WO2020206491A1 - Bee repellent - Google Patents

Abstract

A method of conditioning honeybees in which a deterrent to feeding is provided in conjunction with a signal that can be detected by the honeybee and will be associated with the deterrent. The honeybee will expect the deterrent to be present whenever the signal is detected and so will be conditioned not to feed when the signal is detected. Other honeybees, either other foragers or bees in the hive, are conditioned to associate the signal with the deterrent through social learning Therefore the invention allows for the honeybees in a honeybee population to be repelled from a food source, such as an insecticide-treated crop. The invention provides compositions and kits including the components needed to condition honeybees and ensure they are repelled, and these kits can include alternative signals in case a particular signal is not effective in a specific environment.

Description

BEE REPELLENT

TECHNICAL FIELD

[001 ] The present invention relates to a method of repelling honeybees and to honeybee repellent compositions.

BACKGROUND

[002] Any references to methods, apparatus or documents of the prior art are not to be taken as constituting any evidence or admission that they formed, or form part of the common general knowledge.

[003] The European honeybee, Apis mellifera L., is a valuable insect. The production of honey and related products is worth $A122 million annually in Australia (ABARES, 2019). This value is dwarfed by the contribution of honeybees to pollination of crops. Insect pollination has been estimated to be worth $US153 billion annually to the world, equivalent to 9.5% of food production. Many crops are especially adapted to bee pollination. For these reasons, reported declines in the abundance of pollinators and particularly honeybees have caused great alarm over the last decade or so (Biesmeijer et at. 2006; Klein et al. 2007; Potts et al. 2010; Rhodes 2018).

[004] Honeybees, being insects, are vulnerable to most insecticides. The basic metabolism of bees is very similar to that of caterpillars, beetles, aphids and other pests, and there are too few points of difference to provide target sites for potentially selective insecticides. While bees (in contrast to popular opinion) are not more susceptible to most pesticides (Hardstone & Scott 2010) they are generally not more resistant either.

[005] Older insecticides including organochlorines such as DDT, organophosphates such as parathion and synthetic pyrethroids such as permethrin were highly toxic to honeybees (Estesen et at. 1992). Honeybee deaths near cotton in particular have been associated with the use of these pesticides (Robertson & Rhodes 1992). In recent years however the problem of acute poisoning has been markedly reduced (Kiljanek et al. 2016) partly because many such insecticides have been banned or withdrawn from use, and partly through management practices that separate honeybees from sprayed fields, such as timing of applications when honeybees are not active.

[006] Modern insecticides including neonicotoids such as imidacloprid, clothianidin and thiamethoxam and the phenylpyrazoles such as fipronil pose a different threat. Their acute toxicity is often less than the older insecticides, and they allow a bee that contacts them in the field to survive long enough to carry the insecticide back to the hive. Through the process of trophyllaxis (honeybees sharing nectar in the production of honey) the toxin can spread to other worker bees, and the queen. The sub-lethal levels that result can have many effects on the behaviour and physiology of worker bees, foragers and the queen.

[007] Honeybees can be exposed to pesticides through spray drift or by foraging in treated crops. Spray drift is often blamed, especially by beekeepers, but documented cases are rare. A more common mechanism is exposure of foraging honeybees to pesticides through contact with sprayed foliage or nectar which has been contaminated through the systemic movement of insecticides within the plant. Guttation (the secretion of water or sap forming droplets on leaves which can be consumed by honeybees) is another mechanism (Reetz et al. 2016). The frequency of insecticide contamination in beeswax, and the irregular occurrence of poisoning between hives, support the argument that most bee poisonings arise from foraging rather than spray drift (Robertson & Rhodes 1992; Mullin et al. 2010). In the USA, 98% of hives tested were contaminated with insecticides, with a mean of 6 different pesticides in each (Mullin et al. 2010). This included insecticides used in-hive for pest control as well as those acquired from sprayed crops. Australian data indicate fewer pesticides. In hives placed near canola in Western Australia, Manning (2018) recorded 14 different pesticides (including herbicides and fungicides), with an average of 1 .2 different pesticides per hive.

[008] The route of exposure matters because it determines the effectiveness of strategies to manage honeybee poisonings. Drift varies with application equipment and weather conditions, but in general if best management practices are observed the lethal levels of insecticides do not persist beyond a few tens of metres beyond the sprayed field. Spraying when honeybees are not active, and avoiding known apiary sites, are effective means of countering drift. Restricting contact during foraging is more difficult, given that the foraging range of honeybees is several kilometres and sprayed insecticides may persist for some time (Krupke et al. 2012).

[009] Honeybee poisoning is becoming increasingly important in determining whether insecticides pass registration requirements, and whether they remain on the market. All pesticides that are toxic to bees carry label requirements restricting their use to times when bees are not foraging, and they may be restricted for crops where bees are important pollinators. In recent years the approach of the US EPA has changed so that risk assessments are required to consider the total load of pesticides to which bees may be exposed, not just the risk for any one pesticide (Berenbaum 2016). Registrations for some pesticides such as sulfoxaflor have been withdrawn or severely limited (Vanegas 2017). In Europe many countries have banned some or all of the neonicotinoid insecticides (Carreck 2017) despite conflicting evidence about their role in bee declines (Moritz & Erler 2016). The value of these pesticides to agriculture underlines the potential value of a method for repelling bees from treated crops. [010] There have been numerous attempts in several countries, beginning over 50 years ago to produce a product that can be sprayed on a crop to temporarily repel honeybees. Woodrow et al. (1965) in the USA tested 195 different chemicals or natural extracts by blowing the vapours onto frames extracted from hives. This would have tested their effects on a mixed population of nurse bees, foragers and maybe guard bees. It is not clear whether the aim was to produce materials which would clear the frames to facilitate honey collection, or develop products for spraying on crops, or both. They found 4 compounds that were attractive, and 19 that were repellent, including various amines, acids, acid anhydrides and carbonyl compounds. Woodrow et al. (1965) also note the use of phenol, propionic and butyric acids and anhydrides by beekeepers for clearing bees from frames, but phenol is toxic and the other compounds have unpleasant odours which might contaminate honey. They also provide earlier references on repellence of many older insecticides, but clearly this is not strong enough to prevent poisoning, and in their tests diluted solutions of these compounds were not repellent.

[01 1 ] Also in the USA, Atkins et al. (1975a) tested 143 compounds in the laboratory, using two methods: incorporating the compounds in honey syrup, and by placing disks of silica gel saturated with the compounds around vials containing honey syrup. In both cases the volume of syrup consumed, relative to a control, provided a measure of repellence. They describe the first procedure as testing“gustatory repellence” and the second as testing“spatial repellence”. More properly,“gustatory repellence” should be termed “deterrence”, that is, stopping a process such as feeding after it has begun, whereas “repellence” means preventing the process from occurring at all. Thirteen chemicals were found to have various levels of gustatory repellence, and 10 had high levels of spatial repellence, with several more having lower levels. Some of these chemicals were well known general insect repellents such as N,N-Diethyl-meta-toluamide (DEET), but there were also compounds known to repel other insects that did not work on honeybees. Many compounds were considered too volatile for use as sprays to repel bees from crops. Structurally, compounds containing nitrogen, short side-chain substituted phenyl acetates and/or tolyl compounds showed the most promise. Chemicals with thymol-like odours also appeared promising. In a subsequent study, Atkins et al. (1975b) tested 12 of the most promising compounds by spraying them (diluted in water) on small plots of various crops, and counting foraging honeybees coming from nearby apiaries. Only two chemicals were considered promising. Decylamine gave about 70% reduction in bee visits over about 8 h. Another unidentified proprietary chemical, Rutgers 6-12, provided slightly less protection. Several chemicals were considered either too volatile or too phytotoxic. [012] Mayer et at. (2001 ), also in the USA, tested 240 chemicals and extracts or formulations using small plot trials on apples, buckwheat, clover and dandelions, using application methods similar to those of Atkins et al. (1975). Eleven chemicals significantly reduced honeybee numbers after 1 h, but only one (decylamine) did so after 4 h. Caprylic acid and mixtures of 2-heptanone + decyclamine and 2-heptanone + B-caryophylene + 2 ethyl-1 ,3 hexandiol also reduced bee numbers at 4h.

