*2.2. DNA Extraction and Amplification*

DNA was purified and amplified as described in Symes et al. [40]. Briefly, DNA was purified from dissected digestive tracts following the manufacturer's instructions using the QIAGEN DNeasy Plant Mini Kit (Qiagen 69104; Qiagen, Hilden, Germany) with the following modification. After initial homogenization with micropestles, 50 μL of AP1 buffer was added and the sample was further crushed before adding the 350 μL of AP1 buffer. Primers were utilized to amplify three conserved regions of the plastid genome following the procedure described in [43] (Table S1). Of the three regions, the *rbcL* region has the highest sequence conservation across plant species and is easy to sequence, the *psbA/trnH* region is intermediate in variability, but sometimes difficult to interpret, and the *matK* region is highly variable, but often difficult to amplify. PCR reactions were prepared using 5 μL template, 20 μL GoTaq Green PCR mastermix (Promega M7122; Madison, WI, USA), 0.5 μM F and 0.5 μM R primer, and water to a final volume of 40 μL. Thermocycler (ABI Veriti 96-well, model 9902) conditions for *rbcLa* and *psbA-trnH* were: 95 ◦C for 3 min, 35 cycles (95 ◦C for 30 s, 55 ◦C for 30 s, 72 ◦C for 1 min), 72 ◦C for 10 min and for *matK* were: 95 ◦C for 3 min, 40 cycles (95 ◦C for 30 s, 50 ◦C for 30 s, 72 ◦C for 1 min), 72 ◦C for 10 min. Samples were analyzed with *rbcLa* and *psbA-trnH* primers first, and then PCR was conducted with *matK* primers if needed to achieve a minimum of two successful primer sets per sample.

For positive samples, the remaining 30 μL of PCR product was purified using a Wizard PCR Clean-Up System (Promega A9281; Madison, WI, USA) and resuspended in 50 μL nuclease-free water. 2 μL of purified PCR product was mixed with 25 ng of the appropriate forward primer for the amplicon (*rbc*, rbc\_SI\_For; *psb*, psbA3'f; *matK*, matKfor\_KIM3F) in a 15 μL reaction volume and sequenced using Sanger sequencing through Genewiz (Genewiz, South Plainfield, NJ, USA). Sequences were trimmed using Geneious and a 0.1 error probability limit. Following editing, sequences were exported in FASTA format for analysis with BLAST. Sequence data is available in FASTA format with the Supplementary Materials.
