Methods of determining insecticidal activity and mode of action

Abstract

A method of determining an insecticidal mode of action of at least one substance is provided, the method comprising: exposing at least one insect to the at least one substance; obtaining behavioural response data characterising a response of the at least one insect to the at least one substance; processing the behavioural response data to identify one or more behavioural responses, wherein the one or more behavioural responses are or correspond to one or more respective phenotypes of the at least one insect; generating a behavioural fingerprint based on the one or more behavioural responses; comparing the behavioural fingerprint to a library reference; and identifying the corresponding insecticidal mode of action of the at least one substance A corresponding apparatus and system are also provided.

Description

[0001] A MONITORING APPARATUS

[0002] Field of Invention

[0003] The present disclosure relates to a monitoring apparatus and to methods of using a monitoring apparatus, preferably for monitoring the behaviour of organisms in response to a substance, in particular for the screening of pesticides and the determination of pesticidal modes of actions.

[0004] Background

[0005] There is an urgent need to develop new and sustainable pesticides, in particular insecticides and acaricides. Over time, insect populations evolve resistance to existing chemical controls, diminishing their effectiveness and threatening crop yields. Introducing novel insecticides with distinct modes of action is essential to break resistance cycles and maintain reliable pest management. Climate change further amplifies the need for new solutions, as shifting weather patterns alter pest dynamics and introduce invasive species into previously unaffected regions.

[0006] Efforts in pesticide development and application have increasingly focused on understanding and utilizing pesticides with distinct modes of action to effectively manage pest populations while mitigating resistance. A mode of action refers to the specific biological mechanism through which a pesticide affects or kills a pest and / or the physical route by which the moiety gets to and carries out its action (e.g., systemic, contact, ingestion, suffocation etc). This approach is underpinned by the principle that rotating or combining pesticides with different modes of action can reduce selection pressure on pests and prolong the efficacy of available products.

[0007] However, the success of these strategies has been constrained by numerous critical gaps in knowledge. These include an incomplete understanding of the spectrum of existing modes of action, uncertainty around the precise mechanisms by which many commonly used pesticides exert their effects, and limited insight into how these mechanisms interact with pest biology. Furthermore, there are inadequate systems and methods available to identify new pesticides with novel modes of action, or to classify the modes of action of known pesticides, thereby limiting their integration into pesticide management strategies. These inadequacies are of growing concern due to the alarming rate at which pests are showing resistance to common pesticides.

[0008] 1 Thus, new and improved systems to screen for novel pesticides, preferably with novel modes of action are needed. The present invention addresses this need.

[0009] Brief Description of Figures

[0010] Preferred embodiments of the invention will now be described by way of non-limiting examples with reference to the following drawings, in which:

[0011] Figures 1 and 2 respectively show perspective and exploded views of a plurality of chambers, a barrier, an actuator and a sample retention device of a monitoring apparatus according to an embodiment of the invention;

[0012] Figures 3 and 4 illustrate a method of using a monitoring apparatus according to an embodiment of the invention;

[0013] Figures 5 and 6 respectively show interior and exterior views of a monitoring apparatus according to an embodiment of the invention;

[0014] Figures 7A to 7F illustrate steps of reconfiguring a monitoring apparatus according to the invention;

[0015] Figures 8A to 8C show an alternative screening configuration of a monitoring apparatus according to the invention;

[0016] Figures 9A to 9D show another alternative screening configuration of a monitoring apparatus according to the invention; and

[0017] Figure 10 illustrates a degree of response of microbes with respect to various phenotypes.

[0018] Figure 11 shows the mean proportion of time Drosophila melanogaster larvae spent performing each behaviour (± SE) across time points for four insecticidal chemistries and the control. Colours represent chemistries (Control = grey; Acetamiprid = green; Carbaryl = orange; Coragen = purple; Spinosad = pink). Facets display individual behaviours tracked by the methods of this invention.

[0019] Figure 12 shows the normalised treatment to its corresponding control clarified temporal dynamics and magnitude of response. Each panel shows the ratio of treated to control behaviour values (mean ± 95% Cl) over time for individual behaviours. The dashed line at y = 1 represents control equivalence. Values > 1 indicate stimulation or hyperactivity; values < 1 indicate suppression or paralysis.

[0020] Figure 13 shows radar charts showing mean portional time spent in each behaviour at selected time points (columns) for each chemistry (rows). Polygons represent overall behavioural profiles (fingerprints) relative to controls. Enlarged sectors correspond to behaviours most affected by treatment. Clockwise around the polygon, the behaviours

[0021] 2 were: bend, small action, back up, crawl, roll, hunch, and stop. Definitions of the behaviours are provided in Example 2.

[0022] Figure 14 shows the mean proportion of time Drosophila melanogaster larvae spent performing each behaviour (± SE) across time points for the control and five microbial treatments. Colours denote treatments: Control = grey, BASF X = gold, BB1008 (Pseudomonas sp.) = green, BB1039 (Curtobacterium sp.) = turquoise, BB990 (Bacillus sp.) = blue, and X sp. = magenta.

[0023] Figure 15 shows the ratio of treated-to-control mean behavioural values (± 95% Cl) for Drosophila melanogaster larvae across time. A dashed line at y = 1 denotes equivalence with controls, with values above 1 indicating stimulation or hyperactivity and those below 1 reflecting behavioural suppression or paralysis.

[0024] The figures are not necessarily to scale, and certain features and certain views of the figures may be shown exaggerated in scale or in schematic form in the interests of clarity and conciseness.

[0025] Summary of the Invention

[0026] The Applicant has identified a new system and methods to identify the mode of action of known pesticides, as well as novel pesticides, in a high-throughput pipeline. In particular, the Applicant has identified that different substances induce different behavioural responses (as well as different combinations of behaviours) in insects, arthropods and pests and that these different behavioural responses are indicative of the different modes of action of each substance.

[0027] By analysing the behavioural responses of a single or plurality of insects, arthropods and pests in response to a new substance and cross-referencing these to the behavioural responses for substances with known modes of action, it is possible to deduce if the substance has pesticidal properties, as well as the specific mode of action for the substance. In particular, the present invention introduces a system that captures, quantifies, and analyses multiple sublethal behavioural parameters in parallel and applies computational inference to derive probabilistic predictions of mode of action. Each behavioural variable, for example agitation, avoidance, grooming, feeding inhibition, and paralysis onset, may be measured as a time-resolved quantitative metric and combined into a multidimensional behavioural “fingerprint”. This multi-faceted, information-rich behavioural fingerprint may then be compared, through a trained statistical or machine-

[0028] 3 learning model, to reference profiles generated from known chemical insecticides and from microbial extracts.

[0029] Using this system, it is possible to screen for new pesticides, and / or pesticides with new or specific modes of action in a high-throughput manner.

[0030] Thus, using the methods and apparatus disclosed herein, it is possible to elucidate the modes of action of novel substances. For example, by exposing an organism to a substance, observing its behaviour, determining a behavioural fingerprint (or behavioural responses, multiple objective endpoints or phenotype) and comparing this to the behavioural responses or modes of action associated with known pesticides, it is possible to identify shared behaviour / mode of action signatures between the novel and known substances and classify the mode of action of the novel substance.

[0031] Further, it is possible to identify additional or alternative modes of action of existing substances. For example, the apparatus and methods described herein may re-categorise a substance’s mode of action or may identify an additional mode of action that was previously unknown.

[0032] The methods and apparatus advantageously enable new classes and types of modes of action to be identified in substances, thereby providing a source of novel pesticides. For example, where the behavioural fingerprint (or behavioural responses, multiple objective endpoints or phenotype) does not meet a critical threshold of similarity compared to known pesticides (e.g., a reference pesticide or behavioural profile), the mode of action associated with the behavioural fingerprint (or behavioural responses, multiple objective endpoints or phenotype) may be considered novel.

[0033] A key advantage of the system and methods disclosed herein over the existing prior art is the ability to identify new classes of pesticides (including live microbes and complex extracts) and the predictive discovery of biological insecticides and emergent modes of action. This difference is highly beneficial because knowing a mode of action can help understand risk profiles and further validation, saving significant resources in downstream discovery.

[0034] According to a first aspect of the invention, there is provided a method of determining an insecticidal mode of action of at least one substance comprising:

[0035] 4 a. exposing at least one insect to the at least one substance; b. obtaining behavioural response data characterising a response of the at least one insect to the at least one substance; c. processing the behavioural response data to identify one or more behavioural responses, wherein the one or more behavioural responses are or correspond to one or more respective phenotypes of the at least one insect; d. generating a behavioural fingerprint based on the one or more behavioural responses; e. comparing the behavioural fingerprint to a library reference; and f. identifying the corresponding insecticidal mode of action of the at least one substance.

[0036] In a second aspect of the invention, there is provided a method of screening at least one substance for pesticide activity comprising: a. exposing at least one insect to at least one substance; b. obtaining behavioural response data characterising a response of the insect to the at least one substance; c. processing the behavioural response data to identify one or more behavioural responses, wherein the behavioural responses are or correspond to one or more respective phenotypes of the at least one insect; d. generating a behavioural fingerprint based on the one or more behavioural responses; e. comparing the behavioural fingerprint to a library reference; and f. identifying if the substance has insecticidal activity.

[0037] In one embodiment, the behavioural response data comprises image or video data.

[0038] In one embodiment, the method comprises exposing a plurality of insects to the at least one substance in parallel.

[0039] In one embodiment, the method comprises exposing a plurality of insects to a plurality of substances.

[0040] 5 In one embodiment, step (e) of the method comprises comparing the behavioural fingerprint to a library reference using a statistical or machine-learning model.

[0041] In one embodiment, the library reference is derived from known insecticidal compounds or microbes, microbial cultures, microbial lysates, microbial metabolites or microbial extract.

[0042] In one embodiment, the method further comprises using a statistical or machine-learning model to determine a probabilistic assignment of a mode of action.

[0043] In one embodiment, the statistical or machine-learning model is trained on datasets comprising behavioural fingerprints of chemical insecticides with known I RAC mode-of- action classifications and microbes, microbial cultures, microbial lysates, microbial metabolites or microbial extracts with experimentally confirmed activity, optionally wherein the datasets are labelled datasets.

[0044] In one embodiment, the statistical or machine-learning model outputs a probability distribution across known modes of action and classifies a behavioural fingerprint as novel when the known modes of action do not exceed a similarity threshold.

[0045] In one embodiment, the behavioural fingerprint is stored as a vector of normalised quantitative values.

[0046] In one embodiment, the exposure step comprises simultaneous release of a plurality of insects to synchronise behavioural exposure to the substance.

[0047] In one embodiment, the substance comprises at least one microbe, a microbial culture, a microbial lysate, a microbial metabolite or microbial extract.

[0048] In one embodiment, the method further comprises integrating the behavioural fingerprint with metabolic or genomic data of a microbial source.

[0049] In another aspect of the invention, there is provided an apparatus for monitoring and analysing insect behaviour in response to a substance, comprising: a. a multi-chamber housing configured to contain at least one or a plurality of insect(s); b. a barrier assembly and actuator configured to synchronously open the chambers to expose the at least one or a plurality of insect(s)to the substance;

[0050] 6 c. one or more illumination sources and imaging devices arranged to capture time- resolved video of the at least one or a plurality of insect(s) from above and / or below; and d. a processor configured to extract quantitative behavioural responses from the recorded video.

[0051] In one embodiment, the method further comprises a memory storing instructions executable by the processor to compute a behavioural fingerprint from the behavioural responses and compare it to reference behavioural fingerprints, preferably using a trained machine-learning model or algorithm.

[0052] In one embodiment, the trained model or algorithm is used to determine an insecticidal mode of action of the substance.

[0053] In one embodiment, the multi-chamber housing is constructed of odour-sterilisable materials selected from glass, stainless steel, or aluminium and is sealed to prevent volatile cross-contamination between chambers.

[0054] In one embodiment, the actuator comprises a sliding-barrier mechanism that releases all insects within milliseconds of each other to enable time-synchronised recording.

