatchlings are irregular in position if they are present at all, and the shapes of the endothelial cells lining the oral cavity are highly variable.

The skin, blood vessels and nerves of the snout

In River fish, the epidermis has several layers of cells, and the skin of the snout and lips is particularly thick (Fig. 7a). In section, most of the epithelial cells appear to be transporting cells. Melanocytes invade the epithelium in places. The epithelium is supported by the most superficial layer of the dermis, consisting of a dense continuous layer of collagen, with a layer of melanocytes below the collagen (Fig. 7c, e). The epithelium is ~100 μm thick over most of the head and body, and thicker over the mandible and snout. The skins of hatchlings from the River have numerous goblet cells among the cells of the epidermis and lining the oral cavity. Each goblet cell is long and thin, reaching from the base of the epidermis to the surface, and including material that stains for proteoglycan. The duct from the cell is even in diameter.


Fig. 7.  Sections from the snouts of River and Lake hatchlings, with electroreceptors, olfactory leaflets, nerves, capillaries and lymphatic vessels. (a) Snout skin of a River fish, lymphatics arrowed. (b) Snout skin of a hatchling from Lake Samsonvale, lymphatics arrowed. (c) Section of the olfactory organ and associated nerves, vessels and cartilage in a River hatchling, Stage 52. Lymphatics arrowed. (d) Section of the olfactory organ and associated vessels and nerves in a Lake hatchling, Stage 51. (e) Skin and associated vessels and nerves in a River hatchling, Stage 52. (f) Skin and associated vessels and nerves in a Lake hatchling, Stage 51. b, brain; c, cartilage; cp, capillary; e, electroreceptor; ed, duct of electroreceptor; olf, cavity of olfactory organ; on, olfactory nerve; V, branch of nerve V. Scale bars a, b = 100 µm; cf = 50 µm.
Click to zoom

In sections of skin from Lake hatchlings, some of the skin cells approach normal in appearance, but the thickness of the epithelium varies (Fig. 7b). Arrangement of the cells is not regular. The collagen layer is present in some places and absent in others. The dermis contains isolated connective tissue cells and fibrils, as well as scattered melanocytes, many away from their normal position below the collagen layer. The dermis consists of scattered cells, lymphatic vessels and capillaries containing red blood cells as well as the finer branches of nerves, and connective tissue fibrils, mostly collagen (Fig. 7b, d, f).

Hatchlings from the River develop a system of lymphatic vessels below the epidermis, which ultimately forms the tubule system, including the plexus of lymphatic vessels and the dermal papillae of the head in older fish (Kemp 2014b). Lymphatic vessels in young hatchlings are as wide as the lymphocytes they contain, running a tortuous course in the dermal tissues below the epidermis, among the bulbs of the electroreceptor organs and below the pigment cells and the layer of dermal collagen (Fig. 7a, c, e). They appear initially as the hatchling begins to search for food among water plants growing along the shore. Lymphatic vessels and the lymphocytes they contain increase in number as the hatchling grows, ultimately becoming enclosed in thick connective tissue in the dermis, and entering the epidermis of subadult and adult fish as loops of lymphatic capillaries (Kemp 2014b). In young hatchlings, lymphatic capillaries are confined to the dermis (Table 2). Occasionally, a separate vessel containing red blood cells, flat with a prominent bulge for the nucleus, is present among the dermal tissues.

Young hatchlings from Lake environments have few lymphatic capillaries below the epidermis (Fig. 7b, d, f). The vessels are only partially filled with lymphocytes, and they also contain some erythrocytes (Table 2). The hatchlings die before the system becomes fully formed and functional.

Tissues of the snout in River fish contain connective tissue fibrils as well as capillaries containing red blood cells, lymphatic vessels packed with lymphocytes, particularly prolific around the olfactory organ, and nerves. The olfactory nerve is visible in longitudinal section, emerging from the olfactory bulb above the forebrain and spreading over the sensory tissue of the olfactory organ, among lymphatic vessels (Fig. 7c). Leaflets of the olfactory organ are heavily ciliated. Several branches of nerve V passing towards sensory organs in the snout are visible below the cartilage of the olfactory capsule or among the connective tissue fibrils and blood vessels of the dermis (Fig. 7c, e).

