u l mu; MIMI)!!! gm ll my: mg l ”j; «11 I! This is to certify that the thesis entitled Genetic Variation of Scotch Pine Oleoresin Physiology presented by Michael Raymond Bridgen has been accepted towards fulfillment of the requirements for Ph.D. Forestry degree in l I. /, I .’ .I Major professor June 13, 1979 Date 0-7639 K 22: MW fiMWWv ' LIBRARY Michigan Sate ~ ‘Univcnity OVERDUE FINES ARE 25¢ PER DAY ‘ PER mm Return to book drop to remove this checkout from your record. 1 GENETIC VARIATION OF SCOTCH PINE OLEORESIN PHYSIOLOGY By Michael Raymond Bridgen A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1979 ABSTRACT GENETIC VARIATION OF SCOTCH PINE OLEORESIN PHYSIOLOGY By Michael Raymond Bridgen A rangewide provenance plantation Of Scotch pine was used to study the genetic variation in chemical, physiological and anatomical traits associated with the Oleoresin physiology of the species. Fifty seedlots were examined. The resin acids showed strong positive correlations with each other. Nine of thirteen resin acids and total resin acids showed genetic variation, with seed sources from Scandanavia having the highest concentrations and those from central Europe the lowest. Russian and Siberian varieties had the highest water potential readings (least nega- tive) while the Scandanavian varieties had the weakest Oleoresin pressure and the smallest resin canal cross-sectional areas. No variation was observed in the density of resin canals. The resin acids as a group show evidence of pleiotropy, but each of the physical traits was indepen- dent of the others, there being no correlation between them. The within-tree variation of monoterpenes in the cortical Oleoresin of a single Swedish seedlot Of Scotch pine, certain traits affecting Oleoresin production, and the seasonal variation Of the resin acids in the Oleoresin were observed. The concentrations Of four major resin acids, two minor acids and total resin acids reached a minimum during the May 15 to July l period. Strobic + dehydroabietic acids reached a maximum of 2.3 percent Oleoresin at this time. The monoterpenes did not Michael Raymond Bridgen vary between tissue Of the same age Of different branches although a-pinene tended to increase with the height Of the tree. Oleoresin pres— sure was the only physical trait Observed which varied in the crown, de- creasing from the lower branches to the top of the tree. Possible effects of these variations are discussed. Progenies from twenty—seven seedlots of Scotch pine, planted at two locations in central Michigan, were studied to identify genotype X envir- onment interaction and effects Of planting site on the chemical and physical properties of the Oleoresin system of the species. Thirteen resin acids, water potential, resin canal density and cross sectional areas were examined._ The concentration of resin acids showed no consis- tent variation due to planting site or seedlot, except for Unknown 4l which showed varietal differences. Most of the major resin acids had a high percent of variation due to seed source X plantation interaction. Fifty six percent Of the variation of total resin acids was due to the interaction. Water potential, resin canal area and density also showed high genotype X environment interactions. Plantation differences were Observed in the cross-sectional area Of resin canals. The possible relationships of resistance in Scotch pine to five in- sect pests and the resin acids Of the cortical Oleoresin, water poten- tial, Oleoresin pressure, resin canal area and resin canal density were examined. Simple correlations and multiple regression analyses were used to identify the traits most related to resistance. Models were de- veloped for the European pine sawfly, the pine shoot borer, the white- pine weevil, the Zimmerman pine moth, and the pine root-collar weevil. Only the resin acids showed any association with insect resistance. Re- gression models were produced to predict the expected percentage Of trees Michael Raymond Bridgen attacked by each insect at various levels Of probability. Although these models may work well in a localized area, the presence of genotype X environment interaction in the resin acids limits their usefulness. Recommendations for future work are given. ACKNOWLEDGMENTS The author wishes to recognize the guidance and support given him by Dr. James N. Hanover, who served as chairman and chief mentor, Dr. Jonathan M. Wright for his advice and experience, and Drs. D. I. Dickmann, D. Penner and G. A. Simmons who completed his advisory committee. He is also indebted to the members Of his family for their encouragement throughout his studies. 11 TABLE OF CONTENTS Page LIST OF TABLES ........................... v LIST OF FIGURES ........................... vii INTRODUCTION ............................ l Chapter l. Genetic Variation in Oleoresin Physiology of Pinus sylvestris 3 Abstract .............. _ ..... . ..... 3 Introduction ...................... 4 Methods and Materials .................. 6 Results and Discussion ................. l2 2. Within-tree and Seasonal Variation of Oleoresin Physiology in Scotch Pine . ........... , ............. 2l Abstract .. .................. . . . . . 2l Introduction . ..................... 22. Methods and Materials .......... . ........ 23 Results .......... . ............... 26 Discussion ............... . . . ...... 31 3. Genotype X Environment Interaction Of Oleoresin Physiology of Pinus sylvestris L ............ . ......... 34 Abstract ........................ 34 Introduction ...................... 35 Methods and Materials ............... . . . 37 Results ......................... 39 Discussion ....................... 48 4. An Introduction to the Use Of Indirect Selection for Pest Resistance in Conifers .................... 52 Indirect Selection ................... 52 Indirect Selection of Pest Resistance in Conifers . . . . 57 5. Application Of Oleoresin Physiology of Scotch Pine in an Indirect Selection Program for Insect Resistance ....... 59 Abstract ........................ 59 Introduction ...... a ................ 60 Methods and Materials .................. 62 Results ......................... 63 Discussion ....................... 66 Recommendations for Future Work ............. 71 iii Page SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..74 BIBLIOGRAPHY .............. I . . . . . . ........ 7s TABLE LIST OF TABLES Page Variation Of means Of major and total resin acids between regions and varieties Of Scotch pine . . . . ......... 13 Variation Of means Of minor resin acids between regions and varieties of Scotch pine ................... 14 Variation Of means Of various Oleoresin traits between regions and varieties of Scotch pine (Rose Lake) . . . ........ l5 Simple correlations between resin acids of Scotch pine . . . . 18 Seasonal variation Of the resin acid components of Oleo— resin from Scotch pine seedlot 222 (Sweden) during the _ growing season Of 1978 .................... 27 10. ll. 12. Mean values of OleOresin monoterpene concentrations at three heights within the crown of Scotch pine . . .. ........ 30 Mean values Of four physical traits of Oleoresin physiology as they vary within the crown Of Scotch pine .. ........ 32 Degrees of freedom and variance components Of a seedlot and environment effects study .................. 40 The means of resin acids of regions and varieties Of Scotch pine grown at two locations in Michigan . .......... 42 Components Of variance, as a percent of total variance, for resin acid composition of 25 seed sources of Scotch pine planted at two locations . . .. ................ 45 Degrees of freedom, variance components, and the percent Of total variance in each source of variation for water poten- tial and resin canal density and area measured on 27 seed sources of Scotch pine planted at twO.locations in Michigan . 49 Variation Of several physiological and anatomical traits between varieties Of Scotch pine grown at two locations in Michigan . . . . .. ..... . . . ... . . . . . . ...... 50 TABLE 13. l4. l5. Page Simple correlations of chemical, physical and anatomical traits of Oleoresin physiology with susceptibility patterns of several Scotch pine insect pests .............. 65 Multiple regression models Of resistance to five insect species 67 Results of multiple regression analysis models giving expected percent Of trees attacked in two varieties for five insect pests ...................... _ ....... 69 vi FIGURE 1. LIST OF FIGURES Page Distribution of European seedlots and varieties of Scotch pine used in the study ....... . .............. 7 Distribution of Asian seedlots and varieties Of Scotch pine used in the study ...... , ................ 9 Distribution of resin canal density and resin canal cross- sectional areas of various seedlots and varieties of Scotch pine . . .............. . ............. l9 Seasonal variation of the total resin acids in cortical Oleoresin of‘a Swedish source Of Scotch pine and the rates of elongation of lateral branches . . . ............ 28 Genotype X environment interactions of neoabietic acid in Scotch pine ......................... 46 The theoretical basis for direct selection ...... . . . . 53 vii INTRODUCTION Perhaps the greatest problem in the study Of forest genetics for practical tree improvement is the large amount of time required to get useable results. Two essential facets of any plant breeding program, the initiation of flowering and the selection of progeny based upon traits of mature plants, take many years and may discourage plant breeders from working with forest tree species. Some conifers, including Pinus strobus, P. taeda, and P. contorta may flower as early as 5 to l5 years of age, but maximum cone production does not occur until thirty to fifty years of age and may be as high as 200 in other species, such as Picea engelmannii (Fowells, I965). The geneticist, who is usually interested in traits of full grown, mature trees, is often forced to use older plant- ings whose history and records may not be complete or leave unfinished studies for future workers. These and other problems have led tree breeders to examine indirect selection as an aid in early evaluation. An example of this method is to use traits of seedlings or immature saplings which are strongly correlated with traits of mature trees, to make selections at a much earlier age, thus shortening the breeding pro- cess. Difficult-tO-measure traits may also be selected indirectly by choosing individuals with characteristics which are correlated with the desired trait. Indirect selection can be especially effective if the measured traits are physiological rather than morphological. 2 Resistance to insect and disease pests is a trait that is considered desirable for breeding but which is difficult to measure as it must per- sist throughout the life of a tree. A selected seedling which has re- sisted attack cannot be assumed to be resistant throughout its life. Changes in the physiology and environment of a plant continually affect its ability to resist or tolerate pests. Selections for resistance or tolerance to insects or diseases must, therefore, be made on physiol- ogically mature trees, to be most effective. Oleoresin physiology is a system peculiar to most conifers and esg pecially the pines, spruces, larches, and Douglas fir, which possess normal resin canals. Various factors associated with Oleoresin, includ- ing the chemical composition, anatomy of the resin ducts, and pressure within the canals, may have an influence on the growth and developnent of conifers. Many components of this system have been associated with insect resistance in tree species (Hanover, l975). These components may serve as useful marker traits in an indirect selection program for de- veloping insect resistance. The objectives of this study were (l) to examine the genetic varia- tion of several components of Oleoresin physiology in Scotch pine (Pinus sylvestris L.), (2) determine the effect Of environment on these components, and (3) to compare each component with Observed resistance patterns of several species of insects which attack the species. These factors may be used to practice indirect selection of seedlings for genetic resistance or tolerance to these insects. Chapter l Genetic Variation in Oleoresin Physiology Offfinus sylvestris Abstract A range wide provenance plantation of Scotch pine was used to study the genetic variation in chemical, physiological and anatomical traits associated with the Oleoresin physiology of the species. Fifty seedlots were examined. The resin acids showed strong positive correlations with each other. Nine of thirteen resin acids and total resin acids showed genetic variation, with seed sources from Scandanavia having the highest concentrations and those from central Europe the lowest. Russian and Siberian varieties had the highest water potential readings (least nega- tive) while the Scandanavian varieties had the weakest Oleoresin pressure and the smallest resin canal cross-sectional areas. No variation was ob- served in the density of resin canals. The resin acids as a group show evidence Of pleiotropy, but