What controls the puppets? 20-hydroxyecdysone may not drive manipulation in Leucauge spiders

Data associated with the manuscript: What controls the puppets? 20-hydroxyecdysone may not drive manipulation in Leucauge spiders Ítalo Mendes Delazari, Alexander Gaione-Costa, Anderson Raimundo da Silva Santos, Jenny Dimelza Gomez Arrieta, Jober Fernando Sobczak, Stanislav Korenko & Thiago Gechel Kloss Abstract: Parasites frequently manipulate host behavior to increase their own fitness, but the proximal mechanisms underlying these manipulations remain poorly understood. In interactions between Polysphinctine parasitoid wasps and spiders, hosts are induced to build modified and more resistant webs (cocoon webs). This construction is currently attributed to the activation of the host's innate ecdysis mechanism, based on the cocoon web's similarity to spider molting webs and the association with elevated levels of the hormone 20-hydroxyecdysone (20E) in Araneidae species. However, the only empirical evidence of 20E enhancement is taxonomically restricted, limiting generalization. In this study, we investigated this ecdysis-based manipulation mechanism in a Tetragnathidae spider host. Examining the interaction between the spider Leucauge volupis (Keyserling, 1893) and the polysphinctine wasp Hymenoepimecis cameroni (Townes, 1966), we compared the architecture of normal, molting, and cocoon webs of the spider, and quantified 20E levels across the spider’s group. Cocoon webs shared traits with molting webs, such as the absence of spirals and increased reinforced radii. However, contrary to expectations, 20E levels were not elevated in behaviorally altered spiders compared to the other groups. Therefore, although cocoon webs resemble molting webs, ecdysone does not appear to be the proximate trigger of host manipulation in this system, suggesting that their structural similarity arises from alternative physiological or behavioral pathways yet to be identified. Keywords: Polysphincta, cocoon web, ecdysis, extended phenotype, Darwin wasps Corresponding author: Ítalo M. Delazari, Laboratory of Ecology and Behavior, Department of General Biology, Universidade Federal de Viçosa. E-mail: [e-mail] Paper Citation: Delazari ÍM, Gaione-Costa A, Santos ARS, Arrieta JDG, Sobczak J, Korenko S, & Kloss TG. 2026. What controls the puppets? 20-hydroxyecdysone may not drive manipulation in Leucauge spiders. Behavioral Ecology, xx:xx, xxx-xxx. doi xxxxx About this repository: # DATA FILES AND VARIABLES This repository contains the data files: * Data_Leucauge_volupis_Webs.csv * Data_20E_Leucauge_volupis.csv ## 1. Data_Leucauge_volupis_Webs.csv These file contains data related to different structures of web types built by Leucauge volupis spiders. Variables: * id: Identifier of each observation (replicate; one web per entry) (categorical) * web_type: Type of web (levels: normal, molting, cocoon) (categorical) * spider_age: Developmental stage (levels: adult, juvenile/subadult) (categorical) * spider_sex: Sex of the individual (levels: female, male) (categorical) * spirals: Number of spiral loops, counted along standardized radii (integer) * radii: Number of structural radii converging to the hub (integer) * reinforced_radii: Number of radii reinforced with additional silk (integer) * tangle_thread: Presence and configuration of barrier threads (levels: absent, upper, lower, both) (categorical) * opening_hub: Hub condition (center of the web) (levels: open, closed) (categorical, binary) * hub_spirals: Number of spirals located near the hub region (integer) ## 2. Data_20E_Leucauge_volupis.csv These file contains data on 20-hidroxyecdysone concentrations in analised spiders. Variables: * Trat: Treatment group (character/factor): -Par_spider_mod_behav: Parasitized with normal behavior -Par_spider_norm_behav: Parasitized with modified behavior -Spider_after_moult: Spiders after molting -Non_par_spider: Non-parasitized spider with normal behavior * Ion: The ion of 20E analised (levels: 371 and 445) (character/factor) * Rep: Individual identifier (character/factor) * Conc: Hormone concentration in pg/mg (numerical, counts) # SOFTWARE VERSIONS All statistical analyses were conducted in R software, version 4.5.0. loaded packages: * package: DHARMa (v. 0.4.7)* package: emmeans (v. 1.11.1) # NOTES All variables are described using consistent terminology to facilitate reuse. Missing values (if present) are indicated directly in the dataset files. For methodological details, refer to the main manuscript.