Effect of host plant species and temperature on the development and survival of the plant bug Closterotomus trivialis (Costa) (Hemiptera: Miridae)

and host plant significantly affected the nymphal development of C. trivialis . Specifically, host plant affected the development of nymphs at lower and higher temperatures (15, 20, 27 °C) but not at the optimum (24 °C) for its development temperature. Adults of C. trivialis lived longer on sweet orange, M. annua and S. alba in most tested temperatures compared to U. urens , P. diffusa and olives. Overall, these results suggest a better suitability of M. annua , S. alba and sweet orange compared to U. urens , P. diffusa and olive which were proven to be less suitable host plants, covering partially the nutritional needs of C. trivialis . The estimated lower temperature developmental threshold based on the linear model for C. trivialis was found to be lowest on M. annua (3.30 °C) and highest on P. diffusa (10.7 °C). Τhe assessment of the nymphal development in various host plants and temperatures is particularly important for understanding the biology of C. trivialis and pro-vides useful information to optimize its management strategy under integrated pest management system.

Vol:. (1234567890) Spain, Portugal, Cyprus, Corsica, Malta) (Aukema, 2017;Barbagallo, 1970;Carapezza & Mifsud, 2015;Garcia-Marí, 2013;Yamvrias, 1998). Being a polyphagous species, its presence is common in olive and citrus orchards but also in apricot and peach as well as on a large number of weed species, the most important of which include Mercurialis annua L.  (Barbagallo, 1970;Gerakaki et al., 2007;Kalaitzaki et al., 2013;Varikou & Birouraki, 2014;Yamvrias, 1998). In Italy and Greece, it completes one generation annually (univoltine). The eggs are oviposited in bark cracks and in the exposed wood of the trees at the end of spring. It overwinters at the egg stage and hatching begins in late January and then the nymph completes five instar stages to reach adulthood (Barbagallo, 1970;Yamvrias, 1998).
Early instar nymphs appear in late winter -early spring and start feeding on the young vegetation and flowers of under canopy flora while later instar nymphs and adults of the pest can move on orchard trees causing occasionally serious bud and flower abortion (Barbagallo, 1970;Kalaitzaki et al., 2013;Monaco, 1975;Varikou & Birouraki, 2014). In olive and citrus orchards of Crete (Greece), its populations start to increase from January, peaking from late January to early April while after mid-May, no individuals are recorded in the orchards (Kalaitzaki et al., 2013). Olive and citrus trees are more susceptible to C. trivialis damage from the early emergence of inflorescence until the initiation of blooming. The presence of eight individuals of C. trivialis per shoot has been reported to induce damage on the fruit setting of olive trees (Perdikis et al., 2009), while this density did not cause significant damage in citrus (Perdikis et al., 2010).
Classification of the insect is difficult in terms of pest severity since, although its population levels are usually high on weed host plants in the olive and citrus orchards, only occasional outbreaks have been reported to cause significant losses in olive and citrus production (Barbagallo, 1970;Kalaitzaki et al., 2013;Perdikis et al., 2009;Varikou & Birouraki, 2014). Likely, these outbreaks might be due to the colonization of olive and citrus trees with nymphs and adults of the pest migrating from the under canopy flora i.e. when the latter has been destroyed from herbicides or desiccation (Kalaitzaki et al., 2013;Varikou & Birouraki, 2014). Therefore, studies on host plant suitability of polyphagous insect pests which feed on both crop and non-crop host plants are essential for understanding their crop colonization rates and their damage potential and provide valuable input for the establishment of effective integrated pest management strategies (Panizzi, 1997;Panizzi & Parra, 2012).
Plant species can affect development and survival of polyphagous herbivore insects via their chemical (e.g., nutritional quality, allelochemicals) or physical (e.g., trichomes, tissue hardness) characteristics (Bernays & Chapman, 1994;Panizzi & Parra, 2012;Scriber, 1984). Plant attributes affect various biological parameters of insects: development, survival and fecundity and consequently their population density (Perdikis & Lykouressis, 2000;Rogers & Sullivan, 1986;Scriber & Slansky, 1981). In order to be qualified as a suitable host for a herbivorous insect, a specific plant species must allow insect development and successful completion at least part of its life cycle (Zalucki et al., 1986). The suitability of a particular host plant is often indicated by short developmental time and low mortality rates of an insect species (Awmack & Leather, 2002;Panizzi, 2000).
Evidence on the host plant suitability of C. trivialis can be inferred from previous field monitoring studies in olive and citrus groves in Crete which suggested its preference for M. annua followed by R. raphanistrum, Urtica sp., P. officinalis and S. alba over olive or orange trees as indicated by its population levels (Kalaitzaki et al., 2013;Varikou & Birouraki, 2014). Even though the above-mentioned studies offer valuable information, its nymphal development, mortality and adult longevity have been studied only on R. raphanistrum (Varikou & Birouraki, 2014). Furthermore, knowledge on the potential of each host plant species to support development and survival of the pest at different temperatures would also provide useful information for ranking the host plant suitability, especially focusing on relatively low temperatures occurring during the period of pest presence in the orchards (from January to May). Estimation of the lower temperature developmental threshold and thermal requirements of C. trivialis will also be useful in the prediction of population dynamics, crop colonization and the potential damage, for optimizing its management in olive or citrus orchards.

