Relationship between Land Use/Land-Use Change and Human Health in Australia: A Scoping Study

We undertook a scoping study to map the relevant evidence, summarise the findings, and to help identify gaps in the knowledge base on the relationship between land use/land-use change and human health in Australia. Our systematic search of the scientific literature for relevant articles up to August 2020 identified 37 articles. All 37 articles meeting our inclusion criteria were published after 2003. Zoonotic or vector-borne disease constituted the most common health outcome type studied. Agriculture/grazing was the land use/land-use change type most frequently represented in the literature, followed by coal seam gas extraction and open cut coal mining. The relationship between land use/land use change and human health in Australia, is not conclusive from the existing evidence. This is because of (1) a lack of comprehensive coverage of the topic, (2) a lack of coverage of the geography, (3) a lack of coverage of study types, and (4) conflicting results in the research already undertaken. If we are to protect human health and the ecosystems which support life, more high-quality, specific, end-user driven research is needed to support land management decisions in Australia. Until the health effects of further land use change are better known and understood, caution ought to be practiced in land management and land conversion.


Introduction
The planet's natural systems are being transformed through human activity in ways that are profound, extraordinary, and accelerating [1][2][3][4]. Anthropogenic land use/land-use change (LULUC)-along with climate change and stratospheric ozone depletion-has, and continues to, degrade the environment to the extent that nearly every dimension of human health is implicated [1,2]. Given that the major causes of environmental degradation interact to impact human health, they are in effect, inextricable [1,2]. Informing responsible decision making that protects both human health and the health of the natural systems that support health, nonetheless requires an understanding of the effects of each of the major causes separately, in the relevant local context.
Like much of the globe, the majority of Australia's landscape has been significantly altered (mostly in the last two and a half centuries) to accommodate population growth and growth economies through: urbanisation; agricultural/grazing expansion; forestry; mining; transport; waste management; and other industry, manufacturing, and infrastructure [5][6][7][8][9]. McFarlane et al. (2013) in an extensive examination of the association between infectious disease emergence and land-use and native vegetation change in Australia, reported that 22% of the infectious diseases they reviewed were associated with land conversion [10]. Furthermore, they found that the historical clustering observed in terms of incidence of environmental, vector-borne, and zoonotic disease followed periods of substantial land clearing in Australia [10]. This is supported by findings that suggest that land-use change increases the transmission of henipaviruses (like Hendra virus) in Australia [11,12]. Kessler et al. (2018) and Paez et al.'s (2018) reviews of the evidence on the topic of LULUC and spillover risk of henipaviruses to domestic animals and to humans, exemplifies the way in which LULUC can impact health. In the case of henipaviruses, the altered diet, roosting habitat, and migration behaviours of flying foxes, as a result of changes to their natural habitat, has increased the risk of disease for humans [11,12].
The evidence on the relationship between LULUC and human health in Australia has only relatively recently emerged-within the last two decades. The evidence is neither extensive or comprehensive in terms of coverage of LULUC and health topics, with a predominance of studies on the relationship between LULUC and vector-borne, or zoonotic disease. In order to adequately account for the risks to human health in decision making on land management and proposed land change, up-to-date summaries of the state of the evidence on the relationship between LULUC and health in Australia, are an essential starting point.
In this study, we aimed to assess the current state of evidence on the relationship between LULUC and human health in Australia, and to identify research gaps in the existing literature. We also aimed to draw implications from the existing evidence in order to inform policy and practice. To address this aim, we undertook a scoping study. Scoping studies are an appropriate approach where the 'mapping' of the relevant literature is the main goal [13]. Scoping studies typically address broad topics where different study designs and levels of evidence are applicable [13]. Unlike systematic reviews, scoping studies are less focussed on assessing the quality of the included studies [13]. Scoping studies can have a number of different purposes, but the main aim of this scoping study was to summarise the findings and range of research on the relationship between LULUC and human health in Australia. A summary of the evidence is a necessary starting point in terms of a synthesis of findings and to establish the state of the evidence, but also to help identify gaps in the knowledge base-all of which are necessary to ultimately inform practice and policy on land management and proposed land use change. Land management/proposed land use change decisions need to have at their core, a strong evidence base to support the protection not only of human health and wellbeing, but importantly, of the ecosystems that support life.

Identifying the Research Question
This scoping study answered the following question: What is the current state of evidence on the relationship between land use/land-use change and human health in Australia?

