iiinova — where your digital self keeps evolving

EcoEvoRxiv @ecoevorxiv.bsky.social · 17h

'Expanded Polystyrene Exposure and Nitrite-Related Absorbance in Eisenia fetida' doi.org/10.32942/X2R...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 19h

'A finite-population bound on the selection limit' doi.org/10.32942/X2V...

020

EcoEvoRxiv @ecoevorxiv.bsky.social · 21h

'Plasticity Decay and Attenuated Recovery (PDAR): A Hybrid Dynamical Model for the Historical Accessibility of Evolutionary Space' doi.org/10.32942/X20...

Schematic hierarchical diversification. The rate of major innovation νmajor(t) tracks
the declining capacity bound, while locally-driven minor diversification νminor(t) can continue
to increase over the same interval.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 23h

Land-cover-specific network reconfiguration and spatial trajectories govern abrupt vegetation regime shifts across global drylands doi.org/10.32942/X24...

Spatial distribution of global dryland vegetation step-shift quadrats and associated temporal dynamics of representative spatial metrics.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 23h

Weak and species-specific morphological change in Atlantic Forest birds across two decades of warming doi.org/10.32942/X28...

Modelling sampling turnover more than halves the estimated wing-length decline. (A) Within-species-centred wing length over time across 8,478 records from 72 species. The dashed line represents the baseline model (−1.8 mm/decade); the solid line and band represent the fully adjusted model (−0.7 mm/decade, 95% CrI spanning zero). Inset displays wing records per year. (B) Effects of each capture year on wing length from the baseline (open circles) and fully adjusted (filled circles) models with year as a factor, centred on the record-weighted mean year, with 95% confidence intervals (glmmTMB REML, pooled over 50 phylogenetic trees); the fitted linear trends of panel A are repeated. Unlike raw annual means, these intervals account for the clustering of records within species, contributors and municipalities. Linear trends and their bands are posterior means and 95% credible intervals pooled over 20 phylogenetic trees.

012

EcoEvoRxiv @ecoevorxiv.bsky.social · 02/10/2026

Climate vulnerability assessment framework choice influences species prioritization and prediction of population decline doi.org/10.32942/X2C...

Relationship between the predicted probability of climate-change threat and the probability of observed population decline among Canadian breeding bird species (n=236), modeled from binary response (declined or not). Points represent binned empirical proportions of species that declined (1970–2017), with point size proportional to the number of species per bin. The solid line shows the posterior mean relationship from a Bayesian phylogenetic logistic regression (binary decline response), and the shaded region represents the 95% credible interval. Predicted climate vulnerability strongly increased the probability of decline (β = 2.18, 95% credible interval: 0.98–3.44, log-odds scale), corresponding to an approximately ninefold increase in the odds of decline across a one-unit increase in predicted vulnerability.

020

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Genomic divergence between species identifies taxonomic outliers and establishes calibrated molecular clocks doi.org/10.32942/X2T...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

What, if anything, can studies of adaptation to altitude reveal about responses to climate change? doi.org/10.32942/X2H...

[A] Lowland (red) and highland (blue) populations remain connected by gene flow (arrow), introducing a migration load into the resident population. [B] Migration-selection balance. For a string of subpopulations along an altitudinal gradient, allele 𝐴 (with frequency 1-𝑞) is selected against in the highlands and selected for in the lowlands, and vice versa for allele 𝑎 (with frequency 𝑞). Gene flow continuously reintroduces alleles from the alternate environment. Panel adapted from Mallet (2001) [C] Local adaptation to altitude. The average fitness (𝑊% ) of lowland individuals (red circle) is greater in the lowland than it is for any other population, and vice versa for highland individuals (blue circle).

020

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Genomic repetition in climate adaptation across different crop species doi.org/10.32942/X2W...

Genomic database of crop landraces and wild relatives. Shown is the geographic distribution of the individual samples included in the database for (A) crop landraces and (B) crop wild relatives. Point colors correspond to the different crop species shown on the right, with the respective sample size.

020

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Nonlinear survey-effort adjustment removes the reported bird-diversity effect doi.org/10.32942/X21...