[0055] In one embodiment, the imaging device and illumination system are arranged in sealed optical alignment to provide uniform backlighting suitable for motion-tracking analysis.

[0056] In a further aspect of the invention, there is provided a computer-implemented system for classifying insecticidal activity, comprising:

[0057] (a) a behavioural database containing behavioural fingerprints generated according to the method described herein;

[0058] (b) a processing module configured to train a machine-learning model on behavioural fingerprints corresponding to known insecticidal modes of action; and

[0059] (c) a prediction module configured to receive a new behavioural fingerprint and process the new behavioural fingerprint to output at least one of: a probability of activity, a predicted mode of action, or an indication of novelty.

[0060] In one embodiment, the one or more of the behavioural fingerprints are linked to microbial metabolic profiles or genomic data.

[0061] 7 In one embodiment, the machine-learning model uses one or more of logistic regression, random forest, neural network, or clustering algorithms to perform classification and / or novelty detection.

[0062] In one embodiment, the machine-learning model is a continuous model that continuously retrains as behavioural fingerprints are added to the behavioural database.

[0063] Detailed Description

[0064] It is known to have a chamber for housing an organism for behavioural monitoring.

[0065] According to an aspect of the invention, there is provided a method of monitoring organism behaviour, the method comprising the steps of: exposing at least one organism to a substance; and observing a response of the at least one organism to the substance; comparing the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism; and identifying at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints.

[0066] By “multiple objective endpoint” is meant an observable behavioural response of the organism. Accordingly, “multiple objective endpoints” and “behavioural response” are used interchangeably, and both refer to an observable behavioural response by the organism. This behavioural response may be qualitative or quantitative, and preferably the response is quantitative.

[0067] Behavioural responses and multiple objective endpoints are or correspond to phenotypes. A phenotype can be observed as a plurality of behavioural responses. For example, a paralysis phenotype could be observed as multiple behavioural responses (multiple objective endpoints) for example, frequency of movement, aptitude of movement.

[0068] Various types of phenotypes can be observed using the method according to the invention. Non-limiting examples of phenotypes include:

[0069] • Mortality;

[0070] • Motility, mobility, motor function quality and / or paralysis;

[0071] • Deterrence, attraction, seeking and / or avoidance;

[0072] 8 • Feeding behaviour and / or feeding inhibition;

[0073] • Morphological change, such as colour change and / or size change;

[0074] • Spatial organisation;

[0075] • Fecundity;

[0076] • Cleaning;

[0077] • Agitation;

[0078] • Growth inhibition and / or developmental halting.

[0079] Accordingly, a phenotype may be selected from one or more of: mortality; motility, mobility, motor function quality and / or paralysis; deterrence, attraction, seeking and / or avoidance; feeding behaviour and / or feeding inhibition; morphological change, such as colour change and / or size change; spatial organisation; fecundity; cleaning; agitation; and growth inhibition and / or developmental halting.

[0080] In another aspect of the invention, there is a method of determining a pesticidal mode of action of a substance, the method comprising: exposing at least one organism to a substance; observing a response of the at least one organism to the substance; comparing the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism; and identifying at least one pesticidal mode of action of the substance based on the comparison of the response to the multiple objective endpoints.

[0081] In a preferred embodiment, the method is for of determining an insecticidal mode of action of a substance.

[0082] By “pesticidal mode of action” is meant the mechanism or method through which a substance targets, inhibits, or kills pests, insects and arthropods. Preferably, the pesticidal mode of action of a substance refers to insecticidal mode of action, i.e., the mechanism or method through which a substance targets, inhibits, or kills insects.

[0083] Exemplar modes of action include nervous system disruption, muscle function disruption, growth and / or development inhibition, metabolism disruption, gut disruption, physical disruption, feeding inhibition or direct poisoning (directly toxic upon contact or ingestion) and the like.

[0084] 9 In another aspect of the invention, there is a method of determining a pesticidal mode of action of a substance, the method comprising: a. exposing at least one organism to a substance; b. obtaining behavioural response data characterising a response of the at least one organism to the substance; c. processing the behavioural response data to identify one or more behavioural responses, wherein the one or more behavioural responses are or correspond to one or more respective phenotype of the at least one organism; d. generating a behavioural fingerprint based on the one or more behavioural responses, e.g. defined by a combination of the behavioural responses; e. comparing the behavioural fingerprint to a library reference; and f. identifying the corresponding pesticidal mode of action of the substance.

[0085] In one embodiment, the behavioural response data comprises image or video data, e.g. in a time series. This image or video data may be captured using a recording device such as a camera.

[0086] In one embodiment, the library reference is a library of reference behavioural fingerprints, optionally derived from known pesticidal compounds or microbes, microbial cultures, microbial lysates, microbial metabolite or microbial extract.

[0087] In one embodiment, the method of determining a pesticidal mode of action is a high- throughput method.

[0088] In one embodiment, the high-throughput method comprises exposing a plurality of organisms to at least one substance. Such a method would allow the mode of action to be determined across a range of organisms (pests, insects, arthropods).

[0089] In one embodiment, there is a high-throughput method of determining a pesticidal mode of action of a substance, the method comprising: a. exposing a plurality of organisms to a substance, preferably in parallel; b. obtaining behavioural response data characterising responses of the plurality of organisms to the substance; c. processing the response data to identify one or more behavioural responses, wherein the behavioural responses are or correspond to one or more respective phenotypes of the plurality of organisms;

[0090] 10 d. generating a behavioural fingerprint for the plurality of organisms or each organism forming the plurality of organisms based on the one or more behavioural responses; e. comparing the behavioural fingerprint to a library reference; and f. identifying the corresponding pesticidal mode of action of the substance.

[0091] In a preferred embodiment, the high-throughput method comprises exposing a plurality of organisms to a plurality of substances, optionally in parallel. Such a method would allow the mode of action of multiple substances to be determined for a single organism (pests, insects, arthropods) or across a range of organisms (e.g., different species).

[0092] In one embodiment, there is a high-throughput method of determining a pesticidal mode of action of a plurality of substances, the method comprising: a. exposing a plurality of organisms to plurality of substances, preferably in parallel for each substance; b. obtaining behavioural response data characterising responses of the plurality of organisms to the plurality of substances; c. processing the behavioural response data to identify one or more behavioural responses, wherein the behavioural responses are or correspond to one or more respective phenotype of the plurality of organisms; d. generating a behavioural fingerprint for the plurality of organisms or each organism forming the plurality of organisms based on the one or more behavioural responses; e. comparing the behavioural fingerprint to a library reference; and f. identifying the corresponding pesticidal mode of action of the plurality of substances.

[0093] In one embodiment, step a) of the method comprises exposing at least one organism to a substance under controlled environmental conditions. By controlled environmental conditions is meant under regulated or pre-determined temperature, humidity or light exposure conditions; concentrations or formulations of substance; application methods of substance; exposure times of the substance; type of surface of exposure. This can be achieved, for example, by use of a monitoring apparatus as described herein.

[0094] 11 In one embodiment, step a) comprises simultaneous release of a plurality of organisms to synchronise exposure to the substance. This assists with the alignment of the recording of behavioural responses.

[0095] In one embodiment, step e) of the method comprises comparing the behavioural fingerprint to a library reference derived from known pesticidal compounds or insecticidal compounds or microbes, microbial cultures, microbial lysates, microbial metabolite or microbial extract using a statistical or machine-learning model.

[0096] In one embodiment, the method comprises integrating the behavioural fingerprint with metabolic or genomic data of a metabolic source. For example, the behavioural fingerprint may be stored together with metabolic or genomic data, and / or may include a reference or link to metabolic or genomic data. This enhances the predictive accuracy of pesticidal (insecticidal) activity, and may enable multi-modal prediction of insecticidal potential by cross-referencing between behavioural fingerprints and associated metabolic or genomic data.

[0097] In one embodiment, the method comprises using a statistical or machine-learning model to determine a probabilistic assignment of a mode of action. By “probabilistic assignment” is meant assigning an outcome or category (such as mode of action) based on probabilities e.g., percentage chance across outcomes or categories.

[0098] In one embodiment, the machine-learning model is trained on datasets comprising behavioural fingerprints of chemical insecticides with known I RAC mode-of-action classifications and microbes, microbial cultures, microbial lysates, microbial metabolite or microbial extracts with experimentally confirmed activity, optionally wherein the datasets are labelled datasets.

[0099] By I RAC mode-of-classification is meant the classification system published by the Insecticide Resistance Action Committee. As of writing, the mode of classification consulted is version 11.4 (published May 2025) A link to the classification system is provided here: https: / / irac-online.org / documents / moa-classification / . The entirety of the contents are incorporated by reference.

[0100] The I RAC mode of action classification comprises main groups numbered 1 to 36 and seven non-numbered groups:

[0101] 12 Group 1 - Acetylcholinesterase inhibitors. Site: Nerve action (synapse)

[0102] Group 2 - GABA-gated chloride channel antagonists. Site: Nerve action (chloride channel)

[0103] Group 3 - Sodium channel modulators. Site: Nerve action (axon)

[0104] Group 4 - Nicotinic acetylcholine receptor competitive modulators. Site: Nerve action (synapse)

[0105] Group 5 - Nicotinic acetylcholine receptor allosteric modulators - site I. Site: Nerve action (synapse)

[0106] Group 6 - Glutamate-gated chloride channel activators. Site: Nerve action (chloride channel)

[0107] Group 7 - Juvenile hormone receptor modulators. Site: Growth regulation

[0108] Group 8 - Miscellaneous non-specific nerve inhibitor. Site: Nerve & muscle

[0109] Group 9 - Chordotonal organ TRPV channel modulators. Site: Feeding inhibition

[0110] Group 10 - Mite growth inhibitors CSH1. Site: Growth regulation

[0111] Group 11 - Microbial disruptors of insect midgut membranes. Site: Midgut

[0112] Group 12 - Inhibitors of mitochondrial ATP synthase. Site: Energy metabolism

[0113] Group 13 - Uncouplers of oxidative phosphorylation via disruption of proton gradient.

[0114] Site: Energy metabolism

[0115] Group 14 - Nicotinic acetylcholine receptor channel blockers. Site: Nerve action

[0116] Group 15 - Inhibitors of chitin biosynthesis, CHS1. Site: Growth regulation

[0117] Group 16 - Inhibitors of chitin biosynthesis, type 1. Site: Growth regulation

[0118] Group 17 - Molting disruptors, Dipteran Site: Growth regulation

[0119] Group 18 - Ecdysone receptor agonists. Site: Growth regulation

[0120] Group 19 - Octopamine receptor agonists. Site: Nerve action

[0121] Group 20 - Mitochondrial complex III electron transport inhibitors - Qo site. Site: Energy metabolism

[0122] Group 21 - Mitochondrial complex I electron transport inhibitors. Site: Energy metabolism

[0123] Group 22 - Voltage-dependent sodium channel blockers electron transport inhibitors.

[0124] Site: Nerve action

[0125] Group 23 - Inhibitors of lipid biosynthesis. Site: Lipid synthesis, growth regulation

[0126] Group 24 - Mitochondrial complex IV electron transport inhibitors. Site: Energy metabolism

[0127] Group 25 - Mitochondrial complex II electron transport inhibitors. Site: Energy metabolism

[0128] Group 26 - unassigned Site: Unknown

[0129] Group 27 - unassigned. Site: unknown

[0130] 13 Group 28 - Ryanodine receptor modulators (Diamides). Site: Muscle action

[0131] Group 29 - Chordotonal organ modulators (IR). Site: Feeding inhibition

[0132] Group 30 - GABA-gated chloride channel allosteric modulators. Site: Nerve action Group 31 - Baculoviruses. Site: Host-specific occluded pathogenic viruses

[0133] Group 32 - Nicotinic Acetylcholine Receptor (nAChR) Allosteric Modulators - Site II. Site: Nerve action

[0134] Group 33 - Nicotinic Acetylcholine Receptor (nAChR) Allosteric Modulators - Site II. Site: Nerve action

[0135] Group 34 - Nicotinic Acetylcholine Receptor (nAChR) Allosteric Modulators - Site II. Site: Nerve action

[0136] Group 35 - Chordotonal organ modulators - undefined target site. Site: Nerve action Group 37 - Vesicular acetylcholine transporter (VAChT) inhibitor - Site II. Site: Nerve action

[0137] UN - Compounds of unknown or uncertain MoA

[0138] UNB - Bacterial agents (non-Bt) of unknown or uncertain MoA

[0139] UNE - Botanical essence including synthetic, extracts and unrefined oils with unknown or uncertain MoA

[0140] UNF - Fungal agents of unknown or uncertain MoA

[0141] UNM - Non-specific mechanical and physical disruptors

[0142] UNP - Peptides of unknown or uncertain MoA

[0143] UNV - Viral agents (non-baculovirus) of unknown or uncertain MoA

[0144] An advantage of the instant invention is that the methods and apparatus are optimised to identify and elucidate substances that classify as belonging to groups UN, UNB, UNE, UNF, UNM, UNP, and UNV.