Sources of Literature
We conducted a systematic search of the scientific literature for relevant articles in three electronic databases: two content specific databases-Medline/PubMed (biomedical sciences) and PsychINFO (psychology and related disciplines)-and one multidisciplinary database-Web of Science. We did not conduct a primary search for grey literature but included it where relevant, if it came to our notice in the reference lists of key articles.

Search Terms
Apart from specifying place (Australia), the literature search included LULUC terms related broadly to different aspects and activities of land (and freshwater) use and change, and various types of human health outcomes/conditions/diseases. LULUC terms were derived from preliminary searches of the literature, and from key articles on the links between LULUC and health [2,4,14,16].
Similarly, health terms were drawn from key literature [2,4,10,17] and from Australian Health Department lists of nationally notifiable diseases [18]. Key search terms were mapped to each electronic database prior to the article search, with slight variations on some terms across the databases, depending on whether major terms were searched within the 'subject heading' or 'thesaurus' of a particular database. For example, "natural resources" is the broad subject heading in Medline/PubMed for the natural environment, while "nature (environment)" is the equivalent subject heading or "American Psychological Association Thesaurus" in PsychINFO. Table 1 shows the general terms used, not specific to each database. The search was carried out on March 19, 2019, at which time the citation and abstract for all articles identified via search terms were uploaded to an Endnote (Clarivate Analytics, Philadelphia, U.S.A) X9 library. The search was updated on 12 August 2020. Australia "land use change" OR "land cover change" OR "land clear*" OR "land use" OR "vegetation clear*" OR "land* change*" OR "land degrad*" OR "environment* OR degrad*" OR "eco* degrad*" OR "forest degrad*" OR "forest loss*" OR deforest* OR "environment* change" OR mine OR mining OR agriculture* OR farm OR road OR dam OR urban* OR "natur* environment*" OR "eco* services" OR eco*

Search Process
Our multi-step process for seeking relevant articles to answer the research question is outlined in Figure 1. Our initial search yielded 3155 articles. Two reviewers (the co-authors) were involved in deciding which articles based on title and abstract appeared sufficiently relevant enough to warrant full-text assessment for eligibility, and which articles finally met our inclusion criteria. At each step, the reviewers chose to eliminate or keep articles through a process of consensus. Thirty-seven articles ultimately met our inclusion/exclusion criteria.

Article Selection (Inclusion/Exclusion Criteria)
Publications meeting our criteria included: reports of original research (no reviews or opinion pieces) published up to 19 March 2019; in English; focused on or presenting data from Australia; that refer to a specific health issue; and include the topic of LULUC. The search was updated on 12 August 2020. No further eligible articles were identified.
Inclusion/exclusion criteria specific to the topic of LULUC was defined as either articles that related to one or more aspects of LULUC, or that partly or fully fulfilled the combined (land use, land cover, and land-use change) concept of LULUC as defined by the Intergovernmental Panel on Climate Change (IPCC). The 'land use' (LU) component was defined as, "the total of arrangements, activities, and inputs that people undertake in a certain land cover type", with land cover as, "the observed physical and biological cover of the earth's land, as vegetation or man-made features' [19]. The 'land-use change' (LUC) component was defined as, " . . . human activities which: (a) change the way land is used (e.g., clearing of forests for agricultural use, including open burning of cleared biomass), or (b) affect the amount of biomass in existing biomass stocks (e.g., forests, village trees, woody savannas, etc.)" [19].
Inclusion/exclusion criteria specific to the topic of human health included articles that focused on either direct health outcomes as a result of LULUC, for example an infection or depression; or causative agents of disease/ill health, for example prevalence of mosquitos or pollution exposure, even where no health measurements were taken. While the presence or absence of causative agents does not always result in ill health, we were interested in capturing published literature on increased risk-even if inexact-of disease in humans in relation to LULUC.