In Lake fish, some connective tissue fibrils are present in the dermis, as well as blood vessels (Fig. 7d, f) but branches of nerve V are scanty, as are lymphatic vessels, even around the sensory tissue of the olfactory organ where they are plentiful in River fish (Fig. 7c). The olfactory nerve does not spread over the leaflets of the olfactory organ (Fig. 7d), but passes to one side of the sensory tissue. Leaflets within the olfactory organ have ciliated cells as usual, but the innervation of the olfactory organ appears to be deficient.

In early stages, the external surfaces of the epidermal cells of hatchlings from River environments are highly structured (Table 2). There is some surface modelling depending on their function as epithelial cells, goblet cells, sensory cells or ciliated cells. Most of the cells are polygonal in surface view and cell sizes fall within narrow limits, from 15 to 20 μm. The openings of goblet cells are smaller, ~5 μm in size. The distribution of ciliated cells is regular, and ciliated cells are larger than other epithelial cells (Fig. 8a). Lake hatchlings show a suite of anomalies on the skin surface, all likely to affect skin function (Table 2). Scanning electron micrographs show that the cells vary in size, and there is loss of normal surface modelling (Fig. 8b). Ciliated cells are not evenly distributed, and the cilia are deformed (Fig. 8b).


Fig. 8.  Scanning electron micrographs of skin cells from River and Lake hatchlings. (a) River hatchling, cells even in size, morphology of cilia uniform. (b) Lake hatchling, cells of different sizes, morphology of cilia variable. Scale bars = 20 µm.
F8

Anomalies of the skin and associated structures are universal in Lake hatchlings, and begin to appear in early stages (Table 1). They are unknown in River fish.

Mechanoreceptors and electroreceptors of the head

The skin has many sense organs, and these begin to develop soon after hatching. Electroreceptor (ampullary) pits originate in the epidermis, and remain close to the epidermal/dermal boundary, above the collagen layer of the dermis (Figs 7a, 9b). The mechanoreceptors, based on modified epidermal cells, open at intervals on the surface of the skin (Fig. 9a), later becoming enclosed in canals that sink below the surface of the epidermis, but always enclosed by a layer of dermal collagen. Sense organs in the skin of the snout are innervated by anterior branches of nerve V, which leaves the chondrocranium via the large lateral sphenotic foramen in the lateral pila antotica (Northcutt 1987).


Fig. 9.  Light micrographs of skin sense organs from River and Lake hatchlings. (a) Superficial mechanoreceptor from the snout of a Lake hatchling of Stage 50, showing cells differentiated for stereocilia or as supporting cells. (b) An electroreceptor from the snout skin of a River hatchling of Stage 52, with the sensory bulb under the epidermis and a convoluted channel passing to the external surface of the skin. (c) Superficial mechanoreceptor from the snout of a Lake hatchling of Stage 50 with no differentiation of the cells. (d) Two electroreceptors and a mechanoreceptor from the mandibular skin of a Lake hatchling of Stage 51. Scale bars a, d = 50 µm; b, c = 100 µm.
Click to zoom

In River hatchlings, mechanoreceptors have a cluster of columnar cells derived from the epithelium, with long stereocilia arising from the surface of each cell and extending out of the mechanoreceptor organ, into the canal or onto the surface of the epithelium (Fig. 9a). Modified epithelial cells, also columnar, surround and support the mass of sensory cells. Mechanoreceptors in older hatchlings are enclosed in deep canals that run in a regular pattern over the surface of the head, although the sense organs on the snout are usually superficial (Fig. 9a). The canal is composed of cuboidal epithelial cells supported by collagen, with pigment cells around the collagen fibrils.

Mechanoreceptors in Lake hatchlings are present in roughly correct positions over the surface of the head, and arranged in the usual lines, but with no trace of canals. The external structure of the mechanoreceptors is significantly affected by the structure of the surrounding cells. On the surface they appear to be swollen, with a large corona of cells, and a small orifice, usually with few or no emergent stereocilia. In section, they consist of a large bulb of cells at the base of the epithelium, without any regular structure (Fig. 9c, d). Inside the sense organs, differentiation into sensory cells bearing stereocilia, and columnar cells surrounding and supporting them, is absent (Fig. 9c, d). A small cavity may be present in the organ (Fig. 9d). In Lake hatchlings, all of the mechanoreceptors are abnormal (Table 1).