each of the physical traits was independent of the others, there being no correlation between them. 4 Introduction Scotch pine-(Pinus syZvestri3.L.) is one of the most studied tree species in the world. It has an extenSive natural range in Europe and Asia and can be planted as an exotic widely in other parts of the world. Grown as a timber tree in Europe, it is also used as an ornamental, Christmas tree and rarely as a pulp species in the United States. Some of the physical traits which exhibit genetic variation among seedlots and varieties Of Scotch pine include winter foliage color, needle length, growth rate, flower production and winter injury. The foliage Of Scandanavian varieties changes color from dark green to yellow or gold in the winter. This is caused mainly by a decrease in the chlor- ophyll concentration of these seedlots (GerhOld, 1959). Genetic varia- tion Of the other traits has been described by Wright (l963) and Wright et al (l966). The relative height growth Of seedlots is usually consist- ant between plantations, but some genotype X environment interaCtion is seen (King, l965 a). Sources from central Europe grow faster in the north central states than those from Scandanavia, Siberia or Western and South- ern Europe sources. The production Of flowers correlates positively with height growth, the tallest varieties producing the most cones and the shortest varieties producing the least. There is sOme variation between seedlots within varieties. The Spanish variety, iberica, suffers the most winter damage when planted in Michigan. Central European seedlots, variety haguensis, have the longest needles while seedlots from Spain have the shortest. (Again, a significant interaction for needle length occurs between seedlot and site Of plantation (King, l965 b). Squillace et aZ (l975) have reported the existence Of genetic varia- tion in Scotch pine for resistance to needle cast, Lophodermium pinastri 5 (Schrad. ex Hook.) Chev. Stands from northwestern Europe, especially Holland, had the least needle castdamage. Families accoUnted for 58% Of the total variation in resistance." Genetic variation in resistance to several insect pests has also been reported. The eastern pineshoot borer (Eucosma gloriola Heinrich) causes ‘ greatest damage on varieties from south and western Europe (Steiner, 1974). These same varieties are the least susceptible to the Zimmerman pine moth (Dioryctrt'a zimmemani Giote)(Wright at a1, 1975). Other insects which attack Scotch pine and are affected by its genetic variation are the European pine sawfly (Neodiprion sertifer Geoff.)(Wright at al, 1967) and the pine rOOt collar weevil (Hylobius radicus BuCh)(Wright and Wilson, 1972). The chemical composition of the Oleoresin of Scotch pine is also genetically variable. Ten Of eleven monoterpenes show signifiCant dif— ferences between varieties, but 3-carene and a-pinene are the most useful for distinguishing varieties (TObOlski and Hanover, 1971). This consist- ent variation has been used to identify the variety of an unknown seed source (Bridgen, Hanover, and Noecker, 1979). The diterpenoid resin acids also comprise a portion of the Oleoresin of pines. A previous attempt to identify genetic or geographic variation of the resin acids in a pine species was done by Bridgen, Hanover, and Wilkinson (1979) on eastern white pine (Pinus strobus L.). Both abietic and neoabietic acids showed differences between seed sources, but the variation was randomly distributed. Oleoresin viscosity also showed no variation among seed sources. 0n the assumption that SCOtch pine is a more variable species than eastern white pine, this study was done to determine if genetic variability of the resin acids exists in the species. 6 Several physical traits associated with Oleoresin physiology were also examinedfor genetic variation. Methods and Materials A seventeen-year-Old range-wide provenance plantation growing in the Rose Lake Wildlife Research Area in Shiawassee County, Michigan, was used for the study. The plantation was established in 1961 as part of the NC-99 study Of Scotch pine. Nursery procedures for the collection of seed and planting have been described by Wright and Bull (1963). Fifty seedlots, originating from eighteen varieties across the natural range of Scotch pine, were used throughout the study. The distribution Of seed- lots and varieties used in this experiment is shown in Figures 1 and 2. Between June 27 and July 1, 1978, Oleoresin was collected from the bole of each tree, 1.5 meters from the ground. A 1.6 mm wide, 6.4 mm deep hole was drilled into the bole, and a twenty microliter glass capillary tube was tapped into the hole with a plastic hammer. Silicon rubber around the tube secured it to the tree. The capillary tube was collected the next day and stored at 2 C in nitrogen until analyzed. Two replica- tions of each seedlot were collected, with three trees from each repli- cate. The Oleoresin from all trees in the rep was bulked to shorten the time on the gas chromatograph. The method Of chemical analysis has been described by Bridgen, Hanover, and Wilkinson (1979). Water potential was measured on each of three trees in one repli- cation for fifty seedlots. A 12.7 cm cutting was taken from a lateral branch at 4.5 meters on the south side Of the crown. A 2.5 cm wide strip of bark and phloem was stripped from the base Of the cutting. The branch was inserted into a pressure bomb (Scholander at al, 1965). The Figure 1. Distribution of European seedlots and varieties of Scotch pine used in the study mm 7 I'll 3 mm C n_m2ua<¢3 on u]. 4: /\,\., o I: 3 e b o b /. dd_ (flashing? .25 o O \ <2¢==G<, .ll 1 F\ r11; 2.1% /il /> 52 \..\w . <.U.ZO¢O.— I. so“ > G 2: 0:534» (0.20....(a 00A. o ooufizuaos. .w. ¢_u02< (0:00.» 3 (Damn. me am mm Figure 2. Distribution of Asian seedlots and varieties of Scotch pine used in the study lO cwu c- 99— am an eh a (K .M 41x. (UT-00202 o /jf, __ i XV _g.\a c0 / S m-m2uu<83 00. C / 8 \QB, My / 2 11 pressure in pounds/inch2 at the time that water exuded from the cutting was recorded. Pressure was converted to bars by multiplying by e0.0689 as a conversion factor. All trees were measured between September 6 and September 8, 1978 and remeasured between September 25 and 27, 1978. Measurements were made betweenllzoo AM and 4:00 PM. Oleoresin pressure Was measured in micromanometers. Twenty micro- liter capillary tubes were cut at the calibration mark and fire sealed on the cut end. The Open end was filed_to taper the tube. A 1.6 mm wide hole was drilled into the bole Of each tree at 1.5 meters from the ground. The micromanometers were inserted to a depth Of 6.4mm and secured with silicon rubber. In the morning of the day following insertion (August 31, 1978), the lengths of the air pockets inside the tubes were measured. Oleoresin pressure, measured in atmospheres, was calculatedby dividing the length of the manometer by the length Of the air pocket. Three trees in each Of two replications per seedlot were measured. Branch cuttings were made on each Of three trees in one replication Of each seedlot at_4.5 meters in the crown. A five centimeter piece of stem cutting, at least two years Old, was taken from each branch, labelled, and placed in test tubes, covered with 70% ethanol. After two weeks, the ethanol was drained and the wood samples were air dried. This slices Of the stem were made with a razor blade and stained with two solutions. The first was 0.5 g Safranin 0 in 200 m1 Of water and the second was 1.5 9 K1 + .5 g I in 100 m1 Of water. The minimum and maximum axes of three resin canals in the previous year's growth, the total number Of canals in the ring, and the diameter Of the stem were recorded. Resin canal area was calculated by A = nab where a and b are the lengths Of the semi- axes. Total number of resin canals divided by the circumference Of the 12 stem was used as an estimate of resin canal density. Results and Discussion As existing varietal classifications of Scotch pine sources are good indicators of genetically distinct populations (Wright, 1976), the data from these studies have been listed by variety and groups of varieties (geographic regions) in Tables 1, 2, and 3. The seedlots are subgroups of varieties which are subgroups of the region, and an analysis of var- iance was done using a simple nested design. Tables 4 and 5 show how the concentrations of each resin acid varied between varieties and regions. Total resin acids ranged in percent Of Oleoresin concentration from 8.5 in variety 'East Anglia' to 54.0 in Zapponica. The low concentrations of some varieties may have been due to the time of sample collection, as the resin acid concentration shows seasonal fluctuation with a sharp decrease between May 15 and July 1 (Chapter 2) and varietal differences in magnitude of fluctuation may exist. The Scandanavian varieties had the highest concentrations of total resin acids. The central European varieties had the lowest concentrations of pimaric, abietic, neoabietic, and total resin acids, but the highest of Unknown 28. Tobolski (1968) has reported that the monoterpene fraction of Scotch pine Oleoresin comprises 27 to 30 percent of total Oleoresin. The sum of monoterpenes and resin acids account for 35 to 84 percent, leaving a large amount of Oleoresin unidentified. Genetic variation was also observed in several physical traits (Table 3). The highest water potentials (least negative) were found in the Russian and Siberian varieties. Only the data collected on Septen- ber 6 to 8 are shown, as no relative differences due to time of 13 .umoa macem mpaeopaz m.:eocao )3 am we peacomceu xpaceoee_cmem we: one can—co msem on» ensue: mzpe> cosuoce we Legacy mEem one xn umzop—ow mm=_e> pepmwce> co pe:o_mom eon eN.eN eeone ee.e ne ee.N eon no._ eo oe.~ eeon Ne.ep on ee.N eeeeeneee ee eN.ee to em.m o ee.e eo NP._ no on.. me eN.nN on NN.N .e_nen_m .N .z. eeon n_._m eeeo ee.e ne em.N one ee.o ee Ne._ eeeo no.en e _e.e eeeNeeees N> ee.om N» .e.e N ee.N N No._ N ee.. N eo.N. N en.N eenoenee> eeeoaem wee :ewmmza eon ee.eN eeeon MN.e ne Fe.N one me.o ee me._ eeo Fe.en on ee.N eeNeeeeeneoemoe e mo.en e e_.e o ee.e e ee._ e ee.N e Ne.Ne e oe.e eooeeaaen N ee.Ne N Ne.e N ee.N N» .e.o N oN._ N ON.e_ N ee.N eenoenee> :e_>eceuceom one Ne.N_ one ee.N e me._ one me.o ne NN.o one none one Ne._ eoeeoeeee one ee.e_ one ee.N ne _O.N ne me.o ne Ne.o ne m_.N ne me._ eooeeoeoe eone eo.o~ one me.N e on._ e ee.o ne oe.o eeon .N.N_ one ee._ sooneaeeo one _o.e_ ne ne.N e Ne._ e Nm.o on No.P eone en.o one ee.n eeeeenea eon _e.eN ne eN.N ne en.N ne ee.o ee ee.. tee ON.en one oe._ eooeeeeea x Ne.e_ x ee.N x Ne._ x ON.o > Ne.o > NN.o x Ne._ meeoenees oqoczm Fecocmo eeo eN._e eee ee.e on Ne.e one ee.o ee em._ eeeo nN.eP o NN.N eooeeooonee ne Ne.e_ e on.N e NN._ e Nn.o ne eN.o one eo.e ne Ne.n sooneoe eon ne.nN eone Ne.N ne n_.N e en.e ne eN.o eeon Ne.e_ on eN.N eeeeoeeee one n_.en ne Ne.N e on._ e mm.o no em._ eeone e..__ ne No.2 senesNNe eon ee.NN eone ee.e ne en.N one eN.n e en.o eeeo _e.e_ on NF.N eeeeene e ne.e e oe._ e ne.n one ee.o ne ee.o e Ne._ e eN.o .e__eee .N. ee oe.ee eeeo em.e on eN.e eone ne.e e N_.N to eN.e_ on em.N eeeeoeee one mo.e_ e mP.N e eN._ one nN.o ne Ne.o one No.e one ee._ onesee > ee.eN .>x nN.e > eN.N >x eN.o > en.~ N no.en > No.N menoenee> onenem ccmzpaom wee ccmumoz . I - n u i u u u n n u u . :Pmmcoopo x i u u u u u u u u u u l i u 1 n u . mc_oe =_mm¢ mueuowaeomz mpepmwn< mueces_a mueceseaomH oueepmzpea + mpecee_a .eoen -ooecencem mpeeeeeqo>on .oc_g souoom co mm_umece> use m:o_moc cmozumn mange :_mmc _euou use coves co mceme co :o_ue_ce> ._ m_nep l4 .pmop oooem anonpnoz m.:eo::0 no nn we “coconnno anuceonn -nconm no: one cannoo osem on» onzunz o:_e> cocooce me Lennon oEem on» an ooze—non moone> neoonee> Lo neoonoom ne no.0 no.0 nn.0 ne no.0 No.0 no.0 ;-o on.o enmemNoes e 00.0 o_.0 00._ no on.o N_.0 v_.o con on.o .encoonn .n .z. e 00.0 00.0 no.0 oe no.0 on.0 00.0 n.e no.0 eonNooooe n 00.0 no.0 Nn.0 > No.0 no.0 no.0 > no.0 monoonce> :eneonnm oce :enmmom o No.0 No.0 on.0 no no.0 m_.o 00.0 min no.0 onNeeonsneonmme o «0.0 00.0 on.n o no.0 no.0 nn.o oin no.0 coneommoN N No.0 No.0 no.0 N N_.0 no.0 o_.0 > no.0 monoonce> cen>eceoceon oe no.0 No.0 NN.0 no 00.0 no.0 no.0 .mon on.0 enmemsomm e 00.0 no.0 0N.0 ne no.0 no.0 N_.o : nn.0 ooneoosmn e 00.0 00.0 nn.o oe 0N.o no.0 no.0 e mN.o connomsoo e 00.0 00.0 nN.o e 00.0 No.0 00.0 non nn.o ooneoNon e 00.0 00.0 nn.0 e 00.0 00.0 00.0 one 0N.o oooeoeeon > 00.0 no.0 nN.0 > no.0 n0.o No.0 N on.0 monoonce> oooezn necpcoo e 00.0 no.0 nn.o oe no.0 00.0 Nn.0 con on.o conemNNnnse e 00.0 00.0 nn.o oe no.0 00.0 00.0 o-e no.0 eonnooe o 00.0 no.0 no.0 ne 00.0 No.0 no.0 o-e No.0 cosmooocs e 00.0 00.0 nn.o e 00.0 00.0 00.0 ne no.0 ooneoNNn e 00.0 No.0 no.0 no no.0 n_.0 no.0 ouo no.0 consonn e 00.0 00.0 nn.0 e 00.0 no.0 no.0 :0 mn.0 .ennoce .n. e 00.0 00.0 nN.o e 00.0 mo.0 «0.0 5-0 nn.0 oceansoe e 00.0 NN.o 0n.o no no.0 no.0 m_.0 :uo Nn.0 oceans > 00.0 no.0 on.0 > 00.0 No.0 No.0 > No.0 monoonce> ooocon - - - - - I - - - - - u i - - - u u . :nmocoono & a - I u - u - u - - - - Y Y n - n . .ocno copoon no monuonee> oce moonooe ooozuon monoe :nmoc goons no momma no congenee> no ozocxco no ozooxoo oueuonneoco zoo nm esocxoo + oueoocun Fm czocxco om ozocxco mm czocxco :eocpoon ooe :coumoz .N o—noh 15 Table 3. Variation of means Of various Oleoresin traits be- tween regions and varieties of Scotch pine (Rose Lake). Water Oleoresin Canal Potential Pressure Area Canals (Bars); (atm.) (micron ) per mm Western and Southern Europe varieties -5.08 Z 6.07 Z 1.78 Z 1.50 armena -4.42 abc 6.42 1.68 a-e 1.35 aquitana -3.64 ab 5.80 1.63 a-e 1.41 'E. anglia' -5.31 abcd 3.72 1.96 c-e 1.42 iberica -4.23 abc 6.75 1.85 b-e 1.72 iZZyrica -5.52 abcd 3.94 1.35 ab 1.06 rhodopaea -5.52 abcd 4.57 2.00 C-e 2.35 scotica ' -8.66 d 6.72 1.44 abc 0.75 subiZZyrica -6.1O abcd 7.04 1.93 c-e 1.39 Central Europe varieties -5.26 Z 5.09 Z 1.64 Z 1.29 pannonica -3.08 a 5.89 1.74 a-e 2.14 polonica -6.30 abcd 5.64 2.12 e 1.78 carpatica -7.22 bcd 3.72 1.52 a-d 0.79 hercynica -5.76 abcd 4.76 1.47 abc 1.37 hgguensis -5.33 abcd 5.23 1.66 a-e 1.07 Scandanavian varieties -6.02 Z 3.78 X 1.43 X 1.39 Zapponica .