An evaluation of stringent filtering to improve species distribution models from citizen science data

Establishing the distribution of Carpophilus truncatus in Australia using an integrative approach for an emerging global pest

The nitidulid beetle Carpophilus truncatus is rapidly becoming a major pest of nut crops around the world. This insect first infested Australian almonds in 2013 and has since escalated to be the preeminent insect pest for the industry. Data pertaining to C. truncatus distribution are scant, but without awareness of its origin, distribution, and ecological factors that influence distribution, efforts to understand and manage the insect as a pest are stymied. Here, we employ an integrative approach to gain a multifaceted understanding of the distribution of C. truncatus in Australia. Methods employed were (1) reviewing historical records in insect collections to establish the presence of C. truncatus prior to commercial almond horticulture, (2) field trapping of insects to establish presence in regions of interest, (3) laboratory trials to determine the thermal limits of the organism, and (4) correlative species distribution modelling to describe its current distribution. We find that C. truncatus is more widespread across Australia than was previously known, with historical records preceding commercial almond production in Australia by a century. The methods developed in this study can be applied elsewhere in the world where C. truncatus is an emerging pest, or to novel pest species as they arise with increasing frequency in a globalised and warming world.

Urban biodiversity conservation: A framework for ecological network construction and priority areas identification considering habit differences within species

The construction of ecological networks within the context of urbanization is an effective approach to cope with the challenges of urban biodiversity decline, representing a crucial goal in urban planning and development. However, existing studies often overlook the richness and uniqueness within species communities by homogenizing traits of species in the same class. This study proposes a framework for constructing and optimizing ecological networks focused on differential conservation within the same class. By classifying birds into three groups (specialists of water, forest or urban areas) based on their ecological requirements and urbanization tolerance, we constructed an ecological network tailored to their distinct migratory dispersal patterns. We then identified strategic areas including pinch points, barriers, and breakpoints specific to each bird group. Our findings reveal notable variations in suitable habitat distribution among different bird groups in urban environments. Corridor layouts varied according to habitat preferences and migratory dispersal patterns. Despite these differences, urban built-up areas persist as central hubs for the distribution of suitable habitats for 75% of bird species, with peripheral mountain-plain transition areas constituting 63% of crucial dispersal corridors. This emphasizes the critical role of urban built-up areas in maintaining biodiversity and ecological connectivity. Prioritizing connectivity between central urban areas and distant natural spaces is imperative. Our approach innovatively classifies and constructs networks to identify strategic areas with diverse species-specific attributes, providing valuable spatial information for land planning and guiding solutions to enhance target species. While the primary focus is on bird conservation in Beijing, our framework is broadly applicable to global biodiversity management and green planning under urbanization challenges. Overall, this study offers innovative insights for urban planning development and serves as decision support for prioritizing urban actions.

Boosting biodiversity monitoring using smartphone-driven, rapidly accumulating community-sourced data

Comprehensive biodiversity data is crucial for ecosystem protection. The Biome mobile app, launched in Japan, efficiently gathers species observations from the public using species identification algorithms and gamification elements. The app has amassed >6 million observations since 2019. Nonetheless, community-sourced data may exhibit spatial and taxonomic biases. Species distribution models (SDMs) estimate species distribution while accommodating such bias. Here, we investigated the quality of Biome data and its impact on SDM performance. Species identification accuracy exceeds 95% for birds, reptiles, mammals, and amphibians, but seed plants, molluscs, and fishes scored below 90%. Our SDMs for 132 terrestrial plants and animals across Japan revealed that incorporating Biome data into traditional survey data improved accuracy. For endangered species, traditional survey data required >2000 records for accurate models (Boyce index ≥ 0.9), while blending the two data sources reduced this to around 300. The uniform coverage of urban-natural gradients by Biome data, compared to traditional data biased towards natural areas, may explain this improvement. Combining multiple data sources better estimates species distributions, aiding in protected area designation and ecosystem service assessment. Establishing a platform for accumulating community-sourced distribution data will contribute to conserving and monitoring natural ecosystems.

