Cannabis flower crude extracts as belowground infochemicals shaping the chemotactic responses of parasitic nematodes

Abstract

The widespread use of synthetic pesticides has raised concerns about their impacts on soil biodiversity and ecosystem functioning, increasing interest in plant-derived compounds as environmentally compatible alternatives. Cannabis sativa L. produces a diverse array of secondary metabolites, including cannabinoids and terpenes, which may act as chemical cues in belowground ecological interactions. This study examined the chemotactic responses of five nematode species – Heterorhabditis bacteriophora (Poinar), Steinernema carpocapsae (Weiser), Steinernema feltiae (Filipjev), Oscheius myriophilus (Poinar), and Phasmarhabditis papillosa (Schneider)— to various concentrations of crude flower extracts of C. sativa. Chemotaxis assays were conducted using an agar-based system, and nematode responses were quantified as overall motility and Chemotaxis Index (CI). Cannabinoid and terpene profiles were determined by HPLC–UV/DAD and GC–MS. Nematode species identity and extract concentration were the primary factors influencing motility and chemotactic behaviour, whereas chemotype extract identity had a minor effect. H. bacteriophora, S. feltiae, and P. papillosa showed positive chemotactic responses (higher CI values) in a species- and concentration-dependent manner, whereas S. carpocapsae and O. myriophilus exhibited mostly neutral responses. Chemotactic behaviour was strongly species-specific and concentration-dependent, with extract identity playing a minor role.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-49877-2.

Introduction

The widespread use of synthetic pesticides in modern agriculture has generated substantial concern due to their detrimental impacts on soil health, biodiversity, and ecosystem functioning. Persistent agrochemicals can disrupt soil food webs, alter microbial communities, and impair the ecological services provided by soil biota, prompting growing interest in sustainable pest management strategies that operate in harmony with belowground ecosystems. Botanical biopesticides are increasingly recognized as promising alternatives because of their rapid degradation in soil, reduced accumulation in food webs, and comparatively low toxicity to non-target organisms^1–3. Nevertheless, despite their ecological advantages, biological pesticides still represent only a minute fraction (0.01%) of the approximately 4.1 million tonnes of pesticides applied globally in 2025⁴.

Cannabis sativa L. is a multipurpose crop cultivated for fibre, food, medicinal, and industrial uses and has recently gained attention as a potential source of bioactive compounds relevant to soil-mediated pest regulation^5,6. Beyond its agronomic value, C. sativa produces a diverse array of secondary metabolites that may influence belowground ecological interactions. Extracts from roots, leaves, and inflorescences have been shown to possess insecticidal³, acaricidal⁷, and nematicidal activity⁸, effects attributed to compounds such as alkaloids, flavonoids, lignans, terpenes, and cannabinoids^9,10. Many of these metabolites are volatile or semi-volatile and may act as chemical signals in soil environments, influencing organismal behavior and trophic interactions.

The biochemical diversity of C. sativa is commonly described through its classification into five chemotypes based on the relative abundance of neutral cannabinoids. Chemotype I is dominated by tetrahydrocannabinol (THC; >0.3%), with low cannabidiol (CBD; <0.5%) content, resulting in a THC/CBD ratio greater than one. Chemotype II contains both THC (> 0.3%) and CBD (> 0.5%) at similar concentrations, whereas Chemotype III is characterized by high CBD (> 0.5%) and low THC (< 0.3%) levels^11–14. Two additional chemotypes have been identified: Chemotype IV, in which cannabigerol (CBG) exceeds 0.3%¹⁵, and Chemotype V, characterized by negligible total cannabinoid concentrations^16,17. These chemotypic distinctions reflect profound differences in metabolic profiles that may translate into contrasting ecological functions, particularly in soil-mediated signalling processes.

Among the more than 100 cannabinoids identified, THC, CBD, and CBG are the most abundant and biologically active¹⁸. While cannabinoids are best known for their pharmacological properties, increasing evidence suggests they may also function as semiochemicals, influencing interactions between plants and soil-dwelling organisms. In addition to cannabinoids, cannabis plants emit a complex blend of terpenes that contribute to their volatile organic compound (VOC) profile. Such VOCs are increasingly recognized as key regulators of belowground multitrophic interactions, mediating communication between plants, microbes, herbivores, and their natural enemies^19,20.

Entomopathogenic nematodes (EPNs) represent an important functional group within soil food webs and are widely used as biological control agents. Species belonging to the genera Heterorhabditis and Steinernema infect and kill insect hosts through symbiotic bacteria. In contrast, Phasmarhabditis papillosa is primarily associated with molluscs, whereas Oscheius myriophilus exhibits a broader ecological association, having been reported in both insect-related environments and in association with slugs^21–23. The efficacy of these nematodes in soil environments is strongly dependent on their ability to locate hosts through chemotaxis, guided by gradients of carbon dioxide, root exudates, and plant-derived VOCs diffusing through the soil matrix^24,25.

A growing body of research demonstrates that plant-emitted VOCs play a central role in shaping nematode behaviour and, consequently, pest suppression in the rhizosphere. Volatile cues from crops such as maize, black mustard, and carrot have been shown to attract or repel EPNs, thereby influencing their spatial distribution and biological control efficacy^26–28. However, despite the increasing cultivation of C. sativa and its rich chemical repertoire, the influence of cannabis-derived metabolites on soil-dwelling nematodes remains poorly understood.

In natural systems, plant-derived metabolites can enter the soil environment through multiple pathways, including litter deposition, root exudation, and the leaching of aboveground tissues during rainfall events^19,20. Although floral tissues themselves are less likely to accumulate directly in soil, their chemical constituents may contribute to the pool of bioactive compounds influencing belowground interactions. Once released, these compounds may act as infochemicals guiding the behaviour of soil-dwelling organisms²⁰. However, their persistence is often limited due to microbial degradation, adsorption to soil particles, and volatilisation, resulting in transient but ecologically relevant signalling effects.

Recent studies provide initial evidence that cannabis metabolites may act as behaviour-modifying signals for nematodes. Laznik et al.²⁹ reported that ethanol extracts from different plant organs of different C. sativa genotypes elicited species-specific chemotactic responses in entomopathogenic nematodes. Building on this, Flajšman et al.³⁰ demonstrated that individual cannabinoids can differentially modulate nematode behaviour, with THC and CBG acting as strong attractants for Phasmarhabditis papillosa, Steinernema feltiae, and Heterorhabditis bacteriophora, whereas CBD often induced neutral or repellent responses. These findings suggest that cannabinoids may function as infochemicals in soil, with the potential to influence nematode movement and persistence in the rhizosphere.

