# Environmental Association Analysis Predicts Dispersal and Adaptive Potential in Agricultural Pests: A Landscape Genomics Approach

## Metadata

- **SPORE ID**: SPR-2026-BFF6
- **Domaines**: Genetic diversity and population structure x Genetic and Environmental Crop Studies
- **Date de generation**: 2026-08-26
- **Panel consensus score**: 6.7/10
- **Novelty score**: 0.85/1.0
- **Panel verdict**: publish_brief

## Abstract

If landscape genomics methods (RDA, GEA) are applied to agricultural pest populations, then environmental variables will explain a significant proportion of genetic differentiation beyond geographic distance, and loci associated with environmental gradients will be enriched for adaptive functions, because these methods explicitly model environmental drivers of genomic variation, which are currently ignored in pest population genetics.

The proposed mechanism involves 4 causal steps: (1) Environmental variables (temperature, humidity, host plant availability) exert s -> (2) These selective pressures drive allele frequency changes at loci underlying adap -> (3) Landscape genomics methods (RDA, GEA) detect genotype-environment associations,  -> (4) The spatial distribution of adaptive alleles and environmental gradients reveals.

Literature grounding on 8 verified references yields a novelty score of 0.85 (novel). A 3-phase experimental protocol (budget: €20k-80k, timeline: 8-14 months) is proposed, starting with in silico validation. A panel of 5 expert reviewers reached a consensus score of 6.7/10.

## 1. Hypothese et mecanisme propose

### 1.1 Formulation formelle

If landscape genomics methods (RDA, GEA) are applied to agricultural pest populations, then environmental variables will explain a significant proportion of genetic differentiation beyond geographic distance, and loci associated with environmental gradients will be enriched for adaptive functions, because these methods explicitly model environmental drivers of genomic variation, which are currently ignored in pest population genetics.

### 1.2 Variables

**Variables independantes :**

| Variable | Type | Plage | Unite |
|----------|------|-------|-------|
| Environmental variables (temperature, humidity, host plant availability) | continuous | Temperature: 10-35 °C; Humidity: 30-90%; Host plant availability: 0-100% | °C, %, % |
| Geographic distance | continuous | 0-2000 km | km |

**Variables dependantes :**

| Variable | Type | Direction attendue | Unite |
|----------|------|-------------------|-------|
| Genetic differentiation (FST) | continuous | increase | dimensionless |
| Number of candidate adaptive loci | continuous | increase | count |

### 1.3 Chaine causale

1. Step 1: Environmental variables (temperature, humidity, host plant availability) exert selective pressures on pest populations.
1. Step 2: These selective pressures drive allele frequency changes at loci underlying adaptive traits (e.g., thermal tolerance, host plant adaptation).
1. Step 3: Landscape genomics methods (RDA, GEA) detect genotype-environment associations, identifying these adaptive loci.
1. Step 4: The spatial distribution of adaptive alleles and environmental gradients reveals dispersal corridors and barriers, informing pest management.

**Hypotheses cles :**

- Environmental variables are accurately measured at sampling locations.
- Pest populations are in or near migration-drift equilibrium.
- Genetic markers (e.g., SNPs) are neutral or under selection, and the effects of demography are accounted for.
- The study species has sufficient genetic diversity to detect associations.

**Inconnues identifiees :**

- The relative importance of different environmental variables in driving adaptation.
- The extent to which gene flow homogenizes adaptive divergence.
- The temporal stability of genotype-environment associations.

### 1.4 Conditions aux limites

- **The pest species must have a wide geographic range with environmental gradients.** — Without environmental variation, genotype-environment associations cannot be detected.
- **Sampling must cover at least 20 populations across the environmental gradient.** — Sufficient statistical power to detect associations.
- **Environmental data must be obtained from high-resolution climate models or on-site measurements.** — Inaccurate environmental data reduces power to detect true associations.

### 1.5 Cadre theorique

Landscape genomics and ecological speciation

## 2. Etat de l'art et positionnement

### 2.1 Travaux les plus proches

- **[2021] Redundancy analysis: A Swiss Army Knife for landscape genomics** — [10.1111/2041-210X.13722](https://doi.org/10.1111/2041-210X.13722)
  - Similarite: related
  - Difference cle: This paper reviews RDA methods for landscape genomics in general, but does not apply them to pest populations.
- **[2022] Genome–Environment Associations, an Innovative Tool for Studying Heritable Evolutionary Adaptation in Orphan Crops and Wild Relatives** — [10.3389/fgene.2022.910386](https://doi.org/10.3389/fgene.2022.910386)
  - Similarite: related
  - Difference cle: Focuses on crop wild relatives, not pests, and does not address dispersal patterns.
- **[2012] Spatial Genetic Variation among Spodoptera frugiperda (Lepidoptera: Noctuidae) Sampled from the United States, Puerto Rico, Panama, and Argentina** — [10.1603/AN11111](https://doi.org/10.1603/AN11111)
  - Similarite: related
  - Difference cle: Studies genetic variation in a pest but does not incorporate environmental variables or landscape genomics methods.

