Biodiversity Monitoring Programs Research Articles

High Arctic ecosystem states: Conceptual models of vegetation change to guide long-term monitoring and research

Vegetation change has consequences for terrestrial ecosystem structure and functioning and may involve climate feedbacks. Hence, when monitoring ecosystem states and changes thereof, the vegetation is often a primary monitoring target. Here, we summarize current understanding of vegetation change in the High Arctic—the World’s most rapidly warming region—in the context of ecosystem monitoring. To foster development of deployable monitoring strategies, we categorize different kinds of drivers (disturbances or stresses) of vegetation change either as pulse (i.e. drivers that occur as sudden and short events, though their effects may be long lasting) or press (i.e. drivers where change in conditions remains in place for a prolonged period, or slowly increases in pressure). To account for the great heterogeneity in vegetation responses to climate change and other drivers, we stress the need for increased use of ecosystem-specific conceptual models to guide monitoring and ecological studies in the Arctic. We discuss a conceptual model with three hypothesized alternative vegetation states characterized by mosses, herbaceous plants, and bare ground patches, respectively. We use moss-graminoid tundra of Svalbard as a case study to discuss the documented and potential impacts of different drivers on the possible transitions between those states. Our current understanding points to likely additive effects of herbivores and a warming climate, driving this ecosystem from a moss-dominated state with cool soils, shallow active layer and slow nutrient cycling to an ecosystem with warmer soil, deeper permafrost thaw, and faster nutrient cycling. Herbaceous-dominated vegetation and (patchy) bare ground would present two states in response to those drivers. Conceptual models are an operational tool to focus monitoring efforts towards management needs and identify the most pressing scientific questions. We promote greater use of conceptual models in conjunction with a state-and-transition framework in monitoring to ensure fit for purpose approaches. Defined expectations of the focal systems’ responses to different drivers also facilitate linking local and regional monitoring efforts to international initiatives, such as the Circumpolar Biodiversity Monitoring Program.

Estimation of Ground Dwelling Arthropod Densities Using Pitfall Traps: Individual Based Simulations

Estimation of density of ground dwelling arthropods is important for pest monitoring, landscape restoration, biodiversity and conservation. Existing methods require high amount of time and labor. Considering recent catastrophic decline of invertebrates abundance less invasive methods should be developed. In current paper based on individual based simulations I suggest method for density estimation using pitfall traps. Method is based on “dig in effect”, when traps catch more individuals straight after installation then in following days. Simulation results demonstrated that density of abundant species can be estimated by proposed method with acceptable error range. Suggested method is best suitable for fast estimation of density of abundant species (>2 individuals/m2) or higher taxa, in a large homogeneous landscapes. It can be especially useful in biodiversity monitoring programs and for pest density estimation. Method is not suitable for a long term trapping or for estimation of density of low abundance species and surveying small or narrow habitats (ex. Littoral or riparian zone, edges of habitat). Method presented in current publication can be used for standardization of pitfall trapping and comparison of results from different habitats and periods. Method can be further improved by developing more easy and inexpensive ways of sampling and should be evaluated in the field.

Modelling Distributions of Rove Beetles in Mountainous Areas Using Remote Sensing Data

Mountain ecosystems are biodiversity hotspots that are increasingly threatened by climate and land use/land cover changes. Long-term biodiversity monitoring programs provide unique insights into resulting adverse impacts on plant and animal species distribution. Species distribution models (SDMs) in combination with satellite remote sensing (SRS) data offer the opportunity to analyze shifts of species distributions in response to these changes in a spatially explicit way. Here, we predicted the presence probability of three different rove beetles in a mountainous protected area (Gran Paradiso National Park, GPNP) using environmental variables derived from Landsat and Aster Global Digital Elevation Model data and an ensemble modelling approach based on five different model algorithms (maximum entropy, random forest, generalized boosting models, generalized additive models, and generalized linear models). The objectives of the study were (1) to evaluate the potential of SRS data for predicting the presence of species dependent on local-scale environmental parameters at two different time periods, (2) to analyze shifts in species distributions between the years, and (3) to identify the most important species-specific SRS predictor variables. All ensemble models showed area under curve (AUC) of the receiver operating characteristics values above 0.7 and true skills statistics (TSS) values above 0.4, highlighting the great potential of SRS data. While only a small proportion of the total area was predicted as highly suitable for each species, our results suggest an increase of suitable habitat over time for the species Platydracus stercorarius and Ocypus ophthalmicus, and an opposite trend for Dinothenarus fossor. Vegetation cover was the most important predictor variable in the majority of the SDMs across all three study species. To better account for intra- and inter-annual variability of population dynamics as well as environmental conditions, a continuation of the monitoring program in GPNP as well as the employment of SRS with higher spatial and temporal resolution is recommended.

