Methods
Samples and DNA extraction
North American tissue samples (N=95) were obtained from starlings culled by the United States Department of Agriculture Animal and Plant Health Inspection Service (USDA APHIS) between 2011-2018 at 14 localities across the US (Table 1). Samples were shipped in ethanol and frozen at -20°C upon arrival. For fresh samples collected by the USDA, DNA extraction was performed using the Qiagen Blood and Tissue Kit. Extractions were performed at room temperature, with an overnight incubation at 56°C to completely lyse muscle tissue. Elution buffer was warmed to 56°C and incubated for 30 minutes before the final spin and elution. Extracted DNA samples were stored at -20°C. DNA was also extracted from native range samples (N=2) from National Museums Scotland, Edinburgh.
 Sequences from starlings sampled on other continents were downloaded from GenBank including those from the native range (N=43; Berthouly-Salazar et al., 2013; Rollins et al, 2011; GenBank KF638591–617; HQ263631–42), Australia (N=597; Rollins et al, 2011; GenBank 178 FJ542126.1–FJ542131.1, FJ542133.1, HQ2636230– HQ263630) and South Africa (N=219; Berthouly-Salazar et al., 2013; GenBank KF638591–617). These samples were analysed together with the North American samples (see below). The total number of individuals included in the study was 956 (Table 2).
Amplification and Sequencing
The primers used to amplify the mitochondrial control region in North American specimens were initially designed to analyze the mitochondrial diversity of the Australian population (Rollins et al, 2011). Rollins et al. (2011) designed a series of overlapping primers to be utilized in the amplification of museum specimens or highly degraded samples (Table S1). We used these primers to sequence the control region of North American samples in four overlapping segments. Two of these primers (svCRL1 and svPheH3) amplify most of the mitochondrial control region and also were used to amplify DNA from the starling population in South Africa (Berthouly-Salazar et al., 2013).
For the PCR reactions, PuReTaq Ready-To-Go PCR Beads were rehydrated with 13.5 µl of molecular grade water, 5 µl of 10 µM forward and reverse primers and 1.5 µl of DNA. The thermocycling conditions used here were identical to those described in the original paper (Rollins et al, 2011). This included a 5-min step at 94°C, 30 cycles of 94°C for 30s, 53°C for 15s and 72°C for 30s and a final extension step for 10 min at 72°C. PCR products were sent to GENEWIZ, Inc (South Plainfield, NJ) for PCR clean up via an enzymatic purification. Sequencing reactions were performed by GENEWIZ, Inc using Applied Biosystems BigDye version 3.1 and forward primers. The reactions were then sequenced on an Applied Biosystems 3730xl DNA Analyzer.
 
Population and Expansion Analysis
Overlapping sequences were aligned using the software Geneious 11.1.2 (Kearse et al., 2012) to generate a consensus sequence for each individual from North America. All subsequent alignments including the samples from other continents were generated on Geneious using the standard settings and the Geneious alignment algorithm (Kearse et al., 2012). Median joining haplotype networks were created using Network v10.1.0.0 (Bandelt et al. 1999) and post-processed using the maximum parsimony calculation to remove unnecessary median vectors (Polzin & Daneshmand 2003). Final networks were produced using Network Publisher v2.1.2.5 (Fluxus Engineering, Clare, UK). Networks were constructed for the North American samples (1181bp) and for the full dataset. To construct the latter, we trimmed the full dataset to 928bp to accommodate the continental dataset with the shortest sequence length.
Using the 928bp dataset, we calculated fixation indices (FST values), pairwise nucleotide diversity (π), haplotype diversity (h), Fu’s Fs and Tajima’s D in Arlequin v3.5.1.2 (Excoffier & Lischer 2010). Haplotype richness (R) for each population was calculated using FSTAT v.2.9.4 (Goudet, 2003). Mismatch analyses were also conducted in Arlequin, but with the full dataset from each invasive population (see Table S5).