Electroreceptors are present in the epidermis and superficial areas of the dermis of River hatchlings, first appearing soon after hatching. On the snout and mandible they are single, and scattered. On the dorsal, lateral and ventral surfaces of the head they are arranged in pit lines (Kemp 1999). Rows of electroreceptors run above and below the lateral line of the trunk (Kemp 2012). Despite a close topographic association with the mechanoreceptors, electroreceptors are separate in structure and in function.

Active electroreceptors in the snouts of River hatchlings consist of a double layer of cuboidal cells that form a hollow ball, situated below the collagen of the epidermal/dermal junction or deeper, within the dermis (Figs 7a, 9b). Cells lining the bulb have minute projections from the surface, known as kinocilia (Joergensen 1984). A narrow duct, apparently spiral and lined by a single layer of epithelial cells, extends from the ball of cells to the surface of the epithelium, where it opens as a small circular pit surrounded by tiny papillae. Each electroreceptor contains a diffuse secretion consisting of proteoglycan, and dividing cells are often present.

In Lake hatchlings, electroreceptors are fewer in number than they are in River hatchlings (Table 2), with small bulbs, and wider than normal ducts (Figs 7b, 9d). Proteoglycan secretion within the organ is absent. The ducts retain the spiral arrangement, and their position in the dermis and epidermal/dermal boundary is unchanged. However, openings on the surface of the snout are much more variable in size, and usually wider than they are in River hatchlings.

Abnormal sense organs are found in every Lake hatchling old enough to have developed them, but are unknown in River hatchlings (Table 1).

Gills

In young hatchlings from the Brisbane River, gills are short projections from the branchial arch, covered in a ciliated epithelium (Fig. 10a) and enclosed within a cavity bounded by the operculum. Each gill has a core of cartilage and connective tissue and includes small muscles. As the gill grows, ciliated cells are reduced in number and lamellae appear on each side (Fig. 10b). Gills conforming to this description are found in every River fish examined for the structure of the gills (Table 1). None are abnormal. Gill filaments in Lake hatchlings have few cilia and often appear to be twisted. The epithelium is hyperplastic or even pustular and lamellae do not develop (Fig. 10c). The cartilage supporting the gill is deficient. Anomalies are found in the gills of every Lake hatchling examined for the trait (Table 1).


Fig. 10.  Scanning electron micrographs of gills and vents in River and Lake hatchlings. (a) Gills of a hatchling of Stage 44, River fish with ciliated cells. (b) Gills of a hatchling of Stage 50, River fish with lamellae and few ciliated cells. (c) Gills of a hatchling of Stage 52, Lake fish, deficiency in gill bar arrowed. The epithelium covering the gills is pustular and lacks ciliated cells. (d) The vent of a River hatchling, Stage 48, with developing pelvic fin, tail fin and anterior ventral fin fold. (e) The vent of a Lake hatchling, with hyperplasia and pustular cells in the surrounding skin. (f). The vent of a Lake hatchling, Stage 41, with hyperplastic tissue and multiple orifices, all narrow. Scale bars a, e = 20 µm; b, f = 50 µm; c, d = 100 µm.
Click to zoom

The lung is present as a diverticulum from the gut in young hatchlings but is not sufficiently developed to be functional for air breathing at early stages, so poorly formed gills have serious consequences for the Lake hatchlings. Young lungfish do not rise to the surface to breathe air until they are at least 40 cm long, and Lake hatchlings never reach this stage.

Vents

The rectum in River hatchlings is a straight tube, barely perceptible below the skin, and emerging from the intestine at a slightly obtuse angle. It opens as a small vent, with the tail fin behind and the preanal ventral fin anteriorly (Fig. 10d). The primordium of the pelvic fin is present anterior to the vent.

All of the Lake hatchlings at a similar stage have abnormal rectums and vents (Table 1). The rectum emerges from the posterior intestine at the usual angle, but is swollen (Fig. 10e, f). The vent in one specimen has multiple openings, all minute (Fig. 10e). The vents are often gaping, and surrounded by hyperplastic cells covered in pustules (Fig. 10f). In some Lake hatchlings, the vents are completely occluded or almost closed.