[0145] In one embodiment, a mode of action or behavioural fingerprint is considered novel when no known mode of action exceeds a similarity threshold. In one embodiment, the similarity threshold may be at least 40%, 50%, 60%, 70%, 75%, 80%, 85%, 90% or 95% similarity. Thus, a behavioural fingerprint or mode of action may be classified as novel when no known mode of action exceeds at least 40%, 50%, 60%, 70%, 75%, 80%, 85%, 90% or 95% similarity.

[0146] In one embodiment, a machine-learning model outputs a probability distribution across known modes of action and classifies a behavioural fingerprint as novel when no known mode of action exceeds a similarity threshold.

[0147] 14 In one embodiment, the behavioural fingerprint is stored as a vector of normalised quantitative values permitting cross-experiment comparison and clustering.

[0148] In one embodiment, the behavioural fingerprint is integrated with metabolic and / or genomic data of a microbial source. For example, the behavioural fingerprint may be stored together with metabolic or genomic data, and / or may include a reference or link to metabolic or genomic data. This enhances the predictive accuracy of pesticidal classifications.

[0149] In embodiments of the invention, the method may include the step of recording a video of the response of the at least one organism to the substance. In other embodiments of the invention, the method may include the step of recording an image of the response of the at least one organism to the substance.

[0150] In further embodiments of the invention, the method may include the step of recording a plurality of videos or images of the response of the at least one organism to the substance. Each video or image may be recorded at a respective different time.

[0151] According to another aspect of the invention, there is provided a computer-implemented method of monitoring organism behaviour, the method comprising the steps of: using a recording of a response of at least one organism to a substance, comparing the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism; and identifying at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints.

[0152] The method of the invention may include the step of using the recording of the response, continuously observing a response of the at least one organism to the substance.

[0153] In another aspect of the invention is provided a method of screening at least one substance for pesticidal activity, the method comprising: exposing at least one organism to at least one substance; observing a response of the at least one organism to the at least one substance; comparing the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism; and

[0154] 15 identifying if the at least one substance has pesticidal activity based on the comparison of the response to the multiple objective endpoints.

[0155] In another embodiment, the method comprises: a. exposing at least one organism to at least one substance; b. obtaining behavioural response data characterising a response of the organism to the at least one substance; c. processing the behavioural response data to identify one or more behavioural responses, wherein the behavioural responses are or correspond to one or more respective phenotypes of the at least one organism; d. generating a behavioural fingerprint based on the one or more behavioural responses; e. comparing the behavioural fingerprint to a library reference; and f. identifying if the substance has pesticidal activity.

[0156] The behavioural response data may comprise image or video data.

[0157] In one embodiment, the substance is identified as having novel pesticidal activity. That is, the substance is identified as having pesticidal activity for the first time.

[0158] In one embodiment, the substance is identified as having a novel mode of action. That is, the substance is identified as having a novel mode of action for the first time or a novel additional (second, third, fourth etc.) mode of action.

[0159] In one embodiment, the method of screening at least one substance for pesticidal activity is a high-throughput screening method.

[0160] In one embodiment, the high-throughput screening method comprises exposing a plurality of organisms to at least one substance in parallel.

[0161] In one embodiment, the high-throughput screening method comprises exposing a plurality of organisms to a plurality of substances.

[0162] In another aspect of the invention, there is a substance with pesticidal activity obtained or obtainable from any of the methods disclosed herein.

[0163] 16 Furthermore, the method may include the step of using an artificial intelligence algorithm or model to analyse the observed response of the at least one organism to the substance in order to compare the response to the multiple objective endpoints and identify the at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints.

[0164] According to yet another aspect of the invention, there is provided a computer- implemented method of identifying at least one mode of action of a substance, the method comprising the steps of: collecting a set of data by carrying out the method according to any one of the preceding aspects of the invention and their embodiments, wherein the collected set of data includes the identified at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints; creating a training set including the collected set of data; training a machine learning algorithm or model using the training set; and identifying the at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints based on an output of the machine learning algorithm or model.

[0165] In another aspect of the invention, there is provided a computer-implemented method of identifying at least one pesticidal mode of action of a substance, the method comprising the steps of: a) collecting a set of behavioural response data using a method according to any one of the preceding aspects of the invention and their embodiments; b) creating a training set including the collected set of data; c) training a machine-learning algorithm or model on the training set; and d) using the machine-learning algorithm or model to identify the at least one pesticidal mode of action of the substance.

[0166] In another aspect, there is provided a computer-implemented method of determining at least one pesticidal mode of activity of a substance, comprising:

[0167] (a) collecting a set of behavioural response data by carrying out a method according to any one of the preceding aspects and their embodiments;

[0168] (b) using a computer-implemented method to compare the data to a database containing behavioural fingerprints; and

[0169] (c) using a prediction module to output at least one of: a probability of activity, a predicted mode of action, or an indication of novelty.

[0170] 17 In one embodiment, the method comprises using a statistical or machine-learning model to determine a probabilistic assignment of a mode of action.

[0171] In one embodiment, the behavioural fingerprint is stored as a vector of normalised quantitative values for cross-experiment comparison and clustering.

[0172] In another aspect of the invention, there is provided a computer-implemented method of classifying pesticidal activity of a substance, comprising a) collecting a set of behavioural response data using a method according to any one of the preceding aspects of the invention and their embodiments; b) creating a training set including the collected set of data; c) training a machine-learning algorithm or model on the training set; and d) using the machine-learning algorithm or model to classify the pesticidal activity of the substance.

[0173] In another aspect, there is provided a computer-implemented method of classifying pesticidal activity of a substance, comprising:

[0174] (a) collecting a set of behavioural response data by carrying out a method according to any one of the preceding aspects and their embodiments;

[0175] (b) using a computer-implemented method to compare the data to a database containing behavioural fingerprints; and

[0176] (c) using a prediction module to classify pesticidal activity of a substance.

[0177] In another aspect, there is provided a computer-implemented system of determining at least one pesticidal mode of activity of a substance, comprising: a) a database comprising behavioural fingerprints generated according to the methods disclosed herein; b) a processing module configured to train a machine-learning model on behavioural fingerprints corresponding to known pesticidal modes of action; and c) a prediction module configured to receive a new behavioural fingerprint and process the new behavioural fingerprint to output at least one of: a probability of activity, a predicted mode of action, or an indication of novelty.

[0178] The behavioural fingerprints used to train the machine-learning model may be labelled for use in training.

[0179] 18 In another aspect, there is provided a computer-implemented system for classifying pesticidal activity of a substance, comprising: a) a database comprising behavioural fingerprints generated according to the method disclosed herein; b) a processing module configured to train a machine-learning model on behavioural fingerprints corresponding to known pesticidal modes of action; and c) a prediction module configured to receive a new behavioural fingerprint and process the new behavioural fingerprint to output at least one of: a probability of activity, a predicted mode of action, or an indication of novelty.

[0180] In one embodiment, the system comprises a database comprising behavioural fingerprints that are associated with or linked to microbial metabolic profiles or genomic data.

[0181] In one embodiment, the system facilitates multi-modal prediction of insecticidal potential.

[0182] In one embodiment, the machine-learning model is trained on datasets comprising behavioural fingerprints of chemical insecticides with known I RAC mode-of-action classifications and microbes, microbial cultures, microbial lysates, microbial metabolites or microbial extracts with experimentally confirmed activity.

[0183] In one embodiment, the machine-learning model outputs a probability distribution across known modes of action and classifies a behavioural fingerprint as novel when no known mode of action exceeds a similarity threshold.

[0184] In one embodiment, the machine-learning model uses one or more of logistic regression, random forest, neural network, or clustering algorithms to perform classification and / or novelty analysis.

[0185] In one embodiment, the machine-learning model is a continuous model that retrains as additional data is inputted. The additional data may relate to the organism, substance or microbial profile data, and / or may comprise further behavioural fingerprints.

[0186] In one embodiment, the system integrates the behavioural fingerprint with metabolic and / or genomic data of a microbial source, for example by linking the behavioural fingerprint with the metabolic and / or genomic data. This enhances the predictive accuracy of pesticidal activity.

[0187] 19 According to a further aspect of the invention, there is provided a computer program comprising computer code configured to perform the method of the invention.

[0188] In the invention, the at least one organism is preferably an insect, a pest or an arthropod.

[0189] In a preferred embodiment, the at least one organism is an insect.

[0190] By “insect” is meant six legged arthropods or those belonging to the class Insecta. Examples of insects include Hemiptera (true bugs; sharing a common arrangement of piercing-sucking mouthpart such as aphids, whiteflies, mealybugs) Coleoptera (beetle, weevils), Hymenoptera (ants, bees, and wasps), Lepidoptera (butterflies, moths, armyworms, cutworms, waxworms, diamondback moth), Diptera (flies including fruit flies, gnats, and mosquitoes), Orthoptera (crickets, grasshoppers, and locust). This invention particularly relates to insects which are pests, i.e., those which cause damage to agricultural crops.

[0191] Accordingly, in one embodiment, the insect is selected from an insect in the order Hemiptera, Coleoptera, Hymenoptera, Lepidoptera, Diptera and Orthoptera. In a preferred embodiment, the insect is selected from: an aphid, whiteflies, mealybug, a beetle (optionally a mealworm), a weevil, an ant, wasp, butterfly, moth, armyworm, cutworm, a fruit fly, gnat, mosquito, cricket, grasshopper and locust. In a most preferred embodiment, the insect is selected from: a mealworm, waxworm, diamondback moth, aphid or Drosophila.

[0192] The skilled person will understand that insects which undergo complete metamorphosis undergo four stages of life: egg, larva, pupa, and adult and insects that undergo incomplete metamorphosis undergo three stages of transformation: egg, nymph, and adult. In one embodiment, the insect is at the larva, pupa, nymph or adult stage of life. For example, the mealworm is the larval form of the yellow mealworm beetle, Tenebrio molitor - so the invention is intended to extend to cover both mealworm and Tenebrio molitor As a second example, the diamondback moth larva is the larval form of the Diamondback moth - so the invention is intended to extend to cover both larval and adult form.

[0193] By “pest” is meant a non-arthropod or non-insect organism which is harmful or causes damages to agricultural crops or flora. Examples include molluscs (slugs and snails) and nematodes (roundworms). Accordingly, in one embodiment, the pest is selected from a mollusc or a nematode.

[0194] 20 By “arthropod” is meant invertebrates in the phylum Arthropoda excluding insects. Though the phylum Arthropoda includes insects, insects are discussed separately in this application, therefore as used herein arthropod refers to members of the Arthropoda phylum, excluding insects. Examples of arthropods include arachnids (spiders, ticks, mites), myriapods (such as centipedes and millipedes) and insects (arthropods with six legs, include, for example - aphids, bollworms, whiteflies, grasshoppers, and beetles). Accordingly, in one embodiment, the arthropod is selected from an arachnid and a myriapod.

