The mite Varroa destructorAnderson & Trueman, 2000 (Acari: Varroidae) is a hemophagous ectoparasite of honeybee (Apis mellifera) distributed worldwide (with the notable exception of Australia; Rosenkranz et al., 2010). This parasite is currently one of the greatest menaces for apiculture (Rosenkranz et al., 2010) due to economic losses and high impact on colonies health. This mite infects all the stages of honeybee, having a phoretic phase on adult bees and a reproductive phase in brood cells. Clinical symptoms are not visible at low infestation rates, and in the absence of efficacious control methods, the colonies of A. mellifera can collapse within a few years, in temperate climates (Boecking & Genersch, 2008). Further, V. destructor is responsible for the transmission of viruses to their hosts (Rosenkranz et al., 2010).

In this study two chemical products commonly used for controlling V. destructor were compared: Apivar (a.i. amitraz, a “hard” synthetic chemical) and Apiguard (a.i. thymol, included in the group of essential oils). Amitraz was one of the earliest miticides tested against V. destructor (Hollingworth, 1976). Apivar possesses some advantages, such as the simplicity of application and the low economic cost. However, the main disadvantages of this product are its limited efficacy after continuous use due to the development of resistances (Trouiller, 1998; Elzen et al., 1999) as well as the accumulation of residues in bee products (Martel et al., 2007; Lodesani et al., 2008; Chauzat et al., 2009; Johnson et al., 2009). Nowadays, alternative methods for V. destructor control based on natural products (such as organic acids and essential oils) are being tested. Thymol (2-isopropyl-5-methylphenol) is one of those products of natural origin which have demonstrated to control varroosis both in field and in laboratory conditions (Colombo & Spreafico, 1999; Floris et al., 2004; Akyol & Yeninar, 2008; Gashout & Guzman-Novoa, 2009). Apiguard acts as an efficient miticide with a low risk of residues in bee products and resistance in V. destructor populations (Fries et al., 1991; Floris et al., 2004; Bogdanov, 2006). The aim of the present study was to compare the efficacy of these two commercial products: Apiguard (thymol, essential oil) and Apivar (amitraz, formadimine), during different seasons (spring and autumn-winter) in the Mediterranean climate.

We present here the results of the study conducted in an apiary located at the University of the Balearic Islands in Majorca (39º 38’ 30.20’’N, 2º 38’ 23.57’’ E), Western Mediterranean. The apiary had 15 Langstroth standard colonies of Apis mellifera iberica, naturally infested with Varroa destructor mites. Before the trial, mite infestation level was monitored to obtain three experimental groups, as homogeneous as possible. For testing miticides, treated and untreated (control) hives were randomly selected in the apiary in order to avoid any bias due to position of the hive in the apiary. Five colonies were treated with Apivar, five with Apiguard, and other five remained untreated as control. Distance between hives was the conventional one used in the apiary (1.5-2.0 m from hive to hive).

All treatments were applied following the manufacturer’s instructions. For the Apivar (Veto-Pharma SA, Villebon-sur-Yvette, France) treatment, two plastic strips (21 cm × 3.5 cm each one) were inserted in the brood chamber of each hive and left for 6 weeks. In the case of the Apiguard (Vita Europe Ltd., Basingstoke, UK) treatment, one tray (10 cm diameter) was located in the top of brood frames of each hive and left for 14 days and then it was replaced by a second tray, which was left for 2 weeks more.

The assays of 2010 and 2011 started on 8 April and 11 November, respectively. The treatments were applied on 15 April 2010 and 18 November 2011. Apiguard treatment was repeated on 29 April 2010 and 2 December 2011 and removed on 13 May 2010 and 16 December 2011, whereas Apivar treatment was not repeated and was removed on 27 May 2010 and 30 December 2011. Each colony was equipped with a modified bottom board, provided with a removable tray where a vaseline sticky paper was inserted and removed every 7 days.

Counting fallen mites is a non-destructive method and it is considered good indicator of colony infestation to be used periodically (Dietemann et al., 2013). Fallen mites were counted weekly, starting one week before treatment in order to estimate the initial population of V. destructor in each colony. Two different parameters were studied: (i) the reduction of number of fallen mites (NFM), calculated from the difference between the number of mites obtained at the beginning and at the end of the assay; (ii) the efficacy of treatments was evaluated by using the percentage of reduction of daily fallen mites (%R), according to the following formula (Henderson & Tilton, 1955; Satta et al., 2005):

%R = 100[1 – (Bc * At/Bt * Ac)]

where Bt and At are the data on the average daily mite fall in treated colonies before (Bt) and after treatment (At), and Bc and Ac are the same parameters in the control group. This evaluation method is probably more precise than others using the % change in infestation, because %R incorporates variation in mite population caused by several variables, such as natural mortality, natality, immigration or emigration (Floris et al., 2001).

Data were analyzed by Student’s t-test (for parametric statistics). In the case of spring data, a logarithmic transformation was applied in order to improving the normality of variables (SPSS v. 20.0, SPSS Inc., Chicago, IL, USA). A level of p≤ 0.05 was accepted as significant, and p≤ 0.10 as indicative of tendency.