Article Selection (Inclusion/Exclusion Criteria)
Publications meeting our criteria included: reports of original research (no reviews or opinion pieces) published up to 19 March 2019; in English; focused on or presenting data from Australia; that refer to a specific health issue; and include the topic of LULUC. The search was updated on August 12, 2020. No further eligible articles were identified.
Inclusion/exclusion criteria specific to the topic of LULUC was defined as either articles that related to one or more aspects of LULUC, or that partly or fully fulfilled the combined (land use, land cover, and land-use change) concept of LULUC as defined by the Intergovernmental Panel on Climate Change (IPCC). The 'land use' (LU) component was defined as, "the total of arrangements, activities, and inputs that people undertake in a certain land cover type", with land cover as, "the observed physical and biological cover of the earth's land, as vegetation or man-made features' [19]. The 'land-use change' (LUC) component was defined as, "…human activities which: a) change the way land is used (e.g., clearing of forests for agricultural use, including open burning of cleared biomass), or b) affect the amount of biomass in existing biomass stocks (e.g., forests, village trees, woody savannas, etc.)" [19].
Inclusion/exclusion criteria specific to the topic of human health included articles that focused on either direct health outcomes as a result of LULUC, for example an infection or depression; or causative agents of disease/ill health, for example prevalence of mosquitos or pollution exposure, even where no health measurements were taken. While the presence or absence of causative agents We included studies looking at both positive as well as negative links between LULUC and human health. Notwithstanding the inextricable links between the different forms of environmental degradation, i.e., LULUC and climate change, we excluded studies focusing on climate change/weather extremes and health, because there already exists a relatively large body of evidence on this topic. Articles meeting content/topic criteria were then assessed on the basis of a combination of study objectives and whether the health and LULUC topics were sufficiently the focus of the study and adequately addressed in the results and discussion sections of the articles.

Data Management and Characterization/Charting
We charted data extracted from all articles meeting the inclusion criteria, in a Microsoft Excel (2016) spreadsheet. Included was: author/s; year of publication; title; research objectives; location; study design; type of LULUC and measures; health outcome and measures; and summary of findings. These data points were selected specifically to answer the research question and meet the study purpose.

Analysing, Summarizing, and Reporting the Results
Using the above described charted data as the basis of the analysis, we undertook to quantify the extent, nature and distribution of the studies included in the review (descriptive quantitative analysis). We also organised the literature thematically according to the different types of LULUC and their relationship to human health outcomes (qualitative analysis). Table 2 summarises the extent, nature, and distribution of the 37 articles included in the analysis. Of the articles meeting our inclusion criteria, the earliest was published in 2004. There were four times the number of articles published in the 2016-2019 period (n = 12, 32.4%), compared with that of the 2004-2007 (n = 3, 8.1%) period ( Table 2). The vast majority of articles described studies which had employed quantitative methods (n = 31, 83.8%) ( Table 2). All Australian states and territories, except South Australia and The Australian Capital Territory, were represented in the included articles, but there was unequal representation across and within states and territories. The State most frequently represented was Queensland, constituting 32.4% of all articles (n = 9 focussing just on Queensland, and n = 3 focussing on eastern Australia which includes Queensland and New South Wales). New South Wales (n = 8, 21.6%, including State-specific, Eastern, and Southeast Australia-located studies) and Western Australia (n = 6, 16.2%) were the next most frequently represented locations in Australia (Table 2).

