Exp Appl Acarol (2008) 44:293–306 DOI 10.1007/s10493-008-9151-0
The spatial and temporal distribution of predatory and phytophagous mites in field-grown strawberry in the UK Jean Fitzgerald Æ Xiangming Xu Æ Nicola Pepper Æ Mike Easterbrook Æ Mike Solomon
Received: 19 October 2007 / Accepted: 31 March 2008 / Published online: 16 April 2008 Ó Springer Science+Business Media B.V. 2008
Abstract Extensive sampling of strawberry plants in everbearing and June-bearing strawberry plantations and on potted plants showed that different species of mites were spatially separated. Of the two phytophagous species recorded, Tetranychus urticae was most abundant on old leaves and Phytonemus pallidus on folded leaves and flower/fruit clusters. Predatory phytoseiid mites were found on all plant parts but different species were spatially separated; Neoseiulus cucumeris and N. aurescens were found mostly on folded leaves and clusters, and N. californicus and Phytoseiulus persimilis on old and medium aged leaves. No Typhlodromus pyri were found in the field plantations. These patterns of distribution did not change over sampling dates in summer and early autumn. An understanding of this within-plant zonation of mite species is important when studying predator– prey interactions and when designing sampling strategies for strawberry. A programme to sample the entire mite system on strawberry should be stratified to include all the above mentioned parts of the plant. Different sampling protocols, as appropriate, are required for sampling different pest species and their associated predators. Keywords Strawberry Tetranychus urticae Phytonemus pallidus Phytoseiids Neoseiulus californicus Neoseiulus cucumeris Neoseiulus aurescens Phytoseiulus persimilis Distribution Sampling
Introduction In soft fruit culture two categories of predator are important. Mass cultured, commercially available predators are released artificially into the system as biocontrol agents (e.g. Easterbrook 1992; Easterbrook et al. 2001) but colonisation by naturally occurring predators also takes place (Easterbrook 1998; Cross et al. 2001). If opportunities for biocontrol with these two types of predator are to be optimised, it is necessary to understand the nature of the interactions between the naturally occurring predator species (which J. Fitzgerald (&) X. Xu N. Pepper M. Easterbrook M. Solomon East Malling Research, New Road, East Malling, Kent ME19 6BJ, UK e-mail:
[email protected]
123
294
Exp Appl Acarol (2008) 44:293–306
cannot readily be manipulated) and the artificially introduced predators (the deployment of which is within our control), and their joint impact on pest species. Adding predator species to existing predator–prey systems may not improve suppression of herbivore numbers (Rosenheim 1998) and can lead to an increase in some species (Sih et al. 1985). Biocontrol may be disrupted by intraguild predation (Rosenheim et al. 1995), and by other indirect food–web interactions (Janssen et al. 1998). The major phytophagous mites on strawberry in the UK are the spider mite Tetranychus urticae Koch and the tarsonemid mite Phytonemus pallidus ssp. fragariae (Zimmerman). Large populations of T. urticae feeding on strawberry leaves can lead to a reduction in the quality and quantity of harvested fruit (Sances et al. 1982), and can also affect yield in the following season (Butcher et al. 1987). Phytonemus pallidus feeding can result in stunting of the plants, but they also feed on flowers and developing fruits, reducing the size and number of fruits (Stenseth and Nordby 1976). High population levels of P. pallidus can halt fruit production (Zalom et al. 1990) and populations can build up very rapidly; development time from egg to adult is 8.8 days at 25°C (Easterbrook et al. 2003). The predatory mites that colonise most often in SE England are Neoseiulus cucumeris (Oudemans), N. californicus (McGregor) and Typhlodromus pyri Scheuten; N. californicus is not a native UK species and cannot be released into open field systems as a biocontrol agent. However, it has been released erroneously in the past and has been shown to overwinter successfully in strawberry and blackcurrant plantations (Jolly 2001). Other predatory species, such as N. aurescens Athias-Henriot, are found in plantings in some years. Colonisation of plantings by naturally occurring predators will depend on pesticide usage in the crop and surrounding areas, the availability of local reservoirs of the predators, and the age of the planting; older plantings are likely to contain higher numbers and more species of colonisers. Predatory mite species that can be artificially released in open field plantations are Phytoseiulus persimilis Athias-Henriot (a non-native species recommended for control of T. urticae), which does not survive the winter in UK, and N. cucumeris (a UK native species recommended for control of P. pallidus and for thrips species), which can survive UK winter conditions. In this research we investigated the distribution of predatory and phytophagous mites on strawberry, to enable us to determine if interactions among the species are likely to occur in the field. Samples of leaves of different ages and flower/fruit clusters were taken from potted strawberry plants that were infested with P. pallidus and T. urticae, and from plantations of both everbearing and June-bearing strawberries at East Malling Research (EMR) that had been infested with T. urticae and in which predatory phytoseiid mites were also present, to elucidate the patterns of within- and between-plant distributions of phytophagous and predatory mites. June-bearing and everbearing strawberry plants have different growth habits, with everbearers continuing to grow and fruit until late in the season, so populations of the different mite species on these plants were compared to determine if the plant type has an effect on mite distribution.
