Reviewing the Spectral Variation Hypothesis: Twenty years in the tumultuous sea of biodiversity estimation by remote sensing

Ecological Informatics, Journal Year: 2024, Volume and Issue: 82, P. 102702 - 102702

Published: July 3, 2024

Twenty years ago, the Spectral Variation Hypothesis (SVH) was formulated as a means to link between different aspects of biodiversity and spatial patterns spectral data (e.g. reflectance) measured from optical remote sensing. This hypothesis initially assumed positive correlation variations computed raster in environment, which would turn correlate with species richness: following SVH, areas characterized by high heterogeneity (SH) should be related higher number available ecological niches, more likely host when combined. The past decade has witnessed major evolution progress both terms remotely sensed available, techniques analyze them, questions addressed. SVH been tested many contexts variety sensing data, this recent corpus highlighted potentials pitfalls. aim paper is review discuss methodological developments based on leading knowledge well conceptual uncertainties limitations for application estimate dimensions biodiversity. In particular, we systematically than 130 publications provide an overview ecosystems, characteristics (i.e., spatial, temporal resolution), metrics, tools, applications strength association SH metrics reported each study. conclusion, serves guideline researchers navigating complexities applying offering insights into current state future research possibilities field estimation data.

Language: Английский

---

Spatial resolution impacts projected plant responses to climate change on topographically complex islands

Diversity and Distributions, Journal Year: 2023, Volume and Issue: 29(10), P. 1245 - 1262

Published: July 27, 2023

Abstract Aim Understanding how grain size affects our ability to characterize species responses ongoing climate change is of crucial importance in the context an increasing awareness for substantial difference that exists between coarse spatial resolution macroclimatic data sets and microclimate actually experienced by organisms. Climate impacts on biodiversity are expected peak mountain areas, wherein differences macro microclimates precisely largest. Based a newly generated fine‐scale environmental Canary Islands, we assessed whether at 100 m able provide more accurate predictions than available 1 km resolution. We also analysed future suitability island endemic bryophytes differ depending grids. Location Islands. Time period Present (1979–2013) late‐century (2071–2100). Taxa Bryophytes. Methods compared accuracy using ensemble small models 14 Macaronesian bryophyte species. used two sets: CHELSA v1.2 (~1 km) CanaryClim v1.0 (100 m), downscaled version latter utilizing from local weather stations. encompasses five individual model intercomparison projects three warming shared socio‐economic pathways. Results Species distribution exhibited similar accuracy, but predicted buffered trends mid‐elevation ridges. consistently returned higher proportions suitable pixels (8%–28%) (0%–3%). Consequently, proportion occupy uncertain was with (3–8 species) (0–2 species). Main conclusions The impacted rather performance models. Our results highlight role fine‐resolution can play predicting potential both microrefugia new range under climate.

Language: Английский

---

Optimising occurrence data in species distribution models: sample size, positional uncertainty, and sampling bias matter

Ecography, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 2, 2024

Species distribution models (SDMs) have proven valuable in filling gaps our knowledge of species occurrences. However, despite their broad applicability, SDMs exhibit critical shortcomings due to limitations occurrence data. These include, particular, issues related sample size, positional uncertainty, and sampling bias. In addition, it is widely recognised that the quality as well approaches used mitigate impact aforementioned data depend on ecology. While numerous studies evaluated effects these SDM performance, a synthesis results lacking. without comprehensive understanding individual combined effects, ability predict influence modelled species–environment associations remains largely uncertain, limiting value model outputs. this paper, we review bias, ecology We build upon findings provide recommendations for assessment intended use SDMs.