PSI estimates under alternative survey-effort specifications for four biodiversity measures. Points show PSI coefficient estimates and bars show county-clustered 95% confidence intervals. Richness, Shannon diversity, and Simpson diversity were negatively associated with PSI under the original Duration specification, whereas Evenness was positively associated with PSI. Once total observation time and observer participation were represented explicitly, the PSI estimates became statistically indistinguishable from zero across all four biodiversity measures.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Diurnal temperature fluctuations outweigh population origin in shaping thermal performance and range predictions in European Aedes albopictus doi.org/10.32942/X2N...

Climatic context of the study populations and experimental thermal design. (A) Sampling locations of the three Aedes albopictus populations along the European invasive range, spanning approximately 12 degrees of latitude (Frankfurt, Germany; Trento, Italy; Palermo, Italy). (B) Mean seasonal temperature cycle at each location, calculated from ERA5 reanalysis monthly means (1995–2025); shaded ribbons show ±1 SD across years. (C) Mean seasonal precipitation cycle at each location over the same period, with shaded ribbons showing ±1 SD across years. (D) Interannual coefficients of variation (CV) in annual mean temperature and annual total precipitation for each population, calculated across the 1995–2025 period.

030

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Amplexus Takeover in the Asian Common Toad, Duttaphrynus melanostictus, in Dumka, Jharkhand, India doi.org/10.32942/X25...

Axillary amplexus in Duttaphrynus
melanostictus

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Probabilistic projections of human modification of the terrestrial biosphere from 2020 to 2040 doi.org/10.32942/X2S...

Graphic overview of the human modification (HM) forecasting model architecture, inputs, and outputs. A location encoder converts coordinates into features capturing residual geographic variation. A convolutional long short-term memory (convLSTM) network operates on sequences of three consecutive historical HM maps and dynamic and static covariates to learn joint spatial and temporal patterns. ConvLSTM encodings are passed to a prediction head which outputs a full distribution of potential future HM.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Expert Elicitation on demographic consequences of seabird compensatory measures doi.org/10.32942/X29...

Probabilistic judgements elicited from five experts on the potential demographic gains arising from application of eight alternative compensatory measures for three grouped tern species (common, arctic and sandwich terns). Each probabilistic judgement is depicted by each expert’s value for the median (circle), upper and lower quartiles (thick line), and upper and lower limits (thin dotted line).

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Large differences in thermal tolerance between isolated organs in rainbow trout doi.org/10.32942/X2D...

Relative contraction force of isolated ventricle preparations in organ baths. A) Natural decline in contraction force over time at constant temperature (12°C) and B) temperature dependent decline of contraction force (corrected for the natural decline, based on model predictions of the control samples) in circular ventricle preparations from rainbow trout hearts in an in vitro organ bath set-up.

010

EcoEvoRxiv @ecoevorxiv.bsky.social · 01/10/2026

Toward a General Theory of Heteromorphic Reproduction doi.org/10.32942/X2P...

A hypothetical reproductive portfolio comprising three heteromorphic reproductive outputs. Outputs produced through different mechanisms draw on the same reproductive budget. Their returns are expressed on a common cost and fitness scale, and joint performance is calculated from the joint environmental performance of the complete strategy.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 30/09/2026

Wildlife Responses to a Resident a Family: A Three-Year Camera Census on Restored Prairie–Savanna (2024–2026) doi.org/10.32942/X2X...

Monthly wildlife activity (estimated individuals) by year; 2026 is shown through June.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 30/09/2026

Wolf–Deer and Guild Camera-trap Detections Across Wisconsin’s Statewide Wolf Range (2018–2025), Resolved by Deer Class: Buck Selection in the Hard-Antler Season, Bears and Fawns, and Class-Specific Reference Values for Displacement doi.org/10.32942/X22...

Observed share of bucks among the wolf’s nearest classed adult deer (dots, 95% Wilson intervals) against the share available at the same cameras in the same months (gray dashes), for the all-year record and three hardantler windows. Ratios above each point are observed ÷ available. Data: statewide Snapshot Wisconsin extract, 2018–2025.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 30/09/2026

Reservoir, sink, or reactor? Repositioning marine sponges as candidate indicators for coastal antimicrobial resistance monitoring doi.org/10.32942/X26...