Methods Glasshouse experiment Twenty four potted strawberry plants, cv. Evita, that had become naturally infested with T. urticae in a glasshouse, were infested with 20 adult female P. pallidus taken from laboratory cultures in September 1999 as part of a biocontrol experiment (Easterbrook
123
Exp Appl Acarol (2008) 44:293–306
295
et al. 2001). The mites were transferred to young developing leaves with a fine paintbrush, and the plants were returned to the glasshouse. On 1 October plants were divided into three groups of eight plants and each group received 0, 4 or 20 N. cucumeris per plant (obtained from Biological Crop Protection, Wye, UK). Plants were returned to a glasshouse with heating and supplementary lighting (16L:8D with temperatures between 8 and 14°C). After 4 weeks the plants were dissected and the numbers of T. urticae, P. pallidus and N. cucumeris on each leaf or plant part were recorded under a stereomicroscope. Categories of plant part recorded were: mature leaf (fully expanded); young leaf (part unfolded); folded leaf; flower/fruit cluster; crown. At this time the plants had between 13–16 leaves and 3–5 clusters with flowers and developing fruit. 2001 Field sampling Field sampling in 2001 was done in plantings at EMR. All strawberry plants were grown in raised double row beds through polyethylene mulch. Everbearing planting This contained the cv. Calypso, planted in August 1999. Five plots of 80 plants, planted in two double row beds were used in this experiment. The plants were planted 0.4 m apart between the rows and 0.5 m apart within the rows. The centres of each bed within a plot were 1.8 m apart, and each plot was separated from other plots by 10 m. A release of c. 40 T. urticae per plant (Biological Crop Protection) was made on 26 June 2001. No predatory mites were released in 2001 in this planting; preliminary sampling indicated that naturally occurring phytoseiids had colonised the plants. On 9 July, 6 August, 5 September and 12 October, 20 plants were chosen at random and plant parts of different age were sampled. One old leaf taken from the outside of the plant canopy, one mature leaf taken from inside the plant canopy and one recently expanded leaf, plus one leaf that had not yet unfurled (grouped together as ‘young’ for analysis), and one blossom/fruiting cluster, were taken from each plant. Numbers of the pest mites T. urticae and P. pallidus, and predatory phytoseiid mites found on each plant part were recorded. All adult phytoseiids collected were mounted on slides, cleared, and identified to species under a compound microscope (only adult mites can be identified reliably using taxonomic keys). On 6 August two plants were dug up, and all leaves and blossom clusters inspected for presence of mites. All adult phytoseiids found were mounted on slides and identified to species. June-bearing planting This was planted in August 2000 and contained the cvs Elsanta, Florence and Symphony, planted at the same spacing within the beds as in the everbearing planting described above. Three double row beds were planted with the centres of the beds 1.8 m apart as above. These three beds were separated from adjacent plots of three beds by 2.4 m alleys. There was a natural infestation of T. urticae in this planting and artificial releases were not made. A release of Phytoseiulus persimilis was made in late May 2001. Samples of plant material of different ages were taken on 22 June, 2 July, 16 July, 1 August and 3 September as in the everbearing planting and mites counted and identified as above.
123
296
Exp Appl Acarol (2008) 44:293–306
On 19 July 2001, 10 plants were dug up, and all leaves and blossom clusters inspected for presence of mites. All adult phytoseiids found were mounted on slides and identified to species. 2003 Field sampling On 20 February 2003 two plants and associated dead leaf matter were dug up from the everbearing plantation and divided into crown, old and young green leaves, and developing buds. As these plants were now 4-years old they were large and multi-crowned; no crown thinning had been done on the plants as part of the routine husbandry of the plantings. Leaves and a sub sample of 10 crowns from each plant were examined under a stereomicroscope. Mites were also extracted from a total of 18 crowns and associated plant material and from dead plant material using Tullgren funnels. Numbers of mites were recorded for each plant part, slides were made of any adult phytoseiids present and species present were identified. A similar sample of 10 plants was taken on 6 March. In April 2003 new plots of everbearing strawberries, cv. Calypso, were planted as described above, between the existing everbearing plots, such that plots were now c. 3 m apart (planting 1). A separate new planting (planting 2) was also made using everbearing cv. Bolero; this planting consisted of double row beds with plant spacing as above. Each plot contained 20 plants; there was a 5 m gap between plots along the bed and the centres of the beds were 3.6 m apart. Tetranychus urticae supplied in shaker tubes (BCP) were released on these plots. The required release rate was obtained by counting numbers of mites in a ‘standard’ shake and diluting as required by increasing the bran content in the tube. On 20 June c. 10 T. urticae per plant were released on planting 1 and 20 per plant on planting 2, and on 11 July c. 20 P. persimilis were released per plant. Three sets of leaf samples were taken between 22 July and 1 October on each of the two everbearing strawberry plantings. The first sample unit consisted of 10 old leaflets taken from the outside of the plant canopy (one third of the trifoliate leaf), 10 young fully expanded leaflets taken from inside the plant canopy, 10 folded leaves and 10 flower clusters. Subsequent sample units consisted of five of each plant part. Four replicate samples were taken on each occasion from each planting. Numbers of T. urticae and P. persimilis were counted on each plant part. Any other phytoseiid mites recorded in samples from these new plantings were not identified as this experiment was designed to determine the distribution of P. persimilis on the plants. Data analysis All the statistical analyses were done using GenstatTM (Payne 2002). To determine whether the proportion (p) of leaves or leaflets or other plant parts containing prey or predators was influenced by leaf/leaflet age or plant part a logistic regression analysis was conducted (Cox and Snell 1989), assuming a binomial distribution for p. In this form of analysis, the logit transformation of the proportion, i.e., lnðp=1 pÞ, is expressed as a multiple linear regression on the treatment factors, e.g., leaf age and sampling time. A generalized linear mixed model (GLMM) approach was used for this analysis, where blocks and individual plants were treated as random factors. In the GLMM analysis, the Wald test was used to determine whether the effects of fixed factors (leaf age, sampling time) were significant. If these were significant, the least significant difference (LSD) was used to test the significance of the differences between individual factors using the standard error of the differences estimated from the GLMM analysis.
