ECOLOGY AND POPULATION BIOLOGY Host Selection and Root Colonization of Cyanogenic Stonefruit Species by Capnodis spp. (Coleoptera: )

1 ZVI MENDEL, FABIENNE ASSAEL, AND SHAUL BEN-YEHUDA

Department of Entomology, ARO, The VolcaniCenter, Bet Dagan, 50250, Israel

Ann. Entomol. Soc. Am. 96(2): 127Ð134 (2003) Downloaded from https://academic.oup.com/aesa/article/96/2/127/27983 by guest on 27 September 2021 ABSTRACT The host selection by adults of two ßatheaded rootborers Capnodis tenebrionis L. and C. carbonaria Klug was investigated by examinination of their preference for feeding and oviposition among some of their major host tree taxa. We also studied the resistance to both Capnodis spp. of 10 Prunus rootstocks that were being challenged with the buprestid neonates. The scion selection for feeding and oviposition was related to the cyanide potential in the twig cortex; while potential rootstock resistance was related to the amount of the cyanogenic glycosides in the rootstock cortex. Plum and apricot were preferred by C. tenebrionis and peach was preferred by C. carbonaria. All Prunus rootstocks tested were colonized to some extent by both species. Partial correlation was found between cyanide potential and oviposition for C. carbonaria but not for C. tenebrionis. Nor was signiÞcant correlation found between cyanide potential and host preference for feeding by both species. Peach and plum, which displayed the highest level of cyanide potential, were also the preferred by both species for feeding and oviposition. Apple, with the lowest cyanide potential, was the least preferred for both activities. We found an inverse relationship between resistance to Capnodis and level of cyanogenic compounds in the root cortex. Adults of both Capnodis spp. were not deterred by high levels of cyanogenic compounds in the scion twig cortex. Our Þndings suggest that cyanide potential is not a reliable indicator of the degree of resistance in Prunus spp. rootstocks to Capnodis spp.

KEY WORDS Capnodis tenebrionis, Capnodis carbonaria, Buprestidae Prunus, resistance, host se- lection, rootstock

LARVAE OF THE peach woodborer Capnodis tenebrionis (P. armeniaca L.), cherry (P. vulgaris L.), nectarine L. and the almond woodborer Capnodis carbonaria and peach [P. persica (L.) Batsch], and plum (P. do- Klug destroy the roots of both sapling and mature trees mestica L. and other plum species). Frequent out- of cultivated stonefruits, Prunus spp. C. tenebrionis breaks in Israel during the last decade may have been occurs widely in North Africa, southern and central caused by changes in management practice, including Europe, the Near East and around the Black and the conversion from sprinkler irrigation to drip irrigation Caspian Seas. Damage caused by C. tenebrionis has and the reduction or cessation of irrigation soon after been reported mainly from southern European and fruit picking. Mediterranean areas (e.g., Garrido 1984, Mahhou and The biology of both Capnodis spp. has been thor- Dannis 1992, Tezcan 1995, Ben-Yehuda et al. 2000). oughly studied (Rivnay 1944, 1945, 1946; Christian The distribution of C. carbonaria overlaps much of the 1955; Garrido 1984; Malago´n 1989). Adult be- natural distribution of almond P. amygdalus Batsch come active during the warm months. that ranges from Dalmatia to Asia Minor, the Near As in many species of Buprestidae, young adults East, the Caucasus foothills, and the area between the must feed before mating and ovipositing. Adults of Black and the Caspian Seas. Economic losses due to both species feed on the cortex of twigs and young C. carbonaria have been reported mainly from Israel branches throughout the warm season. Adults may and Egypt (Ben-Yehuda et al. 1997). Outside of the live for 1 yr or more, and a single female may lay cultivated lands, both species are rare and seldom Ͼ1,000 eggs (Rivnay 1944, 1946). Unlike other ßat- found on wild host plants. The species are responsible headed borers that oviposit in cracks or crevices or for the destruction of plantations of almond, apricot glue the eggs to the branch surface (Arnett 1960), eggs of these Capnodis spp. are placed in the ground, usu- ally inserted in cracks of dry soil or positioned under This article reports the results of research only. Mention of a proprietary product does not constitute an endorsement or recom- stones. The neonates penetrate the roots and feed in mendation for its use by USDA. the root cortex. Larvae of many species of ßatheaded 1 E-mail: [email protected]. borers start their feeding period in the cortex and the