[013] In another US study, Collins et at. (1996) tested five compounds as repellents for use by beekeepers against aggressive Africanised bees. None were effective when sprayed on people, but all were effective when sprayed in the air, directly at attacking bees. The most effective were DEET and benzaldehyde. These results however may not be relevant for foraging honeybees.

[014] Another US study is that of Hagler & Buchmann (1993), who noted the reluctance of honeybees to feed on the nectar of some desert plants such as aloe and tamarind which are rich in phenolic compounds. They showed that adding caffeic acid and genistic acid to sucrose solutions reduced feeding by honeybees.

[015] In the UK, Free et al. (1985) tested the alarm pheromones isopentyl acetate and 2-heptanone as repellents on small plots of canola, field beans and sunflowers. This work followed earlier laboratory studies indicating that alarm pheromones suppress appetitive behaviour and appetitive learning. Both substances, alone and in combination, strongly repelled honeybees, but only for 10-20 minutes.

[016] More recently, in India, Mishra & Sihag (2009a, 2010) tested 25 chemicals in“semi field” conditions, by incorporating the chemicals in 30% sucrose in feeders located 10 m from hives and counting the numbers of European honeybees (A mellifera) and dwarf honeybees {A. florea) visiting the feeders over 1 minute. Fifteen compounds (ketones, aldehydes and phenols) showed repellence of greater than 80%, though the duration of the effect was not stated. Mishra & Sihag (2009b) tested these fifteen chemicals in field trials, using sprays on small plots of mustard. Only three (p-bromophenol, m- bromoacetophenone and trimethoxyacetophenone) showed more than 80% repellence for 3 h.

[017] WO 2012/151556 discloses honeybee repellents exhibiting repellent properties which mimic those of 2- heptanone. The repellent effect is said to occur through interaction of a compound with a honeybee odorant-binding protein called OBP2. Some of the specific compounds named include 1 -benzyl-4-(4-methoxy-2,3-dimethylbenzyl) piperazine, 1 -[(6- nitro-1 ,3-benzodioxol-5-yl)methyl]-4- phenylpiperazine, 4-[4-nitro-3-(2- phenoxyethoxy)phenyl]morpholine, 1 -(3-chlorophenyl)-4-[(6-nitro-1 ,3-benzodioxol-5- yl)methyl]piperazine, 1 -(4-methoxybenzyl)-4-(3-methylbenzyl)piperazine, 4-(4-methoxy- 2,3-dimethylbenzyl) morpholine, 1 -benzyl-4-(4-methoxy-3-methylbenzyl) piperazine, 1 -(4- methoxy-2,3-dimethylbenzyl)-4-methyl piperazine, 1 -(4-methoxy-2,3-dimethylbenzyl)-4 methylpiperazine, 1 -(1 ,3-benzodioxol-5-ylmethyl)-4-(4,5-dimethoxy-2- nitrobenzyl)piperazine, 1 -(2,3dimethylphenyl)-4-(4-methoxy-2,3- dimethylbenzyl)piperazine, 1 -(4-methoxy-2,3-dimethylbenzyl)-4 phenylpiperazine, 1 -[3- (4-iodophenoxy)propyl]piperidine and 1 -(4-methoxy-2,3-d imethylbenzyl)-4-(4 methylphenyl)piperazine.

[018] US20160029628 describes a method for repelling or directing an insect away from a structure, article, and/or organism, comprising treating the structure, article, and/or organism with a composition comprising about 0.1 to about 2.5 weight percent of one or more of compounds with any of the Formulae l-XIII set out in therein.

[019] Despite these efforts to develop repellent compositions, there remains a need to develop an effective way to keep honeybees off insecticide-treated crops.

SUMMARY OF INVENTION

[020] The present inventors have formed the view that previous attempts to develop bee repellent compositions have been unsuccessful because they looked for inherent repellence, that is, substances which are“hard-wired” into bees’ brains as things to be avoided. Instead, the present inventors have found that a bee repellent can be produced using a methodology in which learning behaviour of honeybees is exploited.

In one aspect, the invention provides a method of conditioning a honeybee, comprising the steps of:

a) exposing the honeybee to a deterrent to feeding that is unpleasant to a honeybee but not harmful;

b) exposing the honeybee to a signal that can be detected by the honeybee and will be associated with the deterrent by the honeybee;

whereby the honeybee expects the deterrent to be present when the signal is detected and so will be conditioned not to feed when the signal is detected.

[021 ] In another aspect, the invention provides a method of repelling honeybees from a food source, comprising the steps of:

a) conditioning at least one honeybee as described above;

b) providing a signal that can be detected by the honeybees and has become associated with the deterrent; and

c) applying the signal to the food source to repel honeybees.

[022] In another aspect, the invention provides a kit comprising a first composition that provides a deterrent to feeding that is unpleasant to a bee but not harmful and a second composition that provides a signal that can be detected by the bee and will be associated with the stimulus. BRIEF DESCRIPTION OF THE DRAWINGS

[023] Preferred features, embodiments and variations of the invention may be discerned from the following Detailed Description which provides sufficient information for those skilled in the art to perform the invention. The Detailed Description is not to be regarded as limiting the scope of the preceding Summary of the Invention in any way. The Detailed Description will make reference to a number of drawings as follows:

Figure 1 is a graph showing flying, landing and feeding activity on 30% sucrose on a warm sunny day. Landing/flying and feeding/landing ratios are calculated by summing the half-hourly totals over the whole day.

Fig. 2 is a graph showing flying, landing and feeding activity on 30% sucrose on a cool cloudy day. Landing/flying and feeding/landing ratios are calculated by summing the half-hourly totals over the whole day.

Fig. 3 is a graph showing flying, landing and feeding activity of bees feeding on blend EkS2 on the first day. Landing/flying and feeding/landing ratios are calculated by summing the half-hourly totals over the whole day

Fig. 4 is a graph showing flying, landing and feeding activity of bees feeding on blend EkS2 on the second day. Landing/flying and feeding/landing ratios are calculated by summing the half-hourly totals over the whole day

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

[024] The invention allows for learning behaviour of honeybees to be exploited by providing a method of conditioning honeybees. In order to condition a bee, it is first exposed to a deterrent to feeding that is unpleasant to a bee but not harmful and then to a signal that can be detected by the bee and will be associated with the deterrent. In this way the bee is conditioned to associate the signal with the unpleasant sensation they have experienced as a result of exposure to the deterrent, such as by ingesting small amounts of a compound that the bee finds unpleasant or senses as being a hazard. Since the deterrence of feeding and the signal become associated, the bee will be conditioned not to feed when the signal is detected.

[025] Previous efforts to develop bee repellents have looked for inherent repellence, that is, substances which are“hard-wired” into bees’ brains as things to be avoided. However, bee foraging behaviour is complex, sophisticated and highly adapted to changeable environments. Learning is critical. Honeybees can be trained to respond to chemicals which are of no relevance in their normal lives, for example in their use as biosensors (Hadagali & Suan 2017) to detect explosives, corpses and environmental contaminants. Contrary to what is implicit in the prior art teachings, the present inventors have formed the view that it is unlikely that there are many chemicals which are inherently repellent, because this would be maladaptive when learning could associate them with either threats or rewards. Thus the same chemical might be attractive in some circumstances but repellent in others. For example, phenylacetaldehyde was found to be repellent by Mishra & Sihag (2009a) but attractive by other researchers (Blight etal. 1997; Sandoz etal. 2001 ). Almost any sudden change in food odour can produce temporary repellence (Woodrow et ai. 1965).