Within this context, the present study examines the chemotactic responses of five nematode species—Heterorhabditis bacteriophora, Steinernema carpocapsae, Steinernema feltiae, Oscheius myriophilus, and Phasmarhabditis papillosa— to crude flower extracts of different chemotypes of C. sativa. Specifically, we aim to (i) determine how crude flower extracts of four cannabis chemotypes differ in their effects on nematode behaviour, (ii) identify species-specific response patterns relevant to soil food web dynamics, and (iii) evaluate the implications of cannabis-derived metabolites for enhancing nematode-mediated biological control in soil-based agroecosystems.

Materials and methods

Plant material and growth conditions

Four female C. sativa genotypes were used in this study. Three genotypes (S63, S9, and S19), representing chemotypes I, II, and III, respectively, originated from the traditional Slovenian landrace ‘Gorička Simba’. A fourth genotype, selected through an in-house breeding program, was cannabigerol-dominant (called CBG) and represents chemotype IV.

Mother plants were maintained in the glasshouse facilities of the Department of Agronomy, University of Ljubljana (Ljubljana, Slovenia), under combined natural daylight and supplemental high-pressure sodium (HPS) lighting. A photoperiod of 18 h light and 6 h dark (18 L:6D) was applied to sustain vegetative growth. Day and night temperatures were maintained at 26 °C and 18 °C, respectively, with relative humidity ranging between 50% and 70%. Plants were fertigated with a nutrient solution formulated for vegetative growth at an electrical conductivity (EC) of 1.2 mS cm⁻¹, using the fertilizer formulations described below.

All experimental research involving plant material was conducted in accordance with relevant institutional, national, and international guidelines and legislation. The plant material used in this study originated from legally cultivated sources and did not involve wild populations or species at risk of extinction. Therefore, the study does not fall within the scope of the IUCN Policy Statement on Research Involving Species at Risk of Extinction or the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES).

Vegetative propagation by stem cuttings

Vegetative propagation was carried out on 11 May 2021 using lateral shoots collected from healthy mother plants. Stem cuttings measuring 13–16 cm were excised at a 45° angle with sterilized scissors. The basal ends were treated with a commercial rooting gel (Clonex^®, Growth Technology, UK) and inserted into rockwool propagation cubes (Cultilene, 1½″ Large).

To reduce transpirational water loss, all leaves were removed except for two to three apical leaves. Cuttings were placed under a propagation dome maintained at 25 °C, with relative humidity initially set between 95% and 100%. Light intensity was approximately 120 µmol m⁻² s⁻¹ (photosynthetic photon flux density, PPFD), with an 18 L:6D photoperiod. Planting density was 462 cuttings m⁻². Relative humidity was gradually reduced to 85–90% after one week and further lowered to 75–80% by the third week, while temperature, light regime, and photoperiod remained unchanged.

Transplanting and experimental design

Rooted cuttings were transplanted on 1 June 2021 into 0.5 L pots containing a peat-based growing substrate (Kekkilä Brown 540, Kekkilä Oy Professional, Finland) and cultivated for four weeks under the same glasshouse conditions as the mother plants. On 2 July 2021, plants were transferred to an open-field experimental site covered with black polyethylene mulch film. Plant spacing was 2 m between rows and 1 m within rows.

Fertigation was applied throughout the growing season according to the nutrient regime described below, with formulations adjusted to correspond to vegetative or flowering developmental stages.

Nutrient management

During the vegetative growth phase, plants were fertigated with a 1:1 (v/v) mixture of a calcium–magnesium fertilizer (CAL-MAG; 3.5–0–0.5Ca–1.5Mg) and a vegetative growth fertilizer (VEG; 9–3–6–1.5Mg + 6). Nutrients were dissolved in water and the solution pH was adjusted to 6.5 using nitric acid.

The CAL-MAG fertilizer supplied nitrate nitrogen (3.5%), calcium (5%), and magnesium (1.5%), while the VEG formulation provided total nitrogen (9%; 5.4% NO₃⁻-N and 3.6% NH₄⁺-N), phosphorus (3% P₂O₅), potassium (6% K₂O), magnesium (0.5%), and micronutrients (Fe, Mn, Zn, Cu, Mo). EC was gradually increased from 0.6 to 1.0 mS cm⁻¹ during early growth to 1.2 mS cm⁻¹ during later vegetative stages (weeks 1–7). Irrigation frequency increased from once to three times per week in response to plant biomass accumulation.

During flowering, a 1:1 (v/v) mixture of CAL-MAG and a bloom fertilizer (BLOOM; 3–7–7–0.5Mg + 6) was applied. The BLOOM formulation supplied 3% total nitrogen (2.2% NO₃⁻-N and 0.8% NH₄⁺-N), 7% P₂O₅, 7% K₂O, 0.5% Mg²⁺, and the same micronutrient profile as the VEG fertilizer. Solution pH was maintained at 6.5 using phosphoric acid. EC ranged from 1.8 mS cm⁻¹ in early flowering to 2.5 mS cm⁻¹ at peak bloom and was reduced toward harvest. Irrigation frequency during flowering was maintained at three times per week.

Inflorescence sampling

Inflorescences were harvested manually around 30 September 2021, when plants had reached full physiological maturity. Floral material was trimmed to remove excess stems and leaves and dried in a forced-air dryer at 40 °C until constant mass was achieved. Dried samples were stored in dark, dry conditions until further processing.

Cryo-ethanol extraction of inflorescences

Prior to extraction, inflorescences of each chemotype were stored at − 30 °C for 24 h. Cryo-ground flower material (100 g) was extracted with 1 L of cold ethanol (Merck KGaA, Darmstadt, Germany) and gently agitated by hand for 3 min inside a − 30 °C freezer. Plant material was subsequently removed, and the extract was vacuum-filtered to eliminate residual biomass. Ethanol was removed using a rotary evaporator operated at 45 rpm and 50 °C, yielding crude extracts for cannabinoid and terpene analysis.