0013-8746/03/0127Ð0134$04.00/0 ᭧ 2003 Entomological Society of America 128 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 96, no. 2 surface sapwood and at a later stage mine and consume the resistance to larvae of C. tenebrionis was examined the wood. However, the Capnodis spp. studied here in several studies (DÕhallewin et al. 1990, Malago´n and complete their feeding period on the cortex and bore Garrido 1990, Usai and DÕhallewin 1990, Mulas 1994, into the wood only for pupation. Dicentia et al. 1998). These studies suggested that Information on host selection by ßatheaded wood- resistance to the borer was directly related to the borers is extremely limited. Most buprestid borers concentration of cyanogenic glycosides of the root- seek plants whose resistance has been weakened in stock, but a relationship between cyanogenic glyco- some way (e.g., Haak and Slansky 1987). Agrilus spp., side content and host preference by the adults was not which are among the most destructive ßatheaded bor- studied. ers of forest stands, often cause high tree mortality The four objectives of the current study were as after severe drought (Jacquiot 1976, Mercer 1990). follows: (1) to study host selection for feeding and Members of the genus Melanophila, probably the most oviposition choice among tree species by adults of studied buprestids with respect to host Þnding, use C. tenebrionis and C. carbonaria, (2) to determine the Downloaded from https://academic.oup.com/aesa/article/96/2/127/27983 by guest on 27 September 2021 infrared radiation to locate freshly burnt trees for relative resistance of major stonefruit rootstock taxa to reproduction (Schmitz and Bleckmann 1998). Both root colonization by larvae of both Capnodis spp., (3) Capnodis spp. studied here are often observed en to document the variation of cyanogenic glycoside masse on susceptible trees and avoid vigorous trees, content in scion and rootstock of stonefruit taxa, and which their offspring would fail to colonize. It is (4) to examine the relationship of cyanogenic glyco- widely assumed that adult Capnodis accept or reject side content with host selection and root colonization potential hosts by “sensing” physical signals, such as by these buprestids. higher temperature of weakened trees, or chemical signals such as stress compounds that determine the Materials and Methods acceptability of the host (Rivnay 1946, Christian 1955, Garrido 1984). In agricultural habitats, such as stone- . Past attempts to rear buprestids on artiÞcial fruit plantations, the relationship between host ac- or on semi-artiÞcial media have been unsuccessful ceptability and host suitability for larval development (Haak and Slansky 1987). Hence, the beetles used in is not simple, because adults locate and colonize host the tests were obtained from commercial orchards. trees using cues coming from the scion, i.e., the tree Adult C. tenebrionis were collected in apricot and crown. However, successful development of the prog- plum orchards in the Hula Valley and the southern eny is related to host plant vigor, nutritional quality slopes of eastern Galilee. Adult C. carbonaria were and secondary products in the rootstock. Adult C. collected from peach and apricot plantations in the tenebrionis and C. carbonaria were observed feeding western Negev. Beetles were maintained in ventilated on quite a few tree species, belonging to different glass cages (60 by 40 by 50 cm) at 26Ð27ЊC and fed with families, unsuitable for larvae development, such as fresh 6- to 12-mo-old branches of apricot or peach. avocado (Persea spp., Lauraceae) and casuarina (Ca- Petri dishes (9 cm diameter) Þlled with Þne sifted suarina spp., Casuarinaceae) (Z.M., unpublished sandy soil placed in glass cages were provided as ovi- data). Neither the food preference of the adults, nor position substrate. The petri dishes were changed the relationship between host plant feeding and ovi- daily. The eggs were incubated at 27ЊC, 40Ð50% RH, position has been studied for Capnodis spp. and a photoperiod of 14:10 (L:D) h. One-day-old Management of both Capnodis spp. is problematic neonates were used for artiÞcial root infestations. because no natural enemies of Capnodis are Preference and Oviposition Test. Scion stock from not known and there are no fully effective control Þve taxa were used for preference tests: apricot measures available for either adults or neonates. Fre- (Prunus armeniaca L. ÔCaninoÕ), plum (P. domestica L. quent applications of nonselective insecticides or in- ÔRoyal ZeeÕ), white peach (P. persica (L.) Batsch tensive irrigation are required. Control of the larvae ÔRhodesÕ), bitter almond (P. amygdalus Batsch ÔUm that have penetrated the roots is not practical. Bitter El-FahemÕ) and apple (Pyrus malus L. ÔGranny almond rootstocks have been considered for decades SmithÕ). to be a valuable source of resistance to C. tenebrionis, Eight terminal sections from main branches with and some workers have considered breeding rootstock few side branches, each of 1.5Ð1.8 m long were cut for resistance against Capnodis as an essential man- from trees from the Þve scion taxa (n ϭ 40). The base agement tool (Malago´n and Garrido 1990, Salazar et al. of each branch was placed in a plastic container (40 cm 1991, Mulas 1994). However, compared with other in diameter and 25 cm deep) with 20 liter of water. The pests of stonefruit rootstock, such as nematodes (Pi- lower 15 cm of each branch extended into the con- nochet 1997), little has been done to improve the tainer through a 5-cm-diameter hole drilled in a con- resistance to Capnodis. cave lid. The space between the margin of the hole and Many species of Prunus contain the cyanogenic gly- the stem was sealed with cotton wool and the top of cosides, amygdalin, and prunasin. The former diglu- the lid was covered with 2 liters of sifted soil consisted coside occurs chießy in the kernels (Mizutani et al. of grains of 0.2Ð0.4 mm in diameter as an oviposition 1979). The latter, a monoglucoside, occurs in all tissues substrate, which allowed the separation of the soil and including roots and foliage (Conn and Butler 1969, the eggs (the egg is Ϸ1.2 mm in diameter). In late July Santamour 1998). The relationship between the level 1998, all 40 “trees” (the branch and the water con- of cyanogenic glycosides in the root of Prunus spp. and tainer with the soil) were arranged in a complete March 2003 MENDEL ET AL.: HOST SELECTION AND COLONIZATION BY Capnodis 129