Profile of Included Literature
Over one third (n = 13, 35.1%) of all articles focused on zoonotic or vector-borne disease (Hendra virus, Ross River virus, and mosquito-borne disease-general), with a majority of those specific to Hendra virus and Ross River virus, each constituting n = 4, 10.8% of all articles included ( Table 2). The vast majority of zoonotic or vector-borne disease articles (n = 11, 84.6%) measured vector abundance or animal disease as a proxy for human disease, i.e., not as a direct measure of human disease ( Table 2). Just less than one third of all articles related to health and wellbeing in general or multiple health outcomes (n = 12, 32.4%), with the remaining one third of articles dedicated to mental health, ecosystem goods and services, melioidosis, pathogen spillover-general, pollen counts, and respiratory disease ( Table 2). Over half of all articles (n = 21, 56.7%) based their health outcome of interest on observational data, e.g., incidence, prevalence, mortality, counts of bacteria or pollen, or quantity/quality of the relevant ecosystem goods and services (soil, water, biodiversity) ( Table 3). Self-report survey (n = 7, 18.9%), and semi-structured interview (individual and/or focus group) (n = 4, 10.8%) were the next most frequently employed methods for assessing the respective health outcomes of interest ( Table 3).
The most frequent LULUC topic among the included articles was multiple land use/change types (n = 17, 45.9%); in these articles two or more land use/change types were compared in relation to the respective health outcome, and frequently but not always included single or multiple agricultural land use types ( Table 2). Dryland salinity (n = 5, 13.5%), coal seam gas extraction (also n = 5, 13.5%), and open cut coal mine (n = 3, 8.1%) were the next most frequent LULUC topics ( Table 2). Just less than half (n= 16, 43.2%) of all articles based their LULUC topic of interest on existing land use (GIS mapping and satellite images) or field data (observations, field-surveys) ( Table 3). For 15 (40.5%) articles, the LULUC topic of interest did not require data or measurement as it related to either proposed mining or agriculture development or existing mining, contaminated sites, urbanisation, or traditional Indigenous ancestral lands (Table 3). Effects of grazing on ecosystem services were herbivore specific and varied. Critical functions associated with decomposition and nutrients cycling declined with increasing aridity, and these effects were of a greater magnitude than any effects due to grazing. Livestock grazing resulted in reductions in plant biomass, but was associated with greater nitrogen availability (sheep, and to a lesser extent cattle) and decomposition. There was an overall neutral effect of grazing on carbon storage. Phosphorus cycling was one of the ecosystem services less affected by grazing intensity.
Wijesiri et al. 2018 [55] Ningi Creek catchment, North Brisbane Statistical Division, Queensland, Australia Multiple; data from 'Queensland Spatial Catalogue' Water quality; via surface water samples Majority of pollutants released into catchment waters came from 'Water environment' land use type (marsh, wetland, and other water bodies ecosystems). Concentrations of iron and nitrates in downstream water environment may have been due to agricultural activity upstream-there was an expansion in agriculture land and decreases in water environments over study period.   Dryland salinity has strong influence on mosquito community structure and is significantly associated with the abundance of mosquitoes. While interaction between salinity and season is also significant, the influence of salinity remained substantial even when seasonal variation taken into account. Presence of depression was consistently linked to residence in areas with high salinity and the association of asthma, suicide, and heart disease with salinity was likely attributable to the co-morbidity of the conditions with depression. The size of the mangrove stand itself had a significant positive effect on mosquito abundance-explained by resource concentration. Also positive (but not significant) relationship between percentage of industrial land and mosquito abundance. Percentage of parkland and open water had no effect on mosquito abundance and percentage of residential land and bushland in surrounding area had strong negative effect on mosquito abundance in urban mangroves-likely through changes in surface water flow in general but also due to below average rainfall preceding surveys. Mosquito species richness was elevated in anthropogenic grasslands relative to rainforest habitats; the creation of anthropogenic grasslands adjacent to rainforests may increase the susceptibility of species in both habitats to the transmission of novel diseases via observable changes to and mixing of the vector community on rainforest edges.   Measures of neighbourhood environment weakly related to residents' personal wellbeing and level of connection to nature. Some evidence that increases in vegetation density associated with increases in personal wellbeing for certain types of residents. Conversely, many environmental variables strongly related to variation in neighbourhood wellbeing across a range of demographic categories, and residents' satisfaction with their local neighbourhood increased with greater number of bird species, higher proportion of vegetation cover, and lower level of urban development. Nevertheless, demographic variables always the variables most strongly associated with wellbeing or connection to nature.

Risk of Zoonotic Disease
Plowright et al. 2011 [30] Eastern Australia Urbanisation Birth rate, death rate, seasonality, incidence and distribution of disease parameters, and population and meta-population characteristics of grey-headed and black flying foxes; estimated from experimental, captive and field studies Loss to various anthropogenic land use changes of 75% of once contiguous forest cover on east coast of Australia-natural food resource of grey-headed and black flying foxes-provides plausible scenario for recent apparent increased frequency of Hendra virus outbreaks.
Year-round alternative food in expanding urban and peri-urban areas increases number of flying foxes in contact with human and domestic animal populations and decreases bat migratory behaviour, which could lead to decline in bat population immunity, giving rise to more intense outbreaks after local viral reintroduction. Residents' worry about disruptive effect of environmental contamination on health and well-being; via telephone questionnaire survey Female participants, people living with disability or long-term illness, and those living close to contaminated sites were significantly more likely to worry about contamination. Presence of hydrocarbon, metal, and chlorinated solvent were significantly more likely to cause worry about contamination than asbestos. Worries were focused more on how contaminants might disrupt physical health, mental health, and lifestyle, than they were on the effect of the contaminant on flora, fauna, and broader ecosystem.