3
Vol.: (0123456789) Under this light, this study aimed to: (1) study the development and survival of nymphs as well as adult longevity of C. trivialis when reared on 6 different host plants (Olea europaea L. cv. 'Koroneiki' (olive), Citrus sinensis L. cv. 'Washington Navel' (sweet orange), M. annua, U. urens, P. diffusa and S. alba) under four constant temperatures; (2) estimate the lower developmental temperature thresholds using a linear model; and (3) estimate the thermal requirements for development of the insect, on each host plant.

Material and methods
Instar origin, nymphal mortality and development First instar nymphs of C. trivialis (as described by Hiremath & Viraktamath, 1992 for the relative species Calocoris angustatus Lethierry) were collected from six different host plants species (O. europaea cv. 'Koroneiki' (olive), C. sinensis cv. 'Washington Navel' (sweet orange), M. annua, U. urens, P. diffusa and S. alba) from an olive (35°29′12.8"N 24°01′28.0"E) and a citrus (35°29′31.9"N 24°03′00.1"E) orchard in the Chania prefecture (Crete, Greece), from January to April. Aerial parts of C. trivialis host plants were transferred from collection sites to the laboratory and inspected for the presence of first instar nymphs (eggs are not visible since they are inserted into bark cracks of trees) which were collected and transferred individually to plastic Petri dishes (Ø9 cm). Inside each Petri dish, two small stems (7-8 cm in length) with flower buds, flowers and leaves of the respective host plant were placed and the cut ends of stems were wrapped with a watersoaked cotton layer. Over 27 first-instar nymphs (replicates) were used per host plant. To ensure proper ventilation inside each Petri dish, a 3 cm-diameter hole was opened on the lid covers which was sealed with a thin muslin.
The mortality and developmental time of the nymphs were recorded at 12 h intervals until adult emergence. Stems in each Petri dish were replaced at a 24 h basis while the cotton was rewatered. During inspection of the Petri dishes, the presence of exuvium was used as evidence for successful molting and transition from one instar to the next. Upon adult emergence, the sex of the adults was determined. In the estimation of the nymphal development, only data from individuals that reached adulthood were used. Also, since the exact time of hatching of the first stage nymphal instars could not be determined accurately, measurements started after the molting of the nymph to the second instar (the length of the first stage nymphal instars development time is not shown in the respective tables).

Adult longevity
Randomly selected, newly emerged female and male adults (less than 24 h-old), derived from the nymphs used in the previous experiment, were used to determine the effect of host plant on C. trivialis longevity. The newly emerged adults were transferred in pairs in cylindrical transparent plastic cages (9 cm in diameter and 21 cm in height) and reared at the same conditions (temperature, host plant) as in the preimaginal development tests, until their death. In each cage, three small stems (7-8 cm in length) with flower buds, flowers and leaves of the tested host plant were placed in a small plastic jar (25 ml) filled with water. Stems were replaced with fresh ones every 24 h. All cages were inspected daily and mortality of the adults were recorded throughout their lifespan.

Relationship between temperature and developmental rate
Thermal requirements for the development of C. trivialis on each host plant were estimated using the linear model as described by Campbell et al. (1974). The temperature threshold (t) for development was estimated by extrapolating to zero the linear part of the equation representing the relationship between temperature and development rate (the reciprocal of the average development period in days). This relationship is described by the regression equation y = a + bT where y is the rate of development, T is the temperature in o C and a and b are constants estimated with the least squares method using JMP 16.1 (SAS Institute Inc., Cary, NC, USA).
The temperature threshold was calculated as T low = -a/b and the thermal constant K (i.e., the amount of heat units required for development) as K = 1/b degree-days (DD) (Campbell et al., 1974). In these calculations, only the rates of development included in the linear part of development curve were used.

Statistical analysis
Percentage of nymphal mortality on different host plants within each temperature were compared using the chi square test. Furthermore, all possible pairwise comparisons between host plants within each temperature were performed using the chi square test.
The evaluation of the effects of either different temperatures or different host plants on total nymphal developmental time, developmental time of each preimaginal stage, and adult longevity was performed by ad hoc two-way analysis of variance (ANOVA). Analyses were followed by post hoc Tukey's test HSD (honestly significant differences) to compare the significance between the means (p < 0.05). Post hoc analysis on the effect of the sex on nymphal developmental time obtained from the study were compared using Student's t-test, at p < 0.05. Data analysis was carried out using the statistical program JMP 16.1 (SAS Institute, 2021).

