Data

Data - modified hierarchical community distance sampling model

1. Cisticolidae_raw_data.csv: input file to fit a joint species distribution model for 5 species of warblers (Cisticolidae).

   • Variables

• pt: unique identification for each sampling point.
• family: taxanomic rank of birds observed at each sampling point.
• spp_id: unique identification for bird species.
• dist: radial distances (meters) at which observations (birds) were made at sampling points.
• counts: observed counts (birds) at each sampling point.
• elev: elevation data at each sampling point.

2. Virunga_covariates.csv: input file to fit a joint species distribution model for 5 species of warblers (Cisticolidae).

   • Variables

• pt: unique identification for each sampling point.
• elev: elevation data at each sampling point.
• family: taxanomic rank of birds observed at sampling points.
• nha8: unique identification for habitat types recorded at sampling points.

Data - hierarchical community distance sampling model

1. HCDSM_raw_data.csv: input file to fit a hierarchical community distance sampling model for 63 species

   • Variables

• pt: unique identification for each sampling point.
• family: taxanomic rank of birds observed at each sampling point.
• spp_id: unique identification for bird species.
• dist: radial distances (meters) at which observations (birds) were made at sampling points.
• counts: observed counts (birds) at each sampling point.
• elev: elevation data at each sampling point.

2. Virunga_covariates.csv: input file to fit a joint species distribution model for 5 species of warblers (Cisticolidae).

   • Variables

• pt: unique identification for each sampling point.
• elev: elevation data at each sampling point.
• family: taxanomic rank of birds observed at sampling points.
• nha8: unique identification for habitat types recorded at sampling points.

Data to estimate niche overlap indices

1. *_diet.csv: (15 files)- files contain preferred diet categories to estimate diet niche overlap indices for bird species pairs within a family.

   • Variables
     • Species : scientific name of a bird.
     • Diet-Fruit : Estimated percent use of fruits by a bird species in its diet.
     • Diet-Inv : Estimated percent use of invertebrates by a bird species in its diet.
     • Diet-Nect : Estimated percent use of nectar by a bird species in its diet.
     • Diet-PlantO: Estimated percent use of plant materials by a bird species in its diet.
     • Diet-Seed : Estimated percent use of seeds by a bird species in its diet.
     • Diet-Vect : Estimated percent use of reptiles, snakes, amphibians and/or salamanders by a bird species in its diet.
     • Diet-Vend : Estimated percent use of mammals by a bird species in its diet.
     • Diet-Vunk : Estimated percent use of vertebrates (i.e., general or unknown) by a bird species in its diet.

2. *_strata.csv: (15 files)- files contain preferred vertical foraging forest strata categoriesto estimate vertical foraging forest strata niche overlap indices for bird species pairs within a family.

   • Variables
     • Species : scientific name of a bird.
     • ForStrat-ground : Estimated percent foraging on the ground by a bird species in a forest.
     • ForStrat-understory : Estimated percent foraging below 2m by a bird species in a forest.
     • ForStrat-midhigh : Estimated percent foraging the middle strata by a bird species in a forest.
     • ForStrat-canopy : Estimated percent foraging in a forest canopy by a bird species.
     • ForStrat-aerial : Estimated percent foraging above the forest canopy by a bird species.

3. Diet_categories_63species.csv: (1 file)- Conntains the preferred diet categories (i.e., Diet-Fruit, Diet-Inv, Diet-PlantO, Diet-Vect, Diet-Vunk) for all the 63 bird species in our bird community. Data to generate a null distribution (i.e., niche overlap) along the diet niche dimension for species pairs across families (i.e., community).