Pustular cells

Skin degeneration is marked in Lake hatchlings, particularly those from Lake Samsonvale. Many of the hatchlings have pustules (blebs) over the skin surface, particularly on the head (Fig. 11a). Pustules may be so numerous that they occlude skin sense organs, and deform cells that carry cilia on the surface (Fig. 11b). Pustular skin may cover the whole of the hatchling, or be confined to the head and dorsal surface of the trunk. However, most of the skin covering the rectum and the vent in Lake fish is pustular.


Fig. 11.  Scanning electron micrographs of pustular cells in Lake hatchlings. (a) Head of a Lake Samsonvale hatchling, severely affected by pustular cells. (b) Detail of pustular cells on the skin of the head, obliterating the mechanosensory organs and ciliated cells. (c) Pustular cells on the external epithelium and on the endothelium of the oral cavity. (d) Late stage in formation of pustular skin in a hatchling from Lake Samsonvale, culminating in total removal of epidermal cells, exposing collagen in places (arrowed). m, mechanoreceptor. Scale bars a = 1 cm; bd = 50 µm.
Click to zoom

The pustules consist of puffs of epithelial membrane with little discernible structure inside the cell (Fig. 11c). Pustules usually affect the flat epidermal cells, and obliterate the normal fine structure of the external surface of the cell. Within the protrusion there are suggestions of fibres. In some specimens ciliated cells may also be affected, fundamentally changing the arrangement and certainly the function of the ciliated cells (Fig. 11b). Pustular cells have nuclei, and a bleb of clear cytoplasm extending to the surface. It may affect only a few cells in the epithelium, or all of them. In severely affected hatchlings pustules extend to the oral endothelium (Fig. 11c). In some hatchlings the pustular cells are so numerous that they obliterate electroreceptors and surround the mechanoreceptors (Fig. 11b), already dysfunctional because of their altered internal structure.

Pustular cells represent the first sign of complete breakdown of epithelial tissue in hatchlings that survive long enough to develop the trait (Table 1). After pustules form, they grow and expand (Fig. 11c), eventually sloughing off, and leaving collagen exposed on the surface of the hatchling (Fig. 11d). Pustular skin cells are found in all of the altered environments in the water impoundments. Pustular skin is unknown in the River at Lowood before the drought, and common in Lake environments such as Lake Samsonvale.


Discussion

Lake hatchlings are affected by a problem well known in young animals and described as ‘chronic ill thrift’, defined as ‘failure to grow, increase in weight or maintain weight in the presence of apparently adequate food supplies and in the absence of recognisable disease’ (Saunders 2007). There is no doubt that Lake hatchlings are not finding food, eating or digesting it, but the ill thrift is present before the hatchling starts to feed, when it is still dependent on the supply of yolk laid down in the egg by the mother. It is also present in hatchlings that were able to feed, at least for a short time, before they died.

Ill thrift affects Lake hatchlings seriously. Even if the hatchlings start to feed they appear to derive little of nutritional value from their food. Development is slow, and hatching and the onset of feeding is delayed. The delays in hatching and feeding could also relate to lack of nutrition in the eggs, because the embryos are too weak to move and carry out the normal process of hatching. If they are able to break free of the egg case, they are often twisted and not able to swim normally. The muscular system is affected by the chronic ill thrift of the anomalous hatchlings.

The most obvious deficiencies in the Lake specimens concern the epidermis and the skin sense organs, as well as the lining of the oral cavity and the rectum. Initial analysis of the hatchlings using scanning electron microscopy indicated that skin sense organs are poorly formed and likely to be dysfunctional, and ciliated cells in the external epithelium and in the oral cavity are sparse, with deformed or absent cilia (Kemp 2011, 2014a). Histology of Lake hatchlings confirms the deficiencies in the skin sense organs, and indicates that the deformities of Lake hatchlings extend to blood vessels, lymphatic vessels, innervation, the olfactory organ, the gills, the rectum and vent, and bone mineralisation. It is no surprise that the few hatchlings that survive past the early posthatching stages die within two months.

