INJURIES CAUSED BY Frankliniella spp . ( THYSANOPTERA : THRIPIDAE ) ON SEEDLESS GRAPES Injúrias ocasionadas por Frankliniella spp . ( Thysanoptera : Thripidae ) em uva sem semente

Thrips of the genus Frankliniella cause significant reduction on grape production. Little is known about its decision making approach in integrated pest management and information on thrips injuries to the vine is scarce, hindering the control of this pest. This study aimed to determine the species of thrips and injuries that cause damage to berries in seedless grapes clusters. The trial was conducted in an area cultivated with Vitis vinifera L., variety Thompson Seedless, using a randomized block design with five treatments and four replications. The treatments consisted of artificial infestation (zero, two, four and eight thrips per inflorescence), and treatment determined by the monitoring system of the grower. Each replication consisted of three plants with 10 inflorescences, which were artificially infested with thrips and covered with nonwoven fabric bags. During the cluster thinning, pre-harvest cleaning and harvest, the numbers of injuried berries and the berries with injuries were evaluated. The thrips species identified were: Frankliniella schultzei (Trybom), Frankliniella brevicaulis Hood, Frankliniella gemina Bagnall, Frankliniella gardeniae Moulton, and Frankliniella sp. A positive and significant correlation was observed between the number of injuries per fruit and the level of thrips infestation, obtained by the equation y = 2.55 + 48.54x (F= 59.98; P= 0.0001; r2= 0.51). The results suggest that the presence of a whitish halo around a small scar on fruit is associated with the presence of adult Frankliniella spp.

Decision making and information on thrips injuries to the vine is scarce, hindering the control of this pest.In addition, thrips injuries to the berries are difficult to be identified by the grower.Some authors (Jensen;Flaherthy;Luvise, 1981;Botton et al., 2007;Mujica et al., 2007) reported the presence of a whitish halo around a small scar at the place of posture, on the berries.Lopes et al. (2002) described a dark and suberized scar, in some cases causing Niagara grapes to crack and seeds to prolapse.These insects can cause other types of injuries, depending on crop and thrips species.In strawberry plants, F. occidentalis causes injuries in the stamens and floral receptacle, causing tanning on the surface of green and ripe fruit, without leading to deformations (Nondillo et al., 2010).In nectarine trees, this species was reported to cause tanning on the fruit surface and bark deformations (Pearsall, 2000;Tommasini;Ceredi, 2007).Pseudophilothrips sp.causes lesions in guava fruit, which can lead to fungal infections (Picanço et al., 2003).In bunches of bananas, Elixothrips brevisetis (Bagnall) causes rusting on the fruit peel (Lima;Milanez, 2013).
The objective of this study was to determine the injuries caused by different species of thrips in seedless grapes, grown in the Submédio do Vale do São Francisco region, in order to obtain subsidies for the establishment of economic injury level.