Using unstructured crowd-sourced data to evaluate urban tolerance of terrestrial native animal species within a California Mega-City

In response to biodiversity loss and biotic community homogenization in urbanized landscapes, there are increasing efforts to conserve and increase biodiversity within urban areas. Accordingly, around the world, previously extirpated species are (re)colonizing and otherwise infiltrating urban landscapes, while other species are disappearing from these landscapes. Tracking the occurrence of traditionally urban intolerant species and loss of traditionally urban tolerant species should be a management goal of urban areas, but we generally lack tools to study this phenomenon. To address this gap, we first used species' occurrences from iNaturalist, a large collaborative dataset of species observations, to calculate an urban association index (UAI) for 967 native animal species that occur in the city of Los Angeles. On average, the occurrence of native species was negatively associated with our composite measure of urban intensity, with the exception of snails and slugs, which instead occur more frequently in areas of increased urban intensity. Next, we assessed 8,348 0.25 x 0.25 mile grids across the City of Los Angeles to determine the average grid-level UAI scores (i.e., a summary of the UAIs present in a grid cell, which we term Community Urban Tolerance Index or CUTI). We found that areas of higher urban intensity host more urban tolerant species, but also that taxonomic groups differ in their aggregate tolerance of urban areas, and that spatial patterns of tolerance vary between groups. The framework established here has been designed to be iteratively reevaluated by city managers of Los Angeles in order to track the progress of initiatives to preserve and encourage urban biodiversity, but can be rescaled to sample different regions within the city or different cities altogether to provide a valuable tool for city managers globally.

Supplemental structured surveys and pre-existing detection models improve fine-scale density and population estimation with opportunistic community science data

Density and population estimates aid in conservation and stakeholder communication. While free and broadly available community science data can effectively inform species distribution models, they often lack the information necessary to estimate imperfect detection and area sampled, thus limiting their use in fine-scale density modeling. We used structured distance-sampling surveys to model detection probability and calculate survey-specific detection offsets in community science models. We estimated density and population for 16 songbird species under three frameworks: (1) a fixed framework that assumes perfect detection within a specified survey radius, (2) an independent framework that calculates offsets from an independent source, and (3) a calibration framework that calculates offsets from supplemental surveys. Within the calibration framework, we examined the effects of calibration dataset size and data pooling. Estimates of density and population size were consistently biased low in the fixed framework. The independent and calibration frameworks produced reliable estimates for some species, but biased estimates for others, indicating discrepancies in detection probability between structured and community science surveys. The calibration framework produced reliable population estimates with as few as 10 calibration surveys with positive detections. Data pooling dramatically decreased bias. This study provides conservationists and managers with a cost-effective method of estimating density and population.

Quantifying error in occurrence data: Comparing the data quality of iNaturalist and digitized herbarium specimen data in flowering plant families of the southeastern United States

iNaturalist has the potential to be an extremely rich source of organismal occurrence data. Launched in 2008, it now contains over 150 million uploaded observations as of May 2023. Based on the findings of a limited number of past studies assessing the taxonomic accuracy of participatory science-driven sources of occurrence data such as iNaturalist, there has been concern that some portion of these records might be misidentified in certain taxonomic groups. In this case study, we compare Research Grade iNaturalist observations with digitized herbarium specimens, both of which are currently available for combined download from large data aggregators and are therefore the primary sources of occurrence data for large-scale biodiversity/biogeography studies. Our comparisons were confined regionally to the southeastern United States (Florida, Georgia, North Carolina, South Carolina, Texas, Tennessee, Kentucky, and Virginia). Occurrence records from ten plant families (Gentianaceae, Ericaceae, Melanthiaceae, Ulmaceae, Fabaceae, Asteraceae, Fagaceae, Cyperaceae, Juglandaceae, Apocynaceae) were downloaded and scored on taxonomic accuracy. We found a comparable and relatively low rate of misidentification among both digitized herbarium specimens and Research Grade iNaturalist observations within the study area. This finding illustrates the utility and high quality of iNaturalist data for future research in the region, but also points to key differences between data types, giving each a respective advantage, depending on applications of the data.

