Tetranychus urticae colonies

Tetranychus urticae has been collected in 2022 on Fragaria sp. in an organic farm at Balestrate (Palermo, Italy) (38° 1′32.84″N 13° 2′2.29"E) and reared on potted Phaseolus vulgaris L. (Fabaceae) plants, in greenhouse conditions: T° 14–38°C, RH 40–80%. Once a year, T. urticae from weeds or cultivated plants were added to guarantee the genetic variability of the population. New bean plants were infested with the two-spotted spider mites when the cotyledonary leaves appeared.

Neoseiulus idaeus and Phytoseiulus persimilis breedings

Neoseiulus idaeus and P. persimilis were collected in the same year, locality and host plant as T. urticae. Both phytoseiids bred in a conditioned room at 25± 1°C, 16:8 L:D photoperiod and 75±5% RH. Once a year, predatory mites collected in field were added, to avoid genetic depletion of the strains.

Breeding arenas consisted of a black plexiglass plate (100 x 100 x 4 mm). Each element was placed on a plastic container (Ø 7 cm) containing distilled water (Figure 1). A groove of 5 mm wide and 2 mm deep, marked a circular area (cm 6 Ø), which represented the breeding surface (arena). The groove delimiting the arena was filled with a mixture of Canada Balsam + Castor Oil (1:2 respectively) to prevent the mites from escaping. Each arena had a central hole (Ø 5mm) through which a piece of cotton wool passed and became immersed in the water of the underlying container. This serves to provide an additional water support for the phytoseiids, especially when alternative foods are used (i.e. pollens). A 2.5 cm long black cotton thread, frayed in the distal part, was generally placed on the wet cotton wool to encourage the oviposition of females.

FIGURE 1.

Breeding arena: 1—plexiglass plate; 2—groove filled with Canada Balsam + Castor oil; 3—breeding surface; 4—piece of wet cotton wool; 5—cotton wool immersed into the water; 6—plastic container containing distilled water.

Different developmental stages of T. urticae were used as food in the breeding arenas. The two-spotted spider mites were brushed into arenas from infested bean leaves three times a week using a soft painting brush (n. 18). In addition, some Oxalis pes-caprae L. and Carpobrotus edulis (L.) pollens were supplied once a week in N. idaeus breeding arenas. Pollens were collected at the time of flowering and stamens were stored in a freezer (-18°C).

Phytoseiid mites were transferred, every three weeks, into new clean arenas.

Tetranychus urticae eggs obtaining for the experiments

Heavily infested bean leaves were introduced into a plexiglas box custom-made (cm 50 x 50 x 50h), containing water and sodium hypochlorite (1% v/v). Four non-return valves were present at the base of the box connected to an air-compressor (ABAC Sirio 220, Italy) (Figure 2). The compressed air that came out of the valves for about 5 minutes at a pressure of 62.05 kPa, shook and washed the infested leaves and all stages of T. urticae rose to the water surface. Afterwards, adding water slowly, the level reached a side opening connected to a specially constructed funnel which ended with an opening of 8 cm. Four sieves, 400µ, 300µ, 200µ and 100µ were put in column under the funnel. All mites present on the water surface inside the box, fell into the funnel, and were captured by the sieves. In the last sieve, only T. urticae eggs were collected. Eggs were put into Eppendorf tubes (5ml) with distilled water and sodium hypochlorite (0.5%) and maintained in a fridge at 7–8°C for at least a week. Hatching of preserved eggs was of 92% after 3 days and remained intact for about 8 days.

FIGURE 2.

System for collecting Tetranychus urticae eggs: 1—plexiglass box; 2—infested bean leaves by T. urticae; 3—water; 4—air compressor; 5—non-return valve; 6—sieves (400µ, 300µ, 200µ, 100µ from top to down).

Biological and life-table parameters

Experimental unit (plexiglass arena): Tests on post-embryonic development, life table parameters and predation rate, were carried out using black plexiglass elements, like those used for mass-rearing, but smaller in size (60 x 60 x 4 mm) delimiting a circular arena of 3 cm in diameter.