[026] Learnings can be transferred from forager bees to other foragers, or to bees which have never left the hive (Farina et al. 2005). This is known as‘social learning” and can be accomplished by the“waggle dance” or by transfer of information from volatiles chemicals carried by returning foragers, or chemicals shared through shared through trophyllaxis. This means that it is not necessary for all bees in the hive to learn directly through exposure to an unpleasant substance or the signal associated with it. Conditioning of only a proportion of the foraging bees could lead to avoidance by most or all of the others.

[027] Thus repellence depends on context and what honeybees have previously learnt. None of the teachings of the prior art considered these factors, and given the problems of producing naive honeybees (Giurfa et al. 1995), it would be difficult to do so. Accordingly, the present inventors believe that in producing a repellent it will be necessary to exploit learning behaviour rather than working against it.

[028] As used herein, the term“deterrent to feeding” refers to a thing or event that evokes a specific functional reaction in a bee that deters it from feeding but is not harmful. Deterrence occurs as a result of any detectable change in the environment of the bee. A bee may be deterred from feeding by a change in environment detected by touch, vision, smell, taste or sound.

[029] The interaction may be any effect which causes a bee to prefer not to feed but, in an embodiment, is an interaction with a chemical compound that deters feeding. The deterrent effect may be through visual, gustatory and/or olfactory inputs. The honey bee antennae (one on each side) house thousands of sensory organs, some of which are specialized for touch (mechanoreceptors), some for smell (odour receptors), and others for taste (gustatory receptors). Honeybees have a combined mouth parts than can both chew and suck. This is accomplished by having both mandibles which can be opened and closed to chew and a proboscis. The proboscis is mainly used for sucking in liquids such as nectar, water and honey. The interaction may be by way of ingestion of the compound, sensing its taste or odour or any other possible interaction. A compound may deter a bee from feeding in any way possible. For example, it may deter a bee from feeding due to its bad taste or unpleasant odour, or because ingestion is unpleasant in some way that makes the bee uncomfortable. A compound may serve to warn a bee of a hazard or potential hazard. [030] In an embodiment the feeding deterrent is selected from the group consisting of vegetable oils, mineral oils or phenols.

[031 ] In an embodiment the vegetable oil is selected from the group consisting of olive oil, sesame oil, peanut oil, canola oil, cottonseed oil, corn oil, soybean oil, mineral oil, as well as ethoxylated and methylated forms of these oils, or mixtures thereof.

[032] In an embodiment the feeding deterrent is canola oil. Canola oil is a cheap, readily available ingredient which greatly deters honeybees from feeding. While not wishing to be bound by theory, it is believed that oils have a tendency to spread and cover the respiratory spiracles of a honeybee with a thin film that may cause injury or death. Thus ingestion of or contact with a small quantity of oil may warn the bee of its presence and cause it to discontinue feeding at that food source.

[033] As used herein, the term “unpleasant to a bee” means something that causes discomfort to the bee, is unappealing to the bee or which serves as a warning of potential danger or injury.

[034] As used herein, the term“not harmful’ refers to something which will not cause death in a bee, and which preferably is not injurious or dangerous to a bee, most preferably having no effect on the bee’s health or behaviour save for deterring feeding. Whether or not something is harmful will, in some circumstances, depend on how the bee interacts with the substance.

[035] As used herein, the term“signal” refers to a thing or event that conveys information about the behaviour or attributes of some phenomenon.

[036] The signal is detectable by a bee. Accordingly, the signal may be a visual or tactile signal or an odour, taste or sound.

[037] The signal becomes associated with the deterrent. Thus the bee has a specific functional reaction to whatever causes feeding to be deterred and the signal becomes associated with that thing or event. Persistence of the signal may be interpreted by the bee as an indication that the condition that caused the bee to not want to feed may also persist, whether or not it does. Likewise, recurrence of the signal may be interpreted by the bee as an indication that the condition that caused the bee to not want to feed may be present again, whether or not it is. Therefore, persistence or recurrence of the signal causes the bee to endeavour to avoid the unpleasantness associated with feeding, and so the bee may choose not to feed. Thus the signal itself, even in the absence of anything providing the stimulus, deters feeding in honeybees.

[038] Honeybees are notable among insects for the complexity of their societies. Worker bees are divided into many types, each focused on particular tasks, and these include foragers whose role is to search for food and bring it back to the hive. [039] For the purposes of developing deterrents and repellents to reduce pesticide poisoning, the response of foragers is the critical consideration. Even though they are only a proportion of bees in the hive, they are the ones that encounter the pesticide in the field. Thus, while repellents have been developed that work on bees in the hive, to facilitate beekeeping or reduce attacks by aggressive bees (e.g. (Collins et al. 1996) they may not work on foraging bees, or in locations beyond the hive.

[040] In an embodiment foraging bees are conditioned. This can be done by positioning the deterrent to feeding and/or the signal associated with the deterrent at a distance from the hive. In an embodiment the deterrent to feeding and/or the signal associated with the deterrent are positioned at a distance of from 5 metres to 50 metres, advantageously from 15 metres to 40 metres, typically about 20 metres from the hive.

[041 ] In an embodiment the signal is an odour. Typically, the odour is generated by the provision of one or more volatile chemical compounds. The one or more volatiles will overwhelm, and provide a contrast to, the volatiles produced by the crop on which the bees have been foraging. This will provide a short term repellent effect and an olfactory marker for learning about deterrent effects. A wide range of volatiles might fill this function, so choices should take account of pragmatic factors such as cost, regulatory risks, volatility and antimicrobial activity (to extend shelf life).

[042] Honeybees can also generalise their ability to learn odours. They can recognise a chemical in a blend and respond to it in other blends. Sometimes the response to the blend is greater than to the individual chemical, and sometimes it is less (Sandoz 201 1 ). Honeybees can also respond to a range of structurally similar chemicals when they have learned on a representative example (Guerrieri et al. 2005). This flexibility allows them to respond to a much wider range of chemicals and blends than the fixed architecture of their sensory system (different types of neurones and central processing structures) would suggest. It enables them to deal with complex blends of plant volatiles which can change rapidly as resources shift in the field.

[043] It will be appreciated that a wide range of volatiles can serve as a signal, depending on the conditioning of the bee. Moreover, the signal used in conditioning the bee can be changed to suit environmental conditions or to ensure that the conditioning is appropriate. For example, in a particular environment, a volatile used successfully elsewhere as a signal to deter feeding may be naturally present. In these circumstances it would not be desirable to use the compound as a signal as bees could be confused and deterred from feeding in other than the intended locus. For example, the bees may be deterred from feeding at a natural food source when it is intended that they do not attempt to feed from an insecticide-treated crop. [044] It may also be desirable to change the signal from time to time to recondition the bees. This might, for example, ensure that a strong correlation between the signal and deterring feeding at the desired locus is maintained. For example, the signal may be changed to renew the association and then that new signal would be applied to an insecticide-treated crop.

[045] There is no reason why any volatile compound whose odour can be detected by a bee cannot be used. The volatile compounds tested in the prior art should be suitable signals if the bees are conditioned appropriately, even if they have no inherent repellence. Prior publications failed to recognise the ability of bees to learn and adapt, so the compounds described therein could still be suitable for use in the present invention.

[046] In an embodiment the volatile compound is a plant volatile. Plant volatiles are volatile compounds produced in plants as secondary metabolites. Typically plant volatiles are terpenoids, aromatic compounds derived from L-phenylalanine, fatty acid derivatives, and volatiles derived from amino acids other than L-phenylalanine.