### 2.2 Base de preuves

- **[2021] Redundancy analysis: A Swiss Army Knife for landscape genomics** — [10.1111/2041-210X.13722](https://doi.org/10.1111/2041-210X.13722)
  - Type: direct | Citations: 285
  - Provides the methodological framework (RDA) that the hypothesis proposes to transfer to pest populations.
- **[2022] Genome–Environment Associations, an Innovative Tool for Studying Heritable Evolutionary Adaptation in Orphan Crops and Wild Relatives** — [10.3389/fgene.2022.910386](https://doi.org/10.3389/fgene.2022.910386)
  - Type: indirect | Citations: 45
  - Demonstrates the utility of GEA methods in crop wild relatives, supporting the transferability to other taxa.
- **[2016] Environmental versus geographical effects on genomic variation in wild soybean (Glycine soja) across its native range in northeast Asia** — [10.1002/ece3.2351](https://doi.org/10.1002/ece3.2351)
  - Type: indirect | Citations: 34
  - Shows that environmental factors can predict genomic variation better than geography alone in a wild relative, supporting the hypothesis's core premise.
- **[2021] Physical geography, isolation by distance and environmental variables shape genomic variation of wild barley (Hordeum vulgare L. ssp. spontaneum) in the Southern Levant** — [10.1038/s41437-021-00494-x](https://doi.org/10.1038/s41437-021-00494-x)
  - Type: indirect | Citations: 25
  - Provides evidence that environmental variables contribute to genomic variation in a crop wild relative, supporting the methodological transfer.
- **[2019] Landscape genetics reveals that adaptive genetic divergence in Pinus bungeana (Pinaceae) is driven by environmental variables relating to ecological habitats** — [10.1186/s12862-019-1489-x](https://doi.org/10.1186/s12862-019-1489-x)
  - Type: indirect | Citations: 27
  - Demonstrates the use of landscape genomics to identify environmental drivers of adaptive divergence, supporting the approach.
- **[2018] Adaptive genetic differentiation in Pterocarya stenoptera (Juglandaceae) driven by multiple environmental variables were revealed by landscape genomics** — [10.1186/s12870-018-1524-x](https://doi.org/10.1186/s12870-018-1524-x)
  - Type: indirect | Citations: 22
  - Shows that multiple environmental variables drive adaptive genetic differentiation, supporting the use of environmental association analysis.
- **[2018] Genome-wide signatures of local adaptation among seven stoneflies species along a nationwide latitudinal gradient in Japan** — [10.1186/s12864-019-5453-3](https://doi.org/10.1186/s12864-019-5453-3)
  - Type: analogous | Citations: 19
  - Applies genome-environment associations to insects, showing the potential for pests.
- **[2012] Spatial Genetic Variation among Spodoptera frugiperda (Lepidoptera: Noctuidae) Sampled from the United States, Puerto Rico, Panama, and Argentina** — [10.1603/AN11111](https://doi.org/10.1603/AN11111)
  - Type: analogous | Citations: 23
  - Provides baseline genetic data for a pest species, but does not include environmental analysis.

### 2.3 Contre-preuves et limitations connues

- **[addressable] Spatial Genetic Variation among Spodoptera frugiperda (Lepidoptera: Noctuidae) Sampled from the United States, Puerto Rico, Panama, and Argentina**
  - High gene flow and low genetic structure in a pest species may limit the power to detect environmental associations.

### 2.4 Evaluation de nouveaute

- **Score**: 0.85/1.0
- **Verdict**: novel

## 3. Predictions falsifiables

| # | Prediction | Borne quantitative | Methode | H0 | Test statistique |
|---|-----------|-------------------|---------|-----|-----------------|
| 1 | Environmental variables will explain at least 15% of the genetic variation (FST) beyond geographic distance in Spodoptera litura populations. | RDA variance partitioning: environmental fraction ≥ 15% of total genetic variation, after controlling for geographic distance. | Genotyping-by-sequencing (GBS) to obtain SNP data; redundancy analysis (RDA) with environmental predictors and geographic distance as covariate. | H0: Environmental variables explain ≤ 5% of genetic variation beyond geographic distance. | Permutation test (n=999) in RDA, alpha=0.05 |
| 2 | At least 10 loci will show significant genotype-environment associations (GEA) with temperature or humidity, and these loci will be enriched for gene ontology terms related to stress response. | ≥10 significant SNPs (FDR < 0.05) in GEA analysis; enrichment fold > 2 for stress-related GO terms. | GEA using LFMM or Bayenv2; GO enrichment analysis using topGO. | H0: No SNPs show significant association with environmental variables after FDR correction, or no enrichment for stress-related GO terms. | FDR-adjusted p-values (alpha=0.05) for GEA; Fisher's exact test for GO enrichment, alpha=0.05 |