HPLC–UV/DAD identification and quantification of cannabinoids

Inflorescence samples were homogenized to a fine, uniform powder prior to analysis. Cannabinoids were extracted from 500 mg of dried flower material or 50 mg of crude cannabis extract using a chloroform: methanol solvent system (9:1, v/v). Extractions were performed in an ultrasonic water bath for 15 min to enhance compound recovery. Resulting extracts were clarified by filtration through 0.45 μm membrane filters and subsequently diluted with the mobile phase consisting of acetonitrile and 0.1% (v/v) acetic acid at a ratio of 65:35.

Chromatographic separation and quantification were carried out using an Agilent 1100 series high-performance liquid chromatography (HPLC) system (Agilent Technologies Inc., Santa Clara, CA, USA) equipped with a Supelco Ascentis^® Express C18 column (5 μm; Sigma-Aldrich, St. Louis, MO, USA). Analyses were performed at the Faculty of Health Sciences, University of Ljubljana (Ljubljana, Slovenia). The injection volume was 20 µL, the mobile-phase flow rate was set to 1.0 mL min⁻¹, and the column temperature was maintained at 25 °C. Cannabinoids were detected using a UV/DAD detector at a wavelength of 220 nm.

Quantification was based on external calibration curves generated from certified reference standards (Restek Corp., Pennsylvania, PA, USA). Twelve cannabinoids were quantified: cannabidiolic acid (CBDA), cannabidiol (CBD), cannabigerolic acid (CBGA), cannabigerol (CBG), cannabidivarinic acid (CBDVA), cannabidivarin (CBDV), tetrahydrocannabinolic acid (THCA), Δ⁹-tetrahydrocannabinol (Δ⁹-THC), Δ⁸-tetrahydrocannabinol (Δ⁸-THC), tetrahydrocannabivarin (THCV), cannabinol (CBN), and cannabichromene (CBC). Total cannabinoid contents (tCBD, tTHC, and tCBG) were calculated by summing the neutral and acidic forms, applying conversion factors of 0.877 for CBDA and THCA and 0.878 for CBGA.

GC-MS identification and quantification of terpenes

Terpenes were extracted from 100 mg of sample in 9.9 mL n-hexane using ultrasonic extraction for 15 min. Extracts were centrifuged and filtered prior to analysis. GC–MS analysis was conducted using an Agilent 6890 GC coupled to a 5973 mass selective detector with an Rxi-35Sil MS column. Helium was used as carrier gas at 1.2 mL min⁻¹. The oven program ranged from 50 °C to 250 °C at 9 °C min⁻¹. Terpenes were identified using mass spectral libraries and quantified using external standards.

Source and maintenance of parasitic nematodes

Three commercially available species of entomopathogenic nematodes (EPNs), Heterorhabditis bacteriophora, Steinernema carpocapsae, and Steinernema feltiae, were obtained from Koppert B.V. (Berkel en Rodenrijs, The Netherlands) and used in chemotaxis assays. The nematodes were propagated in vivo using final-instar larvae of the greater wax moth, Galleria mellonella L. (Lepidoptera: Pyralidae), according to previously established methodologies²². Infective juveniles (IJs) were collected and stored in M9 buffer at 4 °C at a concentration of 4000 IJs mL⁻¹. Only nematodes younger than two weeks were used in experiments. Population densities were determined following the counting method described by Laznik et al.³¹. Prior to each bioassay, nematode viability was assessed microscopically, and only cultures exhibiting more than 95% viable individuals were used.

In addition to the EPN species, two mollusk-parasitic nematodes, Phasmarhabditis papillosa (GenBank accession no. MT800511.1) and Oscheius myriophilus (GenBank accession no. OP684306.1), were included in the study. These isolates were recently recovered from the invasive slug Arion vulgaris Moquin-Tandon (Stylommatophora: Arionidae) using the protocol described by Laznik et al.³¹. The nematodes were maintained in vivo on freeze-killed slugs. Infective juveniles were extracted ten days post-inoculation using the same isolation procedure applied to the EPN species. Suspensions were stored in M9 buffer at 4 °C, and only cultures less than 14 days old with viability exceeding 95% were selected for experimental use.

Compound preparation and chemotaxis assay

For the chemotaxis assay, cannabis flower crude extracts from four chemotypes were used and prepared as follows: 1 g of each crude extract was dissolved in 1 mL of 96% ethanol to obtain stock solutions defined as 100% concentration. These stocks were then diluted with ethanol to produce working solutions at 50%, 25%, and 12% concentrations. All solutions were vortex-mixed before use to ensure homogeneity. This standardised preparation protocol ensured consistent chemotype extract dosing across experimental replicates and supported the accuracy and reproducibility of the chemotaxis assays³⁰.

Chemotactic responses of nematodes were assessed using an agar-based bioassay adapted from O’Halloran and Burnell³², with subsequent modifications by Laznik et al.^29,31. Assays were conducted in 9 cm diameter polystyrene Petri dishes containing 25 mL of 1.6% technical agar (Biolife, Milan, Italy). The agar was prepared in a buffered solution consisting of 5 mM potassium phosphate (pH 6.0), 1 mM CaCl₂, and 1 mM MgSO₄ to facilitate nematode movement.

On the underside of each Petri dish, three circular reference zones (1 cm diameter) were marked: a central area at the midpoint of the dish and two outer areas positioned symmetrically 1.5 cm from the dish edge (Fig. 1). One outer zone was designated as the treated area, while the opposite zone served as the control area. A 10 µL droplet of the test compound (cannabis flower crude extract solution) at the specified concentration was applied to the treated area, while 10 µL of 96% ethanol was applied to the control area. Immediately afterwards, 100 infective juveniles (IJs) suspended in 50 µL of M9 buffer were placed in the central area. In negative control assays, both the treated and control areas received 10 µL of 96% ethanol.

Fig. 1.

Chemotaxis assay layout. Schematic representation of the chemotaxis bioassay. Each 9 cm Petri dish contained three circular zones (1 cm diameter): a central zone and two outer zones positioned symmetrically at a distance of 1.5 cm from the centre. The treated zone received a 10 µL droplet of cannabis extract solution, while the opposite zone received 10 µL of 96% ethanol (control). One hundred infective juveniles (IJs) suspended in M9 buffer were placed in the central zone. The diagram indicates the position and size of each zone to ensure clarity and reproducibility.