123
Exp Appl Acarol (2008) 44:293–306
297
Statistical tests were also carried out to determine whether the observed counts data could be fitted to a normal distribution or other discrete distributions, using the maximumlikelihood with the goodness-of-fit determined by the v2 test. The distribution analysis indicated that nearly all the T. urticae and P. pallidus counts data could be satisfactorily described by a negative binomial distribution. In contrast, the counts data for phytoseiids were extremely skewed and could not be fitted to any common discrete distributions. Therefore, the effects of leaf age or plant part and sampling time on T. urticae and P. pallidus abundance (counts) were determined using the generalized linear model (GLM) assuming a negative binomial distribution for the data. The significance of each treatment factor was determined by the deviance test in the GLM analysis. If the factor was significant, treatment differences between different levels of this factor were then assessed for statistical significance by comparing their parameter estimates against the standard error of their difference (sed). The sed value was estimated from the variance–covariance matrix from the logistic regression analysis. For the phytoseiids, counts data were analysed using the Kruskal–Wallis nonparametric one-way ANOVA because of their extreme departure from normal distribution. Finally, statistical tests were carried out to determine whether prey and predators tended to coexist on the same leaf or leaflet or other sampling unit using Fisher’s exact test, which provides an exact probability for the observed data.
Results Glasshouse experiment Average numbers of T. urticae, P. pallidus and N. cucumeris, and their incidences, on the different plant parts in the potted plant experiment are given in Tables 1 and 2. There were no significant interactions between N. cucumeris release treatments and plant parts although there were small but significant differences in N. cucumeris abundance between release treatments. Because this paper focuses on the distribution of prey and predators on leaves of different ages and on other plant parts, the data presented are pooled over all three N. cucumeris release treatments (including the untreated control). There were significant (P \ 0.001) differences between the incidences of T. urticae on different plant parts. The incidences of eggs, adults and immatures were much greater (99% of total individuals) on mature leaves than on young leaves, which in turn were greater than on blossoms, crowns and folded leaves; incidences on the latter three plant parts were close to zero. In line with the incidence data, there were significant (P \ 0.001) differences in numbers of T. urticae between different plant parts. Almost all eggs ([99%), immatures ([98%) and adults ([98%) were found on mature leaves (Table 1). For P. pallidus there were similar numbers of eggs, immatures and adults on blossoms, crowns and folded leaves, and numbers on these were significantly (P \ 0.001) higher than on mature and young leaves (Table 1). There were also significant (P \ 0.001) differences between the incidences of P. pallidus on different plant parts. The ranking in the incidences of eggs, adults or immatures was the same: mature leaves (lowest), young leaves, folded leaves, blossom and crown (Table 1); 57% of eggs and 61% of active stages were found on the flower/fruit clusters under the calyx and sepals, where the active stages were feeding on the developing fruit or the receptacles. The incidence of P. pallidus on mature leaves was close to zero.
123
298
Exp Appl Acarol (2008) 44:293–306
Table 1 Average number of T. urticae and P. pallidus per plant part in the potted plant experiment, with the percentage of plant parts with mites present in parentheses Plant part
No. of samples
T. urticae
P. pallidus
Eggs Flower/fruit cluster
101
Immatures
0.1 (1.0)
0.2 (2.0)
Adults 0.0 (0.0)
Eggs 6.3 (52.5)
Immatures 3.7 (56.4)
Adults 9.8 (79.2)
Folded leaves
19
0.0 (0.0)
0.4 (5.3)
0.1 (5.3)
4.9 (21.1)
3.9 (47.4)
4.4 (36.8)
Young leaves
53
1.9 (32.1)
1.7 (39.6)
0.2 (15.1)
0.7 (7.5)
0.8 (26.4)
1.0 (20.8)
Mature leaves
329
94.8 (51.6)
19.6 (47.2)
22
0.0 (0.0)
0.0 (0.0)
Crown
3.3 (36.5) \0.0 (1.2) 0.0 (0.0)
6.4 (87.0)
\0.0 (1.8) 5.2 (76.5)
0.1 (4.3) 12.5 (99.3)
Table 2 Average number of N. cucumeris per plant part in the potted plant experiment, with the percentage of plant parts with predators present in parentheses Plant part
Eggs
Immatures
Adults
Flower/fruit cluster
0.2 (19.8)
0.2 (17.8)
Folded leaves
0.1 (5.3)
0.1 (10.5)
Young leaves
0.1 (9.4)
0.1 (5.7)
\0.1 (1.9)
0.3 (18.8) 0.1 (5.3)
Mature leaves
0.2 (10.0)
0.1 (7.9)
\0.1 (3.6)
Crown
0.0 (0.0)
0.0 (0.0)
0.1 (9.1)
There were significant (P \ 0.001) differences between plant parts for incidences of N. cucumeris eggs and adults (Table 2); the difference was also close to standard statistical significance (P = 0.07) for N. cucumeris immatures. Compared to P. pallidus and T. urticae, the overall incidence of N. cucumeris was very low (Table 2). The highest incidence, close to 20%, occurred on flower/fruit clusters. Non-parametric ANOVA showed that overall there were significant differences in the number of N. cucumeris eggs (P \ 0.001), immatures (P \ 0.05) and adults (P \ 0.001) between different plant parts; numbers were higher on flower/fruit clusters (Table 2). There were significant (P \ 0.001) negative associations between P. pallidus and T. urticae. The associations of N. cucumeris eggs and immatures with T. urticae were not significant but there were negative associations (P \ 0.05) of N. cucumeris adults with T. urticae eggs and immatures. Significant positive associations were also observed between N. cucumeris immatures and adults and P. pallidus. 