Recent climate-driven ecological change across a continent as perceived through local ecological knowledge

Documenting effects of climate change is an important step towards designing mitigation and adaptation responses. Impacts of climate change on terrestrial biodiversity and ecosystems have been well-documented in the Northern Hemisphere, but long-term data to detect change in the Southern Hemisphere are limited, and some types of change are generally difficult to measure. Here we present a novel approach using local ecological knowledge to facilitate a continent-scale view of climate change impacts on terrestrial biodiversity and ecosystems that people have perceived in Australia. We sought local knowledge using a national web-based survey, targeting respondents with close links to the environment (e.g. farmers, ecologists), and using a custom-built mapping tool to ask respondents to describe and attribute recent changes they had observed within an area they knew well. Results drawn from 326 respondents showed that people are already perceiving simple and complex climate change impacts on hundreds of species and ecosystems across Australia, significantly extending the detail previously reported for the continent. While most perceived trends and attributions remain unsubstantiated, >35 reported anecdotes concurred with examples in the literature, and >20 were reported more than once. More generally, anecdotes were compatible with expectations from global climate change impact frameworks, including examples across the spectrum from organisms (e.g. increased mortality in >75 species), populations (e.g. changes in recruitment or abundance in >100 species, phenological change in >50 species), and species (e.g. >80 species newly arriving or disappearing), to communities and landscapes (e.g. >50 examples of altered ecological interactions). The overarching pattern indicated by the anecdotes suggests that people are more often noticing climate change losers (typically native species) than winners in their local areas, but with observations of potential ‘adaptation in action’ via compositional and phenological change and through arrivals and range shifts (particularly for native birds and exotic plants). A high proportion of climate change-related anecdotes also involved cumulative or interactive effects of land use. We conclude that targeted elicitation of local ecological knowledge about climate change impacts can provide a valuable complement to data-derived knowledge, substantially extending the volume of explicit examples and offering a foundation for further investigation.


Introduction
Here we summarize in detail the content of the anecdotes from the Recent Ecological Change in Australia survey. The text is structured from organism to landscape scales (Sections 1-5) according to the ecological impact framework of Bellard et al. (2012) and Scheffers et al. (2016) (adapted for this study in Table 1 main text, Tables A-J in S1 Table). Codes against anecdotes (e.g. 316e) cross reference to Tables A-J in S1 Table (see separate excel file) and to data on the CSIRO Data Access Portal. Attribution of quoted anecdotes is provided using wording requested by the relevant respondents.

Organisms: genetics
While genetic adaptation is expected to be an important response of species to climate change (e.g. Sgro et al. 2011), such changes are not readily observable. Consistent with this, only one anecdote described a case of potential genetic change associated with climate change. This case involved plant hybridization in Mullumbimby, north coast New South Wales: 'Regenerati [ng] (Scheffers et al. 2016).

Organisms: physiology: mortality
Individual organisms may respond to climate stressors through plastic responses in phenotypic traits, including both changes in morphology and physiology. For example, plants can adjust to hotter or drier conditions by altering stomatal conductance, water use efficiency or specific leaf area (Nicotra et al. 2010), and animals may acclimatize to higher thermal optima or adjust their body size (Gardner et al. 2014). While morphological change would potentially have been observable by respondents, none of the anecdotes reported this type of change. Rather, effects of physiological stress on individuals were most evident where physiological thresholds had apparently been breached, resulting in die-back, mortality or increased pests and diseases in individuals. Consistent with this, a range of respondents reported mortality events and changing pest and disease loads in plants and animals, purportedly arising from direct or indirect climate stressors.
Plant mortality. Plant species reported to have been dying as a likely or possible consequence of climate change included 31 species of eucalypts (Eucalyptus and Corymbia spp.) and various species of Banksia, (Allo)Casuarina, Avicennia, Acacia, Melaleuca, Hakea, and Araucaria, as well as tree ferns (Dicksonia/Cyathea sp.) and (often unspecified) herbaceous plants (Table A in S1 Table). This mortality was sometimes reported as being followed by failed recruitment, for example after the death of Avicennia marina (Grey mangrove , 117a), Banksia marginata (Silver banksia, 12a) and Acacia carneorum (Purple-wood wattle, 124a), although replacement 'by young trees of the same species' (174a, Peter Haselgrove) was reported in conjunction with (unattributed) death of red gums and Eucalyptus caliginosa (Broad-leaved stringybark) in otherwise healthy forests of Storm King in the Darling Downs area, Queensland.