[0195] According to a further aspect of the invention, there is provided a monitoring apparatus for monitoring organism behaviour, the monitoring apparatus comprising: at least one chamber configured for housing at least one organism, the or each chamber configurable to selectively expose the at least one organism to a substance; a monitoring device configured to, in use, record a response of the at least one organism to the substance; a processor configured to, using the recording of the response, compare the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism, and identifying at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints.

[0196] In another aspect of the invention, the monitoring apparatus is suitable for determining a pesticidal mode of action of a substance. Preferably the monitoring apparatus is suitable for determining an insecticidal mode of action of a substance.

[0197] Accordingly in one embodiment, there is a monitoring apparatus for determining a pesticidal mode of action of a substance, the apparatus comprising: at least one chamber configured for housing at least one organism, the or each chamber configurable to selectively expose the at least one organism to a substance; a monitoring device configured to record a plurality of behavioural responses of the at least one organism to the substance; a processor configured to, using the recording of the plurality of behavioural responses, compare the plurality of behavioural responses, wherein the behavioural responses are or correspond to one or more respective phenotype of the at least one organism, and identifying at least one mode of action of the substance.

[0198] 21 In one embodiment, the processor is configured to, using the recording of the plurality of behavioural responses, generate a behavioural fingerprint defined by the combination of said behavioural responses.

[0199] In one embodiment, the processor is further configured to compare the behavioural fingerprint to a library reference derived from known pesticidal compounds or microbial extracts, and optionally identifying the corresponding pesticidal mode of action of the substance. For example, the processor is further configured to compare the behavioural fingerprint to a library reference derived from known insecticidal compounds or microbial extracts, and optionally identify the corresponding insecticidal mode of action of the substance.

[0200] In one embodiment, the processor is configured to use a statistical or machine-learning model to determine a probabilistic assignment of a mode of action.

[0201] The monitoring device may be configured to, in use, record a video or image of the response of the at least one organism to the substance.

[0202] The monitoring apparatus may include a memory including computer program code. The memory and computer program code may be configured to, with the processor, enable the monitoring apparatus to carry out various processing functions. The processor and memory may form part of one or more of an electronic device, a portable electronic device, a portable telecommunications device, a mobile phone, a personal digital assistant, a tablet, a phablet, a laptop computer, a server, a cloud computing network, a smartphone, a smartwatch, smart eyewear, and a module for one or more of the same. It will be appreciated that references to a memory or a processor may encompass a plurality of memories or processors.

[0203] According to a still further aspect of the invention, there is provided a monitoring apparatus for monitoring organism behaviour, the monitoring apparatus comprising: at least one chamber configured for housing at least one organism; a barrier configured to retain the or each organism within the or each chamber; an actuator configured to, in use, control the barrier to selectively open and close the or each chamber, the barrier permitting release of the or each organism from the or each chamber when the barrier is configured to open the or each chamber, the barrier

[0204] 22 retaining the or each organism within the or each chamber when the barrier is configured to close the or each chamber; and a monitoring device configured to, in use, monitor the at least one released organism.

[0205] In an alternative aspect, the monitoring apparatus is for determining the pesticidal mode of action of a substrate on at least one organism. Preferably, for determining the insecticidal mode of action of a substrate on at least one insect.

[0206] Preferably, the actuator is configured to, in use, control the barrier to synchronously open some or all of the chambers, thereby releasing the organisms from the opened chambers and exposing the organisms to the sample.

[0207] The Applicants have identified that the use of an actuator configured to synchronously open the chambers to expose a plurality of insect(s) to the substance synchronises the behavioural profile across the plurality of insects. This reduces the “background noise” of any signals.

[0208] In another aspect of the invention, there is provided an apparatus for monitoring and / or analysing the behaviour of at least one organism in response to a substance, comprising:

[0209] (a) a multi-chamber housing configured to contain at least one organism or a plurality of organism(s);

[0210] (b) a barrier assembly and actuator configured to synchronously open the chambers to expose the at least one or plurality of organism(s) to the substance;

[0211] (c) one or more illumination sources and imaging devices arranged to capture time- resolved video of the at least one or plurality of organism(s) from above and / or below; and

[0212] (d) a processor configured to extract quantitative behavioural parameters from the recorded video.

[0213] In one embodiment, the apparatus is for monitoring and / or analysing the behaviour of a plurality of organisms in response to a substance.

[0214] In one embodiment, the monitoring apparatus is for use in a method of screening at least one substance for pesticidal activity, as described herein. In one embodiment, the monitoring apparatus is for use in a method of high-throughput screening at least one substance for pesticidal activity, as described herein.

[0215] 23 In a preferred embodiment, the at least one organism or plurality of organisms is an insect or plurality of insects.

[0216] In one embodiment, the plurality of organisms belong to the same species or class of organism. For example, the plurality of organisms are insects of the same species e.g., all insects are Drosophila species.

[0217] In one embodiment, the plurality of organisms belong to multiple species or class of organism. For example, there may be a plurality of organisms from multiple species, such as a plurality of mealworms, a plurality of waxworms and a plurality of diamondback moth larvae.

[0218] In one embodiment, the apparatus further comprises:

[0219] (d) a memory storing instructions executable by the processor to compute a behavioural fingerprint from said parameters and compare it to a plurality of reference behavioural fingerprints, preferably using a trained machine-learning model or algorithm.

[0220] In one embodiment, the trained model or algorithm is used to determine an insecticidal mode of action of the substance.

[0221] Features of the method of the invention described herein apply mutatis mutandis to the monitoring apparatus of the invention. It will also be appreciated that the method of the invention is not limited to being performed by the monitoring apparatus of the invention.

[0222] It is envisaged that different aspects of the monitoring apparatus (or apparatus for monitoring an organism) of the invention may be used in combination.

[0223] The monitoring apparatus may include a plurality of chambers, each of which is configured for housing at least one respective organism. The barrier may be configured to retain the organisms within the plurality of chambers. The actuator may be configured to, in use, control the barrier to selectively open and close the plurality of chambers, the barrier permitting release of at least one of the organisms from the plurality of chambers when the barrier is configured to open the plurality of chambers, the barrier retaining the organisms within the plurality of chambers when the barrier is configured to close the plurality of chambers.

[0224] 24 The barrier may be formed of a single unitary barrier member configured to retain the organisms within the plurality of chambers.

[0225] The barrier may include a plurality of discrete barrier members. Each barrier member may be configured to separately retain the organisms within a respective one or more of the plurality of chambers.

[0226] The barrier may include a plurality of apertures spaced apart from each other by blocking sections. The apertures may be aligned with the plurality of chambers when the barrier is configured to open the plurality of chambers. The blocking sections may be aligned with the plurality of chambers when the barrier is configured to close the plurality of chambers.

[0227] The actuator may be configured to, in use, control the barrier to synchronously open and close the plurality of chambers. It will be understood that the actuator may be replaced by a plurality of actuators in this embodiment and throughout the specification.

[0228] The barrier may be removably coupled to the actuator. The barrier may be made of a metallic material. The barrier may be configured to be slidable by the actuator to selectively open and close the or each chamber.

[0229] The actuator may be a linear actuator. The actuator may be remotely controllable.

[0230] The monitoring apparatus may further include a programmable controller configured to control the actuator. The programmable controller may be or may include, but is not limited to, a microprocessor or a computer.

[0231] The monitoring apparatus may further include a sample retention device for holding a sample. The barrier may be configured to, in use, separate the or each organism from the sample retention device when the barrier is configured to close the or each chamber. The barrier may be configured to, in use, provide the or each organism with access to the sample retention device when the barrier is configured to open the or each chamber.

[0232] The barrier and the sample retention device may be configured to be in sealing cooperation to provide at least one sealed environment for the or each organism and the sample when the barrier is configured to open the or each chamber. In embodiments employing the use of a plurality of chambers, the barrier and the sample retention device may be configured

[0233] 25 to be in sealing cooperation to provide sealed environments for the organisms and the sample when the barrier is configured to open the plurality of chambers.

[0234] The sample retention device may be a sample clamping device for clamping the sample. The sample clamping device may include at least one clamping member corresponding to the or each chamber. The or each clamping member may be configured to, in use, position the sample against or towards the or each corresponding chamber.

[0235] The sample clamping device may include a first clamping member support structure. The or each clamping member may be attached to, or form part of, the first clamping member support structure.

[0236] The sample clamping device may include a second clamping member support structure. The or each clamping member and the second clamping member support structure may be configured to, in use, clamp the sample between the or each clamping member and the second clamping member support structure.

[0237] The barrier, the or each clamping member and the second clamping member support structure may be configured to be in sealing cooperation to provide at least one sealed environment for the or each organism and the sample when the barrier is configured to open the or each chamber. In embodiments employing the use of a plurality of chambers, the barrier, the clamping members and the second clamping member support structure may be configured to be in sealing cooperation to provide sealed environments for the organisms and the sample when the barrier is configured to open the plurality of chambers.

[0238] The monitoring apparatus may include at least one removable well assembly. The or each well assembly may be configured for holding at least one sample and at least one organism. The monitoring device may be configured to, in use, monitor the at least one organism in the or each well assembly.

[0239] The or each well assembly may be made of a glass material, a metallic material, an aluminium material or a stainless steel material. The monitoring apparatus may include at least one removable tubular conduit. The or each tubular conduit may be configured for holding at least one sample and at least one organism. The monitoring device may be configured to, in use, monitor the at least one organism in the or each tubular conduit.

[0240] 26 The or each tubular conduit may include at least one sample holder arranged at one end of the tubular conduit. The at least one sample holder may be configured for containing a liquid sample. The sample holder may include two membrane layers separated by a spacing member. The or each tubular conduit may be made of a glass material, a metallic material, an aluminium material or a stainless steel material.

[0241] The or each removable well assembly and / or the or each tubular conduit may be configured for use after temporarily removing the at least one chamber, the barrier and the actuator from the monitoring apparatus. This is to provide the monitoring apparatus with one or more alternative configurations for monitoring the or each organism.

[0242] The sample retention device may be made of a metallic material.

[0243] The monitoring apparatus may further include a housing in which the or each chamber, the barrier and the actuator are located. The housing may be described as a multi-chamber housing, where it comprises a plurality of chambers.

[0244] The monitoring apparatus may further include a light source configured to, in use, illuminate the or each chamber.

[0245] The monitoring apparatus may further include one or more light blocking masks arranged to, in use, block part of an illumination light provided by the light source. This is also referred to as an “illumination source”. The or each light blocking mask may be used in combination with the or each well assembly and / or the or each tubular conduit.

[0246] The imaging device and illumination system are preferably arranged in sealed optical alignment to provide uniform backlighting suitable for motion-tracking analysis.

[0247] The or each chamber may be made of a glass material, a metallic material, an aluminium material or a stainless steel material. Preferably, the or each chamber is made of glass, stainless steel, or aluminium.

[0248] The monitoring device may be or may include an imaging device.

[0249] The monitoring apparatus may be an insect, pest or arthropod monitoring apparatus. In a preferred embodiment, the apparatus or methods disclosed herein relate to insects. For example, the apparatus or monitoring apparatus is an insect monitoring apparatus. In another preferred embodiment, the apparatus or monitoring apparatus is for monitoring and / or analysing insect behaviour. In another preferred embodiment, the apparatus or monitoring apparatus is for determining an insecticidal mode of action of a substance.

[0250] According to another aspect of the invention, there is provided a method of using a monitoring apparatus in accordance with any one of the preceding aspects of the invention and its embodiments, the method comprising the steps of: loading at least one organism into the or each chamber; by the barrier, retaining the or each organism within the or each chamber; providing a substance; by the actuator, controlling the barrier to open the or each chamber to release the or each organism from the or each chamber and thereby expose the or each organism to the substance; and by the monitoring device, monitoring a behaviour of the or each organism in response to the substance.

[0251] The substance may be a stimulus-generating substance that produces a stimulus to which the at least one organism may respond. The substance may be an odour-generating substance. The substance may be a leaf. The substance may be an uncut leaf. The substance may be or may include at least one crop material, at least one insect target, at least one microbe, at least one microbial extract, at least one natural product, at least one metabolite, at least one metabolite mixture, at least one chemical active, at last one chemical compound and / or at least one biological material.