2001 Field sampling Everbearing plantation Average numbers of T. urticae and phytoseiids, and their incidences in the everbearing strawberry planting are given in Table 3. No P. pallidus were recorded from these plants on any recording occasion. There were significant differences between plant part and sampling date for incidences of all stages of T. urticae. There were also significant interactions between plant part and sampling dates. However, the interaction accounted for far less variation than the two main factors, particularly the plant part sampled. In addition, the relative order in the incidence between different plant parts did not change over sampling dates. Thus, specific interactions were not considered further. For T. urticae eggs, the incidence was higher
123
90
175
212
109
Flower/fruit cluster
Young leaves
Mature leaves
Old leaves
No. of samples
Plant part
14.3 (62.4)
4.1 (22.8) 9.8 (63.3)
1.9 (22.8)
0.1 (2.7)
\0.1 (0.8) 15.2 (67.3)
2.3 (25.0)
0.1 (2.7) 0.5 (22.9)
0.3 (16.0)
\0.1 (4.0)
\0.1 (1.1)
0.1 (4.4)
\0.1 (3.3) 0.2 (13.7)
All eggs
Immatures
Eggs
Adults
Phytoseiids
T. urticae
0.5 (24.8)
0.2 (16.5)
\0.1 (2.9)
0.4 (27.8)
All immatures
0.1 (10.1)
0.1 (6.1)
\0.1 (4.0)
0.4 (31.1)
N. cucumeris
0.3 (21.1)
0.1 (9.4)
\0.1 (0.6)
\0.1 (1.1)
N. californicus
0.1 (5.5)
\0.1 (1.9)
\0.1 (4.0)
0.2 (13.3)
N. aurescens
Table 3 Average number of T. urticae and phytoseiids per plant part with percentage of plant part with mites present in parentheses (everbearing strawberries, sampled over different dates in 2001)
Exp Appl Acarol (2008) 44:293–306 299
123
300
Exp Appl Acarol (2008) 44:293–306
(P \ 0.01) on the old leaves than on mature leaves, which in turn was greater (P \ 0.05) than on flower/fruit clusters and young leaves. This pattern of differences was similarly observed for immature and adult T. urticae. In line with the incidence data, the number of T. urticae eggs, immatures and adults was higher (P \ 0.01) on the old leaves than on mature leaves, which in turn was greater (P \ 0.05) than on flower/fruit clusters and young leaves (Table 3). Low numbers of N. californicus, N. cucumeris, and N. aurescens adults were identified from samples from the everbearing plantation in 2001; these species had naturally colonised the planting since no artificial releases had been made. Phytoseiids were present on all plant parts, and they persisted through the sampling period. Overall the incidence of phytoseiids was much less than that of T. urticae (Table 3). It is not possible to distinguish between species for phytoseiid eggs, and it is difficult to distinguish between phytoseiid immature stages so these were analysed together. For phytoseiid eggs, immatures and adults, there were significant differences between plant parts but no significant interactions between plant parts and sampling time. There were similar amounts of mature and old leaves with phytoseiid eggs present, and this was greater (P \ 0.05) than incidences of phytoseiid eggs on flower/fruit clusters and young leaves. Similarly, more N. californicus adults were found on mature and old leaves than on flowers and clusters (P \ 0.05). However, for immature phytoseiids the incidence of mites was much higher on clusters compared with egg incidence (P \ 0.05) (Table 3); incidences of immatures were not significantly different on flower/fruit clusters from those on mature and old leaves. It was not possible to identify the species of the eggs and immature stages collected in these samples but it is likely that the species distribution for eggs and immatures will follow that of the adult species. For N. cucumeris and N. aurescens adults, incidence of mites was significantly higher (P \ 0.05) on flower/fruit clusters than on young and mature leaves. Non-parametric ANOVA showed that overall there were significant (P \ 0.001) differences in the number of phytoseiids between different plant parts. There were more phytoseiid eggs and N. californicus adults on old leaves than mature leaves, which was greater than their abundance on flower/fruit clusters and young leaves. On clusters, there were more N. cucumeris and N. aurescens adults than on the old leaves. For phytoseiid immatures, highest numbers were recorded on flower/fruit clusters (possibly N. cucumeris) and on old leaves (possibly N. californicus) (Table 3). There were significant (P \ 0.001) positive associations of phytoseiid eggs, immatures and N. californicus adults with T. urticae (eggs, immatures or adults). There were no significant associations of N. cucumeris adults with T. urticae. There was a negative association of N. aurescens adults with T. urticae eggs (P = 0.06) and immatures (P \ 0.05). Among all pairwise associations among phytoseiid eggs and immatures, and N. aurescens, N. cucumeris and N. californicus adults, only the positive association between phytoseiid immatures and N. californicus adults was significant (P \ 0.01). A total of 10 eggs and 49 active phytoseiid stages were found on the two everbearing plants sampled in August 2001. Twenty five adult mites were identifiable; 20 were N. cucumeris and five were N. californicus; 64% of the identified N. cucumeris were found on the flower clusters and all the N. californicus were on the older leaves. June-bearing planting No P. pallidus were recorded in any samples and very few phytoseiids were found, even though a release of P. persimilis had been made in May; counts of P. persimilis were not analysed. Average numbers of T. urticae and their incidences in the June-bearing plantation in 2001 are given in Table 4. As in the everbearing plantation, there were significant