Table 1. Rootstocks tested and their application for fruit hatched larvae was limited (150Ð200 each day by each production Capnodis species), we infested each time one or two Common Rootstock species saplings from each of the tested taxa. The infestation Fruit name and hybrid was carried out over the summer of 1999 and 2000. The number of treated saplings of each rootstock by each Apricot Prunus armeniaca L. Apricot Bear plum P. ursina Ky. Not used Capnodis species ranged between 17 and 30. Nemagaurd P. persica (L.) Batsch ϫ Peach and nectarine Index of Susceptibility. An index of susceptibility of P. davidiana Franch. the rootstocks was determined, for each Capnodis spe- Mariana P. domestica L. Plum Bitter almond P. amygdalus Batsch Almond and plum cies by comparison with the infestation of apricot, the Baladi P. persica Peach most susceptible taxa, after exposure to C. tenebrionis Citation P. domestica ϫ P. persica Plum or C. carbonaria. The index was calculated using equa- Mahaleb P. mahaleb L. Cherry tion 1, where A ϭ mean number of larvae per sapling ϫ 677 P. persica ϫ P. amygdalus Mainly almond Downloaded from https://academic.oup.com/aesa/article/96/2/127/27983 by guest on 27 September 2021 Hansen P. persica ϫ P. amygdalus Mainly peach percentage of infested saplings of apricot (we exposed Hashabi Pyrus malus L. Apple 28 saplings to C. carbonaria and 30 to C. tenebrionis), L ϭ mean number of larvae per sapling, and P ϭ percentage of infested sapling, which applied to both tested Capnodis species. randomized block in a netted greenhouse. Approxi- mately 350 adult C. tenebrionis (1:1 sex ratio), col- Susceptibility index ϭ (L ⅐ P/A) ⅐ 100. [1] lected in early July, were released in the greenhouse, and allowed to feed and to oviposit for 3 d. Every day Determination of Cyanide Potential. The cyanide at 2-h intervals between 0700 and 1900 hours, we potential was determined by the enzymatic cyanide counted the numbers of beetles that were feeding on assay procedure (Lambert et al. 1975) and expressed the trees. In the evening of day 3, we collected the as mmol cyanide/kg of the biomass subjected to the beetles, sifted the soil, and counted the numbers of procedure (dry weight/dry weight plant material). eggs laid near each tree. The experiment was repeated The live bark (cortex) was removed from 1-yr-old in mid-August 1998, with 340 adult C. carbonaria col- twigs of 10 trees of each of the tested scions as well as lected 2 wk earlier. To avoid the use of individuals of the live bark of the roots of 10 saplings of each of the low reproductive and feeding capacity, we discarded tested taxa (bear plum was not tested). The removed the newly emerged adults, old and damaged, from the cortex was weighed, freeze-dried, ground (using a collected material. The approximate age of collected grinder machine) and stored at Ϫ20ЊC. The samples adults was determined by examination of the sharp- were analyzed according to Lambert et al. (1975) and ness of the mandibles and the condition of the tarsi; old Patton et al. (1997). Twenty-Þve milligrams of each beetles have bruised mandibles and tarsi (Ben-Yehuda sample was put into a disposable 10 ml test tube and and Mendel, 2001). extracted twice with 3 ml phosphate buffer (NaH2PO4 Root Infestation. Rootstock from ten taxa of Prunus ⅐ H O 1 M). The extraction solution (2.6 ml) was were tested (Table 1) including nine taxa routinely 2 transferred into the outer well of a Conway diffusion used by stonefruit producers: Hansen 536, Mahaleb, dish (Bel-Art, Pequannock, NJ) and 1.25 ml of almond 677, bitter almond, apricot, Citation, Baladi, Mariana, meal extract was added to the same well. Almond meal and Nemagaurd. A wild species, bear plum P. ursine extract was prepared by mixing 4.06 g sweet almond Ky., native to east Mediterranean, and apple Pyrus meal with 54 ml water, bubbling with nitrogen, and malus ÔHashabiÕ were also included to compare with ␮ the cultivated stonefruit rootstocks. The plants were Þltering through a 0.45- l glass Þber Þlter using an 2-yr-old saplings growing in polyethylene bags (10 aspirator. Two milliliters of 0.1N NaOH were trans- liter) Þlled with red sandy soil and kept under irriga- ferred to the inner well of the dish. The dishes were Њ covered and sealed with silicon grease. After4hof tion in a greenhouse (22Ð31 C) before being infested. Њ Irrigation was stopped 3 d before the introduction of incubation at 35 C, the solution from the inner well the neonates to make the root more accessible to the was transferred to a volumetric ßask, and the volume neonates and ascertain the suitability of the sapling for was restored to 10 ml with 0.1N NaOH. The solution neonate colonization. Neonates (n ϭ 15) of C. car- was kept refrigerated for the colorimetric determina- bonaria or C. tenebrionis (12Ð24 h old) were placed on tion of the cyanide concentration. The efÞciency of the topsoil 10 cm from the base of the stem. Eight the enzymatic reaction was monitored by preparation weeks later the roots of each sapling were washed and of a solution of amygdalin with a known concentration dissected, and the numbers of larvae per sapling and in the same way as the samples. Two milliliters of the percentage of infested saplings were determined. a solution containing Ϸ100 mg/liter Amygdalin and The 8-wk interval was chosen because accurate count- 600 ␮l phosphate buffer were placed in the outer well ing of the larvae was impracticable on a shorter or of a Conway dish. The process was continued in the longer time period. Less than 6 wk after introduction same way as for the samples. The cyanide concentra- the larvae were too small and too transparent to be tion in the 0.1N NaOH solutions was measured ac- recovered. Wilting of heavily infested saplings occur cording to a colorimetric method used for the deter- about 10 wk after the infestation causing high mor- mination of cyanides in water (Anonymous 1995). tality to the larvae. Because production of newly Absorbance was measured at 580 nm. 130 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 96, no. 2

Table 2. Occurrence of adult Capnodis and eggs as related to scion taxa in multiple choice experiments

Scion Mean adult per tree/observation Ϯ SE Mean number of eggs per tree Ϯ SE (“tree”) C. tenebrionis C. carbonaria C. tenebrionis C. carbonaria Apricot 4.08 Ϯ 0.55ab 0.50 Ϯ 0.14b 121.78 Ϯ 36.67a 20.24 Ϯ 5.20ab Peach 3.62 Ϯ 0.53b 1.29 Ϯ 0.36a 144.67 Ϯ 26.70a 27.89 Ϯ 7.85a Plum 4.68 Ϯ 0.77a 0.54 Ϯ 0.09ab 206.11 Ϯ 56.74a 21.45 Ϯ 5.60ab Almond 2.56 Ϯ 0.42c 1.19 Ϯ 0.32a 12.00 Ϯ 2.60b 10.00 Ϯ 3.40b Apple 0.29 Ϯ 0.05d 0.25 Ϯ 0.09c 9.44 Ϯ 4.35b 0.66 Ϯ 0.32c

Means, within the column, followed by a common letter do not differ signiÞcantly at P ϭ 0.05.

