MQNOTEWPENE CONCENTRATION IN DOUGLAS-FIR IN RELAPlQM TO GEOGWAPHlC LOCATION AND RESISTANCE TO ATTACK BY THE DOUGWS-FIR BEETLE by J. W. Hanover and M. M. Furniss'

The concentration of in Pinus et al. ( 1962 ) showed significant variation in a-pin- monticola Dougl. has been shown to be genetically ene and P- between three populations of controlled ( Hanover, in preparation ) . Genetic con- Pinus radiata in California. Also, Forde and Blight trol of concentration has been implied, ( 1964 ) found that Pinus muricata Don in Calif- also, from analyses of parents or interspecies hy- ornia could be classified into three different groups brids in other species (Bannister et al. 1959; Wil- according to chemical composition. The monoter- liams and Bannister 1962; Smith 1964, and Forde pene cornposition of Pseudotsu-ga has not been 1964 ). Evidence that genes regulate the formation defined adequately although Cvrkal and Janak of has stimulated the use of the ( 1959 ) included Pseudotsuga douglasii Lindl. terpenes as an aid in population studies. Bannister among several they examined.

- We were ~articularlyinterested in measuring At the ti~bethis article was prepared the authors the in oleoresin of Pseudot- weye respectively Geneticist and Entomologist at the monoterpenes Intermountain Forest and Range Experiment statiolL, suga menziesii var. glauca (Beissn.) Franc0 to de- Service, U. S. Department of Agriculture, termine the variation in monoterpenes ( 1) be- Moscow, Idaho. NOW the senior author is Associate (2) Professor, School of Forestry, Yale University, New tween individual , and between groups of Haven, Conn. geographically separated trees. We also wished to determine the sciatisrl between monoterpenes and growth ranged from 1.4 to 41.6 wnm; and crown resistance to attack by the bark beetle, Dendroc- class ranged from suppressed to clouninault, tonus pseudotsuyae Hopk. Finally we studied sea- measurements did not appear to differ significantly sonal variation sf monoterpenes, of , between localities. and the relatiork between rnonoterpenes and sever- al tree cbaraeteristics, Differences in monoterpenes between attacked (resistant) and unattacked trees were tested by statistical analysis. When no significant differences 'llsllgmterials and Methods occurred between groups, both types of trees with- in each area were combined for tests of population Rlonoterpenes were sampled in 1964 at three lo- differences. calities in Idaho and Montana. The three localities ( fig. 1) are sufficiently isolated to permit some Oleoresin samples were collected during the genetic divergence between populations in each latter part of the bark beetle flight period, between area. Oleoresin was sampled from groups of trees June 6 and 19, 1964. Fifteen trees of the Flathead that survived attack by the bark beetles during Lake population were sampled again on October 1962 and also from unattacked trees within Sal- 14, 1964, to determine seasonal variation in their mon River ( A, Valley County ) and Flathead Lake monoterpenes. Samples were collected and ana- (B, Lake County ) locations. At Moscow Mountain lyzed as follows: One day prior to collection of (C, Latah County) only unattacked trees were oleoresin, each tree was chopped lo expose a cross- sampled. The 94 sample trees ranged in age from sectional portion of the , A 38-microliter 81 to 184 years; diameter at breast height ranged sample of oleoresin that exuded from the wood res- from 24.9 to 76.7 cm; the last 10-year radial in ducts was drawn into a calibrated glass capillary tube. The tube of oleoresin was kept in a sealed centrifuge tube under refrigeration until analyzed soon afterwards. Monoterpenes were analyzed qualitatively and quantitatively by gas-liquid chromatography. The oleoresin sample was dissolved in 50 microliters of and a 2-microliter aliquot injected into an F&M Model 5002 gas chromatograph with a flame ionization detector. The chromatograph column was l/4" X 6' stainless steel packed with 10-percent polypropylene glycol on 60-80 mesh Diatoport W- AW. Temperatures were :injection port, 195 C; column, 95 C; detector, 185 C. Helium