Indigenous built structures and anthropogenic impacts on the stratigraphy of Northern Australian rockshelters: insights from Malarakk 1, north western Arnhem Land

Malarrak 1 is currently the northernmost excavated rockshelter on the Australian mainland, located in the Wellington Range in north western Arnhem Land. The site contains a rich late Holocene deposit, with extensive contact rock art, stone artefacts, shell, bone, contact materials, ancestral human remains, and other cultural material. Excavation of the Malarrak 1 rockshelter and analysis of its sediments revealed many impacts on site formation processes within the deposit. We attribute the disturbance to possible erosion or sediment deposition during periods of intense rainfall and also to the construction of timber structures within the site. This is supported by modern and historical observations and is the focus of this paper. The extent of the disturbance to Malarrak 1 provides a cautionary tale for other excavations in the region that may be affected by similar Indigenous site occupation, as these anthropogenic activities enhance the risk of further impacts arising from biological and geomorphological processes that can impinge on the stratigraphic integrity of the cultural deposits.

Archaeology of rock art at Dalakngalarr 1, centralwestern Arnhem Land

In 2011, we began researching the subsurface archaeology, geomorphology and rock art ofDalakngalarr 1, a moderately sized rock shelter on top of the central-western Arnhem Landplateau in Jawoyn Country. Here, four lines of evidence give relative or absolute ages for rockart:1. Archaeological excavations adjacent to a boulder that contains a painting of a red macropodreveal when that boulder attained its present position, so the red macropod must have beenpainted sometime afterwards.2. Paintings of axe/hoes with metal heads indicate that they were painted during the Europeancontact period. A nearby group of X-ray images are painted in comparable pigments,suggesting that they are contemporaneous with the axe/hoes.3. Geomorphological evidence suggests that parts of the site’s ceiling collapsed at datable timesin the past, indicating that the art on that roof must post-date the roof collapse.4. Direct accelerator mass spectrometry (AMS) radiocarbon dates on beeswax art.

Dating painted Panel E1 at Nawarla Gabarnmang, central-western Arnhem Land plateau

IntroductionThe southern Arnhem Land plateau contains a rich mosaic of thousands of rock art sites located in outcrops of Proterozoic Marlgowa Sandstone of the Kombolgie formation (Carson et al. 1999) (Figure 11.1). Within this region in Jawoyn Country can be found Nawarla Gabarnmang, an impressive rockshelter exhibiting a gridded network of pillars that supports a thick ceiling of 10 cm to 40 cm thick cross-beds of hard sandstone and quartzite (Figures 11.2 and 11.3; see also Chapter 10). The inter-layer joints and fissures between these compact and poorly soluble quartz-rich sandstones and quartzites have witnessed geologically slow dissolution of the bedrock, resulting in a hollowing out of the rock in a process known as ‘ghost rock’ formation or ‘phantomisation’ (Quinif 2010), a particular cave-forming process causing the regular gridshaped structure of underground cavities and pillars (for details of site formation processes, see Chapter 13).The remnant pillars supporting ceiling rock strata at Nawarla Gabarnmang are an anthropic cave structure (Delannoy et al. 2013; see Chapter 10): in addition to the slow geological dissolution of the rock along layer planes and fissure lines, people have also entirely or partially removed individual pillars, and possibly ceiling strata, over a period commencing sometime after the site was first occupied around 50,000 years ago (e.g. David et al. 2011, completed manuscript). What catches one’s attention at Nawarla Gabarnmang are the voids between the pillars, typically c. 1–2 m apart in the southwestern corner of the site, but more than 8 m apart in the central eastern portion. In that noticeably more open central-eastern area, a large, sub-horizontal and flat ceiling is supported by some 20 sparsely distributed pillars. Here, as in most other parts of the site, the floor of the sheltered area is generally flat and sub-horizontal, consisting of ashy sand with sparsely scattered, relatively small blocks of rock originating from the ceiling but not in their original fallen positions (these blocks have all, without exception, been moved by people). Within the fill across the site are rich archaeological deposits including stone artefacts, ochre pieces and animal bones, as revealed in the archaeological excavations (David et al. 2011; Geneste et al. 2012). What we see today in the shelter are the results of tens of thousands of years of human occupation, modification of rock surfaces and site use that express well the notion of ‘dwelling’ and ‘inhabitation’ (e.g. David et al. 2013, 2014; Delannoy et al. 2013; Geneste et al. 2010; cf. Ingold 2000; Thomas 2008).

