ORIGINAL ARTICLES
Thrips (Thysanoptera) Resistance of Fig Cultivars and Relationship with the Ostiole Morphology of Young Fruits
2024 Volume 93 Issue 3 Pages 251-262
Details
2024 Volume 93 Issue 3 Pages 251-262
This study investigated how fruit morphology and development, especially ostiole openness of young fruit, affect the thrips (Thysanoptera) resistance of various fig (Ficus carica L.) cultivars. The rate of fruits with thrips (TFR) and the damage (DFR) varied among the 24 cultivars surveyed. TFR and DFR were highest in ‘Masui Dauphine’ (syn. ‘San Piero’), a popular cultivar in Japan, and moderate in ‘Brunswick’, ‘Archipel’, and ‘Saint Jean’, but relatively low to almost zero in many other cultivars. Cultivar differences in DFR were correlated with the difference in the rate of ostiole hole (a hole larger than 0.1 mm in diameter towards the internal pulp) presence. However, thrips were also detected from fruits in which no ostiole hole was observed throughout the growing stages, so they are likely to be able to pass through narrower gaps. Of two predicted morphologies determining ostiole openness, i.e., scale loosening around the ostiole surface and obstruction of the ostiole interior by flowers, only the former was correlated with DFR, so that scale loosening in longitudinal and radial directions relative to the ostiole was actually observed. A general linear model (GLM) for likelihood of thrips detection for individual fruits, with cultivar, scale looseness in longitudinal and radial directions, and interactions between both looseness as explanatory variables, had the best fit at 15 days after fruit set. The results showed that fig cultivars with less loosening of several surface scales at around 15 days after fruit set tended to be less susceptible to thrips invasion.
Thrips (Thysanoptera) are insect pests that invade fig fruits (Ficus carica) and cause browning of the fruit interior (Hansen, 1929); they are a long-standing problem in fig cultivation (Bailey, 1938; Condit, 1947). The first case of thrips damage in Japan was reported in Chiba prefecture (Sawada and Hagiya, 1973). It is difficult to confirm whether thrips are present by examining the fruit exterior, and thorough pest control is necessary to avoid distribution of damaged fruits (Fujimoto et al., 1987). The mainstay of pest control is the use of insecticides, but their repeated use promotes insecticide resistance (Shirotsuka et al., 2019). Physical control methods such as a silver poly mulch (Fujimoto et al., 1987; Morishita, 2002; Takahashi, 1986), red insect-proof nets (Shirotsuka et al., 2018), and covering the fruit ostiole with an adhesive tape (Ichikawa and Naganawa, 2004) are also available, but none is completely effective and all require substantial labour or installation costs. Thrips damage is still a serious problem for fig production worldwide (Ben-Yakir and Costa, 2022; Singh et al., 2022; Wohlfarter et al., 2011).
Thrips damage is said to vary between fig cultivars, and the use of insect-resistant cultivars is a sustainable, low-cost pest control method. However, studies of differences in thrips resistance among fig cultivars are very limited (Horikawa et al., 2023), and much of the understanding remains based on growers’ experience. In most fig fruits, the eye (ostiole) opens as they grow (Fig. 1), which allows fruits to be invaded by various external pests (Eisen, 1901; Himelrick, 1999; Krezdorn and Adriance, 1961), including thrips (Yamauchi and Takahashi, 1984). Differences in ostiole size have been hypothesized to affect fig cultivar differences in the risk of external pest invasion (Doyle et al., 2003; IBPGR, 1986). ‘Summer Red’, a mutant of ‘Masui Dauphine’ (syn. ‘San Piero’), has smaller ostiole openings than ‘Masui Dauphine’ and suffers less damage from thrips (Horikawa et al., 2023). However, whether other fig cultivars show similar patterns, and the relation to degree and timing of ostiole opening are not yet known. This study compared thrips damage in various fig cultivars under experimental conditions, and its relationship to the growth and morphological characteristics of young fruits, especially ostiole characteristics.
Diagram of a longitudinal section of a young fig fruit.