Some of the anecdotes for tree mortality were indicative of deaths at substantial scales. For example a respondent reporting from Harcourt in the Central Highlands of Victoria observed that after many years of drought, 'This past year we have lost a whole VALLEY of large eucalypts and box trees for no apparent reason ' (49b,Anon.), and monitoring in Melaleuca woodland of the Princess Charlotte Bay area (of far north Queensland) revealed that by 2016 'There were tens of thousands of dead trees and in some areas the salt edge had moved inland over 500m ' (27a,Simon Thompson and the Lama Lama Land Trust). Similarly, an anecdote for the Gulf of Carpentaria region of the Northern Territory described that the already well-reported death of mangroves (Duke et al. 2017)   The most common purported drivers in anecdotes describing plant deaths were warming (especially heatwaves) or drying (water stress), and also included salinization associated with sea-level rise (particularly affecting Melaleuca spp.) and other extreme weather events ( Table A in S1 Table). The latter included floods, storms, heavy snow, severe frosts and rainfall-driven grass-fire cycles. For example, death of emergent old Hakea lorea subsp. lorea (Long-leaf corkwood) trees was reported as resulting from hot fires on dry hills invaded by Triodia spp. (Spinifex) and Cenchrus pennisetiformis (Cloncurry buffel grass) after unusually wet years (associated with variation in the northern monsoon, 25a, north-west Queensland).
A number of respondents also reported that climate stressors increased mortality via pests or diseases, including Banksia integrifolia (Coast banksia) succumbing to a virus (89a, Hampton/Sandringham, Melbourne area, Victoria), Banksia robur (Swamp banksia ) succumbing to insect pests after long heatwaves, droughts and heavy storms (326a, Cooloola Cove, Wide Bay-Burnett area, Queensland), fungal parasites (Phellinus sublamaensis, White rot) killing stressed saplings in subtropical riverine rainforest (316a, Mullumbimby, north coast New South Wales), and invertebrate herbivory, gingering, canker and Phytophthora cinnamomi (Phytophthora root rot) in eucalypts. Other respondents reported interactions between climate and land use drivers, including fragmentation, urbanization, livestock grazing, recreational impact, water flow regulation and lowered water tables (Table A in S1 Table).

Plate 2 White gum (Eucalyptus viminalis) gingering and dying at accelerating rates in northern Tasmania (173a). Image credits: Anna Povey.
The most common consequence reported as resulting from plant deaths was habitat loss for animals. For example, observed death of 90% of the woodland mid-storey species Melaleuca decussata (Cross-leaf honey myrtle) was purported to have resulted in declines in small nesting birds such as blue wrens and thornbills, and less food for honeyeaters (243b, Junortoun, central Victoria). Mortality of Banksia robur (Swamp banksia) was perceived as a driver of the disappearance of small possums and birds in Wallum heath (326a, Cooloola Cove, Wide Bay-Burnett area, Queensland), and canopy thinning due to tree mortality was the perceived cause of the disappearance of small birds such as Malurus lamberti and M. melanocephalus (Variegated and Red-backed fairy-wrens), Acanthiza pusilla (Thornbill), Sericornis sp. (Scrubwrens) and robins (Family Petroicidae) (151f, Gunderlong, Clear Mountain, south-east Queensland). One anecdote described a cascade of hydrological and biotic consequences resulting from the unprecedented drought-and heatwavedriven fires killing the keystone tree Eucalyptus delegatensis (Alpine ash), including loss of litter and associated lizards, loss of tree hollows, and warming of alpine streams leading to decline in trout (Family Salmonidae) (see also 3.2.9 'Landscape processes', 186a, North East Catchment Management Region, Victoria).
Animal mortality. Anecdotes of animals dying attributed to climate change-related drivers involved all the major animal groups, including invertebrates, fish, amphibians, reptiles, birds and mammals, with drivers of mortality including drying out of lakes and wetlands, salinization, extreme heatwaves, firestorms, and changed conditions leading to food shortages or increased disease (Table B in S1  Table). For example, warming and increased drying of lakes and wetlands were reported as driving increases in death of fish (Hardyheads, Family Atherinidae) in warm, deoxygenated water of Lake George (75c, Beachport, south-east South Australia); and up to 10,000 Emydura macquarii (Murray River turtle) died after Lake Numalla in western Queensland dried out twice within a 10 year period (193a). Extreme heatwave events were the perceived drivers of mass death of Pteropus poliocephalus and P. alecto (Grey-headed and Black flying fox) in the Hunter and North Coast areas, New South Wales (88b), and two independent reports of mortality in Stagonopleura guttata (Firetail finch) (161f, Lake Burrendong area, central west New South Wales; 239d, Tharwa, Australian Capital Territory). Lack of moisture in eucalypt leaves was a suggested direct driver of death of Phascolarctos cinereus (Koala) (107d, Strathbogie Ranges/Merton, central Victoria, see also impacts on koala disease in 3.2.5), and salinization of ponds and groundwater associated with sea-level rise was reported as driving mass death of tadpoles (e.g. Limnodynastes dumerilii, Banjo frog) in the Derwent River estuary in Tasmania (180a).