MATERIALS AND METHODS
The study was conducted from May to September 2010, in a commercial vineyard located in Casa Nova, Bahia,Brazil (09º15'48.7'' S and 40º 54'38.1'' W).The vines (V.vinifera variety Thompson Seedless) were five years old and the growing area covered 5.2 ha, with 2.5 m x 3.5 m spacing.
A randomized block design with five treatments corresponding to infestation levels of zero (control), two, four and eight thrips per inflorescence and treatment applied by the grower, who relied on monitoring the area using as action level 20% of inflorescences infested with two or more thrips (Haji et al., 2001).During the experiment, this level reached 60% of inflorescences infested, which is equivalent to 12 or more thrips per inflorescence.
The replications were three successive plants, and 10 inflorescences per plant.At 22 days from pruning, each inflorescence was infested with adult thrips from the field and with no specific age established.The thrips were separated manually by using a stereomicroscope, packed in gelatinous capsules and released in nonwoven fabric bags (34.4 cm length, 15.4 cm width at the closed end, 28.3 cm at the open end).The bags containing the number of thrips corresponding to each infestation level were placed involving the inflorescences, attached with flexible wire to the peduncle or to the branch, to prevent the escape of insects.The treatment without thrips (control) was also bagged.Thrips samples were kept in the Laboratory of Entomology at Embrapa Semiárido.
After nine days of infestation, the treatments were opened and the inflorescences sprayed with fungicide and the growth regulator gibberellic acid, afterwards being closed again to avoid reinfestation.The bags were definitively removed for cluster thinning, 43 days after pruning.The discarded material was collected in a white plastic tray, weighted, counted and the number of injuries was determined.This procedure was carried out at preharvest cleaning and harvest, 114 days after pruning.The treatment by the grower included two insecticide applications for thrips control during flowering (at 26 and 33 days after pruning) and one more application of gibberellic regulator.
The number of injuries and thrips infestation were compared through the linear regression model (y = 2.55 + 48.54x; r 2 = 0.51).The means, standard errors and regression were calculated through the PROC MEANS PROC REG by SAS (SAS Institute 2001).The numbers of injuries in berries and berries with injuries, as well as the percentage of damaged berries, were calculated through analysis of variance and the Tukey's test at 5% significance by the software SAS.The original data for the average percentage of injuries per number of berries were transformed into square root (x + 0.5).
The average percentage of berries with injuries, in relation to the total number of berries discarded during cluster thinning and pre-harvest cleaning, was higher in the treatment by the grower (7.0%), differing from the treatments with eight, four, two and zero thrips ( 2.7, 1.7, 2.3 and 0%, respectively) (Table 2).The same result was found at harvest, with 3.7% of berries damaged from the treatment by the grower (Table 2).This result shows that the higher the thrips infestation is, the greater is the number of berries discarded at cluster thinning and pre-harvest cleaning, as well as berries with injuries at harvest.
The number of injuries in berries and the thrips infestation levels showed a significant correlation characterized by the equation y = 2.55 + 48.54x (r 2 = 0.51; F = 59.98;P = 0.0001) (Figure 1).Some authors (Shipp;Wang;Binns, 2000;Wang;Shipp, 2001;Hong-Hyun;Joon-Ho;Ki-Baik, 2007;Tommasini;Ceredi, 2007;Funderburk, 2009;Reisig;Godfrey;Marcum, 2009) have also observed a relationship between thrips damage and intensity of infestation in other crops.In nectarine buds, a positive and significant relationship was observed between the density of F. occidentallis larvae and fruit damage (r 2 = 0.529; P = 0.02) (Pearsall, 2000).As a damaged table grape berry is discarded during cluster thinning, pre-harvest cleaning and harvest, the infestation of two adults thrips per inflorescence (Tables 1  and 2) is sufficient to cause the loss of a berry.This is related to the elimination of damaged berries that compromise the cluster's quality, including visible damages caused by insects, due to high market standards (Lima, 2009).
Espadas ( 2008) explains that eggs placed on the tender tissues of berries in formation and the fluids released by the female at the time of laying, in contact with oxygen, prevent the closure of the lesion, producing the disruption of cells, showing the whitish in the berry tissue.With the growth of the fruit, the skin tends to crack at the point of injury, facilitating the growth of rot in bunches.Jensen, Flaherthy andLuvise (1981), Ripa, Rodriguez andVargas (1993), Botton et al. (2007) and Mujica et al. ( 2007) observed similar injuries.In this research, the fruits at young berries (>2mm) phase showed more visible injuries than those at harvest, probably due to their green coloration.Damaged berries frequently had a single scar; however, up to six scars were found on a single berry.
Other injuries caused by several thrips species have also been observed in grape berries.F. occidentalis produced dark and suberized scars on Niagara grapes, in some cases also causing the cracking of the berries and the prolapse of the seeds (Lopes et al., 2002).In mature clusters, F. occidentalis can cause browning around the contact area of berries or around the insertion of the stalk of the bunch (Espadas, 2008).Drepanothrips reuteri (Uzel) caused linear-shaped tanned stains, and F. occidentalis caused ring-shaped stains on ripe berries of Thompson Seedless (Roditakis;Roditakis, 2007).This variation is possibly related to the different thrips species that occur on the vine and the diversity of grape varieties.These symptoms were not observed in experimental vineyard during the study.
Starfish-shaped injuries were also observed (Figure 2C), corroborating previous studies about Thompson Seedless grapes (Jensen, 1973;Jensen;Flaherthy;Luvise, 1981;Ripa;Rodriguez;Vargas, 1993).In the present work, these scars were more evidenced on the clusters from the treatment by the grower, i.e., those that were not bagged.Probably, this injury results from the persistence of the calyptra, which is fixed at the end or beside the berry, causing the scar.Gibberellic acid application at flowering can increase the number and visibility of these injuries, for stimulating cell growth.Jensen and Luvisi (1973) explained that this is associated to the flowering and fruiting time, and to the occurrence of lower temperatures.During the execution of the study, the temperatures were around 22° C, which are considered mild ones for the semi-arid region.The other treatments, including the treatment without thrips, also showed these scars, but in reduced numbers.As the bagged clusters received one less application of gibberellic acid in the flowering period and were protected from the rain, this could have resulted in a smaller number of scars on berries.
According Jensen and Luvisi (1973) and Ripa, Rodriguez and Vargas (1993), these injuries are aggravated by the larvae of F. occidentalis sheltering in persistent calyptras of Thompson Seedless variety, however, experiments conducted by these authors do not explain the appearance of these scars in treatments without the presence of these insects.Thrips larvae can cause injuries when they feed mainly on pollen and stigma tissues of the calyptra, causing tanning tissue (Ripa, Rodriguez;Espinoza, 2001).However, hitherto, similar symptoms were not found in grape clusters infested or not by thrips larvae in warm seasons or other grape cultivars in the Submédio do Vale do São Francisco region (personal communication).
The results indicate that Frankliniella sp.causes an injury characterized by the presence of a whitish halo around a small scar on the berries.Thus, the information generated during this study can be useful to guide the decision making of the grower as to whether or not to control this thrips complex.Nevertheless, the possibility that other species of thrips may contribute to the occurrence of injuries cannot be excluded, therefore, it is important to continue to track and monitor these populations in grapevine in the Submédio do Vale do São Francisco region.

CONCLUSIONS
Levels of infestation of thrips (Frankliniella schultzei, F. brevicaulis, F. gemina, F. gardeniae and Frankliniella sp.), identified during the cluster thinning, pre-harvest cleaning and harvest of grapes, were positively and significantly correlatade with the number of injuries per fruit and the level of thrips infestation.The results suggest that the presence of a whitish halo around a small scar on fruits is due to the oviposition of thrips, associated with the presence of adult Frankliniella sp.

1Figure 1 -
Figure 1 -Estimates of the number of injuries in Thompson Seedless grape berries, as a function of different levels of thrips infestation.Casa Nova, BA.

Table 1 -
Means (± SE) 1 of the number of injuries per berry and number of berries with injuries at cluster thinning, pre-harvest cleaning and harvest of Thompson Seedlees grapes, subject to different levels of thrips infestation.Casa Nova, BA.Means followed by the same letter, in the column, do not differ by the Tukey's test (P> 0.05). 1

Table 2 -
Means (± SE) 1 of the percentage of injuries, number of injuries per berries and number of berries with injuries at cluster thinning , pre-harvest cleaning and harvested of Thompson Seedless grapes, subject to different levels of thrips infestation.Casa Nova, BA.