Data from: Niche shifts over spread of a biological invasion: unveiling the role of changing habitat preference and density-dependence
Data files
Apr 10, 2026 version files 5.43 MB
-
final_dataset.txt
5.43 MB
-
README.md
5.55 KB
Abstract
Aim: Anticipating the ultimate fraction of a landscape that might be susceptible to invasion is challenging as several species are able to expand the range of environmental conditions used over invasion. Despite its relevance, the more proximate processes underlying observed shifts are not sufficiently understood. Habitat selection theory predicts that as population density increases, individuals start using sub-optimal resources to compensate for the limitation of the preferred ones. However, niche shifts might also occur as a result of changes in habitat preferences over time. Here we tested these alternative hypotheses by investigating nesting resource use and selection over a biological invasion and the relative effect of density dependence on such patterns.
Location: Barcelona, Spain.
Methods: We take advantage of a large dataset recording the occurrence of the invasive Monk Parakeet, Myiopsitta monachus, in Barcelona in the period 1975–2015. We apply generalized linear models to analyse changes in nesting habitat preferences and their interactive effect with conspecific presence.
Results: Temporal changes in nesting habitat use occurred. Palm species were the most commonly used nesting substrate since the beginning of the invasion, but the diversity and frequency of use of other nesting substrates increased over time. Changes in nesting substrate use were consistent with a change in habitat preferences, that mostly involved the less-frequently used resources. Although a general role of conspecific aggregation on occupation patterns was found, it was mostly additive. That is, shifts in nesting substrate did not structure in relation to conspecific distribution.
Conclusions: Shifts in nesting resource selection can occur during the spread stage of biological invasions, probably related to innovation and learning, rather than to optimal habitat selection. These shifts present a significant challenge for forecasting efforts and management. However, changes do not happen rapidly over time, highlighting that there is a management opportunity window before significant shifts occur.
Dataset DOI: 10.5061/dryad.jdfn2z3g7
Description of the data and file structure
The file "final_dataset.txt" contains temporal occurrence data for Myiopsitta monachus in Barcelona. Information on occupied and available locations in different years is provided. For each site habitat characteristics are provided. Description of occurrence data compilation and habitat characterization is provided below.
Description of the compilation of occurrence data: The breeding population of the monk parakeets in the city of Barcelona has been monitored from 1975 [1] with information on location and substrate of breeding sites systematically compiled later by censuses conducted by the Zoology Museum of Barcelona and Museum of Natural Sciences of Barcelona (MNSB) in three years (2001, 2010 and 2015). We focused the analyses on nesting habitat selection in these three years (2001, 2010, 2015). The location of nesting sites and substrate was incorporated to a Geographic Information System (QGIS 3.22.1. program) using a UTM 31 N projection (Datum ETRS89). All points were used for variable compilation, and thus thinned to a 20-m resolution for analyses to avoid pseudoreplication (only one point among duplicated points within 20-m circular radius were retained). A total of 1269 nesting sites were available after thinning.
Description of the compilation of environmental data: We compiled information on habitat variables likely to affect nesting habitat selection of the species. This included (i) nesting substrate (i.e, the tree types) and two variables describing the availability of foraging habitat around nesting sites, namely (ii) minimum Euclidean distance to a green area (dgreen) and (iii) the percentage of green areas in a 200-m radius around the nest (green200). This radius (r) was calculated on the basis of the formula r = (A/π)1/2, where the area (A) is the mean kernel home-range size of the species, 12.4 Ha [2]. The data on tree species was obtained from the official website of the Barcelona City Council updated in 2022 (https://opendata-ajuntament.barcelona.cat/data). We considered 4 main categories for analyses representing the 3 most commonly used tree types used by the species (that is, palms (Phoenix species), pine trees (Pinus species) and plane trees (Platanus species) as well as a category others including other tree types not so frequently used (Table S1, each representing ≤5% of total records)). Human infrastructures were only scantily used by the species in the study area (1.8% of total records) and were not considered in habitat selection analyses due to the difficulty in estimating their availability. Data on green areas was obtained from MNSB (categories Verd privat and Parcs i Jardins Urbans [3].
To assess the effect of conspecific presence on habitat use and selection patterns, we estimated the minimum Euclidean distance from each tree to the nearest occupied tree in the previous census period (NND) and an aggregation index (aggregation) describing the relative position of a given tree within the spatial distribution of the breeding population in the previous period in a 200-m radius. Aggregation was obtained using the formula ∑exp(−dij) with (i ≠ j), with d representing the distance of occupied nests j with respect to a previously occupied tree i. To calculate these variables in our first study year (2001), we considered as previously occupied sites, the pool of nesting locations recorded in previous years 1975-1995. We used this criterion because occupied nests in the population are commonly maintained over time.
References:
[1] Batllori, X., & Nos, R. (1985). Presencia de la cotorrita gris (Myiopsitta monachus) y de la cotorrita de collar (Psittacula krameri) en el rea metropolitana de Barcelona. Misc. Zool, 9(June 2015), 407411. http://other.museucienciesjournals.cat/files/MZ-vol-9-1985-pp-407-411.pdf
[2] Senar, Juan Carlos, Moy, A., Pujol, J., Tomas, X., & Hatchwell, B. J. (2021). Sex and Age Effects on Monk Parakeet Home-Range Variation in the Urban Habitat. Diversity, 13(12), 648. https://doi.org/10.3390/d13120648 [3] GBIF.org (11 May 2020) GBIF Occurrence Downloadhttps://doi.org/10.15468/dl.5a4ax6
[3]RodrguezPastor, R., Senar, J. C., Ortega, A., Faus, J., Uribe, F., & Montalvo, T. (2012). Distribution patterns of invasive Monk parakeets (Myiopsitta monachus) in an urban habitat. Animal Biodiversity and Conservation, 35.1, 107117.
Files and variables
File: final_dataset.txt
Description:
Variables
- year: describes the year of availaibility/occupation of a site
- x: longitud (ETRS89, UTM31N)
- y: latitude (ETRS89, UTM31N)
- hab: nesting substrate
- NND: minimum Euclidean distance to the nearest occupied tree in te previous period in meters
- agreg: aggregation index describing the relative position of a given nesting site within the spatial distribution of the breeding population in the previous census period in a 200-m radius
- dgreen: minimum Euclidean distance to a green area in meters
- green200: surface of green areas in a 200-m radius buffer aroud a nesting site
- occupied: describe whether a site is occupied or not