Nymphal mortality
C. trivialis could successfully complete its development on all tested host plants and temperatures, except in the case of U. urens at 15 and 20 ο C at which no nymph survived to adulthood (Table 1). Mortality rates varied widely among different host plants within each temperature (20.5 -100%) ( Table 1). Specifically, at 15 and 20 °C, the highest mortality (100%) was observed on U. urens and lowest on M. annua and S. alba (15 °C: x 2 = 45.14, df = 5, p < 0.0001; 20 °C: x 2 = 83.58, df = 5, p < 0.0001). At 24 °C, the highest mortality was observed on U. urens and the lowest on S. alba and sweet orange (x 2 = 50.81; df = 5, p < 0.0001). No significant differences were found in mortality rates among the 6 tested host plants at 27 o C (x 2 = 5.11, df = 5, p = 0.402).

Nymphal development
The total period of nymphal development on each host plant at each temperature tested is shown in Table 2. Two-way ANOVA showed that both host plant (F 5, 338 = 20.05, p < 0.0001) and temperature (F 3, 338 = 25.90, p < 0.0001) significantly affected the nymphal development period of C. trivialis. Furthermore, the effect of their interaction on developmental rate was also significant (F 13, 338 = 9.92, p < 0.0001) ( Table 2). The shortest total developmental periods were recorded at 24 °C on U. urens (11.60 days), P. diffusa (12.05 days), S. alba (12.09 days) and sweet Sex did not significantly affect the length of development periods in any of the temperature/host plant combinations tested (t = 1.41, p = 0.16).

Adult longevity
The two-way ANOVA showed that both temperature (F 3,277 = 28.86, p < 0.0001) and host plant (F 5,277 = 16.67, p < 0.0001) significantly affected adult longevity of C. trivialis. Furthermore, the effect of their interaction on adult longevity was also significant (F 13,277 = 3.21, p = 0.0002) ( Table 4). Adults' longevity was significantly longer when they were reared on S. alba, M. annua and sweet orange compared to P. diffusa at 15 °C, and on sweet orange at 20 °C, compared to P. diffusa and olives. At 24 °C, C. trivialis lived significantly longer on S. alba and sweet orange compared P. diffusa and U. urens. At 27 °C, no significant differences were found on adulthood longevity of C. trivialis between the tested host plants.