Comparing avian species richness estimates from structured and semi-structured citizen science data

Citizen science, including structured and semi-structured forms, has become a powerful tool to collect biodiversity data. However, semi-structured citizen science data have been criticized for higher variability in quality, including less information to adjust for imperfect detection and uneven duration that bias the estimates of species richness. Species richness estimators may quantify bias in estimates. Here, we test the effectiveness of Chao1 estimator in eBird (semi-structured) by comparing it to averaged species richness in Breeding Bird Survey Taiwan, BBS (structured) and quantifying bias. We then fit a power function to compare bias while controlling for differences in count duration. The Chao1 estimator increased the species richness estimates of eBird data from 56 to 69% of the average observed BBS and from 47 to 59% of the average estimated BBS. Effects of incomplete short duration samples and variability in detectability skills of observers can lead to biased estimates. Using the Chao1 estimator improved estimates of species richness from semi-structured and structured data, but the strong effect of singleton species on bias, especially in short duration counts, should be evaluated in advance to reduce the uncertainty of estimation processes.

The Big Five: Species Distribution Models from Citizen Science Data as Tool for Preserving the Largest Protected Saproxylic Beetles in Italy

Background. Volunteers’ participation in scientific research has increased in recent decades. Citizen science (CS) data have been used in quantitative ecology to analyse species ranges by means of species distribution models. We investigated the Italian distribution of five large saproxylic beetles (big five), to describe their niche space, paramount areas for their conservation, and conservation gaps. Methods. CS data from two projects, climate and environmental variables were used to produce Habitat suitability (HS) maps for each species and averaged HS maps. The big five’s conservation status was assessed interpolating HS maps with the distribution of protected areas, concomitantly identifying conservation gaps. Results. The pre-alpine and Apennines arcs, north-eastern Sicily and eastern Sardinia, were identified as conservation’s hotspots. Ranking HS levels from minimum to optimal, the extent of conservation gaps decreases as environmental suitability for the big five increases. Conclusions. For the first time in Italy, CS data have been used to investigate niche space of the largest protected saproxylic beetles and analyse the distribution of their suitable habitat. The resulting HS raster maps and vector layers, reporting HS value in all Italian protected areas (n° 3771), were provided and discussed, reporting an application example for conservation purposes.

Measuring Spatial Associations between Environmental Health and Beliefs about Environmental Governance

Research has shown an increasing trend in attempts to integrate social and ecological data that use indicators to improve quality of life. This includes understanding people's beliefs about environmental governance. Understanding patterns in beliefs of environmental governance can be a powerful way to help policy makers take informed actions that meet individuals' needs and expectations. This study examines connections between spatial patterns of beliefs about environmental governance and the health of the environment where people live, measured from both a public health and ecological perspective. Data on people's beliefs about environmental governance were collected in the Puget Sound area of Washington state. Environmental health data include environmental public health disparities including effects and exposures, bird diversity, and tree cover. Results indicate local scale heterogeneity exists within the Puget Sound region. Using AIC model selection, there was strong evidence for effects of canopy cover, environmental effects and exposures, and years of residency, and moderate to strong evidence for the effects on beliefs about environmental governance of race and sex. There was little support for effects of political ideology, income, age, education, or bird diversity. The Akaike Information Criteria (AIC) top model included a negative effect of canopy cover, years of residency, race (i.e., of being non-white), and sex (i.e., of being male), and a positive effect of environmental effects and of environmental exposures. Relating data on environmental health and beliefs about environmental governance generates a more nuanced understanding of determinants of environmental governance success and public support.

Diversity of Palaearctic Dragonflies and Damselflies (Odonata)

More than 1.2 million distribution records were used to create species distribution models for 402 Palaearctic species of dragonflies and damselflies. On the basis of these diversity maps of total, lentic and lotic diversity for the whole of the Palaearctic (excluding China and the Himalayan region) are presented. These maps show a clear pattern of decreasing diversity longitudinally, with species numbers dropping in the eastern half of Europe and remaining low throughout a large part of Russia, then increasing again towards Russia’s Far East and Korea. There are clear differences in diversity patterns of lentic and lotic species, with lentic species being dominant in colder and more arid areas. Areas with a high diversity of species assessed as threatened on the IUCN red list are largely restricted to the Mediterranean, Southwest Asia, and Japan, with clear hotspots found in the Levant and the southern half of Japan. The diversity at species, generic, and family level is higher in the south of Japan than in areas at a similar latitude in the western Mediterranean. This is likely to be the result of the more humid climate of Japan resulting in a higher diversity of freshwater habitats and the stronger impact of the glacial periods in the Western Palaearctic in combination with the Sahara, preventing tropical African lineages dispersing northwards.