Statistical Analysis. The differences among mean Cyanide Potential and Preference Relationships. Downloaded from https://academic.oup.com/aesa/article/96/2/127/27983 by guest on 27 September 2021 number of beetles per tree per observation, mean The endogenous cyanide potential, mmol ⅐ kgϪ1,as number of eggs per tree, mean cyanide potential per measured from the cortex of tested taxa (Table 3) ranged rootstock, and mean larvae per plant for each root- from 37.5 mmol ⅐ kgϪ1 in peach to 0.4 mmol ⅐ kgϪ1 in stock were tested by a parametric one-way analysis of apple. The cyanide potential did not differ signiÞcantly variance (ANOVA). The procedures used were PROC between apricot and almond, which differed from other GLM and PROC MEAN (SAS Institute 1996); the sum tissue sources tested. The correlation of the cyanide of squares was used for computing all F values. Means potential with the number of feeding beetles per tree were transformed into to square root (x ϩ 0.5) to per observation for the tested Capnodis spp. was not stabilize the variances. Differences between means signiÞcant for C. tenebrionis (R2 ϭ 0.43, P ϭ 0.073) and were tested by StudentÐNewmanÐKeuls procedure. for C. carbonaria (R2 ϭ 0.48, P ϭ 0.066). The corre- The percentages were transformed into arcsine. The lation of the cyanide potential and number of eggs laid difference between percentages of sapling infestation near each “tree” for the tested Capnodis spp. was not was subjected to CATMOD procedure (SAS Institute signiÞcant for C. tenebrionis (R2 ϭ 0.18, P ϭ 0.441) but 1996) for maximum likelihood analyses. Differences was signiÞcant for C. carbonaria (R2 ϭ 0.69, P ϭ 0.023). between values for predicted values for response Larval Infestation of Rootstock. Infestation (ex- functions and percentages were tested by chi-square. cluding the apple rootstock) as indicated by larvae per sapling varied signiÞcantly between Capnodis species (F ϭ 3.92; df ϭ 9, 9; P Ͻ 0.05) and plant species (F ϭ ϭ Ͻ Results 12.31; df 9, 9; P 0.01). The mean number of larvae per sapling was 1.21 and 1.05 for C. tenebrionis and Host Preference and Oviposition. The results are C. carbonaria, respectively. The percentages of in- presented in Table 2. Feeding and oviposition activity fested saplings (all taxa combined) were 52.1 and of C. tenebrionis were much more pronounced than 47.3% for C. tenebrionis and C. carbonaria, respec- those of C. carbonaria. The mean “host preference” tively, and there were no signiÞcant differences be- for C. tenebrionis ( per tree per counting) for tween beetle species (␹2 ϭ 3.39, Pr ϭ 0.065). Larval C. tenebrionis as arranged from the highest to the density per plant and percentage of infested plants lowest value per tree was as follows: plum Ͼ apricot Ͼ varied signiÞcantly between rootstock for each tested peach Ͼ almond Ͼ apple. Occurrence on plum was Capnodis species (see below). Among the tested root- signiÞcantly higher than that on the other scions ex- stocks, only infestation density of Mahaleb varied sig- cept apricot. The almond and apple were signiÞcantly niÞcantly between Capnodis species (F ϭ 5.41; df ϭ 1, less attractive than the other scions. The pattern of 38; P Ͻ 0.025). None of the Hashabisaplingswas host preference by C. carbonaria was different from colonized by either Capnodis species. that of its congener: peach Ͼ almond Ͼ plum Ͼ apri- The mean infestation density of C. tenebrionis ar- cot Ͼ apple. Occurrence on peach was signiÞcantly ranged from the highest to the lowest number of higher than on the two latter scions. The mean number larvae per sapling, was as follows: apricot Ͼ bear of eggs of C. tenebrionis per scion, as arranged from the plum Ͼ Nemaguard Ͼ Mariana Ͼ bitter almond Ͼ highest to the lowest value per “tree,” was plum Ͼ peach Ͼ apricot Ͼ almond Ͼ apple. The mean number Table 3. Scion cortex cyanide potential (mean ؎ SD, dry mass of eggs among the plum, peach, apricot and from basis), as determined for five taxa of Rosaceae almond and apple, respectively, did not differ signif- icantly. The mean number of eggs of C. carbonaria per Cyanide potential Scion Ϫ scion as arranged from the highest to the lowest values (mmol ⅐ kg 1) Ͼ Ͼ Ͼ Ͼ per “tree” was peach plum apricot almond Peach 37.54 Ϯ 13.55a apple. The highest number of eggs was recorded on Plum 11.19 Ϯ 6.11b peach differed signiÞcantly from the numbers re- Apricot 7.08 Ϯ 5.41c Ϯ corded for almond and apple (Table 2). A correlation Almond 6.31 3.40c Apple 0.38 Ϯ 0.00d existed between mean number of feeding beetles and 2 ϭ Ͻ mean number of eggs for C. tenebrionis (R 0.86, P Means followed by a common letter do not differ signiÞcantly at P ϭ 0.001) but not for C. carbonaria (R2 ϭ 0.58, P ϭ 0.092). 0.05. March 2003 MENDEL ET AL.: HOST SELECTION AND COLONIZATION BY Capnodis 131

Table 4. Number of larvae per root system (؎SE) and % infested saplings as related to rootstock and Capnodis species