Post-embryonic development: To obtain coetaneous eggs of the predators, 40 females from mass-rearing were transferred in a plexiglas arena with abundant prey for 24 hours. Eggs laid in this time interval were used in these tests. A fresh egg of each predator (max 24h old) was transferred into an arena in the presence of 20 eggs of T. urticae (max 3 days old). Each test consisted of 25 repetitions. The stage of the phytoseiid and the number of T. urticae eggs eaten, were checked daily at the same time until predators reached adulthood. Preyed eggs of T. urticae were counted at 24h intervals and new eggs were added to have 20 eggs of prey per predator available. After 3 days, all prey eggs were replaced with new ones. Tests were carried out in a conditioned room at 25± 1°C, 16:8 L:D photoperiod and 75±5% RH.

Life table parameters

The pre-oviposition, oviposition, post-oviposition periods, survival, and oviposition rate of both predator females, have been ascertained on females obtained by the postembryonic development tests. Each young female was transferred into an arena with a male and observed until the death of the female. Dead males were replaced with new ones from mass-breeding arenas. Various developmental stages of T. urticae, were daily abundantly brushed into arena as food for female and male. Each test was replicated 18 times because this was the number of females obtained from the postembryonic development tests. Observations were made daily and the phytoseiid eggs were counted and removed.

Predation rate of unmated females: To obtain unmated females, 50 eggs of each predator were transferred to as many arenas with abundant prey as food. After reaching adulthood, a young unmated female was transferred into a new arena in the presence of 20 eggs of T. urticae. Preyed eggs were daily detected and new T. urticae eggs were added, to have 20 eggs of prey per day available for each predator. Each bioassay consisted of 25 repetitions and lasted 7 days.

Predation rate of mated females

To obtain coetaneous young females and males, 50 eggs of each phytoseiid species were transferred into an arena with abundant T. urticae as food until attaining adulthood. Twenty new emerged females and 20 males were transferred into a new arena with abundant prey for 3 days, to guarantee fertilization of females. Afterwards, a female of each predator was transferred into experimental arena in the presence of 40 eggs of T. urticae. Preyed eggs were daily counted and new T. urticae eggs were added, to have 40 eggs of prey/day/predator female available. Each test consisted of 20 repetitions and lasted 10 days.

Data analysis

The duration of various developmental stages, oviposition rate, lifespan and predation rate were calculated. The analysis was carried out using days for the period of observations, number of eggs laid per day as the frequency and species as variable (Tsolakis et al. 2016). The Binary Logistic Regression (BLR) was performed on postembryonic development and preoviposition periods because the frequencies were predominantly concentrated on 2 values (1 day and 2 or more days). For the oviposition, post oviposition and lifespan the Survival Analysis for Non-Repairable systems was adopted, because the event (pass from the previous to the next period or to death) occurred only once. For oviposition rate, Survival Analysis for Repairable system was adopted: the female produces, during the observation period, one or several events in each time interval (egg deposition). The analysis was conducted using days for the observation period, number of eggs laid per day as the frequency and species as variable. The Anderson-Darling Goodness of fit test before adopting a parametric analysis has been performed. If the parametric approach was not acceptable, the nonparametric test was adopted (Meeker et al. 2021).

When the distribution of data was spread on several items (i.e. reproductive rate and consumption rate) and after the test for a normal distribution, the General Linear Model (GLM) has been adopted. For all analyses the error distribution is assumed to be normal.

To investigate on the rate of transformation of preyed eggs in laid eggs, we included in the statistical model (GLM) ‘laid eggs/preyed eggs’ as the response variable (qualitative), in respect of ‘preyed eggs’ (quantitative covariate), ‘days’, ‘individuals’ and ‘Phytoseiids’ (categorical covariates). The Box-Cox procedure was adopted to identify the appropriate transformation, indicating the square root transformation and it is that prescribed by the theory as the normalizing transformation. A full second order model has been considered, and a stepwise procedure has been applied, to maintain in the model only significant terms.

The demographic parameters were calculated following Carey (1993). The intrinsic rate of increase (r_m) was calculated based on the Euler equation as adapted to this aim by Dublin and Lotka (1925): . The Lagrange method has been adopted for calculating the intrinsic rate of increase (r_m).

The other demographic parameters were calculated as follows: Finite Rate of Increase λ = e^r_m; Net Reproductive Rate R₀ = ; Gross Reproductive Rate GRR = ; Mean Generation Time T = ; Doubling Time DT = . Where l_x is the proportion of females surviving to age x and m_x is the mean number of female's progeny per adult female at age x.