Satellite-based decadal change assessments of pan-Arctic environments.

Remote sensing can advance the work of the Circumpolar Biodiversity Monitoring Program through monitoring of satellite-derived terrestrial and marine physical and ecological variables. Standardized data facilitate an unbiased comparison across variables and environments. Using MODIS standard products of land surface temperature, percent snow covered area, NDVI, EVI, phenology, burned area, marine chlorophyll, CDOM, sea surface temperature, and marine primary productivity, significant trends were observed in almost all variables between 2000 and 2017. Analysis of seasonal data revealed significant breakpoints in temporal trends. Within the terrestrial environment, data showed significant increasing trends in land surface temperature and NDVI. In the marine environment, significant increasing trends were detected in primary productivity. Significantly earlier onset of green up date was observed in bioclimate subzones C&E and longer end of growing season in B&E. Terrestrial and marine parameters showed similar rates of change with unidirectional change in terrestrial and significant directional and magnitude shifts in marine.

A Review of the Flora and Fauna in the Al Wathba Wetland Reserve in Abu Dhabi, United Arab Emirates

Wetlands play an important role in ecosystem function, and are known to support rich biodiversity, this becomes even more pronounced in hyper-arid climates. The Al Wathba Wetland Reserve in Abu Dhabi Emirate is a wetland mosaic that boasts both brackish and freshwater pools. The levels of nitrate, nitrite, phosphate and total organic carbon is much higher and this can be attributed to various factors such as the lake not having a water outlet, seepage of untreated waste water from an adjacent sewage treatment plant and various industries in the surrounding areas that could lead to these high values. A twenty-year biodiversity assessment and monitoring program, has documented a total of 354 invertebrate species, 16 reptile and amphibian species, 262 bird species, 10 mammalian species, and 39 plant species. These represent 21% of Abu Dhabi Emirate’s terrestrial animal and plant species. Out of a total of 44 marine and terrestrial habitats identified within Abu Dhabi Emirate, 6 of these are represented within AWWR.

Circumpolar patterns of Arctic freshwater fish biodiversity: A baseline for monitoring

Abstract Climate change, biological invasions, and anthropogenic disturbance pose a threat to the biodiversity and function of Arctic freshwater ecosystems. Understanding potential changes in fish species distribution and richness is necessary, given the great importance of fish to the function of freshwater ecosystems and as a resource to humans. However, information gaps limit large‐scale studies and our ability to determine patterns and trends in space and time. This study takes the first step in determining circumpolar patterns of fish species richness and composition, which provides a baseline to improve both monitoring and conservation of Arctic freshwater biodiversity. Information on species presence/absence was gathered from the Circumpolar Biodiversity Monitoring Program's Freshwater Database and used to examine patterns of freshwater fish γ‐, α‐, and β‐diversity across 234° of longitude in the Arctic. The metrics of diversity provided information on species richness and composition across hydrobasins, ecoregions, and Arctic zones. Circumpolar patterns of fish species biodiversity varied with latitude, isolation, and coarse ecoregion characteristics; patterns were consistent with historic and contemporary barriers to colonisation and environmental characteristics. Gamma‐diversity was lower in the high Arctic compared to lower latitude zones, but α‐diversity did not decrease with increasing latitude below 71°N, reflecting glacial history. Alpha‐diversity was reduced to a single species, Arctic charr Salvelinus alpinus, in ecoregions above 71°N, where γ‐diversity was the lowest. Beta‐diversity indicated little variation in the composition and richness of species across the High Arctic; at lower latitudes, ecoregions contained more species, although species composition turned over across large spatial extents. In an analysis of five ecoregions in the circumpolar Arctic, physical isolation, and ecoregion area and topography were identified as strong drivers of γ‐, α‐, and β‐diversity. Physical isolation reduced the γ‐ and α‐diversity, and changes in β‐diversity between adjacent locations were due mainly to losses in species richness, rather than due to differences in species composition. Heterogeneity of habitats, environmental gradients, and geographic distance probably contributed to patterns of fish dissimilarity within and across ecoregions. This study presents the first analysis of large‐scale patterns of freshwater fish biodiversity in the circumpolar Arctic. However, information gaps in space, time, and among taxonomic groups remain. Future inclusion of extensive archive and new data will allow future studies to test for changes and drivers of the observed patterns of biodiversity. This is important given the potential impacts of ongoing and accelerating climate change, land use, and biotic exchange on Arctic fish biodiversity.