   • Variables
     • Diet-Fruit : Estimated percent use of fruits by a bird species in its diet.
     • Diet-Inv : Estimated percent use of invertebrates by a bird species in its diet.
     • Diet-Nect : Estimated percent use of nectar by a bird species in its diet.
     • Diet-PlantO : Estimated percent use of plant materials by a bird species in its diet.
     • Diet-Seed : Estimated percent use of seeds by a bird species in its diet.
     • Diet-Vect : Estimated percent use of reptiles, snakes, amphibians and/or salamanders by a bird species in its diet.
     • Diet-Vend : Estimated percent use of mammals by a bird species in its diet.
     • Diet-Vunk : Estimated percent use of vertebrates (i.e., general or unknown) by a bird species in its diet.
     • Sp_* : Unique identification for species (63) within the community (Full species names are provided in Appendix S5).

4. strata_categories_63species.csv: (1 file)- contains preferred vertical foraging forest strata categories for each bird species by family to generate a null distribution (i.e., niche overlap) along the vertical foraging forest strata niche dimension for species pairs across families (i.e., community).

   • Variables
     • ForStrat-ground : Estimated percent foraging on the ground by a bird species in a forest.
     • ForStrat-understory : Estimated percent foraging below 2m by a bird species in a forest.
     • ForStrat-midhigh : Estimated percent foraging the middle strata by a bird species in a forest.
     • ForStrat-canopy : Estimated percent foraging in a forest canopy by a bird species.
     • ForStrat-aerial : Estimated percent foraging above the forest canopy by a bird species.
     • Sp_* : Unique identification for species (63) within the community (Full species names are provided in Appendix S5).

5. indices.figures.mean.cv.scale.parameter.csv: (1 file)- Contains the mean coefficient of variation (null distribution) values for the detection scale parameter for species pairs across families.

   • Variables

     • col[1] : number of samples (each sample = 60 species pairs) used to generate a null model expectation
     • col[2] : mean coefficient of variation (each sample = 60 species pairs)

6. Virunga_covariates.csv: (1 file)- File used as an input to generate expected abundance.

   • Variables
     • pt: unique identification for each sampling point
     • elev: elevation data at each sampling point
     • nha8: unique identification for habitat types recorded at sampling points

7. HCDSM63spp.RData: (1 file) Contains parameter estimates from the hierarchical community distance sampling model (63 species of birds). The file can be downloaded from Dryad using this link: https://datadryad.org/stash/dataset/doi:10.5061/dryad.fttdz08z8

8. HCDSM_warblers_elev.RData: (1 file)- Model results (with covariate) for the warbler (Cisticolidae) community model extension used to estimate residual correlations among species pairs.

9. HCDSM_warblers_null.RData: (1 file)- Model results (null model) for the warbler (Cisticolidae) community model extension used to estimate residual correlations among species pairs.

10. Species_list.csv: (1 file)- A list of all 63 bird species (from 32 families) that were included in our hierarchical community distance sampling model to estimate abundance variation along an elevation gradient in the Virunga volcanoes.

Data - body mass (63 bird species)

1. Bodymass_63species.csv: File contains data used to calculate the coefficient of variation for 60 random species pairs from a community of 63 species.

   • Variables
     • BodyMass: describes the average body mass for each bird species.
     • Sp_* : Unique identification for species (63) within the community (Full species names are provided in Appendix S5).

Data - null model distributions ( elevation, diet, forest strata)

Folder contents: contains 4 files (i.e., prop.weak.overlap.samples.elev, prop.strong.overlap.samples.elev, prop.weak.overlap.samples.d, prop.strong.overlap.samples.d, prop.weak.overlap.samples.s, prop.strong.overlap.samples.s ).

Purpose: Generated to assess whether the observed niche overlap indices for the 60 species pairs were significantly different from random null expectations (i.e., across families) for elevation, diet and foraging vertical strata niche dimensions.

1. prop.weak.overlap.samples.elev.csv

   • Variables
     • col 1: Proportion of weak elevation niche overlap indices for 60 species pairs. Note: elevation niche partitioning is considered weak when the index is < 0.5
     • Sp_* : Unique identification for bird species (63) within the community (full species names are provided in Appendix S5).

2. prop.strong.overlap.samples.elev.csv

   • Variables
     • col 1: Proportion of strong elevation niche overlap indices for 60 species pairs. Note: elevation niche partitioning is considered strong when the index is ≥ 0.5.
     • Sp_* : Unique identification for bird species (63) within the community (full species names are provided in Appendix S5).