-7.52 cd 1.47 1.24 a 1.13 septentrionalis -5.75 abcd 4.14 1.47 abc 1.45 Russian and Siberian varieties -3.91 X 5.80 Z 1.77 Z 1.43 mongolica -2.84 a 4.97 1.70 a-e 0.89 'N. E. Siberia' -6.83 bcd 2.10 1.92 c-e 1.86 uraZensis -3.96 abc 6.63 2.05 d 1.59 Regional values within the same column followed by the same letter are not different at 5% by Duncan's Multiple Range test. Varietal values within the same column followed by the same letter are not different at 1% by Duncan's Multiple Range test. 16 collection were found. The Scandanavian varieties had the lowest Oleo— resin pressure. The range Of values within varieties was large, however, and no statistically significant differences between varieties were found. These seed sources also had the smallest resin canal cross-sectional areas. No genetic variation in resin canal density was observed, averag- ing 1.4 canals per millimeter of stem circumference. Correlation analysis showed very hgih correlations among many of the resin acids (Table 4). As total resin acids increased, individual compounds also increased, for all but four minor resin acids. This has also been observed in white pine (Bridgen, Hanover and Wilkinson, 1979). Pleiotropy, the influence of one gene or a few closely linked genes on several traits, is probably the reason for this. The compounds which consistently showed low or high concentrations within a region correlated together highly. Water potential correlated positively with Unknown 30 (p=5%) and Unknown 36 (p=1%), but none of the other traits correlated with any Of the resin acids. 0f the four physical traits, only resin canal density and resin canal area correlated. Any relationship between water potential and Oleoresin pressure that may exist cannot be deter- mined as these traits were measured at different times. The correlations of these traits are shown in the following table: Water Oleoresin Resin Canal Potential Pressure Density Water Potential Oleoresin Pressure -.23 Resin Canal Density .10 .Ol Resin Canal Area .03 .18 .51 *** Several traits observed in this study show similar trends on a geo- graphic basis, but do not correlate. For example, the Scandanavian) 17 varieties had the weakest Oleoresin pressure, smallest resin canal areas, and the highest concentration of total resin acids, but none of these traits correlates with the others. On the other hand, resin canal density and area correlated positively, but canal density does not show regional variation and resin canal area does. This is due to the vari- ation of the traits in other parts of the species range and indicates that each physiological trait is independent of the others. Figure 3 illustrates the relationship between resin canal density and resin canal area. The variation in several of these traits indicates a possible rela- tionship with resistance to some insect pests. Scandanavian varieties have the highest resistance to European pine sawfly (Wright et a1, 1967) and eastern pineshoot borer (Steiner, 1974). Also, varieties which have the lowest resin acid concentrations are the most susceptible to these insects. This connection may only be coincidental, but correlation analysis between insect resistance and resin acids for all seedlots show a significant negative relationship. Although the physical traits show apparent relationship with resistance, no strong correlations have been found on range-wide data. This implies that resistance is also independ- ent of the physiological traits. 18 00.1 rrrmm. «rsmm. ¢—.| repv. tom. sxrpw. rrsmm. rsrnm. serum. mm. *mm. mN.u mono< :wmmm Pouch .vn n~.n no.1 nesmo. NP. no.1 no.1 N_.n Fm. no.1 mo. mo. on. no czocxco .mn rssom. FN.I «*mm. mm. *«rww. rsroo. rsrnm. enema. om. cram. n~.u mpnpmwnoomz .NP Po.n «cram. «mm. *xrmu. rrrmm. *«Nmo. sssmm. em. scum. on. muopmwn< .FP mp. wo. wo.1 mp.n NN. o—.: m—. no. xmm. mm czozxca .o— errem. mm. rpm. rsrme. ¥r¥n¢. rnsom. ¥¥¢¢. rssnm. opoumnno nocozcoo + eoenenom .e en. mm. *«NV. em. rrsmn. rspv. om. cm czocxcz .m eeenN. neeNn. .neoN. No.- nN. e_.- onenesneeen .N nrrme. xrsmm. mm. mm. ermm.n wunnumzpoa + opecesnoo>on .0 ...Nn. .om. enN. eN. eoeeesnn sooeeeooem .n eom. neon. on.- ooecoeeoo + oueceena .o nN. Fm. Fm czocxco .m emm. om ozooxoo .N om czooxoo .H on Np FF on 0 o N n n o m N n gonzo: oce oneen .oono souoom no monoe znmoc coozoon moonoe—ossoo oposnm .0 onnen 19 Figure 3. Distribution of resin canal density and resin canal cross- sectional areas of various seedlots and varieties Of Scotch pine Resin Canal 20 Density Nth/mm 4.5 — 0 Central Europe a; Seandanavran 9 Russian and Siberian 0 Western and Southern 3.6 — Eur-ope i o 2.7 ._ o . o O * O 1.8 .— O Q" . . 0 e ,o- . . i o ' o G 0.9 —- ‘ 0 o . o O O o ‘l o e e °~° I I I I I I 1.0 1.3 1.6 1.9 2.2 2.5 2.8 3 2 Resin CanalArea 16 mm Chapter 2 Within-tree and Seasonal Variation of Oleoresin Physiology in Scotch Pine Abstract The within-tree variation of monoterpenes in the cortical Oleoresin of a single Swedish seedlot of Scotch pine (Ednus sylvestris L.), certain traits affecting Oleoresin production, and the seasonal variation of the resin acids in the Oleoresin were observed. The concentrations of four major resin acids, two minor acids and total resin acids reached a mini- mum during the May 15 to July 1 period. Strobic + dehydroabietic acids reached a maximum Of 2.32 percent Oleoresin at this time. The mono- terpenes did not vary between tissue of the same age on different branch- es although a-pinene tended to increase with the height of the tree. Oleo- resin pressure was the only physical trait observed which varied in the crown, decreasing from the lower branches to the tOp of the tree. Pos- sible effects of these variations are discussed. 21 22 Introduction The primary chemical components of the cortical Oleoresin of pines are monoterpenes and resin acids. Other traits which are directly in- volved in Oleoresin physiology include the density and cross-sectional area of the resin canals, Oleoresin pressure and water potential. Know- ledge of any seasonal or within-tree variation of these traits is nec- essary to develop a complete understanding of the effects of this system on insects or diseases. 7 Seasonal variation in the chemical composition of Oleoresin occurs in some pine and spruce Species. Several resin acids of slash pine (Pinus eZZiottii Engelm.) in the phloem, xylem and needle tissue vary in concentrations throughout the year (Finnerty and Falco, 1977). Total resin acids in both the xylem and bark-phloem tissues reach maxima in January-February and May-June. Abietic acid showed the least variability of all the resin acids in the xylem. Squillace et a1 (1971) concluded that levopimaric acid is not affected by seasons in slash pine, although only two sampling dates were used. Moisture stress in loblolly pine (Pdnus taeda L.) causes decreased resin acid and increased monoterpene concentrations in the Oleoresin. During periods of drought, decreased levels of levopimaric + palustric acids and increased levels ofcx-pinene and B-pinene occur (Hodges and Lorio, 1975). Reports of seasonal vari- ation of the monoterpenes have been made for black spruce [Edcea mariana (Mill.)B.S.P.](Ruleff, 1975a), blue spruce (Picea pungens Engelm.) (Rudloff, 1975b), and Sitka spruce {recon sitchensis(Bong.)Carr](Hrutfiord et a1, 1974). There may not be any such variation in the cortical blister Oleoresin in Abies (Zavarin, 1968). Seven monoterpenes in the cortical Oleoresin of Scotch pine vary seasonally, but the variation never exceeds 23 five percent (Tobolski, 1968). The chemical composition also varies between tissue within a tree ‘and with the age of the tissue. Zinkel (1976)-has reported distinct dif- ferences between the resin acids in the needles and cortex of Japanese red pine (Pinus densifiora Sieb. and Zucc.). The upper stems of blue and -white spruce (Picea glauca(Moench) Voss) have higher concentrations of abietic and neoabietic acids that the lower stems (Westfall, 1972). Most of the monoterpenes of Scotch pine vary in concentration between needle, cortex, and xylem tissue, as well as between tissues of differing ages (Tobolski, 1968). Alpha-pinene decreases in concentration with in- creasing height Of crown in both slash (Franklin, 1976) and loblolly (Gilmore, 1975) pines. Most of the other major monoterpenes also vary with height Of the crown, but their directions are not consistent be- tween species. This study was conducted to determine the variation of some chemical, anatomical, and physiological traits of Oleoresin physiology within the crown of a seedlot of Scotch pine and the seasonal variation of the resin acids in the cortical Oleoresin. Methods and Materials The trees in this study were growing in an eighteen-year-old Scotch pine range-wide provenance test on the Rose Lake Wildlife Research Area in Shiawassee County, Michigan. Only one seedlot, number 222 from Sweden of the variety septentrionalis was used. Samples for each of the experi- ments were taken at 1.5, 3.0, and 4.5 meters in the crown of the trees. Seasonal Resin Acid Variation - Beginning April 1, 1978, two 7 mm 0.d. x 50 mm capillary tubes were placed into 14 trees at 1.5 m from the ground 24 in the crown. A 1.6 mm wide hole was drilled at a 45 degree angle into the 1976 branch tissue to a depth of 6.4 mm. The tubes were collected on the following day and were stored at 2 C until analyzed. This was re- peated every fifteen days until October 11. Resin acids were analyzed by gas-liquid chromatography using the same methods described earlier by Bridgen, Hanover and Wilkinson (1979). The length of the terminal and lateral buds or shoots on a selected side branch of each tree was measured at the time Of Oleoresin collection to establish the timing of growth in the trees. Resin canal density and area - Branch cuttings, 12.7 mm in length, from two-year-old tissue, were collected from 14 trees of the seedlot and stored in 70% ethanol. Before measuring, the ethanol was drained and the cuttings were air dried. Microscopic examination was made with a Wild Heerbrugg M-5 stereomicroscope. The diameter of the xylem, the number of resin canals in the past year‘s growth ring and the diameter of five resin canals were recorded. The ratio Of the number of resin canals to the circumference of the stem was used as an estimate of resin canal density. Oleoresin pressure - Small micromanometers, similar to those described by Hodges and Lorio (1968), were constructed from 20 microliter capillary tubes. The tubes were cut at the calibration mark and fire-sealed. The open ends were tapered with a file. On July 12, 1977, 6.4 mm deep hole was drilled into the 4-year—old whorl of a branch at each height with a 1.6 mm drill bit. The micromanometer was tapped into the hole with a plastic hammer, sealed with silicon rubber and left in the trees over- night; a light rain fell during the day of insertion. Between 8:00 and g 10:00 AM on the next day, the length of air pockets inside the sample tubes were measured. Pressure inside the air pockets was calculated by 25 dividing the total length of the micromanometer by the length Of the air pocket, according to BOyle's Law: P V 1=Pv l 2 2 P2 = P1V1/V2 and P = l Atm. 1 water potential - Cuttings (12.7 cm) of lateral branches of the western side of the trees were made at each height within the crown. The base of the cutting was stripped of 2.5 cm of bark and phloem. -Water potential of the xylem was measured by placing the cutting into a pressure bomb (Scholander et at, 1965) and observing the pressure at which water exuded. The pressure was converted to bars by multiplying by a conversion factor of -0.0689. The day of collection was very hot and humid and all measure- ments were made between 10:00 AM and 12:30 PM. Monoterpenes - On July 7, 1977, glass capillary tubes were inserted into eleven trees at each height in the same manner as the micromanometers. The tubes were allowed to set overnight and fill with Oleoresin. They were stored at 2 C until analyzed. Analysis was done on a Hewlett-Packard 5700 gas chromatograph with a hydrogen flame ionization detector and a digital integrator. The column was 6 mm x 240 cm, stainless steel packed with Chromosorb W-AW coated with 10% polypropylene glycol. With this column, limonene and B-phellandrene have the same retention times and are unresolved. The monoterpenes were identified by retention time and a standard curve was used to determine concentrations as a percent Of Oleoresin. 