Citation
Genetic variation in symptomology of slash pine in response to fusiform rust

Material Information

Title:
Genetic variation in symptomology of slash pine in response to fusiform rust
Creator:
Layton, Patricia Adlene, 1954-
Copyright Date:
1985
Language:
English

Subjects

Subjects / Keywords:
Diseases ( jstor )
Fats ( jstor )
Forests ( jstor )
Heritability ( jstor )
Inoculation ( jstor )
Inoculum ( jstor )
Pine trees ( jstor )
Plant gall ( jstor )
Seedlings ( jstor )
Symptomatology ( jstor )
City of Springfield ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
ACY0251 ( ltuf )
13568920 ( oclc )
0021919219 ( ALEPH )

Downloads

This item has the following downloads:


Full Text












GENETIC VARIATION IN SYMPTOMOLOGY OF SLASH PINE
IN RESPONSE TO FUSIFORM RUST









BY


PATRICIA ADLENE LAYTON


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1985









This research is dedicated to my parents, James B. and

Rebecca Layton, who always "think I can." Their love and faith in me have always provided a stable base in my life. Their help was even more concrete in this study in that they gallantly volunteered to help in data collection at the Rust Screening Center.
















ACKNOWLEDGEMENTS


I am deeply grateful to W. Paul Shelley, Jr., Erin C. Shelley and other members of the Shelley family who established the W. P. Shelley, Sr., Graduate Forestry Fellowship which I received during my last two years of graduate study.

I would like to thank the members of my supervisory committee for their guidance and inspiration: Drs. Ray E. Goddard, Ramon C. Littell, Donald L. Rockwood, Robert A. Schmidt, Anthony E. Squillace and Charles J. Wilcox. Additionally I would like to thank Drs. Tom Miller, George Blakeslee, Mitch Flinchum, Ron Dinus, Nancy Pywell, Arnett M1ace, Nick Comerford, Everett Hall and Bro Kinloch for their advice and encouragement.

This project would not have been possible without the help and support of the University of Florida Cooperative Forest Genetics Research Program. I would like to thank the staff at the University, which includes Greg Powell, Harm Kok, B. J. Rabe, Charles Akins, and James Hayes, as well as the industrial members. In addition, SFRC students Penny Parnell and Robert Richens were a great help in collecting the data. Two of the industries, Brunswick









Pulp Land Company and Container Corporation of America, provided materials for this study. I am especially grateful to James Hodges and the research staff at Brunswick for their assistance in collecting the field data.

Portions of the project were conducted at the US Forest

Service's Rust Screening Center in Asheville, North Carolina. I appreciate the advice and help of the staff, especially Robert Anderson and Carol Young.

I wish to thank Martin Marietta Energy Systems for helping in the completion of this project.

I would like to thank my fellow graduate students for their encouragement and support throughout this project. I would especially like to thank Tim Meyer for his friendship and help. Lastly, I want to thank Rick Cantrell, my husband, for his love, patience and understanding.















TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS . . . iii

LIST OF TABLES . . . vi

LIST OF FIGURES . . x

ABSTRACT . . xi

INTRODUCTION . . 1

LITERATURE REVIEW . . . . 4

MATERIALS AND METHODS . . . 14

Field Tests--Experiment 1 . . .14 Artificial Inoculations . . . 19
Experiment 2--RSC Procedure . . . 19
Experiment 3--High Inoculum Concentration . . 21 Experiment 4--Inoculation of 1-0 Seedlings . . 23
Analysis . . . 23

RESULTS AND DISCUSSION . . . 31

Field Tests--Experiment 1 . . .31 Artificial Inoculation . . 54
Experiment 2--RSC Procedure . . . 55
Experiment 3--High Inoculum Concentration . . 64 Experiment 4--Inoculation of 1-0 Seedlings . . 66 Research Perspectives . . .68

CONCLUSIONS . . . 70 LITERATURE CITED . . . . 72

APPENDIX . . 79

BIOGRAPHICAL SKETCH . . . 90















LIST OF TABLES


Table Page

1. Location, year of establishment, site index, fifth year height and rust incidence of Brunswick Pulp Land
Company's diallel progeny tests . . . .15

2. Sets of slash pine crosses used in field and artificial inoculation experiments. . . . .16

3. Percentage of trees infected with fusiform rust for 48 slash pine families (four- to six-years-old)
planted in six progeny tests in southeast Georgia . . . . 17

4. University of Florida rust evaluations, standard errors and number of tests used for evaluation of
slash pine parents. . . . .18

5. Analysis of variance and expected mean squares for a modified half-diallel as computed by DIALL. . .24

6. Analysis of variance and expected mean squares for a four-stage nested design using individual
tree data. . . . .26

7. Analysis of variance and expected mean squares for Experiment 2 using mean number of symptoms
per tray . . . . .29

8. Analysis of variance and expected mean squares for Experiment 4 using individual tree data. . .30

9. Percentage of rust-associated mortality for 48 slash pine families (four- to six-years-old) planted in six
progeny tests in southeast Georgia. . . .32

10. Percentage of trees dead and classified as 'rust bush many
cankers' for 48 slash pine families (four- to six-years-old)
planted in six progeny tests in southeast Georgia . . . . 33








Table Page


11. Results of analyses of variance of the proportion
of trees with different fusiform rust symptoms (arcsin
square root transformation) and the mean number
of these symptoms per tree for family and test
differences . . . 39

12. Significant replication (Reps), general combining
ability (GCA) and specific combining ability (SCA) F values (5% level of significance) obtained from
analysis of variance using DIALL for number of fusiform rust galls of several types/locations
on slash pine (18 analyses for each type) . . 41

13. Significant replication (Reps), general combining
ability (GCA) and specific combining ability (SCA) F values (5% level of significance) obtained from
analysis of variance using DIALL for the proportion
(arcsin square root transformation) of slash pine trees
with various fusiform rust symptoms (18 analyses for
each symptom) . . . . 42

14. Significant site (S), replication (Reps), general
combining ability (GCA), specific combining ability
(SCA),and interaction (GCA x S) F values (5% level of
significance) obtained from analysis of variance using DIALL for number of fusiform rust galls of various types/locations on slash pine, planted at
several sites (nine analyses for each type) . . 45

15. Significant site (S), replication (Reps), general
combining ability (GCA), specific combining ability
(SCA),and interaction (GCA x S) F values (5% level of
significance) obtained from analysis of variance
using DIALL for the proportion (arcsin square root
transformation) of slash pine trees with various rust
symptoms planted at several sites (nine analyses
for each symptom) . . . 46

16. Individual (hi) and family heritibilities (hp)
and their standard errors (s) for rust incidence,
rust bush many cankers (RBMC), rust-associated mortality
(RAM) and the mean number per tree of 10 fusiform
rust gall types for slash pine, four- to
six-years-old . . .50








Table Page


17. Genetic correlations between various fusiform
rust gall types and locations (mean number per
tree), rust incidence and mortality (RAM and RBMC)
for slash pine four- to six-years-old. . . .52

18. General combining ability estimates of the
proportion of trees with fusiform rust galls for
25 slash pine parents based on progeny performance
in six locations in southeast Georgia, aged
four to six years . . . . .54

19. Percentage of seedlings in 20 slash pine families
with various reaction types six months after
artificial inoculation with fusiform rust and
the resistance index of the families . . .56

20. Mean squares for the analyses of variance of the
proportion of slash pine seedlings per family
expressing various disease symptoms resulting from
artificial inoculation with fusiform rust at the
Resistance Screening Center, Asheville, North Carolina . 57

21. Individual heritabilities 0,2) and their standard
errors (s) for artificial inoculation (AI) symptoms
by both proportion of trees with the symptoms and
mean number of symptoms per tree . . .59

22. Pearson product-moment correlations between the
mean number of artificial inoculation (Al) symptoms
per tree of slash pine seedlings inoculated with
fusiform rust, based on the progeny of 20 families . . . . 60