Plate 3 Mass mortality of Pteropus poliocephalus and P. alecto (Flying foxes) due to heat stress in the Singleton area, New South Wales, February 2017 (88a, b). Image credit: Von Naftel.
Mortality from lack of food was perceived to arise through a number of mechanisms. For example, Litoria cyclorhyncha (Spot-thighed frog) mortality in Ewlyamartup, Great Southern Western Australia, was attributed to reduced supply of mosquitos and bugs due to lack of rain (113e), starvation of Rusa unicolor (Sambar deer) was reported due to lack of sufficient food (forest understorey was severely browsed) in combination with extremely cold nights (85a, Yarra Ranges National Park, Central Highlands, Victoria), and starvation of young owls (Order Strigiformes) in Mullumbimby, north coast New South Wales (316b) and of juvenile Pteropus poliocephalus and P. alecto (Flying foxes) in the Hunter and North Coast areas, New South Wales (88a) were independently reported as resulting from changed flowering times. In the case of Pteropus poliocephalus and P. alecto, flowers and fruits were the direct food source, whereas for owls (Order Strigiformes), mortality was attributed to an ecological cascade where earlier flowering times led to earlier breeding of Petaurus breviceps (Sugar glider), with consequent mistimed supply of prey for young owls (316b, Mullumbimby, north coast New South Wales). Trichosurus arnhemensis (Northern brushtail possum) were reported as dying at an unprecedented rate from stress dermatitis in association with warmer conditions (97a, Darwin, Northern Territory).
Importantly, flow-on effects of these animal deaths were also occasionally described. For example, death of Hardyheads (Family Atherinidae) at Lake George (Beachport, south-east South Australia) was reported to lead to a steady decline in migratory wading birds (75c). From a social perspective, starvation of Pteropus poliocephalus and P. alecto (Flying foxes) was proposed as potentially driving their increased colonization of urban areas (88b, Hunter and North Coast areas, New South Wales), but there were otherwise few reported interactions with land use drivers ( Table B in S1 Table). Outcomes of increases in pests and diseases are further discussed under 3.2.10 (Interspecific Relationships).

Populations: phenology
Life history processes of many species, including migration, breeding, leaf emergence and flowering, are closely linked with seasonal and inter-annual patterns in temperature and precipitation. Changes in such phenologies are some of the most commonly reported impacts of climate change worldwide, with evidence derived particularly from agricultural records and documentation of biotic events indicating the arrival of spring in the northern hemisphere (Parmesan 2006 (113d). A suite of anecdotes reported failed or reduced fruiting (and sometimes mortality) due to shortened growing seasons. These included annuals dying due to lack of rain before they could set seed (313b, Stratton, Perth area, Western Australia), and failed fruiting in sedges, orchids and other herbs due to inadequate season length before drought sets in (119b, Tamborine Mountain, southeast Queensland; 49a, Harcourt, Central Highlands Victoria; 32a, Plenty Gorge Park, Bundoora, Melbourne area, Victoria). One respondent described the consequences of shortening at both ends of the growing season as follows:

'Caladenia amoena (Charming Spider-orchid) is re-emerging later in the year, on average by two weeks, and the length of time above ground has reduced with the season also finishing between two to four weeks early consistently. Other plants also showing this pattern, late emerging plant[s] rarely
re-emerge the following year (ie they die). The shortening seasons appear to be impacting on the ability of the plants to replace their tubers, has resulted in reduced flowering, reduced ability to produce seed and a greatly reduced ability to recruit seedlings into the population' (32a, Plenty Gorge Park, Bundoora, Melbourne area, Victoria, Garry French).
Other flow on consequences of changes to flowering or fruiting to animals included food shortages causing starvation in Pteropus poliocephalus and P. alecto (Flying Foxes) and young owls (Order Strigeformes), as described under 3.2.4 'Animal mortality'.
