186 Genetics of Graft Rejection in Douglas-fir DONALD L. COPES1 Pacific Nol'/hll'est Forest and Rangl! E.\'periment Station, U.S.D.A. Forl!sl Sen·ice, Port/am/, Orl!gon97208 Received September 17, 1973 Accepted January 8, 1974 CoPES, D. L. 1974. Genetics of graft rejection in Douglas-fir. Can. J. For. Res. 4, 186-!92. Graft rejection in Douglas-fir was controlled primarily by additive genes. Heritability of graft incompatibility was 0.81. No significance for specific combining ability was found. A breeding program to develop highly compatible stocks should enable tree im­ provement workers to substantially reduce incompatibility losses within clonal orchards. Highly incompatible clones may have different or additional rejection factors which less incompatible clones do not exhibit. CoPES, D. L. 1974. Genetics of graft rejection in Douglas-fir. Can. J. For. Res. 4, 186-192. Le rejet des greffes chez Ps£•udotsuga menziesii fut surtout dO a !'action de genes addi­ tifs, et l'heritabilite de cette incompatibilite se chiffra a 0.81. On obtint aucun resultat significatif quant a Ia capacite specifique d'union. Un programme d'amelioration visant a developper des porte-greffes fortement compatibles devrait permettre aux ameliorateurs forestiers de diminuer les pertes dans les pares a clones, dues a cette incompatibilite. Les clones fortement incompatibles peuvent montrer certains signes de rejet differents ou ajoutes, lesquels passent inapergus chez les clones plus compatibles. [Traduit par le journal] Introduction Breeding programs to develop highly com­ patible stocks for clonal Douglas-fir (Pseudo­ tsuga menziesii [Mirb.] Franco) seed orchards are needed, but some knowledge of the type of inheritance involved in tissue rejection is re­ quired before an efficient program can be ini­ tiated. Although graft rejection in plants has been studied for hundreds of years, the under­ lying mechanisms are still poorly understood. Horticulturists have indicated that close genetic relationship between stock and scion generally increased the likelihood of compatible unions (Bregger 1948, Grasselly 1968, Kester et al. 1965). Incompatibility factors in almonds were transferred from parent to offspring (Kester 1970). In this case, both quantitative and qualitative inheritance were evident. Graft rejection mechanisms in forest conifers are even less well understood, although the transfer of incompatibility factors from parent to progeny appears to occur. Some gain in 1Present address: Forestry Sciences Laboratory, Pacific Northwest Forest and Range Experiment Sta­ tion, U.S.D.A. Forest Service, Corvallis, Oregon 97331. Can. J. For. Res., 4, 186(1974) compatibility resulted from grafting parental clones on related stocks in Pinus taeda (Lantz 1970), P. caribaea (Slee and Spidy 1970), and Pseudotsuga menziesii (Copes 1973). Additive gene action was thought to be the primary type of inheritance present in Douglas­ fir graft rejection, but one of the 12 clones stu­ died was found to deviate greatly from the expected response. The incidence of incom­ patibility was more than twice as great when grafted on related stocks as when grafted on unrelated stocks (Copes 1973). No previous compatibility study has yielded accurate estimates of additive and nonadditive gene effects nor have any estimates of herit­ ability been made. The following grafting study utilized scions from six Douglas-fir parents and their progeny to obtain the estimates for gen­ eral and specific combining ability and for the heritability of graft rejection. Methods Early in 1964, six 20- to 25-year-old Douglas-fir parent trees were selected from a native population in the western foothills of the Oregon Cascade Range. Parent trees were separated from each other by a minimum of 100 m and were selected solely for 187 COPES: GRAFT REJECTION IN DOUGLAS-FIR their ability to produce seeds. In order to determine the average