Data Quality and Participant Engagement in Citizen Science: Comparing Two Approaches for Monitoring Pollinators in France and South Korea

Citizen science has become a mainstream approach for collecting data on biodiversity. However, not all biodiversity monitoring programs achieve the goal of collecting datasets that can be used in robust scientific inquiries. Data quality and the capacity to engage participants in the long-term are the most challenging issues. We compared two methodologies of citizen science programs dedicated to pollinators monitoring in France (Spipoll) and South Korea (K-Spipoll). These programs aimed to launch long-term monitoring at a community-level to better understand environmental effects on the composition and stability of pollinator communities. We assessed, through different metrics, how the two approaches influenced (1) data quality (assessed by “Accuracy in data collection,” “Consistency in protocol relative to volume of sessions contributed by an individual,” “Spatial representation of data,” and “Sample size”), and (2) participant engagement (assessed by “the number of connected days,” “the number of active days,” “the proportion of participant contributing a single session,” “the average number of sessions per participant,” and “the distribution of numbers of contributions per participant in each program.”). On one hand, participants in the Spipoll program abided by the standard protocol more often and provided identification for the photographed insects, leading to efficient ecological analyses. On the other hand, the K-Spipoll program provided more sessions per participant and a lower rate of single participation, with a full session demanding less effort in terms of data input, providing critical data where baseline data have otherwise been unavailable. These differences have emerged through methodology choices: For the Spipoll, the dedicated website favored the emergence of a social network that facilitated identification and increased data quality; for the K-Spipoll, the development of a cell phone application facilitated participation, and regular on-field education sessions motivated participants. We conclude by providing suggestions for the implementations of future citizen science programs to improve both data quality and participant engagement.

Circumpolar terrestrial arthropod monitoring: A review of ongoing activities, opportunities and challenges, with a focus on spiders.

The terrestrial chapter of the Circumpolar Biodiversity Monitoring Programme (CBMP) has the potential to bring international multi-taxon, long-term monitoring together, but detailed fundamental species information for Arctic arthropods lags far behind that for vertebrates and plants. In this paper, we demonstrate this major challenge to the CBMP by focussing on spiders (Order: Araneae) as an example group. We collate available circumpolar data on the distribution of spiders and highlight the current monitoring opportunities and identify the key knowledge gaps to address before monitoring can become efficient. We found spider data to be more complete than data for other taxa, but still variable in quality and availability between Arctic regions, highlighting the need for greater international co-operation for baseline studies and data sharing. There is also a dearth of long-term datasets for spiders and other arthropod groups from which to assess status and trends of biodiversity. Therefore, baseline studies should be conducted at all monitoring stations and we make recommendations for the development of the CBMP in relation to terrestrial arthropods more generally.

THE ROLE OF SPECIAL PROTECTED AREAS OF THE MURMANSK PART OF THE GREEN BELT OF FENNOSCANDIA IN THE DIVERSITY CONSERVATION OF RARE BIRD SPECIES

The contribution of the existing and planned protected areas of the Murmansk part of the Green Belt of Fennoscandia to the conservation of the populations of bird species listed in the Red Data Books of the Russian Federation and the Murmansk Region is analyzed. The history of studies of the avifauna of each protected area is considered. The areas best studied by ornithologists (Ainovy Islands of Kandalakshsky Strict Nature Reserve (zapovednik), Pasvik Strict Nature Reserve, Kaita Regional Nature Reserve (zakaznik), planned zakazniks - Pazovsky and Vuorjema) and the least studied ones (Laplandsky Les, Kutsa regional zakaznik, planned Ionn-N'yugoajv, Spruce Forests of Alla-Akkajarvi zaka-zniks. and others) have been identified. The avifauna of the study area totals 256 species, including 35 red-listed species. It is demonstrated that the protected areas do preserve a high biodiversity of birds and maintain their populations. The best performance is observed for the Whooper Swan Cygnus cygnus, Smew Mergellus albellus, White-tailed Sea Eagle Haliaeetus albicilla, Kestrel Falco tinnunculus, Common Crane Grus grus, Dipper Cinclus cinclus and Great Gray Shrike Lanius excubitor. However, for some protected areas of regional status, knowledge gaps need to be filled through implementation of special biodiversity monitoring programs, research on the state of habitats of rare bird species, and assessment of their numbers.