$Marine sponges as candidate indicators of the coastal resistome. (a) The inherited paradigm: the sponge holobiont has been investigated as a source of bioactive compounds, with ARGs appearing as incidental antibiogram data. (b) The sponge as a candidate resistome integrator: filtration delivers allochthonous, ARG-carrying cells from coastal sources into a resident consortium orders of magnitude denser than the surrounding water, retaining the sub-micrometer fraction that bivalves let pass (inset). Three outcomes remain compatible with the available evidence, each assigning a different meaning to the same measured value (indicator semantics, tinted boxes): reservoir, sink, and reactor. Predictions are tested against paired inhalant water and against distance from the anthropogenic source (gradient bar). (c) Validation pathway: paired sampling of sponge tissue, inhalant and exhalant water, and sediment along a source-to-offshore transect, repeated across seasons, with the analytical methods and the controls required to render the signal interpretable. The figure was generated in BioRender.$

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 30/09/2026

Population density, cooperation, and the evolution of toothed whale societies doi.org/10.32942/X2B...

Population density sets the stage for two alternate socio-spatial strategies. (A) Populations with low density are unlikely to have any social support if individuals spread out to forage. (B) Instead, these populations may form cohesive groups that move collectively. These groups are typically based on female kinship but see exceptions in later discussion on optimal group size. (C) Populations with high density may find themselves closer to non-kin when spreading out to forage. (D) Here, developing affiliative relationships with non-kin can be beneficial, hedging against the possibility of needing help and having no kin nearby. To extend the model to bonds between social groups; e.g., those between female sperm whale units or male Indo-Pacific bottlenose dolphin alliances, silhouettes may represent stable social groups.

010

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Modeling Decomposition: Beyond a Well-Mixed World doi.org/10.32942/X2F...

Implications of organism and environment type on decomposer traits. Depiction of influences of organism (unitary or networked) and environment (well-mixed or patchy) type on decomposer traits that could be integrated into C models. Scenario “A” is status quo with unitary decomposers and a well-mixed environment. As such, light purple dashes represent a status quo response of decomposer traits, whereas green up and red down arrows represent increases and decreases in trait values, respectively. Scenario “B” shows impact of combining unitary organisms with patchy environments, showing the hypothesis that decomposers may be less successful and have reduced carbon use efficiencies (CUE) and decomposition rates with high biomass and necromass turnover rates as a result of stoichiometric mismatch and inefficient C discovery. Scenario “C” shows the impact of combining networked organisms and patchy environments, with orange shading around organisms representing their ability to access resources further from them. In Scenario C, decomposition rates are expected to be high due to stoichiometric balance, while biomass and necromass turnover rates are expected to be low due to structural biomass for network maintenance, and with uncertain impacts on CUE, due to competing influences of greater biomass building and stoichiometric balance.

001

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Common, consequential, and actively maintained: a neo-balance view of genetic variation doi.org/10.32942/X2K...

Figure showing fluctuating selection as revealed by longitudinal sampling across taxa.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Evenness under incomplete sampling: coupled error channels, hidden compensation, and the richness-estimation bottleneck doi.org/10.32942/X2Q...

Within-inventory dependence of the two sampling channels in 3,167 Ecological Register inventories. Median replicate Pearson correlation between Uβ and Vβ under without-replacement rarefaction. The figure shows β = 1. Spearman dependence is exactly β-invariant by Proposition 2; Pearson dependence is not invariant, but its observed differences across β = 0, 1, 2 were small.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Quantifying the probability of evolutionary rescue doi.org/10.32942/X2T...