incompatibility of each parent, scions from each tree were collected in March 1968 and cleft-grafted that same spring on the terminals of 7-year-old Douglas-fir stocks growing in a plantation near Corvallis, Oregon. The stocks were random, nursery-run seedlings from a low elevation Willa­ mette Valley seed source. Each parent clone was grafted on 50-66 different stocks. In September 1969, the grafts were severed from the stocks and the unions preserved in 50% alcohol. The unions were sectioned into transverse sections on a sliding micro­ tome and stained with safranin 0 and fast green (Copes 1967). Microscope preparations were made and examined for the presence (incompatible graft) or absence (compatible graft) of wound-xylem areas in the xylem tissues joining the stock and scion (Copes 1970). A percentage incompatibility value, based on the 50- to 66-graft sample, was then de­ termined for each parent tree (number incompatible unions + total number unions X 100). In April 1964, the six Douglas-fir parent trees were control-pollinated in a 6 X 6 diallel design. A Model I design, where the reciprocal crosses and self-polli­ nation combinations were omitted, was used (Griffing 1956). Fifteen diallel families resulted from this de­ sign. Wind-pollinated families were also collected from each parent tree when the control-pollinated seed was gathered in September 1964. Seedlings were grown from 1965-1967 at Corvallis, Oregon, in cold frames in a randomized block design. In the winter of 1967, they were then planted within the boundaries of the original parent stand. Fifty scions from each diallel and wind-pollinated family were collected in March 1970. Each scion was cut from a different 5-year-old seedling, so no seed­ ling genotype was represented more than once in the study. The scions were cleft grafted in April 1970 on lateral branch tips of 50 9-year-old stocks. Grafting was done by proceeding systematically down each row of stock trees. Each consecutive group of 10 stocks was regarded as a block. Each stock tree held 21 different grafts on 21 of its branch tips: one graft from each of the 15 control-pollinated families and the six wind-pollinated families. Graft technique failures were regrafted in April 1971. The stocks were all part of the same plantation where the six parents had previously been grafted. Grafts were sacrificed 18 months after grafting for microscope study. The staining techniques and incompatibility detection method used for parental grafts were also used for seedling grafts. Heritability of graft rejection was determined directly from the progeny-midparent regression. Diallel analysis of control-pollinated seedling grafts used the procedure described by Griffing as Model I, Method 4 (Griffing 1956). The model used was as follows: xiJ = J.1 + g1 + g1 + siJ + (1/b) E, LJ eiJkt where: X1J = ijth observation; = study mean; g1 and g1 = general combining ability (male and female); SIJ = specific combining ability; eukt = error; and b = number blocks. Percentage data were transformed to arcsins prior to sum of squares analysis. Results Incompatibility percentages of the six parent clones and their progenies are given in Table 1. The mean for all parents was 38.3%, only slightly higher than the 35% mean for a large number of randomly selected clones. In gen­ eral, crosses of more compatible parents gave progeny which were more likely to form com­ patible grafts than were progeny derived from crossing more incompatible parent trees. Cor­ relation between midparent and progeny was significant (r = 0.89), as was the regression of progeny on midparent ( b - 0.81). This b value is also a heritability estimate for graft rejection by the six parents under the condi­ tions of this study. In the diallel analysis, the effect of general combining ability (GCA) was highly significant; the effect of specific combin­ ing ability (SCA) was nonsignificant (Table 2). In