Each treatment included ten technical replicates, and the entire experiment was independently repeated three times. To prevent cross-contamination by volatile compounds, only cannabis flower crude extract solution of one chemotype was tested per experimental run. All dishes were sealed with Parafilm™ (Bemis Company, Inc., Neenah, WI, USA) to limit vapour diffusion beyond individual plates.

Plates were incubated in a controlled-environment chamber (RK-900 CH, Kambič Laboratory Equipment, Semič, Slovenia) at 22 °C and 75% relative humidity in complete darkness. After 24 h, nematode movement was terminated by freezing the plates at − 20 °C for 3 min, preserving their final spatial distribution for analysis.

Nematode positions were recorded using a Nikon SMZ800N stereomicroscope (Nikon Corporation, Tokyo, Japan) fitted with a 4 K UHD HDMI camera (XCAM4K8MPB; Tucsen Photonics Co., Ltd., Fuzhou, China) at 25× magnification. Behavioural responses were expressed as a Chemotaxis Index (CI), calculated according to a modified equation derived from Bargmann and Horvitz³³ and Laznik et al.³¹:

CI = (% IJs in treated zone − % IJs in control zone)/100%.

The CI provided a quantitative measure of attraction or avoidance, with values ranging from + 1.0 (complete attraction) to − 1.0 (complete repulsion). Compounds were categorized based on CI values as attractants (CI ≥ 0.2), weak attractants (0.2 > CI ≥ 0.1), neutral (− 0.1 ≤ CI < 0.1), weak repellents (− 0.2 ≤ CI < − 0.1), or repellents (CI ≤ − 0.2), following established criteria^29,31.

Statistical analysis

Nematode directional responses in the chemotaxis assay, defined as movement from the central zone toward either the treatment or control area, were analysed using paired Student’s t-tests. This approach allowed comparison of nematode distribution between treated and control zones within the same Petri dish, thereby accounting for within-plate variability. Statistical significance was accepted at p < 0.05, indicating a non-random preference or avoidance response. Assumptions of normality were verified using the Shapiro–Wilk test prior to analysis.

To assess differences in behavioural patterns among nematode species, the proportion of infective juveniles (IJs) migrating to peripheral zones or remaining in the central zone was calculated for each replicate. These data were analysed using a three-factor analysis of variance (ANOVA), with nematode species, chemotype extract type, and extract concentration treated as fixed effects. All main effects and interaction terms were included in the model and interpreted accordingly.

Chemotaxis Index (CI) values were analysed using factorial ANOVA to evaluate species-specific responses across chemotype extracts and concentrations. When significant main effects or interactions were detected, differences among means were further explored using Duncan’s multiple range test at a significance level of p < 0.05. Given the relatively liberal nature of Duncan’s test, these pairwise comparisons were interpreted cautiously and used primarily to identify general patterns rather than to draw definitive conclusions about individual treatment differences. Results are presented as mean values ± standard error (SE).

All statistical analyses were performed using Statgraphics Plus for Windows, version 4.0 (Statistical Graphics Corp., Manugistics, Inc., Rockville, MD, USA). Figures were generated using Microsoft Excel 2010.

Results

Cannabinoid profiles of inflorescences

Clear differences in cannabinoid composition were observed among the four Cannabis sativa chemotypes (Table 1). The S63 genotype (chemotype I) was characterized by a high total tetrahydrocannabinol (tTHC) concentration (64.0 mg g⁻¹) and negligible total cannabidiol (tCBD; 0.3 mg g⁻¹). In contrast, S9 (chemotype II) exhibited intermediate concentrations of both cannabinoids, with a tCBD: tTHC ratio of approximately 1.6. The S19 genotype (chemotype III) showed a pronounced dominance of tCBD (53.1 mg g⁻¹), while tTHC levels remained low (1.7 mg g⁻¹). The genotype CBG (chemotype IV) contained the highest total cannabigerol (tCBG) concentration (47.6 mg g⁻¹), whereas THC was not detected. A complete overview of all quantified cannabinoids in inflorescences is provided in Supplementary Table S1.

Table 1.

Concentrations of major cannabinoids in Cannabis sativa inflorescences (mg g⁻¹ dry weight).

Genotype	tCBD (mg g⁻¹)	tTHC (mg g⁻¹)	tCBG (mg g⁻¹)
S19 (Chemotype III)	53.1 ± 3.5	1.7 ± 0.0	1.0 ± 0.0
S63 (Chemotype I)	0.3 ± 0.0	64.0 ± 2.1	2.4 ± 0.3
S9 (Chemotype II)	24.7 ± 1.1	15.0 ± 0.7	1.8 ± 0.6
CBG (Chemotype IV)	3.9 ± 0.0	nd	47.6 ± 1.5

Values are means ± standard deviation (n = 2). nd = not detected.

Cannabinoid composition of crude extracts

Cannabinoid profiles of the crude extracts largely reflected the chemotype-specific patterns observed in the inflorescences, but at substantially higher concentrations (Table 2). Extracts from S19 inflorescences contained the highest tCBD concentration (462.1 mg g⁻¹), while S63 extracts were dominated by tTHC (419.9 mg g⁻¹). The S9 extract retained a balanced cannabinoid profile, with a tCBD: tTHC ratio of 1.8. Extracts from the CBG-dominant genotype showed the highest tCBG concentration (461.2 mg g⁻¹), confirming the chemotypic distinction of this genotype. Detailed concentrations of all quantified cannabinoids in the crude extracts are presented in Supplementary Table S2.

Table 2.

Concentrations of major cannabinoids in crude extracts of Cannabis sativa inflorescences (mg g⁻¹ extract).

Genotype	tCBD (mg g⁻¹)	tTHC (mg g⁻¹)	tCBG (mg g⁻¹)
S19 (Chemotype III)	462.1 ± 2.3	18.2 ± 0.3	36.4 ± 0.3
S63 (Chemotype I)	14.8 ± 1.6	419.9 ± 1.5	14.8 ± 0.3
S9 (Chemotype II)	206.0 ± 4.1	115.5 ± 1.4	35.9 ± 2.7
CBG (Chemotype IV)	29.3 ± 0.3	45.1 ± 0.2	461.2 ± 6.5

Values represent means ± standard deviation (n = 2). Concentrations are expressed per gram of extract.