26 Results Resin acids — Twelve resin acids or pairs of unseparable resin acids were measured in the Oleoresin Of Scotch pine. The concentrations Of these acids and their sum as they vary throughout the eleven sampling dates are shown in Table 5. Four major resin acids, sandaracopimarate, levopimarate + palustrate, strobate + dehydroabietate and neoabietate, two minor acids, unknowns 28 and 38, and the total resin acids showed significant seasonal variation. The combined peak Of strobic + dehydro- abietic acids snowed an increase in concentrations between May 15 and June 15, after which it decreased to levels close to those prior to May 15. All of the other resin acids, and total acids, had a noticeable decrease during this same_time period, extending to July 1. This drop in total resin acid concentrations coincides with the time of most rapid shoot elongation of lateral branches (Figure 4). Monoterpenes - Table 6 lists the concentrations of the eight monoter- penes measured. Alpha—pinene concentrations are higher than expected for a Scandanavian seed source, but this may be due to the seedlot vari- ation within the variety septentrionalis. The concentration of 3—carene is high, which is typical Of all Scandanavian sources of Scotch pine. None of the monoterpenes varied significantly between heights of the crown although depinene tended to increase with height. It would seem that although variation exists between different aged tissue of the same branch (Tobolski, 1968), tissue of the same age on different branches do not vary. Physical trait variation - The mean values of four physical traits at each height in the crown are listed in Table 7. The pressure of Oleoresin in the stem is the only trait that showed statistically significant S. n .... “no. n .. n¢.mF sen.¢ mm.om mo.mF mN.Nm No.0N Nn.w nn.oF Nn.mF nF.nF nm.mm ne.nF mn.mm FoHOF 0F.N st.v wF.¢ «F.n Fn.¢ nm.m nm.m n¢.F nm.F mF.m on.n FN.~ mm.m ouoponooooz nm.F no.m nF.v mo.m o0.m nn.F on.F wo.m mm.m no.0 m0.N ne.m mpoeonn< mm.o eon.m I wF.o on.o oo.o 00.0 mm.o no.0 no.0 Fm.o No.0 No.0 mm czocxco nn.o sson.¢F ¢O.F on.o oo.o «F.F Fm.o mm.N 0o.F on.F mo.F om.F no.0 owopoFooocoxcoo + ouooogum on.F nm.o Om.o 00.0 no.0 00.0 oo.o o0.o 00.0 o¢.o w¢.o NF.0 nm zeocxco mn.m mw.F mm.m NF.N nF.F mn.o N0.F em.F F¢.F Nn.F om.F Nn.F ouoLoEFQOmF nF.w «son.n No.0F 0F.FN on.wF m¢.n F¢.F mm.F n¢.m nw.¢ 00.0F VF.n on.o ouoepmoFoa fin + ooeeeenoo>on mm.o aom.¢ mm.o mm.o mm.o mm.o nv.o 00.0 On.o mm.o wo.F Fm.o Rn.o mponosnoooonoucom Fn.m Fm.F nm.m NN.N ON.F Fm.o NF.F ON.F vv.F mm.N Fm.F oo.N muocoeeou + ooeceena o¢.N mF.o mF.o oo.o mF.o mm.o mN.o mF.o mo.o mm.o Fo.o no.0 Fm :3ocxca mn.F mN.o mm.o 00.0 0N.o vF.o mo.o oF.o no.0 mF.o mo.o wo.o om czocxca Nm. o neon. NF no.0 mm.o o¢.o mo.o no.0 mm.o mm.o mF.F nN.F mm.F em.F mm czocxca em; ooFo>I n poo FFI m0 Feoomeon .n aneF 28 Figure 4. Seasonal variation of the total resin acids in cortical Oleo- resin Of a Swedish source of Scotch pine and the rates of elong- ation of lateral branches 29 Percent of Oleoresin 35 - Growth cm per Day 6 -1.0 IIIIIIIIIIIuanotai Resin Acids Le 4 .3 — .2 - — -Terrninal Growth Lateral Growth t 11% t I1_m. no 11 A V. 11m 1.. e Inlm J m 11 M r 11 o. A SD- 25— _ o 2 15.... ‘IO_ 5 3O Table 6. Mean values of Oleoresin monoterpene concentration at three heights within the crown of Scotch pine. Height Monoterpene 1.5 m 3.0 m 4.5 m Overall mean ------- % Oleoresin - - - - - - - - - - a-Pinene 18.35 23.11 27.43 21.43 Camphene 0.56 O 58 0.64 0.58 B-Pinene 2.20 3.15 1.03 2.22 Myrcene 0 51 1 87 0.79 0.93 3-Carene 9 37 5 36 6.50 7.73 Limonene + B-Phellandrene 0 79 1.41 0.84 0.96 y-Terpinene O 10 0.10 0.07 0.10 Terpinolene 0 93 0.69 0 0.81 .65 Total Monoterpenes 32.80 36.26 37.96 34.76 31 variation. The ranges Of pressures at each level were: Height_ Range 1.07 - 2.67 atm. 4.5 m 3.0 m 1.09 - 6.33 atm. 1.5 m 1.65 - 5.43 atm. The pressure decreased from the top to the base of the crown, and the greatest variation in values occured in the center. Water potential showed no variation between heights. Most of the variation in measurement (71.5%) was due to random variation. Although it is expected that water potential will decrease with increased height in the crown, due to increasing stress of the water column, the trees examined were probably not large enough to show noticeable differences. Resin canal cross sectional areas ranged from 0.28 micron2 to 3.58 micronz. Although the average value increased from the base of the crown to the top, a large amount of variation between trees (OixH=25% total variation) exists. The density of resin canals, also, does not vary within the crown. DiSCussion Scotch pine is attacked by many insect pests. The larvae of the European pine sawfly (Neodiprion sertifer Geoff.) feed on year-old foliage between early May and mid-June. Also during this period, larvae of the eastern pineshoot borer (Eucosma gloriola Heinrich) bore into lateral and terminal shoots, prior to full shoot elongation. The timing of these events corresponds very closely to the decrease of total resin acids in the Oleoresin of branches. It may be only coincidental that the devel- opmental physiology of the insect parallels the change of chemical com- position in the Oleoresin, but it is also possible that the insect has adapted its developmental stages genetically to feed as a time when 32 Table 7. they vary within the Trait Oleoresin Pressure (Atmospheres) Water Potential (Bars) Resin Canal Area (Microns Resin Canal Density (no./mm) crown Of Scotch pine. Height 1.5m 3.0m 4.5m 2.32 2.07 1.82 -4.64 -4.80 ~4.92 0.95 1.26 1.40 1.35 1.55 1.55 Mean value of four physical traits of Oleoresin physiology as F-test of Height Differences 4.63 * 0.06 2.60 0.47 * p<.05 33 undesirable compounds in the host are at a low concentration. A similar situation has been described with the winter moth (Operophtera brumata L.). The tannin concentration of oak leaves may have caused evolution of the winter moth so that its feeding ceaSes before tannins increase in concen— trations in the leaves (Feeny, 1970). The occurrence of such a pheno- menon for these of Scotch pine would be confirmed if resin acid concentra— tions decrease in all species upon which the insects feed. Also, studies in which the resin acid content of tissues fed to the insects is varied could support the hypothesis. A decrease in Oleoresin pressure with increased height of the crown could have an effect on insects which feed on the extremities of the branches such as the shoot borer or the white-pine weevil (Pissodes strobi Peck). Studies are needed to compare the relative Oleoresin pressures of trees attacked by a specific insect to those not attacked to determine if this is true in vivo. Knowing that the largest pressures occur at the base Of the crown, all measurements should be taken at this point to show maximum pressures. The absence Of within-tree variation of monoterpene concentrations, water potential, or resin canal areas or density in l8-year—old ScotCh pine allows measurement of these traits to be made at any point in the CY‘OWIl . Chapter 3 Genotype X Environment Interaction of Oleoresin Physiology of Pinus sylvestris L. Abstract Progenies from 27 seedlots of Scotch pine, planted at two locations in central Michigan, were studied to identify genotype X environment interaction and effects of planting site on the chemical and physical properties of the Oleoresin system. Thirteen resin acids, water poten— tial, resin canal density and cross sectional areas were examined. The concentration of resin acids showed no consistent variation due to planting site or seedlot, except for Unknown 41 which showed varietal differences. Most of the major resin acids had a high percent of varia- tion due to seed source X plantation interaction. Fifty six percent of the variation of total resin acids was due to the interaction. Water potential, resin canal area and density also showed high genotype X environment interactions. Plantation differences were Observed in the cross-sectional area of resin canals. 34 35 Introduction A phenomenon of genetics which has great importance to plant breed« ers is genotype X environment interaction. This situation occurs when the relative responses of two genotypes for a trait differ in two differ- ent environments. Shifts in ranking of several genotypes over planting sites also indicate interaction. The nature of genotype-environment interaction for any trait must be understood by plant breeders so as to determine the geographic areas within which an improved variety may be planted. Growth traits of trees usually show genotype X environment inter- action in studies which are designed to identify it. Interactions of growth traits of Scotch pine (Pinus syZvestris L.) have been well studied. Wright and Baldwin (1938) suggested the existence of a source-site inter- action for growth rate by comparing experimental plantings in Sweden and New Hampshire. Interactions for height growth, needle length and color were found in a study of 122 seed sources and eight planting sites throughout the central United States (King, 1965 a,b). Evaluations of three and four-year old seedlings only one or two years in the field showed the seed source X plantation interaction for height growth to be small. It fluctuated between years and plantations, and was greatest (8.5% total variation) when comparing plantations at Rose Lake and Kel- logg Forest in Michigan. Significant interactions were also found for needle length and color, although these were small in comparison to the seedlot component. These and other traits were examined by Schrum et a1 (1975) on 27 families in nine plantations throughout New York, New Jersey, and Pennsylvania. They concluded that needle color, tree height, branch length and numbers of branches are stable traits, while needle length, branch angle and crown taper are more variable over large areas. 36 Interaction is also small for cone production, but higher for resistance to extreme winter cold and insect and bird pests (Wright et a1, 1976). The chemical composition Of the Oleoresin of Scotch pine has also been studied. Tobolski and Hanover (1971) reported geographic variation in the monoterpenes of a provenance plantation in southern Michigan. Variation in the resin acid component has also been observed (Chapter 1). It has generally been assumed that the monoterpene concentrations are not affected by the environment in which the tree is grown. Mirov (1961) suggested that the monoterpene content is not affected by environment as seedlots of species planted outside Of their native ranges do not differ from trees within their range. Hanover (1966) reported that the mono- terpenes of four clones Of western white pine (Pinus monticoZa Dougl.) showed no variation in their monoterpene composition over three planting sites. He concluded that the monoterpene composition is independent of the environment or interaction with genotype. But Tobolski (1968) reported location differences forII-pinene, myrcene, and B-phellandrene. He also noted a possible seed source X environment interaction for B- phellandrene. It was suggested that this variation was due to defolia- tion by the European pine sawfly (Neodiprion sertifer Geoff.). The purpose of the present study was to determine if genotype X environment interaction exists for the resin acids Of the cortical Oleo- resin Of Scotch pine and several anatomical and physiological traits associated with the Oleoresin system. 37 Methods and Materials Two seventeen-year-Old plantations of a range-wide provenance study of Scotch pine were used to collect both Oleoresin and physiological data. Each had been established in 1961 as part of the NC-99 study of Scotch pine. The Rose Lake plantation, on the Rose Lake Wildlife Experi- ment Station in Shiawassee County, Michigan, has a loamy sand soil with a fine-textured, moisture-retaining layer at about 50 cm deep. The terrain is gently rolling. The second plantation is in the Higgins Lake State Forest, Crawford County, Michigan. It, too, has a loamy sand soil on a level terrain. The following comparative climatic data were recorded at nearby weather stations:‘ Temperatures Annual1 Annual Mean Mean Mean Precipitation Snowfall Annual Maximum Low Higgins Lake 27.88 in. 60.7 in. 43.0 F 55.6 F 30.4 F Rose Lake 29.76 in. 37.0 in. 47.6 F 58.1 F 37.1 F Seed procurement and nursery establishment has been described by Wright and Bull (1963). Twenty-seven seedlots from across the range of Scotch pine were used in the study. The distribution of the seedlots and varie- ties has been given previously (Chapter 1). Oleoresin was collected from the bole of each tree at 1.5 meters from the ground, A 1.6 mm wide, 6.4 mm deep hole was drilled into the bole, and a twenty microliter glass capillary tube was tapped into the hole with a plastic hammer. Silicon rubber was used to secure the tube to the tree. The capillary tube was allowed to fill with Oleoresin overnight, and was collected the next day. The tubes were stored in 1From Climatography Of the United States. No. 11-16. U. S. Weather Bureau Climatic Summary 1931-1952. 