Language: Английский

---

Assessing the utility of SoilGrids250 for biogeographic inference of plant populations

Ecology and Evolution, Journal Year: 2024, Volume and Issue: 14(3)

Published: March 1, 2024

Abstract Inclusion of edaphic conditions in biogeographical studies typically provides a better fit and deeper understanding plant distributions. Increased reliance on soil data calls for easily accessible layers providing continuous predictions worldwide. Although SoilGrids potentially useful source predicted biogeographic applications, its accuracy estimating the characteristics experienced by individuals small‐scale populations is unclear. We used sampling approach to obtain samples from 212 sites across midwestern eastern United States, only at where there was population one 22 species Lobelia sect. . analyzed six physical chemical our compared them with values SoilGrids. Across all species, texture variables (clay, silt, sand) were ( R 2 : .25–.46) than chemistry (carbon nitrogen, ≤ .01; pH, .19). While rarely matched actual field any variable, we able recover qualitative patterns relating means population‐level pH. Rank order mean direct measures much more consistent (Spearman r S = .74–.84; p < .0001) pH .61, .002) carbon nitrogen > .35). Within L. siphilitica , significant association, known measurements, between sex ratios could be detected using data, but large numbers sites. Our results suggest that modeled can caution such as distribution modeling, contents are currently unreliable, least region studied here.

Language: Английский

---

Grassland vertical height heterogeneity predicts flower and bee diversity: an UAV photogrammetric approach

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Jan. 8, 2024

Abstract The ecosystem services offered by pollinators are vital for supporting agriculture and functioning, with bees standing out as especially valuable contributors among these insects. Threats such habitat fragmentation, intensive agriculture, climate change contributing to the decline of natural bee populations. Remote sensing could be a useful tool identify sites high diversity before investing into more expensive field survey. In this study, ability Unoccupied Aerial Vehicles (UAV) images estimate biodiversity at local scale has been assessed while testing concept Height Variation Hypothesis (HVH). This hypothesis states that higher vegetation height heterogeneity (HH) measured remote information, vertical complexity associated species diversity. further developed understand if HH can also considered proxy abundance. We tested approach in 30 grasslands South Netherlands, where an data campaign (collection flower abundance) was carried 2021, along UAV true color-RGB-images spatial resolution). Canopy Models (CHM) were derived using photogrammetry technique “Structure from Motion” (SfM) horizontal resolution (spatial) 10 cm, 25 50 cm. accuracy CHM comparing them through linear regression against LiDAR (Light Detection Ranging) Airborne Laser Scanner completed 2020/2021, yielding $$R^2$$ R 2 0.71. Subsequently, on CHMs three resolutions, four different indices (Rao’s Q, Coefficient Variation, Berger–Parker index, Simpson’s D index), correlated ground-based abundance data. Rao’s Q index most effective reaching correlations (0.44 diversity, 0.47 0.34 abundance). Interestingly, not significantly influenced photogrammetry. Our results suggest used large-scale, standardized, cost-effective inference quality bees.

Language: Английский

---

Assessing the applicability of binary land-cover variables to species distribution models across multiple grains

Landscape Ecology, Journal Year: 2024, Volume and Issue: 39(3)

Published: March 4, 2024

Abstract Context Species distribution models are widely used in ecology. The selection of environmental variables is a critical step SDMs, nowadays compounded by the increasing availability data. Objectives To evaluate interaction between grain size and binary (presence or absence water) proportional (proportion water within cell) representation cover variable when modeling bird species distribution. Methods eBird occurrence data with an average number records 880,270 per across North American continent were for analysis. Models (via Random Forest) fitted 57 species, two seasons (breeding vs. non-breeding), at four grains (1 km 2 to 2500 ) using as variable. Results models’ performances not affected type adopted (proportional binary) but significant decrease was observed importance form. This especially pronounced coarser during breeding season. Binary useful finer sizes (i.e., 1 ). Conclusions At more detailed ), simple presence certain land-cover can be realistic descriptor occurrence. particularly advantageous collecting habitat field simply recording significantly less time-consuming than its total area. For grains, we recommend variables.

Language: Английский

---

CHclim25 – a spatially and temporally very high-resolution climatic dataset for Switzerland

Published: April 8, 2024

Abstract. CHclim25 is a climatic dataset with 25 m resolution for Switzerland that includes daily, monthly and yearly layers temperature, precipitation, relative sunshine duration, growing degree-days, potential evapotranspiration, bioclimatic variables aridity. The downscaled from daily 1 km the Swiss federal agency meteorology using local regressions an elevation model to better account topography complex phenomena. Climatic are provided individual years, 1981–2010 baseline period future periods 2020–2049, 2045–2074, 2070–209. Future incorporate three regional/global circulation models representative concentration pathways. We compare our predictions values observed at independent weather stations show errors minimal in comparison original resolution, more accurate than available global datasets 30’ especially high elevation. improves temporal spatial accuracy of data enables new studies very ecology environmental sciences.