Terpene profiles of cannabis crude extracts

Analysis of terpene composition revealed that caryophyllene oxide, β-caryophyllene, α-humulene, and β-myrcene were the most abundant terpenes across all cannabis crude extracts (Table 3). Among the tested genotypes, S63 (chemotype I) generally exhibited the highest concentrations of these dominant terpenes. The remaining genotypes (S19 and S9) contained detectable but lower levels of several minor terpenes, with some compounds present only marginally above the detection limit. Notably, guaiol was detected exclusively in extracts of the CBG-dominant genotype, distinguishing it from the remaining chemotypes. Caryophyllene oxide represented the most abundant terpene overall, with particularly high concentrations measured in the S63 extract.

Table 3.

Terpene composition of crude extracts of Cannabis sativa inflorescences (mg g⁻¹ extract).

Terpene	S19 (Chemotype III)	CBG (Chemotype IV)	S63 (Chemotype I)	S9 (Chemotype II)
α-Pinene	nd	nd	nd	0.52 ± 0.05
β-Pinene	0.34 ± 0.00	nd	nd	nd
β-Myrcene	0.65 ± 0.02	0.07 ± 0.00	1.11 ± 0.02	0.72 ± 0.12
α-Terpinene	0.45 ± 0.05	nd	nd	0.44 ± 0.08
d-Limonene	0.11 ± 0.04	nd	nd	0.19 ± 0.03
p-Cymene	0.99 ± 0.21	nd	nd	1.14 ± 1.42
γ-Terpinene	0.39 ± 0.05	nd	nd	0.37 ± 0.05
Terpinolene	1.87 ± 0.32	nd	nd	3.48 ± 2.33
β-Caryophyllene	8.72 ± 0.35	9.92 ± 1.01	8.84 ± 1.70	2.08 ± 0.16
α-Humulene	3.97 ± 0.56	3.26 ± 0.56	5.75 ± 0.16	0.45 ± 0.02
Guaiol	nd	5.06 ± 0.27	nd	nd
Caryophyllene oxide	24.89 ± 4.41	25.41 ± 3.29	67.28 ± 16.21	4.89 ± 0.04

Values are means ± standard deviation (n = 2). nd = not detected. The relatively high standard deviations observed for certain compounds (e.g., caryophyllene oxide for CBG and S63) reflect variability between the two replicates and potential compound instability (e.g. oxidation processes).

Nematode motility

In this study, nematode motility was defined as the proportion of infective juveniles (IJs) migrating from the central area towards either the treated area or the control area of the chemotaxis assay plate (Fig. 1). This response variable was used to evaluate the effects of nematode species, chemotype extract identity, chemotype extract concentration, and their interactions (Table 4).

Table 4.

ANOVA results for the directional movement of infective juveniles (IJs) from the inner to the outer segments of the Petri dish.

Factor	Sum of squares	Df	F	P
Nematode species (S)	34183.80	4	108.51	0.0001
Chemotype extract (C)	852.44	3	3.61	0.0135
Chemotype extract concentration (Cc)	6437.66	4	20.44	0.0001
Replicate (technical)	566.90	9	0.80	0.6200
Temporal replication	187.40	2	1.19	0.3100
S x C	1551.70	12	1.64	0.0780
S x Cc	5929.60	16	4.71	0.0001
C x Cc	1442.20	12	1.53	0.1120
S x C x Cc	3691.60	48	0.98	0.5200
Residual	30565.60	389
Total (Corrected)	85527.70	499

Analysis of variance showed that nematode species identity was the main factor influencing IJ motility (F = 108.51, p < 0.0001), indicating marked differences in movement behaviour among the five tested species. Chemotype extract concentration also had a strong and highly significant effect on nematode motility (F = 20.44, p < 0.0001), demonstrating a clear dose-dependent response. Additionally, chemotype extract identity significantly affected IJ movement (F = 3.61, p = 0.0135), although its influence was weaker than that of species identity and concentration. Neither technical replication (F = 0.80, p = 0.62) nor temporal replication (F = 1.19, p = 0.31) significantly affected nematode motility, confirming high reproducibility and temporal stability of the experimental system.

Among the interaction terms, the species × concentration (S × Cc) interaction was highly significant (F = 4.71, p < 0.0001), indicating that nematode species differed in their movement from the central area towards the treated and control areas across chemotype extract concentrations. In contrast, the species × chemotype extract (S × C) interaction was not statistically significant (F = 1.64, p = 0.078), nor was the chemotype extract × concentration (C × Cc) interaction (F = 1.53, p = 0.112). The three-way interaction (S × C × Cc) was also non-significant (F = 0.98, p = 0.52), indicating that the combined effects of species identity, chemotype extract type, and concentration did not further explain variation in motility beyond lower-order interactions.

Overall, these results show that nematode motility from the central area towards the treated and control areas is mainly driven by species-specific traits and chemotype extract concentration, with chemical identity playing a secondary role. The absence of significant replication effects supports the robustness of the experimental design and the reliability of the observed behavioural pattern.

Figure 2 illustrates the mean proportion (± SE) of infective juveniles (IJs) that migrated from the central area to the outer areas (treated and control areas combined) of the Petri dishes for the (A) five nematode species and for the (B) chemotype extract identity. Heterorhabditis bacteriophora showed the highest level of movement toward the outer areas, with approximately 28% of IJs migrating from the central area, followed by Steinernema feltiae (≈ 25%). Phasmarhabditis papillosa exhibited an intermediate level of migration (≈ 13%). In contrast, Steinernema carpocapsae and Oscheius myriophilus displayed the lowest proportions of IJs in the outer areas, with mean values of approximately 9% and 8%, respectively. The control exhibited the lowest percentage (≈ 10%). In contrast, the chemotype extracts S19, S63, and S9 produced the highest responses (≈ 17–18%), representing a significant increase over both the control and the CBG extract, which showed an intermediate response (≈ 14–15%).

Fig. 2.

Mean percentage (± standard error) of infective juveniles (IJs) of five nematode species—Heterorhabditis bacteriophora (HB), Steinernema carpocapsae (SC), Steinernema feltiae (SF), Oscheius myriophilus (OM), and Phasmarhabditis papillosa (PP)—that migrated to the outer areas of the chemotaxis assay dish (A) and for chemotype extract identity (B). Different capital letters above the bars indicate statistically significant differences according to Duncan’s multiple range test (p < 0.05).