## 4. Protocole experimental

**Timeline globale**: 8-14 months
**Budget global**: €20k-80k

### 4.1 Phase 1 — In Silico Validation

**Objectif**: To determine if existing genomic and environmental data can support the detection of genotype-environment associations in agricultural pests, specifically Spodoptera litura, and to refine sampling and analytical strategies.

**Methodologie**: 1. Compile existing genomic datasets (e.g., from NCBI SRA) for Spodoptera litura or closely related species (e.g., S. frugiperda) with geographic coordinates. 2. Obtain high-resolution environmental data (WorldClim, CHELSA) for those locations. 3. Perform redundancy analysis (RDA) using the R package 'vegan' to partition genetic variation (FST or allele frequencies) explained by environmental variables vs. geographic distance. 4. Conduct genotype-environment association (GEA) analysis using LFMM (R package 'LEA') or Bayenv2 on the available SNP data. 5. Assess the proportion of variance explained by environment and the number of significant loci. 6. If data are insufficient, simulate realistic population genetic data under selection using 'msprime' or 'SLiM' to test the power of RDA and GEA under various scenarios.

- Cout: €500-2000
- Duree: 4-6 weeks
- Equipement: Standard computer with R and Python
- Logiciels: R (vegan, LEA, qvalue), Python (msprime, SLiM), PLINK for data conversion

**Criteres de succes :**

- Environmental fraction of genetic variation (RDA): ≥ 10% of total genetic variation (or ≥ 15% if data from pilot study)
- Number of significant GEA loci: ≥ 5 loci with FDR < 0.05 (or simulation shows ≥ 80% power to detect ≥ 10 loci)

- **GO**: Both success criteria are met: environmental fraction ≥10% and ≥5 significant loci (or simulation power ≥80%).
- **NO-GO**: Environmental fraction <5% and no significant loci after FDR correction, or simulations show low power (<50%) to detect expected effects.
- **PIVOT**: If data are insufficient, consider using a different pest species with more available genomic data, or expand sampling strategy in Phase 2.

### 4.2 Phase 2 — Minimal Experimental Validation

**Objectif**: To generate a small empirical dataset from Spodoptera litura populations across an environmental gradient to confirm that environmental variables explain a significant proportion of genetic variation and that candidate adaptive loci are detectable.

**Methodologie**: 1. Select 20-30 sampling sites across a temperature and humidity gradient (e.g., from tropical to temperate regions) within the species' range. 2. Collect 20-30 individuals per site (larvae or adults) and preserve in ethanol. 3. Extract DNA using a standard kit (e.g., Qiagen DNeasy). 4. Perform Genotyping-by-Sequencing (GBS) using restriction enzymes (e.g., PstI) and sequence on Illumina platform (e.g., NovaSeq) to obtain SNP markers. 5. Process raw reads using Stacks or dDocent to call SNPs. 6. Obtain environmental data for each sampling site from high-resolution climate models (e.g., CHELSA at 1km resolution). 7. Perform RDA with environmental variables and geographic distance as covariate. 8. Perform GEA using LFMM and Bayenv2. 9. Annotate significant loci using reference genome if available, or compare to related species.

- Cout: €2k-15k
- Duree: 1-3 months
- Equipement: Field collection kits (vials, ethanol, GPS), DNA extraction kit, PCR and sequencing services (outsourced to genomics facility)
- Logiciels: Stacks or dDocent for SNP calling, R (vegan, LEA), Bayenv2, BWA and SAMtools for alignment if reference genome available

**Criteres de succes :**

- Environmental fraction of genetic variation (RDA): ≥ 15% of total genetic variation explained by environmental variables after controlling for geography.
- Number of significant GEA loci: ≥ 10 SNPs with FDR < 0.05 in LFMM or Bayenv2.
- GO enrichment for stress-related terms: Enrichment fold > 2 for GO terms related to stress response (e.g., heat shock proteins, detoxification).

- **GO**: All three success criteria are met: environmental fraction ≥15%, ≥10 significant loci, and GO enrichment >2-fold.
- **NO-GO**: Environmental fraction <10% or <5 significant loci after FDR correction, or no GO enrichment.
- **PIVOT**: If environmental fraction is moderate (10-15%) but loci are enriched, consider increasing sample size or number of markers in Phase 3.