Conceptual challenges in determining the probability of evolutionary rescue. Panels a and b: Population size trajectories of replicate populations exposed to environmental change (red lines – with evolution: grey lines – without evolution). If the environment does not improve (i.e., the mean fitness of the non-evolving ancestral population, w¯0, remains below 1), all populations go extinct without evolution (panel a). By contrast, if the environment improves after some time, lifting w¯0 above 1, some populations may survive even without evolution (panel b). However, adaptive evolution increases the chance of survival, and more populations persist. Panel c: Fraction of populations extinct at time t with and without evolution over time. Following the first environmental change (blue triangle), all evolving and non-evolving populations persist initially but eventually all non-evolving populations inevitably go extinct, while some evolving populations survive the change. However, the environment does not stay constant indefinitely and at the next environmental change (yellow triangle), more of the evolving populations go extinct. The figure is purely illustrative and not based on an actual model.

063

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Canyoning simplifies benthic macroinvertebrate communities in Mediterranean canyons: evidence from Sierra y Cañones de Guara Natural Park doi.org/10.32942/X2Z...

Map with terrain of the northern part of the Aragón province in Spain indicating the location of the six canyons of the Sierra y Cañones de Guara Natural Park (SCGNP) included in the study.

010

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

A synthesis of bird biodiversity studies in geographically isolated wetlands reveals research synergies and opportunities doi.org/10.32942/X23...

US map showing the number of publications per theme (here, climate on the left and hydrology on the right) from each state in the continental United States. Only publications containing themes that had >5 publications are displayed (n=8). Publications are symbolised by a dot of a size that varies as a function of the number of publications. Hydrology has much fewer points than climate.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Reconsidering interspecific co-occurrence of native and invasive plants in an anthropogenic riparian ecosystem doi.org/10.32942/X27...

Graphical representation of the herbaceous species richness recorded under the three different vegetation types for each season. Values (mean ± SD) were compared between non-native (Fallopia), native (Rubus) and uninvaded area (control). Different minuscule letters indicate significant differences among season, and different capital letters indicate significant differences among vegetation type.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

TaxonBodyMassML: Taxonomy-informed prediction of body mass using gradient-boosted trees doi.org/10.32942/X2B...

Graph showing the predicted vs. actual body mass (g) for TaxonBodyMassML on the held-out test set. Panel A shows results with the full test set, while panel B shows the results for a test set restricted to mass > 0.1 g.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Patterns of wind dispersal within a forest canopy: An epiphyte perspective doi.org/10.32942/X2G...

Conceptual diagram of branch movements and their apparent winds that diverge from the actual winds due to (A) mechanical disturbance and (B) branch sway due to wind loading.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

On the Promise (and Practice) of Deep Comparative Methods: Expressivity, Limitations, and Design doi.org/10.32942/X2M...

An intuition for universal function approximation. A neural network with one hidden layer can form a weighted sum of sigmoid functions applied to weighted and shifted inputs. With that in mind, this figure illustrates the intuition behind the universal approximation results of Cybenko (1989) and Hornik et al. (1989) in one input dimension. More generally, neural networks can approximate mappings between high-dimensional numerical inputs and outputs. Because images, videos, CT scans, text, and conventional trait measurements can all be encoded numerically, this capacity allows neural networks to map complex patterns in any of these data sources onto other numerical representations or predictions.

002

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Building from biodiversity assessment to place-based conservation action across scales of information and governance doi.org/10.32942/X2R...

Diagram of the proposed “Assessment to Action” cycle, highlighting elements of biodiversity assessment initiatives that can enhance their support and connections to place based conservation. The diagram is intended to highlight the themes of this essay rather than provide a comprehensive overview of the assessment process, which necessarily includes additional elements.

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Timelines of vegetation recovery following fire in a shrub-dominated Mediterranean-type ecosystem' doi.org/10.32942/X2V...

Kaplan-Meier plot illustrating timelines of recovery following fire, including censure of points that did not recover before the end of the analysis. Recovery is indexed using the first year in which 4 consecutive years had relative cover > 80% pre-fire levels for Shrubs and Trees, or relative cover < 120% pre-fire levels for Herbaceous/Graminoids (see Methods section 2.5). Early in chronosequences, no points have achieved recovery (unrecovered probability is 100%), and over time vegetation progressively recovers following fire (unrecovered probability goes down). Herbaceous/Graminoid cover returns to baseline first, with a median recovery time of 6 years, followed by Shrubs (9 years), and finally Trees (16 years).

