The Southern Angle-headed Dragon (Lophosaurus spinipes): a systematic review of the literature

Search strategy

Five databases (University of New England library ‘PRIMO’ search, Academic Search Complete, GreenFILE, Scopus, and Web of Science) and Google Scholar were searched for peer reviewed, non-peer reviewed and grey literature published up to November 2023, and supplemented by handsearching and expert input. Search terms were developed using PIO (Population, Interest/Intervention, Outcome), a modified version of the PICO (Population, Interest/Intervention, Comparison and Outcome) framework (Richardson et al. 1995) that is used when there is no comparator. PIO and PICO are useful tools in evidence-based research for providing a framework to structure the research question and facilitate search terms. To capture all variations, alternative terms were combined with the Boolean operand ‘OR’ (Table 1); these searches then were combined using the ‘AND’ operand (Table 1).

The above search yielded only two results. The search was then broadened by removing the ‘Interest’ and ‘Outcome’ search-strings one at a time to retrieve more results, but again few results were retrieved (seven and eight results respectively). Due to the limited results obtained, the decision was then made to conduct a very broad search using only the ‘Population’ search terms (Table 2).

Inclusion/exclusion criteria

All peer reviewed, non-peer reviewed and grey literature (except commentaries, letters to the editor, and field guides) that mentioned Lophosaurus spinipes (or its previous classifications of Gonocephalus spinipes Kaup 1825 or Hypsilurus spinipes Peters, 1867 were included. This included mentions of their single Australian congener, Lophosaurus boydii, when there were data relevant to similarities in biology or behaviour. Publications needed to be available in full text.

Publications were excluded if they were commentaries, letters to the editor, field guides, or if the full text was not available.

Selection of studies

The process of selecting studies and the extraction of data were conducted according to the 2020 PRISMA framework (Page et al. 2021) (Fig. 2). The results were downloaded to Microsoft Excel for review and duplicates manually removed. The first and second authors independently screened the titles and abstracts against the criteria for inclusion/exclusion. Results were compared and disagreements were settled by discussion and consensus. The reviewers then screened the full text of the remaining articles in parallel. Reasons for exclusion were documented at each stage. The lead author then searched the reference lists of the included publications to find any that may not have been captured in the search, and expert input was also sought.

Fig. 2.

PRISMA flowchart of the search criteria and selection process.

Synthesis of data

Using a standardised form for extraction, the lead author extracted the following from all publications: title, author, date, aims, type of study, methods, results, and conclusions. Meta-analysis was not possible due to the heterogeneity of the types of study. Data were analysed narratively into themes using Braun and Clarke’s (2006) six-step thematic analysis framework, which is a popular framework to help identify patterns (or themes) in data sets. Appraisal of quality was not conducted on the included articles due to the limited results.

Biology

Nine publications (50%) were categorised in the biology theme. These included publications on behaviour (Webb 1984; Manning and Ehmann 1991; Rummery et al. 1995; Torr 2003), threat displays (Manning and Ehmann 1991), diet (Webb 1984; Manning 1991, 1992; Traynor and Houston 2018), nesting behaviours and duration of incubation (Manning 1991; Shea et al. 1991; Harlow 2001), and thermoregulation (Manning 1991; Manning and Ehmann 1991; Rummery et al. 1995; Torr 2003).

The detection and capture of L. spinipes is difficult throughout the year due to its effective camouflage and arboreal lifestyle (Shea et al. 1991). Most sightings occur during spring and early summer (late November and early December) when gravid females are seeking sites for nesting along roadside verges (Shea et al. 1991; Manning 1991; Manning and Ehmann 1991; Rummery et al. 1995).