General Health
Werner et al. 2016 [48] Queensland, Australia Coal seam gas extraction, coal mining, rural/agricultural Hospitalization rates as measured by ICD-10-AM codes; from Queensland Hospital Admitted Patient Data Collection Coal seam gas area showed increases in hospitalization rates (compared to rural area, not to coal mining area) for neoplasms and blood/immune diseases. This descriptive-analytic study provided preliminary assessment of hospitalization rates only and did not assess causality. Association, Ag Force, and Lock the Gate-concern about coal seam gas impacts on human health, community, and environment (constituting 'off-farm'), was rated, along with weather and economic viability, as most stressful factor for farmers, and explained significant amount of unique variance in farmers' depression and stress reactivity (after controlling for more traditional agricultural stressors). Coal seam gas impacts 'on-farm'-operations, profitability, and personal privacy-were less of a concern to farmers. Transformation of the regional landscape challenged many participants' sense of place, identity, physical and mental health, and general wellbeing. Participants also felt powerless to influence the change process. These responses resonate with the dominant components of 'solastalgia', which was experienced even by some who worked in the relevant mining and power industries.

Mining development proposal: large open-cut coal mine
Perceived psychological impacts for individuals and community; via semi-structured interviews Semi-structured interviews with range of individuals representing various roles and service industries within the community indicated that the proposed mineral development was source of psychological stress, in farmers in particular. Impacts at individual and family level included a sense of powerlessness; unknown future; interrupted succession plans for young and old; threat of disruption to relationship with the land and landscape; and changed financial circumstances. The proposal also created tensions within the community regarding differential impacts, with some seeing opportunity and others perceiving threat. Community well-being declined with increased clearing of endangered ecological communities, loss of highly significant Aboriginal sites, and displacement of rural families from affected villages, but increased with length of time that mine provides employment and with planting or protection of endangered ecological communities as offsets.

General Health and Wellbeing
Johnston et al.
2007 [22] Arnhem land, northeast corner of Northern Territory, Australia Traditionally owned Indigenous ancestral lands Perspectives of health and wellbeing; via observation and unstructured and semi-structured interviews Major themes which emerged from the observations and interviews with Aboriginal community members, included: the need that both land and people have for each other, for the well-being of both; that traditional lands provided access to traditional food and medicines (and physical activity to get them); a way of escaping from stresses; a means to educating young people; and deeper connection to the past and therefore to Aboriginal identity via traditional stories and beliefs. Health and wellbeing; via semi-structured interviews Semi-structured interviews identified that Caring for Country builds self-esteem, fosters self-identity, maintains cultural connection, and enables relaxation and enjoyment through contact with the natural environment. Caring for Country-a traditional practice of over 50,000 years-constitutes a key determinant of the health and wellbeing of Indigenous people in Australia.

Agriculture/Grazing
Two articles focussed specifically on agricultural land use [21,41]. An additional seven articles did not focus on agricultural land use/change alone, but included agriculture/grazing as one of a number of land use types used for comparison in relation to the respective health outcome of interest. (Table 3) [25,28,33,35,39,50,55].

Dryland Salinity
Secondary salinity occurs over time as a result of agricultural activity in dry environments such as southwestern Western Australia, where the native vegetation has been cleared and replaced with shallow-rooted crops and grazing grasses [26]. Shallow-rooted vegetation draws less on groundwater, which can lead to a rise in the water table, transporting land and surface water-contaminating salts to the surface [26]. Although the phenomenon of secondary salinity necessarily relates to agricultural land use/change, the five articles which focused on various health outcomes related to dryland salinity-'mental health'; mosquito abundance'; and 'other chronic disease'-are presented in their own category because of the specific conditions (and health outcomes) of salinity as a LULUC type (Table 3). Table 3 summarises findings of the eleven articles which measured 'converted' versus 'unconverted' land, [52] or multiple land use types, in relation to 'mosquito abundance'; 'pollen counts'; 'respiratory disease'; 'vector-borne and infectious disease'; and 'wellbeing' [29,34,42,43,45,47,49,51,53,54].

Urbanisation
One article focussed specifically on urbanisation as a LULUC type, and assessed the risk of Hendra virus outbreaks in Australia as a result of year-round alternative food for flying foxes in expanding urban-and peri-urban settings and a contraction of natural food sources (Table 3) [30].

Contaminated Sites
The only article addressing the impacts of contaminated sites (not related to contamination from mining operations) assessed wellbeing as an outcome (Table 3) [56].