Language: Английский

---

Unraveling the characteristic spatial scale of habitat selection for forest grouse species in the boreal landscape

Forest Ecology and Management, Journal Year: 2024, Volume and Issue: 563, P. 122008 - 122008

Published: May 25, 2024

The characteristic spatial scale at which species respond strongest to forest structure is unclear and species-specific depends on the degree of landscape heterogeneity. Research often analyzes a pre-defined when constructing distribution models relating variables with occupancy patterns. This limitation, as characteristics shape use habitat multiple scales. To explore drivers this relationship, we conducted an in-depth investigation into how scaling biologically relevant scales affects grouse in boreal forest. We used 4,790 observations (broods adults) collected over 39,303 stands for 15 years four (capercaillie, black grouse, hazel willow grouse) obtained from comprehensive Finnish wildlife triangle census data Airborne Laser Scanning satellite originally sampled 16 m resolution. fitted Generalized Additive Mixed Models linking presence/absence stand composition. estimated effects predictor aggregated three reflecting landscape: local level scale, home range 1 km radius, regional 5 radius. Multi-grain considering forest-species relationships were evaluate whether there specific best predict occupancy. found that affected predictive capacity selection was same (i.e., scale) among species. Different exhibited varying optimal prediction. Forest more important than compositional diversity predicting irrespective scale. A limited number predictors related availability multi-layered vegetation suitable thickets explained patterns all different In conclusion, modeling using can inform managers about perceive landscape. evidence calls integrated multiscale approach modelling

Language: Английский

---

Species distribution models affected by positional uncertainty in species occurrences can still be ecologically interpretable

Ecography, Journal Year: 2023, Volume and Issue: 2023(6)

Published: April 27, 2023

Species distribution models (SDMs) have become a common tool in studies of species–environment relationships but can be negatively affected by positional uncertainty underlying species occurrence data. Previous work has documented the effect on model predictive performance, its consequences for inference about remain largely unknown. Here we use over 12 000 combinations virtual and real environmental variables species, as well case study, to investigate how accurately SDMs recover after applying known errors We explored range predictors with various spatial heterogeneity, species' niche widths, sample sizes magnitudes error. Positional decreased performance all modeled scenarios. The absolute relative importance shape species–environmental co‐varied level uncertainty. These differences were much weaker than those observed overall especially homogenous predictor variables. This suggests that, at least example conditions analyzed, negative did not extend strongly ecological interpretability models. Although findings are encouraging practitioners using reveal generative mechanisms based spatially uncertain data, they suggest greater applications utilizing distributions predicted from positionally such conservation prioritization biodiversity monitoring.

Language: Английский

---

Dealing with area‐to‐point spatial misalignment in species distribution models

Ecography, Journal Year: 2024, Volume and Issue: 2024(5)

Published: March 22, 2024

Species distribution models (SDMs) are extensively used to estimate species–environment relationships (SERs) and predict species across space time. For this purpose, it is key choose relevant spatial grains for predictor response variables at the onset of modelling process. However, environmental often derived from large‐scale climate a grain that can be coarser than one variable. Such area‐to‐point misalignment bias estimates SER jeopardise robustness predictions. We virtual approach, running simulations different levels seek statistical solutions problem. specifically compared accuracy predictive performances, assessed degrees heterogeneity in conditions, three SDMs: GLM, GLM Berkson error model (BEM) accounts fine‐grain within coarse‐grain cells. Only BEM accurately relatively data (up 50 times grain), while two GLMs provide flattened SER. all perform poorly when predicting data, particularly environments more heterogeneous training conditions. Conversely, decreasing relative dataset reduces biases. Because predictions made covariate‐grain displays lower performance GLMs. Thus, standard selection methods would fail select best SERs (here, BEM), which could lead false interpretations about drivers distributions. Overall, we conclude BEM, because robustly grain, holds great promise overcome misalignment.

Language: Английский

---