Reconstructing rock art chronology with transfer learning: A case study from Arnhem Land, Australia

In recent years, machine learning approaches have been used to classify and extract style from media and have been used to reinforce known chronologies from classical art history. In this work we employ the first ever machine learning analysis of Australian rock art using a data efficient transfer learning approach to identify features suitable for distinguishing styles of rock art. These features are evaluated in a one-shot learning setting. Results demonstrate that known Arnhem Land Rock art styles can be resolved without knowledge of prior groupings. We then analyse the activation space of learned features and report on the relationships between styles and arrange these classes into a stylistic chronology based on distance within the activation space. By generating a stylistic chronology, it is shown that the model is sensitive to both temporal and spatial patterns in the distribution of rock art in the Arnhem Land Plateau region. More broadly, this approach is ideally suited to evaluating style within any material culture assemblage and overcomes the common constraint of small training data sets in archaeological machine learning studies.

Determining the age of paintings at JSARN-113/23, Jawoyn Country, central-western Arnhem Land plateau

Western Arnhem Land in northern Australia has the rare distinction, both at national and global scales, of containing a vast landscape of many thousands of rockshelters richly decorated with art, some of which was probably made tens of thousands of years ago, others as recently as a few decades ago. Yet the challenge remains as to how to date this art, how to find out how old it is. While relative dating methods have been commonly applied, in particular patterns of superimposition and changing faunal themes supposedly signalling changing environmental conditions, we still lack a clear understanding of the age of almost all the region’s art styles or conventions.Other chapters in this volume report direct dates for Arnhem Land art using radiocarbon determinations on beeswax figures with the likelihood that the ‘art event’, the time when a beeswax figure was made, is at most a few years different from the ‘carbon event’, the time of the last biological capture of atmospheric carbon, which is the actual date measured by radiocarbon. But many, in fact most, sites have no beeswax figures or other ways directly to date the art. Sometimes, as again reported in this volume, there is some indication of date when a radiocarbon determination is obtained on, for instance, charcoal in an archaeological deposit that can be related to the art. Often that route is also blocked: many a painted surface without beeswax figures is in no close relation to a deposit that might so be dated. What can be done then? Here we present results of investigations at a small rockshelter in Jawoyn Country, in the centralwestern part of the Arnhem Land plateau. Since its art cannot be directly dated, we follow a different path. In the first instance, we aim to understand the history, and antiquity, of the decorated rock surfaces, since the exposed surfaces of the boulder have undergone repeated transformations over a long time. Determining when now-decorated rock surfaces were formed can give us maximum possible ages for the art, since we can date when the surface first was available. Taken with related archaeological evidence from deposits, such as ochre fragments with signs of use, we can arrive at some indications for the age of the art, or at least how the range of possible dates is constrained. This approach is akin to that used at other sites in Jawoyn Country (see Chapters 11 and 15)

Archaeology of JSARN-124 site 3, central-western Arnhem Land: Determining the age of the so-called ‘Genyornis’ painting

In 2011, Gunn et al. published the discovery in a remote part of the western Arnhem Land plateau of a rock painting that closely resembles the most recent reconstructions of an extinct megafaunal bird, the dromornithid Genyornis newtoni. Characteristics of the painting distinguish it from depictions of other bird species and support its identification as G. newtoni: a deep convex bill, unlike the shallow bill of an emu or cassowary; a globular cranium and relatively thick neck; indication of a crop (emus and cassowaries lack crops); non-pendulous posture of the wing (unlike the pendulous posture of emus); the proportions of the pelvic limb showing long tibiotarsi and stout tarsometatarsi; the short, broad toes that appear to terminate in blunt claws; and a dorsal profile paralleling that of reconstructed dromornithids and unlike an emu or any species of cassowary, in which the vertex of the back is more anterior. The several points of special resemblance between the painting and reconstructions of the extinct bird based on paleontological evidence led Gunn et al. (2011:10) to conclude, ‘on the basis of probability the painting is indeed a representation of Genyornis newtoni’. This finding brings a conundrum. If the painting is indeed a contemporaneous depiction of G. newtoni, it becomes the oldest painting known in the world, for the bird is thought to have become extinct around 50,000 +/- 5000 years ago (Miller et al. 1999; Roberts and Brook 2010; Roberts et al. 2001) or even earlier (see Grellet-Tinner et al. 2016). Or that timing for the extinction of Genyornis is wrong; or a relic population survived longer on the Arnhem Land plateau (e.g. Murray and Vickers-Rich 2004), perhaps until the Last Glacial Maximum (LGM), which commenced around 25,000 years ago. And, as Gunn et al. (2011) acknowledged, there are still other alternatives. The painting might be of a ‘mythological’ animal: either one rooted in ancient memory of G. newtoni, as has been suggested for the ‘mihirung’ of southern Australian Aboriginal peoples (Vickers-Rich 1987) – although in societies without writing, social memory has not been found anywhere in the world to reliably extend in recognisable form over very long periods of time (cf. Bradley 2002) – or of a creature without a material counterpart that fortuitously resembled the fossil bird. The painting occurs on a vertical rock wall under a shallow overhang. Could the motif really have survived there since Pleistocene times, whether 25,000 or 50,000 years ago? And the painting shows a speared bird, so it could be the first evidence for the hunting of extinct megafauna in Australia. With these enigmas in mind, the Jawoyn Association asked us in 2010 to study the ‘Genyornis’ site, to investigate the age of the art and its archaeological and geomorphological context. Our major aims were to ask whether the painting dates, or theoretically could date, to the time of G. newtoni, or whether it must be more recent, and to uncover contextual ancient cultural information relating to the artworks and to occupation activities at the site in the past. Here, we present the results of these archaeological and geomorphological investigations along with a summary of the chemistry of the rock surface that houses the painting of the large bird thought by Gunn et al. (2011) to be of a Genyornis.