### 4.3 Phase 3 — Full Experimental Protocol

**Objectif**: To rigorously validate the hypothesis with a comprehensive dataset, including functional validation of candidate loci, and to produce publishable results on the role of environmental variables in shaping adaptive genetic variation and dispersal in Spodoptera litura.

**Methodologie**: 1. Expand sampling to at least 50 populations across the full environmental gradient (temperature, humidity, host plant availability) and geographic range (up to 2000 km). 2. Collect 30-50 individuals per population for robust allele frequency estimates. 3. Generate high-density SNP data using whole-genome resequencing (WGS) at low coverage (e.g., 5-10x) for a subset of individuals (e.g., 10 per population) or use ddRAD for all. 4. Obtain comprehensive environmental data including host plant availability (e.g., remote sensing data, land use maps). 5. Perform landscape genomics analyses: RDA with variance partitioning, GEA using multiple methods (LFMM, Bayenv2, RDA-based), and also test for isolation-by-environment (IBE) using partial Mantel tests. 6. Annotate candidate loci using reference genome (if available) or de novo assembly. 7. Perform functional validation: (a) Gene expression analysis (qPCR or RNA-seq) for selected candidate genes under controlled temperature/humidity stress in common garden experiments; (b) If feasible, perform CRISPR-Cas9 knockout or RNAi to confirm phenotypic effects. 8. Model dispersal corridors and barriers using resistance surfaces based on environmental suitability and adaptive allele frequencies. 9. Integrate results to inform pest management strategies.

- Cout: €15k-100k+
- Duree: 6-12 months
- Equipement: Field sampling kits, DNA extraction and library preparation equipment, Access to high-throughput sequencing (Illumina NovaSeq), Controlled environment chambers for stress experiments, qPCR machine or RNA-seq services
- Logiciels: Stacks or dDocent for SNP calling, R (vegan, LEA, gdm, topGO), Bayenv2, SLiM or msprime for simulations, ArcGIS or R for spatial analysis

**Criteres de succes :**

- Environmental fraction of genetic variation (RDA): ≥ 20% of total genetic variation explained by environmental variables after controlling for geography.
- Number of significant GEA loci: ≥ 20 SNPs with FDR < 0.05 in at least two GEA methods (e.g., LFMM and Bayenv2).
- Functional validation: At least 3 candidate genes show differential expression (fold change >2, p<0.05) under stress conditions or altered phenotype in knockout/RNAi lines.
- Dispersal model predictive accuracy: AUC > 0.80 for species distribution model based on adaptive alleles.

- **GO**: All success criteria are met: environmental fraction ≥20%, ≥20 significant loci, functional validation of ≥3 genes, and dispersal model AUC >0.80.
- **NO-GO**: Environmental fraction <15% or <10 significant loci, or functional validation fails for all candidate genes.
- **PIVOT**: If environmental fraction is moderate but functional validation is successful, consider focusing on the adaptive role of specific loci rather than genome-wide patterns.

### 4.4 Quick start : comment demarrer aujourd'hui

- **Peut demarrer maintenant**: Oui
- **Premiere action**: Search NCBI SRA for existing Spodoptera litura or S. frugiperda genomic datasets with geographic coordinates, and download environmental data from WorldClim for those locations.
- **Outils**: NCBI SRA, WorldClim, R (vegan, LEA)
- **Donnees ouvertes**: NCBI SRA, WorldClim, CHELSA

## 5. Analyse d'impact

### 5.1 Impact scientifique

Novelty score: 0.85/1.0 (novel)

### 5.2 Applications industrielles et marche

**Score industriel**: 6.5/10

**Forces :**
- Addresses a critical gap in pest management: current methods ignore environmental drivers of genetic variation, leading to suboptimal control strategies. This approach can provide actionable insights for predicting dispersal and adaptive potential, directly impacting crop protection.
- The protocol is cost-effective and time-efficient (8-14 months, €20k-80k) compared to traditional field trials, making it attractive for agrochemical companies and biotech firms seeking rapid, data-driven solutions.

**Faiblesses :**
- The market may be niche: agricultural pest management is dominated by chemical solutions, and adoption of genomic tools may be slow due to lack of expertise and infrastructure in many regions.
- The validation relies on a single pest species (Spodoptera litura), which may limit generalizability. Competitors like Bayer and Syngenta already invest heavily in pest genomics, and they may have existing in-house capabilities that could replicate this approach quickly.

**Recommandation**: Pursue a phased development: first, validate the approach with a proof-of-concept on Spodoptera litura, then partner with a major agrochemical company (e.g., Syngenta, Bayer) to co-develop a commercial service for pest risk assessment. Focus on building a proprietary database of genotype-environment associations for key pests, which can be licensed to crop protection companies. Target a timeline of 2-3 years to market, with a clear ROI based on licensing fees and consulting services.