Dermal and epidermal structures are important for the health of a young hatchling that shares its environment with innumerable small settling organisms and masses of particulate matter. A normal epidermis, as found in River hatchlings, has epidermal cells of even size, with many goblet cells that secrete mucus, and cells with cilia. Most of the epithelial cells are modified as a transporting epithelium. The dermis, in addition to mesenchyme cells and connective tissue fibrils, contains a system of lymphatic vessels that increase in density as the River hatchling grows, but this system is virtually absent in hatchlings from water impoundments. In Lake fish, poor development of the epidermis, including mucus glands, and missing or deformed cilia on the cells, results in ineffective epithelial function, affecting water transport, mucus secretion and clearing of the skin surface of settling organisms and infective agents. The same applies to the epidermal covering of the gills, with few ciliated cells and deficient structure in Lake fish. In addition, the gills lack properly formed gill arches and support structure within the gill filaments and no lamellae are formed.

Electroreceptors are small, confined to the superficial layers of the dermis or the epidermis (Joergensen 1984), and have a single innervation (Northcutt 1987). They are sensitive to weak electric currents in the water, possibly given off by prey animals (Roth 1973; Roth and Tscharntke 1976; Bullock et al. 1983; Northcutt 1986; Kramer 1996; Collin and Whitehead 2004). The commonest form of electroreceptor has a double layer of cells within the bulb, with traces of cell division, and is lined by sensory cells. The bulb enters a narrow and convoluted duct, occasionally closed with a valve at the surface of the epithelium. Lake hatchlings have fewer electroreceptors than River fish, with a smaller bulb and wider openings on the skin surface. They do not look grossly abnormal and may have some function.

Mechanoreceptors in Lake hatchlings follow the usual arrangement of sense organs in lines over the head (Kemp 1999), although they are always superficial and never enclosed in canals like the mechanoreceptors of older River hatchlings. The sensory tissue is poorly formed, with sensory cells that have no stereocilia, or deformed stereocilia, and there is less differentiation of the supporting cells. There is no mound of differentiated cells surrounding the opening and the sense organ is often severely affected by anomalies in the surrounding skin. Mechanoreceptors are unlikely to be functional in Lake fish.

Vents in Lake fish are always abnormal and have hyperplastic cells. The opening can be closed or almost occluded, or may be too wide. Pustular cells are common in Lake fish, although the cause is not known, and are a sign of skin degeneration. A large number of the hatchlings from Lake Samsonvale have pustules in the skin cells, sometimes so numerous that they obscure skin sense organs, particularly the mechanosensory cells, already dysfunctional because of their altered structure. In severely affected hatchlings, pustular skin on the head extends into the oral epithelium as well.

The chondrocranium of Lake hatchlings is affected less than dermal and epidermal structures, sense organs or blood vessels. The structure is basically normal, in the few hatchlings that were able to develop to late stages, and have a normal head shape, although the shapes of the component parts may be distorted. If the hatchling is deformed so is the chondrocranium. Anomalies in Meckel’s cartilage and in the ceratohyal and jaw articulation, which occur in River hatchlings as well (Kemp 2003a), are also found in the few older Lake hatchlings that have been assessed. These malformations are not immediately life threatening, unlike the universal problems associated with the skin and skin sense organs found in Lake fish.

Skull roofing bones are usually late to develop in Lake hatchlings, and when they are present are poorly mineralised. The dentition, which has a dual endothelial/mesenchymal origin (Kemp 2003b), does not develop normally, with deformed tooth plates, absent or enlarged cusps, and a lack of endothelial cover affecting the formation of the enamel. Prearticular and pterygopalatine bones are affected by abnormalities in the dentition, but the posterior processes of the bones that support the jaw articulation are not affected. The parasphenoid, which also plays a part in support of the jaw articulation, is present but poorly ossified. Perichondral ossification in the hyoid appears normal, but perichondral elements elsewhere in the head are absent. Otoliths in the Lake fish are heavily mineralised, and abnormal in shape. They are oblong, with fenestrations, when they should be flat and circular (Retzius 1881). Mineralisation of otoliths in River fish is not precocious, and they do not show in stained preparations of hatchlings.

Olfactory openings in Lake hatchlings are not normal. The anterior opening is an elongate slit, when it should be round. The posterior opening is missing or occluded, although it should also be rounded. Normal olfactory openings from River fish are heavily ciliated, but few ciliated cells are present in the openings of Lake hatchlings. Formation of leaflets within the olfactory organ appears normal, with stereocilia present as they should be, but the innervation is not. In transverse sections of River fish, the olfactory nerve spreads over the surface of the olfactory tissue, among the lymphatic vessels that surround it (Greil 1913). Olfactory nerves in the Lake fish do not spread over the organ but slide down one side. Anomalies in the external openings of the olfactory organs, and in the unusual ossification of the hatchling otoliths, suggest that neural structures may be affected also by poor development. Numbers of branches of nerve V, which innervate the anterior skin sense organs, are also deficient, with two or three instead of eight visible in sections of the snout.