C. tenebrionis C. carbonaria Rootstock Larvae per sapling % infested saplings Larvae per sapling % infested saplings Apricot 2.48 Ϯ 0.45a 88.0a 2.50 Ϯ 0.65a 87.5a Bear plum 1.80 Ϯ 0.90ab 80.0a 1.33 Ϯ 0.30abc 66.5b Nemagaurd 1.62 Ϯ 0.44abc 66.7ab 1.36 Ϯ 0.15ab 81.8a Mariana 1.40 Ϯ 0.25ab 80.0a 1.00 Ϯ 0.25abcd 58.3b Bitter almond 1.11 Ϯ 0.30abc 52.6bc 0.50 Ϯ 0.21cde 35.0bc Baladi0.99 Ϯ 0.28bc 72.0a 1.83 Ϯ 0.41ab 70.8ab Citation 0.96 Ϯ 0.31bc 40.9bc 0.65 Ϯ 0.35bcde 47.1bc Mahaleb 0.77 Ϯ 0.32bc 36.5c 0.06 Ϯ 0.15e 6.0d 677 0.63 Ϯ 0.21bc 40.7bc 0.29 Ϯ 0.20de 21.5cd Hansen 0.44 Ϯ 0.16c 33.3c 0.40 Ϯ 0.42de 26.7cd

Hashabi0 0 0 0 Downloaded from https://academic.oup.com/aesa/article/96/2/127/27983 by guest on 27 September 2021

Means, within the column, followed by a common letter do not differ signiÞcantly at P ϭ 0.05.

Baladi Ͼ Citation Ͼ Mahaleb Ͼ 677 Ͼ Hansen (Table of larvae per sapling, was: Apricot Ͼ Mariana Ͼ bitter 4). Infestation of the apricot rootstock was signiÞ- almond Ͼ bear plum Ͼ Baladi Ͼ Citation Ͼ Nema- cantly different from the Þve latter rootstocks. The gaurd Ͼ Hansen Ͼ Mahaleb Ͼ 677 (Table 4). Infes- percentage of C. tenebrionis infested saplings per root- tation of the apricot rootstock was signiÞcantly dif- stock arranged from the highest to the lowest number ferent from the seven latter rootstocks. 677, the least of larvae per sapling, was as follows: apricot Ͼ bear infested rootstock, did not differ signiÞcantly (for plum Ͼ Mariana Ͼ Baladi Ͼ Nemagaurd Ͼ Bitter both mean infestation density and percentage of in- almond Ͼ Citation Ͼ 677 Ͼ Mahaleb Ͼ Hansen (Table fested sapling) from Mahaleb and Hansen (Table 4). 4). Infestation of the apricot rootstock was signiÞ- Arrangement of the rootstocks according to their in- cantly different from the Þve latter rootstocks. Hansen dex of susceptibility to C. carbonaria is displayed in was the least infested rootstock (mean infestation Table 5. density and percentage of infested sapling) and dif- Cyanide Potential. The endogenous cyanide poten- fered signiÞcantly from apricot, bear plum, Nema- tial, as measured from the tested rootstock taxa (Table gourd, and Mariana (Table 4). Arrangement of the 5), ranged from 89 mmol ⅐ kgϪ1 in apricot to Ͻ 0.01 rootstocks according to their index of susceptibility to mmol ⅐ kgϪ1 in Mahaleb and Hashabi, a level which C. tenebrionis is displayed in Table 5. was too low for accurate measurement by the current The mean infestation density of C. carbonaria as procedure. Cyanide potential in the apricot roots was arranged from the highest to the lowest number of signiÞcantly higher than that of the other tested taxa. larvae per sapling did not differ much form that ob- The cyanide potential of Baladi, Mariana, bitter al- served for C. tenebrionis and was: apricot Ͼ bear mond, Nemagaurd and 677 was more or less the same. plum Ͼ Nemaguard Ͼ Mariana Ͼ bitter almond Ͼ The cyanide potential of Citation was lower than the Baladi Ͼ Citation Ͼ Mahaleb Ͼ 677 Ͼ Hansen (Table latter taxa but did not signiÞcantly differ from 677. The 4). Infestation of the apricot rootstock was signiÞ- cyanide potential of Hansen was lower than Citation cantly different from the Þve latter rootstocks. The but did not signiÞcantly differ from Mahaleb and percentage of C. cabonaria infested saplings per root- Hashabi. A clear linear direct relationship between stock arranged from the highest to the lowest number the index of susceptibility and cyanide potential was determined for both tested Capnodis spp.; R2 ϭ 0.71 (P Ͻ 0.0013) and R2 ϭ 0.77 (P Ͻ 0.0005), for Table 5. Index of susceptibility for two species of Capnodis and C. tenebrionis and C. carbonaria, respectively. root cortex cyanide potential (mean ؎ SD, dry mass basis), as determined for 11 taxa of Rosaceae