The results of the treated groups showed a similar pattern in both seasons (Fig. 1). We observed nnly one peak for Apivar, in the first week of product application (NFM was multiplied by 4.8 in spring and by 6.2 in autumn), showing a tendency between Apivar and control groups in the autumn assay (t = -1.447, p = 0.095). Whereas Apiguard showed two high points, one the first week (increased in 2.4 in spring and 5.7 in autumn) and other in the third week (multiplied by 2.2 in spring and by 2.9 in autumn), which coincides with the replacement of the treatment. Significant differences between Apiguard and control groups in the autumn assay were found for the first week (t = -3.681, p = 0.004) and tendency was found for the third week (t = -1.817, p = 0.053). Moreover, in the autumn assay significant differences were found between Apivar and Apiguard groups in the third week (t = -1.594, p   0.007).

Figure 1. (a) Spring and (b) autumn assays. Data represent the means of number of fallen mites (NFM) ± SEM, before, during and after treatments. Values of weekly average NFM at t0 (week before treatment) were set at 100, and the values of average NFM during the following weeks were represented as percentage of the NFM for week t0. A.U.: arbitrary units. ** indicates significant differences at the p<0.05 level with control group. * indicates tendency at the p<0.1 level with control group. ¤ indicates significant differences at the p<0.05 level with Apivar group.

At the end of the treatments in the post-treatment week, the average NFM during spring assay was reduced 89.8% in Apiguard treatment and 64.3% in Apivar treatment, being the difference between Apiguard and control colonies statistically significant (t = 3.302, p = 0.008), and it showed a tendency (t = 1.492, p = 0.089) in the case of Apivar and control (Fig. 1). Regarding to autumn assay, the average NFM in Apivar colonies was reduced a 17.9%, meanwhile in Apiguard-treated ones the reduction was higher, a 30.8%. However, in the untreated group the average NFM was increased a 5%. There were not statistically significant differences between groups, but there was a tendency (t = 1.405, p = 0.099) between control and Apiguard group (Fig. 1).

Table 1 shows the %R calculated for both treatments. Thymol-based miticide product (Apiguard) showed higher efficacy compared to amitraz-based product (Apivar). In spring the former showed a %R = 76.9%, meanwhile for the same period, Apivar showed only a %R = 65.1%, although there were not statistical significant differences between treatments (t = -0.587, p = 0.286 in spring assay and t = -0.565, p = 0.294 in autumn). Means for both treatments were ~40% lower in autumn respect to spring. The %R was also variable among colonies (Table 1).

The present work is the first conducted in the Western Mediterranean to test the efficacy of commonly used anti-varroosis treatments. Our results obtained for thymol (Apiguard) either in spring and autumn agreed with other published works where thymol showed efficacies varying from 40% to more than 90% (Marchetti & Barbattini, 1984; Chiesa, 1991; Imdorf et al., 1995; Gregorc & Planinc, 2005; Loucif-ayad et al., 2010). We detected a drop of efficacy of thymol in late autumn (November-December), probably due to the low external temperatures (13ºC of average daily external temperature) that hampered an optimal volatilization of the product. This observation was also reported by Ellis (2001), showing that Apiguard has a good performance at 15-20⁰C, and lose efficacy when the temperatures are lower. In regards to amitraz, our results obtained in spring are similar to those published by Floris et al. (2001) and Pohorecka et al. (2011), who achieved an efficacy ~70%. Other authors have reported higher efficacies ranging from 82% (Chuda-Mickiewickz et al., 2007; Loucif-ayad et al., 2010) to 99.5% (Faucon et al., 2007). One possible explanation to this lower efficacy of amitraz that needs to be tested in future works is the possibility that V. destructor populations in the Balearics are resistant to amitraz due to the continuous application of this product during the last 30 years (first years with wooden frames impregnated with amitraz).

Our results also showed a high variability of treatment efficacy among colonies. This fact was already observed by other authors in Italy, Swiss and Poland (Marchetti & Barbattini 1984; Imdorf et al., 1995; Colombo & Spreafico, 1999; Floris et al., 2004; Chuda-Mickiewickz et al., 2007). Variable efficacy among colonies could be due, in the case of thymol, to microclimatic conditions related to each colony that cause different volatilization rate and thus dispersal inside the colony; in the case of amitraz, there is emergence of resistance of V. destructor to this product (Trouiller, 1998; Elzen et al., 1999).

Based on the results of the present study, thymol could represent an alternative in integrated control strategies against V. destructor, having a low risk of residues and resistance, although it may be more expensive than amitraz products (in Spain Apiguard is 86% more expensive than Apivar). Our results also pointed out that spring is the most appropriate time to apply thymol treatments to colonies managed under Mediterranean climate, whereas late autumn (November-December) seemed to be unappropriated time for an effective use of thymol. Finally, future works may include the test of other essential oils to compare with the anti-varroa effect of thymol. Further studies focused on the resistance status of Varroa destructor population in the Balearic Islands to widely products such as amitraz are also needed.