23. Pearson product-moment correlations between
field galls and artificial inoculation symptoms
(proportions) and the resistance index of 17 slash
pine families and fusiform rust . . .62


v ii i








Table Po


24. Results of artificial inoculation with high
concentrations of inoculum (2.0 x 106 spores/ml)
on 11 slash pine families, Resistance Screening Center,
Asheville, North Carolina. . . .65

25. Comparison of family means for several fusiform rust
disease symptoms on artificially inoculated 1-0 slash
pine seedlings. . . . .67

26. Least squares estimates of the mean number of fusiform
rust galls per tree by gall type or location for 48 slash pine families (four- to six-years-old) planted
in six progeny tests in southeast Georgia. . .79

27. Least squares estimates of the mean proportion
of trees infected with fusiform rust galls by gall type or location for 48 slash pine families (fourto six-years-old) planted in six progeny tests in
southeast Georgia . . . .81

28. Variance components and their standard errors for
replication, general combining ability (GCA),
specific combining ability (SCA) and error effects
for the mean number of galls per tree for various fusiform rust gall types or locations as computed
by the FORTRAN program, DIALL for each crossing
set and test . . . . .83

29. Specific combining ability for fusiform rust incidence
and fat galls based on the proportion of trees
diseased or having fat galls. Values reflect
performance of 48 families (four-to six-yearsold) in six progeny tests in southeast Georgia. . .88
















LIST OF FIGURES


Figure Page

1. Gall types commonly found on field-grown slash
pine trees infected with fusiform rust. . . .20

2. Galls found on slash pine seedlings six months
after artificial inoculation with fusiforrn rust. . .22

3. Frequency distribution for slash pine trees
by the total number of fusiform rust galls, in six
progeny tests in southeast Georgia, at four-, five-,
and six-years-old. . . . .35

4. Frequency distribution for two slash pine families,
0098 x 0088 (resistant) and 0350 x 0354
(susceptible), for the total number of fusiform
rust galls at age five years, planted in a progeny
test in Camden County, Georgia . . .38
















Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




GENETIC VARIATION IN SYMPTOMOLOGY OF SLASH PINE IN RESPONSE TO FUSIFORMr RUST

By

Patricia Adlene Layton

December, 1985

Chairman: R. E. Goddard
Major Department: Forest Resources and Conservation

Genetic variation in the reaction of slash pine to fusiform rust was investigated in field and greenhouse situations. In the field, four- to six-year-old plantings of five crossing sets, each involving five or six parents, were studied. Each crossing set was planted at two or four sites. Rust incidence in the tests ranged from 46% to 89%. Survival and stem form of each tree was determined. Galls on infected trees were counted and classified as to location and type.

Variances in traits measured were analyzed by crossing set, crossing sets among sites and all crossing sets at all sites. Significant differences among families were found for some gall types, gall locations, total number of galls per tree, incidence,










mortality and bush stem form. Additive genetic effects were found in many traits, but the number of fat galls had non-additive effects. Total number of galls was the best field measure of susceptibility, having high genetic correlations with mortality and bush stem form.

Seedlings from 17 of the field-tested families were

artificially inoculated. After six months, there were differences among families for the proportion of seedlings and number per tree with seven resistance and susceptibility indicators. Although heritabilities were often low, those for proportion of trees with galls, purple spots and adventitious shoots were 0.23, 0.28 and

0.33, respectively.

Following artificial inoculation with high concentrations of rust inoculum, families varied in the number of red spots at 10 and 48 days.' Correlations with family field performance were low.

Galls developing on artificially inoculated one-year-old

seedlings resembled field gall types more than those on younger seedlings. Families differed in percentage of trees with rust and the mean number of fat galls. Three of four phenotypically resistant males increased family resistance to a level equal to or better than a rust-resistant orchard check.

These studies indicated genetic variation for specific disease reactions in the field and confirmed such variation in artificial inoculation studies. Proportion of trees infected, total galls per tree, mortality, and bushstem form had the highest heritabilieies and should be considered in screening for rust resistance.
















INTRODUCTION


The major pest of southern commercial forests is fusiform rust (Cronartium quercuum (Berk.) Miyabe ex Shirai f. sp. fusiforme) (Cqf). In a 1979 survey of five states (South Carolina, North Carolina, Virginia, Georgia and Florida), more than one-third of the acreage of slash (Pinus elliottii Engelm. var. elliottii) and loblolly (P. taeda L.) pine had more than I04 of the trees infected (Robert Anderson, personal communication 1985), representing an annual loss of $32,894,000. Losses throughout the South are expected to be much greater.

Disease incidence has been increasing over time (Schmidt et al. 1974; Griggs and Schmidt 1977). Management practices, such as mismatching species and site, fertilization and intensive site preparation (Powers et al. 1981), are implicated in this increase. Another reason for the increase may be the abundance of rust-susceptible Quercus spp. (alternate host of the fungus) in southern forests (Squillace et al. 1978). Modifying these management practices as well as utilizing more effective silvicultural techniques for suppressing Quercus spp. may help











reduce the epidemic. Planting rust-resistant pines is an effective way of significantly decreasing disease losses (Schmidt et al. 1981; Schmidt et al. 1985).

The disease is caused by a heteroecious, macrocyclic rust

fungus. Pycnia and aecia occur on pine while uredia and telia occur on oak. Basidiospores produced from telia on oak infect pine during favorable environmental conditions. Most southern red oaks, especially water oak (Q. nigra L), are susceptible to the fungus (Dwinell 1974). Although much is known about the life cycle (Owinell 1977), knowledge is lacking about the genetics of the fungus. There is a wide range of variation in the ability of Cqf to infect pine (Snow and Kais 1970; Snow et al. 1975; Powers et al. 1977). Spores from galls on resistant families evidenced increased virulence in comparison to other spore collections when used to reinoculate the resistant families (Griggs and Walkinshaw 1982; Powers et al. 1978; Snow and Griggs 1980).

Various review articles have summarized family resistance in slash and loblolly pine (Goddard and Wells 1977; Powers et al. 1981; Schmidt et al. 1981). Certain loblolly pine geographic seed sources also are resistant to infection by Cqf (Wells and Wakely 1966).

In the past, family resistance to Cqf has been evaluated by percentage incidence in both field and greenhouse studies. This method is effective in screening families for resistance, but does little to identify resistance mechanisms or their inheritance. More recent methods have evaluated disease resistance by examining










symptoms resulting from artificial inoculation (Carson 1984, Framptom et al. 1983; Walkinshaw et al. 1980).

The objectives of this study were as follows:

1. To examine and describe qualitative characteristics

(symptoms) of the response of slash pine to Cqf, both

in the field under natural conditions and in the

greenhouse using artificial inoculation techniques;

2. To determine heritability and genetic correlations of

the traits; and

3. To correlate field and greenhouse observations and

suggest symptoms that may lead to a more precise

prediction of field response by greenhouse testing.















LITERATURE REVIEW


As fusiform rust became epidemic, tree improvement programs incorporated disease resistance into their selection and breeding programs. Although not a total solution, resistance is the primary pest management strategy in intensively managed forests (Schmidt et al. 1981; Schmidt et al. 1985). Original mass selections were essentially random for rust resistance, but approximately 7.5% had some degree of resistance when evaluated using artificial inoculation techniques (Goddard and Arnold 1966). Goddard and Arnold (1966) reported 0.199 as an estimate of heritability from artificial inoculation. Estimates of heritability for percentage rust incidence in field tests of slash pine varied greatly. Sohn and Goddard (1979) stated that they varied with the incidence levels in the tests examined. Individual heritabilities varied from 0.054 to 0.238 (Rockwood and Goddard 1973) in several progeny tests. Sohn (1977) estimated individual heritabilities for incidence ranging from 0.067 to 0.285 in other progeny tests. The relationship between incidence and heritability in these studies supported Sohn and Goddard's (1979) conclusion.










