J. K. Trigg* and N. Hill

An evaluation of a new plant-based biting insect repellent (PMD) with the principal active ingredient p-menthane-3,8-diol is presented. Laboratory trials testing 50% active ingredient formulations of the repellent on a human forearm show good repellency against the malaria vector Anopheles gambiae Giles s.s. with an average ED_M value of 0.68 μL/cm² and effective protection from biting for around 5 h. These results compare favourably with a 20% deet product, (ED₉₀ value of 0.48 μL/cm² with 5 h protection), and are far superior to the best known existing plant-based product, citronella (ED₉₀ value of 1.37 μL/cm² and 2 h protection). PMD was also found to give total protection against the biting midge, Culicoides variipennis Coquillett, for at least 6 h at 035 μL/cm². When evaluated against the deer tick, Ixodes ricinis L., attachment and subsequent feeding of /. ricinis nymphs was significantly reduced on rabbit ears treated with 036 μL/cm² of PMD. Against the stable fly, Stomoxys calcitrans L, an application of 0.5 mL PMD to a human forearm reduced biting to 6% of the control biting level. It is concluded that PMD is a broad spectrum repellent with considerable potential as means of personal protection against nuisance biting and insect-borne disease.

Insect repellents are widely used as a means of personal protection against biting arthropods, the main motive usually being avoidance of nuisance and discomfort. Personal protection measures, including the use of repellents, are also important in reducing the risk of contracting insect vector borne disease (Curtis, 1992).

The majority of commercial insect repellent preparations contain the chemical diethyltoluamide (deet), first synthe­sized in 1954 (McCabe et al., 1954). There are disadvantages associated with deet usage which stem from its activity as a solvent of paints, varnishes and certain plastics and synthetic fabrics. These, together with concerns arising from reports of the occasional toxicity of deet especially in children (Zadikoff, 1979; Roland et al., 1985), have highlighted the need for alternatives to deet. There has been much research into natural plant extracts, both prior to (Granett, 1940) and since (Opuku et al., 1986: Sharma and Dhiman, 1993) the advent of synthetic repellents. Citronella oil has long been used in commercial preparations and is still popular in India, though generally rated less effective than repellents with synthetic active ingredients (Curtis et al.. 1987). The Chinese repellent, quwenling, first described by Li et al. (1978) was reported to be more effective than deet (data cited by Lu Baolin in Curtis et al., 1987), though subsequent studies by Schreck and Leonhardt (1991), Collins et al. (1993) and Curtis et al. (1994) suggested that whilst an effective repellent, it is somewhat less persistent than deet. Quwenling is made from the waste distillate after extraction of oil from the lemon eucalyptus plant {Euca­lyptus maculata citriodon). The principal active component of quweniing is p-menthane-3,8-diol (Li et al., 1978; Schreck and Leonhardt, 1991).

More recently, an extraction process developed at University College, London, utilizing eucalyptus oil has yielded a new repellent preparation (trade name 'Mosiguard Natural'). The active component (50%) is principally p-menthane-3,8-diol (PMD) with additional isopulegol and citronellol and the repellent is formulated as a patented mixture of isomers of each. The repellent, referred to as PMD in this paper, was evaluated in the laboratory against four biting arthropods.

Mosquitoes. Three formulations of PMD (liquid, stick and gel) were evaluated, in comparison with an Autan™ stick (20% deet, Bayer AG, Germany) and citronella oil (50% active ingredient) by a dose response method (modified from Curtis et al., 1987) using laboratory reared Anopheles gambiae. Twenty hungry female mosquitoes were placed in each of three netting cages (45 x 45 x 45 cm), held under optimal mosquito environmental conditions. The experi­menter's untreated bare forearm, with the hand protected by a latex glove, was introduced into the first cage. The numbers of mosquitoes probing on the skin surface after 30 s were counted, then shaken off before they had taken a blood feed. This control test was repeated using the second and third test cages. A small measured dose of repellent was then applied to one forearm and the procedure repeated. Incremental doses of repellent were applied and tested until no bites were recorded. This minimum dose for 100% repellency then remained untouched on the skin and was re-tested at hourly intervals to give a measure of repellent longevity. As a check for the continued feeding avidity of the mosquitoes, the untreated control arm was introduced into the cages at intervals between the introductions of the treated arm. For each experiment, the data were used to compute the dose which would give 90% protection (i.e. 10% of the normal number of bites) using probit analysis.