Nematode motility varied significantly among species and chemotype extract treatments, as shown in Fig. 3. Across all chemotype extract treatments, distinct species-specific movement patterns from the central area to the outer areas of the chemotaxis assay were observed. Heterorhabditis bacteriophora consistently showed the highest levels of migration towards the outer areas across most concentrations and treatments, indicating a strong tendency for movement under the tested conditions. Similarly, Steinernema feltiae displayed pronounced motility in response to all chemotype extracts, with relatively high proportions of infective juveniles reaching the outer areas across a wide range of concentrations.

Fig. 3.

Chemotactic responses of nematodes to cannabis crude extracts, expressed as the mean percentage (± SE) of infective juveniles (IJs) migrating from the central area to the outer areas in the chemotaxis assay. Responses to genotypes CBG (A), S9 (B), S63 (C), and S19 (D) were measured at four concentrations (100%, 50%, 25%, and 12%). A 96% ethanol treatment served as the control in all panels. Different capital letters above the bars indicate statistically significant differences among nematode species within the same treatment and concentration (Duncan’s multiple range test, p < 0.05). Species abbreviations: Heterorhabditis bacteriophora (HB), Steinernema carpocapsae (SC), Steinernema feltiae (SF), Oscheius myriophilus (OM), and Phasmarhabditis papillosa (PP).

In contrast, Steinernema carpocapsae exhibited substantially lower migration from the central area under all chemotype extract treatments, with only slight increases at intermediate concentrations. Oscheius myriophilus also showed consistently low motility, with minimal variation across chemotype extract treatments and concentrations. Phasmarhabditis papillosa displayed intermediate responses, with migration levels varying according to both the chemotype extract treatment and its concentration.

Chemotype extract concentration significantly affected nematode motility across all treatments, although the direction and magnitude of these effects varied among species. Higher and intermediate concentrations generally produced stronger responses in the more motile species, whereas lower concentrations and control treatments led to uniformly reduced movement across all taxa. These results show that nematode motility in response to chemotype extract is both species-specific and concentration-dependent.

Chemotaxis index

The directional movement of infective juveniles (IJs) in response to chemotype extract was quantified using the chemotaxis assay, and Chemotaxis Index (CI) values were analysed by analysis of variance (Table 5). The analysis showed that nematode species identity was the most influential factor affecting chemotactic responses (F = 44.45, p < 0.0001). Chemotype extract concentration also had a strong and highly significant effect on CI values (F = 11.54, p < 0.0001), indicating concentration-dependent modulation of nematode behaviour. In contrast, chemotype extract identity did not significantly affect CI values (F = 1.54, p = 0.2032).

Table 5.

ANOVA results for the Chemotaxis Index values.

Factor	Sum of squares	Df	F	P
Nematode species (S)	1.30	4	44.45	0.0001
Chemotype extract (C)	0.03	3	1.54	0.2032
Chemotype extract concentration (Cc)	0.34	4	11.54	0.0001
Replicate (technical)	0.03	9	0.41	0.9300
Temporal replication	0.02	2	1.22	0.3000
S x C	0.12	12	1.40	0.1650
S x Cc	0.60	16	5.10	0.0001
C x Cc	0.07	12	0.77	0.6850
S x C x Cc	0.49	48	1.39	0.0510
Residual	2.87	389
Total (Corrected)	5.85	499

Neither technical replication (F = 0.41, p = 0.93) nor temporal replication (F = 1.22, p = 0.30) had a significant effect on CI values, confirming the reproducibility and temporal stability of the experimental system. Among the interaction terms, the species × concentration (S × Cc) interaction was highly significant (F = 5.10, p < 0.0001), demonstrating that nematode species differed in their concentration-dependent chemotactic responses. In contrast, the species × chemotype extract (S × C) interaction (F = 1.40, p = 0.165) and the chemotype extract × concentration (C × Cc) interaction (F = 0.77, p = 0.685) were not significant. The three-way interaction species × chemotype extract × concentration (S × C × Cc) approached, but did not reach, statistical significance (F = 1.39, p = 0.051).

Overall, these results indicate that the chemotactic behaviour of nematodes towards chemotype extract is primarily driven by species-specific traits and chemotype extract concentration, while the identity of the chemotype extract itself plays a comparatively minor role. The relatively low residual variance suggests that the statistical model accounted for a substantial proportion of the biologically relevant variation in CI values.

The Chemotaxis Index (CI) values for five nematode species in response to crude extracts derived from Cannabis sativa genotypes CBG, S9, S63, and S19 are summarized in Table 6.

Table 6.

Chemotaxis Index (CI ± standard error) of five nematode species—Heterorhabditis bacteriophora (HB), Steinernema carpocapsae (SC), Steinernema feltiae (SF), Oscheius myriophilus (OM), and Phasmarhabditis papillosa (PP)—in response to varying concentrations (100%, 50%, 25%, and 12%) of crude extracts derived from Cannabis sativa genotypes CBG, S9, S63, and S19, including a 96% ethanol control.