### 5.3 Opportunites de financement

| Programme | Agence | Fit | Budget type | Taux succes |
|-----------|--------|-----|-------------|-------------|
| ANR Jeunes Chercheuses et Jeunes Chercheurs (JCJC) 2026 | Agence Nationale de la Recherche (ANR) | 0.9 | €200k-€400k | ~15% |
| ERC Starting Grant 2026 | European Research Council | 0.8 | €1.5M-€2M | ~13% |
| Horizon Europe - EIC Pathfinder Open 2025 | European Innovation Council | 0.7 | €3M-€5M | ~5% |

- **ANR Jeunes Chercheuses et Jeunes Chercheurs (JCJC) 2026** (Agence Nationale de la Recherche (ANR)): Supports early-career researchers with innovative projects; the hypothesis fits the 'Génétique, Génomique et Écologie' theme. The phased approach and moderate budget are well-suited for JCJC.
- **ERC Starting Grant 2026** (European Research Council): Ideal for ambitious, ground-breaking research. The hypothesis has high risk/high gain potential, and the ERC values interdisciplinary approaches. The larger budget would enable comprehensive validation and functional studies.
- **Horizon Europe - EIC Pathfinder Open 2025** (European Innovation Council): Supports high-risk, breakthrough technologies with societal impact. The landscape genomics approach could lead to innovative pest management tools, aligning with EIC's mission. However, the TRL is low, so a more applied angle would be needed.

**Recommandation financement**: Target the ANR JCJC for initial in silico and minimal experimental validation, then leverage ERC Starting Grant for comprehensive validation and functional studies. Build a consortium with entomologists, landscape ecologists, and bioinformaticians to strengthen the proposal.

## 6. Panel Review Summary

| Reviewer | Score | Verdict | Point cle |
|----------|-------|---------|-----------|
| methodologist | 7.2/10 | accept | The protocol clearly defines falsifiable predictions with quantitative thresholds (e.g., ≥15% variance explained, ≥10 loci) and specifies null hypotheses, which is a strong foundation for rigorous testing. |
| domain_expert | 8.0/10 | accept | The hypothesis is theoretically coherent, leveraging established landscape genomics methods (RDA, GEA) that have been validated across diverse taxa, and applies them to a novel context (agricultural pests). |
| contrarian | 4.0/10 | weak_reject | The hypothesis addresses a real gap in pest population genetics by integrating environmental drivers, which is timely and relevant for management. |
| industrialist | 6.5/10 | accept | Addresses a critical gap in pest management: current methods ignore environmental drivers of genetic variation, leading to suboptimal control strategies. This approach can provide actionable insights for predicting dispersal and adaptive potential, directly impacting crop protection. |
| funding_strategist | 7.5/10 | accept | Originality: Integrates landscape genomics into pest management, a novel approach with high potential for applied impact. |

**Consensus score**: 6.7/10
**Verdict final**: publish_brief

### 6.1 Consensus

- The hypothesis addresses a critical gap in pest management by integrating environmental drivers into population genomics, with strong potential for applied impact.
- The phased protocol (in silico, minimal, full) is pragmatic and well-structured, allowing for iterative refinement and risk reduction.
- The use of established landscape genomics methods (RDA, GEA) provides a solid methodological foundation, though careful control for population structure is essential.

### 6.2 Points de desaccord

- The contrarian reviewer argues that the assumption of migration-drift equilibrium is likely violated in agricultural pests, leading to spurious associations, while other reviewers consider the approach feasible with appropriate corrections.
- There is disagreement on the expected effect sizes and the realism of the prediction of ≥10 loci with enrichment for stress-related GO terms, with the contrarian deeming it unrealistic for many pests.

### 6.3 Critical path

The most critical factor is the control for population structure and demographic history in the GEA analyses, as this directly impacts the validity of the genotype-environment associations and the overall conclusions.

**Recommandation finale**: The panel recognizes the hypothesis as innovative and timely, with a well-designed protocol that balances rigor and practicality. While the contrarian raises valid concerns about demographic confounding and effect sizes, the consensus is that these can be addressed through careful methodological adjustments, such as incorporating population structure covariates and performing power analyses. The phased approach allows for early detection of issues, and the moderate budget and timeline make it feasible. Therefore, the panel recommends publishing the brief to proceed with the in silico phase, with the expectation that the sharpening agent will incorporate the feedback on population structure control and power calculations.

## 7. Gap Manifest residuel

### 7.1 Data gaps

- Empirical validation of landscape genomics methods on pest species with high gene flow is needed.
- Integration of dispersal corridors and adaptive potential into pest management strategies remains untested.