Stable fly. The spray formulation of PMD was tested on a human forearm, using a similar method to that described above, for repellency against laboratory reared Stomoxys calcitrans. The method differed in three ways from the mosquito trial; (1) 40 insects instead of 20 were put in each of three cages, (2) the arm was placed into each cage for a period of 1 min rather than 30 s for each test and (3) due to the painful nature of stable fly bites, the repellent was applied to the forearm at a measured dose from the onset of the experiment rather than using the dose response method. Repellent was first applied at a dose of 0.35 mL which approximated to the ED₉₀ dose computed in the mosquito trial. In a second test, the dose was increased to 0.5 mL. On the day of testing, the flies were 15 days old and had been fed daily on blood since emerging as adults, but had not fed on the day of the test.

Midges. The spray formulation of PMD was tested on a human forearm for repellency against laboratory reared Culicoides variipennis. An area measuring 6cmxl5cm was marked out on the ventral surface of each forearm. One arm was treated with repellent whilst the other arm remained as an untreated control. 0.032 mL was applied evenly over the 90 cm² area on one arm (at a dose rate of 0.36 u.L/cm²).

Tests were made hourly using three batches of 10 hungry female midges for each arm. The midges were housed in clear plastic pots (4 cm diameterx6cm depth) which had been prepared for the experiment with a small hole made in the side of each pot and then plugged with a cotton wool bung. The open end of each pot was covered with a square of netting, (through which the midges could feed) and was secured with an elastic band. For each test, 10 C. variipennis were introduced into each of six pots using an aspirator and left to settle for 10 min. Three pots were sequentially held firmly against the arm at intervals along the treated area for a period of 3 min each. The number of bites received during each time period was recorded and the process then repeated with the remaining three pots, held at intervals along the marked area on the untreated control arm.

In initial experiments, testing commenced immediately after the applied repellent had dried and continued hourly for up to 6 h post-application. The results from these initial trials suggested that the repellent gave 100% repellency for the full 6 h after application with an average control biting rate of 81%. On this basis, later experiments commenced 5 h post-application and continued up until 9 h had elapsed, or alternatively, the repellent was applied at a lower dose of 0.016 mL and tests lasted for 6 h.

Ticks. The PMD spray formulation was evaluated at the Central Veterinary Laboratory (CVL), Addlestone. Surrey, against Ixodes ricinis nymphs collected from tick infested areas of southern England. Repellent was applied to laboratory rabbits' ears which had been shaved 24 h earlier and were found to have an average surface area of 91 cm^:. 0.032 μL repellent was applied to each ear of three rabbits (at a rate of 0.36 μL/cm²). Repellent was mixed with alcohol for spreading, and administered with a micropipette then spread evenly with the fingertips over inner and outer ear and then left to dry.

An adaptation of the CVL standard tick rearing method was used for this trial. For each rabbit, a cotton earbag was fitted over each ear and attached securely to the base of the ear with surgical tape. 20 /. ricinis nymphs were then introduced into each bag before folding over the open end and sealing it with surgical tape. Earbags were taped

Test repellent PMD liquid (50% ai) PMD gel (50% ai) PMD stick (50% ai) Autan stick (20% deet) Citronella liquid (50% ai)

together to prevent removal by grooming. For repellent treated rabbits, earbags were attached and ticks were introduced 45 min after repellent application. The rabbits were checked twice over the next 36 h period to ensure that they were comfortable and that earbags were still attached. /. ricinis nymphs will usually take about a day to attach and start feeding on their host, remaining attached until fully engorged (up to 7 days), then dropping off. After 43 h, each rabbit was inspected in turn. The earbags were opened and gently rolled down and a count was made of the number of ticks attached, feeding or fed. Any dead ticks were recorded and removed from the earbag. This process was repeated daily until all ticks, alive or dead, had been removed.

RESULTS

The ED₉₀ values for each repellent are given in Table 1. Persistence of the minimum dose of each repellent applied to the forearm is shown in Fig. 1. The three formulations of PMD were slightly less effective per unit volume than deet but were all at least twice as effective as citronella and

201-

0 12 3 4 5

Time / Hours

Figure 1. Repellent longevity. The decline in repellency with time following an application, to the forearm, of the minimum dose of repellent required to achieve total protection against caged An. gambiae biting. + PMD stick; ■ PMD gel; * PMD liquid; ♦ citronella liquid; x DEET stick.