Population dynamics and pathogens of the invasive yellow crazy ant (Anoplolepis gracilipes) in Arnhem Land, Australia

<p>Though many populations of introduced species have been observed to collapse, the reasons behind these declines are seldom investigated. Anoplolepis gracilipes is considered among one of the top six most economically and ecologically damaging invasive ant species in the world. However, introduced populations of A. gracilipes have been observed to decline. My overall aims in this thesis were to document A. gracilipes population declines, to investigate the possibility that pathogens were playing a role in the observed population declines, and to identify putative pathogens infecting A. gracilipes as potential candidates for biocontrol agents. I documented the observed A. gracilipes population declines that were the driving force for this project. I detailed large-scale reductions in the spatial extent of four populations with before and after survey data. I also presented data on three populations that were recorded as present, but disappeared before they could be spatially delimited. I speculated on the possible reasons for these declines and explained why I do not think other explanations are likely. I then investigated the hypothesis that a pathogen or parasite is affecting A. gracilipes queens in declining Arnhem Land populations. I did this in three ways: 1) based on preliminary findings, I looked at the effect of an artificial fungal infection on A. gracilipes reproduction. I compared reproductive output between control colonies and those treated with either a fungal entomopathogen (Metarhizium anisopliae) or fungicidal antibiotics. There was no correlation between either treatment and the number of eggs, larvae, pupae or males a colony produced after 70 days. I found queen number had no effect on colony reproductive output, suggesting that queens are able to adjust their egg-laying rate in the presence of other queens. I found no evidence that M. anisopliae affected reproductive output at the tested concentrations; 2) I explored the hypothesis that a pathogen that kills or affects the reproductive output of A. gracilipes queens is the mechanism or reason behind the population declines. I measured queen number per nest, egg-laying rate, fecundity and fat content and compared them between sites in different stages of decline or expansion (population types, consisting of low, medium and high-density populations). I discovered that 23% of queens had melanized nodules, a cellular immune response in insects, in their ovaries or fat bodies. The presence of nodules was correlated with a 22% decrease in the number of oocytes per ovary; however, nodule presence was not associated with population type, suggesting that though there are clearly pathogens or parasites capable of penetrating the cuticle of A. gracilipes, they are unlikely to be responsible for the observed population declines; 3) I compared microbial communities (bacteria and viruses) between queens from different population types. I found viral sequences that match to the Dicistroviridae family of viruses in low and medium-density populations. I found no differences in bacterial community structure between population types. The presence of sequences similar to the entomopathogens Rhabdochlamydia and Serratia marcescens, as well as the reproductive parasite Cardinium in A. gracilipes, deserves further investigation. Though introduced species’ populations have been observed to decline, this is one of the first studies to quantitatively examine, document, and investigate a mechanism behind such a decline. Understanding the mechanisms by which an invader declines may have important implications for invasive ant management worldwide.</p>

Population dynamics and pathogens of the invasive yellow crazy ant (Anoplolepis gracilipes) in Arnhem Land, Australia