[047] In an embodiment the signal comprises a terpenoid. The terpenoids are a large and diverse class of naturally occurring compounds derived from terpenes by introduction of a functional group. Most are multicyclic structures with oxygen-containing functional groups. Terpenoids often contribute to the scent associated with a plant. For example, terpenoids provide the characteristic scent of eucalypts. Terpenoids exhibit several carbon skeletons and are extremely variable in chemical structure, yet share a common feature of biosynthesis which involves addition of isoprene units. Accordingly, the terpenoids can be classified according to the number of isoprene units that comprise the parent terpene. Using this scheme, a hemiterpenoid is a terpenoid derived from a terpene comprising a single isoprene unit, a monoterpenoid is a terpenoid derived from a terpene comprising two isoprene units, a sesquiterpenoid contains three isoprene units, a diterpenoid contains four isoprene units, a sesterterpenoid contains 5 isoprene units, a triterpenoid contains 6 isoprene units, a tetraterpenoid contains 8 isoprene units and a polyterpenoid contains more than 8 isoprene units. The terpenoids may contain one or more cyclic moieties. Advantageously the terpenoid is a volatile compound, generally a hemiterpenoid, a monoterpenoid or a sesquiterpenoid.

[048] In an embodiment the terpenoid is selected from the group consisting of E or Z- 3,7- dimethyl-2,6-octadienal, menthol, camphor, a-pinene, cineole and D-limonene.

[049] In an embodiment the terpenoid is selected from the group consisting of a-pinene, cineole, citral, geraniol, linalool and limonene.

[050] In an embodiment the signal comprises a volatile aromatic compound. Typically, aromatic compounds classed as plant volatiles comprise an oxygen-containing functional group. These aromatic compounds are frequently elaborated in floral tissues. They are often derived from L-phenylalanine in biosynthetic pathways involving chain-shortening of trans-cinnamic acid and structures co-opted from lignin biosynthesis to form benzoids. Aromatic compounds can also be formed independently of the L-phenylalanine pathway using type-lll polyketide synthases from various coenzyme-A conjugates.

[051 ] In an embodiment the signal is a volatile aromatic compound comprising a hydroxyl, ether, ester or carbonyl group.

[052] In an embodiment the signal is a volatile aromatic compound comprising an oxygen-containing functional group selected from the group consisting of anisyl alcohol, benzaldehyde, benzyl alcohol, phenylacetaldehyde, methyl salicylate, butyl salicylate, 2- phenylethanol, anethole, asarone, chavicol, and eugenol.

[053] In an embodiment the signal is a green leaf volatile. A green leaf volatile produces the odour that one detects when a leaf is crushed or damaged. Typically, these are C₆ alcohols, aldehydes, ketones and esters that are derived from Cis fatty acids. Examples of green leaf volatiles include E-2-hexanol, E-2 hexanal and Z-3-hexenyl acetate.

[054] Honeybees produce at least 30 different pheromones (defined as chemicals which communicate information between members of the same species). Alarm pheromones are released by guard bees and either recruit other bees to defend the hive, or induce them to attack natural enemies such as hornets or intruder bees from other hives which may seek to rob the resources of the hive. Alarm pheromones are produced in two locations, in mandibular glands in the mouthparts and in the sting apparatus. Several chemicals are produced in both locations, but the most abundant and best known are 2- heptanone in the mandibular gland and isopentyl acetate in the sting apparatus.

[055] In an embodiment the signal is a pheromone.

[056] In an embodiment the signal is an alarm pheromone.

[057] In an embodiment the signal is 2-heptanone or a functional analogue thereof. A functional analogue of 2-heptanone is a compound that mimics the interaction of 2- heptanone with a honeybee odorant-binding protein called OBP2. Functional analogues of 2-heptanone are described in WO2012/151556, the content of which is incorporated herein by reference.

[058] In an embodiment the signal is isopentyl acetate or an analogue thereof. Analogues of isopentyl acetate include n-pentyl acetate, 2-methylbutyl acetate, isobutyl acetate, 1 -methylbutyl acetate, 4-methyl-2-pentyl acetate, isopropyl acetate, ethyl acetate and isoamyl alcohol.

[059] Odours are not the only stimuli which lead to appetitive or aversive learning. Colour (e.g. Giurfa et al. 1994), patterns and texture (Erber et al. 1998) can also be learned, and responses to these stimuli can interact with odours (Dotterl & Vereecken 2010), and with endogenous circadian rhythms of foraging behaviour (Moore et al. 1989; Lehmann et al. 201 1 ). This leads to learned associations that are time and place-specific, so that bees “know what to do, and when” (Zhang etal. 2006). Some colours have an intrinsic attraction to honeybees, such as blue. Therefore, advantageously, apparatus and compositions used to put the method of the invention into effect will be a colour which has no intrinsic attraction to honeybees e.g. grey or black rather than blue.

[060] In an embodiment the signal is a colour, pattern or texture. A grey or black dye such as carbon black or activated charcoal powder might be used to colour compositions that contain a deterrent to feeding that is unpleasant.

[061 ] In an embodiment the signal is an odour in conjunction with a colour, pattern or texture. In this case a grey or black dye such as carbon black or activated charcoal powder might be used to colour compositions that contain a deterrent that is unpleasant and a signal such as a volatile chemical to reinforce the chemical signal.

[062] The present invention also relates to a method of repelling honeybees from a food source. In order to repel the signal is applied to the food source to repel honeybees. This may be in the presence or absence of the deterrent. In the former case the signal and the deterrent may be applied sequentially or may be formulated together into an agriculturally acceptable formulation.

[063] The method may be achieved by applying a composition comprising a first compound that provides a deterrent to feeding and a second compound that provides a signal that can be detected by the bee and will be associated with the deterrent.

[064] In an embodiment one or more additional compounds that provide a signal that can be detected by the bee are incorporated in the composition. These additional compounds may work together to create a signal stronger than the sum of contributions, or may simply work together in an additive fashion to create a signal.

[065] An insecticide may be included in the composition. In an embodiment the insecticide is selected from the group consisting of fipronil, sulfoxaflor, imidacloprid, acetamiprid, clothianidin, thiamethoxam, synthetic pyrethroids, carbamates, and organophosphates.

[066] In an embodiment a slow release mechanism is employed to prevent overly rapid loss of the volatiles. This might be microencapsulation or the use of activated charcoal powder to absorb a proportion of the volatiles. The volatile compounds may also be incorporated in a solid substrate, such as clays, diatomaceous earth, silica, polyvinyl chloride, polystyrene, polyurethanes, ureaform aldehyde condensates, and starches. Other useful solid support matrices include expanded vermiculite and paraffinic wax. Oil in water emulsions provide some slow-release characteristics, especially if they are applied in large droplets and combined with a humectant such as sugar, glycerol, glycols or polydextrose, which allows the formulations to re-liquefy overnight. Addition of a humectant may allow a composition comprising volatiles remain active for 4-6 days.

[067] Excipient ingredients such as thickeners, carriers and antioxidants may be incorporated in compositions according to the invention. Other components which may be included in the formulation include antimicrobial agents, emulsifiers, film forming polymers and mixtures thereof. Thickeners include starches, vegetable gums and synthetic polymers such as carbomer. Oils incorporated as feeding deterrents may also function as carriers. Additional carriers including polyols, esters, methylene chloride and alcohols could be incorporated. Mixtures of carriers are envisaged in the present invention and, for example, an aqueous/oil mixture in which the volatile compounds dissolve in a miscible vegetable oil for subsequent admixture with a water are envisaged. Antioxidants include natural antioxidants such as ascorbic acid and tocopherols, as well as synthetic antioxidants such as propyl gallate, tertiary butylhydroquinone, butylated hydroxyanisole and butylated hydroxytoluene.