[0252] In a preferred embodiment, the substance consists or comprises of at least one microbe, a microbial culture, a microbial lysate, at least one microbial metabolite, or at least one microbial extract. As shown in Example 2, we have shown that substances from microbes are able to elicit a behavioural response that can be used to screen for pesticides with novel modes of action, and identify the modes of action for existing pesticides.

[0253] By microbial extract is meant a component derived from a microbial culture or lysate that can be fractionated from a bulk mixture of microbial cultures or microbial lysate.

[0254] The substance may be classified as an adjuvant. By adjuvant is meant a substance added to a pesticidal product or mixture to improve its effectiveness and application properties.

[0255] 28 In one embodiment, at least one substance means at least one substance, preferably two substances, preferably three substances, or preferably at least four substances.

[0256] Through the incorporation of multiple substances, it is possible to study the interaction of substances (being tested / classified) with other substances which may or may not be known to be pesticides or have a known mode of action. For example, analyse the interaction between a substance known to have pesticidal activity and a substance not known to have (but under investigation for) pesticidal activity. As another example, study the interaction between a known microbial compound with a known mode of action with a known adjuvant to learn of additional modes of action.

[0257] Accordingly, in one embodiment, all the substances (i.e., the at least one, two, three substances) are anticipated or known to have pesticidal activity.

[0258] In another embodiment, a first substance is known to have pesticidal activity and a second (or more) substance(s) is (are) analysed for pesticidal activity or for determining modes of action as described herein.

[0259] In one embodiment, where two or more substances are present, the first substance is a substance anticipated or known to have pesticidal activity and the second (or more) substance(s) does not have known pesticidal activity.

[0260] The step of providing a substance may include adding a substance as a sample to the sample retention device prior to the barrier being controlled to open the or each chamber.

[0261] The method may further include the step of controlling the barrier to synchronously open the plurality of chambers to release the organisms from the plurality of chambers and thereby expose the organisms to the substance at the same time. This may be achieved by a sliding-barrier mechanism. The skilled person will understand that a sliding-barrier mechanism refers to when the barrier slides in and out of closed / open position.

[0262] According to yet another aspect of the invention, there is provided a monitoring apparatus comprising: at least one tubular conduit configured for holding at least one sample and at least one organism, wherein the or each tubular conduit includes at least one sample holder for

[0263] 29 holding the at least one sample, the sample holder including two membrane layers separated by a spacing member, a monitoring device configured to, in use, monitor the at least one organism in the or each tubular conduit.

[0264] In a preferred embodiment, the monitoring apparatus is suitable for or for use in a method of determining the pesticidal mode of action of a substrate on at least one organism.

[0265] In a preferred embodiment, the monitoring apparatus is suitable for or for use in a method of determining the insecticidal mode of action of a substrate on at least one insect.

[0266] The or each tubular conduit may include at least one sample holder arranged at one end of the tubular conduit. The at least one sample holder may be configured for containing a liquid sample. The tubular conduit may be made of a glass material, a metallic material, an aluminium material or a stainless steel material.

[0267] It will be appreciated that the use of the terms “first” and “second”, and the like, in this patent specification may be used to help distinguish between similar features (e.g., the first and second clamping structures) and is not intended to indicate the relative importance of one feature over another feature, unless otherwise specified.

[0268] Within the scope of this application it is expressly intended that the various aspects, embodiments, examples and alternatives set out in the preceding paragraphs, and the claims and / or the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. That is, all embodiments and all features of any embodiment can be combined in any way and / or combination, unless such features are incompatible. The applicant reserves the right to change any originally filed claim or file any new claim accordingly, including the right to amend any originally filed claim to depend from and / or incorporate any feature of any other claim although not originally claimed in that manner.

[0269] Preferred embodiments of the invention will now be described by way of non-limiting examples with reference to the following drawings, in which:

[0270] Figures 1 and 2 respectively show perspective and exploded views of a plurality of chambers, a barrier, an actuator and a sample retention device of a monitoring apparatus according to an embodiment of the invention;

[0271] 30 Figures 3 and 4 illustrate a method of using a monitoring apparatus according to an embodiment of the invention;

[0272] Figures 5 and 6 respectively show interior and exterior views of a monitoring apparatus according to an embodiment of the invention;

[0273] Figures 7A to 7F illustrate steps of reconfiguring a monitoring apparatus according to the invention;

[0274] Figures 8A to 8C show an alternative screening configuration of a monitoring apparatus according to the invention;

[0275] Figures 9A to 9D show another alternative screening configuration of a monitoring apparatus according to the invention; and

[0276] Figure 10 illustrates a degree of response of microbes with respect to various phenotypes.

[0277] The figures are not necessarily to scale, and certain features and certain views of the figures may be shown exaggerated in scale or in schematic form in the interests of clarity and conciseness.

[0278] A plurality of chambers 20, a barrier 30, an actuator 40 and a sample retention device 50 of a monitoring apparatus according to an embodiment of the invention are shown in Figures 1 and 2. The sample retention device 50 is optional, and it is envisaged that the sample retention device 50 may not be included in all embodiments of the invention.

[0279] Each of the plurality of chambers 20 may house an organism. In further embodiments, each chamber 20 may house more than one organism. In some embodiments, the organism may be an insect or an agricultural pest or an arthropod.

[0280] The plurality of chambers 20 may be in the form of a well assembly (e.g. a 256-well assembly) in which each chamber 20 is a well. In other embodiments, the well assembly may be removable for replacement by another well assembly. In further other embodiments, the plurality of chambers 20 may be in the form of multiple well assemblies in which each well assembly includes some of the chambers 20.

[0281] The plurality of chambers 20 may be made of a metallic material. In alternative embodiments, the plurality of chambers 20 may be made of another material which does not retain or can be cleansed of stimulus-generating substances, odour-generating substances or odours. Accordingly, in one embodiment, the plurality of chambers 20 are odour sterilisable. By ‘odour sterilisable’ is meant that the material does not inherently

[0282] 31 retain an odour or can be cleaned of odour- or stimulus-generating substances. For example, and not limitation, a chamber 20 may be made of glass, a non-porous material, or a specialised synthetic material or polymer. Preferably, the chamber housing comprises or is constructed of an odour sterilisable material selected from glass, stainless steel or aluminium.

[0283] In one embodiment, at least a portion, preferably all, of the apparatus of the invention can be sealed. In one embodiment, the chamber is sealed. Preferably, the chamber is sealed to prevent cross contamination between chambers.

[0284] The barrier 30 is configured, in its closed configuration, to retain the organisms within each of the plurality of chambers 20. In some embodiments, the barrier 30 may be formed of a plurality of discrete barrier members 32. In other embodiments, the barrier 30 may be formed of a single unitary barrier member. The barrier 30 may be at least partially comprised of a metallic material, glass, a non-porous material, or a specialised synthetic material or polymer.

[0285] The material of the barrier 30 preferably does not retain odours or must be completely sanitisable to remove any substances, stimuli or odours, e.g. by UV light, boiling water, solvents or heating (e.g. using autoclaves). In one embodiment, at least a portion, preferably all, of the apparatus of the invention is odour sterilisable. In one embodiment, at least the material of the barrier is odour sterilisable. Preferably, the barrier comprises or is constructed of an odour sterilisable material selected from glass, stainless steel or aluminium.

[0286] The barrier 30 may be formed of a plurality of discrete barrier members 32 which are configured to separately retain organisms within a subset of the plurality of chambers 20. In some embodiments, the subset may be a single “row” of chambers 20, or any number of rows of chambers 20. In the pictured embodiment, each row of chambers 20 is closed by the barrier 30 which is formed of a plurality of long, single-row barrier members 32 (e.g. sliding keys). Each single-row barrier member 32 takes the form of a strip having a plurality of apertures 34 along its length so that the apertures 34 are spaced apart from each other by blocking sections.

[0287] The individual barrier members 32 are controlled by an actuator 40. The barrier members 32 may be permanently or detachably coupled to the actuator 40. The actuator 40 controls the barrier 30 to selectively retain and selectively release the organisms in the plurality of

[0288] 32 chambers 20. The actuator 40 controls the movement of the barrier 30 from a closed configuration to an open configuration, and from an open configuration to a closed configuration. The barrier 30 permits release of the organisms from the chambers 20 when the barrier 30 is configured to align the apertures 34 with the plurality of chambers 20 and thereby open the chambers 20 (i.e. the open configuration). The barrier 30 retains the organisms within the chambers 20 when the barrier 30 is configured to align the blocking sections with the plurality of chambers 20 and thereby close the chambers 20 (i.e. the closed configuration).

[0289] In the pictured embodiment, the actuator 40 is a linear actuator which is configured to synchronously control all barrier members 32 in the barrier 30 to open all chambers 20 simultaneously. In one embodiment, the actuator is configured to synchronously control all barrier members 32 in the barrier 30 to open all chambers 20 within less than 5 seconds or in milliseconds. Preferably, 1 , 2, 3 or 4 seconds, more preferably less than 1000 milliseconds, most preferably 100-500 milliseconds. In one embodiment, the actuator comprises a sliding-barrier mechanism that releases all insects within milliseconds of each to other. This facilitates synchronised exposure to the substance and recording of behavioural responses.

[0290] A guide rail may be employed to guide the linear movement of the actuator 40. Guide rails 36 may be employed to aid the linear sliding movement of the barrier members 32 as controlled by the actuator 40. In other embodiments, the actuator 40 may control the barrier 30 to open some chambers 20 non-synchronously. For example, the actuator 40 may be configured to control a number of barrier members 32 to open the chambers 20 “row by row”, or in specified regions. In yet other embodiments, there may be a further actuator or actuators to control different barrier members 32.

[0291] In some embodiments, the actuator 40 may be controlled remotely by a programmable controller and / or a human operator. The remote control prevents the human operator from directly affecting the organisms inside the chambers 20, e.g. by odour contamination. The programmable controller may be programmed or configured to control the actuator 40 to selectively open or close the barrier 30 after a predetermined time or in response to a predetermined signal or command. In other embodiments a human operator may use their judgment to control the actuator 40 to selectively open or close the barrier 30.

[0292] A sample retention device 50 may be included in some embodiments. The sample retention device 50 may be comprised of a single unitary piece or multiple pieces.

[0293] 33 An embodiment of the sample retention device 50 is shown in Figures 3 and 4. In the pictured embodiment, the sample retention device 50 is a sample clamping device comprising first and second clamping member support structures 52, 54. Both clamping member support structures 52, 54 are in the form of plates respectively. The second clamping member support structure 54 further includes a receptacle surrounded by a frame wall, whereby the receptacle is shaped to receive the first clamping member support structure 52.

[0294] In some aspects, the sample retention device 50 (or specifically the second clamping member support structure 54) may be configured to be removably attached to a casing 56 storing the plurality of chambers 20. The attachment may be performed by an external clamping device or by a clamping device which is part of the sample retention device 50. In some embodiments the attachment may be performed by a plurality of removable screws or a different form of fastener or securing mechanism. In the pictured embodiment, the second clamping member support structure 54 is attached to the casing 56 by way of a plurality of screws 56.

[0295] In use, a sample 58 is placed on the second clamping member support structure 54, as shown in Figure 3. The first clamping member support structure 52 is lowered onto the second clamping member support structure 54 so as to clamp the sample 58 between the first and second clamping member support structures 52, 54, as shown in Figure 4. The first clamping member support structure 52 may be held in place simply by the weight of the first clamping member support structure 52, or may be held in place either by complementary fit with the second clamping member support structure 54 or by another fastening or securing mechanism.

[0296] The sample retention device 50 may include a plurality of clamping members, each corresponding to a respective one of the plurality of chambers 20 and configured to position the sample 58 against or towards the corresponding chamber 20. The clamping members may be in the form of projections. The clamping members may form part of or be attached to the first clamping member support structure 52 and may be configured to position and retain the stimulus-generating substance against openings in the second clamping member support structure 54 corresponding to the chambers 20.