flow rate was 165 rnl per minute at the injection port. Monoterpenes were tentatively identified by comparing relative retention times of the un- knowns with those of known compounds. Identity was also determined by the addition of known monoterpenes to the plant samples to enhance the corresponding peaks. This procedure utilized both the polar polypropylene glycol column and a non- polar Apiezon-L column packing. Agreement be- tween two such columns is considered sound quali- tative determination. With Apiezon-L, temperatures were: injection port, 195 C; column, 135 C; detec- tor, 195 C. Helium flow rate was 35 ml per minute. concentration was expressed as a percent of total oleoresin by use of standard curves relating peak area (disc integrator values) and monoterpene concentration. Because no curve was available for the unknown monoterpene, this com- ponent of the oleoresin was omitted in determining FIGURE1. - Location of the three geographic areas and groups of trees within areas from which samples total concentrations of the monoterpenes. of Douglas-fir oleoresin were obtained. The areas are: (A) Salmon River in Valley County, Idaho; IB) Flathead Lake in Lake County, Montana; and Use of trade name is for identification and does (C) Moscow Mountain in Latah County, IlaEo. not constitute endorsement by the Forcst Scrviee. Terpenes Detected and Their Ccxlcentraticns 10 0 Seven terpenes were detected in the Douglas- fir wood oleoresin ( fig'. 2 ), Six of these were tentatively identified, in the order of their reten- tion times, as a-pinene, , P-pinene, myr- 80 cene, 3-carene, and , The remaining com- pound, number 7, has not bccn. identified. In the 94 trees of this study, a-pinene was always the pre- dominant rnonoterpene. ft ranged from 10.4 to 52 percent of the oleoresin content, exclusive of the ," 60 unknown. Other rnorioterpenes were present in the C 0 following concentrations : P-pinene, trace to 16.2 B percent; , 0.0 to 4.2 percent; 3-carene, 0.0 oeu to 7.0 percent; and limonene, 0.3 to 8.4 percent. E Because camphene was always present in but min- 6 40 ute amounts, it was excluded from statistical an- Y alyses. 0 0 01 K Muraoterpenes and Resistance to Doughs-Fir Beetle 20 Monoterpenes did not differ qualitatively be- tween trees that survived attack and those that were unattacked (table 1). However, some quan- titative differences occurred, At Flathead Lake, concentration of 3-carene differed signif icantly be- 0 4 8 12 tween trees that resisted attack ( groups 3 and 4) and unattacked trees ( group 5 ), while concentra- Time (rnin) tion of the unknown monoterpene differed between FIGURE2. - Typical chromatogram of monoter~enes groups 3 and 5, These two monoterpenes also in wood oleoresin of F'seudotszi,ga rnenxiesii var. usually were more highly concentrated in unat- glauca dissolved in acetone (see text). tacked trees at Salmon River and Moscow Moun- tain, but the resistance of trees at Moscow Moun- tain is not known. more, probably not all the unattacked trees in this study were susceptible to beetle attack. The beetle The data in table 1 do not directly compare characteristically attacks groups of trees without beetle-resistant and susceptible trees. Several prob- evident selection. Our inventories of these groups lems prevent such a comparison. For example, have shown that each contained some survivors. death after attack was used as proof of suscept- Twenty-two percent of 740 trees attacked in the ibility. Lack of oleoresin exudation pressure, and Flathead Lake area were still alive when examined physiological changes accompanying death made in 1963, 1 year after attack. Thus, any group of resin sampling impossible or undesirable. Also, the unattacked trees selected at random for sampling trees that survived may have been attacked by too is likely to contain some trees that would resist few beetles to kill them; or the unsuccessful attacks beetle attack. Therefore, our sample of unattacked may have disrupted the physiology of the sample trees is probably confounded by the presence of trees and affected their monoterpenes. Further- resistant trees in the sample.