Relationship between temperature and developmental rate
The linear model fitted well the data on developmental rate on all tested host plants (excluding U. urens) ( Table 5). The estimated temperature threshold for the nymphal development of C. trivialis was found to be lowest on M. annua and highest on P. diffusa. Specifically, the temperature thresholds for development were estimated to be 3.30 °C on M. annua, 5.97 °C on olive, 7 °C on sweet orange, 8.13 °C on S. alba and 10.7 oC on P. diffusa (Table 5). Lowest numbers of degree-days were found on P. diffusa and highest on M. annua. In particular, the degree-days for the completion of nymphal development were 271, 238, 212, 189 and 153 DD on M. annua, olive, sweet orange, S. alba and P. diffusa, respectively (Table 5).
Interestingly, and in contrast to the results of a previous study reporting that Urtica sp. was the second most preferable host of the insect nymphs from end of February to early April (Kalaitzaki et al., 2013), our results showed that U. urens did not support the completion of C. trivialis nymphal development at 15 and 20 °C. In addition, C. trivialis had a prolonged nymphal development period on P. diffusa at 15 and 27 °C. Likely, C. trivialis development is obstructed on species belonging to the Urticaceae family at these temperatures. A possible explanation could be the lower suitability of plant sap for nymphal development in plant tissues (e.g., low levels of nitrogen) (Awmack & Leather, 2002;Panizzi, 2000). In addition, species of the Urticaceae family produce secondary metabolites that either kill or retard the development of insects (War et al., 2012). It is possible that these weed species serve as hosts for the pest only for a short period, probably covering part of its nutritional needs or providing shelter against natural enemies (Panizzi, 2000).
Our study found that both temperature and host plant significantly influenced the nymphal development of C. trivialis. Plant host species affected the development of nymphs at lower and higher temperatures but not at the optimum for its development temperature. This interacting effect of diet quality with temperature on the insect performance has been observed in other cases in the literature. For example, differences on the length of the developmental period of the mirid Macrolophus pygmaeus (Rambur) (Hemiptera: Miridae) were observed only at low and high temperatures between two eggplant cultivars (Lykouressis et al., 2001). Additionally, Combs and Valerio (1980) have observed differences on the development period of Spodoptera frugiperda (Smith) (Lepidoptera: Noctuidae) on different varieties of Cynodon dactylon L. (Poaceae) at 20 and 25 °C, but not at 30 °C which was proven to be a favorable temperature. According to Quispe-Tarqui et al. (2021) the interaction between temperature and diet observed in the development rate of Copitarsia incommoda Walker (Lepidoptera: Noctuidae) can be attributed to the effect of temperature on food intake rate. In the study of Larsen et al. (1990) johnsongrass appeared to be more suitable than oats or maize for Although the growing period of all tested weed host plants coincides with the occurrence of the C. trivialis (winter until late spring) in the field (Kalaitzaki et al., 2013), the estimated temperature threshold based on thermal requirements for the nymphal development of C. trivialis differed greatly in different host plants. Specifically, it was found to be lowest on M. annua (3.30 °C) and highest on P. diffusa (10.7 o C). The temperature thresholds for the nymphal development on M. annua, olive, sweet orange, and S. alba, were below the winter mean minimum temperature of Chania during the coldest months (January-March) which is approximately 9 °C . Similarly, a threshold of 7 °C was estimated for C. trivialis on R. raphanistrum by Varikou and Birouraki (2014).
The degree-days needed for the completion of nymphal development were estimated to be highest on M. annua (271 DD) and lowest on P. diffusa (153 DD) while the respective DD value recorded by Varikou and Birouraki (2014) on R. raphanistrum was 212.7 DD.
Host plant species and temperature had a significant effect on adulthood longevity. Adult longevity was higher at 15 °C on all tested host plants and was reduced dramatically at higher temperatures. Similarly, adult longevity was highest at 15 o C (45.28 days) on R. raphanistrum and remained high up to 25 °C, while significant reduction was observed at 30 and 32.5 °C (Varikou & Birouraki, 2014). Adults of C. trivialis lived longer when they were reared on sweet orange, M. annua and S. alba in most tested temperatures compared to the other weed species or olive trees, suggesting a good suitability of these plant species as hosts for adults. However, field observations showed that only M. annua support significantly higher adult populations (Kalaitzaki et al., 2013). In U. urens, the longevity of adults was very low even in higher temperatures in which nymphal development was completed successfully. This result is in accordance with previous studies reporting relatively low abundance of adults on Urtica sp. on both olive and citrus orchards in Crete (Kalaitzaki et al., 2013), suggesting that this weed species is not a suitable host for adult feeding. Also, when reared on olive, adults lived less compared to alternative weed hosts such as M. annua or S. alba, in most temperatures.
M. annua was found to be the most suitable host plant for C. trivialis based on the low temperature threshold, short nymphal developmental period, low nymphal mortality and relatively high adult longevity especially at low temperatures typically observed in the field during the period of pest appearance which is 15 days earlier than in the other host plants (Kalaitzaki et al., 2013).
On olive, nymphal survival and adult longevity of C. trivialis was low, clearly indicating that this economically important crop is a less suitable host serving as an occasional source of nutrients and an overwintering site. This is in agreement with field observations reporting extremely low C. trivialis populations from late January to late spring, and limited injury levels caused by the insect on olives early in the growing season and only in cases where the surrounding weed hosts had been destroyed by herbicide applications or had begun to desiccate (Kalaitzaki et al., 2013;Perdikis et al., 2009;Varikou & Birouraki, 2014).
In case of citrus, the low preference of C. trivialis nymphs for this host over most of the weeds tested under field conditions (Kalaitzaki et al., 2013;Varikou & Birouraki, 2014) cannot sufficiently be justified in association with the tested biological parameters (shortest nymphal development period at 15 °C, low mortality at 24 °C and high adult longevity at 20 and 24 °C). This might be due to the non-synchronization of the appearance of the nymphs with the appropriate phenological stage of the sweet orange suitable for their feeding. In natural settings, under typical Mediterranean weather conditions, there is only a small time-window where the appropriate phenological stage of sweet orange (flower buds and flowers) (from March to mid-May) for the nutrition of nymphs coincides with the population peak of C. trivialis nymphs (late January to early April) forcing them to feed on different flowering plant species.
In any case, the relationship between host preference of weed species, present in a crop orchard and the injury levels inflicted to the cultivated plants is complex. Presence of suitable weed hosts in the appropriate growth stage seems to mitigate the loss of flower buds and blossom of olive and citrus since the insect prefers weeds over the crop trees. Furthermore, observations in olive orchards with the non-host plant Oxalis sp. demonstrated extremely low or even the absence of the C. trivialis population both in weed and olive tree canopy (Kalaitzaki unpublished data). In the latter case, insecticide applications are unnecessary and should be avoided.