The Jackknife method was used to create pseudo-values to estimate the variability in all life table parameters (Meyer et al. 1986). For the Jackknife procedure, one of the n replicate females (the i^th female, i = 1, 2, . . ., n) from the original data set was omitted and demographic parameters were recomputed using data from the remaining n - 1 females. A total of 1,444 replicates were run to create pseudo-values. The results, after the test for a normal distribution, were analysed with Analysis of Variance (ANOVA), assuming a normal distribution of the error, followed by Tukey's test and Student's t test using Minitab^® 17 software.

Postembryonic development

The first three stages - egg, larva and protonymph - lasted approximately one day with no significant differences between the two species. The majority of deutonymphs lasted one day too, but more than 30% of those requires 2 days. The binary logistic regression (BLR) showed that only the factor Stage presents significant differences (χ²=24.85; df=3; p<0.001) and from the analysis of the Odds Ratio it emerges that the stage significantly different from the others is the deutonymphal one (Table 1). No differences were noted between the female and male postembryonic development for both species (χ²=0.00; df=1; p=0.980) and no differences between species (χ²=0.12; df=1; p=0.732). All individuals of N. idaeus reached the adult stage (100%), while a slight mortality was noted in P. persimilis (8%). The sex ratio was 2.57:1 and 3.6:1 (female: male) for N. idaeus and P. persimilis, respectively (Table 1).

TABLE 1.

Postembryonic development in days of Neoseiulus idaeus and Phytoseiulus persimilis fed on eggs of Tetranychus urticae. Binary logistic regression was applied on the data. Different letters or asterisks denote significant differences (p<0.05).

Preoviposition, oviposition, postoviposition and longevity

Statistical analysis (BLR) showed that preoviposition period was significantly longer for N. idaeus females (χ² = 16.45; df=1; p<0.001) (Table 2). Regarding the oviposition period, the survival analysis for non-repairable systems, showed no significant differences between the two species (χ² = 3.51; df=2; p=0.173), while the post oviposition one was longer for N. idaeus (χ² = 97.07; df=2; p<0.001). Consequently, the longevity was longer for the latter species (χ² = 37.71; df=2; p<0.001). Neoseiulus idaeus laid fewer eggs per female than P. persimilis considering both the whole oviposition period (F=19.25; df=1, 34; p<0.001) and the daily oviposition rate (eggs/females/day) (F=73.38; df=1, 1299; p<0.001) (Table 2).

TABLE 2.

Oviposition rate and longevity of Neoseiulus idaeus and Phytoseiulus persimilis females obtained by the postembryonic development tests (18 females) on various stages of Tetranychus urticae (Mean ± SE). Different letters denote significant differences between the two species for each biological parameter. (1) - Binary Logistic Regression; (2) – Weibull; (3) - ANOVA.

Demographic parameters

The life table parameters of N. idaeus and P. persimilis fed on various T. urticae stages are reported in Table 3. The intrinsic rate of increase (r_m), and the finite rate of increase (λ), were significantly higher in P. persimilis (t=-108.21; df=36; p<0.001), as well as the Gross and the Net reproductive rates (t=-86.22; df=36; p<0.001 and t=-77.83; df=36; p<0.001 for the latter parameters respectively). Doubling time (DT) and mean generation time (T) were shorter in P. persimilis (Table 3).

The daily fecundity m_x (female eggs/female/day) reached a peak 3 days after reaching the adult stage, with 2.5 and 3.5 eggs/female for N. idaeus and P. persimilis respectively, maintaining a plateau for about 7 days; for another week the daily offspring production was more than 2 female eggs/female/day (Figure 3). Afterwards, oviposition dramatically decreased, remaining at low levels (less than one egg/female/day), up to 41^st and 43^rd day for N. idaeus and P. persimilis respectively.

The oviposition period lasted about 45 days for both species but the survival rate during this period was significantly different between the two phytoseiids. Almost all females of N. idaeus, survived the whole oviposition period while only half of P. persimilis females reached the end of this period (Figure 3). The last two females of N. idaeus died after 78 days while P. persimilis females concluded their lifespan after 49 days (Figure 3). The high rate of fecundity during the first 20 days in both species and the following dramatic decline at low ovipositional level that lasted for a similar period (about 20 days), yielded data that cannot be adapted in a parametric model. The nonparametric growth curve yielded by the Repairable System Analysis (Meeker et al. 2021) showed statistical differences for the cumulative fecundity between the two species, being that of P. persimilis higher (Figure 4).

TABLE 3.