Capillaries are present in the snouts of Lake fish, but lymphatics are deficient when compared with hatchlings from the River. There are few lymphatics in the dermis of Lake fish, and the vessels are often almost empty. Lymphatics provide important protection against infection in lungfish (Kemp 2014b).

Eggs, embryos and hatchlings from water impoundments in south-east Queensland, which comprise the majority of current lungfish habitats, do not develop normally and few, if any, develop into juveniles and are recruited into the adult population (Kemp 2011, 2014a). Some die because they are grossly abnormal, and others die because they are unable to hatch. None of the hatchlings survive for long, and most fail to react to the presence of food or attempt to eat any. The few hatchlings that survive and feed have obvious signs of chronic ill thrift (Saunders 2007) and die within a few weeks. Scanning electron microscopy reveals that the skin ciliation is so poor that the normal cleaning function of the ciliated epithelium in inoperable, exposing the hatchling to infection (Kemp 2011, 2014a). In addition, the skin sense organs of the hatchlings are so deformed that the hatchlings will be unable to sense the presence of food. It is likely that the problems encountered in the laboratory are reflected in the natural environment, because the small number of late stage embryos and hatchlings found in the field are as abnormal as the young reared in the laboratory. This suggests that recruitment in most water impoundments in south-east Queensland since 2009 may have failed completely.

Efforts are currently underway to restore and improve the riverine environments inhabited by lungfish, mostly badly damaged by the long drought of 2001–08 and by subsequent severe flooding (2009–13), affecting all of the rivers of south-east Queensland (Foster, SEQwater, pers. comm.). Landowners who lost soil from the banks of the rivers during the outflow of floodwater from the reservoirs are being encouraged to plant trees along the river, and Creek Catchment groups are monitoring the plants and animals in the water. Unfortunately, these efforts do not yet extend to the creation of protected areas of the rivers where lungfish are found. Riverine environments are under continual threat from the existence of water impoundments, and the management of these reservoirs is such that the environments cannot be improved because of fluctuating levels when water is required for use. Refuges and food for young lungfish, and food for adults, are lost every time the water level drops in the water impoundment.

The loss of suitable environments for native animals such as the lungfish has reached critical levels. The current human population of south-east Queensland, and the needs of agriculture and industry for water, means that water impoundments are unfortunately essential for the future. However, it should be possible to preserve the unaltered parts of the rivers below or above the water impoundments as aquatic reserves, and prevent, as far as possible, continued degradation in the environment where endangered species live. A fence could be built along the banks of the river to prevent access by large destructive animals such as cattle, and people could still be allowed to use the river for swimming or kayaking. The submerged plants could be protected, or reintroduced if they are absent, and trees could be restored to the banks of the rivers. Small animals on which lungfish depend for food would find shelter among the submerged plants, and these would also provide refuges for hatchling lungfish. Although such a reserve is not a complete solution to the problem of floods and droughts, creation of a protected area along the banks of a river should be possible because Government authorities have legal control of a wide strip of land on either side of the river. Clearly, conservation of the threatened Australian lungfish will depend on the success of efforts to improve the environment in south-east Queensland, and this should be given priority.




Acknowledgements

Thanks are due to staff in CARF at the Queensland University of Technology, Gardens Point, Brisbane, for advice and help with scanning electron microscopy, and to friends and family who assisted with the collection of eggs. Scanning electron microscopy was partially funded by a grant from the Linnean Society of New South Wales. This work was carried out with the permission of the Queensland Government Department of Primary Industries and Fisheries, permit no. 160372, and the Animal Ethics Committee of Queensland University, permit no. CMS/350/09, and Griffith University, permit no. ENV/03/11/AEC.