[068] A signal that deters feeding in the manner described herein serves to repel honeybees. While not wishing to be bound by theory, communication of the perceived threat that the signal provides prevents other bees from returning to the area where the perceived threat is located. Thus the number of bees returning to the area is greatly reduced. Those bees that do pass by an area treated with a signal such a volatile chemical are much less likely to land or feed than they are in an untreated area.

[069] The food source may be any source of food for honeybees. Commonly the food source is a crop. It may be an orchard crop, oilseed crop, fibre crop, grain legume or pasture crop. Typically, the crop is one that honeybees favour as a food source.

[070] In an embodiment the food source is an oilseed or fibre crop selected from the group consisting of canola, sunflower, safflower, sesame, hemp, cotton and mustard.

[071 ] In an embodiment the food source is grain legume crop selected from the group consisting of fava beans, soybeans, lentils, lupins, chick peas, field peas, pigeon peas and peanuts.

[072] In an embodiment the food source is pasture crop selected from the group consisting of lucerne, clovers, medics, trefoils, vetches, lablab and Desmodium.

[073] In an embodiment the food source is an orchard crop selected from the group consisting of apples, mangoes, kiwi fruit, plums, peaches, nectarines, guava, pomegranates, pears, black and red currants, cashews, apricots, avocados, almonds, passion fruit, cherries, coffee, walnut, lychee, macadamias, citrus, persimmons, hazelnut, Brazil nuts and papaya.

[074] In an embodiment the food source is a vegetable crop selected from the group consisting of strawberries, onions, green beans, lima beans, kidney beans, celery, carrots, cucumber, zucchinis, watermelon, cantaloupe, pumpkins, beets, broccoli, cauliflower, cabbage, Brussels sprouts, bok choy (Chinese cabbage), turnips, potatoes, capsicum, eggplant, raspberries, elderberries, cranberries, blueberries, tomatoes and grapes. In an embodiment the food source is an ornamental plant.

[075] The signal may be applied to an insecticide-treated crop, either alone or in conjunction with the feeding deterrent. The easiest way to repel honeybees from the crop would be to incorporate a first composition that is a feeding deterrent, and a second composition that provides a signal into a spray. The honeybees would be conditioned as they ingest the feeding deterrent, and would be deterred from feeding thereafter for as long as the signal persists.

[076] An alternative way to repel honeybees from the crop would be to incorporate the second composition that provides a signal into an insecticidal spray without the first composition, provided that the honeybees have been conditioned to recognise the second composition as a signal. The honeybees might be conditioned prior to insecticide application by providing nearby apiaries with feeders containing a composition comprising both the first composition and the second composition.

[077] A spray of the repellent (comprising either first and second compositions or just the second composition) could be applied shortly before the insecticide. This would add to the cost but the application technique could be adapted to suit the repellent rather than the insecticide. It might be possible to treat only part of the field in this way, such as the edges.

[078] The crop may be a crop which is treated in anticipation of later application of an insecticide treatment. The treatment with insecticide may follow immediately after application of the signal. Alternatively, the food source may be treated with an insecticide at a later time. In an embodiment the insecticide may be applied from 30 minutes to 1 week after application of the signal.

[079] The compositions may be provided in the form of a kit. A kit may comprise a first composition that provides a deterrent to feeding, a second composition that provides a signal that can be detected by the honeybee and will be associated with the deterrent. The kit will generally include instructions for use. A kit could further comprise additional compositions that provide a signal that can be detected by the honeybee and will be associated with the deterrent. These could be used as alternatives in case one signal is less effective than others. The kit may include an insecticide.

EXAMPLES

Example 1 Feeding stations and behavioural observation

[080] Three beehives stocked with honeybees ( Apis mellifera L., subspecies ligustica) were established in an urban backyard at Armidale, NSW, Australia (30°29’46” S, 151°417Έ)). Starter cultures were obtained from a local beekeeper, and hives were maintained using standard beekeeping practices.

Feeding stations

[081 ] Two level platforms 1 .6 x 1.6 m constructed of plywood were established at distances of approximately 15 m from the hives. This distance was determined so that the behaviour of foraging bees (as distinct from nurse and guard bees within the hive) could be observed. The platforms were covered with black builder’s plastic which was replaced at regular intervals to minimise contamination with formulations that may have been spilt or transferred by bees.

[082] Formulations were presented in feeders made from plastic pot dishes (Artevasi Round Saucer, Artevasi DNA, Argoncihle, Portugal) 30 cm in diameter. A perforated metal plate (Williamstown Metals, Melbourne) which had 2.41 mm holes spaced at 6.36 mm (open area approximately 12%) was cut to fit inside the base. There were approximately 1200 holes in each disk, and each hole could accommodate one feeding bee. The plates were supported by legs made from furniture protectors, so that they sat approximately 15 mm from the bottom of the dish. Eight radial segments were marked on each dish to facilitate counting.

[083] Candidate formulations were weighed out (approximately 540 g) and placed in the dish so that the level would be just below the metal disc, which was then placed over the formulation. The full feeder was then weighed. Bees could feed through the holes to a depth of about 7 mm, the length of their proboscis, so they could potentially consume up to half of the formulation supplied. After each run the feeder was weighed again to determine the amount that had been consumed. Another feeder covered by steel mesh so bees could not access it was set up to control for evaporation losses. When sugar solutions were being fed, a valve made from a modified bee entrance feeder (Tamworth Beekeeping Supplies, Tamworth) was fitted so that it kept the level of sugar solution just below the metal plate. A PET 1.25 litre soft drink bottle was screwed into the feeder to continuously supply sugar solution so bees could potentially consume approximately 1 .5 litres of sugar solution (the bottle plus about 250 ml of the initial dish fill), or about 1950 g. Behavioural observations

[084] Time lapse video cameras (Brinno TLC200 Pro, 4F, Taipei,) fitted with 19 mm lenses were suspended approximately 80 cm above the centre of the observation platforms, so that the image encompassed the whole platform. They were set to record one frame every two seconds. Recordings began shortly after sunrise, before bees became active, and ceased around 1500 - 1600 h, after all activity had ceased or (in the case of sugar solutions) all the sugar had been consumed. [085] The videos were scored using VLC Media Player software v. 2.0.8 (VideoLan Organisation,) which could replay them at varying speeds and freeze individual frames. At 30 minute intervals, exactly on the hour and half-hour, counts were made on the video frame of the following behaviours:

• Flying : all the airborne bees on the platform (not just the feeder dish) were counted.

• Landed : Bees alighted on the metal plate disc were counted. It was sometimes difficult to separate landed bees from those flying low over the plate, but bees that cast a shadow, or had visibly extended wings, were counted as flying rather than landed.

• Feeding : Bees that remained on the same hole between the scored frame and the next frame (2 seconds after) were scored as feeding.

Example 2 Formulations

[086] A base formulation was made up using the ingredients listed in Table 1 .

Table 1

[087] Volatile components were added to the base formulation and/or modified the base formulation. To do this we made up the base using an industrial blender (Robot Coupe Blixer 6VV, Robot-Coupe Australia Pty. Ltd., Artarmon) and selectively added volatiles or base components, or selectively omitted those being investigated.

[088] In this way the following formulations were made up:

(a) Base formulation variations

1. Base: as per Table 1 , with no volatiles.

2. Thin base: as per Table 1 but omitting Rhodapol and with no volatiles 3. A low oil base: as per Table 1 but with only 25 g of canola oil instead of 380 g and no emulsifiers and no volatiles. 50 ml ethanol was used to dissolve the antioxidants for this blend.