Traditionally Owned/Ancestral Indigenous Land Managed by Indigenous Communities
Two articles identified as part of our search process address the importance of access to ancestral lands for the general health and wellbeing of Australian Indigenous communities (Table 3) [22,27].

Discussion
Of the 37 articles meeting our inclusion criteria (not restricted by publication date), all were published after 2003. Most of the research focussed on the eastern states of Queensland and New South Wales, and on specific regions within the states and territories. Zoonotic or vector-borne disease constituted the most common health outcome type studied-well over a third of all articles. Agriculture/grazing was by far the LULUC type most frequently represented in the literature we reviewed, followed by coal seam gas extraction and open cut coal mining.

Land Use/Land-Use Change and Human Health in Australia
Agriculture/grazing was reported by the range of studies in the review as having negative, inconclusive, and neutral relationships with health outcomes in Australia. In the literature, agriculture/grazing was related to decreased wellbeing [39,40]; increased abundance of disease-transmitting mosquitoes [20,23,24,28]; decreased ESGs [36]; occurrence of the melioidosis bacterium B. pseudomallei in soil [25,33]; and hospitalisations for depression [26,31]. The relationship between agriculture/grazing and soil quality [51]; and agriculture/grazing and water quality, Ref. [55] was mixed/inconclusive. There was no association between mental health scores in women, and soil salination, in a study conducted in the rural south-western corner of Western Australia [41] The mixed and sometimes conflicting findings epitomise the complexity of the relationship between LULUC and human health, and the challenge of synthesising findings, largely because of the variety of methods used. For example, studies looking at the association between mental illness and dryland salinity in the same area in Western Australia came to different conclusions on the basis of different methods. Dryland salinity can have economic and social impacts on communities, with changes in agricultural productivity leading to financial stress and potentially mental illness, but Fearnley et al. (2014)-using self-reported survey data as part of an established longitudinal study of a randomly recruited study sample from the national Medicare health insurance database, which covers all citizens and permanent residents of Australia-found no associations between salinity and mental health scores for women across three age cohorts [41]. By contrast, Speldewinde et al. (2009) and Speldewinde et al. (2011) reported findings which indicated that after adjustment for major socioeconomic and demographic factors, an elevated risk of hospitalisations for depression was associated with residence in areas proportionately more affected by dryland salinity [26,31]. In explaining findings inconsistent with their own, Fearnley et al. (2014) proffered that area level measurements of aspects of the physical environment-such as salinity-may be poor indicators for individual health outcome analyses due to 'ecological fallacy', which is commonly observed in spatial analysis, i.e., the individuals being admitted to hospital may not be the same individuals exposed to dryland salinity [41]. Differences in results between Fearnley's et al. (2014) and the two Speldewinde (2009, and 2011) studies on salinity and depression may also be due to the use of different measures for mental health: the Mental Health Component Score (MCS) versus the use of hospital admission data, respectively [41]. While one would expect some correlation, the MCS-a general measure of mental health-is unlikely to capture exactly the same phenomenon as hospital admissions data, [41] in terms of both severity and specificity of mental health conditions.
Other anthropogenic LULUC types such as industry, non-native grasslands, and urbanisation-all of which necessarily impact biodiversity-were negatively associated with health in terms of increased mosquito abundance and species richness [34,47,49,54]; Hendra virus spillover to horses and humans [30,45,51]; and respiratory disease [54]. However, as with agriculture/grazing, the evidence indicated a nuanced relationship between the 'other anthropogenic LULUC types' and human health. For example, the highest risk of pathogen spillover (not disease specific, in the case of Faust's et al. (2018) simulation study) from wildlife to domestic animals, to humans, occurs at intermediate levels of habitat loss [52]. Interspecies contact and host populations not only vary with the proportion of land converted; time since initial habitat loss is also likely to drive changes in infectious disease transmission [52]. The working explanation for the dynamic relationship between land change and pathogen spillover-using mosquito-borne disease as an example-is that declining resources for reservoir hosts (wildlife) in converted landscapes drive wildlife from their natural habitat, leaving infected vectors (mosquitoes) to attain blood meals from the (now) more readily available human hosts [52]. Pathogen spillover is complex and influenced by multiple processes, including pathogen dynamics in reservoir hosts, environmental processes determining pathogen survival, carriage of the pathogen beyond hosts, and the behaviour and susceptibility of recipient hosts [10][11][12]57]. Each one of these processes may respond to changing landscapes and determine the particular relationship between land conversion and disease emergence [52].