Observations were conducted in 2019–2022 in an experimental field of the Toyo Institute of Food Technology, Hyogo, Japan (34°49'N; 135°24'E). The field site was 945 m2 in area, 38 m above sea level, and dressed by decomposed granite soils on a lowland paddy soil. As shown in Table 1, 74 trees of 24 fig cultivars planted in 2008, 2009, 2017, or 2020 at a density of 95 trees per 1,000 m2 were used for the survey. Representative names of the cultivars were identified by integration of their seedling label names based on the literature (Condit, 1955; Simonet et al., 1945) as reported (Hosomi and Takahashi, 2023). These representative names are used in the following descriptions, but as exceptions, the name ‘Masui Dauphine’, a common synonym in Japan, is used for ‘San Piero’, and the label name ‘Dreamy Sweet’ is used for its unidentified representative name. All of the trees were base-form trained, except for two trees each of ‘Bordeaux’, ‘Du Japon’, and ‘Précoce de Barcelona’, which were trained by straight-line training (Kabumoto et al., 1985). Thirty to 35 shoots per tree elongated as fruit-bearing shoots and all other shoots were removed. Each of the fruit-bearing shoots was pinched out when it had elongated beyond ca. 180 cm above ground. Pest control was carried out in late July, to avoid peak thrips occurrence; only acetamiprid 20% water soluble powder (Mospilan; Nihon Nohyaku Co., Tokyo, Japan) and triflumizole 30% wettable powder (Trifmin; Nippon Soda Co., Tokyo, Japan) were applied in 2020 and 2021, and only azoxystrobin 30% wettable powder (Amistar 10; Syngenta Japan, Co., Tokyo, Japan) in 2022. Other details of cultivation, such as irrigation and fertilization, followed typical practices for fig orchard management.
Fig cultivars surveyed.
The seasonal prevalence of thrips was monitored each year through the use of blue sticky traps (Horiba Blue; Arista Life Science Co., Tokyo, Japan). Four traps were set with one in the centre of each quarter section of the field, at 1.5 m above the ground. The traps were replaced every ca. 10 days from early May to late October. The number of thrips of each species classified according to Chiwaki et al. (1994) attached to each trap was counted.
Mature fruits were collected from the 1st–5th nodes of randomly selected shoots for each tree in 2019, and from the 1st–5th, 6th–10th, 11th–15th, and 16th–20th nodes of 4 to 6 moderately growing shoots for each tree in 2020, 2021, and 2022. Because ‘Panachée’ had an extremely low fruit set rate, fruits from this cultivar were randomly collected from all shoots without regard for node segmentation. Fruits were halved lengthwise with a kitchen knife and observed with the naked eye and with a dissecting microscope (SMZ-10-1; Nikon Co., Tokyo, Japan) under magnification at 10–40×. Fruits with browning and an irregular glossy yellow pattern on the fruit pulp, which could easily be distinguished with the naked eye, were considered to have thrips damage (Takahashi and Yamauchi, 1986). The percentage of fruits with damage symptoms was recorded as the thrips-damaged fruit rate (DFR) for each node range of each shoot. The presence of thrips was confirmed by microscopy in marked basal fruits (see next section) in 2019, and in all DFR survey fruits (including marked basal fruits) in 2020, 2021, and 2022. The percentage of fruits with thrips (including any species, adult or larval, alive or dead) was recorded as the thrips-detected fruit rate (TFR) for each node range of each shoot. If the number of surveyed fruits for each node range of each cultivar in each year was lower than 8, calculations of DFR and TFR were omitted.
