Concern about the impacts of biological invasions has generated a great deal of interest in understanding factors that determine invasion success. Most of our current knowledge comes from static approaches that use spatial patterns as a proxy of temporal processes. These approaches assume that species are present in areas where environmental conditions are the most favourable. However, this assumption is problematic when applied to dynamic processes such as species expansions when equilibrium has not been reached. In our work, we analyse the roles played by human activities, climatic matching, and spatial connectivity on the two main underlying processes shaping the spread of invasive species (i.e., colonisation and extinction) using a dynamic modelling approach. For this, we used a large dataset that has recorded the occurrence of two invasive bird species -the ring-necked and the monk parakeets-  in the Iberian Peninsula from 1991 to 2016. 

Temporal occurrence data for the monk and ring-necked parakeets were obtained from a comprehensive database of exotic birds in mainland Spain and Portugal, which compiled records of exotic species observed in the wild in both countries from 1912 to 2012 through a systematic review of scientific and grey literature and observations from local experts [1]. This dataset was updated until 2016 using the same methodology and complemented with ‘human observation’ data from the Global Biodiversity Information Facility [2,3].

Locations were incorporated to a Geographic Information System (GIS) using a cylindrical equal-area projection at 10 km resolution to fit the maximum daily distances covered by the species. We used as sampling sites for analyses the 10-km grid cells in the Iberian Peninsula. The occurrence data in each sampling sites was classified in surveys seasons and replicate observation periods within seasons using the date of the records. To account for potential variation related to the criteria used to classified the data, we considered three alternative sampling schemes: (1) survey seasons of one calendar year with two replicate observation periods (Jan-Jun and Jul-Dec), (2) survey seasons of two calendar years with two replicate observation periods (each of 1 calendar year) and (3) survey seasons of three calendar years with two replicate observation periods (each of 1.5 years). To account for potential detection biases related to an uneven sampling effort across time and space, we included an estimate of sampling effort as a survey-specific covariate of detection probability in models. This variable was computed as the cumulative value of observation records of both native and alien bird individuals retrieved from GBIF (‘human observation’ category [4]) in a particular sampling site and observation period considered. 

As sampling site covariates, we calculated the climatic similarity between each of the sites in the study area and the species native ranges using multivariate environmental similarity surfaces (MESS) and compiled information on three variables describing human-transformed environments: (i) the Global Human Influence Index [5] and two more specific descriptors of anthropogenic habitats known to affect invasions, the percentage of ii) urban environments (including urban and built-up areas) and iii) farmland. These two land-use variables were derived from data provided by the USGS Land Cover Institute (LCI) (https://landcover.usgs.gov/) at 500m resolution using ArcMap 10.5.

More detailed description of methods can be found in the manuscript.

References:

[1] Abellán, P., Carrete, M., Anadón, J. D., Cardador, L., & Tella, J. L. (2016). Non-random patterns and temporal trends (1912-2012) in the transport, introduction and establishment of exotic birds in Spain and Portugal. Diversity and Distributions, 22(3), 263–273. https://doi.org/10.1111/ddi.12403

[2] GBIF.org (03 June 2020) GBIF Occurrence Download https://doi.org/10.15468/dl.mwj59n

[3] GBIF.org (11 May 2020) GBIF Occurrence Download https://doi.org/10.15468/dl.5a4ax6

[4] GBIF.org (03 June 2020) GBIF Occurrence Download https://doi.org/10.15468/dl.mudmhx

[5] Sanderson, E. W., Malanding, J., Levy, M. A., Redford, K. H., Wannebo, A. V., & Woolmer, G. (2002). The human footprint and the last of the wild. BioScience, 52, 891–904. https://doi.org//10.1641/0006-3568(2002)052[0891:THFATL]2.0.CO;2

Description of the files:

File "README.txt": Description and metadata associated with data files and scripts uploaded

File "input_data_myimon.csv": Temporal occurrence data for Myiopsitta monachus in the Iberian Peninsula from 1975 to 2016.

File "input_data_psikra.csv": Temporal occurrence data for Psittacula krameri in the Iberian Peninsula from 1970 to 2016.

File "myimon_1y2sampl_data_with_vars.csv": Detection history, sampling effort, site covariates and testing data for Myiopsitta monachus according to survey seasons of 1 year with two replicate observation periods.

File "myimon_2y2sampl_data_with_vars.csv": Detection history, sampling effort, site covariates and testing data for Myiopsitta monachus according to survey seasons of 2 years with two replicate observation periods.

File "myimon_3y2sampl_data_with_vars.csv": Detection history, sampling effort, site covariates and testing data for Myiopsitta monachus according to survey seasons of 3 years with two replicate observation periods.

File "psikra_1y2sampl_data_with_vars.csv": Detection history, sampling effort, site covariates and testing data for Psittacula krameri according to survey seasons of 1 year with two replicate observation periods.

File "psikra_2y2sampl_data_with_vars.csv": Detection history, sampling effort, site covariates and testing data for Psittacula krameri according to survey seasons of 2 years with two replicate observation periods.

File "psikra_3y2sampl_data_with_vars.csv": Detection history, sampling effort, site covariates and testing data for Psittacula krameri in the period 1991-2013 according to survey seasons of 3 years with two replicate observation periods.

File "Prepare_datasets_for_PRESENCE.R": Code from R to set the data in the format for software PRESENCE.

File "Run_multiseason_occupancy_models.R": Code from R to conduct multiseason occupancy models using the package RPresence in R.