Decreases or increases in growing season length were also occasionally reported as driving decreases or increases in ecosystem productivity. For example, reduced growth of grasses due to shorter cool-season growth periods was reported for the Lake Burrendong area, central west New South Wales (161c), and a notable increase in pasture productivity was reported in the Kangaroo Valley (south coast New South Wales): Climate stressors can influence plant and animal population dynamics directly or via effects on mortality, disease or phenology as described above, to result in observable effects on recruitment, population age structure, abundances and sex ratios (Scheffers et al. 2016). Anecdotes revealed numerous examples of increasing and decreasing levels of recruitment and abundance. Although only one anecdote referred explicitly to population age structure (mass mortality in Emydura macquarii, Murray River turtle, at Lake Numulla, south-west Queensland, led to significant change in the age demographics, 193a), and none referred to sex ratios; examples regarding recruitment and abundance may inherently involve such changes.
Plant recruitment and abundance. Anecdotal information included numerous cases of unusual levels of plant establishment or changes in plant abundance that respondents attributed as potentially due to climate change ( Table F in S1 Table). These anecdotes were commonly associated with other changes, including plants dying or new species arriving in the area. Notably, anecdotes describing declining plant establishment pertained solely to native species, particularly trees such as Eucalyptus delegatensis (Alpine ash), Eucalyptus populneus (Poplar box), Eucalyptus camaldulensis (assumed, Red gum), mangroves, Acacia carneorum (Purple-wood wattle), Banksia marginata (Silver banksia), Callitris collumellaris (Murray pine) and Allocasuarina luehmannii (Buloke). The native grasses Imperata cylindrica (Blady grass) and Triodia sp. (Spinifex) and unnamed shrub species were also observed to be regenerating poorly. Drier and/or hotter conditions, sometimes combined with land use drivers or exotic invasions, were the most common stated causes of declining plant recruitment. Mistletoe (and with it Dicaeum hirundinaceum, the Mistletoebird) was also seen to be declining in association with the millennium drought; as for a range of cases nominating the millennium drought, it is uncertain whether this was exacerbated by climate change, 188d, Bendigo, north central Victoria).  Cascading ecological consequences of these changes in plant establishment or abundance were occasionally mentioned. Examples included a perceived decline in Macropus rufogriseus (Rednecked wallaby) and Wallabia biocolor (Swamp wallaby) due to sparser understoreys associated with poor establishment of Imperata cylindrica (Blady grass) and other native grasses (151e, Gunderlong, Clear Mountain, south-east Queensland), and disease and decline in Phascolarctos cinereus (Koala) associated with limited availability of young, nutritious eucalypt leaves, in turn attributed to prolonged droughts and heatwaves limiting seed germination and survival of Eucalyptus populnea (Poplar box) in the Oakey area of south-east Queensland (192a).

Image 6 Phascolarctos cinereus (Koala) in a Eucalyptus populnea (Poplar box) tree in the Oakey district, where
Phascolarctos cinereus were reported to have declined substantially, partly attributed to prolonged droughts and heatwaves limiting Eucalyptus populnea seed germination, recruitment and leaf quality (192a). Image credit: N. Laws.
Animal abundance. Anecdotes describing changes in animal abundance were often directly attributed to warming and/or drying, including numerous cases of decline in aquatic species due to drying out of wetlands or dams (insects, fish, frogs, lizards), and more direct negative impacts of drying and warming on moths (Order Lepidoptera), mosquitoes (Family Cuculidae), leeches (Subclass Hirudinea), snakes (Suborder Serpentes), Tiliqua rugosa (Bobtail/Shingleback), Pogona minor (Western bearded dragon), at least 13 bird species, and mammals such as possums (Family Phalangeridae), Tachyglossus aculeatus (Short-beaked echidna) and Antechinus spp. (Antechinus) ( Table G in S1 Table). For example, decline in moths (Order Lepidoptera) in the Bulga, Hunter Valley area, New South Wales, region was attributed to early emergence on unseasonably warm winter days, followed by mortality with return of cooler weather (300a). Another direct climate-change driver involved increased salinity in coastal wetlands leading to decline in birds (180b, Derwent River Estuary, southern Tasmania; Table J in S1 Table).