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 29/09/2026

Aligning evidence and action: The Research-to-Decision (RED) Framework to address modern environmental and ecological challenges. doi.org/10.32942/X20...

The Research-to-Decision (RED) Framework addresses the persistent science-policy challenges of aligning timelines, managing uncertainty, and ensuring accountability by integrating transdisciplinary research, science-policy interfaces, AI-enabled ecology, and responsible AI. The framework is operationalised through embedding AI heterogeneously across the entire research cycle; coloured circles indicate stages of the research cycle where domain expertise leads (green), AI accelerates (orange), and AI-human synergy is critical (blue). The iterative cycle incorporates reflection (solid black arrows) and feedback (hollow black arrows and dashed line) to support evidence-based decision-making. The cycle ends when a challenge is solved or continues if the challenge needs to be iterated or redefined.

030

EcoEvoRxiv @ecoevorxiv.bsky.social · 27/09/2026

Disentangling the drivers of parasite occurrence in a native and a non-native small mammal across a land-use gradient. DOI: doi.org/10.32942/X24...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 27/09/2026

Why nature cannot be replaced: the limits of substitution in biodiversity conservation. DOI: doi.org/10.32942/X28...

010

EcoEvoRxiv @ecoevorxiv.bsky.social · 27/09/2026

Regional range restriction, not national endemism, predicts plant extinction risk: operationalising Conservation Imperatives for under-assessed tropical floras. DOI: doi.org/10.32942/X2C...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 27/09/2026

We Have Never Been Hosts: Microbial Niche Weaving and the Evolution of Multicellular Life. DOI: doi.org/10.32942/X2H...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 27/09/2026

Why abundance-based indices of niche breadth are biased, and what can be done to improve them. DOI: doi.org/10.32942/X2N...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 27/09/2026

Evolutionary potential of natal dispersal in multiple blue tit populations while accounting for spatial autocorrelation among relatives. DOI: doi.org/10.32942/X2S...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 26/09/2026

Global Evidence for Biotic Differentiation Amid Contrasting Freshwater β-Diversity Responses to Human Impacts. DOI: doi.org/10.32942/X2W...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 26/09/2026

Different Beta-Diversity Frameworks Reveal Contrasting Roles of Species Abundance Distributions and Spatial Aggregation. DOI: doi.org/10.32942/X21...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 26/09/2026

Individual recognition and number prediction for Manx shearwater (Puffinus puffinus) by passive acoustic monitoring. DOI: doi.org/10.32942/X25...

010

EcoEvoRxiv @ecoevorxiv.bsky.social · 26/09/2026

The marine annelid trait manifesto: A call to rethink functional traits for ecological and conservation research. DOI: doi.org/10.32942/X29...

011

EcoEvoRxiv @ecoevorxiv.bsky.social · 26/09/2026

INSTAR: reporting items for invertebrate welfare in research. DOI: doi.org/10.32942/X2D...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 26/09/2026

Spatial structure before algorithmic complexity: evidence-based guidance for marine species distribution modelling where data are limited. DOI: doi.org/10.32942/X2J...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 26/09/2026

Caenorhabditis elegans and Serratia marcescens host-parasite coevolution model strains derived through mutagenesis and experimental evolution. DOI: doi.org/10.32942/X2P...

000

EcoEvoRxiv @ecoevorxiv.bsky.social · 25/09/2026

A Geospatial informed Conservation Strategy for Sustainable Management of Dalbergia sissoo (Roxb.) Population in India amidst Climate Change and Pathogen Induced Mortality Stress. DOI: doi.org/10.32942/X2T...

120

EcoEvoRxiv @ecoevorxiv.bsky.social · 25/09/2026

Coordinated interannual change in NEON soil bacterial communities across continental North America. DOI: doi.org/10.32942/X2X...

110

EcoEvoRxiv @ecoevorxiv.bsky.social · 25/09/2026

Global assemblage stability is driven by asynchrony shaped by environmental heterogeneity. DOI: doi.org/10.32942/X22...

000