Data-integration of opportunistic species observations into hierarchical modeling frameworks improves spatial predictions for urban red squirrels

The prevailing trend of increasing urbanization and habitat fragmentation makes knowledge of species’ habitat requirements and distribution a crucial factor in conservation and urban planning. Species distribution models (SDMs) offer powerful toolboxes for discriminating the underlying environmental factors driving habitat suitability. Nevertheless, challenges in SDMs emerge if multiple data sets - often sampled with different intention and therefore sampling scheme – can complement each other and increase predictive accuracy. Here, we investigate the potential of using recent data integration techniques to model potential habitat and movement corridors for Eurasian red squirrels (Sciurus vulgaris), in an urban area. We constructed hierarchical models integrating data sets of different quality stemming from unstructured on one side and semi-structured wildlife observation campaigns on the other side in a combined likelihood approach and compared the results to modeling techniques based on only one data source - wherein all models were fit with the same selection of environmental variables. Our study highlights the increasing importance of considering multiple data sets for SDMs to enhance their predictive performance. We finally used Circuitscape (version 4.0.5) on the most robust SDM to delineate suitable movement corridors for red squirrels as a basis for planning road mortality mitigation measures. Our results indicate that even though red squirrels are common, urban habitats are rather small and partially lack connectivity along natural connectivity corridors in Berlin. Thus, additional fragmentation could bring the species closer to its limit to persist in urban environments, where our results can act as a template for conservation and management implications.

Identifying engaging bird species and traits with community science observations

Identifying rates at which birders engage with different species can inform the impact and efficacy of conservation outreach and the scientific use of community-collected biodiversity data. Species that are thought to be “charismatic” are often prioritized in conservation, and previous researchers have used sociological experiments and digital records to estimate charisma indirectly. In this study, we take advantage of community science efforts as another record of human engagement with animals that can reveal observer biases directly, which are in part driven by observer preference. We apply a multistage analysis to ask whether opportunistic birders contributing to iNaturalist engage more with larger, more colorful, and rarer birds relative to a baseline approximated from eBird contributors. We find that body mass, color contrast, and range size all predict overrepresentation in the opportunistic dataset. We also find evidence that, across 472 modeled species, 52 species are significantly overreported and 158 are significantly underreported, indicating a wide variety of species-specific effects. Understanding which birds are highly engaging can aid conservationists in creating impactful outreach materials and engaging new naturalists. The quantified differences between two prominent community science efforts may also be of use for researchers leveraging the data from one or both of them to answer scientific questions of interest.

Large-bodied birds are over-represented in unstructured citizen science data

Citizen science platforms are quickly accumulating hundreds of millions of biodiversity observations around the world annually. Quantifying and correcting for the biases in citizen science datasets remains an important first step before these data are used to address ecological questions and monitor biodiversity. One source of potential bias among datasets is the difference between those citizen science programs that have unstructured protocols and those that have semi-structured or structured protocols for submitting observations. To quantify biases in an unstructured citizen science platform, we contrasted bird observations from the unstructured iNaturalist platform with that from a semi-structured citizen science platform-eBird-for the continental United States. We tested whether four traits of species (body size, commonness, flock size, and color) predicted if a species was under- or over-represented in the unstructured dataset compared with the semi-structured dataset. We found strong evidence that large-bodied birds were over-represented in the unstructured citizen science dataset; moderate evidence that common species were over-represented in the unstructured dataset; strong evidence that species in large groups were over-represented; and no evidence that colorful species were over-represented in unstructured citizen science data. Our results suggest that biases exist in unstructured citizen science data when compared with semi-structured data, likely as a result of the detectability of a species and the inherent recording process. Importantly, in programs like iNaturalist the detectability process is two-fold-first, an individual organism needs to be detected, and second, it needs to be photographed, which is likely easier for many large-bodied species. Our results indicate that caution is warranted when using unstructured citizen science data in ecological modelling, and highlight body size as a fundamental trait that can be used as a covariate for modelling opportunistic species occurrence records, representing the detectability or identifiability in unstructured citizen science datasets. Future research in this space should continue to focus on quantifying and documenting biases in citizen science data, and expand our research by including structured citizen science data to understand how biases differ among unstructured, semi-structured, and structured citizen science platforms.