Decreases in animal abundance were also indirectly attributed to climate change through ecological cascades, via impacts on vegetation, food supplies or other processes. For example, decline in Thynine wasps (Family Thynnidae) in Box Ironbark forests of Victoria was attributed to declines in the local Hakea and ground-layer plants (32b, Plenty Gorge Park, Bundoora, Melbourne, Victoria), decline in trout (Family Salmonidae) in the North East Catchment Management Region, Victoria, was attributed to warming of mountain streams resulting from extreme drought-and fire-driven loss of Eucalyptus delegatensis (Alpine ash) forests (see above, 186a), decline in Artamus cyanopterus (Dusky woodswallow) was attributed to competition from other Artamus (Woodswallow) species seeking refuge from persistent warming (161e, Lake Burrendong area, central west New South Wales) and decline in butterflies (Order Lepidoptera) was attributed to disconnects between the timing of food plants' flowering and butterfly emergence (316b, Mullumbimby, north coast New South Wales). While most reported decreases in animal abundance involved native species, declines in Sturnus vulgaris (European starling) and Vulpes vulpes (European red fox) were also reported (28b, Cape Paterson, Gippsland, Victoria; 87b, Logan, south-east Queensland). Perceived Vulpes vulpes decline in Logan, for example, was also attributed to an ecological cascade: 'The warmer weather has led to an increased season for paralysis tick [Ixodes holocyclus] and this has resulted in a marked decline in fox numbers' (87b, Anon.).
Anecdotal reports of increases in animal abundance were notably less common than decreases in abundance (Figure 4 main text). They typically involved increases in common, larger birds and mammals such as parrots (Licmetis sp., Corella; Eolophus roseicapilla, Galah; Cacatua galerita, Sulphur-crested cockatoo), Ocyphaps lophotes (Crested pigeon); Nycticorax caledonicus (Nankeen night heron), Rusa unicolor (Sambar deer), Macropus giganteus (Eastern grey kangaroo) and Wallabia bicolor (Swamp wallaby), typically due to warming and/or drying. A notable exception was a purported increase in small birds, small marsupials and reptiles, and subsequent increase in raptors, attributed to improved plant productivity (see 3.2.6, 240a, Gardners Bay, southern Tasmania; Table J in S1 Table). Other increases in animal abundance involved a suite of invertebrates, including beetles (Order Coleoptera), locusts (Family Acrididae), psyllids (Family Psyllidae) and scale (Superfamily Coccoidea) purported to be increasing due to warming or drying, increased numbers of termite (Infraorder Isoptera) nests due to increased drought and heat stress in eucalypts (see also 3.2.5), and swarming of insects due to larger rainfall events and localized flooding.
While cascading effects were common drivers of changes in fauna abundance, changes in fauna abundance were also occasionally reported as drivers of further ecological change. This includes increases in raptors due to increases in small prey animals as mentioned earlier (240a, Gardners Bay, southern Tasmania; Table G in S1 Table), and disappearance of Westralunio carteri (Freshwater mussel), due to decline in fish they rely on for dispersal and reproduction (86a, Swan Coastal Plain, Western Australia). Another such cascade involved thynine wasps (Family Thynnidae) as an intermediary to effects on plant pollination: 'Key species required for the associated thynine wasp pollinator (local Hakea and the grassy field layer) are also being impacted upon through the drier conditions, thus potentially reducing the ability of the orchid [Caladenia amoena (Charming Spiderorchid)] to be naturally pollinated.' (32b, Plenty Gorge Park, Melbourne, Victoria, Garry French).
Although respondents commonly attributed changes in animal abundance to both land use and climate-change drivers in Part 1, explicit mention of interactions among drivers in anecdotal text was only occasional. These included potential exacerbation of reptile decline in the Perth region by poachers (313a, Stratton, Western Australia), exacerbated decline in small birds and wallabies in Gunderlong, Clear Mountain, south-east Queensland, due to increased traffic and human habitation (151e), and interactions between climate drivers, fire suppression and cat predation contributing to the decline of the Pachycephala rufogularis (Red-lored whistler) in the Riverland Biosphere Reserve, South Australia (201b). Reported interactions with climate change resulting in increases in animal abundances include increased grain spillage benefiting large parrots (Eolophus roseicapilla, Galah; Cacatua galerita, Sulphur-crested cockatoo; 7c, Bungendore, south-east New South Wales), potential interactions between removal of cattle grazing and increasing Rusa unicolor (Sambar deer) in treeless alpine vegetation (82a, Central Highlands, Victoria), and logging and clearing contributing to increases in Macropus giganteus (Eastern grey kangaroo) and Wallabia bicolor (Swamp wallaby, 220a, Moruya, south coast New South Wales).