most cases GCA ranking of parents cor­ related closely with rankings of parents by their incompatibility values. But, the actual incom­ patibility values of the seedling grafts from five of the six parents average 4.9% lower than their parents. SCA values ranged from +10.9 to -15.7% and GCA values ranged from 26.7 to 39.5% (Table 1). The three largest SCA values all came from crosses of parent 1 1. If results from this study are to be applicable to _seed orchard conditions, it is important that the stocks be similar to those grafted in the orchards. In order to determine if they were similar, the average incompatibility of each stock was calculated from the 2 1 unions grafted .on each stock. Each stock value was plotted into one of 10 incompatibility classes and a histogram drawn (Fig. 1). The histo­ gram illustrated approximately the same distri­ bution pattern encountered in numerous past tests from orchards when many clones were grafted on unselected stocks. In addition, the mean incompatibility value from this study was approximately the same as the average normally encountered in orchards with larger · ...... 00 00 TABLE 1. Estimates of percentage incompatibility (IC) for each parental clone, midparents, control, and wind-pollinated families and their general (GCA) and specific (SCA) combining abilities Parent clone numbers Average IC' of parent clone Percentage incompatibility 5 56 32 58 9 35 8 27 11 29 10 25 Mean 38.3 Parent clone numbers 5 9 8 11 10 Average of mid parents Average of wind-pollinated progeny 39.5 38.3 36.5 30.6 28.8 26.7 33.4 45.4 46.2 37.0 33.8 34.7 33.0 38.3 36 43 34 33 22 33 35.3 () > z X 44 45 34 45 30 Estimate of specific combining abilityc 5 X 32 +5.1 9 +7.0 8 -1.1 II +10.9 10 -3.1 aJc denotes incompatjbility. 32 Average of progeny with common parent (GCA)" X 41 43 40 X 37 29 23 34 X +3.6 +8.5 +6.5 -9.5 X 14 27 X 18 X :-- '11 0 :00 ;:d til :" < 0 r- ::;; -..1 X +3.4 -3.7 +2.4 """ X -15.7 -1.6 X -9.8 bGCA denotes general combining ability (average incompatibility value of the five progeny families per parent clone). <Calculated as individual cross incompatibility values-(GCA o + GCA )/2. X 189 COPES: GRAFT REJECTION IN DOUGLAS-FIR TABLE 2. Analysis of variance for graft rejection in control-pollinated families. Percentages were transformed to arcsins before analysis Source of variation Total Blocks Families General combining ability Specific combining ability Error 5 Degrees of freedom 74 4 (14) 5 9 56 Mean squares Fvalues 487.225 365.093 620.253 223.337 184.437 2.64** 1.97* 3,36** 1.2"·'· Coefficient variation 41% 25 a; Q. ::E 0 u i!!: :I: u < w i!!:"' "' "' ,.,..: u-' ou ... "' u. 0 "' "' z w u 0: w .. 0 0 90 10% 100 INCOMPATIBILITY CLASSES FIG. 1. Histogram showing the distribution of 50 stocks according to their average incom­ patibility values. samples (33.4 versus 35%, respectively). Stocks grafted in this study were typical of the type of stock normally grafted in seed orchards at this time. Preliminary inspection of the data indicated that the relative incompatibilities of stock and scion might affect their interaction with each other. Stocks, therefore, were separated for analysis into five groups of 10 trees according to their incompatible values. The incompatibil­ ity ranges within the groups were 0-19%, 20-30%, 30-38%, 38-48%, and 48-72%. Grafts on stocks in the. range 0-48% incom­ patibility exhibited similar inheritance patterns -progenies of the most compatible parents pr<'duced the most compatible grafts (Table 3, Fig. 2). But a different response was observed on stocks of the most incompatible group ( 48­ 72% ), where progenies of the most incom­ patible parents produced the more compatible grafts. Variation between blocks was found to be highly significant, but I believe this to have been merely a chance occurrence. Examina­ tions of average compatibility of stocks within blocks revealed that the fourth block was responsible