Evidence for the relationship between mining-another anthropogenic LULUC type-and health in Australia also signified a complex dynamic, with competing vested interests, and differential impacts and benefits. Coal seam gas extraction was found to be associated with depression and stress in farmers [46]; increased population hospitalisation rates for neoplasms (tumours) and blood/immune disease [48]; and various health complaints-skin and eye irritations, headaches, paraesthesia, fatigue, and difficulty concentrating [36]. However, these findings were contentious, with as many studies on CSG extraction and open cut coal mining reporting inconclusive or mixed findings with regard to impacts on health in Australia [32,37,38,44]. For example, The McCarron (2013) investigation which reported various health complaints of residents in a CSG extraction area in Queensland, [36] was in response to a State health department report which concluded no clear link between residents' health complaints and the local CSG activities [38] McCarron (2013) surveyed 113 residents from 38 households who had previously expressed concerns about health impacts of the CSG activity, and collated their self-reported health status [36]. As part of the study, neither clinical nor environmental sampling assessments were undertaken, and the sample was not representative of the whole population [36]. While the Queensland Health report indicated that no medical staff visited the CSG site as part of their investigations, and only 15 people were clinically examined, they reported that there was nothing in their environmental assessments on air, water, or soil to indicate a risk to human health [38]. McCarron's (2013) conclusions do align more closely, than do the Queensland Health (2013) report, to a summary of literature from Australia and overseas regarding human and environmental health concerns from unconventional gas mining [58]. Notwithstanding incomparable methods, and methodological limitations in the two reviewed articles reporting conflicting findings, they serve to highlight the challenges in undertaking studies involving CSG exploration and health outcomes in Australia, including the small sizes of the affected populations and highly politically-charged situations.
Another complexity highlighted in the available evidence on the relationship between land use type, and human health in Australia, was the finding of 'opportunity-threat' differentials within affected groups. For example, while a proposed open cut-coal mine development in Queensland was a source of stress ('threat') to a range of community members-but especially to farmers and their families-some in the community saw 'opportunity', and this differential created tension between groups [37]. Similarly-albeit related to agricultural development-Adams et al. (2014) identified differences in responses by residents to the proposed development of a river catchment area in the Northern Territory. Indigenous respondents were more likely to value socio-cultural and biodiversity factors over the 'commercial' factors of the proposed development, while respondents who earned an income from agriculture were more likely to value the economic gains from which they were likely to benefit [40]. There was also evidence in the literature we reviewed, of differential impacts on human health and wellbeing over time in relation to land use change. This was epitomised by the outcomes of a choice modelling exercise with residents in the Hunter Valley Coalfields of New South Wales, in relation to the proposed extension of an open-cut coal mine. Community well-being declined with increased clearing of endangered ecological communities, loss [59] of highly significant Aboriginal sites, and displacement of rural families as a result of the mine's extension, but increased with the length of time that the mine provided employment and with the planting or protection of endangered ecological communities as offsets [32].
Both articles on the relationship between traditionally-owned ancestral land and health showed a positive association [22,27]. Traditionally-owned ancestral lands are unlike other LULUC categories in this scoping study in that they are not focussed on a change from unconverted to converted landscapes/ecologies/ecosystems which may impact human health and wellbeing. Rather, they evaluate the relationship between access to this 'land use type,' and health. While all of Australia is technically 'traditionally-owned'-ancestral land was never ceded to colonists when they settled in Australia in the late eighteenth century-only a proportion of the continent is currently recognised in common law as being traditionally-owned. Access to ancestral lands for Aboriginal and Torres Strait Islander peoples is either through a grant of freehold or perpetual lease title (land rights) or through native title which arises as a result of the "recognition of pre-existing Indigenous rights and interests according to traditional laws and customs" [60]. In the case of native title, Indigenous groups may have customary rights to areas with other land tenure, such as agriculture or National Park land [22]. By implication, conversion of traditionally-owned ancestral land, while it may not alter access, would have implications for the health of the traditional owners themselves, given the inextricable link between the health of the land and the health of Indigenous Australians [59,61,62].

Limitations of the Scoping Study Methods
We used a systematic and comprehensive search process, and adhered to an established scoping study framework. Notwithstanding this, the broad nature of the research question, and the fact that the research topic necessarily involves multiple scientific disciplines-often employing different terminology-meant that it was possible that some relevant articles were not captured in our search process. Within those limits, we nonetheless were able to fulfil the aims of our scoping study to summarise the range of current research reports and the relationship between LULUC and health in Australia, and identify broad gaps and implications for practice and policy.