Relationship between thrips damage and fruit growth characteristics, including ostiole opennessIn 2019, an experiment using 46 trees of 13 cultivars, ‘Barnissotte’, ‘Blanche’, ‘Bordeaux’, ‘Bourjassotte Grise’, ‘Bourjassotte Noire’, ‘Brown Turkey’, ‘Brunswick’, ‘Dottato’, ‘Du Japon’, ‘Franciscana’, ‘Ischia’, ‘Masui Dauphine’, and ‘Précoce de Barcelona’ was conducted. Six shoots with medium basal width were selected for each tree, and the earliest or second-earliest young fruit on each shoot was chosen at fruit set (ca. 3 mm fruit diameter) and designated as a marked basal fruit (MBF). An enlargement curve was drawn for each MBF by measuring fruit width (diameter of the equatorial plane) with digital callipers (AD-5765A-150; A&D Co., Tokyo, Japan) every 3–4 days from ca. 5 mm until maturity, as reported in Hosomi and Takahashi (2023). From each enlargement curve, growth stages of rapid enlargement (Stage I), stagnation (Stage II), and rapid re-enlargement (Stage III) were classified as in Crane (1948), and the fruit widths at 10, 15, 20, 25, 30, and 35 days after fruit set and at the end of Stage I were read again. Presence/absence of an ostiole hole larger than a pinhole (ca. 0.1 mm diameter) to the internal pulp of the fruit, was also recorded under observation through a loupe (5×). The percentage of MBFs with an ostiole hole from 10 ± 2, 15 ± 2, 20 ± 2, 25 ± 2, 30 ± 2, and 35 ± 2 days after fruit set in each cultivar was calculated and recorded as the “ostiole hole rate”. Fruit widths of mature MBFs were measured, and fruits were halved lengthwise to measure the area ratio of the inner cavity cross section (cavity area). The cavity area ranks were scored as (0) almost none, (1) < 5%, (2) 5–20%, or (3) > 20% of the total cross section. The presence of thrips was recorded as in the previous section.
More precise ostiolar morphology of growing MBFs was also observed. As it is difficult to observe the interior of a growing fruit non-destructively, it was estimated from the width of other fruits collected for calibration as follows. For each cultivar, a total of 36 or 72 young fruits were collected from the 2nd, 4th, 6th, 8th, 10th, and 12th nodes of 6 or 12 shoots, about 40 days after the beginning of fruit set. Fruit width was measured as above, and then longitudinal sections cut with a razor blade were observed through the loupe (5×). Observed looseness of several surface layers of scales (ostiole surface openness) and degree of underdevelopment of flowers filling ostiole interior (ostiole interior openness) were ranked (0–3) against the rubric shown in Figure 2. For each cultivar, logistic regression equations were produced to predict logit of ostiole openness (response variable) from calibration fruit width (explanatory variable) by a general linear model (GLM) with “binomial” as the linear predictor and “logit” as the link function, in R ver. 4.1.2 software (R Core Team, 2021). The logistic regression equations were inverse transformed as:
Rubric of ostiole openness as affected by scale loosening and flower development in young fig fruits. The example plate is a cross-section of a young fruit ostiole of ‘Masui Dauphine’. Ostiole surface openness is ranked by the looseness of several surface layers of scales, as (0) tightly interlocking, (1) loosening in radial and longitudinal directions relative to ostiole, opening a tiny passageway, (2) loosening radially and collapsing, opening a passageway, and (3) fully open. Ostiole interior openness is ranked by the degree of underdevelopment of filling by flowers deep in the ostiole, as (3) underdeveloped, (2) slightly developed, (1) developed but a gap remains, or (0) fully filling the gap.
where PR denotes the predicted rank of the ostiole openness, w is measured fruit width (mm), a is the partial regression coefficient, and b is the intercept, subscripts 1 and 2 correspond to ostiole surface and interior, respectively. The width of each MBF at 10, 15, 20, and 25 days after fruit set (that at 30 and 35 days was excluded because increments in w were negligible) was used to predict PR1 or PR2 on each date.
Spearman’s correlation coefficient between the growth parameters above (fruit set date, end date of Stage I, maturity date, fruit width, ostiole hole rate in each date range, PR1, and PR2 on each date) of MBFs in 2019 and DFR of the shoot basal (1st–5th) nodes of each cultivar in 2019 (see previous section) was calculated using Microsoft Excel software (Microsoft Corp., Redmond, WA, USA, 2016).
Relationship between thrips damage and scale loosenessIn 2020, loosening of several surface layers of growing fruit scales was observed on 54 trees of 16 cultivars, ‘Barnissotte’, ‘Blanche’, ‘Bordeaux’, ‘Bourjassotte Grise’, ‘Bourjassotte Noire’, ‘Brown Turkey’, ‘Brunswick’, ‘Dottato’, ‘Drap d’Or’, ‘Du Japon’, ‘Franciscana’, ‘Houraishi’, ‘Ischia’, ‘Masui Dauphine’, ‘Précoce de Barcelona’, and ‘Verdone’. A young fruit on any of the basal (1st–5th) nodes of 6 moderately growing shoots for each tree of each cultivar was selected as MBF in 2020. Scales around the MBF ostiole were observed through a loupe (5×) every 3–4 days from fruit set to maturity, and the degree of scale looseness in both the radial and longitudinal directions relative to the ostiole was ranked from 0 to 3 as shown in Figure 3. Scale looseness of each fruit at 5, 10, 15, 20, 25, and 30 days after fruit set was ranked and if data for a measurement day were missing, the rank was approximated by 0.5 units from the rankings before and after.