for the significance noted. Two of the 10 stocks in that block were highly incom­ patible. If more stocks per block had been used, much of the block variation would have been removed. Discussion Incompatibility symptoms in Douglas-fir (Copes 1970) are identical to the 'contact type' described by Mosse in 1962. Fruit tree studies (Mosse 1962) and my own work with Douglas-fir (unpublished data) all indicate that reciprocal grafts, in species typefied by the contact type of incompatibility, yield identical compatibility tests. It makes little difference whether a clone is grafted as the stock or as the scion. Thus, compatibility values deter­ mined in this study from grafting parental CAN. J. FOR. RES. VOL. 4, 1974 190 TABLE 3. Graft incompatibility of seedling scions grafted on stocks which were I (0-19%), II (20-30%), III (30-38%), IV (38-48%), and V (48-72%) incompatible Stock group number Parent clone nu mber Percentage incompatible I 5 II III IV v Average incompatibility of progeny with common parent (GCA) 5 9 8 11 10 X 32 9 8 11 10 30 10 0 5 32 9 8 11 10 X 44 11 60 40 30 0 22 X 5 32 9 8 11 10 50 30 70 38 50 5 32 9 8 11 10 70 78 50 60 33 5 32 9 8 11 10 32 X X 10 70 30 30 33 X 10 11 20 11 X 33 56 67 25 X 40 22 38 33 X 50 70 50 22 X 71 56 33 30 X 11 0 0 X 10 0 14 X 40 30 44 X 30 30 30 X 90 80 80 clones and seedling families as scions can be used directly to approximate how those same individuals would react as stocks. It appears possible for breeding programs to develop highly compatible stocks which can be used with scions from most plus trees. With mainly additive gene effect (heritability was 0.81), considerable gains in compatibility can be made through stringent parent selection and cross-breeding among the best parents. A likely area to begin this approach is in crossing the most highly compatible orchard clones. I started a program of this type in 1971 and information on the resulting stock compatibility should be known about 1975. The most incompatible trees in the popula­ tion may have rejection factors not found in X 0 0 X 0 13 X 0 30 X 10 30 X 67 67 X 0 X 0 X 11 X 10 X 60 (Mean= 8%) X 18.7 14.3 6.1 2.0 12.5 4.3 (Mean= 26%) X 40.8 48.9 19.6 20.4 23.9 15.4 (Mean= 35%) X 47.9 37.0 36.0 32.7 22.2 34.0 (Mean = 43%) X 58.3 53.1 42.9 38,0 32.0 25.0 (Mean = 55%) X 34.7 37.8 78.7 61.7 55.3 54.2 less incompatible trees. Results from grafting on stocks with known incompatibility values suggest that rejection factors in about 80% of the stocks (0-48% incompatible) can be overcome by grafting on stocks grown from seed of crosses of moderately compatible parents. It should be remembered that none of the six diallel parents had extremely high com­ patibilities; the best clone was only 75% compatible. When the 10 most incompatible stocks were selected out of a population of 50, the incompatibility factors of the stocks may have exceeded the compatibility limits of the parents. If two 95% compatible parents had been included in the study and crossed, their progeny may have had a broad enough com­ COPES: GRAFT REJECTION IN DOUGLAS-FIR 100 90 >t: ...J iii ::. 0 u !i: ..: 0: (!) ..: u (!) Gr o up 1 stocks Group 4 Group 2 Group 3 stocks 191 Group 5 stocks stocks stocks 80 70 60 50 40 30 20 10 0 0 70 Group 1 moan mean mean mean 80 mean AVERAGE STOCK INCOMPATIBILITY (%) FIG. 2. General combining ability (GCA) of the six parents on five classes of increasingly incompatiblt< stocks (Groups 1-V). patibility spectrum to encompass the rejection factors of most Douglas-fir trees. Ortet age appeared to have an effect on average incompatibility. Grafts from 5-year­ old ortets were uniformly lower in average incompatibility than their 24- to 29-year-old midparents (33.4 versus 38.3% ). A similar case of lowered incompatibility associated with juvenility was noted for grafts of peach on plum stocks (Herrero and Tabuenca 1969). It is hypothesized that Douglas-fir tissue becomes