Demographic parameters of Neoseiulus idaeus and Phytoseiulus persimilis females calculated on the 18 females obtained by the postembryonic development tests, using various stages of Tetranychus urticae as food.

FIGURE 3.

Age-specific survival rate (l_x) (broken lines) and age-specific fecundity (m_x) (solid lines) of Neoseiulus idaeus (A) and Phytoseiulus persimilis (B) fed on various stages of T. urticae.

FIGURE 4.

Cumulative number of eggs/female (mean± 95% CI) of Neoseiulus idaeus (green triangles) and Phytoseiulus persimilis (orange circles) fed on various stages of T. urticae at 25 ± 1 °C, 75 ± 5% RH and 16L:8D photoperiod.

TABLE 4.

Consumption rate of T. urticae eggs by females and males during their postembryonic development and by unmated and mated females of Neoseiulus idaeus and Phytoseiulus persimilis. Different letters or symbols denote significant differences between different stages and species.

Predation of different stages of Neoseiulus idaeus and Phytoseiulus persimilis on Tetranychus urticae eggs

Predation rate during the postembryonic development of the two predators showed significant differences between the two species (F=10.94; df=1, 47; p= 0.002), but no differences were noted regarding sex (F= 0.04; df=1, 47; p= 0.844). Data and the interaction Species x Sex factors (F=2.89, df=1, 47, p<0.096). It is noteworthy that 32% of the N. idaeus larvae transformed into protonymph without feeding.

As expected, mated females consumed a higher number of prey eggs in comparison to the unmated ones (F= 3140.91; df=1, 482; p<0.001) and P. persimilis females showed more voracity than those of N. idaeus (F= 291.67; df= 1, 482; p<0.001), regardless their status (unmated or mated) (Table 4). In the first two days the predation rate of unmated females was the same for the two phytoseiids, but afterwards the number of preyed eggs decreased drastically for N. idaeus, whereas the predation rate of P. persimilis remained almost at the same level for the entire observation period (Figure 5A). A similar constant predation rate was registered also for the mated females of the latter species, whereas mated N. idaeus females reached the plateau at the 2^nd day (Figure 5B). In these experiments, P. persimilis produced a mean of 3.81±0.083 eggs/female/day (mean of period) and this value was statistically higher to that registered with N. idaeus females (2.87±0.062) (F=95.61; df=1, 380; p<0.001) (Figure 6). In four out of ten days, P. persimilis females produced more eggs than those of N. idaeus, but during the remaining period no differences were noted between the two phytoseiids, even if the daily oviposition rate of the latter species has been always lower (Figure 6). To verify if the 40 eggs of T. urticae used in the above-mentioned experiments provided the required energy for the maximum daily egg production, we compared the latter parameter between females fed on all stages of T. urticae (life-table experiments) and females fed on 40 tetranychid eggs/day. No differences were noted between the daily oviposition rate obtained on 40 eggs and all stages of T. urticae within each species, but significant differences registered between the two species (F=53.36; df=1, 755; p<0.001) (Figure 7). This datum indicated that 40 eggs of prey furnish sufficient energy for the maximum egg laying.

FIGURE 5.

Predation of unmated (A) and mated females (B) of N. idaeus and P. persimilis on T. urticae eggs. Curves reported the daily predation rate (mean ± S.E.) and histograms the mean of the period. Different letters denote significant differences; *-not significative (p<0.05).

FIGURE 6.

Oviposition rate of N. idaeus and P. persimilis females on 40 eggs of T. urticae/predator female/day. Curves indicate the daily oviposition rate and histograms the mean (±SE) of the period. Different letters denote significant differences; *-not significative (p<0.05).

FIGURE 7.

Oviposition rate of N. idaeus and P. persimilis females on 40 eggs and on all stages of T. urticae/predator female/day. Different letters denote significant differences (p<0.05).

Figure 8 reports the relation between the predation efficiency of the two species, calculated as ratio between laid and preyed eggs (conversion index), vs prayed eggs, according to the model singled out from the GLM, previously described in 2.4.

In both species the conversion index is decreasing when increasing the number of preyed eggs, but the slope is statistically different between the two species, being the decrement of N. idaeus quicker than that of P. persimilis (F=12.75; df=1, 378; p<0.001). Moreover, the average number of preyed eggs of the latter species is significantly higher than the former one (F=80.03; df=1, 378; p<0.001).

FIGURE 8.

Conversion index (CI) (laid eggs/preyed eggs) vs preyed eggs. The General Linear Model has been performed on data (p<0.05).