38 nitrogen at 2 C until analyzed. Two replications of each seedlot in each plantation were collected, with three trees from each replicate. The Oleoresin from all trees in the replication was bulked to minimize time on the gas chromatograph. Samples were collected between June 27 and July 1, 1978, on the Rose Lake plantation and between July 3-4, 1978, at Higgins Lake. The method of chemical analysis has been described by Bridgen, Hanover and Wilkinson (1979). Water potential was measured on each of three trees in one replica- tion for each seedlot. A 12.7 cm cutting was taken from a lateral branch at 4.5 meters on the south side of the crown. A 2.5 cm strip Of bark and phloem was stripped from the base of the cutting. The branch was inserted into a pressure bomb (Scholander et a1, 1965). The pressure in pounds/inch2 at the time that water exuded from the cutting was re- corded. Pressure was converted to bars by multiplying by -0.0689 as a conversion factor. All trees at Rose Lake were measured between Septem- ber 6 and September 8, 1978 and those at Higgins Lake between October 12 and 13, 1978. All measurements were made between 11:00AM and 4:00PM. Oleoresin pressure was measured with micromanometers. Twenty-micro- liter capillary tubes were cut at the calibration mark and fire-sealed at the cut end. The Open end was filed to taper the tube. A 1.6 mm wide hole was drilled into the bole Of each tree at 1.5 meters from the ground. The micromanometers were inserted to a depth of 6.4 mm and secured with silicon rubber. 0n the morning of the day following insertion, the ‘ lengths Of the air pockets inside the tubes were recorded. Oleoresin pressure, measured in atmospheres, was calculated by dividing the length of the manometer by the length of the air pocket. Three trees in each of two replications per seedlot were measured. 39 Branch cuttings were made on each of three trees in one replication of each seedlot at 4.5 meters in the crown. A 5 cm piece of stem cutting, at least two—years-Old, was taken from each branch, labelled, and stored in test tubes, covered with 70% ethanol. After two weeks, the ethanol was drained and the wood samples were air dried. Thin slices of the stem were made with a razor blade and stained with two solutions. The first was 0.5 g Safranin 0 in 200 ml of water and the second was 1.5 g KI + 0.5 g I in 100 ml of water. The minimum and maximum axes Of three resin canals in the previous year's growth, the total number of canals in the ring, and the diameter of the stem were recorded. Resin canal area was calculated by Area =1Tab where a and b are the lengths of the semiaxes. Total number of resin canals divided by the circumference of the stem was used as an estimate of resin canal density. Results Resin acids - The design of the analysis of variance and breakdown Of variance components for the resin acids experiment is shown on Table 8. With this design, differences in the means Of the plantations or seedlots could be identified, as well as the presence Of an interaction between seedlots and planting location. Differences among replications in the same plantation were negligible. The distribution Of concentrations of each resin acid, as a percent of total Oleoresin, and the sum of all resin acids as they varied between plantations and varieties, is shown in Table 9. Individually, the com- pounds are highly positively correlated, and the total resin acid figures adequately represent the trends of each resin acid. Total resin acids ranged from 6.1 to 65.0 percent of Oleoresin in the Rose Lake plantation 4O Table_8. Degrees Of freedOm and variance components of a seedlot and environment effects study. Source *' D.F. . _ .E.M.S. . 2 '2 2 Plantation . (p-l) oe + sor + srop Rep within plantation (r-l)p o: + sci 2 2 2 Seed source (s 1) oe + ropS + rpoS Plantation X Seed source (s-1)(P-1) a: + ragS Error (s-1)(r-1)p o: 02 = component of variance due to plantation site Q Q fDN UNMN‘N'C O O U! - - component of variance due to block component of variance due to seed source component of variance due to interaction of plantation with seed source component of variance due to error 41 and from 6.2 to 81.0 in the Higgins Lake plantation. The Russian and Siberian varieties had the highest concentrations while the Central Europe varieties had the lowest in the Rose Lake plantation. At the Higgins Lake plantation, however, the Western and Southern Europe varie- ties had the highest concentrations while Central Europe and the Russian and Siberian varieties had the lowest. The distribution of the percent of variation for each component within the study is given in Table 10. There were no differences between plantations for individual or total resin acids. Over all seedlots, Rose Lake trees had an average total resin acid composition of 23.6 percent Oleoresin while the Higgins Lake trees averaged 23.1 percent. Similarly, over both plantations, no differences between seedlots were consistent, except for Unknown 41, which had its highest concentration (0.2%) in variety septentrionalis and its lowest in Central Europe varieties. The plantation X seed source interaction term, however, accounted for a significant portion of the main effects terms. F-tests showed significance at the 1% level for all of the major resin acids, as well as Unknowns 28 and 36, compounds in very low concentrations. For total resin acids, this source accounted for 56.0% of the total variation. Figure 5 illustrates the extent of the seed source X plantation in- teraction for neoabietic acid. Only eight varieties are shown to keep the figure simple. Higgins Lake plantation is used as the poor environ- ment as the Scotch pine grow so slowly on the site. Varieties mongolica, uralensis, and carpatica have higher concentrations Of neoabietic acid in the Rose Lake plantation while all of the other varieties have lower levels than at Higgins Lake. nn.0 nn.F nn.0 Fo.m 00.0 00.0 nF.0 0F.o on.0 on.0 onoeoNoN: 0m.F NF.F nm.m Fn.e oF.o ow.o 0F.o nF.0 on.0 nn.0 .eFxoan .n .z. no.0 no.0 on.0 nF.N mm.o 0F.0 no.0 Fo.o nn.0 no.0 oonNooeoE Fn.0 nN.F nn.F oo.m no.0 NF.0 NF.o 00.0 on.0 Fn.0 monoonxe> :enxoan one :emezm nn.0 on.0 nn.F no.m nN.0 no.0 on.0 NF.0 on.0 on.0 onNeeoosneonmoo nn.0 on.0 nn.F no.m nm.0 no.0 on.0 NF.0 on.0 nn.0 moFquxe> cen>eoeoceon Nn.o nn.0 nn.F nn.F nF.0 00.0 nF.0 no.0 nn.0 nn.0 onoeosoeo nn.0 nn.0 N¢.F 0N.F om.o no.0 nF.0 mw.0 on.0 No.0 ooneooson on.0 nn.0 no.0 nn.F no.0 no.0 00.0 no.0 Fn.0 mm.o consensus on.0 on.0 nn.F 0¢.F nF.0 no.0 0N.o FF.0 on.0 nn.0 monoonxe> 42 oooxon Fexpcoo no.F nn.0 on.m F¢.m nF.0 om.o o~.o nF.0 No.0 00.0 soomooofiN no.0 on.0 nn.F nm.F nF.0 nF.0 nF.0 oo.o on.0 mm.o eonsoNNn oo.F on.0 Fm.N no.F nF.0 oo.o FF.0 no.0 nw.o nn.0 consonw nn.F nn.0 vv.m no.0 nF.0 no.0 mm.o no.0 nn.0 nn.0 oeonnzoo nn.F F0.o mm.m nn.F 0F.o v0.0 mm.o nF.0 nn.0 Nn.o assess wF.F mm.o mn.N mm.F nF.0 FF.0 mF.0 mo.o Nm.o on.0 moFquce> oooxzm ccoeuzon ooe :xoomoz I I I I I I I I I I I I I I I I I I :FmoxooFo no & I I I oxen oxen oxen oxen oxen oxen oxex oxen oxen oxen mcFooF: omoa monoonz omom menoonz omom mcFooF: omom mcFooF: omom opexeeFo ope0355oo + Fm czocxoo om czooxoo om ozocxco Iooeeeocem ooexeene .eeennonz :F moonueooF can we czoxo oonq :ouoom no mononxe> oce mcoFoox no moFoe :Fmox no mceoe ocF .0 anen 43 oxen monoonz omom no ozoexco 0F.o nn.0 mo.o on.0 oo.o on.0 no.o Nn.o 00.0 NN.F o0.o NN.F No.o on.0 oF.o nn.0 oo.o mn.o no.0 nn.0 no.0 nn.o oo.o NF.F No.0 em.F oo.o nw.o nn.0 NF.F NF.o nF.F oxen oxen nn.0 mo.F FN.F F0.F no.F nn.0 on.0 no.0 nn.0 Fn.0 oxex monoonz omom wN.o mm.o no.o mv.o mm.o oF.o I I :FmoxooFo no oxen oxen menooF: omom o\° NF.F F0.F mm.m ow.F nn.0 nn.F nn.F nn.F MN.N nn.F mm.m nn.F FF.~ Nn.0 nn.F mm.o nm.o om.o oF.F nn.0 nn.F no.0 nn.F mm.F Fm.F nn.0 mn.F Fm.o nm.N no.0 0N.F nn.0 oxen oxen monooF: omom ooeponneoeoacoo + eoeneeom nm czocxco m#MLoEmemH N¢.F on.0 mN.F mo.n om.n nn.n nm.F ¢¢.n 0w.mF Fe.n Fm.NF F¢.wF N0.nF nw.mF oxen F mm.nF nm.nm ov.nF nn.wF on.mF on.0F Fn.m ne.n Fn.NF nn.m FF.mm mF.FF no.0F nn.0 nn.m om.NF oxen monoonz omom opexomsFeo + ooeceenoo>ox .enxonnm .m .z. ooeNomeos monooer> :eFxoan oce cemezx onNosoosnsoomoo monoonxe> wa>ocowcwom onososmon nonsmoeox donoomsoo monoer> oqoxzn Fexpcoo doomoeoea eoeennee consoan cronnxoc 6§m§ monoonxe> oooxsn ccozooon oce :eoomoz .oozcnocoo 0 anen 44 F¢.Fm mo.0m NN.0N on.0m w¢.nm ow.Fm oxen on.00 nm.¢¢ nn.Fm on.0m mm.nw mm.nm mn.mm nF.oF 00.0N nN.oF nF.mF nn.NN oxex monoonz omom meno< nneee neoeF no.o no.0 oxen enneonn eooe Fe czocxco 00.0 0¢.N nN.m oo.F nm.m 00.0 0m.n om.n mm.n nm.e 00.0 mF.N m¢.e nn.F oo.m 00.0 on.m mm.e 0n.N 0o.m No.0 on.e nn.m ve.m N¢.N No.0 nn.e nn.m oe.m N¢.N F0.o nF.0 en.m e0.m nn.F 00.0 Fo.m on.0 oo.N om.N 00.0 nn.F on.N 0n.F 0n.F 00.0 on.m on.N nn.N no.F Fo.o nn.n NN.¢ nN.m oo.m 00.0 nn.m N¢.N on.~ 0n.F 00.0 on.n nF.m mm.e n0.F 00.0 mn.n n0.F oe.e no.F 00.0 No.n mo.N n0.m nn.F oo.o on.n on.m no.e F0.F I :FmoxooFo no R I I I I I I I I I I I I oxex oxen oxen oxen oxen monooF: omom moFooF: omom opeoonneooz opeoono< onoeoNes: .eFLonFm .n .z. dooNomsoE moFooer> cenxoonn ooe :enmmzz onNoeonsesonmoo monoonxe> ceF>eceoceon ono:o3mo£ enormosmx eonncmeoo moFquLe> ooocsn Fexocoo coonoooxs eoeennne consoee osonosoo gmtg moFooFLe> oooxzn :xogpzon oce coopmoz .oozonpcoo o oFoen 45 Table 10. Components of variance, as a percent of total variance, for resin acid composition Of 25 seed sources of Scotch pine planted at two locations. 2 2 2 2 C2 Resin Acid Ep 31 is fps _2 Unknown 28 5.0 15.1 10 2 35.6 ** 34.2 Unknown 30 2.8 10.1 3 5 0.0 83 7 Unknown 31 6.3 10.6 0 0 0.0 83.1 Pimarate + Communate 0.0 0.0 O 0 54 2 ** 45.8 Sandaracopimarate 0.0 220 4 9 39 2 ** 53.9 Levopimarate + Palustrate 2.7 8.3 7.9 58.8 ** 22.4 Isopimarate 2.9 16.0 3.9 42.2 ** 35.0 Unknown 36 4.2 11.8 0.0 42.0 ** 42.0 Strobate + Dehydroabietate 0.0 5.6 1.1 10.9 82.5 Unknown 38 0.0 7.4 0.0 0.0 92.6 Abietate 7.1 18.8 4.9 29.9 ** 39.3 Neoabietate 6.2 13.5 0.0 56.7 ** 23.6 Unknown 41 0.6 6.4 29.2 ** 0.0 63.8 Total Resin Acids 0.0 1.6 4 3 56.0.** 38.1 ** = F—test significant at P < 1% with seed source 0' a: = component Of variance due to plantation site oi = component Of variance due to block 0: = component of variance due to seed source 035 = component of variance due to interaction of plantation 2 e = component Of variance due to error 46 Figure 5. Genotype X environment interactions of neoabietic acid in Scotch pine 47 Neoabietic Acid Percent of Oleoresin a '1 C 7 — E _ C C ‘ N. E. Siberia 5 — 0 mongolica rhodopaea 4 .— \ o , o ‘ uraZenszs 0 ibemfica :3 - I ’ carpatica ° c iZZyrica E — ° aquitcma 1 I l Poor Better Quality of Environment 48 water potential and resin canal anatomy - The experimental design for the analysis of these traits is given in Table 11. As there was only one replication sampled per plantation, the plantation X seedlot interaction term was used as the error term for plantations. Table 12 shows the mean values for each trait as they varied between varieties and plantations. Water potential ranged from -1.8 to -13.1 bars in the Rose Lake planta- tion and from -1.4 to -11.9 at Higgins Lake with mean values of -5.1 and -5.6 respectively. No variation was accounted for by either seed source or plantation, but the interaction term accounted for 38.4 percent (Table 11). Resin canal area did vary between plantations. The Rose Lake trees had canals that averaged 1.7 microns2 in area and the H1991115 Lake trees averaged 1.5 micronz. The interaction term was significant, accounting for 8.4 percent. This was large enough to mask any seedlot differences within plantations. Resin canal density averaged 1.5 canals per millimeter over all seedlots and both plantations. A very high inter- action term accounted for 77.6 percent of the total variation. Discussion Many environmental factors may be responsible for causing the great variation of traits between planting sites. Campbell and Sorensen (1978) reported interaction for bud burst and bud set phenology, height, diameter, and dry weight of forty populations of Douglas-fir (Pseudotsuga menziesii (Mirb.)Franco) planted in eight locations. They attributed air and soil temperatures and soil fertility as several causes of the interactions, coupled with the change of photoperiods between the source locations and planting sites. Other environmental factors are probably involved with interaction, but they are not always easily determined. Water stress on the Higgins Lake site may be the major one for water potential, resin 49 Table 11. Degrees of freedom, variance components, and the percent of total variance in each source of variation for water potential and resin canal density and area measured on 27 seed sources of Scotch pine planted at two locations in Michigan. Resin Resin Water Canal Canal D.F. E.M.S. Potential Area Density