Table 1.--Mean concontratlons of monoterpenes rn oleoresln from trees that resisted attack and unattacked trees

. unltnown Geographic Number : Resis'ted(R) : Percent of oleoresin and standard error for- . (integrator area and ; of or : units and group . trees :unattacked(U): o-pinene 8-pinene Myrcene 1 3-carene : Limonene : Total : standard error)

Salmon River A 1 25 R 28.2b.08 2.1*.3 0.8*.1 0.1~.1 2.4%.2 33.63.1 10.69.5 A 2 10 U 26.7*.43 3.6d.0 0.53.1 0.6ib.4 2.59.4 33.93.2 14.0i6~0

Flathead Lake B 3 11 R 21.79.5 7.13.2 0.9*.1 0.13.1 2.9%.1 32.6i3.8 4.53.9 B 4 15 R 25.19.0 6.9L1.2 0.9%.2 0.49.2 3,2%.4 36.5+1.4 11.5%.6 B 5 22 U 22.15.2 4.720.6 0.4b.7 1.620.5 2.9z0.3 31.63.1 17.02.3 Unfortunately, we know no suitable method ual trees or in studying rrionoterpcnes in relation of testing resistance or susceptibility of Douglas- to physiological trails, it is desirable to restrict fir to bark beetle attack. Until such a method is sampling of o-leoresin~of these species to a uniform available, biochemical or similar comparisons of time sf the year. trees will be inconclusive. Monoterpenes anci Tree &I\lleasurements Seasonal Vaiation Simple correlatioxr cocfficdents of a41 monoter- Concentration of a-pinene and p-pinene in- penes with 10-year radial growth, diameter, and creased significantly between June and October age were computed for 94 tre~sby locality, Qnly ( table 2 ). In some other species, concentration of combined data for all localities are presented in monoterpenes has remained constant between sea- table 3. Highest positive correlations existed be- sons (Smith 1964; Mirov 1961). However, Ban- tween p-pinene and limonene, and 3-carene and nister et al. f 1959 ) found that the ratio of a-pinene the unknown. Within localities, strong positive to ppinene in resin samples from Pinus radiata correlations also existed between P-pinene and varied seasonally ( seasons not specified ) but to a limonene and between myrcene and limonene. lesser degree than between trees. Analyses of Monoterpenes were not strorlgfy correlated with variance between and within seasons showed that tree diameter, age, or growth. the seasonal variation in amounts of a-pinene and Although only 11 trees were involved at Flat- P-pinene in Douglas-fir was also significantly less head Lake ( Group B-3) it is interesting that all of than the between-tree variation. However, the rela- the identified monoterpenes were highly corre- tive concentrations of monoterpenes - especially lated (P < 0.01 ) either positively or negatively tt-pinene - in individual trees did change slightly with one another in the group. This result was between seasons. Concentrations of the three other unique and may reflect a physiological response of identified rnonoterpenes were relatively constant trees that repelled bark beetle attack. Possibly, between June and October, but the error in meas- the so-called resistant trees of group 3 were more uring low concentrations could mask small seasonal resistant to attack than the other "resistant" differences. Monoterpenes in Pinus monticota groups. Greater resistance of group B-3 may also ( Hanover, in preparation ) show a seasonal vari- be indicated by its low content of 3-carene and ation pattern, similar to that of Douglas-fir. There- unknown terpene as compared with other groups fore, in comparing monoterpene content of individ- (table 1).

Table 2.--Seasonal variation in the concentrations of five monoterpenes in individual Douglas-fir trees at Flathead Lake

Percent of oleoresin for- Tree : a-pinene : I?-pinene : Myrcene : 3'carene : Limonene : Total Oct.: June Oct.: Jw kt-: Jue OctL

'aired t-test P < 0.01 P < 0.01 NS NS NS P K0.01

* Not determined Table 3.--Simple correlation coefficients for M~glaS-fi~monoterpenes and tree characteristics (93 d.f.).