Various estimates of rust resistance are reported for loblolly pine. Kinloch and Stonecypher (1969) reported family heritabilities for incidence of 0.65 to 0.85 over four different environments. Blair (1970) estimated heritabilities for number of galls per tree, severity index and percentage incidence. In one field test, these estimates were 0.29, 0.22 and 0.20, and in another they were 0.09,

0.04 and 0.12, respectively. In the same tests as Blair, but at an older age, Barker (1973) estimated the heritabilities of the severity index to be 0.11 and 0.35. These heritabilities supported the concept that substantial gains in rust resistance were possible.

Screening selections for rust resistance became a priority. Inoculation techniques were developed to screen families for incorporation in breeding programs (Jewell 1960; Goddard and Schmidt 1971). Schmidt (1972) describes early techniques and their deficiencies were identified by Laird and Phelps (1975) as

1. temperature and/or humidity not adequately controlled;

2. inoculation time not controlled;

3. inoculum concentration not controlled;

4. techniques not applicable to large numbers of

seedlings;

5. results not repeatable; and

6. results not quickly obtainable.

Laird and Phelps (1975) tested three methods of artificial inoculation and determined that the concentrated basidiospore spray system (CBS) (Matthews and Rowan 1972) was best for minimizing these problems.










In 1973, the United States Forest Service established a

screening center (RSC) for rust resistance in loblolly and slash pines from throughout the South at Asheville, North Carolina, utilizing the CBS system. There, in a controlled environment, seedlings could be inoculated with various pathogen sources at optimum inoculum levels (Hubbard and Anderson 1980). The percentage of trees galled at six to nine months of age was used to rank family resistance. Resistant and susceptible checks were used in every test.

Screening by artificial inoculation (all methods) was plagued by lack of adequate correlation with family field performance. The University of Florida Cooperative Forest Genetics Research Program began field testing families rather than relying solely on artificial inoculation procedures. Short-term field progeny tests were established to mass screen selections in their tree improvement program (Goddard et al. 1972). Later, other tests were established to retest the most resistant families in areas of high rust incidence (Goddard et al. 1984). Many families with high degrees of resistance were found with these methods. However, open-pollinated families varied tremendously in their reactions. Powers and Zobel (1978) found significant differences in infection occurred on open-pollinated seedlings of the same clones produced in different orchards. These results illustrate the strong influence of the pollen parent in resistance. To understand the genetics of resistance better, testing of full-sib families (artificial inoculations and field tests) must be performed.










Artificial inoculation methods were useful in exploring host variation in response to infection. Families were found that were resistant to Cqf from some regions, but not to spores originating in other areas (Snow and Kais 1970). Gall form (ratio of length to width) varied by host family and inoculum source (Snow et al. 1982). Resistance was overcome in some families by particular isolates of the pathogen (Griggs and Walkinshaw 1982; Powers and Dwinell 1978). Selection for increased virulence in the pathogen has been found in slash pine families (Dinus et al. 1975; Griggs and Walkinshaw 1982; Snow and Griggs 1980). Snow et al. (1976) stated that the greater virulence of inocula from resistant trees was evidence of pathogenic specialization.

Virulence varies greatly in the fungus. Powers (1980) found as much variation in 10 single aeciospore inoculations from one gall as he did in stand- or county-wide collections of aecia. Powers and Owinell (1978) reported that pathogen virulence had not increased within time when comparing isolates from galls originating in 1945 and 1970; however, the 1970 isolates were more variable in their response.

The fungus infects pine in a haploid (IN) state, while the aeciospore, collected from pine for testing, is a dikaryon (N+N). The biological mechanism for this change is not known. This was studied in the related species C. ribicola by flirt (1964), but he found no conclusive evidence of sexual recombination. Until the sexual processes of Cqf are understood, the exact nature of pathogenic variation will be unknown.










Increasing the accuracy of assessment of host variability is a continuing goal. Walkinshaw et al. (1980) developed a scoring procedure for open-pollinated slash pine families tested by the RSC to better predict resistance. They examined twelve traits. Correlations between RSC performance and field evaluations were increased by utilizing the proportion of trees in a family with 1) smooth galls, 2) symptoms without swelling and 3) fat galls. These variables accounted for 62% of the variation in field performance evaluations. Predicting field performance with similar methods for loblolly pine has not been reported. Walkinshaw and the staff at the IRSC continue to test methods for this species (C. H. Walkinshaw, personal communication 1984).

Walkinshaw and Anderson (1983) described seven fusiform rust symptoms found on seedlings in the greenhouse and eight on trees in the field. In the greenhouse, the most resistant symptom found was a purplish discoloration on the stem and/or a needle base, called SYMNO. Fungal tissue was not found in the area of discoloration. A short gall of less than 25 millimeters and galls with a rough textured, slightly discolored surface were also indicative of resistant reactions. More susceptible reactions included fat galls (gall diameter is twice that of the stem below the gall), typical galls (gall is fusiform shape and not abnormally large or small), smooth galls (gall surface is smooth and generally green on six-months-old slash pine seedlings) and baseball bat galls (oase of the gall is constricted, but there is a typical fusiform shape at the top). Baseball bat, fat and typical galls were also found on










trees in the field. In addition, field symptoms included thin galls (only slightly swollen), twisted galls (stem twists in the area of the gall), sunken galls (a depression in the gall surface) and witches' broom galls (a loss of apical dominance). The authors proposed using these symptoms to increase the prediction capabilities.

Prediction methods rely on the premise that resistant or susceptible reactions in seedling progeny of a particular open-pollinated family can predict response of other family members grown in the field. Day (1974) advised that possible limitations of this premise include the following:

1. The observed resistance or susceptibility may be a

feature of the juvenile tissues (cotyledons and

unsuberized stems).

2. Observed reactions may reflect test environment as

opposed to host genotype.

Other problems with relating field and artificial inoculation tests are pathogen related. Composite inoculum from the area where host families are to be outplanted eliminates (or complicates) many of the problems of pathogenic variability. Inoculum concentrations are extremely variable in the field, but are fairly constant in artificial inoculations. Lundquist et a]. (1982), utilizing extremely high inoculum concentrations, found breakdowns in resistance for some slash pine families--a product of both inoculum concenfiration and host genotype. In their tests, families were ranked by observations of pigmented spots on seedlings and the










rankings were Similar to those achieved by the CBS method (percentage galled). The authors suggested that two types of resistance may exist: one that persists under high inoculum loads and one that does not. They did not consider that high inoculum concentrations significantly increase the probability of a virulent spore's landing on a seedling. Further testing of this method is needed.

Lundquist and Luttrell (1982) reported that pigmentation

patterns found in response to rust inoculation were similar to those observed when seedlings were wounded or exposed to ultraviolet light. They suggested this may have been a specious effect or a reaction by-product. Timing of pigmentation varied with host-parasite interaction and did not vary with other stress-related reactions, suggesting it was related to a resistance mechanism.

Miller et al. (1976), utilizing different methods of

inoculation from Lundquist and Luttrell (1982), suggested there might be four general types of host-parasite interactions at work. The first interaction was the inability of germinated basidiospores to penetrate and/or infect host epidermal cells. This was considered total incompatibility of host and pathogen. The second interaction, subliminal infection, resulted in infection and a sparse spreading of mycelium witm a few scattered haustoria. Host cells did not show the distortion typically found in infected trees. Miller et al. (1976) suggested that the pathogen in this interaction was virulent. The authors described three hypersensitive reactions as a third interaction. Two of these










hypersensitive reactions resulted from stem infections and produced darkly stained reaction zones. Purple stem lesions were visible for both reactions. They differed primarily in size and time of development. The third hypersensitive stem reaction resulted only from primary needle infection. Internal reaction zones developed and were stabilized from enlargement within three months. The fungus was confined to the reaction zone and the seedlings recovered from infection. There was no evidence of resistance in the needle tissue; resistance occurred only after the fungus reached the cambial area and began to spread. The fourth host-parasite interaction was typical gall development commonly found with susceptible pines and virulent Cqf.