### 7.2 Competence gaps

- Phase 1: Population genomics analysis
- Phase 1: Landscape genomics methods (RDA, LFMM)
- Phase 1: Basic programming in R/Python
- Phase 2: Field sampling of insects
- Phase 2: DNA extraction and library preparation
- Phase 2: Bioinformatics for GBS data processing
- Phase 2: Landscape genomics analysis
- Phase 3: Population genomics
- Phase 3: Landscape genomics
- Phase 3: Functional genomics (gene expression, CRISPR)
- Phase 3: Bioinformatics
- Phase 3: Spatial analysis

### 7.3 Epistemic gaps

- The relative importance of different environmental variables in driving adaptation.
- The extent to which gene flow homogenizes adaptive divergence.
- The temporal stability of genotype-environment associations.

## References

[1] Thibaut Capblancq, B. Forester (2021). *Redundancy analysis: A Swiss Army Knife for landscape genomics*. DOI: [10.1111/2041-210X.13722](https://doi.org/10.1111/2041-210X.13722)
[2] A. Cortés, F. López-Hernández, et al. (2022). *Genome–Environment Associations, an Innovative Tool for Studying Heritable Evolutionary Adaptation in Orphan Crops and Wild Relatives*. DOI: [10.3389/fgene.2022.910386](https://doi.org/10.3389/fgene.2022.910386)
[3] L. Leamy, Cheng-Ruei Lee, et al. (2016). *Environmental versus geographical effects on genomic variation in wild soybean (Glycine soja) across its native range in northeast Asia*. DOI: [10.1002/ece3.2351](https://doi.org/10.1002/ece3.2351)
[4] Che-Wei Chang, E. Fridman, et al. (2021). *Physical geography, isolation by distance and environmental variables shape genomic variation of wild barley (Hordeum vulgare L. ssp. spontaneum) in the Southern Levant*. DOI: [10.1038/s41437-021-00494-x](https://doi.org/10.1038/s41437-021-00494-x)
[5] Xue-Xia Zhang, BaoTing Liu, et al. (2019). *Landscape genetics reveals that adaptive genetic divergence in Pinus bungeana (Pinaceae) is driven by environmental variables relating to ecological habitats*. DOI: [10.1186/s12862-019-1489-x](https://doi.org/10.1186/s12862-019-1489-x)
[6] Jia‐Xin Li, Xiuhong Zhu, et al. (2018). *Adaptive genetic differentiation in Pterocarya stenoptera (Juglandaceae) driven by multiple environmental variables were revealed by landscape genomics*. DOI: [10.1186/s12870-018-1524-x](https://doi.org/10.1186/s12870-018-1524-x)
[7] Maribet Gamboa, Kozo Watanabe (2018). *Genome-wide signatures of local adaptation among seven stoneflies species along a nationwide latitudinal gradient in Japan*. DOI: [10.1186/s12864-019-5453-3](https://doi.org/10.1186/s12864-019-5453-3)
[8] D. Belay, Pete L. Clark, et al. (2012). *Spatial Genetic Variation among Spodoptera frugiperda (Lepidoptera: Noctuidae) Sampled from the United States, Puerto Rico, Panama, and Argentina*. DOI: [10.1603/AN11111](https://doi.org/10.1603/AN11111)

## Annexes

### A. Detailed Reviewer Reports

#### Methodologist

- **Score**: 7.2/10 | **Verdict**: accept | **Confidence**: 0.8
- **Strengths**: The protocol clearly defines falsifiable predictions with quantitative thresholds (e.g., ≥15% variance explained, ≥10 loci) and specifies null hypotheses, which is a strong foundation for rigorous testing.; The phased approach (in silico, minimal, full) is pragmatic and allows for iterative refinement, reducing the risk of costly failures in later stages.; The inclusion of multiple GEA methods (LFMM, Bayenv2) and a functional validation step (gene expression, CRISPR) strengthens the evidence for adaptive significance.
- **Weaknesses**: The protocol lacks explicit controls for population structure in the GEA analyses. While geographic distance is used as a covariate in RDA, GEA methods like LFMM can still be confounded by demographic history; the protocol does not mention using principal components or admixture proportions as covariates.; The power analysis is only mentioned in the in silico phase, but there is no formal power calculation for the minimal and full phases. Sample sizes (20-30 per site, 30-50 per population) are stated, but justification based on expected effect sizes and allele frequencies is missing.; The protocol does not address potential biases in environmental data (e.g., measurement error, spatial autocorrelation) or in the selection of sampling sites. A stratified random sampling design is not described, which could introduce selection bias.
- **Questions**: How will you control for population structure in the GEA analyses? Will you include principal components or admixture proportions as covariates in LFMM/Bayenv2?; What is the statistical power to detect the expected effect sizes (e.g., ≥15% variance explained, ≥10 loci) given the proposed sample sizes and marker densities? Have you performed a priori power calculations?; How will you ensure that sampling sites are representative of the environmental gradient and not biased by accessibility or other factors? What is the spatial distribution of sites relative to the environmental variables?
- **Recommendation**: The protocol is well-structured and has a strong logical flow from in silico to full validation. However, to strengthen the methodological rigor, I recommend incorporating explicit controls for population structure in GEA, performing formal power calculations for each phase, and detailing the sampling design to minimize selection bias. With these adjustments, the protocol would be more robust and the conclusions more reliable.