<p>Though many populations of introduced species have been observed to collapse, the reasons behind these declines are seldom investigated. Anoplolepis gracilipes is considered among one of the top six most economically and ecologically damaging invasive ant species in the world. However, introduced populations of A. gracilipes have been observed to decline. My overall aims in this thesis were to document A. gracilipes population declines, to investigate the possibility that pathogens were playing a role in the observed population declines, and to identify putative pathogens infecting A. gracilipes as potential candidates for biocontrol agents. I documented the observed A. gracilipes population declines that were the driving force for this project. I detailed large-scale reductions in the spatial extent of four populations with before and after survey data. I also presented data on three populations that were recorded as present, but disappeared before they could be spatially delimited. I speculated on the possible reasons for these declines and explained why I do not think other explanations are likely. I then investigated the hypothesis that a pathogen or parasite is affecting A. gracilipes queens in declining Arnhem Land populations. I did this in three ways: 1) based on preliminary findings, I looked at the effect of an artificial fungal infection on A. gracilipes reproduction. I compared reproductive output between control colonies and those treated with either a fungal entomopathogen (Metarhizium anisopliae) or fungicidal antibiotics. There was no correlation between either treatment and the number of eggs, larvae, pupae or males a colony produced after 70 days. I found queen number had no effect on colony reproductive output, suggesting that queens are able to adjust their egg-laying rate in the presence of other queens. I found no evidence that M. anisopliae affected reproductive output at the tested concentrations; 2) I explored the hypothesis that a pathogen that kills or affects the reproductive output of A. gracilipes queens is the mechanism or reason behind the population declines. I measured queen number per nest, egg-laying rate, fecundity and fat content and compared them between sites in different stages of decline or expansion (population types, consisting of low, medium and high-density populations). I discovered that 23% of queens had melanized nodules, a cellular immune response in insects, in their ovaries or fat bodies. The presence of nodules was correlated with a 22% decrease in the number of oocytes per ovary; however, nodule presence was not associated with population type, suggesting that though there are clearly pathogens or parasites capable of penetrating the cuticle of A. gracilipes, they are unlikely to be responsible for the observed population declines; 3) I compared microbial communities (bacteria and viruses) between queens from different population types. I found viral sequences that match to the Dicistroviridae family of viruses in low and medium-density populations. I found no differences in bacterial community structure between population types. The presence of sequences similar to the entomopathogens Rhabdochlamydia and Serratia marcescens, as well as the reproductive parasite Cardinium in A. gracilipes, deserves further investigation. Though introduced species’ populations have been observed to decline, this is one of the first studies to quantitatively examine, document, and investigate a mechanism behind such a decline. Understanding the mechanisms by which an invader declines may have important implications for invasive ant management worldwide.</p>

Genetic Factors Associated with Variation in Abundance of the Invasive Yellow Crazy Ant (Anoplolepis gracilipes)

<p>A key component of successful invasion is the ability of an introduced population to reach sufficient abundance to persist, spread, and alter or dominate the recipient biological community. Genetic diversity is one of many factors that may contribute to population dynamics, but has important ramifications for biological fitness, and thus invasion success in the long term. I explored genetic factors associated with variation in abundance (i.e., differential invasion success) of the yellow crazy ant Anoplolepis gracilipes in the Indo-Pacific region, primarily focussing on Arnhem Land in Australia's Northern Territory. I explored five aspects that I hypothesised could contribute to variation in the abundance of this ant: 1) I investigated the unusual reproductive mode of A. gracilipes, and tested whether it involved dependent-lineage genetic caste determination (DL GCD) in Arnhem Land. In DL GCD systems populations require hybridisation between genetically distinct groups to produce both reproductive and worker castes. Asymmetry in the ratio of different lineages may result in low abundance and population collapse. I found no evidence for a DL GCD system in A. gracilipes, and thus its abundance in Arnhem Land does not appear to be constrained by any lineage ratio asymmetry. Worker reproduction (either of males or asexual production of other workers) also appeared unlikely. The reproductive mode of the species remains fascinating but enigmatic; 2) I explored whether multiple source populations were responsible for the observed variation in abundance in Arnhem Land (i.e., is abundance associated with propagule pressure, or populations from different sources), and if the population has diverged since introduction. The A. gracilipes population in Arnhem Land stemmed from a single source, and thus propagule pressure was apparently not responsible for variation in abundance. In contrast to many invasive ants, population divergence has occurred since introduction; 3) I tested the hypotheses that genetic variation was associated with variation in abundance in Arnhem Land, and that ecological success was density-dependent. While the population divergence found in Chapter 3 was not related to variation in abundance, genotypic diversity was higher in more abundant nest clusters. These more abundant nest clusters were in turn associated with lower native ant species diversity, and a difference in composition of the invaded ant community (i.e., greater ecological success); 4) I revisited the invasion of the yellow crazy ant in Tokelau to determine whether a haplotype that was linked to greater abundance and dominance of the ant community has increased in distribution. Although ants of the inferred dominant haplotype were implicated in most new invasions, their abundance was substantially lower than previously observed in Tokelau; 5) I conducted a preliminary analysis of the metagenomic diversity of A. gracilipes endogenous parasites and symbionts among populations from Christmas Island, Okinawa, Samoa and Arnhem Land. Bacterial community composition and diversity differed between the study populations. Variation in abundance among A. gracilipes populations in Arnhem Land was not due to parasite load on populations with low abundance. However, low abundance of A. gracilipes was correlated with lower microbial diversity overall, and higher prevalence of some groups, notably two that confer antibiotic properties. Together, my findings suggest that propagule pressure, reproductive mode and haplotype-specific effects do not appear to be associated with variation in A. gracilipes abundance. Other genetic factors I investigated do appear to be associated with the variation in A. gracilipes abundance and effects on the invaded ant communities. Genotypic diversity was positively related to the abundance of A. gracilipes in Arnhem Land, and this relationship may be affected by population divergence through population bottlenecks. In addition, differences in bacterial diversity among populations highlighted several candidate bacteria that could be associated with variation in abundance, which would be a topic of future work. Although genetic factors are often implicated in the successful establishment of invasive species, my thesis demonstrates that genetic factors may also be associated with post-establishment population dynamics.</p>