Resistant zones similar to those identified by Miller et al.

(1976) were described by Jewell et al. (1982). Fungal tissues were effectively isolated in resistant pines. In 1982, Jewell et al. reported that wounding trees produced reactions similar to those produced by fusiform rust infection. Resistant seedlings differentiated tissues incompatible to rust thereby limiting the spread of the fungus. Susceptible seedlings produced initial resistant reactions, but were unable to produce incompatible tissues.

Walkinshaw (1978) found seedlings that had necrotic areas surrounded by a periderm. His discussion did not point out differences between resistant and susceptible seedlings, but centered mainly on the number of necrotic, tannin-filled areas.

Lundquist and Miller (1984) described a relationship between macroscopic pigmentation patterns and microscopic events in resistant lesions. These lesions resulted from artificial










inoculation with high concentrations of inoculum. Developing phellem cells formed a boundary that effectively checked the spread of the disease and distorted infected cells, pushing them toward the epidermis. Prior to or in conjunction with periderm formation, an impermeable layer was formed.

Mullick (1977) discussed three non-specific processes that are initiated during pathogen attack: 1) phellogen restoration, 2) vascular cambium restoration and 3) sapwood conduction blockage. The process triggered was dependent on depth of attack, with processes one, two and three occurring with increasing depth of injury. They may act independently or collectively depending on the attack. The author used microscopic techniques not previously used in observation of Cqf infections in pine tissues. His processes appeared to describe the observations of Frampton et al. (1983), Gray and Amerson (1983), Jewell et al. (1982), Lundquist and Miller (1984), Miller et al. (1976), and Walkinshaw (1978). Mullick (1977) proposed that host resistance or susceptibility is a function of the successfulness of completing the periderm under the pathogen's influence. During restoration, the pathogen is exposed to a host of chemicals that might influence its ability to successfully survive (e.g. phytoalexins). Virulence results from the pathogen's 1) not triggering the host response, 2) blocking the triggered response or 3) growing and spreading faster than phellogen formation can respond.

The microscopic differences in host response as discussed by

the authors above should be represented by corresponding differences







13


in macroscopic symptoms. Carson (1984) found family differences in macroscopic symptoms on artificially inoculated loblolly pine. Walkinshaw and Anderson (1983) described different macroscopic symptoms on slash pine in field studies. However, they did not present evidence of family variation for these symptoms.















MATERIALS AND METHODS


Four experiments were conducted to examine fusiform rust

resistance variation in slash pine. Three experiments contained full-sib families in five sets of crosses. The crossing sets, belonging to Brunswick Pulp Land Company (BPL), were designed as four- or five-parent half-diallels with an additional cross between the females and a male from another geographic area (i.e. a wide cross). Field tests of the crossing sets ranged from four- to six-years-old. Seed from these sets were sent to the RSC for a standard artificial inoculation study and a high inoculuin concentration technique study (Lundquist et al. 1982), Experiments 2 and 3, respectively. The fourth experiment involved artificially inoculating a group of 1-0 nursery grown seedlings at RSC. These seedlings were donated by Container Corporation of America (CCA) and are part of a rust-resistant factorial mating design.


Field Tests--Experiment 1

Six progeny tests containing four diallel crossing sets each were planted by BPL near Brunswick, Georgia. Progeny tests were randomized complete block plantings with three to five blocks and seven- to ten-tree row plots for each family. Crossing sets one, two, four, five and nine were examined. Test sites were cleared, windrowed and bedded prior to machine planting. Tests 34 and 35










Table 1. Location, year of establishment, site index, fifth year
height and rust incidence of Brunswick Pulp Land
Company's diallel progeny tests.


Crossing 5th Year Rust Site
Test Location/County Planted Set Mean Ht. Incidence Index.a/

(Georgia) (yr) (feet) M%


34 Wayne 1978 1,3,4,5 14.5 89 60
35 Wayne 1978 1,3,4,5 14.8 61 55
36 Camden 1979 1,2,6,10 10.4 46 65
37 Camden 1979 1,2,6,10 12.1 58 60
40 Wayne 1980 2,4,5,9 _.b/ 54 65
41 Camden 1980 2,4,5,9 _.Y 66 65



A./Average height of dominant and codominant trees at age 25. WYFifth year heights were unavailable.



were disked before bedding. All tests sites were in the coastal plain. Tests 34, 35 and parts of 37 were planted on well-drained ultisols. Tests 36 and 40 were planted on somewhat poorly to poorly drained spodosols. Test 41 and portions of Test 37 were planted on somewhat poorly drained entisols. Site index at age 25 and fifth year heights (Tests 34 to 37) are presented in Table 1, which contains details of the tests and their locations. The parents and crosses made are shown in Table 2. Not all crosses were represented at all possiole locations. A listing of crosses in each test may be found in Table 3.

The 25 parents in the crosses were evaluated previously for rust resistance in open-pollinated progeny tests by the University





of Florida Cooperative Forest Genetics Research Program. Parental evaluations (Table 4) are expressed as weighted mean standard deviations away from progeny test mean incidence levels. The weighting factors were for incidence level (Sohn and Goddard 1979)


in field and artificial


Table 2. Sets of slash pine crosses used
inoculation experiments.


Males


Female


Set I


0098
1,2
1
1


0096
1
1



0001
1
1,2,3
1


0060 1,2,3
1



0295
1,2 1,2


0088 1,3
1,2
1,3


0065
1
1
1


0159 1,3 1,2,3
1


0287 1,2,3 1,2,3
1


0355
1
1,2
1


0048 1,2,3 1,2,3
1


0071
1
1



0050
1
1
1


0047 1,2,3
1
1


0064 1,2 1,2,3
1


0350
1 2, 3a/




0146
1

1,3


0352
1



0141
1




0050
1
1,2,3


0354 0350 0098


Set 2


0157
0146 0096


Set 4


0288 0352 0001


Set 5


0019
0141 0060


Set 9


0284 0270 0295


A/I=Cross included in Experiment
Experiment 2; 3=Cross included


1; 2=Cross included in in Experiment 3.










Table 3. Percentage of trees infected with fusiform rust for 48
slash pine families (four- to six-years-old) planted in
six progeny tests in southeast Georgia.


Least
Family Progeny Tests Squares
Female Male 34 35 36 37 40 41 Mean
-------------------------- % --------------------------


97.7
64.9 58.5 66.5

64.8 70.1 66.1 95.2

37.8 25.0

56.9 79.5
74.3 72.4


0354 0350
0098 0088
0048

0350 OU98
0088
0048 0354


100.0 96.7 93.3 90.0

100.0 100.0 89.6


100.0 55.8 65.2 63.3

52.6 68.5 56.7


97.1
48.2 38.2
49.1

47.6 48.1 37.9 78.6


95.8 58.9
43.7 69.6

57.3 50.2 70.5 97.1


0098 0088 76.7 31.1 31.7 23.3
0048 63.3 13.3 17.3 22.3


42.9 52.3 56.2
24.8


51.9
82.4 75.1 83.1


59.4 70.4 68.3 62.5


38.8 78.8 61.1 82.2


0157 0146
0096 0065 0071

0146 0096
0065 0071

0096 0065
0146

0288 0352
0001 0159 0050

0352 0001
0159 0050

0001 0159
0050

0019 0141
0060 0287
0047

0141 0060
0287
0047


42.9
37.1 46.8 27.8 45.4


45.6
25.0 50.8 51.4 40.3 62.5 53.5


46.2 56.2 56.8 68.2 65.2
46.4 61.7 59.8


40.9 56.9 71.5 48.7

41.1 64.6 31.0

73.6


91 .9 66.9 66.9
45.0

40.8 43.5 50.0


58.7 53.8 96.1 30.0


62.9 63.7 82.9 57.7


80.4 96.7 97.0 100.0

73.4 81.7 83.3

96.7


100.0 96.7 93.0 88.9


86.7 66.7 79.3
45.8

40.7 51.1
27.4

73.3
46.3

100.0 89.6 71.7


95.2


30.4 41.4 42.8 54.9
41.7


77.9
45.1

95.8 82.9 70.5 67.3


92.8 85.5 65.5 80.0


76.8 36.7 47.6
66.4


91.7










Table 3--continued.