(b) Volatile blends

Volatile Blend 1 -three terpenoids: a-pinene 5.88g/L, cineole 5.07 g/L and D-limonene 1.88g/L) were added to the base blend in T able 1 .

Volatile Blend 2 - aromatic components: anisyl alcohol 5.20g/L, phenylacetaldehyde 9.08g/L and butyl salicylate 10.4 g/L were added to the base blend in Table 1 .

[089] Six formulations made up using the base in T able 1 , with 10g/L of one of: a-pinene, anisyl alcohol, butyl salicylate, cineole, D-limonene and phenylacetaldehyde and are referred to herein as Volatile Blends 3 to 8

Three formulations were made up using the base in Table 1 as follows:

1. Base as in Table 1 plus 10g/L 2-phenylethanol

2. Base as in Table 1 plus 10g/L 2-heptanone (an alarm pheromone compound)

3. Base as in Table 1 plus 10g/L isopentyl acetate (an alarm pheromone compound) These are referred to as Volatile Blends 9 to 1 1 .

[090] A further eight blends were made up as follows:

1. Blend Ek1 - Base as in Table 1 . Added volatiles were 2.5g/L each of phenylacetaldehyde, cineole, isopentyl acetate and 2-heptanone.

2. Blend EkF1 - Base as in Table 2. Twice the canola oil concentration (15.9%) and twice the emulsifier concentration (2.0% each of Alkamuls V02003 and RC) as the base from Table 1. Carbon black food dye (Queen Fine Foods) at 0.13% was substituted for blue food dye. Added volatiles were 20g/L anisyl alcohol plus 20g/L a-pinene plus 20g/L isopentyl acetate.

3. EkS1 - Base as in Table 3 with 1 .5 times normal canola oil and 1.8 times normal Alkamuls (emulsifier) concentration. Added volatiles were 20g/L D-limonene plus 20g/L phenylacetaldehyde plus 20g/L 2-heptanone.

4. EkS2 - Base as in Table 3 except the black food dye was increased to 40 g. Added volatiles were 2-phenylethanol 14.8g/L, cineole 20g/L and 2-heptanone 20g/L.

5. EkS3 - Base as in Table 3. Added volatiles were anisyl alcohol 20g/L plus cineole 20g/L plus 2-heptanone 10g/L plus isopentyl acetate 10g/L.

6. EkS3b - Base as in Table 4, similar to Table 3, but with canola oil levels reduced to standard base concentrations. Volatile components as for EkS3. This formulation has not yet been scored but has been compared in side-by side videos with 30% sucrose.

7. EkS4 - Base as in Table 5, with no sugar. Added volatiles were anisyl alcohol 20g/L, cineole 20g/L, 2-heptanone 10g/L and isopentyl acetate 10g/L. 8. EkS5 - Base as in Table 6, no sugar but polydextrose added as a humectant, and black food dye replaced by activated charcoal powder. Added volatiles were anisyl alcohol 20g/L, cineole 20g/L, 2-heptanone 10g/L and isopentyl acetate 10g/L.

Table 2. Base for blend EkF1

Table 3 Base for blends EkS1 and EkS2

Table 4 - Base for blend EkS3b

Table 5 - Base for blend EkS4

Table 6 - Base for blend EkS5

[091 ] All of these 8 blends had the following characteristics:

They were an oil in water emulsion with concentrations of canola oil varying from 7.9 to 18% and total emulsifier content varying from 2.0 to 4.0%

Except for blends EkS4 and EkS5 they had sugar (sucrose) at concentrations of around 30% (after addition of the volatiles)

They had minor components of antioxidants (0.28 - 0.42%) and dyes (0.1 to 1 .0% except for blend EkS5 where the dye was replaced by activated charcoal powder which had a similar colouring effect).

They had 3 or 4 volatile components at total concentrations ranging from 10 to

60g/L

Example 3 Feeding trials

[092] Blends were trialled after at least two training runs when 30% sucrose solution was supplied using the bottle attached to the feeder dish. A feeder was placed on each platform shortly after sunrise, before foraging activity started.

[093] Following the training runs, the blends were supplied in the feeders, usually without bottles as it soon became apparent that for most blends consum ption would be very low. Feeders containing approximately 540 g of each blend were placed on both platforms shortly after sunrise for two consecutive days. They were removed late in the afternoon, by which time feeding had ceased and little or no bee activity was occurring. Consumption was determined by weighing the dishes and correcting for the loss in the evaporation control.

[094] The videos were scored at intervals of 30 minutes and the numbers of flying, landed and feeding bees recorded.

[095] Bees always consumed the total quantity of sugar solution supplied (about 1950 g), but the rate at which they did so varied between days (Figs. 1 and 2). Factors affecting the rate at which the sugar was consumed included the weather (bee activity was reduced on cool or cloudy days) and the strength of the hives (the experiments took about 4 months to complete and during this time the hives were generally becoming stronger, with increased

[096] For these reasons comparing the absolute numbers of bees exhibiting different behaviours could be misleading, so we calculated two ratios: the landing/flying ratio and the feeding/landing ratio. We totalled the numbers of bees exhibiting each behaviour over all single frames at 30 min observations for the day in calculating these ratios. For both ratios, where the denominator was zero we added one to it, to enable calculation of a ratio. We also calculated the total activity for the day (total numbers of bees seen flying and landed over all observations for the day). Since feeding was a subset of landing, feeding bees were not counted in the total activity measure.

[097] On warm fine days the numbers of bees landing on the feeder increased rapidly for the first 2-3 h, then fell rapidly because the sugar solution was exhausted (all the solution in the bottle had been consumed, and the level in the feeder had fallen to the point where proboscis of a bee could not reach it). On such days the numbers of bees landed on the platform greatly exceeded those flying around it, and most of those which landed were feeding. Observing individual bees indicated that a bee typically fed at the same hole for 1 -2 minutes before flying back to the hive. However, when the sugar solution was running low or had been exhausted, bees probed a number of holes for much briefer periods, and after 3-4 h the numbers of bees still visiting the feeder were very low.

[098] On cooler or cloudy days there was often a delay to the start of feeding and the numbers landing and feeding were lower throughout the day. There were often frequent temporary falls in the extent of feeding, coinciding with the passage of heavy cloud and leading to overall lower landing/flying and feeding/landing ratios. However, landing and feeding continued later into the day, which meant that total activity figures for the day were similar, and the total supply of sugar solution was always consumed.

The pattern of activity when blend EkS2 was placed in the feeder was quite different (Fig. 3).

[099] The total activity was far less than for sugar (about 0.7% of the total number of flying and landing bees for the day; note the scale difference between Fig. 3 and Figs 1 & 2). The number of bees seen flying was always greater than those which landed, leading to much lower landing/flying ratios than for sugar. In the first 2 h there was an increase in the number of flying bees, but very few of them landed after 1 .5 h, and the feeding/landing ratio was less than half of that for sugar. Observations indicated that bees frequently hovered over the feeder (1 -10 cm above the plate), and if they landed they fed only for a few seconds, and frequently moved between holes on the feeder as if seeking better food. [100] On the second day of feeding on blend EkS2 total activity was reduced even further (Fig. 4). There was hardly any landing and no feeding.

These observations suggested that several mechanisms were involved:

a repellent effect that discouraged bees from landing on the plate

a deterrent effect that limited the feeding of those few bees that did land a learned response, and possibly communication between early foragers and those still in the hive, which curtailed activity early on the first day, and led to very little activity on the second day.

[101 ] Table 7 summarises the total activity, landing/flying and feeding/landing ratios, and quantity of formulation consumed, for all the formulations we tested.

Table 7. Total activity, landing/flying ratios, feeding/landing ratios, and weight of formulation consumed for various formulations. Figures are the mean of two replicates, totals for the whole day, based on analysis of single video frames at 30 min intervals.

feeding/landing ratios are followed by (a), the numbers were low and caution is required when comparing the ratios.