Genetic Factors Associated with Variation in Abundance of the Invasive Yellow Crazy Ant (Anoplolepis gracilipes)

<p>A key component of successful invasion is the ability of an introduced population to reach sufficient abundance to persist, spread, and alter or dominate the recipient biological community. Genetic diversity is one of many factors that may contribute to population dynamics, but has important ramifications for biological fitness, and thus invasion success in the long term. I explored genetic factors associated with variation in abundance (i.e., differential invasion success) of the yellow crazy ant Anoplolepis gracilipes in the Indo-Pacific region, primarily focussing on Arnhem Land in Australia's Northern Territory. I explored five aspects that I hypothesised could contribute to variation in the abundance of this ant: 1) I investigated the unusual reproductive mode of A. gracilipes, and tested whether it involved dependent-lineage genetic caste determination (DL GCD) in Arnhem Land. In DL GCD systems populations require hybridisation between genetically distinct groups to produce both reproductive and worker castes. Asymmetry in the ratio of different lineages may result in low abundance and population collapse. I found no evidence for a DL GCD system in A. gracilipes, and thus its abundance in Arnhem Land does not appear to be constrained by any lineage ratio asymmetry. Worker reproduction (either of males or asexual production of other workers) also appeared unlikely. The reproductive mode of the species remains fascinating but enigmatic; 2) I explored whether multiple source populations were responsible for the observed variation in abundance in Arnhem Land (i.e., is abundance associated with propagule pressure, or populations from different sources), and if the population has diverged since introduction. The A. gracilipes population in Arnhem Land stemmed from a single source, and thus propagule pressure was apparently not responsible for variation in abundance. In contrast to many invasive ants, population divergence has occurred since introduction; 3) I tested the hypotheses that genetic variation was associated with variation in abundance in Arnhem Land, and that ecological success was density-dependent. While the population divergence found in Chapter 3 was not related to variation in abundance, genotypic diversity was higher in more abundant nest clusters. These more abundant nest clusters were in turn associated with lower native ant species diversity, and a difference in composition of the invaded ant community (i.e., greater ecological success); 4) I revisited the invasion of the yellow crazy ant in Tokelau to determine whether a haplotype that was linked to greater abundance and dominance of the ant community has increased in distribution. Although ants of the inferred dominant haplotype were implicated in most new invasions, their abundance was substantially lower than previously observed in Tokelau; 5) I conducted a preliminary analysis of the metagenomic diversity of A. gracilipes endogenous parasites and symbionts among populations from Christmas Island, Okinawa, Samoa and Arnhem Land. Bacterial community composition and diversity differed between the study populations. Variation in abundance among A. gracilipes populations in Arnhem Land was not due to parasite load on populations with low abundance. However, low abundance of A. gracilipes was correlated with lower microbial diversity overall, and higher prevalence of some groups, notably two that confer antibiotic properties. Together, my findings suggest that propagule pressure, reproductive mode and haplotype-specific effects do not appear to be associated with variation in A. gracilipes abundance. Other genetic factors I investigated do appear to be associated with the variation in A. gracilipes abundance and effects on the invaded ant communities. Genotypic diversity was positively related to the abundance of A. gracilipes in Arnhem Land, and this relationship may be affected by population divergence through population bottlenecks. In addition, differences in bacterial diversity among populations highlighted several candidate bacteria that could be associated with variation in abundance, which would be a topic of future work. Although genetic factors are often implicated in the successful establishment of invasive species, my thesis demonstrates that genetic factors may also be associated with post-establishment population dynamics.</p>