Webb (1984) originally suggested an omnivorous diet for L. spinipes. However, observational studies (Manning and Ehmann 1991; Manning 1992), trials offering various sources of foods (Manning 1992), and direct inspection of the gut contents of museum specimens (Traynor and Houston 2018), found Insecta (insects), Arachnida (spiders), and unidentified unsegmented roundworms assumed to be gut-parasites. These confirmed the species as a predator of invertebrates. Although Chilopoda (centipedes) and Diplopoda (millipedes) were not observed in any of these studies of L. spinipes, both taxa were found in the guts of the related species L. boydii and L. dilophus. Most soft-bodied prey are usually broken down very rapidly, thereby making them indistinguishable in the gut contents, but Clitellata (earthworms) and Gastropoda (slugs and snails) have been identified in the field being ingested in some cases (Webb 1984; Manning and Ehmann 1991). Many species of ants, or their parts, have been noted in most guts of Lophosaurus species (Traynor and Houston 2018).

Manning (1991) and Shea et al. (1991) observed nesting behaviours in the field, where the females were seen digging test nests along the dirt roads in the area. Several days later the females were seen digging shallow (2–4 cm) nests in which they proceeded to deposit the eggs. Upon completion of deposition, they covered (or partially covered) the eggs with soil. The postoviposition duration of incubation has been reported with variable results: 74–75 days in situ (N = 1 nest) (Manning 1991), and between 68 and 125 days in the laboratory (temperature-dependent) (Harlow 2001). Harlow (2001) trialled a range of temperatures on developing eggs, and found incubation to last 125 days at 20°C, about 68 days at 26°C, and mortality increased substantially when incubated at a constant 27°C.

Observational studies on the movements of L. spinipes using the spool tracking technique indicated that they move in short bursts, over short distances, relying on crypsis as their defensive strategy (Manning and Ehmann 1991; Manning 1992). An experimental study showed that males were territorial with clear separation of their defended areas, with territories of between 300 m² and 700 m² (Manning and Ehmann 1991). Males frequently move over the ground rather than arboreally, whereas females move substantially shorter distances on the ground, and tend to remain on the abundant saplings and vines of different species within the forest (Manning and Ehmann 1991). Females show no preference for particular species of tree or vine (Rummery et al. 1995). Females are not harassed when crossing males’ territories, but males are actively attacked or threatened upon entering another male’s territory (Manning and Ehmann 1991; Rummery et al. 1995).

Males and females tend to perch above the ground at an average of 2.6 m and 3 m respectively (Manning and Ehmann 1991). The diameters of perches were variable: Manning and Ehmann (1991) noted a range of 20–60 mm, whereas Rummery et al. (1995) indicated <20 mm even though a greater range of perch-diameters was available. Both studies suggested that the perch was chosen to provide a better grip for climbing, to increase camouflage, and to widen the animal’s field of vision (Manning and Ehmann 1991; Rummery et al. 1995).

Lophosaurus spinipes was originally thought to exhibit some heliothermy (that is, utilising sunlight as a source of heat), as apparent basking was observed (Manning 1991; Manning and Ehmann 1991). However, a later study demonstrated that this species does not actively shuttle to maintain a precise range of temperature, and the species is considered to be a thermoconformer (Rummery et al. 1995). Its habitat of closed-forest offers fewer basking sites than those of many other agamids, and constitutes an area of prohibitive costs either for heliothermy or thigmothermy (that is, absorbing heat from the substrate) (Rummery et al. 1995). It was noted that females may increase their thermoregulatory effort during gravidity (Rummery et al. 1995), but this topic requires further investigation. In contrast to most other Australian agamids, L. spinipes is active over a wider range of temperatures: 11−26°C (Rummery et al. 1995); 17.8−25.1°C (Manning and Ehmann 1991). Although this species is reported to have a broad set-point range, it was seen to actively avoid temperatures outside of that range whenever and wherever possible (Rummery et al. 1995). The lower set-points and temperatures that the lizard voluntarily accepts during its activity period can be compared with those of most other agamids (Manning and Ehmann 1991; Rummery et al. 1995) that inhabit more arid and open environments. It was suggested that choice of microhabitat (the height and diameter of perches and the openness of their location), may result in an imprecise level of thermoregulation (Rummery et al. 1995). It is important to note that the study by Rummery et al (1995) was carried out late in the seasonal activity cycle of L. spinipes (April of 1992), and may not have entirely captured temporal changes in thermoregulatory effort.