Species: distribution
Another common and rapid response of species to climate change is to track changing environments by shifting their distributions. This is especially apparent for mobile species, with many examples pertaining to birds, freshwater fish and marine taxa, whereas there can be greater time lags between climate change and plant responses (Scheffers et al. 2016). From the perspective of observing change in specific area or location, shifts in species distributions (including total extinction) are evident through disappearance of species previously known to be present, and appearance of new species.
Species disappearing. More than forty different species, encompassing invertebrates, amphibians, reptiles, birds, mammals, plants and fungi (to different levels of taxonomic resolution) were reported anecdotally as having disappeared in recent years potentially in association with climate change-related drivers ( Table H in S1 Table). Similar to observations of decline in native but not exotic plants, these reports pertained solely to native species.
Disappearances and concurrent appearances of species associated with rainfall decline were also implied in an anecdote from south-western Australia 'Forest species have moved west, wheat belt species are moving in from the east. Rainfall has gone from 21 inches a year to 14 inches a year.' (17d, Trevor Bunce).
Warming alone was attributed as a driver for apparent disappearance of the Western Australian endemic shrub Pimelea spectabilis (Bunjong,191a), and of Pogona sp. (Bearded dragon) in Junortoun, central Victoria: 'Bearded dragons were always around previously but not sighted at all for a few years. Climate change by way of warming seems to be the major culprit.' (243c, Anon.), but warming was more commonly considered to act in concert with drying. This included interactions with intense fire, such as disappearance of lizards due to loss of the litter layer after repeated unprecedented fires across Alpine Ash landscapes (186a, North East Catchment Management Region, Victoria), and catastrophic extinctions of Thylogale stigmatica (Red-legged pademelon) and Notamacropus dorsalis (Black-striped wallaby) in hot dry years after unprecedented wildfire in the Forty Mile Scrub National Park, south-east Queensland (166a). Warming and drying effects on animals were also commonly mediated by direct effects on plants, including disappearance of small birds such as Malurus melanocephalus (Red-backed fairywren) and Malurus lamberti (Variegated fairywren) at Gunderlong, Clear Mountain, south-east Queensland due to drought and heatwaveinduced thinning of bushland habitat (151f), disappearance of Dicaeum hirundinaceum (Mistletoebird) from the same location and Bendigo, north central Victoria, when mistletoes disappeared (151g, 188d), and disappearance of gliders (Family Petauridae) from Strathbogie Ranges/Merton, central Victoria, due to lack of moisture in stressed eucalypts (on which they feed) (107d).
New species arriving. Anecdotal evidence describing new plant species arriving in association with potential climate-change drivers pertained almost exclusively to exotic species (13 species, Table I in S1 Table), consistent with the contention that climate change is expected to increase exotic invasions (Walther et al. 2009 On the other hand, appearance of new animal species most commonly pertained to native species (16 native species, Table I in S1 Table), with only one exotic bird (Acridotheres tristis, Indian myna, 7c, Bungendore, south-east New South Wales) and two presumed exotic insects (a moth, 170a, Eppalock, north central Victoria; Vespula germanica, European wasp, 292d, Mt Lofty Ranges, South Australia) reported as arriving in association with climate change. Most of the native species arriving were birds (13 species) including the Trichoglossus moluccanus (Rainbow lorikeet) in Bullen Range Nature Reserve, Australian Capital Territory, and Bungendore, south-east New South Wales, attributed potentially to milder winters, urbanization, and/or early eucalypt flowering (5b, 7c), and Lichmera indistincta (Brown honeyeaters) and Aprosmictus erythropterus (Red-winged parrots) in Dubbo, central west, New South Wales: 'Brown Honeyeaters and Red-winged Parrots now regular migrants and residents (respectively). Not the case 20 years ago' (163a, Tim Hosking). Other animals arriving included four invertebrate species ̶ a Chickweed-dependent moth (Order Lepidoptera), Vespula germanica (European wasp), fireflies (Family Lampyridae) and Ixodes holocyclus (Paralysis tick), the Litoria caerulea (Green tree frog), and several mammal species (Phascolarctos cinereus (Koalas), Flying foxes (Family Pteropodidae), Oryctolagus cuniculus (European rabbit) ( Table I in S1  Table). For example: National Park prior to 2009(1974-2009. However, following heavy rains in 2010, green tree frogs were observed to colonise and successfully breed at the creeks and ephemeral water holes on the park. They appear to have established in the area in the within the last decade. This may be a result of changes in temperature and rainfall patterns associated with climate change' (185a, Anon.).