Status and trends of terrestrial arthropod abundance and diversity in the North Atlantic region of the Arctic.

The Circumpolar Biodiversity Monitoring Programme (CBMP) provides an opportunity to improve our knowledge of Arctic arthropod diversity, but initial baseline studies are required to summarise the status and trends of planned target groups of species known as Focal Ecosystem Components (FECs). We begin this process by collating available data for a relatively well-studied region in the Arctic, the North Atlantic region, summarising the diversity of key terrestrial arthropod FECs, and compiling trends for some representative species. We found the FEC classification system to be challenging to implement, but identified some key groups to target in the initial phases of the programme. Long-term data are scarce and exhibit high levels of spatial and temporal variability. Nevertheless, we found that a number of species and groups are in decline, mirroring patterns in other regions of the world. We emphasise that terrestrial arthropods require higher priority within future Arctic monitoring programmes.

The use and utility of surrogates in biodiversity monitoring programmes

The use and utility of surrogates in biodiversity monitoring programmes

Environmental DNA (eDNA) metabarcoding reveals strong discrimination among diverse marine habitats connected by water movement.

While in recent years environmental DNA (eDNA) metabarcoding surveys have shown great promise as an alternative monitoring method, the integration into existing marine monitoring programs may be confounded by the dispersal of the eDNA signal. Currents and tidal influences could transport eDNA over great distances, inducing false-positive species detection, leading to inaccurate biodiversity assessments and, ultimately, mismanagement of marine environments. In this study, we determined the ability of eDNA metabarcoding surveys to distinguish localized signals obtained from four marine habitats within a small spatial scale (<5km) subject to significant tidal and along-shore water flow. Our eDNA metabarcoding survey detected 86 genera, within 77 families and across 11 phyla using three established metabarcoding assays targeting fish (16S rRNA gene), crustacean (16S rRNA gene) and eukaryotic (cytochrome oxidase subunit 1) diversity. Ordination and cluster analyses for both taxonomic and OTU data sets show distinct eDNA signals between the sampled habitats, suggesting dispersal of eDNA among habitats was limited. Individual taxa with strong habitat preferences displayed localized eDNA signals in accordance with their respective habitat, whereas taxa known to be less habitat-specific generated more ubiquitous signals. Our data add to evidence that eDNA metabarcoding surveys in marine environments detect a broad range of taxa that are spatially discrete. Our work also highlights that refinement of assay choice is essential to realize the full potential of eDNA metabarcoding surveys in marine biodiversity monitoring programs.

Species turnover reveals hidden effects of decreasing nitrogen deposition in mountain hay meadows.

Nitrogen (N) deposition is a major threat to biodiversity in many habitats. The recent introduction of cleaner technologies in Switzerland has led to a reduction in the emissions of nitrogen oxides, with a consequent decrease in N deposition. We examined different drivers of plant community change, that is, N deposition, climate warming, and land-use change, in Swiss mountain hay meadows, using data from the Swiss biodiversity monitoring program. We compared indicator values of species that disappeared from or colonized a site (species turnover) with the indicator values of randomly chosen species from the same site. While oligotrophic plant species were more likely to colonize, compared to random expectation, we found only weak shifts in plant community composition. In particular, the average nutrient value of plant communities remained stable over time (2003–2017). We found the largest deviations from random expectation in the nutrient values of colonizing species, suggesting that N deposition or other factors that change the nutrient content of soils were important drivers of the species composition change over the last 15 years in Swiss mountain hay meadows. In addition, we observed an overall replacement of species with lower indicator values for temperature with species with higher values. Apparently, the community effects of the replacement of eutrophic species with oligotrophic species was outweighed by climate warming. Our results add to the increasing evidence that plant communities in changing environments may be relatively stable regarding average species richness or average indicator values, but that this apparent stability is often accompanied by a marked turnover of species.