[0297] The barrier 30 in its closed configuration separates the organisms within the chambers 20 from the first and second clamping member support structures 52, 54 and the clamped

[0298] 34 sample 58. The barrier 30 in its open configuration provides the organisms within the chambers 20 with access to the first and second clamping member support structures 52, 54 and the clamped sample 58. In this way, when the apertures 34 of the barrier members 32 are aligned with the plurality of chambers 20, the organisms in the chambers 20 are exposed to the clamped sample 58 via the openings in the second clamping member support structure 54.

[0299] Preferably, in the open configuration of the barrier 30, the barrier 30 and the first and second clamping member support structures 52, 54 are engaged with each other so as to be in sealing cooperation with each other and thereby create sealed environments that include the plurality of chambers 20. This beneficially creates environments for at least part of the sample 58 and the organisms that is sealed against external stimulus such as unwanted odours. In other embodiments, in the open configuration of the barrier 30, the barrier 30 and the first and second clamping member support structures 52, 54 may be engaged with each other to physically contain at least part of the sample 58 and organisms within an enclosed environment, without necessarily being in sealing cooperation with each other.

[0300] The clamping members may be transparent. When the barrier 30 is in the open configuration, the transparent clamping members and the openings in the second clamping member support structure 54 permit light from a light source to pass through the chambers 20, the openings in the second clamping member support structure 54, the sample 58, and the transparent clamping members in that order. A monitoring device 70 may be arranged to receive the light that has passed through the chambers 20, the openings and the transparent clamping members. Further details of the monitoring device 70 are described elsewhere in this specification.

[0301] Components of the sample retention device 50 may be wholly or partially comprised of a metallic material, glass, a non-porous material, or a specialised synthetic material or polymer. The material of the components of the sample retention device 50 preferably does not retain odours or must be completely sanitisable to remove any substances, stimuli or odours, e.g. by UV light, boiling water, solvents or heating (e.g. using autoclaves). In one embodiment, at least a portion, preferably all, of the apparatus of the invention are odour sterilisable. In one embodiment, at least the sample retention device is odour sterilisable.

[0302] 35 The light source 60 may be located beneath the chambers 20 to provide backlighting through a glass panel 90. Alternatively the light source 60 may be located to the side or in another location and be configured to illuminate the chambers 20 from below, for example and not limitation by directing light beneath the chambers 20. The monitoring apparatus may further include one or more mirrors positioned to redirect the light to illuminate the chambers 20, or the light may be redirected in another manner. The light source 60 may be configured to illuminate all or some of the chambers 20.

[0303] The organism to be monitored by the monitoring apparatus may be an insect, an agricultural pest or an arthropod. The monitoring apparatus may be suitable for use with other insects or organisms and especially for monitoring the response of agricultural pests to a stimulus. The stimulus may be, for example and not limitation, a sample in the form of a leaf (which may be untreated), a leaf coated in or treated with a stimulus-generating substance, or a stimulus-generating substance. The stimulus-generating substance may be an odour-generating substance. The stimulus-generating substance may be a powder or liquid and may be coated on or otherwise applied to a leaf. In some aspects, a whole uncut leaf by itself, or the whole leaf coated in or treated with the stimulus-generating substance, may be retained by the sample retention device 50 for exposure to the organisms. It will be appreciated that the reference to the leaf is to help illustrate the working of the invention, and that the invention is applicable to other types of stimulusgenerating substances, such as or including at least one crop material, at least one insect target, at least one microbe, at least one microbial extract, at least one natural product, at least one metabolite, at least one metabolite mixture, at least one chemical active, at last one chemical compound and / or at least one biological material.

[0304] In the pictured embodiment in Figure 5, the monitoring apparatus 70 is located above the sample retention device 50 and arranged to face downwards towards the first clamping member support structure 50 so as to enable the monitoring apparatus 70 to receive the light that has passed through the chambers 20, the openings in the second clamping member support structure 54, the sample 58, and the transparent clamping members in that order. In other embodiments, the monitoring apparatus may further include one or more mirrors positioned to redirect the light to the monitoring apparatus, or the light may be redirected in another manner.

[0305] The monitoring device 70 may be or include a camera, an imaging device, or another lightsensitive sensor. The monitoring device 70 may further include a programmable controller which may be pre-programmed by a user or manufacturer to control the monitoring device

[0306] 36 70. The monitoring device 70 may be configured to begin monitoring the behaviour or movement of the organisms upon detection of a predetermined cue such as, for example and not limitation, a user pressing a key or button, the light source 60 switching to an “on” state, light from the light source 60 being received by the monitoring device 70, a threshold amount of light being received by the monitoring device 70, a predetermined time limit elapsing, the activation of the actuator 40 or the opening of the barriers 30. The predetermined cue may be a combination of the aforementioned examples or another cue.

[0307] The monitoring device 70 may be configured to detect movement of the organisms, for example by way of a detected change in the light (e.g. change in cast shadow) detected by the monitoring device 70. In this way the monitoring device 70 may detect whether the organisms are repelled by the stimulus, attracted to the stimulus, are consuming the sample or have another behavioural reaction to the stimulus. In some embodiments the monitoring device 70 may be configured to transmit data for analysis (using wired or wireless connection) and / or store the data using an onboard memory. In further embodiments the monitoring device 70 may transmit the data in a “live feed” or to be viewed in real time by the user. The data may include still image data and / or video data. The data may include time stamps and / or date stamps.

[0308] The monitoring apparatus may further comprise a housing 100 inside which the other components including the plurality of chambers 20, the barrier 30, the actuator 40, the sample retention device 50, the light source 60 and the monitoring device 70 are housed. In further embodiments, the light source 60 and / or the monitoring device 70 may be located outside the case or enclosure. The housing 100 may be sealable to be, for example, lighttight, odour-tight and / or stimulus-tight. The housing 100 may insulate the organisms and sample from the external environment to ensure that the observed behaviour by the organisms is a result of exposure to the target sample, stimulus or stimulus-generating substance, and not any other sample, stimulus or stimulus-generating substance. An exemplary embodiment of a monitoring apparatus inside a sealable housing 100 is shown in Figures 5 and 6. A cutout of an embodiment of a monitoring apparatus including a sealable housing 100 is shown in Figure 5 to illustrate exemplary locations of the plurality of chambers 20, the barrier 30, the actuator 40, the sample retention device 50, the light source 60 and the monitoring device 70. Figure 6 shows the exterior of the housing 100, which may include a display screen 102 for showing still images or videos transmitted from the monitoring device 70.

[0309] An exemplary method of using the monitoring apparatus comprises the following steps:

[0310] 37 a) Using the actuator 40 or manually, slide the barrier members 32 to open the chambers 20, and load organisms into the opened chambers 20. The chamber opening and the organism loading can be carried out simultaneously or row- by-row. b) Using the actuator 40 or manually, slide the barrier members 32 to close the chambers 20 so as to retain the organisms within the chambers 20. c) Place the second clamping member support structure 54 over the casing 56 and use the screws to attach the second clamping member support structure 54 to the casing 56. d) Place the sample 58 over a selection of the openings of the second clamping member support structure 54. e) Use the first clamping member support structure 52 to clamp the sample 58 between the first and second clamping member support structures 52, 54. f) Turning on the light source 60 to apply backlighting to the chambers 20. g) Using the actuator 40, slide the barrier members to synchronously open some or all of the chambers 20, thereby releasing the organisms from the opened chambers 20 and exposing the organisms to the sample 58. h) Using the monitoring device 70, monitoring a behaviour of the organisms exposed to the sample 50. Each chamber 20 will be seen as a lit circle, with each organism being observed as a shadow. The movement of the shadow is recorded by the monitoring device 70. i) After completing the monitoring step, removing the organisms from the chambers 20 by performing steps a) to e) in reverse.

[0311] The foregoing features of the monitoring apparatus result in a standalone platform with a high throughput screening configuration for monitoring and tracking the behaviour of live insects exposed to a stimulus in a controlled environment. The features of the monitoring apparatus, particularly the controlled actuation of the barrier members 32 to synchronously release the insects, enable high levels of repeatability which is useful for producing reliable behavioural and efficacy data collection for high numbers of insects exposed to same or different types of stimuli. This is in contrast to manually exposing the insects to a stimulus, which is prohibitively slow and lacks controllability over exposure timing and duration.

[0312] In addition to the high throughput screening configuration described above, the monitoring apparatus of the invention may be reconfigured into one or more alternative screening configurations. The or each alternative screening configuration allows for lower throughput screening, which may be suitable for rapid early screening (especially if high numbers of

[0313] 38 insects are not required) or for running different monitoring experiments (e.g. using larger or smaller organisms, different types of organisms, different types of stimulus, different environments). This is particularly advantageous due to the lower setup time and the shorter loading times. As the throughput is lower (i.e. the number of insects tested per run is lower), synchronised exposure to the stimulus in the same rune is less critical because a time difference between the first insect and last insect being loaded will be more acceptable.

[0314] Thereafter, the high throughput screening configuration may be used to confirm the results of the rapid early screening by obtaining reliable statistical data using a high number of insects.

[0315] Non-limiting examples of alternative screening configurations are described as follows. The barrier 30, the actuator 40 and the sample retention device 50 are not essential for the alternative screening configurations.

[0316] Figures 7A to 7F show a deconstruction sequence of a monitoring apparatus according to an exemplary embodiment of the invention, as follows:

[0317] 1) [Figure 7A] The first clamping member support structure 52 is removed from the second clamping member support structure 54.

[0318] 2) [Figure 7A] The screws are removed, and the second clamping member support structure 54 is removed from the casing 56.

[0319] 3) [Figure 7C] The barrier members 32 are decoupled from the actuator 40.

[0320] 4) [Figure 7D] The barrier members 32 are removed from the casing 56.

[0321] 5) [Figure 7E] The guide rails 36 for the barrier members 32 are removed.

[0322] 6) [Figure 7F] The well assembly defining the chambers 20 are removed from the casing 56, leaving the backlighting glass panel 90 exposed.

[0323] Following the deconstruction sequence, the monitoring apparatus is then reconfigured to an alternative screening configuration.

[0324] A first alternative screening configuration based on removable well assemblies 110 is shown in Figures 8A to 8C. In the pictured embodiments, each well assembly 110 includes 6 wells, but it is envisaged that other numbers of wells may be used.

[0325] In Figure 8A, a light blocking mask 112 is placed over the glass panel 90. The light blocking mask 112 has an opaque region that acts to block part of an illumination light provided by

[0326] 39 the light source 60, and has openings that permit the transmission of light therethrough. In Figure 8B, well assemblies 110 (such as well plates) are placed over the openings in the light blocking mask 112. Figure 8C shows four well assemblies 110 in situ, completing the alternative screening configuration. Each well 110 contains a sample and an organism. The well assemblies 110 may be pre-loaded with the organisms, or the organisms may be loaded into the well assemblies 110 after the alternative screening configuration is set up. Preferably each well 110 is larger than the wells 20 used in the high throughput screening configuration. The monitoring device 70 detects the transmitted light and thereby detects the shadows cast by the organisms within the wells 110.

[0327] The first alternative screening configuration permits the use of standardised well assemblies. Furthermore, different well assemblies can be used to allow for a range of different insect retention environments. The different well assemblies may have, for example, different types of wells, different numbers of wells, different sizes of wells and / or different footprints of well assemblies.

[0328] A second alternative screening configuration based on tubular conduits is shown in Figures 9A to 9D. Each tubular conduit is in the form of a specimen tube 120.