F am Ii y
Female Male


Least
Progeny Tests Squares
34 35 36 37 - 40 41 Mean


0060 0287
0047

0284 0050
0295 0355
0064

0270 0050
0295 0355
0064

0295 0355
0064 0270


48.6
85.0 46.7


27.5 25.6

43.3 57.4 59.2 70.8

36.9
45.0 66.4 70.8

78.2 82.3
41.9


52.5 45.8
45.9


51.4 61.3
70.4 83.7

59.4 85.5 62.5 97.2

100.0 88.5 58.6


55.7 65.2 71.3 75.7

54.7 71.7 72.2
91.4

96.6 92.7 67 .0


Table 4. University
and number
parents.


Florida rust evaluations, standard errors tests used for evaluation of slash pine


No. of Rusta'/ Standardk/ No. of Rust Standard
Parent Tgsts Eval. Error Parent Tests Eval. Error
0157 9 .19 .12 0287 7 -.06 .35
0146 19 1.20 .14 0047 18 1.10 .17
0096 10 .86 .16 0284 13 .75 .18
0065 4 .82 .11 0270 7 .39 .14
0071 9 .64 .10 0295 8 .10 .18
0288 11 -.03 .18 0355 14 .31 .13
0352 10 .68 .18 0064 4 .53 .26
0001 10 .41 .13 0354 11 .12 .13
0159 14 .45 .20 0350 8 .49 .31
0050 12 1.07 .25 0098 12 1.15 .24
0019 10 -.41 .21 0088 9 .29 .14
0141 11 -.08 .09 0048 21 1.53 .17
0060 7 .07 .36 0621 14 .35 .18
0298 12 .31 .17 0618 18 .98 .14
0620 12 .60 .21


!/R ust Evaluation = L Infection We iht x Progeny Mean)/(Error Mean Square)1 .


Lot Weight x ((Test Mean -


P./standard Error = (Variance of a family rating/number of tests)1/2.










and number of families in a test (Lewis 1973). The standard error (Table 4) of an evaluation was based on the unweighted standard deviations.

Data were collected from the tests in either March or June, 1984. Gall development was assumed not to change dramatically during this time. However, each test was measured within a three-day period and four of the six tests were measured during a two-week period in June. Tests 36 and 37 were measured during a one-week period in March. Data collected from each tree included survival, stem form, and the number of 1) stem, 2) limb, 3) stemto-limb, 4) fat, 5) thin, 6) typical, 7) basally truncated fusiform (BTF) (referred to as baseball bat by Walkinshaw and Anderson (1983)), 8) witches' broom, 9) twisted, and (10) sunken galls. Gall types are illustrated in Figure 1.


Artificial Inoculations

Experiment 2--RSC Procedures

Seventeen families (Table 2) were tested at RSC using routine inoculation procedures described by Anderson et al. (1983). The inoculum originated in the Nassau County, Florida, area which is approximately 30-75 miles south of the field test sites. This same inoculum source was used by Walkinshaw et al. (1980) in developing their rust resistance prediction system. Two runs (replications) of 24-60 seedlings from each family were inoculated with a concentration of 25,000 spores per milliliter of solution. The variable number of seedlings was due to germination and survival differences. Three standard check lots were included with this

















4Q


Typical


Witches's Broom


Twi sted


Thin


Sunken


Rust Bush Many Cankers


Figure 1. Gall types commonly found on field-grown slash pine trees
infected with fusiform rust.










test. Data collected from each seedling included the number of 1) symptoms without swelling (SYMNO), 2) galls present, 3) rough galls, 4) fat galls, 5) galls less than 25 millimeters long (LT25M), 7) BTF galls and 8) sunken areas on the gall and the occurrence of adventitious shoots in the galled area. Gall types are illustrated in Figure 2. Data were collected with the assistance of an RSC staff member wno routinely collects data on seedlots tested there.

Presence of a pink blush on seedlings 10 days after inoculation was also recorded. This symptom may be related to the red spots reported by Lundquist and Luttrell (1982). It consisted of a faint, irregular spot on the hypocotyl one to three weeks after inoculation. The relationship of this symptom to resistance is unknown.


Experiment 3--High Inoculum Concentration

Fifteen full-sib families from the BPL crosses (Table 2) were inoculated following techniques of Lundquist et al. (1982). The inoculum concentration of 2.0 x 06spores per milliliter was extremely high, allowing only 20 seedlings per family to be inoculated with the limited quantity of available basidiospores. Eleven of the families were used in Experiment 2 (Table 2). Numbers and types of spots after 10 and 48 days were tallied for each seedling in the test. Personnel at RSC transferred the seedlings to different trays between the two measurement dates, preventing observation of the increase in spots per tree over time.























SYMNORug


Ii*


Thin


Gall with Adventitious
Shoots


Gall with Sunken Area


Figure 2. Galls found on slash pine seedlings six months after artificial
inoculation with fusiform rust.


LT25M


7, r






Rough









Experiment 4--Inoculation of 1-0 Seedlings

This experiment contained six full-sib families and two check lots grown in the CCA nursery near Archer, Florida. The seedlings were lifted in January 1984, transplanted in one gallon pots filled with sterilized medium and transported to the RSC. There were two seedlings per pot and 20 pots per family or check lot. The seedlings were inoculated after their first flush of growth with the CBS system using 35,000 spores per milliliter. The pathogen source was from Louisiana as the trees flushed earlier than expected and the Nassau County pathogen source was unavailable. The six families were 1) 0298 x 2002, 2) 0298 x 2005, 3) 0298 x 2013, 4) 0620 x 2002, 5) 0618 x 2007 and 6) 0621 x 2005. The four male parents (2002, 2005, 2007 and 2013) were phenotypically rust resistant, selected in stands exposed to very high levels of natural inoculum. Most trees in these stands were severely infected. The male parents are presumed likely to have genetic resistance (Goddard et al. 1975). The female orchard selections and their rust evaluations are shown in Table 4. The check lots were groups of seedlings from a rust-resistant orchard and an orchard where there was no selection for rust resistance. Observations on these seedlings were the same as those in Experiment 1.


Analysis

To determine least squares estimates of family performance across tests, the GLM procedure in SAS (SAS Institute Inc. 1982b) was used. The model for the analysis was Yij u + Ti + F i + eii, where T i was the effect of the i th test, F i was the










effect of the jth full-sib family and e. was the error representing the family x test interaction. The analysis was performed on the means of gall types and locations, and the proportion of trees (arcsin square root transformation) having a particular gall type or location.

Each diallel crossing set was analyzed using DIALL, a FORTRAN program (Schaffer and Usanis 1969). This program allows computation of general (GCA) and specific (SCA) combining ability, site, replication and GCA by site effects (Table 5) by a least squares approach to the analysis of variance. The computation method used an abbreviated forward Doolittle solution rather than direct matrix inversion. Estimates of the variance components were calculated by inverting the matrix of coefficients of the expected mean squares


Table 5. Analysis of variance and expected mean squares for a
modified half-diallel as computed by DIALL.


Source df Expected Mean Squares


SiteA/ 1/ Ve + c9VGCAxS + clOVR + CllQS

Replication within Site 4 Ve + c7VGCAxS + c8VR

General Combining Ability 4 Ve + c4VGCAxS + c5VSCA + C6VGCA
(GCA)

Specific Combining Ability 4 Ve + c2VGCAxS + c3VSCA
(SCA)
GCA x Site 4 Ve + clVGCAxS

Error 36 Ve


. /Analysis for one site follows a similar model except for site
and GCA x site effects.