123
Exp Appl Acarol (2008) 44:293–306
301
Table 4 Average number of T. urticae per plant part with percentage of plant parts with mites present in parentheses (June-bearing strawberries, sampled over different dates in 2001) Plant part Flower/fruit cluster Young leaves
No. of samples
Eggs
Immatures
Adults
53
21.2 (32.1)
2.6 (32.1)
1.9 (35.8)
227
1.9 (13.2)
0.6 (10.4)
1.0 (16.2)
Mature leaves
237
56.0 (58.1)
9.6 (43.6)
8.5 (55.8)
Old leaves
118
74.3 (73.2)
33.8 (66.1)
19.8 (67.8)
Table 5 Average number of T. urticae per plant part when sampled on 19 July 2001, with percentage of plant parts with mites present in parentheses (June-bearing strawberries) Plant part
No. of samples
Eggs
Immatures
Adults
Flower/fruit cluster
27
7.0 (29.3)
7.0 (40.7)
3.8 (44.4)
Young leaves
37
1.2 (13.5)
2.5 (27.0)
2.5 (24.3)
Mature leaves Old leaves
54
144.4 (80.8)
47.4 (78.3)
29.3 (85.1)
113
140.9 (89.2)
111.0 (90.1)
67.4 (90.7)
(P \ 0.001) differences in the numbers and incidence of T. urticae between plant parts and sample dates for eggs, immatures and adults. There were also significant (P \ 0.001) interactions between plant parts and sampling dates. However, the interaction accounted for far less variation than the two main factors, particularly the plant part sampled. In addition, the relative order in the incidence between different plant parts did not change over sampling dates. Thus, as in the analysis of the data from the everbearing planting, specific interactions were not considered further. The ranking of the incidences of T. urticae eggs, immatures and adults was the same: young leaves (lowest), flower/fruit clusters, mature and old (highest) leaves (Table 4). As in the incidence data, for T. urticae abundance there were significant (P \ 0.001) differences between plant parts and sampling dates as well as their interactions for all T. urticae stages. The ranking was also the same as for the incidence data except that the differences in numbers of immatures and adults between flower/fruit clusters and young leaves were much smaller (Table 4). In the whole plant samples from the June-bearing planting taken in July 2001 there were significant differences between plant part and sample date for all T. urticae stages (Table 5). For T. urticae eggs, there were no significant differences in incidence between mature and old leaves or between young leaves and flower/fruit clusters. The former two were significantly (P \ 0.001) greater than the latter two. This pattern of distribution was similarly observed for T. urticae immatures and adults (P \ 0.001). As with the incidence data, the number of T. urticae eggs, immatures and adults was higher (P \ 0.01) on the old and mature leaves than on flower/fruit clusters and young leaves. There were more (P \ 0.01) T. urticae eggs and immatures on flower/fruit clusters than on young leaves whereas there were no significant differences between these two plant parts for T. urticae adults. Both P. persimilis and N. cucumeris were identified in very low numbers on these plants; these data were not analysed. 2003 Field sampling Phytoseiids were found to be active on the developing leaves in the winter samples taken in February and March 2003. A mean of four phytoseiids was found on the crowns and
123
302
Exp Appl Acarol (2008) 44:293–306
associated leaf material in the sample taken on 20 February; a total of 129 adult female phytoseiids were extracted from the dead plant material associated with these two plants. The most abundant species was N. cucumeris; only one N. californicus was identified from these samples. In the samples taken on 6 March, a total of 51 phytoseiids were extracted from leaves, 74 from crowns and 234 from dead plant material; these differences were significant at P \ 0.001 for leaves and dead plant material, and P \ 0.01 for leaves and crowns. Most of the phytoseiids identified were N. cucumeris. Three specimens of N. californicus and one of Amblyseius potentillae (Garman) were also identified. No T. urticae or P. pallidus were found in these samples. In the summer sampling on new planting 1, the incidences of T. urticae eggs, adults and immature stages were greater on mature leaves than on young or folded leaves, or clusters (P \ 0.001); incidences on the latter three plant parts were close to zero (Table 6). Significantly more eggs, immature stages and adults were found on old than on young leaves, which in turn was higher than on folded leaves and clusters (P \ 0.01). Almost all eggs, immatures and adults ([98%) were found on mature leaves (Table 6). This was further confirmed by non-parametric ANOVA of counts data within each sampling date. Very low incidences of P. persimilis eggs or adults were found on all plant parts (Table 6). However, there were significantly higher numbers of P. persimilis immature stages (P \ 0.001) on the old leaves. Non parametric ANOVA showed that overall there were no significant differences in the numbers of P. persimilis eggs and adults within each sampling date, but that there were significant differences in numbers of immature stages within each sampling date (P \ 0.05); higher numbers of P. persimilis were found on the old leaves than on folded leaves or clusters. There was a significant positive association of P. persimilis immature stages with T. urticae immatures and eggs (P \ 0.01). No other pairwise relationships were statistically significant. In planting 2 the incidences of T. urticae eggs, adults or immature stages were greater on mature leaves than on young leaves, which in turn were higher than on folded leaves or clusters (P \ 0.001) (Table 7). Significantly more T. urticae eggs and adults were found on old and young leaves than on folded leaves and clusters (P \ 0.01) (Table 7). For immature stages, more were recorded on old leaves than on clusters or young leaves (P \ 0.01), with lower numbers on folded leaves (P \ 0.01). More T. urticae adults were found on old leaves than on clusters (P \ 0.05), which in turn were higher than on young or folded leaves (P \ 0.05). This was further confirmed by non-parametric ANOVA of counts data within each sampling date. For P. persimilis eggs and immature stages, incidences were significantly higher on old or young leaves (P \ 0.001) than on folded leaves or clusters (Table 7). No statistically