#### Domain Expert

- **Score**: 8.0/10 | **Verdict**: accept | **Confidence**: 0.85
- **Strengths**: The hypothesis is theoretically coherent, leveraging established landscape genomics methods (RDA, GEA) that have been validated across diverse taxa, and applies them to a novel context (agricultural pests).; The proposed causal chain is plausible: environmental selective pressures drive allele frequency changes at adaptive loci, and landscape genomics can detect these associations, providing insights into dispersal and adaptive potential.; The evidence base includes key methodological papers (e.g., Capblancq & Forester 2021) and analogous studies in insects and plants, supporting the feasibility and potential success of the approach.
- **Weaknesses**: The hypothesis assumes pest populations are near migration-drift equilibrium, which may not hold for many agricultural pests with complex demographic histories (e.g., recent invasions, pesticide-driven bottlenecks).; The evidence base lacks direct applications to agricultural pests; most cited studies are on wild species or crops, and the transferability to pests with strong human-mediated dispersal may be limited.; The hypothesis does not address the potential confounding effects of neutral processes (e.g., isolation by distance) beyond geographic distance, and the methods (RDA, GEA) require careful correction for population structure to avoid false positives.
- **Questions**: How will the hypothesis account for the strong influence of human-mediated dispersal (e.g., agricultural trade, transport) on pest population structure, which may overwhelm natural environmental drivers?; What specific environmental variables are most likely to drive adaptive divergence in agricultural pests, and how will the study ensure these are measured at appropriate spatial and temporal scales?; Given that many pests have large effective population sizes and high gene flow, how will the study detect loci under selection when the signal may be swamped by neutral variation?
- **Recommendation**: The hypothesis is well-founded and timely, with strong potential to advance pest management through landscape genomics. However, it requires careful consideration of demographic history and human-mediated dispersal, and validation on empirical pest datasets. I recommend acceptance with minor revisions to address these caveats and strengthen the evidence base with pest-specific examples.

#### Contrarian

- **Score**: 4.0/10 | **Verdict**: weak_reject | **Confidence**: 0.75
- **Strengths**: The hypothesis addresses a real gap in pest population genetics by integrating environmental drivers, which is timely and relevant for management.; The proposed methods (RDA, GEA) are well-established and have been successfully applied in other systems, providing a solid methodological foundation.
- **Weaknesses**: FAIL REASON #1: The assumption of migration-drift equilibrium is likely violated in agricultural pests, which often experience recent range expansions, bottlenecks, and high gene flow due to human-mediated dispersal. This can lead to spurious genotype-environment associations and inflated false positive rates.; FAIL REASON #2: The hypothesis ignores the confounding effect of population structure and isolation-by-distance. Even with geographic distance as a covariate, RDA and GEA can produce false positives if environmental gradients are spatially correlated with population history. The proposed bound of 15% variance explained is optimistic given the typically small effect sizes of individual loci in polygenic adaptation.; FAIL REASON #3: The prediction of at least 10 significant loci with enrichment for stress-related GO terms is unrealistic for many agricultural pests, which may have low genetic diversity due to recent bottlenecks or strong selection. Moreover, the functional annotation of non-model species is often incomplete, making enrichment analysis unreliable.
- **Questions**: How will you account for the confounding effects of population structure and demographic history in your GEA analyses, given that standard methods like LFMM and Bayenv2 are sensitive to these factors?; What is the expected effect size of individual loci under selection in your study species, and is your sample size sufficient to detect such effects with adequate statistical power after multiple testing correction?
- **Recommendation**: To strengthen the hypothesis, the authors should first simulate realistic demographic scenarios (e.g., range expansion, admixture) to assess the false positive rates of their proposed methods. They should also incorporate additional genomic data (e.g., whole-genome resequencing) to improve functional annotation and use methods that explicitly model demography (e.g., bayescenv with population structure correction).