Plantation (p-l) o: + rags + rso: 0.0 4.7 * 5.0 2 2 2 Seed source (s-l) oe + rapS + rsop 0.0 0.0 0.0 PlantationX 2 2 seed source (s-1)(p—l) Ce + ropS 38.4 ** 8.4 * 77.6 ** Within plot (r-l)ps o: 61.6 86.9 17.4 **, * = F-test significant at p < 1% and p < 5% respectively 0: = component of variance due to plantation site a: = component of variance due to seed source 0:5 = component of variance due to interaction of plantation Wlth seed source a: = component of variance due to random variation between plantations and seed sources 50 Table 12. Variation of several physiological and anatomical traits between varieties of Scotch pine grown at two locations in Michigan. Water Potential Resin CanalZArea Resin Canal Density Ajfiarslg_ (micron ) Ajcanals/mnl Rose Higgins Rose Higgins Rose Higgins Lake Lake Lake Lake Lake Lake Western and Southern Europe varieties -3.99 —6.02 1.77 1.49 1.50 1.70 armena -3.92 -7.41 1.76 1.80 1.35 2.14 aquitana -3.64 -3.39 1.63 1.42 1.41 0.99 iberica -3.57 -6.14 1.78 1.40 1.72 2.00 rhodopaea -6.48 —6.39 2.08 l 18 2.35 1.32 Central Europe varieties -5.75 -5.54 1.55 1.49 1 29 1.78 carpatica -7.22 -3.61 1.52 1.58 0.79 1.65 hercyntca -5.14 -8.24 1.35 1.44 1.37 1.56 haguensis -5.79 -4.55 1.65 1.49 1.07 1.89 Scandanavian varieties -6.99 —4.89 1.47 1.55 1.45 1.60 septentrionalis -6.99 -4.89 1.47 1.55 1.45 1.60 Russian and Siberian varieties -4.66 -5.03 1.91 1.41 1.43 1.16 mongolica -2.87 -5.13 1.55 2.35 0.89 1.79 'N. E. Siberia' -6.83 ~5.52 1.92 0.65 1.86 0.36 uralensis -4.83 —4.74 2.03 1.47 1.59 1.35 51 canal area and density, as it has a direct role in controlling growth processes in plants. Both plantations have been attacked by several species of insects, including European pine sawfly, but it is unknown what effect this would have on the Oleoresin chemistry. Whatever the reason, environmental variation does have a major influence on these traits. This fact must be considered if resin acids or anatomical fea- tures are to be used as marker traits for indirect selection. Chapter 4 An Introduction to the Use of Indirect Selection for Pest Resistance in Conifers Indirect Se Zection Indirect selection is the choosing of individuals or populations for a trait other than the one for which improvement is sought. The trait being used for selection, the secondary trait, must be highly correlated with the desired or primary trait. Indirect selection is especially useful if greater or more rapid success can be had by selecting for a correlated trait instead of a primary character. The theoretical back- ground for indirect selection has been outlined by Falconer (1960). Weissenberg (1976) has described the theory, advantages, disadvantages, and use of indirect selection in breeding for various traits in forest tree species. The mathematical basis for direct selection is illustrated in Figure 6. With direct selection, a number of individuals from a population are chosen to be the parental stock for a future generation. The intensity of selection is defined as the ratio of the mean phenotypic value of the individuals chosen, S, to the phenotypic standard deviation, op, i = S/op Heritability, hZ, is defined as the ratio of the mean phenotypic value of the progeny of selected parents, R, to the mean phenotypic value of the parents. Heritability is also the ratio of additive genetic variance to total phenotypic variance of the population. 52 53 Figure 6. The theoretical basis for direct Selection 54 Original Population Distribution Selected Parents‘ . EDisfribuhon s I Pro en Disfgribuyfion i I i A _R~_. i =Selection Intensity = S/a-p h2 =Heritabillty = R/s = O’g/a'g R = hzs R /ch = 112 S/cp R = “16° Rx: ihx"'¢:u< 55 h2 = R/S = 02/02 a P The expected gain or change from the mean of the original population, R, is the product of the intensity of selection times the square root of the heritability of the trait times the square root of the additive gene- tic variance of the selected trait. R = ihoa By using indirect selection, a secondary trait is chosen which is highly correlated with the primary trait. The expected gain by this method is calculated by multiplying the expected gain of the secondary trait and a value derived from the regression of the two traits, baxy (Falconer, 1960). This value is the product of the additive genetic correlation coefficient times the square root of the ratio of the addi- tive genetic variance for the primary trait and that of the secondary trait. baxy = Ir‘aOaX/an The gain for the primary trait, then, is the product of the intensity of selection for the secondary trait, the additive genetive correlation coefficient, and the square root of the heritability of the secondary trait times the additive genetic variance of the primary trait. GX = iyhyraoaX The relative efficiency of using indirect instead of direct selection is shown by a ratio of the expected gain of indirect selection to that expected with the direct method. This ratio is the correlation coeffi- cient times the selection intensity and square root of heritability of the secondary trait divided by the selection intensity and square root of heritability of the primary trait. e 1' h _X , rau RX lxhx Indirect selection can be a superior method to direct selection only if: 1) a high genetic correlation exists between the two traits, 2) the secondary trait has a higher heritability than the desired character, 3) a higher selection intensity can be applied on the secondary trait than on the primary trait. There are several situations in which indirect selection would be more useful than direct selection: 1) the errors of measuring the primary trait are so large as to reduce the predicted gain from direct selection 2) the primary trait can only be measured in one sex, while the secondary trait is measureable in both sexes 3) larger numbers of individuals can be measured more quickly or less expensively than the primary trait, thus raising the selection intensity of the secondary trait (an example of this is the use of routine biochemical analysis marker traits, e.g. Weissenberg, 1976) 4) generations of selection can be completed more quickly than by direct selection 5) the secondary trait is actually a component of the primary, complex trait 6) a secondary juvenile trait can be measured earlier in the life of a long-lived species. Weissenberg (1976) pointed out several problems that are involved with indirect selection. Marker traits may be difficult to find and unless they are determined as offshoots from concurrent studies, they 57 may be too expensive to search out. In addition, genetic correlations and estimates of heritabilities may be very inaccurate. If long-range breeding programs are used with several generations, the genetic correla- tions will decrease, unless pleiotropy is involved. Pleiotropy exists when one gene or several closely linked genes influence different traits. Heritabilities will also change with each generation. Finally, while several traits may be involved in a breeding program, indirect selection can best be used for improving only one trait. Indirect Selection of Pest Resistance in Conifers Indirect selection for pest resistance may be applicable on a wide range of plant species. Weissenberg (1973) has suggested that it may be useful to select for rust resistance in wheat, root-knot nematode resistance in tobacco, insect resistance in cotton, and in many other plant-pest systems. This method has application in breeding for disease resistance in forest tree species. Schfitt (1964) demonstrated that the rate of mycelial growth of Lophodermium pinastri on agar containing sap from needles of Scotch pine differed proportionally to the degree of resistance shown by the tree. Resistance to the rust fungus, Cronartium fusiforme, is in- dicated by the concentration of B-phellandrene, a monoterpene, in the cortical tissue of pines. High levels of B-phellandrene denote resis- tance in Pinus elliottii (Rockwood, 1974) while low levels do so in P. taeda (Rockwood, 1973). Concentrations of an l8-C fatty acid with two double bonds, found in the shoots of loblolly pine seedlings, also indi- cate resistance to fusiform rust (Weissenberg, 1973). There is a large potential of traits which could be used to select 58 for insect resistance in trees. For example, the phenolic contents of foliage in Picea abies increase‘ with resistance to the eastern spruce gall aphid, Adelges abietis (Tjia, 1973). Glaucous needles of ponderosa pine are resistant to a resin midge, Retinodiplosis sp., while glabrous or viscous needles are attractive (Austin et al, 1945). Many other correlative systems exist as possible bases for indirect selection. Oleoresin of coniferous trees is the most studied physiological system in tree species that affects insect resistance. Each tree species, variety, and individual tree has a varying chemical composition in its several tissues and each insect species may be attracted or repelled by qualitative or quantitative differences in concentrations of these chemicals. Many insect species have evolved mechanisms of overcoming the chemicals or of using them to their own advantage. This system should certainly be studied in any coniferous tree species to determine the role that the chemical components of oleoresin may have in insect resis- tance. Chapter 5 Application of Oleoresin Physiology of Scotch Pine in an Indirect Selection Program for Insect Resistance Abstract The possible relationships of resistance in Scotch pine to five insect pests and the resin acids of the cortical oleoresin, water poten- tial, oleoresin pressure, resin canal area and resin canal density were examined. Simple correlations and multiple regression analyses were used to identify the traits most related to resistance. Only the resin acids showed any association with insect resistance. Regression models were produced to predict the expected percentage of trees attacked by each insect at various levels of probability. Although these models may work well in a localized area, the presence of genotype X environment inter- action in the resin acids limits their usefulness. Recommendations for future work are given. 59 60 Introduction Both insects and forest tree species have been evolving for hundreds of millions of years. Fossil records indicate that the earliest angio- sperms occurred 100 million years ago, during the Cretaceous Period, while the Coniferales originated during the Carboniferous Period, about 300 million years ago. Insects existed at the time the first conifers evolved and probably had for the previous 200 or 300 million years. During geo- logic time, forest insect pests have adapted to the physiology and pheno- logy of their hosts, feeding upon specific tissues. Trees have developed chemical and physical methods to repel insect pests or to minimize their damage. The degree of complexity which has been achieved by these two co-evolving organisms is illustrated with ponderosa pine (Pinus pondero- sa Laws) and the black pineleaf scale (Nuculaspis califbrnica Coleman). Edmunds and Alstad (1978) have reported their observations of this system. Scale-free trees remain uninfested over several years even though infested branches from nearby trees provide a source of inoculum. Grafted twigs on infested trees retain the high or low levels of insects that occurred on the donor trees. Using inter- and intra-tree transfers of scale pop- ulations, they presented evidence for three hypotheses: 1) individual trees vary in their phenotypic resistance to the insects 2) as the scale population develops over several years, it becomes adapted to the defensive mechanisms of the individual tree 3) natural selection of insects for fitness on one tree hampers the development of fitness on other trees. In essence, the scale develops differentiated demes which are specific for their host trees. 