Ranacapa: An R package and Shiny web app to explore environmental DNA data with exploratory statistics and interactive visualizations

Environmental DNA (eDNA) metabarcoding is becoming a core tool in ecology and conservation biology, and is being used in a growing number of education, biodiversity monitoring, and public outreach programs in which professional research scientists engage community partners in primary research. Results from eDNA analyses can engage and educate natural resource managers, students, community scientists, and naturalists, but without significant training in bioinformatics, it can be difficult for this diverse audience to interact with eDNA results. Here we present the R package ranacapa, at the core of which is a Shiny web app that helps perform exploratory biodiversity analyses and visualizations of eDNA results. The app requires a taxonomy-by-sample matrix and a simple metadata file with descriptive information about each sample. The app enables users to explore the data with interactive figures and presents results from simple community ecology analyses. We demonstrate the value of ranacapa to two groups of community partners engaging with eDNA metabarcoding results.

Novel approaches to sampling pollinators in whole landscapes: a lesson for landscape-wide biodiversity monitoring

ContextBiodiversity monitoring programs require fast, reliable and cost-effective methods for biodiversity assessment in landscapes. Sampling pollinators across entire landscapes is challenging, as trapping needs to cover many habitat types.ObjectivesWe developed and tested a landscape-wide sampling design for pollinators. We assessed the predictability and stability of pollinator biodiversity estimates in agricultural landscapes, and tested how estimates were affected by sampled habitat, landscape composition and spatial scale.MethodsWe sampled pollinators using pan traps at 250 locations in 10 replicated landscapes measuring 1 × 1 km and calculated bee richness predictions based on different sample sizes. Traps were placed regularly in each landscape, sampling each habitat proportionally to its area. Landscapes contained semi-natural habitats, crop fields and forests and differed in the amount of a mass-flowering crop (oilseed rape).ResultsRegular sampling reflected local habitat amount. Compared with cereal fields, significantly more pollinators occurred in oilseed rape, and fewer in forests. Sampling in only one habitat type led to biased estimates of landscape-wide bee species richness, even when sample size was increased. The spatial scale of best predictions depended on the sampled habitat. Species richness was overestimated when sampling was limited to semi-natural habitats and underestimated in oilseed rape fields. Precision increased with the number of sampling points per landscape.ConclusionsTo study landscape-wide pollinator biodiversity, we suggest to sample multiple sites per landscape in a broad range of resource-providing habitat types, with sample sizes proportional to habitat amount. Our approach will also be useful for biodiversity monitoring programs in general.

Comparative use of active searches and artificial refuges to detect amphibians in terrestrial environments

AbstractArtificial refuges (cover boards) are commonly used to survey and monitor herpetofauna in many parts of the world. Despite the extensive use of artificial refuges in mesic environments, their effectiveness for detecting amphibians in temperate zones has rarely been examined. We compared amphibian detection probabilities between two survey methods; active searches of natural habitat and artificial refuges of three different types (corrugated steel, roofing tiles and timber railway sleepers). Our study area included five bioregions encompassing a 1180‐km latitudinal gradient across a modified, temperate eucalypt woodland vegetation community in south‐eastern Australia. We deployed 14 778 artificial refuges in terrestrial environments, within patches of remnant vegetation, and collected presence and abundance data on herpetofauna between 1999 and 2017. We used Bayesian logistic regression to identify the most effective survey method for detecting frog species across all bioregions. We modelled frog detections by fitting survey method, time since refuge deployment and rainfall prior to each survey. We detected 3970 individuals from 18 frog species. Overall, we found active searches and timber substrates most effective for detecting a broad range of species, although detection rates were driven by the numerically abundant spotted marsh frog Limnodynastes tasmaniensis. Timber refuges were effective for detecting several burrowing species, whereas active searches were effective at detecting habitat generalists. Quadratic effects of rainfall prior to survey as opposed to linear effects of time since artificial refuge placement was important in explaining frog detection rates in some bioregions. Active searches, timber railway sleepers and sheets of corrugated steel provide complimentary survey methods for detecting amphibians, although detection rates are influenced by rainfall patterns. Artificial refuges provide a time‐effective and standardized method for studying amphibians in their non‐breeding terrestrial environment and should be incorporated into future surveys and biodiversity monitoring programmes.

Ranacapa: An R package and Shiny web app to explore environmental DNA data with exploratory statistics and interactive visualizations.