#### Industrialist

- **Score**: 6.5/10 | **Verdict**: accept | **Confidence**: 0.7
- **Strengths**: Addresses a critical gap in pest management: current methods ignore environmental drivers of genetic variation, leading to suboptimal control strategies. This approach can provide actionable insights for predicting dispersal and adaptive potential, directly impacting crop protection.; The protocol is cost-effective and time-efficient (8-14 months, €20k-80k) compared to traditional field trials, making it attractive for agrochemical companies and biotech firms seeking rapid, data-driven solutions.
- **Weaknesses**: The market may be niche: agricultural pest management is dominated by chemical solutions, and adoption of genomic tools may be slow due to lack of expertise and infrastructure in many regions.; The validation relies on a single pest species (Spodoptera litura), which may limit generalizability. Competitors like Bayer and Syngenta already invest heavily in pest genomics, and they may have existing in-house capabilities that could replicate this approach quickly.
- **Questions**: What is the specific commercial application: will this be sold as a service (e.g., consulting for pest risk assessment) or as a product (e.g., genomic markers for breeding resistant crops)?; How will we protect the IP? The methods (RDA, GEA) are well-established, so the value lies in the specific application and data. Can we patent the biomarkers or the analytical pipeline?; What is the willingness to pay? Will agrochemical companies pay for this information, or will they expect it as part of a larger integrated pest management package?
- **Recommendation**: Pursue a phased development: first, validate the approach with a proof-of-concept on Spodoptera litura, then partner with a major agrochemical company (e.g., Syngenta, Bayer) to co-develop a commercial service for pest risk assessment. Focus on building a proprietary database of genotype-environment associations for key pests, which can be licensed to crop protection companies. Target a timeline of 2-3 years to market, with a clear ROI based on licensing fees and consulting services.

#### Funding Strategist

- **Score**: 7.5/10 | **Verdict**: accept | **Confidence**: 0.8
- **Strengths**: Originality: Integrates landscape genomics into pest management, a novel approach with high potential for applied impact.; Feasibility: Phased protocol with clear go/no-go criteria reduces risk and allows for incremental validation, appealing to funders.; Interdisciplinary: Combines population genomics, environmental modeling, and bioinformatics, fostering collaboration across fields.
- **Weaknesses**: TRL: Currently at low TRL (conceptual/in silico), requiring validation before applied use, which may deter some applied funders.; Consortium: No explicit consortium proposed; need for partnerships with agricultural stakeholders and genomic facilities.; Budget: Relatively small for full experimental validation; may need to leverage existing data and collaborations to stay within limits.
- **Questions**: How will you access existing genomic and environmental data for Spodoptera litura, and what is the quality and coverage?; What specific environmental variables and spatial scales will be considered, and how will you account for confounding factors like population structure?; How will you ensure the functional validation of candidate loci is feasible within the budget and timeline?
- **Recommendation**: Target the ANR JCJC for initial in silico and minimal experimental validation, then leverage ERC Starting Grant for comprehensive validation and functional studies. Build a consortium with entomologists, landscape ecologists, and bioinformaticians to strengthen the proposal.

### B. Semantic Scholar Search Queries

- [novelty] `genotype-environment association crop wild relatives pests` — Direct search for the core hypothesis of applying landscape genomics methods to pests
- [novelty] `landscape genomics pest populations dispersal adaptive potential` — Search for existing studies combining landscape genomics with pest population dynamics
- [novelty] `environmental association analysis insect pests gene flow` — Check if environmental association methods have been applied to insect pests for gene flow inference
- [novelty] `redundancy analysis genetic differentiation pests environmental variables` — Look for use of RDA in pest population genetics to model environmental drivers
- [evidence] `microsatellite markers Bayesian inference pest population structure` — Evidence for standard methods in pest population genetics, often ignoring environment
- [evidence] `isolation-by-distance Spodoptera litura genetic structure` — Specific evidence for the S. litura study focusing on IBD only
- [evidence] `genotype-environment association methods crop wild relatives` — Evidence for robust statistical frameworks in crop landscape genomics
- [evidence] `environmental drivers genomic variation redundancy analysis` — Evidence for RDA and similar methods to model environmental drivers of genomic variation
- [cross_domain] `landscape genomics agriculture pest management` — Interdisciplinary precedent linking landscape genomics to pest management
- [cross_domain] `environmental association analysis adaptive loci insects` — Precedent for identifying adaptive loci in insects using environmental association
- [cross_domain] `dispersal corridors genetic connectivity environmental gradients` — Precedent for mapping dispersal corridors using environmental gradients in other taxa
- [cross_domain] `climate change pest population genetics landscape genomics` — Precedent for integrating climate change and landscape genomics in pest studies

---

*Generated by SPORE (Systeme de Production d'Opportunites de Recherche par Exploration) on 2026-08-26*