Limitations of the Literature Reviewed
The evidence on the relationship between LULUC and human health in Australia is necessarily complex, with a multiplicity of land use types, health outcomes, and methods used. Some studies compared converted and unconverted land use types in relation to a health outcome, while others focussed on the relationship between one already converted land use-type/proposed land use change, in relation to a health outcome. Primarily because of the nature of the topic, there were no studies assessing causation, only association, which limits the strength of the conclusions that can be drawn from the findings. Human health states have multiple, interacting causes and comorbidities, and effects of a particular type of LULUC may only be relevant to that particular place, with findings unable to be generalised. The study of LULUC is also inherently complex, both in terms how it interacts with other factors-especially climate change-to impact health, and the ways in which it can impact health. Landscape changes can both exacerbate climate change, for example, deforestation which reduces carbon dioxide sinks can contribute to warming, and/or LULUC can be exacerbated by climate change (for example, loss of vegetation from drought) [3,4]. In combination, LULUC, climate change, and stratospheric ozone depletion can impact human health and wellbeing either directly (for example, through heatwaves or exposure to pollutants), indirectly (for example, through livelihood loss or population displacement), and/or through the mediation of compromised ecosystems and ecosystem 'goods and services' (EGSs), such as reduced food yields or altered infectious disease risk [2,17]. Unless climatic factors (which measured at a single point in time may not necessarily be representative of anthropogenic climate change) are measured alongside LULUC, it may be difficult to confidently attribute the measured health impacts to LULUC alone. Two of the studies we reviewed, did also measure climatic factors. Haberle et al. (2014) assessed the macroecology of airborne pollen in urban areas, Ref. [42] and reported that the only statistically significant factor explaining the difference between airborne pollen in each of the 11 sites studied was minimum temperature and mean annual precipitation [42]. Similarly, it was found that the critical ecosystem services functions associated with decomposition and nutrients cycling declined with increasing aridity (dryness/lack of precipitation), and that the effects of aridity were of a greater magnitude than any effects due to grazing [50].

Gaps and Implications for Research, Practice, and Policy
The current scoping study adopted a broad, all-LULUC-all-health outcomes-in-Australia approach, with the primary purpose of summarising and disseminating the findings and the full range of research. It also sought to identify research gaps; the fulfilment of which might ultimately inform policy and practice. In deliberations over a specific land management case-for example, land clearing/deforestation to facilitate agricultural development-practice and policy is likely to be best served by a synthesis of evidence of the impacts on health of that particular LULUC type. While there were 14 papers that described the health impacts of agriculture and grazing (including dryland salinity, a consequence of both), the studies were largely cross-sectional and therefore unable to ascribe causation. In spite of the large impact that agriculture has on water availability and water quality, this remains a large gap in the literature. For example, we found no studies examining the impact of loss of biodiversity, wetlands and rivers on health, in spite of these being major consequences of irrigation [63,64]. We also did not find any studies examining the health impacts of agriculture on microbiological water quality (for example, Cryptosporidium parvum contamination), [65] or any studies investigating the link between agriculture, land degradation and human dust exposure [66][67][68].
Similarly, more prospective studies in relation to the impact of mining are required, together with those that examine longer term impacts. Studies related to the health impacts of urban incursions into natural areas are also limited. General research gaps include geographical representation across and within all states and territories in Australia, and work on the health effects of LULUC which impacts food yield, quality of food, livelihood loss, and population displacement.
The work is specific to the Australian context, but the scoping study methods are necessarily relevant to all contexts. General findings from the current study may be reflective of the state of the evidence for other contexts and would depend on the nature of the research question; for example, whether a broad all LULUC-types and health outcomes approach was taken-as in the current study.

Conclusions
The relationship between land use/land use change and human health in Australia is not conclusive from the existing evidence. This is because the topic has not been investigated comprehensively across the whole continent. We also found conflicting results between some of the studies, perhaps reflecting the complexities of the relationships between LULUC and human health. More high-quality, specific, end-user driven research is needed to support land management decisions in Australia-in particular, to consider the human health and the ecological implications of land use changes. During a time of rapid deforestation in many parts of Australia, this work is critically important.