more hypersensitive to foreign tissues in developing from the seedling to older ju­ venile or adult stages and, as a result, exhibits a decrease in overall graft compatibility. Ana­ tomical examinations of 5- to 10-year-old grafts have indicated that the compatibility of unions between old scions and young stocks do not change as the stocks and scions grow older (unpublished data). Thus, some chemical adjustment or immune response between con­ tiguous stock and scion cells may occur through the years which allows them to maintain viable cell contact even after the noncontiguous stock and scion cells become physiologically incom­ patible. At this later date, grafts made between noncontiguous stock and scion tissues develop into incompatible grafts (unpublished seed orchard data). There appears to be a difference between clones in the amount of compatibility change caused by maturation. If it can be assumed that GCA values are fair estimates of parental clone performance when young, it appears that the most incompatible clones become more incompatible with increased age than do the less incompatible clones. A comparison of difference between GCA and parent incom­ patibility values illustrates this point. Clones C-5 and C-32 were the most incompatible parents and the difference between their GCA and parent values was much greater than the differences for the other four less incompatible parents and progenies. Grafting tests of stocks grown from wind­ pollinated seeds can be used to obtain usable GCA estimates for selecting parents for breed­ ing. But in wind-pollinated progeny, it is not known if only one or two different trees or if many trees contributed their genes to a ma­ jority of the progeny. Even with this limita­ tion, the wind-pollinated GCA values were all within 7% of those obtained with diallel seed­ lings. BREGGER, J. T. 1948. Peach variety incompatibilities on seedlings of a Yunnan understock. Proc. Am. Soc. Hortic. Sci. 52, 141-142. CoPES, D. L. 1967. A simple method for detecting incom­ patibility in 2-year-old grafts of Douglas-fir. U .S.D.A. For. Serv., Pac. Northwest For. Range Exp. Stn., Portland, Oreg. Res Note PNW-70. -- 1970. Initiation and development of graft incom­ patibility symptoms in Douglas-fir. Silvae Genet. 19, 101-107. 192 CAN. J. FOR. RES. VOL. 4, !974 1973. Inheritance of graft compatibility in Douglas-fir. Bot. Gaz. 134(1), 49-52. GRASSELLY, C. 1968. The plums Prunus domestica root­ stocks for peach trees. Study and improvement of graft­ ing compatibility of their seedlings. (Fr.) Ann. Amer­ lior. Plantes 18(1), 59-73. In Bioi. Abstr. SO, 133870. GRIFFING, B. 1956. Concept of general and specific com­ bining ability in relation to diallel crossing systems. Aust. J. Bioi. Sci. 9, 463-493. HERRERO, J., and TABUENCA, M. C. 1969. Incom­ patibiJidad entre patron e injerto. X. Comportamiento de Ia combinacion melocotonero/mirobohin injertado en estado cotiledonar. (Incompatibility between root­ stock and scion. X. Behavior of the combination peach/myrobalan grafted at the cotyledonary stage.) An. Estac. Exp. Auld Dei. 10, 937-945. In Hortic. Abstr. 41, 5898. KESTER, D. E. 1970. Graft-compatibility of almond seed­ ling populations to Marianna 2624 plum rootstock. (Abstr.) Hortic. Sci., Sec. 2, 5(4), 349. KESTER, D. E., HANSEN, C. J., and PANETSOS, C. 1965, Effect of scion and interstock variety on incompatibil­ ity of almond on Marianna 2624 rootstock. Proc. Am. Soc. Hortic. Sci. 86, 169-177. LANTZ, C. W. 1970. Graft incompatibility in loblolly pine. Ph.D. thesis, N.C. State Univ. at Raleigh. Univer. Microfilms, Ann Arbor, Mich., Int. Diss. Abstr. 32(1), 16-B. MossE, B. 1962. Graft incompatibility in fruit trees. Com­ monw. Bur. Hortic. Plant. Crops, Tech. Commun. 28. SLEE, M. U., and SPIDY, T. 1970. The incidence of graft incompatibility with related stock in Pinus caribaell Mor. var. hondurensis B. et. G. Silvae Genet. 19(5-6), 184-187.