-n/The degrees of freedom here are an example of those generated
by the program.










and multiplying the computed mean squares for each variable by this inverse. The standard error of a variance component was computed by the equation (Anderson and Bancroft 1952)

std err = (L". (2a . MS i2 /DF i+2)) 1/2,

where a.= the coefficients of the linear combination of the mean squares used to estimate the variance component,

MS.i= the i thmean square,

DF i= the degrees of freedom for the i th effect.

In calculation of the F test, when the denominator mean square was smaller than a lower order effect containing that variance component in its expected mean square, the lower order effect was used as the denominator. For example, in an analysis of one site, if the SCA mean square was smaller than the error mean square, the error mean square was used as the denominator in the F test computation for GCA. DIALL did not adjust variance component computations to reflect this problem. All V GCA were computed by subtracting the niean square for SCA and dividing by the coefficient of V GCA.

Traits analyzed included total number of galls, number of galls by location on the tree, and all six gall types found on trees in the field. Plot means were used in this analysis. Additionally the proportion of trees in a family having a certain gall type or gall location was analyzed. Proportion data were transformed by the arcsin square root transformation (Steel and Torrie 1960). GCA by site interaction was tested by pairing only those diallel crossing sets planted in the same year to avoid confounding year effects.










Estimates of heritability and variance components were computed oy pooling all families at all sites as if they were in a nested design (Table 6). This method made use of all data available to compute the estimates. The procedure, while computationally feasible, would oe expected to produce biased estimates of variance components. The analysis accounted for the 18 crossing set-site groups (CSS), males within CSS, females within males within CSS and replications within females within males within CSS (Table 6). The NESTED procedure in SAS (SAS Institute Inc. 1982b) was used to analyze individual tree data. Incidence, rust-associated mortality and rust bush many cankers were analyzed as binary data.


Table 6. Analysis of variance and expected mean squares for a
four-stage nested design using individual tree data.


Source df Expected Mean Squares


Crossing Set-Site (CSS) 17
Males (CSS) 58 Ve + 9.20VR(FMCSS) + 31.73VF(MCSS)+
59.23VM(CSS)

Female (Males, CSS) 75 Ve + 9.05VR(FMCSS) + 30.80VF(MCSS)

Replication (Females, 405 Ve + 8.28VR(FMCSS)
Males, CSS)

Error 4179 Ve









Individual tree heritabilities were calculated by the equation (Wright 1976)

2= 4VM(CSS)
I Ve + VR(FNCSS) + VF(MCSS) + VM(CSS)

where Ve = variance due to trees within plots,

VR(FMCSS) = variance due to replications nested within females, males and CSS, VF(MCSS) = variance due to females nested within males and CSS,

VM(CSS) = variance due to males nested within CSS. Family heritabilities were calculated similarly by the equation

.2 V
h2 VM(CSS) .
F V e/59.33 + VR(FMCSS)/31.73 + V F(MCSS)/9.20 + VM(CSS)
e R(FMCSS)F(MCFM(CSS)

Standard errors(s) for individual heritabilities were calculated by the equation (Wright 1976)

(1 - h 2/4) (1 + Kh2 /4)
s I
S [K(M - 1)/2]1/2

and for family heritabilities by the equation (Wright 1976)

s = (1 - t)(l + Kt)
[K(M - 1)/211/2
where t = h/4
hj= individual heritability,

K = number of trees in a male half-sib family,

M = number of male half-sib families.

The NESTED procedure calculates a covariance analysis.

Utilizing the covariance and variance estimates of the male source










of variation, an estimate of 1/4 of the additive genetic variance, genetic correlations were calculated by the equation


COVxY
rg (VX x Vy)112


where covxy = the covariance between two traits for males nested within CSS,

VX = VM(CSS) as described above for a trait,

Vy = VM(CSS) as described above for another trait.

General combining ability estimates for traits were computed using the SAS MATRIX procedure (SAS Institute Inc. 1982b). The computer program for this was developed by J. P. van Buijtenen (personal communication 1983).

Data in Experiment 2 were analyzed using the GLM procedure in SAS. Factors in the random model were runs, family, runs x family and trays within runs x family was the error component (Table 7). This model was used because the results from it would be similar to results obtained by RSC. The RSC does not account for male and female effects in their screening procedure analysis. Tray means for the traits observed were used in the calculation.

Pink blush was analyzed with a one-way analysis of variance on three to six tray means for each of 20 families.

Variance components were determined using the VARCOMP procedure (IIIVQUEO method) (SAS Institute Inc. 1982b) for females (VF), males (VM), run x males (VRM), run x females (VRM), females x males (V MF), trays within runs x females (VT(RF)), trays within runs x males (VT(RM)) and error (VE). This model accounted for










male and female effects ana was a more complete model than the previous one (Table 7).

Individual tree heritabilities were calculated by the equation


2 2(VF + VM1)
I V E + VT(RM) + VT(RF)+ VMF + VRM+ VF + VM



Table 7. Analysis of variance and expected mean squares for
Experiment 2 using mean number of symptoms per tray.



Source df Expected Mean Squares!/


Run 1 Ve + c4 VR x F + c5 VR

Family_/ 19 Ve + c2 VR x F + c3 VF

Run * Family 19 Ve + Cl VR x F

Trays(Run Family) 112 Ve


A/Due to the unbalanced design
expected mean squares were b/Includes three checklots.


coeffic ents unequal.


(c) in the true


and standard errors were calculated similarly. Correlations between variaoles were calculated by the SAS CORR procedure (SAS Institute Inc. 1982a).

Data from Experiment 3 were analyzed to determine within and between family differences for the number of spots found 10 and 48 days after inoculation using the model for a one-way analysis of variance for a completely randomized design. There were 15 families with approximately 20 trees per family.










Data in Experiment 4 also were analyzed using the SAS GLM

procedure. Independent variables in the model were family and pots within family (Table 8).



Table 8. Analysis of variance and expected mean squares for
Experiment 4 using individual tree data.


Source df Expected Mean Squares



Family 7 Ve + l.8Vp(F) + 35.5VF

Pots(Family) 146 Ve + l.9Vp(F)

Error 140 Ve















RESULTS AND DISCUSSION


Field Tests--Experimient1

Rust incidence in the six field tests ranged from 46% to 89% (Table 1). Test 34 had the highest percentage incidence and rust-associated mortality (RAM) (Tables 3 and 9). All 30 trees of family 0354 x 0350 were dead from rust and each had a rust bush form. More than 30 cankers were observed on eacn of these trees with a maximum of 75 on some trees. To facilitate measurements, data were not collected from dead trees with a rust bush form and many galls (RBMC), but a classification was established to account for these trees. Total gall number was assumed to be 31 and the number of distorted, stein, limb and stem-to-limb galls was set at one. All other trees dead due to rust were classified as having one stem gall; this was true on most tree remains found. Although several of the other tests contained R.BMC trees, Test 34 had the highest percentage (Table 10). A comparison of Tables 9 and 10 indicates that as RAM increased, there was an associated increase in the number of trees having a RBMC form.

The frequency distribution of number of galls per tree is found in Figure 3. Distribution of galls per tree for individual families did not always follow this same pattern.










Table 9. Percentage of rust-associated mortality for 48 slash pine
families (four- to six-years-old) planted in six progeny
tests in southeast Georgia.