[0329] In Figure 9A, the specimen tubes 120 are placed over the glass panel 90. The light blocking mask 112 act to block part of an illumination light provided by the light source 60, as shown in Figures 9A and 9B. The tubes 120 are placed over the openings in the light blocking mask 112. In Figure 9C, a sample holder is arranged at the top of each tube 120. The sample holder comprises two parafilm membrane layers 122 separated by a spacing member 124 in the form of an O-ring. A space between the two parafilm membrane layers 122 contains a liquid active, or another type of solid or liquid stimulus-generating substance. An organism is loaded into the tube 120 from the bottom of the tube. In Figure 9D, the organism has reached the top of the tube 120 and feeds on the liquid active through the lower parafilm membrane layer 122. The monitoring device 70 detects the transmitted light through the membrane layers 122 and thereby detects the shadows cast by the organisms within the tubes 120.

[0330] Furthermore, two sample holders may be arranged at respective top and bottom ends of each tube 120, thus enabling the testing of two stimulus-generating substance in the same run in a controlled way. The two stimulus-generating substances may be the same or different.

[0331] 40 Conventionally insect screening is carried out manually, which is time-consuming and has low throughput. Such screening is based on the monitoring of a single end point, involves the use of subjective scoring, and has poor temporal resolution. As a result, the conventional insect screening provides limited information on a mode of action of a given substance, which is defined as a functional or anatomical change to an organism.

[0332] In a preferred method of monitoring organism behaviour, one or more organisms (preferably insects, pests or arthropods) is exposed to a stimulus-generating substance, for example, by using any of the aforementioned screening configurations. A video (or a plurality of videos recorded at different times) of a response of the or each organism to a stimulus provided by the stimulus-generating substance is recorded using a video monitoring device. Using the recorded video, it becomes possible to compare the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism; and identify one or more mode of actions of the stimulus-generating substance based on the comparison of the response to the multiple objective endpoints. Such comparison and identification are performed with the assistance of a processor and, optionally, an artificial intelligence algorithm or model. It is envisaged that the video monitoring device may be replaced by an image monitoring device.

[0333] Phenotypes observable by the method of the invention include:

[0334] • Mortality (including mortality rates);

[0335] • Motility, mobility, motor function quality and / or paralysis (including paralysis progression);

[0336] • Deterrence, attraction, seeking and / or avoidance;

[0337] • Feeding behaviour and / or feeding inhibition;

[0338] • Morphological change, such as colour change (e.g. pigmentation change, melanisation pattern) and / or size change;

[0339] • Spatial organisation;

[0340] • Fecundity;

[0341] • Cleaning;

[0342] • Agitation;

[0343] • Growth inhibition and / or developmental halting.

[0344] Accordingly, a phenotype may be selected from one or more of: mortality; motility, mobility, motor function quality and / or paralysis; deterrence, attraction, seeking and / or avoidance;

[0345] 41 feeding behaviour and / or feeding inhibition; morphological change, such as colour change and / or size change; spatial organisation; fecundity; cleaning; agitation; and growth inhibition and / or developmental halting.

[0346] Morphological change is defined as the physical alteration in organism appearance or structure. This can provide an indication of, for example, metabolic disruption, molting issues, or physiological stress. This provides an early indicator of sublethal effects or insect pest modifying behaviour before mortality.

[0347] Paralysis is defined as loss of voluntary movement ability, either partial or complete. This can provide an indicator of, for example, neurotoxic effects or muscle function disruption. This is a critical safety measure as paralysis often precedes mortality.

[0348] Mortality is defined as the death of the organism. This provides, for example, a measure of acute toxicity.

[0349] Developmental halting is defined as interruption of normal life cycle progression. This can provide an indication of, for example, interference with molting, metamorphosis, or growth. This provides long-term population control without acute toxicity.

[0350] Feeding behaviour is defined as changes in food consumption patterns or feeding cessation. This can provide an indication of, for example, appetite suppression or inability to feed. This is critical to crop protection without the involvement of direct mortality. For example, feeding behaviour may include food consumption patterns (such as consumption volume, duration, frequency), or the proboscis extension response (PER).

[0351] Spatial organization relates to how insects position themselves relative to each other and environment. This can indicate, for example, changes in social behaviour or environmental response. This can reveal repellent or attractant effects.

[0352] Fecundity is defined as reproductive capacity and egg-laying behaviour. This can indicate, for example, the impact on population growth potential. This can provide long-term pest control through reproductive suppression.

[0353] Seeking is defined as active search behaviour for food, mates, or habitat. This can show, for example, changes in motivation or ability to locate resources. This can reveal disruption of essential behaviours.

[0354] 42 Avoidance is defined as active movement away from specific stimuli or areas. This can indicate, for example, recognition and rejection of treated areas. This is key for repellentbased crop protection strategies.

[0355] Motor function quality is defined as quality and coordination of movement. This can indicate, for example, neuromuscular system impacts. This can reveal subtle effects missed by mortality screening.

[0356] Cleaning is defined as grooming and maintenance behaviours. This can indicate, for example, changes in basic behavioural patterns. This provides an early indicator of neurological or physiological stress. Accordingly, in one embodiment, the phenotype is grooming and maintenance behaviours.

[0357] Agitation is defined as increased movement or erratic behaviour. This can indicate, for example, nervous system disruption or stress response. This provides an early warning of insecticide effects.

[0358] Other phenotypes include developmental transitions, behavioural patterns, time-to-effect and dose response curves.

[0359] Non-limiting examples of modes of action identifiable by the invention include:

[0360] 1) Nervous system disruption

[0361] Action: Interference with the insect's nervous system, leading to paralysis or hyperactivity, followed by death.

[0362] Examples of substances related to this mode of action include:

[0363] • Acetylcholinesterase inhibitors: Organophosphates (e.g., malathion) and carbamates (e.g., carbaryl);

[0364] • Sodium channel modulators: Pyrethroids (e.g., permethrin);

[0365] • Nicotinic acetylcholine receptor (nAChR) agonists / antagonists: Neonicotinoids (e.g., imidacloprid).

[0366] 2) Muscle Function Disruption

[0367] Action: Targets the muscles, preventing contraction or causing excessive contraction.

[0368] Examples of substances related to this mode of action include:

[0369] • Ryanodine receptor modulators: Diamides (e.g., chlorantraniliprole).

[0370] 43 3) Growth and Development Interference

[0371] Action: Disruption of normal insect growth or molting processes.

[0372] Examples of substances related to this mode of action include:

[0373] • Chitin synthesis inhibitors: Benzoylureas (e.g., diflubenzuron);

[0374] • Juvenile hormone analogs: Pyriproxyfen (used in mosquito control).

[0375] 4) Energy Metabolism Inhibition

[0376] Action: Disruption of energy production, affecting cellular respiration.

[0377] Examples of substances related to this mode of action include:

[0378] • Mitochondrial electron transport inhibitors: Rotenone;

[0379] • ATP synthase inhibitors: Spirotetramat.

[0380] 5) Feeding Inhibition

[0381] Action: Stops the insect from feeding by affecting behaviour or physiology.

[0382] Examples of substances related to this mode of action include:

[0383] • Antifeedants: Azadirachtin (from neem).

[0384] 6) Gut Disruption

[0385] Action: Damage to the insect's gut lining, leading to starvation or septicaemia.

[0386] Examples of substances related to this mode of action include:

[0387] • Bacterial toxins: Bacillus thuringiensis (Bt) toxins.

[0388] • Toxins from spinosyns: Spinosad.

[0389] 7) Physical Disruption

[0390] Action: Physically harms or suffocates the insect.

[0391] Examples of substances related to this mode of action include:

[0392] • Oils and soaps: Disrupt the waxy cuticle.

[0393] • Desiccants: Diatomaceous earth.

[0394] The invention is applicable to a wide range of organisms, including insects, pests and arthropods.

[0395] The comparison of the observed response to multiple objective endpoints facilitates the generation of more data that enables not only accurate identification of the mode(s) of action of the stimulus-generating substance but also new modes of action that cannot be identified through traditional screening, and thereby provides more information on subtle

[0396] 44 changes in insect behaviour that cannot be detected by the conventional insect screening method. For example, a combination of observed behaviours would be indicative of a particular type of mode of action, while another combination of observed behaviours would be indicative of another type of mode of action. Such data can be used to map out a response of a given organism in terms of different phenotypes and / or different types of substances. An exemplary map is illustrated in Figure 10 that shows the degree of response of microbes with respect to various phenotypes. This level of information is not available by simply monitoring a single objective endpoint and identifying a single mode of action which conventionally is mortality.

[0397] A machine learning algorithm or model may be trained to improve the identification of at least one mode of action of a stimulus-generating substance. The training may involve the following steps:

[0398] • collecting a set of data that includes the identified at least one mode of action of the stimulus-generating substance;

[0399] • creating a training set including the collected set of data;

[0400] • training a machine learning algorithm or model using the training set; and identifying the at least one mode of action of the stimulus-generating substance based on the comparison of the response to the multiple objective endpoints based on an output of the machine learning algorithm or model.

[0401] Preferences and options for a given aspect, feature or parameter of the invention should, unless the context indicates otherwise, be regarded as having been disclosed in combination with any and all preferences and options for all other aspects, features and parameters of the invention.

[0402] The listing or discussion of an apparently prior published document or apparently published information in this specification should not necessarily be taken as an acknowledgement that the document is part of the state of the art or is common general knowledge.

[0403] In a non-limiting example, the invention may be encompassed by the following statements:

[0404] 1. A method of monitoring organism behaviour, the method comprising the steps of: exposing at least one organism to a substance; and observing a response of the at least one organism to the substance;

[0405] 45 comparing the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism; and identifying at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints.

[0406] 2. A method according to statement 1 including the step of recording a video or image of the response of the at least one organism to the substance.

[0407] 3. A method according to statement 1 or 2 including the step of recording a plurality of videos or images of the response of the at least one organism to the substance, wherein each video or image is recorded at a respective different time.

[0408] 4. A computer-implemented method of monitoring organism behaviour, the method comprising the step of: using a recording of a response of at least one organism to a substance, comparing the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism; and identifying at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints.

[0409] 5. A method according to any one of statements 1 to 4 wherein the at least one phenotype includes mortality.

[0410] 6. A method according to any one of statements 1 to 5 wherein the at least one phenotype includes motility, mobility, motor function quality and / or paralysis.

[0411] 7. A method according to any one of one of statements 1 to 6 wherein the at least one phenotype includes deterrence, attraction, seeking and / or avoidance.

[0412] 8. A method according to any one of statements 1 to 7 wherein the at least one phenotype includes feeding behaviour and / or feeding inhibition.

[0413] 9. A method according to any one of one of statements 1 to 8 wherein the at least one phenotype includes morphological change.

[0414] 46 10. A method according to statement 9 wherein the morphological change includes colour change and / or size change.

[0415] 11. A method according to any one of the preceding statements wherein the at least one phenotype includes spatial organisation.

[0416] 12. A method according to any one of the preceding statements wherein the at least one phenotype includes fecundity.

[0417] 13. A method according to any one of the preceding statements wherein the at least one phenotype includes cleaning.

[0418] 14. A method according to any one of the preceding statements wherein the at least one phenotype includes agitation.

[0419] 15. A method according to any one of the preceding statements wherein the at least one phenotype includes growth inhibition and / or developmental halting.

[0420] 16. A method according to any one of the preceding statements including the step of continuously observing a response of the at least one organism to the stimulus provided by the substance.

[0421] 17. A method according to any one of the preceding statements including the step of using an artificial intelligence algorithm or model to analyse the observed response of the at least one organism to the stimulus provided by the substance in order to compare the response to the multiple objective endpoints and identify the at least one mode of action of the substance based on the comparison of the response to the multiple objective endpoints.

[0422] 18. A computer-implemented method of identifying at least one mode of action of a substance, the method comprising the steps of: collecting a set of data by carrying out the method according to any one of the preceding statements, wherein the collected set of data includes the identified at least one mode of action of the substance associated with the multiple objective endpoints; creating a training set including the collected set of data; training a machine learning algorithm or model using the training set; and identifying the at least one mode of action of the at least one organism based on the

[0423] 47 comparison of the response to the multiple objective endpoints based on an output of the machine learning algorithm or model.

[0424] 19. A computer program comprising computer code configured to perform the method of any one of statements 4 to 18.