Susceptibility index Cyanide Discussion Rootstock potential C. tenebrionis C. carbonaria (mmol ⅐ kgϪ1) The Capnodis species studied differed in their host plant preference for feeding and oviposition. C. tenebrio- Apricot 100.0 100.0 88.98 Ϯ 11.38a Bear plum 66.0 35.6 Ñ nis preferred plum and apricot, whereas C. carbonaria Nemagaurd 59.3 41.6 41.10 Ϯ 20.46b preferred peach. As expected, apple was not suitable for Mariana 46.2 32.4 48.43 Ϯ 15.74b reproduction, and both species infrequently chose it Ϯ Bitter almond 34.0 18.7 47.71 15.40b for feeding, and laid the lowest number of eggs next to Baladi23.9 29.4 48.50 Ϯ 18.40b Citation 17.9 14.0 26.84 Ϯ 7.46bc it. The correlation coefÞcient between mean “host se- Mahaleb 14.4 0.6 Ͼ0.01e lection for feeding” and oviposition was signiÞcant for 677 10.5 0.8 36.93 Ϯ 9.85c C. tenebrionis but not for C. carbonaria, mainly due to Ϯ Hansen 536 6.8 4.9 11.71 7.34d the latterÕs high occurrence on almond. By contrast, the Hashabi0.0 0.0 Ͼ0.01e number of eggs C. carbonaria laid near the almond Means followed by a common letter do not differ signiÞcantly at P ϭ “trees” was relatively small. We cannot rule out the 0.05. possibility that the breeding host tree of the beetles, 132 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 96, no. 2 that had been used in the experiment, affected the resistance. Our results showed a different picture, i.e., pattern of host selection for feeding and oviposition the resistance was inversely proportional to the cya- purposes. We could not detect any gradient effect of nide content with signiÞcant correlation between the the cyanide potential in the scion cortex on host feed- indices of susceptibility of both tested Capnodis spe- ing preference of either species or oviposition of C. cies. Thus, the highest level of this monoglucoside tenebrionis. Egg laying by C. carbonaria coincided with occurred in the apricot rootstock, and the lowest was the high level of cyanide potential in peach, apricot, recorded in Hansen and Mahaleb, taxa displaying the and plum. However, it is interesting to note that peach strongest resistance. and plum that displayed the highest level of cyanide Cyanogenic monoglucosides are a major compo- potential also were most preferred by the Capnodis nent in the resistant mechanism of many plant species species studied. Although both species showed signif- and cultivars against phytophagous insects. However, icant preference for feeding on and oviposition near the content of these allelochemicals alone was not certain scion taxa, they fed on and laid eggs near all enough to fully explain the resistance itself. Study of Downloaded from https://academic.oup.com/aesa/article/96/2/127/27983 by guest on 27 September 2021 tested scions, even apple. ArtiÞcially introduced ne- the sorghum crop varieties resistant to shoot ßy Atheri- onates failed to develop in the apple rootstock. Feed- gona soccata Rond. and stem borer Chilo partellus ing of both the Capnodis spp. on hosts unsuitable for Swinhoe showed that the cyanogenic glucoside ex- reproduction occurred frequently. During outbreaks isted in greater quantities in susceptible cultivars swarms of adults of both species appeared in avocado (Rizvi 1992). Among eight clover cultivars, feeding plantations causing severe injury to the tree crowns. damage caused by larvae of Hypera postica (Gyll.) was However, there is no evidence of beetlesÕ oviposition least for a cultivar with high cyanide content (Ellsbury in those circumstances (Z.M. and S.B.-Y., unpublished et al. 1992). data). Data herein suggest that adults of both C. tenebrionis The index of susceptibility of the Prunus rootstock and C. carbonaria were not deterred by high levels of to both C. tenebrionis and C. carbonaria revealed that cyanide in the scion twig cortex, and resistance of the apricot rootstock is the most susceptible to both spe- rootstocks was not due to the toxicity of prunasin to cies of borers. Strong resistance to the borers has been the larvae. Among phytophagous insects, phloem and displayed by Mahaleb, P. mahaleb, and by two crosses sapwood borers cause the highest potential impact on of P. persica ϫ P. amygdalus, 677 and Hansen. The level tree growth and reproduction (Mattson et al. 1988). of resistance to borers in our results agreed with Death of a stonefruit tree is usually inevitable once Salazar et al. (1991) and Mulas (1994). Our results colonization of the roots by Capnodis larvae has taken indicate that 677 maintains a resistance against the place. A constitutive defense, such as incorporation borers higher than that of bitter almond, and do not into the tissue of cyanogenic glycosides, is aimed at coincide with the Þndings of Mulas (1994), suggesting intoxicating or interfering with the development of that 677 was signiÞcantly more infested than peach the phytophage. Similar to other allelochemicals, cy- and plum rootstocks by C. tenebrionis. The moderate anogenic glycosides may be effective against gener- resistance of bitter almond was unexpected because alist phytophagous insects, whereas these chemicals other workers (Malago´n and Garrido 1990, Salazar et probably encourage the evolution of specialists, be- al. 1991) have reported bitter almond as a highly re- cause they exert a strong selection pressure (Berry- sistant rootstock for C. tenebrionis. Data presented by man 1988). This may explain the results of Patton et al. Mulas (1994) showed that several clones of bitter (1997) showing that feeding intensity of adult Japa- almond, containing a similar amount of cyanide, varied nese beetle, Popillia japonica Newman (feeding on considerably in their resistance to C. tenebrionis. Ͼ275 host plant species), on leaves of 27 taxa of Prunus The resistance pattern was similar for both tested spp. decreased exponentially as endogenous cyanide Capnodis species, except for Mahaleb which was potential increased. However, a signiÞcant positive signiÞcantly more susceptible to C. carbonaria than to correlation was found between prunasin content and C. tenebrionis. One possible explanation may be re- emergence of the peachtree borer Synanthedon exi- lated to the high level of coumarins in P. mahaleb tiosa Say (feeding on Prunus spp.) (Brown et al. 1991). (Fung and Herrebout 1987) affecting C. carbonaria Both tested Capnodis spp. develop solely on members more than C. tenebrionis. Infestation density of Ma- of the genus Prunus, suggesting that accumulation of haleb rootstock saplings by the larvae was signiÞcantly prunasin is not meant to deter them, but instead, acts lower for C. carbonaria than for C. tenebrionis (Table against general phytophagous insects while preform- 4). Bear plum, the sole noncultivated taxon tested was ing other functions as well. Therefore, we assume that highly susceptible to both Capnodis spp., contradict- the level of prunasin content is not a reliable indicator ing our original expectation that, being a wild species, for resistance against these Capnodis spp. it should display high resistance to these borers. We cannot ignore other research reports showing a The endogenous cyanide potential varied substan- positive relationship between the resistance to Cap- tially among the tested rootstock taxa. Although a nodis and prunasin in the roots of stonefruit trees different spectrum of rootstock was tested, both (DÕhallewin et al. 1990, Malago´n and Garrido 1990, Malago´n and Garrido (1990) and Mulas (1994) dem- Mulas 1994). However, we suggest that, in several onstrated that the resistance to C. tenebrionis was clones and cultivars, the level of prunasin may be directly related to the cyanide content of roots and linked to other chemicals or mechanisms that provide suggested that prunasin is probably involved in the the plant with the means to deter colonization of the March 2003 MENDEL ET AL.: HOST SELECTION AND COLONIZATION BY Capnodis 133 root by Capnodis. Thus, for example among several Brown, S. L., C. C. Reilly, J. R. McVay, C. S. Gorsuch, R. cultivars of almond tested by Mulas (1994), few dif- Gentry, and J. A. Payne. 