Observing the Observers: How Participants Contribute Data to iNaturalist and Implications for Biodiversity Science

The availability of citizen science data has resulted in growing applications in biodiversity science. One widely used platform, iNaturalist, provides millions of digitally vouchered observations submitted by a global user base. These observation records include a date and a location but otherwise do not contain any information about the sampling process. As a result, sampling biases must be inferred from the data themselves. In the present article, we examine spatial and temporal biases in iNaturalist observations from the platform's launch in 2008 through the end of 2019. We also characterize user behavior on the platform in terms of individual activity level and taxonomic specialization. We found that, at the level of taxonomic class, the users typically specialized on a particular group, especially plants or insects, and rarely made observations of the same species twice. Biodiversity scientists should consider whether user behavior results in systematic biases in their analyses before using iNaturalist data.

A quantitative assessment of site-level factors in influencing Chukar (Alectoris chukar) introduction outcomes

Chukar partridges (Alectoris chukar) are popular game birds that have been introduced throughout the world. Propagules of varying magnitudes have been used to try and establish populations into novel locations, though the relationship between propagule size and species establishment remains speculative. Previous qualitative studies argue that site-level factors are of importance when determining where to release Chukar. We utilized machine learning ensembles to evaluate bioclimatic and topographic data from native and naturalized regions to produce predictive species distribution models (SDMs) and evaluate the relationship between establishment and site-level factors for the conterminous United States. Predictions were then compared to a distribution map based on recorded occurrences to determine model prediction performance. SDM predictions scored an average of 88% accuracy and suitability favored states where Chukars were successfully introduced and are present. Our study shows that the use of quantitative models in evaluating environmental variables and that site-level factors are strong indicators of habitat suitability and species establishment.

Ecological impact assessment of climate change and habitat loss on wetland vertebrate assemblages of the Great Barrier Reef catchment and the influence of survey bias

Wetlands are among the most vulnerable ecosystems, stressed by habitat loss and degradation from expanding and intensifying agricultural and urban areas. Climate change will exacerbate the impacts of habitat loss by altering temperature and rainfall patterns. Wetlands within Australia's Great Barrier Reef (GBR) catchment are not different, stressed by extensive cropping, urban expansion, and alteration for grazing. Understanding how stressors affect wildlife is essential for the effective management of biodiversity values and minimizing unintended consequences when trading off the multiple values wetlands support. Impact assessment is difficult, often relying on an aggregation of ad hoc observations that are spatially biased toward easily accessible areas, rather than systematic and randomized surveys. Using a large aggregate database of ad hoc observations, this study aimed to examine the influence of urban proximity on machine-learning models predicting taxonomic richness and assemblage turnover, relative to other habitat, landscape, and climate variables, for vertebrates dwelling in the wetlands of the GBR catchment. The distance from the nearest city was, by substantial margins, the most influential factor in predicting the richness and assemblage turnover of all vertebrate groups, except fish. Richness and assemblage turnover was predicted to be greatest nearest the main urban centers. The extent of various wetland habitats was highly influential in predicting the richness of all groups, while climate (predominately the rainfall in the wettest quarter) was highly influential in predicting assemblage turnover for all groups. Bias of survey records toward urban centers strongly influenced our ability to model wetland-affiliated vertebrates and may obscure our understanding of how vertebrates respond to habitat loss and climate change. This reinforces the need for randomized and systematic surveys to supplement existing ad hoc surveys. We urge modelers in other jurisdictions to better portray the potential influence of survey biases when modeling species distributions.