Least Squares Mean


Family Progeny Tests
Female Male 34 35 36 37 40 41
-------------------------- ----------------------------


0354 0350
0098 0088
0048 0350 0098
0088
0048 0354 0098 0088
0048 0157 0146
0096 0065 0071 0146 0096
0065 0071
0096 0065
0146 0288 0352
0001 0159 0050 0352 0001
0159 0050 0001 0159
0050 0019 0141
0060 0287
0047 0141 0060
0287
0047 0060 0287
0047 0284 0050
0295 0355
0064 0270 0050
0295 0355
0064 0295 0355
0064 0270


100.0 30.0 17.2 9.1 32.1 0.0 16.0 3.3 53.3 10.3 6.7 0.0 40.0 10.0 17.6 0.0 36.7 7.1 5.9 4.2 41.4 20.7 8.0 3.7 34.5 3.3 3.1 8.7
15.2 14.3 10.3 0.0 13.3 0.0 3.8 0.0 3.3 0.0 14.3 10.3
3.1 0.0 10.0 0.0 0.0 0.0 5.3
5.7 9.1 5.0 12.5 0.0 0.0


27.7
2.8 5.2 5.1 3.0 7.1 2.3
10.2 0.3 0.1 7.3
2.8 4.5 2.2 5.9 5.8
4.7 2.8 6.5 6.2 11.1
8.4 7.5 5.7
4.6 3.3 9.8 6.7 16.3 7.3 5.0 7.7
1.3
4.8 4.4 2.2 1.7 6.5 7.1
6.5
4.8 6.5
4.8 4.8 8.4 8.4 4.8 4.8


45.8 56.7 50.0
54.5 50.0 28.6 30.0
48.3

86.7 53.3
48.3 37.9


20.0 23.3 10.3 8.3 10.3 10.3 6.9
20.0 11.5 11.1 9.1 7.1 26.7 0.0


2.7 2.7
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.6 0.0 0.0 2.8 0.0 0.0 0.0 0.0 0.0 0.0


45.8 3.8
0.0
14.3 20.0










Table 10. Percentage of trees dead and classified as 'rust bush many
cankers' for 48 slash pine families (four- to six-years-old)
planted in six progeny tests in southeast Georgia.

. Least
Family Progeny Tests Squares
Female Male 34 35 36 37 40 41 Mean
-------------------------- -------- ----------------------


100.0 30.0 6.9 4.6 32.1 0.0 0.0 0.0 53.3 10.3 6.7 0.0 40.0 10.0 2.9 0.0 36.7 5.1 0.0 0.0 41.4 20.7 0.0 3.7 31.0 3.3 3.1 4.3
9.1 10.7 6.9 0.0 6.7 0.0 3.7 0.0 0.0 0.0 5.7 6.9 0.0 0.0 0.0 0.0 0.0 0.0 5.3
5.7 6.1 0.0 0.0 0.0 0.0


24.3 7.7
5.9 3.7
2.0
5.7 1.9 8.3 0.1 0.1
4.8 2.5 2.5 2.5 6.2 5.2 2.5 2.5 3.6 5.0 8.8
4.0 4.9 4.2 2.4 1.7 6.7 0.6 37.1 3.5 2.6 5.3 0.7
3.6 3.5
1.4 0.4 3.6 3.6 3.6 3.6 3.6 3.6 3.6 3.6 5.1 3.6 3.6


0354 0350
0098 0088
0048 0350 0098
0088
0048 0354 0098 0088
0048 0157 0146
0096 0065 0071
U146 0096
0065 0071
0096 0065
0146 0288 0352
0001 0159 0050 0352 0001
0159 0050 0001 0159
0050 0019 0141
0060 0287
0047 0141 OU60
0287
0047 0060 0287
0047 0284 0050
0295 0355
0064 0270 0050
0295 0355
0064 0295 0355
0064 0270


0.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0



2.9 0.0 0.0 0.0

0.0

0.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.9 0.0 0.0


20.0 23.3 10.3 8.3 10.3 10.3 6.9
20.0
3.8
7.4 4.5 7.1 26.7 0.0


45.8 56.7
40.0 50.0
46.2 21.4 23.3 37.9

86.7
46.7 34.5 34.5


45.8 3.9
0.0
14.3 13.3























Figure 3. Frequency distribution for slash pine trees by the total number of fusiform rust galls,
in six progeny tests in southeast Georgia, at four-, five-, and six-years-old.










2
1.9
187
1.7 -7

1.6
1.5
1.4

Q~)1.2LLC 1.1

0 0.9i.- 0.8
0.7
z
0.6

0.57
0.4

0.2
oor,

0.1 177: 7J6 79
0 ('V('(~ (/I(WII7W177I7
0 1 2 3 4 5 6 7 8 9 10 1 1 31


GALLS PER TREE










Families relatively resistant to the disease had a skewed distribution with few trees with multiple galls, whereas more susceptible families often had many trees with many galls per tree. Differences in distribution patterns are illustrated by two families planted in Test 37 (Figure 4). Family 0098 x 0088 was fairly resistant (average percentage rust incidence of 37.8) and exhibited a strongly skewed pattern, whereas family 0350 x 0354 (average percentage rust incidence of 95.3) followed a different pattern (Figure 4). Griggs and Dinus (1977) found that distribution patterns reflected resistance.

Results of the analyses of full-sib family means are presented in Table 11. This analysis provided least squared means for famili-es for all traits. Twisted galls were infrequent and were deleted from this and all further analyses. All variables analyzed, with the exception of average number of sunken galls per tree and proportion of trees having sunken galls, differed significantly among tests. Sunken galls were rare with only 201 galls found. Significant differences among families were found for proportional data for gall location, the gall types witches' broom and typical, and rust incidence. (In this and further discussion the term incidence refers to the presence of any galls.) However, mean gall count per tree varied among families only for witches' broom galls and total galls per tree (Table 11). Families did not vary significantly for most gall types. Except for witches' broom galls, there did not appear to be any unique pattern of gall development associated with families across all planting sites. It is notable




























Figure 4. Frequeiicy distribution for two slash pine families, 0098 x 0088 (resistant) and
0350 x 0354 (susceptible), for the total number of fusiform rust galls at age five
years, planted in a progeny test in Camden County, Georgia.









14

13

.12

11

lo w 9w
8-/











.GALLS PER TREE
S98xOO8 50x54

0 1 2 3 4 5 6 7 8 9 10 31

GALLS PER TREE 0098x0088 0350x0354










Table 11. Results of analyses of variance of the proportion of
trees with different fusiform rust symptoms (arcsin
square root transformation) and the mean number of
these symptoms per tree for family and test differences.


Mean Squares
Galls Family Test


Proportion of trees:
Locations
Any 0.117** 0.723**
Stem 0.061** 0.980**
Limb 0.093** 0.799**
Stem-Limb 0.061** 1.370**
Types
Typical 0.088** 0.868**
Witches' Broom 0.077** 0.752**
BTF/ 0.019 0.139**
Fat 0.014 0.183**
Thin 0.014 0.038*
Sunken 0.017 0.136
Mean number per tree:
Locations
Total 13.166* 561.985**
Stem 9.216 401.512**
Limb 0.862 13.933**
Stem-Limb 0.073 0.792**
Types
Typical 0.823 13.696**
Witches' Broom 0.016* 0.755**
BTF/ 0.016 0.128**
Fat 0.006 0.075**
Thin 0.005 0.014*
Sunken 0.003 0.004


A/Basally truncated fusiform gall.

* Significant at 0.05 level.

** Significant at 0.01 level.


that there was a significant tendency for some families to have a

higher incidence of stem or stem-to-limb galls than others, although

the number of galls at a specific location on the tree did not vary










significantly by family. Some bias in these results was possible because the analysis used did not allow consideration of genetic relationships among half-sib families.

An analysis considering all genetic relationships was conducted using DIALL (Table 5). Crossing sets were analyzed individually at each planting site. As each crossing set involved only five parent trees, restricted and considerable differences in ranges of genetic expression were to be expected. Results of the average number of galls by type or location as well as the percentage of trees having that characteristic are presented in Tables 12 and 13. As each of the five crossing sets were planted at two or four sites, there were 18 individual analyses for each gall type or location.