1991. Evaluation of peach cul- fered signiÞcantly in the content of prunasin while tivars as peachtree borer (Lepidoptera: Sesiidae) hosts. displaying the same susceptibility index. A possible Environ. Entomol. 20: 1108Ð1112. explanation may lie in the level of ␤- glycosidase ac- Christian, P. 1955. Le Capnode noir des Rosace´es. Service tivity and/or other enzymes that are responsible for de la De´fense des Ve´ge´taux Originaux No 6. Protectorat the release of small toxic molecules (such as HCN), de la Re´publique Franc¸aise au Maroc, Rabat, Morocco. while the invading larvae cause injuries to the root Conn, E. E., and G. W. Butler. 1969. The biosynthesis of cyanogenic glycosides and other simple nitrogen com- cortex. pounds, pp. 47Ð74. In H. Swain (eds.), Perspective in Both borers pose a continuous threat to stonefruit phyto-chemistry. Academic, San Diego, CA. plantations in warm countries. Integrated manage- D’hallewin, G., M. Malus, and G. Pellizzaro. 1990. Results of ment of these Capnodis spp. that does not rely on two year observation on bitter almond rootstock selec- application of synthetic insecticide is mainly based on tion. EUR 14081: Ame´lioration ge´ne´tique de deux espe`ces Downloaded from https://academic.oup.com/aesa/article/96/2/127/27983 by guest on 27 September 2021 resistant rootstock. The high resistance to both Cap- de fruits secs Me´diterrane´ens. lÕAmandier et le Pistachier nodis spp. of Mahaleb, Hansen, and 677 suggest that 1990: 229Ð232. crosses with bitter almond and possibly with P. ma- Dicentia, F., A. Leo´n, P. Martinez-Go´mez, A. Lacasa, A. Soler, haleb have a potential for breeding resistant rootstock. V. Berenguer, N. Grane, and M. L. Martin. 1998. Possi- Our research points to the lack of information and bilities of breeding almond resistant to Capnode (Cap- confusion concerning the mechanism of resistance of nodis tenebrionis). FAO Nucis Newsl. 7: 11Ð12. Prunus taxa to Capnodis. This knowledge is needed for Ellsbury, M. M., G. A. Pederson, and T. E. Fairbrother. 1992. rootstock breeding and improvement programs of tra- Resistance to foliar-feeding hypergine weevils (Co- ditional selection as well as for genetic engineering leoptera: Curculionidae) in cyanogenic white clover. J. procedures of Prunus taxa expressing new toxins and Econ. Entomol. 85: 2467Ð2472. Prunus resistant mechanisms. Enhanced Capnodis resistance Fung, S., and W. M. Herrebout. 1987. Coumarins in mahaleb and its herbivore, the small ermine moth of Prunus taxa offers a great potential to breed root- Yponomeuta mahalebellus. J. Chem. Ecol. 13: 2041Ð2047. stocks that may express in practice immunity against Garrido, A. 1984. Bioecologõ´adeCapnodis tenebrionis L. the borer neonates using transgenic plants. (Col. Buprestidae) y orientaciones para su control. Bull. San. Veg. Plagas 10: 205Ð221. Acknowledgments Haak, R. A., and F. Slansky. 1987. Nutritional ecology of wood-feeding Coleoptera, Lepidoptera and Hymenop- We express our gratitude to all extension ofÞcers, growers, tera, pp. 449Ð486. In F. Slansky and J. G. Rodriguez (eds.), and technicians for their help in the collection and rearing of Nutritional ecology of insects, mites and spiders. Wiley, the beetles. We thank AmiZehaviand NitzaShaphir(KKL New York. Forests Department) for help in obtaining the bear plum Jacquiot, C. 1976. Tumors caused by Agrilus biguttatus Fab. saplings, Cynthia A. (Patton) Pierce (NCSU) for the Conway attacks on the stem of oak trees. Marcellia 39: 61Ð67. diffusion dishes, Thomas G. Ranney (NCSU) for the valuable Lambert, J. L., J. Ramasamy, and J. V. Paukstiles. 1975. Sta- comments on an early version of the manuscript, and Ruth ble reagents for the colorimetric determination of cya- Marcus and Miri Zarhi (ARO) for assistance in the statistical nide by modiÞed Ko¨nig reactions. Ann. Chem. 47: 916Ð analysis of the results. The Fruit Marketing Board and the 918. Chief Scientist Research Fund Þnanced the study as project Mahhou, A., and F. G. Dannis. 1992. The almond in Mo- No. 0131-0637. rocco. HortTechnology 2: 488Ð492. Malago´n, J. 1989. Bioecologõ´adeCapnodis tenebrionis L. References Cited (Col. Buprestidae) e inßuencia de ciertos factores abõ´- oticos sobre sus estados inmaduros, en el momento de la Anonymous. 1995. Standard methods for the determination eclosio´n del huevo y su penetracio´n en huespedes de of water and wastewater, 19th ed. pp. 4Ð24. Water Pol- interes agrõ´cola. Ph.D. dissertation, Polytechnic Univer- lution Control Federation, Washington, DC. sity of Valencia, Spain. Arnett, H. R. 1960. The beetles of the United States. Cath- Malago´n, J., and A. Garrido. 1990. Relacio´n entre el con- olic University American Press, Washington, DC. tenido de glicosidos cianogenicos y la resistencia a Cap- Ben-Yehuda, S., and Z. Mendel. 2001. A method to deter- nodis tenebrionis L. en frutales de hueso. Bol. San. Veg. mine the adult age of two species of Capnodis (Co- Plagas. 16: 499Ð503. leoptera: Buprestidae) and its implication in determining Mattson, W. M., R. K. Lawrence, R. A. Haak, D. A. Herns, the seasonal population age. Phytoparasitica 29: 85Ð86. and P. J. Charles. 1988. Defensive strategies of woody Ben-Yehuda, S., F. Assael, and Z. Mendel. 1997. Recent out- breaks of phloem- and wood-boring insects in deciduous plants against different -feeding guilds in relation to orchards in Israel. Phytoparasitica 25: 163Ð164. plant ecological strategies and intimacy of association Ben-Yehuda, S., F. Assael, and Z. Mendel. 2000. Improved with insects, pp. 3Ð38. In W. M. Mattson, J. Levieux and chemical control of Capnodis tenebrionis L. and C. car- C. Bernard-Dagan (eds.), Mechanism of woody plants bonaria Klug (Coleoptera: Buprestidae) in stonefruit defenses against insects: research for pattern. Springer, plantations in Israel. Phytoparasitica 28: 27Ð41. New-York. Berryman, A. A. 1988. Toward a uniÞed theory of plant Mercer, C.W.L. 1990. Prospects for integrated pest man- defense, pp. 39Ð55. In W. M. Mattson, J. Levieux, and agement in forestry in Papua New Guinea. Brighton Crop C. Bernard-Dagan (eds.), Mechanism of woody plants Prot. Conf. Pest Dis. 1: 385Ð390. defenses against insects: research for pattern. Springer, Mulas, M. 1994. Almond genetic resources and resistance to New York. Capnodis tenebrionis. Acta Hortic. 373: 41Ð48. 134 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 96, no. 2