Postembryonic development

Neoseiulus idaeus completes the postembryonic development as quickly as the specialized P. persimilis, which is noted to have a very short postembryonic development in comparison to other phytoseiid taxa (Takafuji & Chant 1976; Escudero & Ferragut 2005; Moghadasi et al. 2016). Using a Chilean strain of N. idaeus (reported as C. picanus), on T. urticae eggs, Ragusa et al. (2000) registered a similar juvenile development (3.9 and 4.0 days for female and male respectively), and similar postembryonic development has been reported by Collier et al. (2007) and de Moraes et al. (1994) with a Brazilian strain of the species. A slightly longer development on various T. urticae stages was reported by Reichert et al. (2017) (5.23 and 5.18 days for females and males respectively) for another Brazilian strain of N. idaeus. This latter discrepancy could be due to the egg stage lasting more than two days, whereas in our tests the embryonic development concluded in about one day.

Our tests showed a shorter juvenile period to that registered on other congeneric species such as Neoseiulus californicus (McGregor) and Neoseiulus fallacis (Garman) (Monetti & Croft 1997; Williams et al. 2004; Escudero & Ferragut 2005). For example, 5.1 and 4.9 days (for female and male respectively) have been reported for the congeneric type II predator Neoseiulus californicus (McGregor) = chilenensis (see Beaulieu & Beard 2018; Beaulieu et al. 2019), fed on T. urticae eggs (Ma & Laing 1973), and also for a Mediterranean strain (5.64 and 5.45 days for female and male respectively) (Ragusa et al. 2009) and for a Japanese strain of the latter species (5.06 and 4.62 days for female and male respectively (Canlas et al. 2006). The sex-ratio is generally female-biased in Phytoseiidae, but temperature, population density and quality of food heavily influence this parameter. The emergence of females from the eggs used in our experiments showed values (72% and 78% for N. idaeus and P. persimilis respectively) similar to those reported for the same or other phytoseiid species fed on optimal food at the optimal temperature condition (van Dihn et al. 1988; Escudero & Ferragut 2005; Collier et al. 2007).

Preoviposition, oviposition, postoviposition and longevity

During the adulthood the two phytoseiids showed significant differences in all but one (oviposition period) of parameters taken into consideration (Table 2). N. idaeus showed a 5-fold longer postoviposition period to that occurred in P. persimilis. Ragusa et al. (2000) reported a duration of about two weeks for both oviposition and post-oviposition periods with a Chilean strain of N. idaeus, and similar data are reported for a presumably Brazilian strain of this species (van Dihn et al. 1988). However, shorter oviposition and post-oviposition periods were reported with another Brazilian strain of the species (≈8 and ≈2.5 days for oviposition and post-oviposition periods respectively) (de Moraes et al. 1994; Collier et al. 2007) (for morphological differences between the two strains see Tixier et al. 2011). The Sicilian strain used in our experiments is like the Chilean one (presence of the solenostomes gd2) indicating both morphological and biological similarities between the two strains. Differences among strains of different geographical origin were reported for other phytoseiid species (Gotoh et al. 2004; Furtado et al. 2007). Vangansbeke et al. (2013) reported a slightly longer oviposition period for P. persimilis (29.6 days) compared to our data and a higher fecundity (82.7 eggs/female and 2.81 eggs/female/day) at a lower constant temperature (20°C); similar values are also reported from the above-mentioned authors for N. californicus (26.2 days, 55.5 eggs/female and 2.12 eggs/female/day). Popov and Kondryakov (2008) reported similar reproductive parameters (52.6 eggs/female and 2.92 eggs/female/day) for P. persimilis fed on Tetranychus atlanticus McGregor but a shorter lifespan (24 days). Ji et al. (2007) showed that longevity and post-oviposition period were negatively correlated with fertility of Neoseiulus cucumeris (Oudemans) and our data confirm this finding.

