Several types of bone development are present in the Australian lungfish, Neoceratodus forsteri, the only extant member of the family Neoceratodontidae. In this species, dermal and parachondral bones form around the chondrocranium and mandible, to protect the brain and sense organs, to support the dentition and to facilitate oral function. The notochord persists throughout life. The quadrate and Meckel’s cartilage remain cartilaginous, as does the chondrocranium. Anterior elements of the hyoid arch, the basihyal and hypohyals, do not ossify. The ceratohyal, which articulates with the hypohyals, ossifies perichondrally, as do the exoccipital bone and the ribs of the trunk, including the cranial rib. The exoccipital bone is embedded in the chondrocranium in the adult fish, and the cranial rib is immobile. Some elements of the skeleton, such as the pectoral and pelvic fins, and the pelvic girdle, remain cartilaginous, and the skeletal elements in these fins resemble the structure of the tail. Fully developed bone is trabecular or lamellar, and does not include vascular elements. Endochondral ossification does not occur in the living Australian lungfish.

Across Australia’s monsoon tropics and vast arid zone isolated regions or ‘islands’ of upland or rocky habitat are home to disjunct populations of many taxa of plants and animals. Comparative analyses of lineages that occur across these habitat islands provide opportunities to understand when and how environmental change drove isolation and diversification across arid Australia. Here we present an analysis of mitochondrial genetic diversity across disjunct populations of geckos in the Nephrurus asper group. Dating analyses suggest that disjunct and genetically divergent populations spanning the northern half of Australia diverged through the Plio–Pleistocene. Based on the timing of divergence and current habitat associations we hypothesise that species in this lineage were isolated by the expansion of unsuitable arid-zone habitats from the late Pliocene onwards. Across most areas, these barriers appear to be sandy or stony deserts. However, in eastern Australia genetically divergent populations are separated by grassland on flat vertisol-dominated soils (‘blacksoils’), suggesting that these habitats also expanded during the late Pliocene aridification. Finally, we show that western Queensland populations formerly referred to N. asper are genetically divergent and diagnosable on the basis of colour pattern and, herein, recognise these populations as a distinct species.

The use of chemical pesticides to manage locust populations in natural ecosystems is likely to impact non-target arthropods and their predators. However, the relative effects of different locust control applications on Australian birds are unknown. Aerial applications of fipronil and fenitrothion are examples of two pesticides used in locust control in semiarid Australia. To test the relative impacts of pesticides on non-target fauna, pesticides were applied to replicate sites using aerial ultra-low-volume application methods. The body condition and biomarkers of pesticide exposure in resident white-winged fairy wrens (Malurus leucopterus leuconotus) at treatment and control sites were measured for two weeks before and after treatments. No measures suggested negative impacts of pesticide applications. However, birds monitored at treatment sites gained mass, possibly due to indirect impacts of pesticides on bird feeding patterns or the availability or behaviour of insect prey. Bird mass measures remained high at fipronil sites, whereas the mass of birds at fenitrothion sites returned to baseline levels within one week. As this study was conducted during dry conditions, when locust plagues are less likely, future insecticide research should also consider the availability of insect prey, its effect on insectivore feeding behaviour and the interaction of rainfall events.