Reproduction

Four publications discussed reproductive processes of related species of dragons. Three discussed observations and anecdotes (Webb 1984; Manning 1991; Shea et al. 1991), and one was an experimental study (Harlow 2001).

Observations in the field of L. spinipes’ on nest site selection, nesting behaviours, clutch sizes, and hatching success, in two different latitudinal areas of New South Wales, Australia, appeared in the early 1990s (Manning 1991; Shea et al. 1991). Lophosaurus spinipes selects more open areas such as roads and tracks for oviposition in very shallow nests (Manning 1991; Shea et al. 1991; Harlow 2001), making them more easily detected by predators such as brush turkeys (Alectura lathami) and lace monitors (Varanus varius) (Webb 1984; Manning 1991). An experimental study (Harlow 2001) recorded the fates of embryos of L. spinipes incubated at different temperatures, and showed that mortality of developing embryos during incubation at temperatures above 26°C was greater than for those incubated at lower temperatures.

Phylogeny and taxonomy

Five publications discussed the phylogeny and taxonomy of L. spinipes (Ota et al. 1992; Manthey and Denzer 2006; Hugall et al. 2008; Pyron et al. 2013; Denzer and Manthey 2016). The classification of the genus Lophosaurus and its phylogenetic placement have changed considerably over the years, with some revisions based solely on morphology, others on both morphology and molecular analysis, and some on molecular analysis alone (Manthey and Denzer 2006; Hugall et al. 2008; Pyron et al. 2013; Denzer and Manthey 2016). A comparison of the karyology of two Bornean species of Gonocephalus with that of Australian Gonocephalus (Lophosaurus) spinipes, demonstrated a clear difference and a possible zoogeographic dichotomy between Asian and Australian species (Ota et al. 1992). This finding would seem to add further weight to the argument for an early, endemic agamid radiation in Australia (Ota et al. 1992).

The inherent interrelatedness of three Hypsilurus (Lophosaurus) species from Australia and New Guinea (dilophus, boydii, and spinipes), along with the fact that they demonstrated few similarities with any other members of the genus, led to their placement within the dilophus subgroup (Manthey and Denzer 2006). A dichotomous key for the genus was developed. The rise of sequencing and analysis of DNA created incongruities within the morphologically-based phylogenies (Pyron et al. 2013). To reduce these conflicts, additional groupings and subcategories were added to taxonomic nomenclature. The use of both morphological and molecular data to revise the classification of squamate reptiles resulted in a new, large-scale phylogenetic revision (Pyron et al. 2013). Although reviews of phylogeny had been previously conducted at many taxonomic levels, a large-scale phylogeny had not been attempted. The familial and subfamilial classifications within squamate reptiles were reviewed and altered to match the newly presented phylogeny; Hypsilurus was retained, and the Australian–New Guinean species remained in the dilophus grouping (Pyron et al. 2013).

In 2008, the Australasian members of Hypsilurus (Lophosaurus) were once again revised as the newer molecular phylogenies were still found to be highly incongruent with the morphological scheme (Hugall et al. 2008). This revision, based upon mitochondrial deoxyribonucleic acid (mtDNA) sequences, provided acceptance for the molecular phylogeny, and added support to the idea that lizards from mesic-forest came onto the Australian plate from the north during the Miocene epoch. This theory has since been rescinded and the idea of an Australian radiation moving south to north has been adopted (Pyron et al. 2013; Denzer and Manthey 2016). The resultant phylogeny places the genera Physignathus (Intellagama) and Hypsilurus (Lophosaurus) into a “… paraphyletic basal assemblage to the more derived Australian forms such as Amphibolurus and Ctenophorus …” (Hugall et al. 2008, p. 343). Finally, the taxonomy and nomenclature of Hypsilurus was revisited after an examination of voucher specimens’ sequences held on GenBank (Denzer and Manthey 2016). Some confusion was noted in assignation of sequences to voucher specimens and with respect to some of the phylogenies. The resultant outcome was to place the Australian and New Guinean species (formerly in the dilophus group) into the resurrected genus Lophosaurus Fitzinger 1843, where they reside to this date.