Examples of surface scales of a young fig fruit with various degrees of looseness. Degree of looseness was ranked (0–3) in radial and longitudinal directions relative to the ostiole. Isolated scales that formed far from the center of the ostiole were ignored.
Spearman’s correlation coefficient between the mean scale looseness rank at each date in each cultivar and DFR of the shoot basal (1st–5th) nodes of each cultivar in 2020 was calculated using Microsoft Excel. The relationship between scale looseness and thrips detection of individual MBFs in 2020 was analysed by GLM in R ver. 4.1.2. The model was constructed with “cultivar”, “radial scale looseness”, “longitudinal scale looseness”, and the interaction between the two looseness variables as explanatory variables; thrips detection (1, presence; 0 absence) as the response variable; “binomial” as the linear predictor; and “logit” as the link function. P values of each explanatory variable and fitness of the model, based on Akaike’s information criteria (AIC), were compared between dates.
Thrips were found in the traps throughout the monitoring period, and classified into 5 species that are known to invade fig fruits and 7 other species. The abundant thrips species were Thrips tabaci Linderman, Frankliniella intonsa Trybom, T. hawaiiensis Morgan, T. flavus Schrank, and T. coloratus Schmutz, all of which are known as fig invading species (Miyazaki and Kudo, 1988). Seasonal changes in the total numbers of the fig-invading thrips trapped is shown in Figure 4. Although thrips abundance varied from year to year, most individuals were trapped from May to July, with a peak in the first half of June. A similar trend was reported for thrips in fig orchards in Japan (Fujimoto et al., 1987; Morishita, 2002), and the experimental field was likely exposed to invasion pressure by thrips from May to July every year.
Seasonal prevalence of thrips as monitored by blue sticky traps hung in the experimental field (1 trap per 263 m2) roughly every 10 days. The total number of Thrips tabaci, Frankliniella intonsa, T. hawaiiensis, T. flavus, and T. coloratus, all of which are known to invade figs, per trap per day is shown.
Thrips damage by cultivar is shown in Table 2. DFR and TFR were closely linked and varied by fig cultivar and by shoot node range. In general, thrips damage decreased towards the apical nodes of shoots. This pattern is well known in ‘Masui Dauphine’ and is considered to result from the time lag between the fruit growth stage that allows thrips to invade and the seasonal peak of thrips abundance (Fujimoto et al., 1987; Yamauchi and Takahashi, 1984). Here, too, thrips invasion pressure likely decreased as the season progressed, resulting in less damage to more apical fruits.
Thrips damage in various fig cultivars (2019–2022).
Continued
DFR and TFR varied greatly among cultivars, and cultivars differed somewhat between years. Roughly, DFR and TFR were extremely high in ‘Masui Dauphine’; moderate in ‘Archipel’, ‘Brunswick’, and ‘Saint Jean’; relatively low in ‘Barnissotte’, ‘Bourjassotte Grise’, ‘Bourjassotte Noire’, ‘Brown Turkey’, ‘Dottato’, ‘Drap d’Or’, ‘Du Japon’, ‘Genoa’, and ‘Malta’; and low or almost zero in ‘Blanche’, ‘Bordeaux’, ‘Dreamy Sweet’, ‘Franciscana’, ‘Houraishi’, ‘Ischia’, ‘Panachée’, ‘Précoce de Barcelona’, and ‘Verdone’. Damage to ‘Conadria’ and ‘Longue d’Août’ was also low or almost zero, but because these trees were younger, this result requires further investigation.