HB	CBG 100%	CBG 50%	CBG 25%	CBG 12%	Control
	0.13 ± 0.03 Cb	0.23 ± 0.05 Dc	0.03 ± 0.03 Ba	0.15 ± 0.03 Cb	0.03 ± 0.03 Aa
OM	0.02 ± 0.02 Bb	0.00 ± 0.01 Bb	0.00 ± 0.00 Bb	−0.03 ± 0.02 Aa	0.04 ± 0.02 Ab
PP	0.06 ± 0.02 Bb	0.06 ± 0.02 Cb	0.10 ± 0.01 Cc	0.10 ± 0.02 Cbc	0.00 ± 0.02 Aa
SC	−0.07 ± 0.04 Aa	−0.03 ± 0.02 Aa	−0.04 ± 0.03 Aa	0.00 ± 0.00 Bb	0.01 ± 0.03 Aab
SF	0.17 ± 0.05 Cb	0.08 ± 0.05 Cab	0.15 ± 0.02 Db	0.17 ± 0.05 Cb	0.01 ± 0.03 Aa
	S9 100%	S9 50%	S9 25%	S9 12%	Control
HB	0.16 ± 0.04 Bb	0.13 ± 0.04 Cb	0.05 ± 0.03 Ba	0.18 ± 0.03b	0.03 ± 0.03 Aa
OM	0.03 ± 0.03 Aa	0.02 ± 0.02 Aa	0.05 ± 0.01 Ba	0.03 ± 0.04a	0.04 ± 0.02 Aa
PP	0.13 ± 0.02 Bc	0.06 ± 0.01 Bb	0.16 ± 0.02 Cc	0.13 ± 0.03c	0.00 ± 0.02 Aa
SC	0.02 ± 0.02 Ab	0.00 ± 0.03 Ab	−0.07 ± 0.02 Aa	0.03 ± 0.03b	0.01 ± 0.03 Ab
SF	0.11 ± 0.04 Bb	0.17 ± 0.03 Cb	0.16 ± 0.04 Cb	0.16 ± 0.04b	0.01 ± 0.03 Aa
	S63 100%	S63 50%	S63 25%	S63 12%	Control
HB	0.24 ± 0.03 Db	0.02 ± 0.03 Aa	0.03 ± 0.01 Aa	−0.02 ± 0.04 Aa	0.03 ± 0.03 Aa
OM	0.05 ± 0.01 Bb	0.00 ± 0.02 Aab	0.00 ± 0.01 Aa	0.06 ± 0.01 Bb	0.04 ± 0.02 Ab
PP	0.09 ± 0.02 Cb	0.09 ± 0.01 Bb	0.07 ± 0.01 Cb	0.10 ± 0.02 Cb	0.00 ± 0.02 Aa
SC	−0.07 ± 0.05 Aa	0.04 ± 0.02 Ab	0.05 ± 0.01 Bb	0.03 ± 0.02 Bb	0.01 ± 0.03 Ab
SF	0.29 ± 0.02 Dc	0.17 ± 0.06 Bb	0.20 ± 0.04 Db	0.24 ± 0.03 Dbc	0.01 ± 0.03 Aa
	S19 100%	S19 50%	S19 25%	S19 12%	Control
HB	0.13 ± 0.02 Cb	0.13 ± 0.03 Bb	0.11 ± 0.02 Bb	0.13 ± 0.02 Bb	0.03 ± 0.03 Aa
OM	0.06 ± 0.02 Bb	0.01 ± 0.02 Aa	0.00 ± 0.02 Aa	0.05 ± 0.02 Ab	0.04 ± 0.02 Aab
PP	0.10 ± 0.02 Cb	0.15 ± 0.03 Bb	0.09 ± 0.03 Bb	0.12 ± 0.01 Bb	0.00 ± 0.02 Aa
SC	0.01 ± 0.02 Aa	0.01 ± 0.01 Aa	0.01 ± 0.02 Aa	0.08 ± 0.02 Ab	0.01 ± 0.03 Aa
SF	0.14 ± 0.02 Cb	0.14 ± 0.03 Bb	0.29 ± 0.06 Cc	0.13 ± 0.02 Bb	0.01 ± 0.03 Aa

Different capital letters within columns denote statistically significant differences (p < 0.05) among nematode species at the same treatment and concentration, whereas different lowercase letters within rows indicate significant differences (p < 0.05) among concentrations for the same nematode species, according Duncan’s multiple range test.

Chemotactic responses varied significantly among nematode species and treatments, revealing distinct species-specific and concentration-dependent patterns. Heterorhabditis bacteriophora consistently showed positive responses to CBG, with the strongest attraction at 50% CBG (CI = 0.23 ± 0.05), which is classified as a strong attractant. Positive but weaker responses were observed at 100% and 12% CBG (CI = 0.13–0.15), while responses at 25% CBG were neutral. Similarly, Steinernema feltiae showed pronounced attraction to CBG, particularly at 100% and 12% concentrations (CI = 0.17 ± 0.05 for both), indicating weak attractant effects, whereas intermediate concentrations produced more variable responses.

Phasmarhabditis papillosa responded positively to CBG across multiple concentrations, with weak attractant effects at 25% and 12% CBG (CI ≈ 0.10), while responses at higher concentrations were slightly lower but still positive. In contrast, Steinernema carpocapsae generally exhibited neutral to weakly repellent responses to CBG, with CI values ranging from − 0.07 to 0.00, indicating limited sensitivity to this compound. Oscheius myriophilus showed consistently neutral or weakly repellent behaviour across all CBG concentrations, with CI values close to zero or slightly negative.

Responses to crude extracts from Cannabis sativa genotypes S9 (Chemotype II), S63 (Chemotype I), and S19 (Chemotype III) followed similar species-specific trends. H. bacteriophora and S. feltiae showed the most pronounced attraction across genotypes, particularly to S63 and S19 extracts. For example, S. feltiae was strongly attracted to S63 at 100% concentration (CI = 0.29 ± 0.02), clearly exceeding the strong attractant threshold, while H. bacteriophora responded positively to S63 at 100% (CI = 0.24 ± 0.03). Extracts from genotype S19 also consistently elicited weak attractant responses in these species across all concentrations (CI ≈ 0.11–0.14).

Phasmarhabditis papillosa displayed moderate, concentration-dependent attraction to all three chemotypes (I - III), with weak attractant responses most apparent at intermediate concentrations (e.g., S9 25%: CI = 0.16 ± 0.02; S19 50%: CI = 0.15 ± 0.03). In contrast, S. carpocapsae generally exhibited neutral responses to cannabis extracts, with occasional weak attraction at lower concentrations (e.g., S19 12%: CI = 0.08 ± 0.02). Oscheius myriophilus consistently showed CI values near zero across all chemotype extract and concentrations, indicating no clear chemotactic response.

Across all treatments, ethanol controls yielded CI values between − 0.03 and 0.05, all within the neutral response category, confirming that the solvent itself did not influence nematode movement. Collectively, these results demonstrate that nematode chemotaxis towards chemotype extract is strongly species-specific and concentration-dependent, with H. bacteriophora, S. feltiae, and P. papillosa showing the greatest sensitivity.

Discussion

The present study demonstrates that chemotactic responses of parasitic nematodes to Cannabis sativa crude flower extracts are primarily driven by nematode species identity and extract concentration, whereas differences among chemotypes play a comparatively minor role. This finding suggests that behavioural responses are governed more strongly by species-specific sensory ecology and stimulus intensity than by variation in chemotype-specific chemical profiles.