Mizutani, F., M. Yamada, A. Sugiura, and T. Tomana. 1979. aspects. Miscellaneous publication. USDA-ARS, Gaines- The distribution of prunasin and amygdalin in Prunus ville, FL. species. Mem. Coll. Agric. Kyoto Univ. 113: 53Ð65. Salazar, D. M., M. Miro, and S. Garcia. 1991. Rootstocks for Patton, C. A., T. G. Ranney, and J. B. Burton. 1997. Natural dry region apricot tree faced with Capnodis tenebrionis L. pest resistance of Prunus taxa to feeding by adult Japanese Acta Hortic. 293: 401Ð404. beetles: role of endogenous allelochemicals in host resis- Santamour, F. S. 1998. Amygdalin in Prunus leaves. Phyto- tance. J. Am. Hortic. Sci. 122: 668Ð672. chemistry 47: 1537Ð1538. Pinochet, J. 1997. Breeding and selection for resistance to SAS Institute. 1996. SAS/STAT userÕs guide, version 6.11. root-knot and lesion nematodes in Prunus rootstocks SAS Institute, Cary, NC. adapted to Mediterranean conditions. Phytoparasitica 25: Schmitz, H., and H. Bleckmann. 1998. The photomechanic 271Ð274. infrared receptor for the detection of forest Þres in the Rivnay, E. 1944. Physiological and ecological studies on the beetle Melanophila acuminata (Coleoptera:Buprestidae). species of Capnodis, in Palestine (Col., Buprestidae): I. J. Comp. Physiol. 182: 647Ð657. Studies on the eggs. Bull. Entomol. Res. 35: 235Ð242. Tezcan, S. 1995. Investigations on the harmful species of Downloaded from https://academic.oup.com/aesa/article/96/2/127/27983 by guest on 27 September 2021 Rivnay, E. 1945. Physiological and ecological studies on the Buprestidae (Coleoptera) of cherry orchards in Kemal- species of Capnodis, in Palestine (Col., Buprestidae): II. pasa (Izmir) district (Turkey). Turk. Entomol. Derg. 19: Studies on the larvae. Bull. Entomol. Res. 36: 103Ð119. 221Ð230. Rivnay, E. 1946. Ecological and physiological studies on Usai, M., and G. D’hallewin. 1990. Cyanogenic glucosides Capnodis spp. (Col., Buprestidae) in Palestine: III. Studies contained in different organs of bitter and sweet almond. on the adult. Bull. Entomol. Res 37: 273Ð280. EUR 14081: Ame´lloration ge´ne´tique de deux espe`ces de Rizvi, V. 1992. Biochemical selection of sorghum crop va- fruits secs me´diterrane´ens. lÕAmandier et le Pistachier rieties resistant to sorghum shoot ßy (Atherigona soccata) 1990: 233Ð236. and stem borer (Chilo partellus): role of allelochemicals, pp. 101Ð117. In H. Alborn, G. Stenhagen, K. Leuschner, Received for publication 19 October 2001; accepted 4 April and S.J.H. Rizvi (eds.), Allelopathy: basic and applied 2002.