Table 6 2003 Field experiment-planting 1. Average number of T. urticae and P. persimilis per plant part, with percentage of plant parts with mites present in parentheses Plant part
No. of samples
T. urticae Eggs
P. persimilis Immatures
Adults
Eggs
Immatures
Adults 0.0 (0.0)
Cluster
60
0.1 (1.7)
0.1 (8.3)
0.0 (0.0)
0.0 (0.0)
0.1 (1.7)
Folded leaf
60
0.1 (6.7)
0.1 (5.0)
0.2 (3.3)
0.0 (0.0)
0.1 (1.7)
0.0 (0.0)
Young leaflet
39
4.4 (28.2)
1.2 (30.8)
0.3 (18.0)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
Old leaflet
60
22.8 (83.3)
13.4 (91.7)
1.0 (46.7)
0.1 (3.3)
0.3 (16.7)
0.0 (0.0)
123
Exp Appl Acarol (2008) 44:293–306
303
Table 7 2003 Field experiment-planting 2. Average number of T. urticae and P. persimilis per plant part with percentage of plant parts with mites present in parentheses Plant part
No. of samples T. urticae Eggs
P. persimilis Immatures
Adults
Eggs
Immatures Adults
Cluster
60
2.2 (18.3)
3.3 (23.3) 1.0 (15.0) 0.1 (10.0) 0.2 (8.3)
Folded leaf
60
0.7 (16.7)
0.5 (26.7) 0.2 (16.7) 0.0 (0.0)
0.1 (1.7)
0.0 (0.0)
Young leaflet 60
4.4 (38.3)
2.0 (58.3) 0.5 (26.7) 0.1 (3.3)
0.2 (15.0)
0.0 (0.0)
Old leaflet
80
24.4 (71.3) 20.8 (85.0) 2.2 (57.5) 1.0 (36.3) 1.2 (36.3)
0.2 (1.7)
0.2 (15.0)
significant differences were detected in incidences of adults. Non-parametric ANOVA showed that there were significant differences in the number of P. persimilis eggs and immature stages within each of the first two sampling dates (P \ 0.01): more were recorded on old leaves. For the last sampling occasion, counts of adult P. persimilis were greater on old leaves than on young leaves (P \ 0.05). There was a significant positive association of T. urticae eggs and of T. urticae adults with P. persimilis eggs and adults, and of T. urticae immature stages with P. persimilis eggs and immatures (P \ 0.01). All other pairwise relationships were not statistically significant.
Discussion Results from the potted plant experiments where both T. urticae and P. pallidus were present showed that there was a negative association between these two species, with nearly all T. urticae found on older leaves and P. pallidus present on the folded leaves and flower/fruit clusters. Eggs, immature stages and adults of each species were distributed in the same pattern. These results for P. pallidus distribution are in line with those found by Croft et al. (1998) who reported that this species was more abundant in developing buds and crowns than on other plant parts on strawberry. In our experiments there was a positive association of N. cucumeris with P. pallidus and a negative association with T. urticae, showing that N. cucumeris was found most frequently on the same plant parts as P. pallidus. Results of sampling of everbearing and June-bearing strawberry plantations over two years during the summers of 2001 and 2003 confirmed the results of the potted plant experiments for T. urticae and showed that this species was present in greatest numbers on old leaves and lowest numbers on the young leaves and fruit clusters, with all stages distributed in the same pattern. There was a similar distribution of T. urticae in everbearing and June-bearing plants and the pattern of distribution did not change over the summer sampling periods. It did not prove possible to rear sufficient P. pallidus to release on the field plots, and none were found naturally occurring on these plots, so we were unable to confirm the distribution patterns of this species found in the potted plant experiment. However in previous research we have noted that higher numbers of P. pallidus are generally found on the folded leaves (Easterbrook unpublished results). In our field experiments three species of predatory phytoseiid mites, N. cucumeris, N. californicus and N. aurescens, were identified as naturally occurring in the planting, while P. persimilis were artificially released. In laboratory experiments (Fitzgerald and
123
304
Exp Appl Acarol (2008) 44:293–306
Easterbrook 2003), N. aurescens, N. cucumeris and N. californicus were all able to develop to the adult stage when fed on P. pallidus active stages. Neoseiulus aurescens has been reported as a predator of P. pallidus on strawberry in California (Strand 1994) and is one of the most common naturally occurring phytoseiid species found in Latvian strawberry fields (Petrova et al. 2004). However, in a small scale field experiment in strawberry in UK (Fitzgerald and Easterbrook 2003) there was no evidence of N. aurescens controlling P. pallidus numbers. Neoseiulus cucumeris has been shown to give control of P. pallidus in potted plant experiments in USA (Croft et al. 1998), and in UK both N. cucumeris and N. californicus reduced numbers of this pest (Easterbrook et al. 2001). Neoseiulus californicus has also been shown to be an effective biocontrol agent for T. urticae on strawberry in several countries (Oatman et al. 1977; Zalom et al. 1990; Garcia-Mari and Gonzalez-Zamora 1999; Greco et al. 1999; Easterbrook