[0425] 20. A method according to any one of the preceding statements wherein the at least one organism is an insect, a pest or an arthropod.

[0426] 21. A monitoring apparatus for monitoring organism behaviour, the monitoring apparatus comprising: at least one chamber configured for housing at least one organism, the or each chamber configurable to selectively expose the at least one organism to a substance; a monitoring device configured to, in use, record a response of the at least one organism to a stimulus provided by the substance; a processor configured to, using the recording of the response, compare the response to multiple objective endpoints, wherein the multiple objective endpoints are or correspond to respective phenotypes of the at least one organism, and identifying at least one mode of action of the substance based on the comparison of the response to one or more of the multiple objective endpoints.

[0427] Examples

[0428] The invention is exemplified in non-limiting examples:

[0429] Example 1 - Behaviour analysis of Drosophila melanogaster larvae exposed to insecticidal chemistries isolates

[0430] We have determined that insects show distinct activity patterns following exposure to different insecticidal chemistries that are indicative of the chemistry’s mode of action. Accordingly, it is possible to screen for insecticidal substances and identify their mode of action by observing an insect’s behavioural response following exposure to a substance. It is further possible to identify new modes of action of a substance, or identify the mode of action for a new substance. The behavioural responses can be observed using the apparatus disclosed herein.

[0431] 48 This example provides proof-of-concept data to support that insects show distinct activity patterns following exposure to different insecticidal chemistries that are indicative of the chemistry’s mode of action. In particular, this example relates to behavioural analysis of Drosophila melanogaster larvae exposed to insecticidal chemistries isolates.

[0432] Twenty-five larvae were placed per observed in a multi-chamber housing unit, with at least three biological replicates per treatment and observed using the apparatus and methods disclosed herein. The insects were exposed to four substances, namely four reference chemistries from distinct IRAC (Insecticide Resistance Action Committee) classes, tested independently against a negative control. The method quantified the proportion of time each larva spent performing specific behaviours at defined time points from short video recordings (a few minutes in duration). The seven tracked behaviours are summarised below.

[0433] These behaviours capture a range of neuromuscular outcomes, from excitation and aversion to paralysis, enabling inference of mode of action (MoA) signatures.

[0434] Each chemistry therefore represents a distinct neuromuscular target and is expected to generate a unique behavioural signature. Initial exploratory plots comparing mean behavioural proportions over time revealed clear differences between treatments and controls.

[0435] Figure 11 shows that Acetamiprid (green) and Carbaryl (orange) induced early increases in bend and back-up behaviours, followed by a progressive rise in stop events, consistent with cholinergic overstimulation and subsequent paralysis. Coragen exposure (purple) produced gradual increases in hunch behaviour, reflecting tonic muscle contraction linked to disrupted calcium release. Spinosad (pink) caused pronounced roll and small-action spikes prior to immobilisation, aligning with its mixed nicotinic / GABAergic hyperexcitation profile.

[0436] Figure 12, showing the normalised results for each treatment, reveals that each chemistry (reflected by colour) shows distinct temporal fingerprints consistent with their known modes of action (e.g. rapid excitation for Acetamiprid and Carbaryl; tonic contraction for Coragen; hyperactive rolling for Spinosad).

[0437] Figure 13 shows a collection of radar plots, which illustrate the mean proportional time spent in each behaviour for each substance. The unique shape of each radar plot - as well as the collection of the radar plots - reflects the different behaviours induced by each substance. The differences in behavioural responses can be attributed to and is indicative of the different modes of actions of each substance.

[0438] Thus, behavioural phenotyping of D. melanogaster larvae successfully distinguished four insecticidal chemistries with contrasting I RAC classifications. Each substance produced a unique temporal and proportional pattern of movement consistent with its known neurophysiological target. The approach demonstrates the potential for rapid mode-of- action profiling of known chemistries with different I RAC modes of action which may allow for novel microbial candidates to be screened and classified as similar or different to known mode of actions.

[0439] Example 3 - Behaviour analysis of Drosophila melanogaster larvae exposed to microbial isolates

[0440] It is known that some microbes per se or microbial products have insecticidal activity. However, the modes of action through which these microbes and microbial products exert their insecticidal effects are yet to be fully determined. Further, there are presently inadequate methods to screen for new microbes / microbial products with insecticidal activity.

[0441] We demonstrate in this Example that the apparatus and methods disclosed herein can be used to resolve these issues to identify whether microbes or their products have insecticidal activity, the degree of their insecticidal activity and determine the insecticidal mode of action.

[0442] The methods of Example 2 were replicated to expose a plurality of Drosophila melanogaster larvae to microbial isolates, to determine if bacterial exposure elicits behavioural responses consistent with insecticidal effects. In particular, each treatment consisted of 25 larvae per chamber, with at least two biological replicates per isolate. Both non-insecticidal and insecticidal bacterial treatments were tested alongside negative

[0443] 51 controls. The pipeline quantified time spent in seven behavioural states: back-up, bend, crawl, hunch, roll, small action, and stop (defined above).

[0444] Figure 14 shows that across treatments there were clear differences in behavioural responses between insects that were not exposed to bacterial isolates (control) bacterial isolate), exposed to non-insecticidal bacterial isolates and exposed to insecticidal bacterial isolates. It was unknown whether the BB990 isolate of Bacillus sp. and microbe X were insecticidal or not.

[0445] Larvae not exposed to a bacterial isolate (control; grey line) maintained stable crawling and bending activity with minimal immobility, reflecting a healthy locomotor baseline. Insects exposed to isolates of Pseudomonas sp, (BB1008) and Curtobacterium sp. (BB1039) followed broadly similar trajectories close to parity within controls, with only minor fluctuations in bending and small-action behaviours, consistent with mild irritancy rather than acute toxicity. Therefore, it was possible to deduce that Pseudomonas sp, (BB1008) and Curtobacterium sp. (BB1039) have no or low insecticidal activity at this concentration.

[0446] In contrast, exposure to the unknown microbe X, BASF bacterial isolate from an insecticidal product and the BB990 Bacillus sp triggered clear reductions in crawl frequency alongside marked increases in stop behaviour over time. This indicative of neuromuscular disruption and potential lethality. In particular, early increases in bend and small-action ratios for BASF X and microbe X suggest transient hyperactivity prior to paralysis. In contrast, BB990 showed marked suppression of crawl and bend behaviours from the outset, consistent with rapid neuromuscular failure seen in the chemistry previously analysed. Notably, insects exposed to BASF X and BB990 show late-stage surges in the stop ratio reaching more than tenfold above control levels. This observation confirms near-complete immobilisation and early lethality in these treatments. This is consistent with the known insecticidal activity of BASF X and suggests that BB990 isolate and microbe X are also insecticidal strains. Further, based on the pattern of behaviour it is possible to deduce that both BB990 and microbe X has a mode of action impacting the neural system with a signature of transient hyperactivity prior to paralysis.

[0447] Although the comparisons were conducted using a reference behavioural fingerprint obtained experimentally, this could be completed using an online reference or by using a statistical or machine-learning model or computer-implemented system to identify a mode of action.

[0448] 52

Claims

53CLAIMS1. A method of determining an insecticidal mode of action of at least one substance comprising: a. exposing at least one insect to the at least one substance; b. obtaining behavioural response data characterising a response of the at least one insect to the at least one substance; c. processing the behavioural response data to identify one or more behavioural responses, wherein the one or more behavioural responses are or correspond to one or more respective phenotypes of the at least one insect; d. generating a behavioural fingerprint based on the one or more behavioural responses; e. comparing the behavioural fingerprint to a library reference; and f. identifying the corresponding insecticidal mode of action of the at least one substance.

2. A method of screening at least one substance for pesticide activity comprising: a. exposing at least one insect to at least one substance; b. obtaining behavioural response data characterising a response of the insect to the at least one substance; c. processing the behavioural response data to identify one or more behavioural responses, wherein the behavioural responses are or correspond to one or more respective phenotypes of the at least one insect; d. generating a behavioural fingerprint based on the one or more behavioural responses; e. comparing the behavioural fingerprint to a library reference; and f. identifying if the substance has insecticidal activity.

3. The method of claim 1 or 2, wherein the behavioural response data comprises image or video data.53544. The method of any preceding claim, wherein the method comprises exposing a plurality of insects to the at least one substance in parallel.

5. The method of any preceding claim, wherein the method comprises exposing a plurality of insects to a plurality of substances.

6. The method of any preceding claim, wherein step (e) of the method comprises comparing the behavioural fingerprint to a library reference using a statistical or machine-learning model.

7. The method of any preceding claim, wherein the library reference is derived from known insecticidal compounds or microbes, microbial cultures, microbial lysates, microbial metabolite or microbial extract.

8. The method of any preceding claim further comprising using a statistical or machine-learning model to determine a probabilistic assignment of a mode of action.

9. The method of claim 8 wherein the statistical or machine-learning model is trained on datasets comprising behavioural fingerprints of chemical insecticides with known I RAC mode-of-action classifications and microbes, microbial cultures, microbial lysates, microbial metabolites or microbial extracts with experimentally confirmed activity, optionally wherein the datasets are labelled datasets.

10. The method of claim 8 or 9 wherein the statistical or machine-learning model outputs a probability distribution across known modes of action and classifies a behavioural fingerprint as novel when the known modes of actions do not exceed a similarity threshold.11 . The method of any preceding claim wherein the behavioural fingerprint is stored as a vector of normalised quantitative values.

12. The method of any preceding claim wherein the exposure step comprises simultaneous release of a plurality of insects to synchronise behavioural exposure to the substance.545513. The method of any preceding claim wherein the substance comprises at least one microbe, a microbial culture, a microbial lysate, a microbial metabolite or microbial extract.

14. The method of any preceding claim further comprising integrating the behavioural fingerprint with metabolic or genomic data of a microbial source.

15. An apparatus for monitoring and analysing insect behaviour in response to a substance, comprising: a. a multi-chamber housing configured to contain at least one or a plurality of insect(s); b. a barrier assembly and actuator configured to synchronously open the chambers to expose the at least one or a plurality of insect(s) to the substance; c. one or more illumination sources and imaging devices arranged to capture time- resolved video of the at least one or a plurality of insect(s) from above and / or below; and d. a processor configured to extract quantitative behavioural responses from the recorded video.

16. The apparatus of claim 15, further comprising a memory storing instructions executable by the processor to compute a behavioural fingerprint from the behavioural responses and compare it to reference behavioural fingerprints, preferably using a trained machine-learning model or algorithm.

17. The apparatus of claim 16, wherein the trained model or algorithm is used to determine an insecticidal mode of action of the substance.

18. The apparatus of any of claims 15 to 17. wherein the multi-chamber housing is constructed of odour-sterilisable materials selected from glass, stainless steel, or aluminium and is sealed to prevent volatile cross-contamination between chambers.

19. The apparatus of claim 15 to 18, wherein the actuator comprises a sliding-barrier mechanism that releases all insects within milliseconds of each other to enable time-synchronised recording.555620. The apparatus of any of claims 15 to 19, wherein the imaging device and illumination system are arranged in sealed optical alignment to provide uniform backlighting suitable for motion-tracking analysis.

21. A computer-implemented system for classifying insecticidal activity, comprising:(a) a behavioural database containing behavioural fingerprints generated according to the method of any of claims 1 to 14;(b) a processing module configured to train a machine-learning model on behavioural fingerprints corresponding to known insecticidal modes of action; and(c) a prediction module configured to receive a new behavioural fingerprint and process the new behavioural fingerprint to output at least one of: a probability of activity, a predicted mode of action, or an indication of novelty.

22. The system of claim 21 , wherein one or more of the behavioural fingerprints are linked to microbial metabolic profiles or genomic data.

23. The system of claim 21 or 22, wherein the machine-learning model uses one or more of logistic regression, random forest, neural network, or clustering algorithms to perform classification and / or novelty detection.

24. The system of any of claims 21 to 23, wherein the machine-learning model is a continuous model that continuously retrains as behavioural fingerprints are added to the behavioural database.56

Images