[102] It is noted, firstly, that the landing/flying ratios are based on the numbers of bees shown in the total activity column for the corresponding blend and day. The feeding/landing ratios are based on those numbers multiplied by the landing/flying ratios, which will be lower. In general, where the total activity on Day 2 is less than 20 the ratios may be inaccurate (due to the small numbers) but the trend is apparent nevertheless. Secondly, the methods for determining the weight of formulation consumed depended on weighing the dish before and after each day’s run, and comparing the weight loss with that of an evaporation control shielded from feeding. Errors in weighing or microclimate differences from the evaporation control could affect the results, so comparisons between readings in the range of ±20 g (about 4% of the initial weights) should all be treated as indicating no, or almost no, consumption.

[103] Some conclusions can be drawn:

the only formulation apart from sugar where bees consumed all that was supplied was the base with low oil. Even here the bee activity was lower than for sugar (about 30% of the sugar total), and the bees took most of the day to finish the supply. While on the first day this might have been due to the weather, this explanation does not fit the second day.

Apart from the low oil formulation, the only formulations that were consumed at all were very small quantities of the thin base and possibly the normal base on the first but not the second day.

After the low oil base, the highest levels of total activity, and the highest landing/flying ratios, were seen on the various modifications of the base, without volatiles. Whenever volatiles of any type were added, these measures were lower. The only exception to this was with the base + butyl salicylate formulation.

Whenever volatiles were added, total activity on the second day was reduced, often markedly so.

The lowest levels of total activity, largest differences between day 1 and day 2, and lowest landing/flying ratios were noted with the blends, EkF1 , EkS2 and EkS3. These blends all contained one or both of the alarm pheromone components isopentyl acetate and 2-heptanone. This trend may be stronger than the data in Table 7 suggests, because these blends were tested last in the season, when the hives were stronger with larger numbers of bees.

[104] The only way we could get bees to consume variations of the base formulation was to omit most (about 94%) of the canola oil. Even then, there were some reductions in activity and feeding. This suggests that canola oil is a powerful feeding deterrent.

[105] The thickening agent Rhodapol (xanthan gum) may also have deterrent effects, as omitting it led to slight increases in activity and landing/flying ratios. However, the effect is much weaker than for canola oil.

[106] The addition of any volatile or blend of volatiles (with the exception of butyl salicylate) led to reductions in total activity and landing/flying ratios compared to base formulations with no volatiles. This was probably due to the short term repellent effect which any major change in odour produces, combined with the deterrent effects of the oil. The hovering flight a few cm above the disc and the reluctance to land were indicators of this repellence.

[107] The reductions in total activity and landing/flying ratio on the second compared with the first day were due to learning, and this required the presence of one or more volatiles. It did not occur with the base only or thin base formulations. It is likely that during the training phase bees located and fed on the sugar solution through visual stimuli, i.e. the shape and colour of the feeder. These stimuli were still present when the base formulations were supplied, but because the volatility of the base ingredients is very low, there were no olfactory signals to counteract their learning, and the bees did not forget that sugar had been present only two days before. However, the addition of the volatiles provided an olfactory signal indicating the unsuitability of the formulations and this information was learned and probably communicated to other bees in the hive, through social learning.

[108] Although almost any volatiles might serve as indicators of unsuitability, the highest levels of repellence were provided by formulations that contained one or both of the alarm pheromones. When these pheromones were present a characteristic behaviour we termed“scrummaging” was often observed. Bees congregated on the PVC filler cap above the disc, or on the plastic film adjacent to the dish (especially on the shady side), but never on the disc. They then milled about in small and tight groups for periods of a few seconds to a minute. No bees were attacked in this process and nor were the human observers. We formed the impression that the bees were preparing for defence as they would if the hive was threatened, but when no threat emerged the aggregations broke up. Scrummaging bees were not counted as either landing or flying, so this behaviour was a distraction which led to lower levels of those activities.

[109] In compliance with the statute, the invention has been described in language more or less specific to structural or methodical features. The term “comprises” and its variations, such as“comprising” and“comprised of” is used throughout in an inclusive sense and not to the exclusion of any additional features.

[1 10] It is to be understood that the invention is not limited to specific features shown or described since the means herein described comprises preferred forms of putting the invention into effect.

[1 1 1 ] The invention is, therefore, claimed in any of its forms or modifications within the proper scope of the appended claims appropriately interpreted by those skilled in the art.

REFERENCES

The following documents are referred to herein, and their disclosure is incorporated herein by reference:

ABARES (Australian Bureau of Agricultural Resources and Sciences) 201 9. Agricultural overview. December quarter 2018. http://www.aariculture.aov.au/abares/research-topics/aaricultural-commodities/dec-

2018/aqriculture-overview. Accessed 28 February 2019.

Atkins, E. L, R. L. Macdonald, et al. (1975a). Repellent additives to reduce pesticide hazards to honeybees: laboratory testing. Journal of Apicultural Research 14(2): 85- 97.

Atkins, E. L., R. L. Macdonald, et al. (1975b). Repellent additives to reduce pesticide hazards to honey bees: field tests. Environmental Entomology 4(2): 207-210.

Berenbaum, M. R. (2016). Does the honey bee "risk cup" runneth over? Estimating aggregate exposures for assessing pesticide risks to honey bees in agroecosystems. Journal of Agricultural and Food Chemistry 64(1 ): 13-20.

Biesmeijer, J. C., S. P. M. Roberts, et al. (2006). Parallel declines in pollinators and insect-pollinated plants in Britain and the Netherlands. Science 313(5785): 351 -354.

Blight, M. M., M. LeMetayer, et al. (1997). Identification of floral volatiles involved in recognition of oilseed rape flowers, Brassica napus by honeybees, Apis mellifera. Journal of Chemical Ecology 23(7): 1715-1727.

Carreck, N. L. (2017). A beekeeper's perspective on the neonicotinoid ban. Pest Management Science 73(7): 1295-1298.

Collins, A.M. (1996) Use of insect repellents for dispersing defending honeybees (Hymenoptera: Apidae). Journal of Economic Entomology 89, 608-613.

Dotterl, S. and N. J. Vereecken (2010). The chemical ecology and evolution of bee- flower interactions: a review and perspectives. Canadian Journal of Zoology 88(7): 668-697.

Erber, J., S. Kierzek, et al. (1998). Tactile learning in the honeybee. Journal of Comparative Physiology A-Sensory Neural and Behavioral Physiology 183(6): 737- 744.

Estesen, B. J., N. A. Buck, et al. (1992). Residual life and toxicity to honey-bees (Hymenoptera: Apidae) of selected insecticides applied to cotton in Arizona, Journal of Economic Entomology 85(3): 700-709.

Farina W. M., C. GrOter and P. C. Diaz (2005) Social learning of floral odours inside the honeybee hive. Proceedings of the Royal Society Series B. 272: 1923 - 1928. Free, J. B., J. A. Pickett, et al. (1985). Repelling foraging honeybees with alarm pheromones. Journal of Agricultural Science 105: 255-260.

Giurfa, M., J. Nunez, et al. (1994). Odor and color information in the foraging choice- behaviour of the honeybee. Journal of Comparative Physiology A -Sensory Neural and Behavioral Physiology 175(6): 773-779.

Giurfa, M., J. Nunez, et al. (1995). Color preferences of flower-na^'ive honeybees. Journal of Comparative Physiology A-Sensory Neural and Behavioral Physiology 177(3): 247-259.

Guerrieri, F., M. Schubert, et al. (2005). Perceptual and neural olfactory similarity in honeybees. Plos Biology 3(4): 718-732.