61 It appears that individual trees vary in their ability to resist insect pests. The varying chemical composition of plants gives them in- dividuality. There are two functional classes of plant compounds which are defensive against insects (Feeny, 1976). The first is a group which are toxic in low concentrations, but which may be countered by coevolving insects. These are probably most common in unapparent plants, that is, plants which grow, reproduce, and die very rapidly, such as annuals. The second group of compounds form complexes with proteins, thus reducing the insect's ability to digest the plant. Chemicals which have been shown to have the latter effect include tannins, alkaloids, nonprotein amino acids, glycosides, fatty acids, and many others (Rhoades and Gates, 1976). The tannins in oak leaves have been shown to cause reductions in larval growth rates and pupal weights of the winter moth, Operoptera brumata L. (Feeny, 1968). This system exemplifies the coevolutionary process that an insect may take to overcome a defensive compound. The tannin content is lowest in new, developing leaves and increases steadily to five percent of the dry leaf weight by the end of September (Feeny and Bostock, 1968). The moth has apparently altered its timing of egg hatch so that the larvae feeds as early as possible after bud break (Feeny, 1970). A high mortality of larvae occurs each year as eclosion occurs prior to bud burst, and the larvae cannot feed on closed buds. Those oak trees which break bud the earliest in the spring are the most heavily damaged by the insect. One can envision the gradual changing of egg hatch phenology as early—bud burst trees are eventually removed from the popu- lation of oaks and only the latest emerging larvae can feed and survive to reproduce the population of insects. The Pinaceae is the largest family within the order Coniferales. The 62 most important genus of this family is Pinus. Each species of Pinus has many insect pests which cause damage ranging from insignificant or moder- ate to severe or lethal. As members of this genus have been subjected to insect attacks for hundreds of millions of years, it is reasonable to assume that chemical defenses have evolved to permit the genus to survive. One of the most unique chemical systems in pines is the oleoresin system. It is also important in members of Picea, Pseudotsuga, and Larix which also develop normal resin canal systems. The chemical composition of the oleoresin has been associated with causing resistance (Hanover, 1975; Edmunds and Alstad, 1978; Bordasch and Berryman, 1977) or attractance (Berryman, 1976; Hughes, 1974; Selander et al, 1974) to various insect pests. The purpose of this paper is to compare the known genetic varia- tion of resistance to insect pests in Scotch pine with the reported varia- tion in the resin acids and associated physiological traits. Methods and Materials The collection of oleoresin chemical, physical, and anatomical data has been described in earlier chapters. Only data collected at the Rose Lake plantation were used for comparisons. Insect susceptibility, as percent of trees attacked by an insect, were recorded on several pro- venance plantations throughout Michigan. European pine sawfly attacks were measured on plantations in Kellogg Forest in Kalamazoo County and the Rose Lake plantation over a three year period (1963-1965). The eastern pineshoot borer damage was measured at the Rose Lake and Allegan County plantations; white-pine weevil at the Higgins Lake plantation; Zimmerman pine moth (Dioryctria zimmermani Grote) at the Fred Russ Forest in Cass County, and the pine root collar weevil (Hylobius radicis Buch) 63 in Newaygo County. Results of these studies have already been published by Wright et al (1967), Steiner (1974), Wright et al (1976), Wright et al (1975), and Wright and Wilson (1972). Simple correlations and multiple regressions were calculated for each insect and resin acid combination. Only simple correlations with insect resistance were made for the physical and anatomical traits. All calculations were run on a CDC 6500 computer using the MSU STAT statistical system. Results Correlations - Simple correlations of chemical, physical and anatomical traits of oleoresin physiology with susceptibility patterns of five insect pests are given in Table 13. A negative value indicates that as suscep- tibility increases (resistance decreases), the variable decreases. Each insect species showed strong negative correlations with two to six resin acid values. Data from-two plantations were available for the European pine sawfly and the pineshoot borer, and the pattern of resin acids which correlate with resistance were very similar for both plantations. This is due to the strong genetic control of resistance at the varietal level which is little affected by environmental differences. Resistance to the sawfly, the Zimmerman pine moth and the pine root-collar weevil showed negative correlations with pimarate, levopimarate + palustrate, iso- pimarate, abietate, neoabietate, and total resin acids. The similarity of patterns between these insects was unexpected as the most resistant sources of Scotch pine to the sawfly are in the Scandanavian varieties, while the southern Europe varieties are most resistant to the pine root collar weevil and the Zimmerman pine moth. It should be expected that each insect would show a unique pattern as each is genetically indepen- dent from the others and is interacting with Scotch pine independently. 54 Resistance to the pine shoot borer showed strong correlations with sand— aracopimarate, isopimarate, Unknown 36, abietate and neoabietate. West- ern and southern Europe varieties of Scotch pine are most susceptible to the ‘insect, while Scandanavian varieties have the highest levels of re- sistance. The white—pine weevil, which has no recognized patterns in Scotch pine, was correlated negatively with Unknown 31 and positively with isopimarate. Resistance patterns showed no correlations with water potential, oleoresin pressure, resin canal area or resin canal density. It is unlikely that these traits are involved with any mechanisms causing resistance. Multiple Regression - Backward elimination of all resin acids and total resin acids was run against resistance patterns for each insect species. The physiological and anatomical data was not placed into the regression as no strong correlations were found. A variable was eliminated from the model if the significance of the partial F values for the variable dropped below 0.25, 0.15, or 0.10. The partial F values were calculated with the following equation: _ R(xi/x0,x],...xi_],xi+1,...xn) 2 s where the numerator is the amount of variation due to regression of all variables minus that amount due to all variables except x1. 52 is the error mean square for a complete model with fourteen variables. The models which were produced by the regression analysis for each insect species and each level of testing are shown in Table 14. As the required level of significance to remove a variable decreased, the number of variables in the final equations decreased. The coefficients of mul- tiple determination, r2, which describe that proportion of total variability 65 &_ v a r« Nm v a * mo. o—. mp. Po. we. we. we. we. Po. m—. mm. me. ea. am. we. mm. no. mm. mo.u —~.u MN.| mm. po.u $0.: F_.u ¢N.: mm. No.1 mo. vp.n mo. mo. MN. mo. mo.| no.1 —o. mm.: a om.| «s mm.: mo. m_.i NN.: ¥« NF.| m—.u N~.| op. PF. NF.I m_.| NP.| op. pp. r Fm.u * mm.n mow: PN.| « ov.| rs mp.u ¢F.n o—.: mo. we. ¢—.: ON.| mo.u o—.n ¢~.| o—.u om.u ON.| « mm.u « Nm.| m~.: r om.n rs o¢. mm.| ¥¥ Fm.u s* s om.n * om.u No. . NF.| om.n «* FN.I MN.| mo. «m.: s mm.n « ¢M.I ¥¥ mm.n —o.: «N.I NN.I ¥« FN.I w—.: «« m¢.v e0. ¢—. m—.u cm.n Np.u F0. mo.| mm. N—. No.1 no.1 no. _..P>603 Lm—FOU SHOE ®C_.n_ _._.>w¢3 wv—Gx— wmom cwmw—F< mwv—MA wwom IHOOK mzwm :QELwEEwN wcwmlwwwzwé LwLom POOLW 0:?”— m_.- >u_m:mo chcu :_mo¢ m—.- m—.: mo. ** mm.u mo. mo. ** mm.: we. m—.u emg< _o:mu :wmmm mesmmmea :wmmeowpo _e_u:mpoa Lopez mewo< cwmmm quoe Fe czocxzz muaum_nwomz mumpwwn< mm egocxcz mpmpmwnwogezsma + mammogum mm czocxcz mpmemeeaomH mueaumzpma + mpcecewqo>m4 maeeeewaoomgmucmm mum::EEou + opmeaewa Fm czocxzz om czocxcz mm czocxcs .mumma pomm:_ ween cupoum Fmgo>mm to mcemupmq Auwywnwpamomzm saw; amo—o_m>;a :_mm;om~o to muwueu Pmo_soum:m ecu ~mowmxga .Fmo_Em;u to mco_uc—mggoo mpaewm .mp m_acp 66 in the dependent variable being explained by all of the independent var- iables left in the model, also decreased as the numbers of variables be- came smaller. The highest r2 values were obtained in the sawfly and pineshoot borer models, indicating that the resin acids may be used to identify resistance for these insects with greater success than the’other pests. As the models became more demanding on the variables for significance, the most effective variables became apparent. Total resin acids was in the final models for the sawfly, the pineshoot borer, and the pine root collar weevil. Isopimaric acid was the most significant variable in the final equations for both the white-pine weevil and Zimmerman pine moth. Multiple regression models are of the form: E(y) = 81x1 + 82x2 + .. 8an where xi are values of independent traits being placed into the equation (in this case, they are concentrations of individual resin acids), 8i are the coefficients of the equation (given in Table 14) and E(y) is the expected value of the dependent trait (insect resistance). Resin acid concentrations from varieties armena and septentrionalis (Tables 2,3) were used to demonstrate the models. The expected percent of trees attacked for each variety is given in Table 15. Generally, the models which re- quired significance at fifteen percent gave the most reasonable results. With these models, one can determine the probability of an attack by any of these insects on a tree growing in the Rose Lake area after analyzing the resin acid composition of the tree's oleoresin. Discussion The data which have been presented seem to indicate that the resin 67 es 653 me.um :.mwx —wao» _p.—- ow.we mm. ow. me. an. o—.o mm. aw. mo. mm. m—.o mm. ow. we. mm. m~.o pw>w— oozeu,e_:o_w aw...uwam ecu an m_sc_=o—mo acz o_.o Nn. mp. Np. m_.o pm. cm. mo. mw.o sue: m:_e =mEcmEE.~ n_. No. mm. c_.o mp. No. em. mp.o .~. - av. - o_. mo. we. m~.o __>6a: 6=.m-sumem cw. - mm. - so. .o. No. o~.o ON. x mm. 1 —o. —o. _o. m5. m..o cs. - ea. - om. - mm. <0. cm. m~.o Auxoe wmoxv cocoa aoozmw:.m :Lmummu ow. eo. NN. c_.o n~.~ ~e.~ o~.~ w—.~ Km. m~. an. m—.o m~.n me.m mm.m wo.m am. mw. mm. m~.o waee omozv a—ezmm m:.a :mwacc=u nmwm...1mwm...| .n IIAmWI,iume11 ammwm .Imfllaui. In ..wmmamwi usme_e -:=EEou czozxcs excess: crocxca .cou be wocmo N 09 mucmu lounge + was -,e.:a—m -.e.:u—m -c=~m -cme_; 69 Table 15. Results of multiple regression analysis models giving expected percent of trees attacked in two varieties for five insect pests. Model Significance armena septentrionalis European Pine Sawfly .25 99 24 0.15 92 24 0.10 60 40 (Observed) (35) (14) Eastern Pineshoot Borer 0.25 -3 0.15 47 23 0.10 52 22 (Observed) (51) .(21) White-pine Weevil 0.25 19 36 0.15 21 25 0.10 21 25 (Observed) (25) (30) Zimmerman Pine Moth 0.25 13 2 0.15 20 14 (Observed) (37) (30) Pine Root-collar Weevil 0.25 49 0.15 30 49 0.10 30 49 (Observed) (12) (38) 7O acid composition of the oleoresin of Scotch pine may be involved in mech- anisms of resistance to insect pests. Seasonal variation in the quantity of the resin acids exists and parallels the larval feeding peaks of the European pine sawfly and the eastern pineshoot borer (Chapter 2). Also, seed sources of Scotch pine which are the most susceptible to these insects have the lowest concentrations of resin acids (Chapter 1). The many nega- tive correlations of insect susceptibility With resin acids (Table 13) support the statement that resistance is greatest when the resin acid composition of the oleoresin is highest. Statistical models may be de- veloped (Table 14) to predict levels of resistance to various insect species using the concentrations of these acids. Resistance to insect pests in Scotch pine is under genetic control and patterns of resistance are consistent over a variety of planting sites. The concentrations of resin acids, however, are affected by the environ- ment in which the trees are growing (Chapter 3). In the Rose Lake plan- tation, Scandanavian varieties had the highest concentrations of total resin acids and the central European varieties had the lowest concentra- tions of pimaric, abietic, neoabietic, and total resin acids. The Higgins Lake study showed that the Western and Southern Europe varieties had the highest concentrations while the Central Europe and the Russian and Sib- erian varieties had the lowest. The correlations with insect resistance are only phenotypic, not genetic correlations. Obviously, mathematical models for insect resistance based upon the Rose Lake study will not be of use for trees grown in the Higgins Lake area. It is not known to what extent the Rose Lake data may be applied. The cortical tissue of boles of Scotch pine trees was used as a source of oleoresin. The insects which attack this species, however, feed on a 71 variety of tissues. The European pine sawfly feeds on foliage; the white- pine weevil and pineshoot borer tunnel through terminal shoots, one only on the leader at the top of trees and the second only on lateral branches; the Zimmerman pine moth and pine root-collar weevil feed on cambial tissue at various parts of the tree. Previous studies have shown that resin acids vary in concentrations between different tissues within a tree (Zinkel, 1976; Westfall, 1972). Any resin acid variation that may exist among the tissues of Scotch pine will have to be identified before an accurate model of resistance to any of these insects can be made. The existance of both environmental and within-tree variation for each resin acid needs to be quantified before any general hypothesis of resin acid interaction with insect pests can be accepted. Recommendations for Future Work Genotype X Environment Interaction - The existance of an interaction in the resin acids between seed source and plantation site has not previously been reported. It has generally been assumed that none of the monoterpenes and probably higher terpenes are subject to environmental variation. Hodges and Lorio (1975) have shown that moisture stress can alter the ter- penoid