'According the NSW NPWS Atlas, no Green Tree Frogs [Litoria caerulea] were recorded at Willandra
About one third of new arrivals applied to species appearing at higher elevation due to warming. This included two of the invertebrate species, in particular, one respondent reporting on Lowanna, north coast New South Wales, noted: 'Over the past 20 years have watched the fireflies migrate up our mountain' (272a, Anon.), and another noted that Ixodes holocyclus (Paralysis tick, predominantly a coastal lowland species) is 'now found in our area of the tablelands' (296d, Mongarlowe, Southern Tablelands, New South Wales, Anon.). Other species observed arriving at higher elevation included Artamus personatus and Artamus superciliosus (Masked and White-browed woodswallows) 'seeking water and habitat in a cooler environment' (161e, Lake Burrendong area, central west New South Wales, Long term observations by Neville Mattick/Hargraves NSW), and Phascolarctos cinereus (Koala) moving up the mountain away 'from the increasingly hotter and habitat-depleted lowlands below' (95e, Tamborine Mountain, south-east Queensland, Anon.). Notably, Menura sp. (Lyre birds) were reported as moving in the opposite direction: 'arriving from higher on the mountain only in the last few years, moving to moist gullies' due to drier, more erratic rainfall (33b, Grassy Head, north coast, New South Wales, Anon.

Communities and ecosystems
Changes other than those described in relation to the prior nine primary questions pertained mostly to community, ecosystem or biome level outcomes described in the climate change impact frameworks of Bellard et al. (2012) and Scheffers et al. (2016), representing some of the more complex ecological impacts of climate change that can be difficult to predict (Walther 2010). Here we discuss these higher-level changes under four main categories: impacts on interspecific and intertrophic interactions, community productivity, ecosystem structure and composition, and landscape-scale ecological processes.
Interspecific and intertrophic relationships. Direct effects of climate change on individuals, populations and species are expected in turn to impact (positively or negatively) on other species that they interact with in ecological communities (Parmesan 2006, Bellard et al. 2012). Text anecdotes elucidated over 50 cases of potentially climate change-induced changes to interspecific or intertrophic relationships, through descriptions of ecological cascades (Table J in S1 Table), a range of which have already been highlighted in earlier sections. Most involved a single cascade (i.e. a primary climate-change driven change leading to a secondary change), although 11 cases involved two cascades (three biological changes), and one case involved three cascades (four levels of biological change: warming and eutrophication-induced algal blooms leading to decline in lizards at Lake Goegorup, Western Australia, with subsequent change to insect populations, in turn affecting birds, 129a).
The reported cascades implied changes to at least eight different types of ecological interactions (Table 7 main text), the most commonly-reported type involving plant-herbivore interactions (  Table).
Changes to competitive interactions, predator-prey interactions and loss of synchronization were other occasionally reported types of change to interspecific relationships. The former typically involved suppression of local native plants by invading native or exotic plants (e.g. Ligustrum sinense (Small leafed privet) producing viable seed up to 3 months earlier in Upper Coopers Creek, north coast New South Wales, leading to increased competitiveness over native plants, 153a), but included two cases of native or exotic birds displacing native birds (e.g. Artamus personatus and A. superciliosus (Masked and White browed woodswallows) displacing Artamus cyanopterus (Dusky woodswallow) in the Lake Burrendong area, central west New South Wales, 161e). While not strictly reported as changes to interspecific relationships, a number of other species replacements were also reported (see section 'Ecosystem structure and composition', below), that may involve a competitive dynamic.
Perceived changes in predator-prey relationships included both primary climate-driven changes to predators resulting in impacts on prey (e.g. arrival of predatory bird species (Cracticus sp., Butcher bird; Dacelo novaeguineae, Laughing kookaburra) and subsequent drastic reduction or disappearance of a suite of native and exotic small birds ( Community productivity. Scheffers et al. (2016) considered changes in ecosystem productivity to be one of the most critical impacts of climate change on ecosystems, with the global trends to date resulting in a net increase in terrestrial plant growth. Respondents of our survey occasionally alluded to changes in plant growth rates or length of growing season in association with changing climate or weather conditions, implying changes in ecosystem (or sub-stratum) productivity (Table J in S1 Table). These included several cases suggesting reduced productivity, including 'poor growth of understorey species such as native grasses' due to hotter, drier conditions (151e, Gunderlong, southeast Queensland, Dr J. Blok), 'less grass growth' due to elevated growing season temperatures (161c, Lake Burrendong area, central west New South Wales, long term observations by Neville Mattick/Hargraves NSW), and reduced pasture productivity due to frosts in Logan, south-east Queensland (87a). The remaining six examples involved increasing productivity. These included increased algal blooms due to warming and eutrophication (129a)  'While rainforest encroachment into eucalypt communities in wet tropical Australia is not new, in the last 20 years it has accelerated. While fire has traditionally been used to limit rainforest establishment, I suspect there is another factor involved -increased atmospheric CO2. A suite of fire resistant, suckering rainforest species is now taking over a large proportion of the wet eucalypt communities' (100a, Anon.).

