Glyphosate Weakens Spider Hunting Skills, but Antibiotics Reveal a Hidden Role for Microbes

A new laboratory study reports that a widely used glyphosate-based herbicide significantly reduces the hunting efficiency of wolf spiders at field-relevant spray rates, and that this sublethal damage depends strikingly on the internal physiological state of the animals. When adult female Pardosa spiders were pre-treated with the antibiotic tetracycline, they captured more prey after herbicide exposure than spiders that received the herbicide alone. The finding, published open access in the journal Ecotoxicology by Jana Niedobová and Radek Michalko, suggests that antibiotic-sensitive processes, most plausibly the host-associated microbial communities that live inside spiders, can modulate how a non-target predator responds to one of the world’s most heavily applied pesticides.

The significance of the result lies in what it implies for biological pest control. Spiders are abundant generalist predators in agricultural landscapes, and their appetite for pest insects makes them important contributors to sustainable crop production. Any pesticide that quietly degrades their predatory performance without killing them outright could erode a natural ecosystem service long before population declines become visible. The new study adds to a growing body of evidence that survival statistics alone underestimate the ecological footprint of herbicides, and it introduces a method designed to capture those subtler effects with greater precision.

Glyphosate is the most widely used herbicide on the planet, formulated in more than 750 commercial products and applied at roughly 700,000 tonnes per year across an estimated 36 million square kilometres of land. Its known biochemical target, the shikimate acid pathway, exists in plants and some microorganisms but not in animals, which is why glyphosate was long regarded as largely harmless to wildlife. However, environmental exposure typically involves commercial glyphosate-based herbicides rather than the pure active ingredient. These formulations contain surfactants and other co-formulants that may be biologically active in their own right or interact additively or synergistically with glyphosate, complicating any attempt to attribute observed effects solely to the active molecule.

Spiders have emerged as a sensitive group in this context. Previous studies on wolf spiders and other species have reported changes in locomotion, reproduction and predatory efficiency following exposure to glyphosate-based herbicides, even though the classic molecular target of glyphosate is absent from animal tissues. Ground-dwelling wolf spiders are especially likely to encounter herbicide residues, since they forage on the soil surface where fresh spray and dried deposits accumulate. What has remained unclear is the mechanism behind these behavioural changes, and why different studies, species and formulations have produced such inconsistent results.

The new experiment was designed to probe one possible source of that variability: the physiological background of the animal itself. The researchers collected 240 adult female spiders of the Pardosa lugubris group, two morphologically near-identical co-occurring species, from a chemically untreated forest near Drnovice in the Czech Republic on 9 May 2024, a timing chosen to coincide with the main glyphosate application window in Central European orchards. The spiders were randomly assigned to four groups of sixty: an untreated control, a herbicide-only group, a tetracycline-only group, and a group pre-treated with tetracycline before herbicide exposure. Tetracycline is a standard experimental tool for perturbing arthropod microbial communities, including the intracellular bacterial endosymbionts such as Wolbachia, Cardinium, Rickettsia and Spiroplasma that spiders and other arthropods carry, although the authors note that they did not directly characterise the microbiota in this study.

The herbicide tested was Roundup Klasik Pro, diluted at the manufacturer’s recommended field rate, and applied with a Potter spray tower that delivered a field-equivalent deposit of approximately 2 milligrams of spray solution per square centimetre, corresponding to a nominal 2.16 kilograms of glyphosate active ingredient per hectare. This setup simulated direct, label-rate spray exposure of the kind that might strike foraging spiders during orchard or vineyard treatments. Mortality was low and did not differ significantly among the treatments, confirming that the hazard under investigation was sublethal rather than lethal.

Predatory performance was measured with functional response analysis, which tracks how many prey an individual predator kills across a range of prey densities. Rather than offering a single fixed number of flies, as most earlier studies did, the researchers gave each spider one of five densities of Drosophila fruit flies: 1, 3, 6, 9 or 12 individuals. Prey killed were counted hourly during the first five hours after exposure, capturing acute effects, and then daily from the second to the fifth day, capturing persistent changes. Killed flies were immediately replaced with live ones to keep prey availability constant, and spiders that had moulted around the experiment were excluded, since spiders do not feed before moulting. The final analyses included 48 to 58 spiders per group.

The results were consistent across both time windows. In the short term, prey capture increased with prey density in all groups, as expected, but treatment had a significant additive effect. Control spiders achieved the highest capture rate, followed closely by the tetracycline-only group, while spiders exposed to the herbicide alone showed the lowest rate of any treatment. Spiders that received both tetracycline and the herbicide captured significantly less prey than untreated controls but significantly more than the herbicide-only group. Over the medium term the pattern sharpened: the herbicide-only spiders again had the lowest capture rate, while the antibiotic, combined and control groups did not differ significantly from one another. In statistical terms, the spiders displayed a type II functional response, the decelerating, saturating curve typical of spiders, and the herbicide depressed that curve.

Why would an antibiotic blunt the harm of a herbicide? The authors offer several candidate explanations while stressing that identifying a mechanism was not the goal of the study. One plausible route is indirect toxicity through microbial disruption: growing evidence, particularly from pollinators, shows that glyphosate can perturb gut microbiota with consequences for immunity and behaviour, so spiders whose internal microbes had already been altered by tetracycline may simply have had less to lose from that pathway. A second possibility involves microbiota-gut-brain signalling, which could mediate behavioural impairments. A third is that tetracycline itself has direct antioxidant properties at low concentrations, which might mitigate oxidative stress caused by herbicide exposure. Tetracycline alone showed only a non-significant trend toward reduced capture, and the authors emphasise that their microbiome interpretation remains a hypothesis requiring direct molecular validation.

The broader message is that pesticide risk assessment may need to look inside the organism as well as at it. Reviews and meta-analyses of arthropod natural enemies have already shown that herbicides can reduce predation even when acute mortality is negligible, and comparable feeding impairments have now been documented in fish larvae, honey bees and Drosophila. The new study extends this pattern up the food web, showing that sublethal herbicide effects can disrupt predator-prey interactions themselves, not merely herbivore and detritivore performance. If the internal microbial state of a predator determines how badly a herbicide impairs its hunting, then two populations of the same species exposed to the same spray could suffer very different consequences depending on their prior history of stress, diet or symbiont infections. Niedobová and Michalko argue that such context-dependent hazards, and the microbiome-mediated effects that may underlie them, should be explicitly considered in future pesticide regulation, because ignoring them risks underestimating the true cost of herbicide use to the natural enemies that agriculture quietly depends upon. Further work characterising spider microbial communities directly will be needed to confirm which of the candidate mechanisms explains the attenuation observed here.

Subject of Research: Context-dependent sublethal effects of a glyphosate-based herbicide on the predatory functional response of wolf spiders and the modulating role of antibiotic-sensitive physiological processes

Article Title: Context-dependent sublethal hazard of glyphosate to non-target predators revealed by functional response analysis

Article References: Niedobová, J., & Michalko, R. (2026). Context-dependent sublethal hazard of glyphosate to non-target predators revealed by functional response analysis. Ecotoxicology, 35(9), Article 198. https://doi.org/10.1007/s10646-026-03179-5

Image Credits: AI Generated

DOI: 10.1007/s10646-026-03179-5

Keywords: glyphosate, glyphosate-based herbicides, wolf spiders, Pardosa, functional response, sublethal effects, tetracycline, microbiome, endosymbionts, biological pest control, ecotoxicology, non-target organisms