Hadagali, M. D. and C. L. Suan (2017). Advancement of sensitive sniffer bee technology. Trends in Analytical Chemistry 97: 153-158.

Flagler, J. R. and S. L. Buchmann (1993). Honey-bee (Hymenoptera: Apidae) foraging responses to phenolic-rich nectars. Journal of the Kansas Entomological Society 66(2): 223-230.

Hardstone, M. C. and J. G. Scott (2010). Is Apis mellifera more sensitive to insecticides than other insects ? Pest Management Science 66(1 1 ): 1 171 -1 180.

Kiljanek, T., M. Niewladowska, et al. (2016). Pesticide poisoning of honeybees: a review of symptoms, incident classification and causes of poisoning. Journal of Apicultural Science 60(2): 5-24.

Klein, A. M., B. E. Vaissiere, et al. (2007). Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society B-Biological Sciences 274(1608): 303-313.

Krupke, C. H., G. J. Hunt, et al. (2012). Multiple Routes of Pesticide Exposure for Honey Bees Living Near Agricultural Fields. Plos One 7(1 ).

Lehmann, M., D. Gustav, et al. (201 1 ). The early bee catches the flower - circadian rhythmicity influences learning performance in honey bees, Apis mellifera. Behavioral Ecology and Sociobiology 65(2): 205-215.

Manning, R. (2018). Chemical residues in beebread, honey, pollen and wax samples collected from bee hives placed on canola crops in Western Australia. Journal of Apicultural Research 57(5): 696-708.

Mayer, D. F., J. D. Lunden, et al. (2001 ). Field evaluation of non-pesticide chemicals as honey bee repellents. In Hazards of Pesticides to Bees. Ed. L. P. Belzunces, C. Pelissier and G. B. Lewis: pp. 159-168.

Mishra, S. and R. C. Sihag (2009a). Efficacy of some chemicals as repellents against two honey bee species, Apis mellifera L. and Apis florea F. in semi-field trials. Journal of Apicultural Science 53(1 ): 53-66. Mishra, S. and R. C. Sihag (2009b). Efficacy of some chemicals as repellents against two honey bee species, Apis mellifera L. and Apis florea F. in field trials. Journal of Apicultural Science 53(2): 21 -29.

Mishra, S. and R. C. Sihag (2010). Efficacy of some chemicals as repellents against two honey bee species, Apis mellifera L. and Apis florea F. in semi-field trials. Journal of Apicultural Science 54(1 ): 21 -34.

Moore, D., D. Siegfried, et al. (1989). The influence of time of day on the foraging behavior of the honeybee Apis mellifera. Journal of Biological Rhythms 4(3): 305-325.

Moritz, R. F. A. and S. Erler (2016). Lost colonies found in a data mine: Global honey trade but not pests or pesticides as a major cause of regional honeybee colony declines. Agriculture Ecosystems & Environment 216: 44-50.

Mullin, C. A., M. Frazier, et al. (2010). High levels of miticides and agrochemicals in North American apiaries: Implications for honey bee health. Plos One 5(3).

Potts, S. G., J. C. Biesmeijer, et al. (2010). Global pollinator declines: trends, impacts and drivers. Trends in Ecology & Evolution 25(6): 345-353.

Reetz, J. E., W. Schulz, et al. (2016). Uptake of neonicotinoid insecticides by water foraging honey bees (Hymenoptera: Apidae) through guttation fluid of winter oilseed rape. Journal of Economic Entomology 109(1 ): 31 -40.

Rhodes, C. J. (2018). Pollinator decline - an ecological calamity in the making? Science Progress 101 (2): 121 -160.

Robertson, L. N. and J. W. Rhodes (1992). Honey bee ( Apis mellifera L.) deaths near sprayed cotton and observations on bee foraging in flowering cotton (Hymenoptera: Apidae). Journal of the Australian Entomological Society 31 : 243-246.

Sandoz, J. C. (201 1 ). Behavioral and neurophysiological study of olfactory perception and learning in honeybees. Frontiers in Systems Neuroscience 5: 98.

Sandoz, J. C., M. H. Pham-Delegue, et al. (2001 ). Asymmetrical generalisation between pheromonal and floral odours in appetitive olfactory conditioning of the honey bee ( Apis mellifera L.). Journal of Comparative Physiology A-Sensory Neural and Behavioral Physiology 187(7): 559-568.

Vanegas, M. (2017). The silent beehive: How the decline of honey bee populations shifted the Environmental Protection Agency's pesticide policy towards pollinators. Ecology Law Quarterly 44(2): 31 1 -341 .

Woodrow, A. W., N. Green, et al. (1965). Evaluation of chemicals as honey bee attractants and repellents. Journal of Economic Entomology 38: 1094-1 102.

Zhang, S. W., S. Schwarz, et al. (2006). Honeybee memory: a honeybee knows what to do and when. Journal of Experimental Biology 209(22): 4420-4428.

Claims

1 . A method of conditioning a honeybee, comprising the steps of:

c) exposing the honeybee to a deterrent to feeding that is unpleasant to a honeybee but not harmful;

d) exposing the honeybee to a signal that can be detected by the honeybee and will be associated with the deterrent by the honeybee;

whereby the honeybee expects the deterrent to be present when the signal is detected and so will be conditioned not to feed when the signal is detected.

2. A method according to claim 1 wherein the honeybee is exposed to the deterrent to feeding by way of ingestion of a compound and ingestion is unpleasant.

3. A method according to claim 2 wherein the compound is selected from the group consisting of vegetable oils, minerals oils and phenols.

4. A method according to claim 3 wherein the vegetable oil is selected from the group consisting of olive oil, sesame oil, peanut oil, canola oil, cottonseed oil, corn oil, soybean oil, mineral oil, as well as methylated and ethoxylated forms of these oils, or mixtures thereof.

5. A method according to claim 4 wherein the vegetable oil is canola oil.

6. A method according to any preceding claim wherein the signal comprises one or more volatile compounds.

7. A method according to claim 6 wherein the signal comprises a terpenoid.

8. A method according to claim 6 wherein the signal comprises a volatile aromatic compound comprising a hydroxyl, ether, ester or carbonyl group.

9. A method according to claim 6 wherein the signal comprises a green leaf volatile.

10. A method according to claim 6 wherein the signal is an alarm pheromone.

1 1 . A method according to claim 10 wherein the signal is 2-heptanone or an analogue thereof.

12. A method according to claim 10 wherein the signal is isopentyl acetate or an analogue thereof.

13. A method according to any preceding claim wherein a slow release mechanism is employed to prevent overly rapid loss of the volatile compounds.

14. A method according to any preceding claim, wherein other honeybees are conditioned to associate the signal with the deterrent through social learning.

15. A method of repelling honeybees from a food source, comprising the steps of: a) conditioning at least one honeybee by a method according to any one of claims 1 to 14;

b) providing a signal that can be detected by the honeybees and which has become associated with the deterrent; and

c) applying the signal to the food source to repel honeybees.

16. A method as claimed in claim 15, wherein honeybees are conditioned and the signal is applied thereafter.

17. A method as claimed in claim 16 wherein insecticide is applied together with the signal.

18. A method as claimed in claim 15 wherein the signal is applied together with a deterrent to feeding to condition and then repel honeybees with a single application.

19. A method according to claim 15 wherein the food source is a crop.

20. A method according to claim 16 wherein the crop is an insecticide-treated crop.

21 . A kit comprising a first composition that provides a deterrent to feeding, a second composition that provides a signal that can be detected by the honeybee and will be associated with the deterrent, and instructions for use.

22. A kit according to claim 21 further comprising additional compositions that provide a signal that can be detected by the honeybee and will be associated with the deterrent.

23. A kit according to claim 21 or 22 further comprising an insecticide.