Environmental DNA (eDNA) metabarcoding is becoming a core tool in ecology and conservation biology, and is being used in a growing number of education, biodiversity monitoring, and public outreach programs in which professional research scientists engage community partners in primary research. Results from eDNA analyses can engage and educate natural resource managers, students, community scientists, and naturalists, but without significant training in bioinformatics, it can be difficult for this diverse audience to interact with eDNA results. Here we present the R package ranacapa, at the core of which is a Shiny web app that helps perform exploratory biodiversity analyses and visualizations of eDNA results. The app requires a taxonomy-by-sample matrix and a simple metadata file with descriptive information about each sample. The app enables users to explore the data with interactive figures and presents results from simple community ecology analyses. We demonstrate the value of ranacapa to two groups of community partners engaging with eDNA metabarcoding results.

Higher-taxon and functional group responses of ant and bird assemblages to livestock grazing: A test of an explicit surrogate concept

Biodiversity monitoring programs are routinely established to quantify changes in biotic communities in response to land management. Surrogacy is implicitly used in many such monitoring programs whereby the measurement of a component of biodiversity is used to infer responses of broader biodiversity. Yet rarely is this surrogacy validated by demonstrating that measured variables and the target variable of interest have matching responses to management treatments. Here we examined the responses of higher-taxon and functional groupings of ants and birds (our surrogate variables) two years after the implementation of experimental livestock grazing treatments, and compared these with the responses of total ant and bird species richness (our target variables) to the same treatments. We found significant and strong correlations between surrogate and target variables, but this did not predict corresponding similar response to treatments. For ants, we found that the genus Monomorium had a negative response to the grazing exclusion treatment, but there was no matching response of species richness, and so no surrogacy was identified. For birds, total species richness had a weak positive response to spring/summer grazing exclusion, and the abundance of honeyeaters (Meliphagidae) showed a similar positive response, suggesting surrogacy. Our study highlights that correlations among variables do not necessarily lead to surrogacy, and indeed that different sub-components of biotic assemblages can respond in ways that contrast with overall species richness. Careful assessment of the matched responses of surrogate and target variables to management can provide a simple and robust way to critically assess biodiversity surrogacy.

Macrobenthic fauna from an upwelling coastal area of Peru (Warm Temperate South-eastern Pacific province -Humboldtian ecoregion)

BackgroundA total of 162 species and subspecies of marine macroinvertebrates were recorded in the submerged soft and hard substrates around the PERU LNG marine terminal and surrounding area, in the central coast of Peru, 167 km south of Lima, Peru. The collection of specimens was carried out from June 2011 to June 2015 as part of the research studies conducted by the Biodiversity Monitoring and Assessment Program (BMAP) around the marine terminal. The area is part of the Humboldt Current Large Marine Ecosystem, one of the most important upwelling systems in the world.New informationWe identified specimens belonging to 83 families and seven phyla. The list was assembled from the taxonomic identifications made by the BMAP. We identified species and subspecies belonging to phyla Annelida, Arthropoda, Brachiopoda, Cnidaria, Echinodermata, Bryozoa and Mollusca. Phyla Annelida (60 spp.), Arthropoda (47 spp.)and Mollusca (45 spp.) exhibited the largest number of species.

Control history, longitude and multiple abiotic and biotic variables predict the abundances of invasive brushtail possums in New Zealand forests

The factors affecting variation in the abundances of invasive mammal species have seldom been evaluated at large spatial scales. We used data collected in a large-scale biodiversity monitoring program to identify variables predicting abundances of common brushtail possums (Trichosurus vulpecula; ‘possum’), a marsupial folivore introduced to New Zealand in the 1800s. We sampled possum abundances, vegetation structure and composition, and abiotic variables at 395 spatially representative sites in natural forest on the North, South and Stewart islands. Records of possum control were obtained from management agencies. The large sample size enabled us to consider the relative importance of 27 variables for possum abundances. Twelve variables explained 80% of the contributions to the final boosted regression tree model. Control history had the largest contribution to the model, with sites subject to aerial 1080 poison baiting either once or more than once in the previous 7 years having possum abundances 64% and 74% lower, respectively, than sites without control. Possum abundances were also negatively related to humidity, stem density, tree cover density, slope, elevation and distance to pasture/crop. Abundances increased with increasing food plant species richness, Weinmannia racemosa (a preferred food tree) cover, and solar radiation. There was an interaction between longitude and control history. We conclude that invasive possum abundances in New Zealand’s natural forests are determined by multiple biotic and abiotic factors, sometimes in interaction. Recent control operations, however, have greatly reduced possum abundances. Increasing the spatial extent of aerial 1080 poison baiting will further reduce possum abundances in New Zealand forests.