Plants & fungi / Behavioral experiment

When another organism responds

Source year 2026 · Note reviewed 2026-10-24

In controlled choices, female Egyptian cotton leafworm moths (Spodoptera littoralis) changed where they laid egg clusters when playback of drought-stressed tomato-plant clicks was present. The preference depended on the other cues: playback attracted oviposition in a plant-free arena, while moths preferred the silent side when healthy tomato plants were offered on both sides. Deafening removed the plant-free playback preference; male-moth click controls did not produce a detected preference. A separate long-arena experiment used a lower 30-click/min rate and found egg clusters concentrated near the playback and feeder rather than the silent edge. A checksum-matched aggregate audit of the single Mendeley data.xlsx file also matched the five main Table 1 tuples for repetitions, paired arena/date observations and egg-cluster totals; this is a source-table consistency check, not a statistical-model reproduction.

What we reviewed · Full-text methods, results and public review record reviewed; one checksum-pinned source table checked in aggregate

Full version-of-record text reviewed through the NCBI PMC copy and eLife PDF, including main methods, figures, public reviewer comments and author responses. The public Mendeley Data v1 inventory lists three folders with 32 files (24 CSV, four R, two MATLAB and two XLSX); its page and file panels were reviewed separately from the article. On October 24, one 51,160-byte data.xlsx file was acquired to ignored external storage and matched the Mendeley file-panel SHA-256. A standard-library aggregate-only audit matched all five main Table 1 repetition, paired-observation and egg-cluster totals. No spreadsheet formulas or deposited code were executed, no inferential model was fit, the other files and publisher ZIP remain uninspected, and no independent specialist review was obtained.

Not yet independently reviewed by a scientific specialist.

What this does — and doesn’t — establish

This is one moth species, one plant species and a controlled laboratory/greenhouse study, not evidence that plants intentionally communicate or that animals can be translated by AI. The authors characterize drought sounds as a likely cue rather than an evolved message. The main choice experiments used 60 clicks/min; the authors measured at most 20 clicks/min from 45 drying seedlings in a 50-hour recording and argue that a dense patch could raise exposure. The later 30-click/min gradient setup differs in arena and subject design, so it is not a direct replication of the main choice trials. Preferences also reverse with plant context; other visual, olfactory and tactile cues complicate interpretation. The Mendeley inventory contains tabular and XY-point files, not listed audio recordings; it cannot serve as an animal-audio benchmark. Although one workbook was checksum-verified and parsed, this does not verify the remaining files, cross-file joins, dependencies or correspondence to the publisher ZIP. Reviewer comments raised ecological realism, effect-size presentation, availability of raw data/code and species generalization; Talk2Nature has not independently reproduced or reanalysed the results. The workbook contains two additional condition labels outside the five Table 1 groups compared here. In particular, its separate Bat oviposition group is not mapped to a reported oviposition outcome in the reviewed article, which mentions bat playback only in a flight-response validation; exclude that group from interpretation pending source clarification. The lower-rate workbook condition is not a direct replacement for the main 60-click/min choices. The full eLife source-data ZIP and the rest of the Mendeley files remain unacquired.

What we can build on

A receiver's controlled response can show that a sound is a usable cue in a specific setting. That is a stronger behavioral result than detecting a signal, but it does not by itself establish a sender's intent, a shared message or conversation.

The next reading task

Obtain qualified insect-bioacoustics and statistical review of the experimental units, table mapping and analysis before any inferential reproduction. Resolve the workbook Bat condition against an explicit source description; do not infer a plant-response result from its label. Keep the separate publisher ZIP unacquired unless exact comparison becomes necessary and its artifact terms are reviewed. This behavioral table is not an audio/model-training benchmark; ordinary phone microphones are not ultrasonic recorders.

Go to the sources

Moth responses to plant sounds ↗
Seltzer et al. Female moths incorporate plant acoustic emissions into their oviposition decision-making process. eLife; version of record 5 January 2026. The article is CC BY 4.0; earlier reviewed preprints appeared in 2024 and 2025.
Israeli research connection ↗
American Friends of Tel Aviv University. Plant/animal auditory interaction (14 July 2025). Identifies the Yossi Yovel and Lilach Hadany research groups.
Moth–plant acoustic study: full text and public review record ↗
NCBI PubMed Central full text of the eLife version of record, including methods, figures, reviewer comments and author responses.
eLife figures and source data for the moth–plant acoustic study ↗
The publisher supplement is listed as Source data 1, a 1.2 MB ZIP described by PMC as raw data tables and analysis code; the figures page also lists raw scatter-plot data for Figure 1. The linked archive itself was not downloaded or inspected.
Mendeley Data v1: moth oviposition tables and analysis code ↗
Author-linked data deposit for the eLife study; its record describes the content as codes and statistics and displays a CC0 1.0 license.
Mendeley Data v1 file inventory and integrity metadata for the moth study ↗
The public file browser lists 32 files in three folders: basic_arena_fig1_S2 (6), coridor_arena_fig2_S3_S4 (5), and Location_fig_3 (21). By displayed extensions these are 24 CSV, four R, two MATLAB and two XLSX files. No audio extension appears in the inventory. The 50 KB data.xlsx preview shows experiment, date, arena, repetition, control, speaker side, female count, examined side and egg-cluster columns. A partial run_stats_whole.m preview reads data.xlsx. The Download All ZIP is displayed as 1.22 MB with SHA-256 0cce953171fcf416ead21696f30b3234e7bee70e93f7b5a6a8bc403765156c8a.

Original editorial note by Talk2Nature contributors. AI-assisted, source-grounded synthesis; not a peer review or an independent replication. Linked works retain their own rights.