et al. 2001). Neoseiulus cucumeris also reduced T. urticae numbers in experiments in the UK (Easterbrook et al. 2001). Phytoseiulus persimilis is widely used as a biocontrol agent for T. urticae in a range of crops; its effectiveness on strawberry in UK was first demonstrated by Easterbrook (1992); it is a specialist predator of tetranychids and has been shown to aggregate on cucumber leaves with abundant T. urticae even where both mites were patchily distributed (Nachman 2006). In field experiments, releases of either P. persimilis or N. cucumeris significantly reduced T. urticae numbers (Fitzgerald et al. 2007); however, there was a significant interaction between these predator species, leading to poorer control of T. urticae when both species were released together. Phytoseiids were found on all plant parts in the sampling programme but the species were spatially separated, with most N. cucumeris and N. aurescens found on flower clusters and folded leaves, N. californicus on old and mature leaves, and P. persimilis on the older leaves; these distribution patterns remained the same thoughout the sampling periods. Thus, N. cucumeris and N. aurescens were found on the plant parts where most P. pallidus occurred and N. californicus and P. persimilis on plant parts where T. urticae were more abundant. This was highlighted by a significant positive association between N. californicus adults and T. urticae, and a negative association between both N. cucumeris and N. aurescens and T. urticae. There was a significant association between P. persimilis immature stages and T. urticae eggs and immatures and P. persimilis eggs and adults with T. urticae eggs and adults. This indicates that P. persimilis adults were laying eggs within developing colonies of T. urticae and then moving on to new colonies, leaving prey available for the immature stages; this is in line with results described by Vanas et al. (2006). Our results are the first to report the distribution of multiple species of phytoseiids on field grown strawberry, and show the co-incidence of different species of predatory mite with what is believed to be their preferred prey. The results are in line with previous findings of single predator: prey distribution patterns (e.g. Zhang and Sanderson 1995; Greco et al. 1999, 2004; Nachman 2006). However, in laboratory interaction experiments Fitzgerald et al. (2007) found that when given a choice of T. urticae or P. pallidus on leaf arenas, N. cucumeris and N. californicus showed no preference for either prey type. Also, Easterbrook et al. (2001) showed that N. californicus could reduce P. pallidus numbers, and N. cucumeris could reduce T. urticae numbers on strawberry. Thus mite prey distribution on the plant may not be the main reason for choice of habitat by the predators; availability of alternative food sources, such as nectar or pollen, microclimate or other physical aspects of the plant part are also likely to be important (e.g. Roda et al. 2000). Temperate zone phytoseiids overwinter as diapausing adult female mites (Overmeer 1985). Diapause is induced by short day conditions (e.g. Overmeer 1985; Fitzgerald and
123
Exp Appl Acarol (2008) 44:293–306
305
Solomon 1991). The winter sampling results showed that in established strawberry plantations large numbers of phytoseiids can overwinter successfully. The most abundant species in the winter sampling was N. cucumeris although low numbers of N. californicus were also found, confirming the results of Jolly (2001) who showed that UK collected strains of N. californicus were able to enter diapause. Most phytoseiids were found in the dead plant material; thus if dead plant material is removed as part of routine husbandry of strawberry plantings, large numbers of predators will be lost to the system. Phytoseiids were active on the plants in March; 14% of the phytoseiids collected were on plant leaves at this time, but 65% were still in the leaf litter. This is in line with experiments on diapausing Typhlodromus pyri (Fitzgerald and Solomon 1991), where mites were seen feeding on plants before diapause terminated and egg laying commenced. No T. urticae were found in these winter samples. This may indicate that the adult female mites diapaused off the plant, but more detailed sampling would be necessary to confirm this, as numbers of T. urticae in the planting in the 2002 season had not been recorded. Where P. persimilis were released in early July, numbers of adult mites continued to increase until the end of September, when sampling ended. This species does not overwinter successfully in the field in UK since it does not diapause, and continued to produce eggs until temperatures declined. Knowledge of the within-plant zonation of mite species is important when designing sampling strategies for mites. In Impatiens, a strategy for sampling T. urticae was developed based on the distribution of the mites on different plant parts (Alatawi et al. 2005). A presence-absence sampling strategy was developed for T. urticae and N. californicus on strawberry in Argentina (Greco et al. 2004) based on the reported spatial coincidence of the two species in glasshouse grown strawberry crops (Greco et al. 1999). Our results show that a programme to sample the entire mite system on strawberry should be stratified to include all parts of the plant. However, a different sampling protocol would be appropriate if sampling for each of the different pest species and their associated predators. Acknowledgements This research was funded by the UK Department of Environment Food and Rural Affairs (Defra). We thank Gael Perlet and Jennifer Butcher for technical assistance.