Consistent with previous research on entomopathogenic nematodes (EPNs), strong interspecific differences in chemotactic behaviour were observed. Heterorhabditis bacteriophora and Steinernema feltiae exhibited pronounced attraction to the tested extracts, whereas Steinernema carpocapsae and Oscheius myriophilus showed predominantly neutral responses. Such variability aligns with established differences in foraging strategies among nematodes, where “cruiser” species such as Heterorhabditis spp. and S. feltiae rely more heavily on chemical gradients, while “ambusher” species such as S. carpocapsae are less responsive to volatile cues^24,25. These behavioural strategies are closely linked to the evolution of chemosensory systems that enable nematodes to locate hosts or resources in heterogeneous soil environments.

The strong concentration-dependent responses observed in this study further highlight the importance of signal intensity in mediating nematode behaviour. Increasing extract concentrations generally enhanced chemotactic responses, particularly in H. bacteriophora and S. feltiae, suggesting that these species are highly sensitive to chemical gradients. Similar dose-dependent effects have been reported for plant-emitted volatile organic compounds (VOCs), which can attract or repel nematodes depending on concentration thresholds^26,29. This supports the concept that nematode orientation is governed by gradient detection rather than simple presence or absence of chemical cues.

However, the relationship between concentration and response was not strictly linear. In several cases, intermediate concentrations elicited weaker responses than both lower and higher concentrations. Such non-monotonic patterns may reflect threshold effects in chemosensory perception, behavioural saturation at higher stimulus intensities, or complex interactions among compounds within crude extracts, including synergistic or antagonistic effects. Similar non-linear dose–response relationships have been reported in studies of chemically mediated behaviour in soil organisms^27,28. Nevertheless, the precise mechanisms underlying these responses remain unclear and warrant further investigation.

Importantly, although significant differences were observed in nematode motility, these did not always translate into significant differences in chemotaxis index (CI). This indicates that increased movement does not necessarily imply directional preference, highlighting the importance of distinguishing between general activity and true chemotactic behaviour. Similar discrepancies between locomotion and directional response have been reported in behavioural assays of soil invertebrates, where stimulation can increase activity without inducing oriented movement²⁵. Consequently, CI provides a more reliable measure of behavioural orientation than overall motility alone.

Despite pronounced differences in cannabinoid and terpene composition among the tested chemotypes, the overall lack of a strong chemotype effect on CI suggests that nematode responses are not driven by chemotype identity per se. One possible explanation is that key bioactive compounds are shared across chemotypes, resulting in overlapping chemical signals that elicit similar behavioural responses. Alternatively, the complexity of crude extracts may obscure the effects of individual compounds due to additive or antagonistic interactions. Previous studies have shown that individual cannabinoids such as THC and CBG can differentially influence nematode behaviour³⁰, yet such effects may be less distinct in complex mixtures.

From an ecological perspective, plant-derived metabolites such as cannabinoids and terpenes may function as infochemicals in soil ecosystems, mediating interactions between plants and soil-dwelling organisms. These compounds can enter the soil through litter decomposition, leaching, or root exudation, where they contribute to the chemical landscape that guides nematode behaviour²⁶. However, the ecological relevance of flower-derived extracts should be interpreted cautiously. In natural environments, these compounds are typically present at lower concentrations and are subject to rapid degradation by soil microorganisms, volatilisation, and adsorption to soil particles^34–36. These processes limit their persistence and may reduce their behavioural impact under field conditions.

In addition to behavioural effects, some compounds present in cannabis extracts may exert toxic or sublethal effects on nematodes, particularly at high concentrations. Cannabinoids and terpenes are known to possess bioactive properties, including antimicrobial and insecticidal activity^5–8. However, the present study focused exclusively on chemotactic responses and did not assess mortality or physiological stress. Future research should therefore aim to disentangle behavioural attraction from potential toxic effects and to identify the specific compounds responsible for observed responses.

Overall, our findings support the hypothesis that plant-derived metabolites from C. sativa can function as infochemicals influencing nematode behaviour^37,38. However, the dominant role of species-specific traits and concentration gradients indicates that behavioural responses are shaped more by ecological and physiological factors than by chemotype identity alone. These insights contribute to a growing body of evidence highlighting the role of plant secondary metabolites in structuring belowground interactions and may inform future strategies for enhancing nematode-mediated biological control in agroecosystems.

Future studies should focus on isolating and testing individual compounds and defined mixtures, as well as evaluating nematode responses under more realistic soil conditions^39,40. Integrating chemical ecology with soil biology will be essential for understanding how plant-derived signals influence multitrophic interactions in complex soil environments.

Conclusions

This study demonstrates that chemotactic responses of parasitic nematodes to Cannabis sativa crude flower extracts are primarily influenced by nematode species identity and extract concentration, whereas differences among chemotypes play a comparatively minor role. Behavioural responses were strongly species-specific, with H. bacteriophora, S. feltiae, and P. papillosa showing consistent positive responses, while S. carpocapsae and O. myriophilus exhibited mostly neutral behaviour.

Importantly, increased motility did not always correspond to directional chemotaxis, highlighting the need to distinguish between general activity and true behavioural orientation when interpreting nematode responses. These findings indicate that concentration-dependent chemical cues, rather than chemotype-specific differences, are the dominant drivers of nematode behaviour in this system.

While the results support the role of cannabis-derived metabolites as potential infochemicals, their ecological relevance under natural soil conditions remains uncertain due to likely differences in concentration, rapid degradation, and interactions with soil components. Further research is therefore needed to identify the specific compounds responsible for these effects and to evaluate their activity under more realistic environmental conditions.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Ž.L.: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, visualization, writing – original draft preparation, writing – review and editing. S.T.: formal analysis, validation, writing – review and editing. L.B.: data curation, investigation, visualization. R.Š.: data curation, formal analysis, investigation, methodology, visualization. M.F.: conceptualization, funding acquisition, methodology, project administration, resources, supervision, visualization, writing – original draft preparation, writing – review and editing. All the authors approved the final version of the manuscript. All the authors are accountable for all aspects of the work.

Funding

This research was supported by the Slovenian Research Agency through project J4-50135 and research programmes P4-0431 and P4-0077. The field experiment was conducted at the Laboratory Field, supported by the infrastructural centre IC RRC-AG (IO-0022-0481-001). Additional funding was provided through the Professional Tasks from the Field of Plant Protection programme, financed by the Ministry of Agriculture, Forestry and Food of the Republic of Slovenia (Phytosanitary Administration).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.