3. prop.weak.overlap.samples.d.csv

   • Variables
     • col 1: Proportion of weak diet niche overlap indices considering 60 species pairs. Note: diet niche partitioning is considered weak when index is < 0.5.
     • Sp_* : Unique identification for bird species (63) within the community (full species names are provided in Appendix S5).

4. prop.strong.overlap.samples.d.csv

   • Variables
     • col 1: Proportion of strong diet niche overlap indices considering 60 species pairs. Note: diet niche partitioning is considered strong when index is < 0.5.
     • Sp_* : Unique identification for bird species (63) within the community (full species names are provided in Appendix S5).

5. prop.weak.overlap.samples.s.csv

   • Variables
     • col 1: Proportion of weak vertical foraging forest strata niche overlap indices considering 60 species pairs. Note: diet niche partitioning is considered weak when index is < 0.5.
     • Sp_* : Unique identification for species (63) within the community (Full species names are provided in Appendix S5).

6. prop.strong.overlap.samples.s.csv

   • Variables
     • col 1: Proportion of strong vertical foraging forest strata niche overlap indices considering 60 species pairs. Note: vertical foraging forest strata niche partitioning is considered strong when index is < 0.5.
     • Sp_* : Unique identification for species (63) within the community (Full species names are provided in Appendix S5).

Data - rank abundance

1. Rank.sp44.csv: Relative abundance per species - derived from this study which comes from Parc National des Volcans in Rwanda and Derhé et al. 2020 using a multi-year data set collected in Parc National des Volcans.

   • Variables

• Species: scientific name of a bird.
• This.study: Relative abundance of birds (only Parc National des Volcans).
• Derhe.et.al.2020: Relative abundance of birds derived from Derhé et al. 2020.
• Total.no.spp-5yrs: Relative abundance of birds from Derhé et al. 2020 across five years (i.e 10 sampling occasion).

2. Rank.sp51.csv: Relative abundance per species - derived from this study which comes from three national parks (i.e., Mgahinga Gorilla National Park in Uganda, Parc National des Volcans in Rwanda, Parc National des Virunga in the Democratic Republic of Congo). and Derhé et al. 2020 using a multi-year data set collected in Parc National des Volcans.

   • Variables

• Species: scientific name of a bird.
• This.study: Relative abundance of birds (only Parc National des Volcans).
• Derhe.et.al.2020: Relative abundance of birds derived from Derhé et al. 2020.
• Total.no.spp-5yrs: Relative abundance of birds from Derhé et al. 2020 across five years (i.e 10 sampling occasion).

Data - spatial site index

1. Community.63sp.csv: File contains lantent abundance of birds along an elevation gradient -- generated from a hierarchical community distance sampling model.

   • Variables
     • column 1: sampling points.
     • Subsequent columns- Unique identification for each species (i.e., latent abundance at each sampling points in the study area).

2. Raw_counts.csv: File contains raw counts to estimate the spatial site index.

   • Variables
     • column 1: sampling points.
     • Subsequent columns- Unique identification for each species (i.e., raw counts at each sampling point in the study area).

3. spatial.site.index.mean.cv.scale.parameter.csv: File contains data on the mean coefficient of variation (null distribution) for the detection scale parameter (σ_js) for species pairs across families.

   • Variables
     • col[1] - number of samples (each sample = 60 species pairs) used to generate a null model expectation.
     • col[2] - mean coefficient of variation (each sample = 60 species pairs).

4. mean.site.indx.samples_latent.csv: File contains data to estimate the spatial site index based on mean latent abundance (null distribution).

   • Variables
     • col[1] - number of samples (each sample = 60 species pairs) used to generate a null model expectation.
     • col[2] - mean latent abundance (each sample = 60 species pairs).

Data attributes

1. Data attributes.txt: Files containing metadata (mostly duplicated in this README). References a slightly different file structure but also has info on file size.