References

Bullock, T. H., Bodznick, D. A., and Northcutt, R. G. (1983). The phylogenetic distribution of electroreception: evidence for convergent evolution of a primitive vertebrate sense modality. Brain Research. Brain Research Reviews 6, 25–46.
The phylogenetic distribution of electroreception: evidence for convergent evolution of a primitive vertebrate sense modality.CrossRef |

Collin, S. P., and Whitehead, D. (2004). The functional roles of passive electroreception in non-electric fishes. Animal Biology 54, 1–25.
The functional roles of passive electroreception in non-electric fishes.CrossRef |

Frentiu, F. D., Ovenden, J. R., and Street, R. (2001). Australian lungfish (Neoceratodus forsteri: Dipnoi) have low genetic variation at allozyme and mitochondrial DNA loci: a conservation alert? Conservation Genetics 2, 63–67.
Australian lungfish (Neoceratodus forsteri: Dipnoi) have low genetic variation at allozyme and mitochondrial DNA loci: a conservation alert?CrossRef | 1:CAS:528:DC%2BD3MXmsl2gu7c%3D&md5=f554c2772623ce447ba74dffcf605310CAS |

Fuiman, L. A., and Falk, C. K. (2013). Batch spawning facilitates transfer of an essential nutrient from diet to eggs in a marine fish. Biology Letters 9, 20130593.
Batch spawning facilitates transfer of an essential nutrient from diet to eggs in a marine fish.CrossRef |

Furuita, H., Ohta, H., Unuma, T., Tanaka, H., Kagawa, H., Suzuki, N., and Yamamoto, N. (2003). Biochemical composition of eggs in relation to egg quality in the Japanese eel, Anguilla japonica. Fish Physiology and Biochemistry 29, 37–46.
Biochemical composition of eggs in relation to egg quality in the Japanese eel, Anguilla japonica.CrossRef | 1:CAS:528:DC%2BD2cXlvFCgurw%3D&md5=85ff1bd4092da607297fcc4e2ad173b5CAS |

Greil, A. (1913). Entwickelungsgeschichte des Kopfes und des Blutgefässsystems von Ceratodus forsteri. II. Die epigenetischen Erwerbungen während der Stadien 39–48. Denkschriften der Medicinisch-Naturwissenschaftlichen Gesellschaft zu Jena 4, 935–1492.

Joergensen, J. M. (1984). On the morphology of the electroreceptors of the two lungfish Neoceratodus forsteri Krefft and Protopterus annectens Owen. Videnskabelige Meddelelser Dansk Naturhistorisk Forening 145, 77–86.

Kelly, W. L., and Bryden, M. (1983). A modified differential stain for cartilage and bone in whole mount preparations of mammalian foetuses and small vertebrates. Stain Technology 58, 131–134.
A modified differential stain for cartilage and bone in whole mount preparations of mammalian foetuses and small vertebrates.CrossRef | 1:CAS:528:DyaL3sXlsFClt7g%3D&md5=61c33d037051cacfe75aff77d7f9c187CAS |

Kemp, A. (1981). Rearing of embryos and larvae of the Australian lungfish, Neoceratodus forsteri (Krefft) under laboratory conditions. Copeia 1981, 776–784.
Rearing of embryos and larvae of the Australian lungfish, Neoceratodus forsteri (Krefft) under laboratory conditions.CrossRef |

Kemp, A. (1982). The embryological development of the Queensland lungfish, Neoceratodus forsteri (Krefft). Memoirs of the Queensland Museum 20, 553–597.

Kemp, A. (1986). The Biology of the Australian lungfish, Neoceratodus forsteri (Krefft 1870). Journal of Morphology 190, 181–198.
The Biology of the Australian lungfish, Neoceratodus forsteri (Krefft 1870).CrossRef |

Kemp, A. (1994). Pathology in eggs and embryos of Neoceratodus forsteri (Osteichthyes: Dipnoi). Copeia 1994, 935–943.
Pathology in eggs and embryos of Neoceratodus forsteri (Osteichthyes: Dipnoi).CrossRef |

Kemp, A. (1996). Role of epidermal cilia in development of the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi). Journal of Morphology 228, 203–221.
Role of epidermal cilia in development of the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi).CrossRef |

Kemp, A. (1999). Ontogeny of the skull of the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi). Journal of Zoology 248, 97–137.
Ontogeny of the skull of the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi).CrossRef |