Monoterpenes : 8-pinene : Myrcene : 3-Carene : Lirnonene : Unknown : radial : Diameter : Age ---- - : growth : - CI-pinene -0.430+* 0.222+ -O.I$8 -0,315** 0.030 -0.018 0.111 0.197 8-pinene -- - ,1311 - ,001 .657++ ,058 - .005 -.172 -.226 Myrcene -- - .386** .J?8** - ,024 - ,010 - .002 - .049

3-carens -- .046 .557** .lo2 - .242* - .230* Lirnonene -- .I70 .125 - .064 -.245* Unknown -- - .037 - .lo1 -.091 -Total .077 -.034 -.059 * P < "05, r = .202 +* +? < .01, r zz .264

Table 4.--Quantitative variation in Douglas-fir monoterpenes associated with geographic location of trees L/

-- -- Percent of oleoresin and standard error for- : Unknown Geographic - : (integrator area a-plncne : 8-pinene .: Myrcene 1 3-carene I Limonene : : units and - : monoterpenes : standard error)

C-MOSCOWMtn. 21 99.4 kd.8 1 -* 0.5510.17 1.103.61 b1%.59 33.0b.6 13.124.5

- Bracketed values differ from unbracketed values at the 0.05 or lower probability levels.

Differs significantly (P < .05)from Salmon River only.

Oleoresin viscosity varied considerably in sam- than to the Salmon River population. Further spec- ples collected for the analyses. However, "very ulation about their evolutionary patterns, based viscous" and "very fluid" resin did not appear to upon monoterpene data, would require more sam- differ in rnonoterpene composition or content. pling than is reported here. As pointed out by Bannister et al. ( 1962 ), interpretation of statistical analyses of population data assumes random sam- Geographic Dif f erenees pling and homogeneous variances within the sam- pled populations. We lack proof of this. Also, we If monoterpenes in Douglas-fir are controlled assume a lack of environmental influence on the genetically, one could expect an evolution of popu- traits analyzed. Nonetheless, the present study in- lation differences for these compounds. Such dif- dicates that monoterpenes may be useful in more ferences might be detectable between populations intensive study of insect resistance, genetics, and that are widely separated geographically or by physiology of Douglas-f ir. some other barrier to natural crossing. The three populations sampled in this study differed signifi- cantly in certain components of their monoterpene composition (table 4). The Salmon River popula- tion is distinguished from both Flathead Lake and Moscow Mountain populations by having compar- Literature Cited atively higher levels of a-pinene but lower levels of P-pinene, 3-carene and limonene. However, mean Bannister, M. H.; M. V. Rrewerton; and I. R. C. Mc- total concentration of monoterpene was remark- Ilonald. 1959, Vapour-phase chromatography in a ably constant for all localities. study of hybridism in Pinz~s.Svensk Papperstidning 62: 567-573. If these results are representative, they support ; Allette L. Williams; I. R. C. McDonald; and our hypotheses that the Flathead Lake and Moscow Margot B. Forde. 1962. Variation of Mountain populations have diverged but little, or conlposition in five population samples of Pinus that they are more closely related to one another yadiata. New Zeal. J. Sci. 5 (4) : 486-495. Cvrkal, N. and J. Janak. 1959. Anwendung der Gas- Mirov, N. T. 1961. Composition of gum of chromatographie zur Identifizierung einiger Terpene . U. S. Dep. Agr. Tech. Bull. 1239, 158 pp. aus antherischen Olen von Nadelbaumen (- ae). (Application of gas chromatography for identi- Smith, Richard H. 1964. Variations in the monoterpene fying some terpenes from volatile oils of conifers.) composition of ponderosa wood oleoresin. 1[3, S. Coll. Czech. Chem. Commun. 24: 1967-1974. Forest Serv. Res. Paper PSW-15, 17 pp. Pacific Southwest Forest and Range Exp. Sta., Berkeley, Forde, Margot B. 1964. Inheritance of turpentine com- Calif. position in Pinus attenuata X radiata hybrids. New Zeal. J. Bot. 2 (1): 53-59. Williams, Allette L. and M. R. Bannister. 1962, Corn.- and Margaret M. Blight. 1964. Geographical position of gum turpentines from twenty-two species variation in the turpentine of Bishop pine. New Zeal. of pines grown in New Zealand. J. Pharm. Sci. 51: J. Bot. 2(1) : 44-52. 970-975.