References Alatawi FJ, Opit GP, Margolies DC, Nechols JR (2005) Within plant distribution of two spotted spider mites (Acari: Tetranychidae) on impatiens: development of a presence–absence sampling plan. J Econ Entomol 98:1040–1047 Butcher MR, Penman DR, Scott RR (1987) The relationship between two-spotted spider mite and strawberry yield in Canterbury. N Z J Exp Agric 15:367–370 Cox DR, Snell EJ (1989) Analysis of binary data. Chapman and Hall, London Croft BA, Pratt PD, Koskela G, Kaufman D (1998) Predation, reproduction and impact of phytoseiid mites (Acari: Phytoseiidae) on cyclamen mite (Acari: Tarsonemidae) on strawberry. J Econ Entomol 91:1307–1314 Cross JV, Easterbrook MA, Crook AM, Crook D, Fitzgerald JD, Innocenzi PJ, Jay CN, Solomon MG (2001) Biocontrol of pests of strawberry in Northern and Central Europe. Biocontrol Sci Technol 11:165–216 Easterbrook MA (1992) The possibilities for control of two-spotted spider mite Tetranychus urticae on fieldgrown strawberries in the UK by predatory mites. Biocontrol Sci Technol 2:235–245 Easterbrook MA (1998) The beneficial fauna of strawberry fields in the south-east of England. J Hortic Sci Biotechnol 73:137–144 Easterbrook MA, Fitzgerald JD, Solomon MG (2001) Biological control of strawberry tarsonemid mite Phytonemus pallidus and two-spotted spider mite Tetranychus urticae on strawberry in the UK using species of Neoseiulus (Amblyseius) (Acari: Phytoseiidae). Exp Appl Acarol 25:25–36
123
306
Exp Appl Acarol (2008) 44:293–306
Easterbrook MA, Fitzgerald JD, Pinch C, Tooley J, Xu X-M (2003) Development times and fecundity of three important arthropod pests of strawberry in the UK. Ann Appl Biol 143:325–331 Fitzgerald JD, Easterbrook MA (2003) Phytoseiids for control of spider mite, Tetranychus urticae, and tarsonemid mite, Phytonemus pallidus, on strawberry in UK. Bull IOBC/WPRS 26(2):107–111 Fitzgerald JD, Solomon MG (1991) Diapause induction and duration in the phytoseiid mite Typhlodromus pyri. Exp Appl Acarol 12:135–145 Fitzgerald JD, Pepper N, Easterbrook MA, Pope T, Solomon MG (2007) Interactions among phytophagous mites and introduced and naturally occurring predatory mites on strawberry in the UK. Exp Appl Acarol 43:33–47 Garcia-Mari F, Gonzalez-Zamora JE (1999) Biological control of Tetranychus urticae (Acari: Tetranychidae) with naturally occurring predators in strawberry plantings in Valencia, Spain. Exp Appl Acarol 23:487–495 Greco NM, Liljesthrom GG, Sanchez NE (1999) Spatial distribution and coincidence of Neoseiulus californicus and Tetranychus urticae (Acari: Phytoseiidae, Tetranychidae) on strawberry. Exp Appl Acarol 23:567–580 Greco NM, Tetzlaff GT, Liljesthrom GG (2004) Presence-absence sampling for Tetranychus urticae and its predator Neoseiulus californicus (Acari:Tetranychidae; Phytoseiidae) on strawberries. Int J Pest Manage 50:23–27 Janssen A, Pallini A, Venzon M, Sabelis MW (1998) Behaviour and indirect interactions in food webs of plant-inhabiting arthropods. Exp Appl Acarol 22:497–521 Jolly RL (2001) The status of the predatory mite Neoseiulus californicus (McGregor) (Acari:Phytoseiidae) in the UK, and its potential as a biocontrol agent of Panonychus ulmi (Koch) (Acari:Tetranychidae). Ph.D. thesis, University of Birmingham, 173 pp Nachman G (2006) The effects of prey patchiness, predator aggregation, and mutual interference on the functional response of Phytoseiulus persimilis feeding on Tetranychus urticae (Acari: Phytoseiidae, Tetranychidae). Exp Appl Acarol 38:87–111 Oatman ER, McMurtry JA, Gilstrap FE, Voth V (1977) Effects of releases of Amblyseius californicus on the two-spotted spider mite on strawberry in southern California. J Econ Entomol 70:638–640 Overmeer WPJ (1985) Diapause. In: Helle W, Sabelis MW (eds) Spider mites, their biology, natural enemies and control, vol B. Elsevier, Amsterdam, pp 95–102 Payne R (2002) The guide to GenStatÒ Release 6.1—part 2: statistics. VSN International, Oxford Petrova V, Salmane I, Cudare Z (2004) The predatory mite (Acari, Parasitiformes: Mesostigmata (Gamasina); Acariformes: Prostigmata) community in strawberry agrocenosis. Acta Univ Latviensis Biol 676:87–95 Roda A, Nyrop J, Dicke M, English-Loeb G (2000) Trichomes and spider-mite webbing protect predatory mite eggs from intraguild predation. Oecologia 125:428–435 Rosenheim JA (1998) Higher-order predators and the regulation of insect herbivore populations. Ann Rev Entomol 43:421–447 Rosenheim JA, Kaya HK, Ehler LE, Marois JJ, Jaffee BA (1995) Intraguild predation among biological control agents-theory and evidence. Biol Control 5:303–335 Sances FV, Toscano NC, Lapre LF, Oatman ER, Johnson MW (1982) Spider mites can reduce strawberry yields. Calif Agric 36:15–16 Sih A, Crowley P, Mcpeek M, Petranka J, Strohmeier K (1985) Predation, competition, and prey communities: a review of field experiments. Ann Rev Ecol System 16:269–311 Stenseth C, Nordby A (1976) Damage and control of the strawberry mite, Stenotarsonemus pallidus (Acarina: Tarsonemidae), on strawberries. J Hortic Sci 51:49–54 Strand LL (1994) Integrated pest managment for strawberries. Univ. California Statewide IPM Proj, Publ. 3351, 142 pp Vanas V, Enigl M, Walzer A, Schausberger P (2006) The predatory mite Phytoseiulus persimilis adjusts patch-leaving to own and progeny needs. Exp Appl Acarol 39:1–11 Zalom FG, Pickel C, Welch NC (1990) Recent trends in strawberry arthropod management for coastal areas of the western United States. In: Bostanian NJ, Wilson LT, Dennehy TJ (eds) Monitoring and integrated management of arthropod pests of small fruit crops. Intercept, Andover, pp 239–259 Zhang ZQ, Sanderson JP (1995) Two spotted spider mite (Acari: Tetranychidae) and Phytoseiulus persimilis (Acari:Phytoseiidae) on greenhouse roses-spatial distribution and predator efficacy. J Econ Entomol 88:352–357
123