Morphology and morphometry

Four papers discussed morphology and morphometry. These included an article of popular science (Torr 2003); a descriptive analysis of the development of the ornamentation of scales (Alibardi 1999); an experimental study on female agamids and their offspring (Harlow 2001); and an account of the gross morphometry of museum specimens (Houston and Traynor 2018).

The developmental stages and the layers involved in the micro-ornamentation on scales in Hypsilurus (Lophosaurus) spinipes and Physignathus (Intellagama) lesueurii were shown to follow the same process, and was also noted to be similar in process to that of other species of lizards (Alibardi 1999). Consensus on the morphometric details of L. spinipes was reached (Harlow 2001; Torr 2003; Houston and Traynor 2018). However, morphometry of the three members of the genus Lophosaurus taken from museum holdings at the Queensland Museum and the Australian Museum, yielded inconclusive results regarding morphometric differences between the sexes. There was a lack of any significant sexual dimorphism in size (but sexual dichromatism was noted in L. spinipes) (Manning and Ehmann 1991; Shea et al. 1991). Although Torr (2003) also included morphometric measurements of L. spinipes that demonstrated an overall difference in size between these two congeners, the article lacked correlative analysis and added little to the morphometry of L. spinipes.

Ecology

Only one publication reviewed the overall ecology of Lophosaurus spinipes. This was a government report (Webb 1984).

Concerns raised about the status of L. spinipes by the Forestry Commission of NSW and NSW National Parks and Wildlife Services, led to a comprehensive report being produced by the Forestry Commission of NSW (Webb 1984). Fragmentation of habitat due to heavy logging, clearing for houses and infrastructure, and natural dwindling of rainforests noted south of the Hunter River Valley (resulting in disjunct localised populations), caused a reduction in sightings of the species. Further concerns grew with the proposal for the extensive forest-operation in the Hastings River catchment at that time, the lack of records in the area, and the unknown status of this species in general. Webb (1984) noted the paucity of data on this species, and used these scant records along with anecdotal evidence to conclude that the northern populations were well represented, but numbers were of concern in the areas at lower latitudes.

Vulnerability

Two publications highlighted the vulnerability of L. spinipes. These were an analysis of existing data sets (Hagger et al. 2013), and a governmental report (Kavanagh et al. 2022).

Hagger et al. (2013) evaluated the vulnerabilities to climatic change in 38 species in subtropical rainforests, and focussed on endemic and/or threatened species, using two aspects of vulnerability to evaluate each species: (1) resistance, defined by indicators of rarity; and (2) resilience, defined by indicators of a species’ potential to recover (Hagger et al. 2013). This study deemed L. spinipes’ vulnerability to be high; however, its resilience was also measured as high, thus offsetting vulnerability. The species was therefore listed as less vulnerable than many others due to its suspected ability to recolonise after disturbance (Webb 1984). Hagger et al. (2013) also concluded that montane, subtropical rainforest (L. spinipes’ habitat) deserves the highest protection status as habitat for vulnerable taxa.

Kavanagh et al. (2022) concluded that climatic change and fire were the key drivers of change and threat to forest-dependant species, which included L. spinipes. There were too few records for L. spinipes to conduct any modelling on distribution and occupancy.

Strengths and limitations

This review was undertaken using rigorous systematic methods. It includes a wide range of sources to comprehensively capture the available evidence. It is the first synthesis of information on L. spinipes to ascertain what data are available to help guide decisions on whether the species is vulnerable to climatic change, and to provide recommendations for future research.

This review has some limitations. The selected databases searched were chosen as they contained the most relevant and up-to-date information on the topic. However, it is possible that publications catalogued in other databases could have been missed. Due to the two-part PRISMA screening process, even if the inclusion criteria were present in the full text, it is possible that the publication was screened out in the Title/Abstract Screening stage because the criteria were not present there. Only published literature was included, and publication-bias may be present. Due to the limited number of results, appraisal of quality of the included publications was not conducted, and the results should be interpreted with caution.