‘Masui Dauphine’, the cultivar most susceptible to thrips invasion in this study is a popular fig cultivar in Japan, accounting for 70% of Japan’s production (Aljane et al., 2022) because of its outstanding fruit size and yield. It is also known as ‘San Piero’ (Condit, 1955), and it is grown worldwide under names such as ‘Seoungjeongdoupin’ in Korea (Lim et al., 2018), ‘Brown Turkey’ in the USA (Ferguson et al., 1990), and ‘Roxo de Valinhos’ in Brazil (Aksoy and Flaishman, 2022). The severity of thrips damage to this cultivar has been reported not only in Japan, but also in Korea (Kim et al., 2014). Of all cultivars surveyed here, ‘Masui Dauphine’ showed the worst thrips damage. As many cultivars suffered little or no damage, there are many options for a thrips-free cultivar if some reduction in fruit size and yield are acceptable.
Relationship between thrips damage and cultivar differences in fruit growth and ostiole morphologyTable 3 shows the ostiole hole rate for MBFs, and the correlation between these characteristics and DFR of basal (1st–5th) nodes in 2019. DFR was significantly correlated with the ostiole hole rate at 20 ± 2, 25 ± 2, and 30 ± 2 days after fruit set. Thrips are known to invade young fig fruits when the ostiole opens (Fujimoto et al., 1987; Sawada, 1977; Yamauchi and Takahashi, 1984); the present results are consistent with this observation. The hole size required for thrips invasion has not been reported, but Fujimoto et al. (1987) and Ichikawa and Naganawa (2004) stated that ostiole opening starts in ‘Masui Dauphine’ at fruit widths of 20–30 mm (15–20 days after fruit set) and 20 mm, respectively, which is consistent with the trends in fruit width and ostiole hole observed here (Table 3). The observer’s perception of “ostiole opening” in previous reports does not seem to differ significantly from the ostiole hole criteria used here. However, Wohlfarter et al. (2011) reported a case of thrips damage in fruits without ostiole opening. Morishita (2002) also observed thrips emerging from fruit that had not yet opened, and inferred that thrips can invade even when ostiole opening has not yet been confirmed. Here, thrips were also detected in fruits in which no ostiole hole was observed throughout the growth stages (from fruit set to the end of Stage II), e.g., in 4 out of 5 ‘Masui Dauphine’ fruits and 6 out of 11 ‘Brunswick’ fruits (Table 3). Thrips that invade figs, for example T. tabaci, are minute, with bodies ca. 1.0 mm long and 0.13 mm wide (Murai and Toda, 2002). Other species are similar in size (Saikawa, 1988), so they are probably able to pass through even narrow gaps. An ostiole hole large enough to be recognized by an observer is an indicator, but not a requirement, for thrips damage susceptibility.
Growth characteristics (date, size and ostiole hole) of young fig fruits and correlation with thrips-damaged fruit rate (2019).
The ostiole openness, i.e., the state of the thrips pathway, is understood to be when several surface layers of scales are loose and the flowers have not yet obstructed the ostiole interior (Sawada, 1977). The openness of both the surface and interior of the ostiole changed with growth as shown by regression equations derived from calibrated fruits. Mean predicted values for each cultivar and the relationship with DFR at the 1st–5th nodes in 2019 are shown in Table 4. Predicted ostiole interior openness tended to decrease with fruit growth, and there was no significant correlation with DFR. Obstruction of the ostiole interior by flowers was predicted to be complete simultaneously (around 20 days after fruit set) in many cultivars, which may have resulted in the low correlation between ostiole interior openness and DFR. On the other hand, predicted ostiole surface openness tended to increase with fruit growth and was significantly correlated with DFR from 15 to 25 days after fruit set. Thus, only the surface openness of the ostiole, i.e., the entrance, was related to cultivar difference in thrips invasion.
Predicted ostiole openness of young fig fruit and correlation with thrips-damaged fruit rate (2019).
Ostiole surface openness could be assessed only by external observation without dissecting the fruit, so in the following year’s study, surface openness was measured by actual observation of the radial and longitudinal looseness of several surface layers of scales. Results are shown in Table 5. Looseness in both directions tended to increase with days after fruit set, and ranking on the same days after fruit set differed among cultivars. Radial or longitudinal scale looseness was significantly correlated with DFR at the 1st–5th nodes in 2020 (Spearman’s correlation coefficient), but over a wide range of days (15–25 days) after fruit set with no specific timing. GLM analysis was conducted using “cultivar”, “radial scale looseness”, “longitudinal scale looseness”, and the interaction between the two looseness variables as explanatory variables to identify the factors contributing to individual fruit variation in thrips detection. Cultivar effect was significant at several times and between several pairs of cultivars, but was omitted to avoid complexity, and only the effect of scale looseness, the main target of this analysis was assessed. At 15 days after fruit set, the model had the best fit (lowest AIC), and the effects of radial and longitudinal scale looseness were both significant.