fractions of oleoresin in Pinus taeda, but did not report any varia- tion of the levels of stress due to the genotype of the trees. The extent of interaction should be quantified. Within Michigan there are at least sixteen provenance tests of Scotch pine which may be sampled to determine the geographic distances within which interaction is minimal. It is pos- sible that each plantation may have a unique pattern of resin acid varia- tion. A more effective approach to the study of interaction would be to establish clonal plantations of several seedlots over a range of 72 environments. Observances of climatic and edaphic variation may provide some explanation of the causes of this interaction. Genetics of Resin Acids - No work has yet been done with the modes of inheritance of resin acids. The monoterpenes are apparently under simple gene control (Hanover, 1966b, 1971) and can be readily manipulated within a population with properly planned breeding programs. Expected gains in changing the concentrations of resin acids by controlled crosses are un- known. If these compounds are under multiple gene control, being qualita- tive rather than quantitative traits, they may be of no use in an insect resistance program. Half-sib family analysis, full sib families with controlled pollinations, and multiple clonal plantations as previously described, need to be made in order to determine inheritance mechanisms. Seasonal Variation - Seasonal fluctuations in the resin acid composition should be examined for other varieties of Scotch pine, at several planting sites, and for other pine species. It was observed that one seedlot of variety septentrionalis planted at Rose Lake had seasonal variation, but it is not known if this variation would be consistent among other varieties. Also, the timing of variation may not be consistent in different environ- ments. It is possible that phenological differences between the Rose Lake and Higgins Lake plantations are responsible for the great amount of genotype X environment interaction that was observed. Perhaps if samples are taken in the late fall or early winter months, less apparent interaction will be observed. Other pine species should be examined to determine if the variation has a physiological basis. European Pine Sawfly - Tree resistance studies of this insect pest should be continued for several reasons. It is a major pest in Michigan, it has demonstrated genetic control for host selection, and it is 73 available in large numbers for easy collection. Another advantage to this insect is that the female lays her_eggs in patches on the same needle. When the larvae hatch and begin to feed, all of the larvae in the colony are part of a half-sib family. If collections are made soon after eclosion and prior to larval movement within the trees, genetically similar groups of larvae may be isolated. This fact can be used to determine if the insect is genetically adapted to the tree upon which it is feeding. Mem- bers of colonies may be placed on different trees and observed for sur- vival, growth rates, fecal resin acid composition, and so on. Demonstrated genetic adaptation for specific hosts could alter theories of colonial development and population expansion. The chemical composition of the needles of Scotch pine requires greater study for sawfly resistance. Ikeda et al (1977) have reported that a fatty acid, l3-keto-8(l4)-podocarpen-l8-oic acid, and dehydro- abietic acid in the foliage of Jack pine repel larvae of Neodiprion rugifrons and N. swainei. These may also be present in the needles of Scotch pine and may be less affected by environmental variation. SUMMARY, The objectives of this study were (1) to examine the genetic varia- tion of several components of oleoresin physiology in Scotch pine, (2) determine the effect of environment of these components, and (3) to com- pare each component with observed resistance patterns of several species of insects which attack the species. Genetic variation among regions and varieties were observed for nine resin acids, oleoresin pressure, water potential and resin canal cross-sectional areas. No variation was observed in the density of resin canals. The plantation site did affect each of these traits. Genotype X environment interaction accounted for 56% of the variation of total resin acids, 38.4% for water potential, 8.4% of resin canal area variation, and 77.6% of resin canal density variation. On trees studied at the Rose Lake plantation, many of the resin acids showed positive correlations with resistance to five insect pests. No physical or anatomical trait showed such correlations. Multiple regres- sion models were develOped for each insect to predict susceptibility of a tree given its resin acid composition. Such models, however, are de- pendent upon the plantation site due to the presence of environmental variation. If any of these traits are to be used as markers for indirect selection of insect resistance, they will only be useful within a limited planting area. 74 BIBLIOGRAPHY BIBLIOGRAPHY Austin, L., 0.5. Yeill, and K. Brecheen. 1945. Use of shoot character- istics in selecting ponderosa pines resistant to resin midge. Ecology 26:288-296. Berryman, A.A. '1976. Theoretical explanation of mountain pine beetle dynamics in lodgepole pine forests. Environ Ento 5:1225-1233. Bordasch, R.P. and A.A. Berryman. 1977. Host resistance to the fir engraver beetle scolytus ventralis (Coleoptera: Scolytidae) II. Repellency of Abies grandis resins and some monoterpenes. Can Ent 109:95-100. \ Bridgen, M.R., J.W. Hanover, and N. Noecker. 1979. Varietal identi- fication of eight year old Scotch pine trees by monoterpene analysis. Amer Christmas Tree Jour 23:33-34. Bridgen, M. R, J. W. Hanover, and R. C. Wilkinson. 1979. Oleoresin char- acteristics of eastern white pine seed sources and relationship to weevil resistance. For Sci 25: 175- 183. Campbell, R.K. and F.C..Sorensen. 1978. Effect of test environment on expression of clines and on delimitation of seed zones in Douglas- fir. Theor Appl Genet 51:233-246. Edmunds, G.E. Jr. and D.N. Alstad. 1978. Coevolution in insect herbi- vores and conifers. Scale insects adapt to intraspecific variation in pine defenses with a differentiated deme on each tree. Science 199:941-945. Falconer, 0.5. 1960. Introduction to quantitative genetics. The Ronald Press Company, New York. 365p. Feeny, P.P. 1968. Effect of oak leaf tannins on larval growth of the winter moth Operophtera brumata. Jour Insect Physiol 14:805-817. Feeny, P.P. 1970. Seasonal changes in oak leaf tannins and nutrients as a cause of spring feeding by winter moth caterpillars. Ecology 51:565-581. Feeny, P.P. 1976. Plant apparency and chemical defense. In Recent Advances in Phytochemistry Volume 10:1-40. Feeny, P.P. and H. Bostock. 1968. Seasonal changes in the tannin con- tent of oak leaves. Phytochemistry 7:871-880. 75 76 Finnerty, W.R. and P.K. Falco. 1977. Seasonal variation of resin acids and other metabolites in Pinus elliottii. In Proc Lightwood Res Coordinating Council Ann Mtg pp 112-120. Fowells, H.A. 1965. Silvics of forest trees of the United States. U.S.D.A. Agric Handbook 271, 762 p. Franklin, E.C. 1976. Within-tree variation of monoterpene composition and yield in slash pine clones and families. For Sci 22:185-191. Gerhold, H.D. 1959. Seasonal variation of chloroplast pigments and nutrient elements in the needles of geographic races of Scotch pine. Silvae Genet 8:113-123. Gilmore, A.R. 1975. Composition of monoterpenes in loblolly pine at different stem heights. Trans Illinois State Acad Sci 68:26-28. Hanover, J.W. 1966 a. Environmental variation in the monoterpenes of Pinus monticola Dougl. Phytochem 5:713-717. Hanover, J.W. 1966 b. Genetics of terpenes I. Gene control of monoter: pene levels in Pinus monticola Dougl. Heredity 21:73-84. Hanover, J.W. 197l. Genetics of terpenes II. Genetic variances and interrelationships of monoterpene concentrations in Pinus monti- cola. Heredity 27:237-245. Hanover, J.W. 1975. Physiology of tree resistance to insects. Ann Rev Ento 20:75-95. Hodges, J.D. and P.L. Lorio, Jr. 1968. Measurement of oleoresin exuda- tion pressure in loblolly pine. For Sci 14:75-76. Hodges, J.D. and P.L. Loria, Jr. 1975. Moisture stress and composition of xylem oleoresin in loblolly pine. For Sci 21:283-290. Hrutfiord, B.F., S.M. Hopley, and R.I. Gara. 1974. Monoterpenes in Sitka spruce: within tree and seasonal variation. Phytochem 13:2167-2170. Hughes, P.R. 1974. Myrcene: a precurser of pheromones in Ips beetles. J Insect Physiol 20:1271—1275. Ikeda, T., F. Matsumura, and D.M. Benjamin. 1977. Chemical basis for feeding adaptation of pine sawflies Neodiprion rugifrons and Neodi- prion swainei. Science 197:497-499. King, J.P. 1965 a. Seed source X environment interactions in Scotch pine. I. Height growth. Silvae Genet 14:105-115. King, J.P. 1965 b. Seed source X environment interactions in Scotch pine. II. Needle length and color. Silvae Genet 14:141-148. 77 Mirov, N.T. 1961. Composition of gum turpentines of pines. U.S.D.A. Tech Bull No. 1239. 158p. Rhoades, D.F. and R.G. Cates. 1976. Toward a general theory of plant antiherbivore chemistry. In Recent Advances of Phytochemistry 10:168-213. Rockwood, D.L. 1973. Monoterpene-fusiforme rust relationships in loblolly pine. Phytopathology 63:551-553. Rockwood, D.L. '1974. Cortical monoterpene and fusiforme rust relation- ships in slash pine. Phytopathology 64:976-979. Rudloff, E. von. 1975 a. Seasonal variation in the terpenes of the foli- age of black spruce. Phytochemistry 14:1695-1699. Rudloff, E. von, 1975 b. Seasonal variation of the terpenes of the leaves, buds, and twigs of blue spruce (Picea pungens). Can J Bot 53:2978-2982. Scholander, P.F., H.T. Hammel, E.D. Bradstreet and E.A. Hemmingsen. 1965. Sap pressure in vascular plants. Science 148:339—346. Schrum, G.M., H.D. Gerhold, R.F. West, and L.S. Hamilton. 1975. Genetic variances of Scotch pine: environment and age effects. For Sci 21:330-339. Schfitt, P. 1964. Eine mykologische Methode der Resistenzprfifung, entwickelt fur den Lophodermium - Befall der Kiefer. Phyto- pathologische Zeitschrift 51:62-72. (English summary). Selander, J.,.P,=Kalo, E. Kangas and V. Pertunnen. 1974. Olfactory be- haviour of Hylobius abietus L. (Col., Curculionidae). I. Response to several terpenoid fractions isolated from Scots pine phloem. Ann Ent Fenn 40:108-115. Steiner, K. 1974. Genetic differences in resistance of Scotch pine to eastern pineshoot borer. Great Lakes Ento 7:103-107. Squillace, A.E., J.G.A. La Bastide, and C.L.H. van Vredenburch. 1975 Genetic variation and breeding of Scots pine in the Netherlands. For Sci 21:341-352. Tjia, B. 1973. Relation between phenol content and eastern spruce gall aphid resistance in Norway spruce. HortScience 8:279-280. Tobolski, J.J. 1968. Variations in monoterpenes in Scotch pine (Pinus sylvestris L.). Ph.D. Diss Michigan State Univ, East Lansing, MI. 84 p. (Univ Microfilm 69:5974; Diss Abstr 29135678). Tobolski, J.J. and J.W. Hanover. 1971. Genetic variation in the mono- terpenes of Scotch pine. For Sci 17:293-299. 78 Weissenberg, K. von. 1973. Indirect selection for resistance to fusi- form rust in loblolly pine.' Acta For Fenn 134:1-46. Weissenberg, K. Von. 1976. Indirect selection for improvement of de- sired traits. In Modern methods in forest genetics. J. P. Miksche, ed. Springer-Verlag. 288 p. Westfall, R.D. 1972. Developmental and genetic variation in the cortical terpenes of species of Pinus and Picea. Ph.D. Thesis Michigan State Univ. East Lansing, MI. 113p. (Univ Microfilm 72:30065; Diss Abstr 33:1877). Wright, J.W. 1963. Genetic variation among 140 half-sib Scotch pine families derived from 9 stands. Silvae Genet 12:83-89. Wright, J.W. 1976. Introduction to forest genetics. Academic Press, New York. 463 p. Wright, J.W. and H.I. Baldwin. 1957. The 1938 International Union Scotch pine provenance test in New Hampshire. Silvae Genet 6:2-14. Wright, J.W. and W.I. Bull. 1963. Geographic variation in Scotch pine. Results of a 3-year study. Silvae Genet 12:1-25. Wright, J.W., W.A. Lemmien, J.N. Bright, M.W. Day, and R.L. Sajdak. 1976. Scotch pine varieties for Christmas tree and forest planting in Michigan. M.S.U. Agric Exp Sta Res Rep 293. 16 p. Wright, J.W., S.S. Pauley, R.B. Polk, J.J. Jokela, and R.A. Read. 1966. Performance of Scotch pine varieties in the north central region. Silvae Genet 15:101-110. Wright, J.W. and L.F. Wilson. 1972. Genetic differences in Scotch pine resistance to pine root collar weevil. M.S.U. Agric Exp Sta Res Rep 159. 6 p. Wright, J.W., L.F. Wilson, and J.N. Bright. 1975. Genetic variation in resistance of Scotch pine to Zimmerman pine moth. Great Lakes Ento 8:231-236. Wright, J.W., L.F. Wilson, and W.K. Randall. 1967. Differences among Scotch pine varieties in susceptibility to European pine sawfly. For Sci 13:175-181. Zavarin, E. 1968. Chemotaxonomy of the genus Abies - II. Within tree variation of the terpenes in cortical oleoresin. Phytochemistry 7:00-107. Zinkel, D.F. 1976. Diterpene resin acids of Pinus densiflora needles and cortex. Phytochemistry 15:1073-1074. ‘ "I111111.111111111111111115