Solenopsis invicta workers were collected from a nest in Tucheng District, New Taipei City, Taiwan (23624°58′41.1″ N, 121°26′49.6″ E). The ants were kept in a plastic tray containing a piece of damp cotton wool and reared on honey water (ca. 10%) in a room at around 25 °C and 70% RH in Taipei City, Taiwan.
As a fumigant, we used 50 g of microencapsulated AITC pellets containing in total 1 g of AITC (the weight of each pellet was 0.01 g, and contained 2% of AITC by weight); the pellets were obtained from PRD Co., Ltd. (Osaka, Japan). A transparent gas-barrier bag (91 cm width × 60 cm length × 40 cm depth, 80 L) used for fumigation consisted of a poly(ethylene terephthalate) film (12 μm thickness) deposited on a film of linear low-density polyethylene (80 μm thickness). The gas-barrier bags were developed by Moriya Industry Co., Ltd. (Osaka, Japan).
.For the ease of handling, 10 workers of S. invicta were placed in a polypropylene snap-cap vial (50 mL, 3.5 cm diameter, and 7.3 cm depth) with a piece of damp cotton wool. The vial had a cap with a mesh-covered 3 cm diameter opening, which prevented the ants from escaping and allowed air to flow into the inside (Fig. 1a). For fumigation treatment, we placed the vial in a cardboard box (15 cm width × 29 cm length × 13 cm depth) and then put the box in a gas-barrier bag with microencapsulated AITC pellets holding in a tea filter bag (Fig. 1b, c). In the control treatment, the bag contained no pellets. In both treatments, S. invicta were removed from the bag after 24 h of exposure, and mortality was determined by observation. Ants were considered dead if they did not respond to probing. They were considered moribund when locomotion was not possible and the movement of appendages was feeble and uncoordinated. We counted the number of moribund ants as alive but confirmed whether they showed any subsequent recovery of activity during the next day. Using four gas-barrier bags, we performed two trials in each of the control and fumigation treatments daily and were repeated for 5 consecutive days (i.e., 10 trails per treatments).
For statistical analysis, to compare mortality rates between control and fumigation treatments, we used generalized linear mixed models (GLMM) with binomial distribution and a logistic regression link. To control for potential differences in gas-barrier properties among the bags and conditions of trial days, we included “bag differences” and “trail days” as random intercept effects in the GLMM. Analysis of the GLMM was performed using R package lme4 version 1.1-21 (Bates et al. 2015; R Core Team 2019).
AITC release kinetics in the gas-barrier bag
To evaluate the ability of the gas-barrier bag to maintain AITC vapor, we measured the kinetics of AITC release from the pellets in the bag. As in the fumigation treatment, we put 50 g of the microencapsulated AITC pellets together with a cardboard box containing a piece of damp cotton wool into a gas-barrier bag, but without S. invicta. Release kinetics were measured using the colorimetric gas detector tube (No.149, Gastec Corp., Ayase, Japan; measuring range 5–500 ppm). The detector tubes are graduated glass tubes filled with chemical reagents that produce a color stain when exposed to the target gas. The length of the color change is proportional to the concentration of the gas being tested (Noda et al. 2001). Gas in the bag was pulled out by a gas sampling pump (Ap-1, Komyo Rikagaku Kogyo KK, Japan) and passed through a detector tube with a needle from a hole in the bag. The hole was tightly closed with sealing tape to avoid gas leakage during the experiment. Gas samples were taken at 1, 3, 6, and 24 h and 2, 5, and 13 days. The measurements had three replicates. We assessed AITC release kinetics in a laboratory of PRD Co., Ltd. in Osaka City, Japan, at around 25 °C and 70% RH.
We observed ant behavior in responses to the vapor. A single worker of S. invicta and a microencapsulated AITC sheet (40 mm × 40 mm, 1 mm thick, AITC content 0.01 g) was placed in a covered glass Petri dish (64 mm diameter, 20 mm height) with a piece of damp cotton wool (Fig. 1d). The behavior of the worker within the arena was observed and video-recorded on a smartphone camera (iPhone XS produced by Apple Inc, USA).