How citizen science could improve species distribution models and their independent assessment

Species distribution models (SDM) have been increasingly developed in recent years, but their validity is questioned. Their assessment can be improved by the use of independent data, but this can be difficult to obtain and prohibitive to collect. Standardized data from citizen science may be used to establish external evaluation datasets and to improve SDM validation and applicability.We used opportunistic presence-only data along with presence-absence data from a standardized citizen science program to establish and assess habitat suitability maps for 9 species of amphibian in western France. We assessed Generalized Additive and Random Forest Models' performance by (1) cross-validation using 30% of the opportunistic dataset used to calibrate the model or (2) external validation using different independent datasets derived from citizen science monitoring. We tested the effects of applying different combinations of filters to the citizen data and of complementing it with additional standardized fieldwork.Cross-validation with an internal evaluation dataset resulted in higher AUC (Area Under the receiver operating Curve) than external evaluation causing overestimation of model accuracy and did not select the same models; models integrating sampling effort performed better with external validation. AUC, specificity, and sensitivity of models calculated with different filtered external datasets differed for some species. However, for most species, complementary fieldwork was not necessary to obtain coherent results, as long as the citizen science data were strongly filtered.Since external validation methods using independent data are considered more robust, filtering data from citizen sciences may make a valuable contribution to the assessment of SDM. Limited complementary fieldwork with volunteer's participation to complete ecological gradients may also possibly enhance citizen involvement and lead to better use of SDM in decision processes for nature conservation.

The role of passive surveillance and citizen science in plant health

The early detection of plant pests and diseases is vital to the success of any eradication or control programme, but the resources for surveillance are often limited. Plant health authorities can however make use of observations from individuals and stakeholder groups who are monitoring for signs of ill health. Volunteered data is most often discussed in relation to citizen science groups, however these groups are only part of a wider network of professional agents, land-users and owners who can all contribute to significantly increase surveillance efforts through "passive surveillance". These ad-hoc reports represent chance observations by individuals who may not necessarily be looking for signs of pests and diseases when they are discovered. Passive surveillance contributes vital observations in support of national and international surveillance programs, detecting potentially unknown issues in the wider landscape, beyond points of entry and the plant trade. This review sets out to describe various forms of passive surveillance, identify analytical methods that can be applied to these "messy" unstructured data, and indicate how new programs can be established and maintained. Case studies discuss two tree health projects from Great Britain (TreeAlert and Observatree) to illustrate the challenges and successes of existing passive surveillance programmes. When analysing passive surveillance reports it is important to understand the observers' probability to detect and report each plant health issue, which will vary depending on how distinctive the symptoms are and the experience of the observer. It is also vital to assess how representative the reports are and whether they occur more frequently in certain locations. Methods are increasingly available to predict species distributions from large datasets, but more work is needed to understand how these apply to rare events such as new introductions. One solution for general surveillance is to develop and maintain a network of tree health volunteers, but this requires a large investment in training, feedback and engagement to maintain motivation. There are already many working examples of passive surveillance programmes and the suite of options to interpret the resulting datasets is growing rapidly.

Data-derived metrics describing the behaviour of field-based citizen scientists provide insights for project design and modelling bias

Around the world volunteers and non-professionals collect data as part of environmental citizen science projects, collecting wildlife observations, measures of water quality and much more. However, where projects allow flexibility in how, where, and when data are collected there will be variation in the behaviour of participants which results in biases in the datasets collected. We develop a method to quantify this behavioural variation, describing the key drivers and providing a tool to account for biases in models that use these data. We used a suite of metrics to describe the temporal and spatial behaviour of participants, as well as variation in the data they collected. These were applied to 5,268 users of the iRecord Butterflies mobile phone app, a multi-species environmental citizen science project. In contrast to previous studies, after removing transient participants (those active on few days and who contribute few records), we do not find evidence of clustering of participants; instead, participants fall along four continuous axes that describe variation in participants' behaviour: recording intensity, spatial extent, recording potential and rarity recording. Our results support a move away from labelling participants as belonging to one behavioural group or another in favour of placing them along axes of participant behaviour that better represent the continuous variation between individuals. Understanding participant behaviour could support better use of the data, by accounting for biases in the data collection process.