Observed ostiole surface openness of young fig fruit and correlation with thrips-damaged fruit rate or thrips detection (Yes/No) (2020).
Table 3 also shows the date of fruit set and subsequent growth stages, as well as fruit width and cavity area for MBFs. DFR was correlated with fruit width at all stages of growth and with the cavity area rank of mature fruit: wider fruits and those with a greater cavity area had higher DFR. Although fruit width was unlikely to have directly affected thrips invasion, the internal cavity area may contribute to damage by providing space for thrips activity within the fruit (Horikawa et al., 2023). However, even if cavity area does affect thrips damage, this factor comes into play only after thrips have invaded the fruit and is unlikely to affect thrips presence or absence. Horikawa et al. (2023) reported that the rate of cavities correlates with the ostiole opening rate. I also found that fruit width and cavity area rank were significantly correlated with the ostiole hole rate at 25 ± 2 days after fruit set (Table 3). Here, it is likely that fruit width and cavity area rank were only collinear with ostiole opening.
Dates of fruit set, end of Stage I, and maturation had no significant correlation with DFR (Table 3), and there was no conspicuous peak period (scatter diagram, not shown). As mentioned in the previous section, thrips damage can occur only if the fruit growth stage that permits thrips invasion coincides with their presence, which is why fruits on basal nodes of shoots have a higher damage rate. Ostiole opening, which permits thrips invasion, occurs during Stage I of fruit growth (Fujimoto et al., 1987). Stage I of the MBFs in this study ranged from 22 May–13 June (earliest–latest) to 26 June–15 July, which coincides with peak thrips abundance (Table 3). It is concluded that there was no significant difference in the timing of thrips invasion pressure among the MBFs of all cultivars.
In conclusion, fig cultivars with less scale loosening (ostiole surface openness) at around 15 days (not exceeding 20 days) after fruit set tended to be less susceptible to thrips invasion. Entry of insect pests through the fruit ostiole has long been a problem in fig cultivation (Eisen, 1901). One goal of breeding programs has been to keep the ostiole as small as possible to prevent insect pest invasion (IBPGR, 1986; Krezdorn and Adriance, 1961; Storey and Condit, 1969; Stover et al., 2007). ‘Conadoria’, for example, was developed with this in mind (Condit and Warner, 1956). Fruits with scale looseness, especially radial looseness, with a rank of < 1 at 15 days after fruit set suffered little damage (Table 5). As mentioned above, the conventionally recognized ostiole hole is too large for thrips invasion, but a smaller ostiole gap may serve as a physical threshold for thrips invasion if 15 days after fruit set is absolute timing for thrips invasion, and this factor can potentially be used to assess resistance to thrips.
However, the cultivar differences in thrips invasion are not be entirely determined by the morphological condition of the ostiole. For example in 2019, ostiole hole detection of ‘Brown Turkey’ (not a synonym for ‘San Piero’), was earlier than ‘Masui Dauphine’ (Table 3), but ‘Brown Turkey’ fruits had substantially lower DFR and TFR (Table 2). Other experiments with ‘Masui Dauphine’ figs showed no thrips invasion of young fruits prior to 15 days after fruit set, even if the ostiole was clearly open (Hosomi, unpublished). These facts suggest that thrips do not invade very young fruits (prior to 15 days after fruit set), even if the ostiole is open. Even if ostiole openness is a major factor in cultivar differences in terms of thrips damage, the preference of thrips for fruits in a certain physiological state in different growth stages should also be considered. Further investigations, focusing on thrips preferences and invasion behaviour, should be conducted.
Special thanks to Kanako Shirotsuka and Dr. Manabu Shibao for their expert professional advice on thrips identification and survey methods. I am also grateful to Toru Takahashi and Eijiro Hoshiko for their kind support throughout the study.