Landscape processes
While our study focused on biotic change, anecdotes also highlighted direct and indirect impacts of climate change on landscape scale processes, in particular, fire regimes, salinization and erosion (Table J in S1 Table). Examples of potentially climate-driven changes in fire regimes typically pertained to one of two types: strengthening of rain-growth-fire cycles in arid Australia, and increase in large, intense fires in mountainous landscapes of south-eastern Australia. The former were commonly associated with increases in Cenchrus ciliaris (Buffel grass) and Triodia sp. (Spinifex) productivity in increasingly wet years, and resulting in structural vegetation change (25a, far north Queensland; 90a, Simpsons Gap National Park, Northern Territory; 211a, Alice Springs, Northern Territory: see earlier descriptions). It was not always clear whether respondents considered intense fires in mountainous landscapes to be a consequence of climate change, but terminology such as 'unprecedented' and 'catastrophic', and comments including 'in Tasmania dry lightning strikes were almost unknown' (14b, Deborah Hunter, Mole Creek Caving Club), suggests this is likely or feasible (Table I in S1 Table). Indeed, a respondent working in Eucalyptus delegatensis (Alpine ash) forest of the Australian Alps noted 'I believe this is a climate change impact as there are many ecological indicators that this type of fire is unprecedented' (186a, Jim Blackney). Drivers of these unprecedented intense fires were reported to include increased dry lightning incidence (especially in Tasmania, 14b, 91a), extremely warm and/or dry summer conditions (186a, 223b), and in one case, exacerbated by the large extent of forest logging (223b). These fires were usually interpreted as leading to unprecedented vegetation change or loss, including 'dry lightning over 2 days burned. Additional examples of apparent changes in fire regime included escalating fire-invasion cycles after invasion by Acacia longifolia (Sallow wattle) into the Wimmera Region of south-west Victoria (73a), suggesting indirect effects of climate change on fire in fragmented landscapes, via its effects on land use: 'The drier conditions have allowed broad acre cropping to be undertaken in areas that were once considered unviable due to water logging or poorer soil...we are witnessing a lot more burning off as a tool to handle crop residues and in the process paddock trees and roadside vegetation is being irreparably damaged' (73c, Anon.).
Respondents occasionally noted increases in soil, stream or beach erosion in association with potential impacts of climate change on vegetation or fauna. This included soil erosion as a consequence of climate change impacts on vegetation cover, e.g. 'This overall decrease in bush density has made it more vulnerable to erosion' (151f, Gunderlong, south-east Queensland, Dr J. Blok), 'more and longer dry periods reduce the amount of feed, bake the crust and result in erosion when rain occurs' (306b, Moormbool, Central Highlands Victoria, Anon.), 'loss of understorey in bush areas...results in the following cascade: failure to hold moisture-flash floods, erosion, loss of seed bank, simplification of habitat' (164f, Stoney Creek Nature Conservation Reserve/Kara Kara National Park, south-west Victoria, Anne Hughes, President, St Arnaud Field Naturalist Club and covenanted landholder). On the other hand, stream and beach erosion were mostly reported as potential drivers of biological impacts. In particular, stream erosion was reported where increased flooding was perceived, e.g. 'Major storm/flash flooding events in summer 2012 and 2013 gouged out the creek line with major soil erosion and loss of chains of ponds' (239c at Tharwa, Australian Capital Territory, Anon.), leading to perceived consequences such as weed invasion or decline in waterway plant and animal diversity. Beach erosion was similarly noted to result in impacts on sea birds and dune vegetation (180b, 198a), but it was not clear whether respondents felt the erosion was due to sealevel rise or normal levels of beach erosion.
Finally, reports of impacts on landscape hydrology related largely to coastal examples of salinization in association with sea-level rise. These included direct contact of some vegetation with increasingly high levels of spring tides (198a), examples of mangrove invasion or retreat described above (27b, 117a, 198b), and salinization of soils, ground water, ponds and wetlands, affecting trees, birds and tadpoles (27a, 180a, b).