Demographic parameters

The high fecundity of specialised phytoseiids, usually corresponds to high values of demographic parameters (Zhang 1995; McMurtry & Croft 1997). However, the data reported in literature for both species used in our experiments show a great variability probably depending both on the different experimental conditions adopted and the origin of strains (Takafuji & Chant 1976; Badii & McMurtry 1984; Collier et al. 2007; Ragusa et al. 2000; Reichert et al. 2017; van Dihn et al. 1988) (Table 5). As expected, all demographic parameters of P. persimilis outperformed those of N. idaeus, but the data on the latter species are of particular interest because they are like those obtained for N. californicus, a phytoseiid mite widely used for the biological control of T. urticae (Gotoh et al. 2004; Escudero & Ferragut 2005). The Sicilian strain of N. idaeus used in our experiments showed similar demographic parameters to those reported for some Brazilian strains (van Dihn et al. 1988; de Moraes et al. 1994; Reichert et al. 2017) but significantly higher and lower parameters have also been reported in literature for this species (Ragusa et al. 2000; Collier et al. 2007) (Table 5).

Predation of different stages of N. idaeus and P. persimilis on T. urticae eggs

It was commonly accepted that larva in many Phytoseiid species is a facultative feeding stage (de Moraes & McMurtry 1981; Zhang & Croft 1994; Palevsky et al. 1999; Schausberger & Croft 1999), but Chitteden and Saito (2001) showed three different behaviors of phytoseiid larvae: obligatory feeding larvae (OFL), facultative feeding larvae (FFL) and no feeding larvae (NFL). However, one-hour interval observations needed to ascertain the larvae feeding behavior, and probably for this reason some authors did not consider this stage in their experiments on predation (Alipour et al. 2016; Moghadasi et al. 2016). Regarding P. persimilis larva, various authors reported it as a no-feeding stage or as a usually no-feeding stage (Laing 1968; Sabelis 1981; McMurtry & Croft 1997; Chitteden & Saito 2001). In our experiments all P. persimilis larvae consumed about two prey eggs per day, but it should be mentioned that within the time interval we adopted for observations (24h), all larvae were transformed in protonymphs. As a consequence, the predation rate observed should be attributed to early protonymphal stage rather than to the larval one.

TABLE 5.

Demographic and biological parameters of N. idaeus and P. persimilis reported in literature.

Deutonymphs of both species were the more voracious juvenile stage with similar daily predation rate to the unmated females as also reported by various authors in similar experiments (Ragusa et al. 2000; Moghadasi et al. 2016; Reichert et al. 2017; Saemi et al. 2017). This fact may be due to the greater need of energy in deutonymphs, for the formation of the reproductive system as stated by Sabelis (1985). The lower voracity of P. persimilis during the postembryonic development period, considering that the two species showed the same development time, indicate a higher conversion rate in this species in comparison to N. idaeus. On the contrary, the daily consumption rate was significantly higher for both unmated and mated females of the former species. It is known that the mated females of phytoseiids allocate an important fraction of the food ingested in the production of eggs (Sabelis 1985; Sabelis & Janssen 1994). We observed a different behaviour between the number of eggs preyed and eggs laid in the two phytoseiids. Specifically, N. idaeus consuming only the 46.3% of eggs preyed by P. persimilis females, reached the 75.3% of the oviposition rate of the latter species. From an energy point of view, N. idaeus showed higher performance than P. persimilis, because the number of laid eggs is greater at the same level of predation rate, indicating that the former species could be more competitive at low prey densities in comparison to the specialised P. persimilis. The lower conversion index registered for P. persimilis confirms its greater efficiency in controlling tetranychid populations in a short time period. Similar behaviour was also observed in another congeneric phytoseiid Phytoseiulus macropilis (Banks) (Ferla et al. 2011).

According to Sabelis and Bakker (1992), specialised predators of spider mites have longer dorsal setae to facilitate their movement within the dense webs produced by tetranychid mites, whereas generalist predators generally have shorter dorsal setae, making their movement within the webs more difficult. However, although N. idaeus is commonly classified as a type II predator, it has longer dorsal setae in comparison to other type II phytoseiid species (Denmark & Muma 1973; Ragusa 2000; Tixier et al. 2011; Tsolakis & Ragusa 2016). This characteristic could be interpreted as a specialised adaptation to tetranychid species that produce a complicated web (C-W type) (Saito 1985; Croft et al. 1998). Indeed, our predation results offer interesting insights, confirming the ability of this phytoseiid to prey inside C-W type spider mite colonies.

In conclusion, our study shows that N. idaeus populations in the Mediterranean area have all the necessary characteristics to effectively control two-spotted spider populations. This predator could be particularly advantageous compared to the predator P. persimilis due to its ability to adapt to alternative food sources, i.e. pollen of weeds or arboreal plants (Ragusa et al. 2000; Rioja & Vargas 2009) and consequently contain infestations at an early stage.