Kemp, A. (2003a). Anomalies in the developing neural and visceral head skeleton of the Australian lungfish, Neoceratodus forsteri. Annals of Anatomy 185, 121–134.
Anomalies in the developing neural and visceral head skeleton of the Australian lungfish, Neoceratodus forsteri.CrossRef |

Kemp, A. (2003b). Ultrastructure of developing tooth plates in the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi). Tissue & Cell 35, 401–426.
Ultrastructure of developing tooth plates in the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi).CrossRef |

Kemp, A. (2011). Comparison of embryological development in the threatened Australian lungfish, Neoceratodus forsteri, from two sites in a Queensland river system. Endangered Species Research 15, 87–101.
Comparison of embryological development in the threatened Australian lungfish, Neoceratodus forsteri, from two sites in a Queensland river system.CrossRef |

Kemp, A. (2012). Formation and structure of scales in the Australian lungfish, Neoceratodus forsteri. Journal of Morphology 273, 530–540.
Formation and structure of scales in the Australian lungfish, Neoceratodus forsteri.CrossRef |

Kemp, A. (2014a). Abnormal development in embryos and hatchlings of the Australian lungfish, Neoceratodus forsteri, from two reservoirs in southeast Queensland. Australian Journal of Zoology 62, 63–79.
Abnormal development in embryos and hatchlings of the Australian lungfish, Neoceratodus forsteri, from two reservoirs in southeast Queensland.CrossRef |

Kemp, A. (2014b). Skin structure in the snout of the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi). Tissue & Cell 46, 397–408.
Skin structure in the snout of the Australian lungfish, Neoceratodus forsteri (Osteichthyes: Dipnoi).CrossRef | 1:STN:280:DC%2BC2M%2FmslCgsQ%3D%3D&md5=1ba7499cc8e4a90594e6e56eb6884f59CAS |

Kramer, B. (1996). Electroreception and communication in fishes. (Ed. W. Rathmayer.) Progress in Zoology 42, 1–119. (Gustav Fischer: Stuttgart.)

Lissone, I. (2003). Conservation genetics and the Australian lungfish Neoceratodus forsteri (Krefft 1870); a spatio-temporal study of population structure. M.Sc. Thesis, University of the Sunshine Coast, Sippy Downs, Queensland.

Lissone, I., Shapcott, A., and Ovenden, J. (2001). The use of RAFs enables determination of genetic structure within and among catchments in the Australian lungfish Neoceratodus forsteri. Abstract, Genetics Society of Australia, Melbourne.

Northcutt, R. G. (1986). Electroreception in non-teleost bony fishes. In ‘Electroreception’. (Eds T. H. Bullock and W. Heiligenberg.) Chapter 9, pp. 257–285. (John Wiley and Sons: New York.)

Northcutt, R. G. (1987). Lungfish neural characters and their bearing on sarcopterygian phylogeny. Journal of Morphology 1, 277–297.

Peleteiro, J. B., Lavens, P., Rodriguez-Ojea, G., and Inglesias, J. (1995). Relationship between egg quality and fatty acid content of various turbot broodstocks (Scophthalmus maximus L.). ICES Marine Science Symposia 201, 51–56.

Retzius, G. (1881). ‘Das Gehörorgan der Wirbelthiere. Morphologisch-histologische Studien. I. Das Gehörorgan der Fische und Amphibien.’ (Centraldruckerei: Stockholm.)

Roth, A. (1973). Electroreceptors in Brachiopterygii and Dipnoi. Naturwissenschaften 60, 106.
Electroreceptors in Brachiopterygii and Dipnoi.CrossRef | 1:STN:280:DyaE3s7ktl2gsQ%3D%3D&md5=b04bbe7397491990e20866e30a663834CAS |

Roth, A., and Tscharntke, H. (1976). Ultrastructure of the ampullary electroreceptors in lungfish and Brachiopterygii. Cell and Tissue Research 173, 95–108.
Ultrastructure of the ampullary electroreceptors in lungfish and Brachiopterygii.CrossRef | 1:STN:280:DyaE2s%2FkslSqtQ%3D%3D&md5=77dbc5b8ea0130984597fa6422932059CAS |

Saunders (2007). Saunders Comprehensive Veterinary Dictionary. 3rd edn. Available at: http://medical-dictionary.thefreedictionary.com/ [accessed 10 August 2016].


Abstract PDF (2.1 MB) Export Citation