For all variables, one or more major sources of variation (replication, general combining ability or specific combining ability) in at least one of the analyses was significant (Tables 12 and 13). No gall type or location had a consistent pattern of differences across all sets and planting sites. Significant differences among replications were infrequent as expected because the tests were designed with replications creating blocks to reduce variation in growth resulting from soil-site variation and not to block for differences in fusiform rust infection.







41


Table 12. Significant replication (Reps) general combining ability
(GCA) and specific combining ability (SCA) F values (5%
level of significance) obtained from analysis of
variance using DIALL for number of fusiform rust galls of
several types/locations on slash pine (18 analyses for
each type).


Source of Variation
Reps GCA S CA
Galls --- number of significant F valuesLocat ion
Total 1 3 3
Stem 0 31
Limb 0 4 2
Stem-Limb 1 4 2
Types
Typical 3 2 3
Witches' Broom 0 2 3
B TFI/ 1 1 2
F at 1 0 2
Thin 2 01
Sunken 0 01


A/Basally truncated fusiform galls.


Average number of galls per tree of each type more often had

significant SCA effects than GCA (12 versus 5). Fat, thin and

sunken galls per tree had no significant GCA effects, hence no

expression of additive genetic variance. However, their SCA

effects were sometimes significant suggesting the presence of

nonadditive genetic effects. The average number of typical,

witches' broom and BTF galls per tree had both significant GCA and

SCA effects.

The mean number of galls per location had the opposite pattern

from gall types as there were more than twice as many significant

GCA effects (11) than SCA effects (five). Total number of galls

per tree had an equal number of significant GCA and SCA effects.









Taole 13. Significant replication (Reps) general combining ability
(GCA) and specific combining ability (SCA) F values (5%
level of significance) obtained from analysis of
variance using DIALL for proportion (arcsin square root
transformation) of slash pine trees with various
fusiform rust symptoms (18 analyses for each symptom).


Source of Variation
Reps GCA SCA
Galls ---number of significant F values-Locations
All 1 8 0
Stem 0 3 2
Limb 0 3 1
Stem-Limb 0 4 0
Types
Typical 0 4 1
Witches' Broom 2 2 1
BTFI/ 0 1 3
Fat 0 0 3
Thin 1 1 0
Sunken 0 0 1
Mortality
RAMU/ 1 4 1
RBMC-_/ 0 2 2


2./Basally truncated fusiform gall.
b/Rust-associated mortality.

L/Rust bush with many cankers.



The proportion of trees with each gall type had three more

significant GCA effects than the mean number of gall types. Two of

these were for the proportion of trees with typical galls and the

other for thin galls. Typical and witches' broom galls had more

significant GCA effects than SCA effects. There were nine

significant SCA effects for the proportion of gall types, and each










type had at least one except for thin galls. There was no evidence of additive genetic effects for the proportion of trees with fat and sunken galls.

The proportion of trees with each gall location (stem, limb and stem-to-limb) had more than three times the number of significant GCA effects than SCA effects. The proportion of trees with stem-to-limb galls had no significant SCA effects. The proportion of trees galled (rust incidence) also had no significant SCA effects, but did have eight significant GCA effects. RAM had four significant GCA effects while RBMC had two.

These results suggested that gall locations (mean or

proportion) were more controlled by additive genes than gall types were, with the exception of typical and witches' broom galls. There appeared to be more additive genetic effects for the incidence of a gall type than for the average number of gall types per tree.

In crossing set two, there were significant GCA or SCA effects for only four analyses, fewer than any other crossing set. In contrast, crossing set one had significant GCA or SCA effects in 41 analyses. Such widely different results possibly are because of differences in the extent of genetic variation within crossing sets as only five parent trees were involved in each crossing set. Substantial differences among sets were expected.

Lack of consistency for significant F values for the trait's variance components implies that the variances are different among the eighteen subpopulations. Each analysis can be considered a










subpopulation because at a test there are two to four crossing sets. Also, at each test although the host genotype component is controlled, the fungal component is not. The host-parasite interaction is implicitly different, thereby producing subpopul at ions.

Variance components and their standard errors fcr each crossing set in each test as computed by DIALL (Table 5) are presented in Table 28. There are discrepancies in the significance of a variance component when judged by the F test or the standard error. Large amounts of variation are inherent in the estimation of both variance components and their standard errors. This variation, in addition to the small sample size and unbalanced data, may have contributed to the discrepancies. As an example, in crossing set 1, in Test 34 a highly significant F value was calculated for the GCA effect of the mean number of stem galls per tree; however, the large standard error of V GCA makes this value appear non-significant. The GCA mean square (0.431) when divided by the SCA mean square (0.00796) equaled 54.1, a highly significant F value (4 and 5 degrees of freedom). Subtracting the SCA mean square from the GCA mean square and dividing by the coefficient for V GCA (10.05) yielded 0.0421 as an estimate of V GCA. The standard error of the estimate calculated by the equation on page 25 was [(2(1/10.05) 2(0.431) 2/ (4+2)) + (2(1/10.05) 2(0.00796) 2/ (5+2))]l1/2 = 0.0245 which is slightly more than half of V GCA.










Each crossing set was planted at two sites in a single year.

Year effects were not estimable as tests were not planted at the

same site in different years. Site effects, without confounding age

or year differences, were estimated by analyzing each crossing set

at two sites planted in the same year. In general, sites were

different for mean gall number per tree by type or location

(Table 14). GCA by site interaction (GCA x S) was significantly

different in 21 of 90 (10 traits in each of nine crossing set-site



Table 14. Significant site (S), replication (Reps), general
combining ability (GCA), specific combining ability
(SCA) and interaction (GCA x 5) F values (5% level of
significance) obtained from analysis of variance using
DIALL for number of fusiforin rust galls of various types!
locations on slash pine, planted at several sites (nine
analyses for each type).


Source of Variation
Site Reps GCA SCA GCAxS
Galls ----number of significant F values ------Location
Total 8 0 1 4 3
Stem 5 0 2 0 3
Limb 5 0 1 2 2
Stem-Limb 9 0 1 0 2
Types
Typical 9 0 1 3 3
Witches' Broom 5 0 1 0 2
bTFI/ 3 0 1 1 1
Fat 5 0 0 1 1
Thin 1 2 0 01
Sunken 2 1 0 0 4


A~/Basally truncated fusiform gall.



combinations) different analyses for mean number of galls per tree (all

types and locations). Crossing set four had significant F-values for










GCA x S for every gall type or location in at least one analysis, which

is approximately half of the significant GCA x S effects. Confounding

pathogenic variation with environmental variation in the site effect

obscured the cause of the significant GCA x S effect.

Site effects were significant in 40 of the 90 analyses of the

proportion of trees having the various gall types or locations

(Table 15). Two-thirds of the analyses for proportion of trees with



Table 15. Significant site (S), replication (Reps), general
combining ability (GCA), specific combining ability (SCA)
and interaction (GCA x S) F values (5% level of
significance) obtained from analysis of variance using
DIALL for the proportion (arcsin square root
transformation) of slash pine trees with various rust symptoms planted at several sites (nine analyses for
each symptom).


Source of Variation
Site Reps GCA SCA GCAxS
Galls number of significant F values-----Location
All 7 1 4 0 2
Stem 6 0 1 0 2
Limb 6 1 4 1 1
Stem-Limb 8 0 2 0 1
Types
Typical 6 0 4 1 1
Witches' Broom 7 2 1 2 2
BTFI/ 2 0 0 1 1
Fat 4 0 0 1 1
Thin 0 1 0 1 1
Sunken 1 0 0 0 3
Mortality
RAMb-/ 5 0 1 1 2
RBMC_./ 3 0 1 0 2


. /Basally truncated fusiform gall.
b/Rust-associated mortality.

c/Rust bush with many cankers.










limb galls had significant GCA effects. Proportion of trees with stem-to-limb galls was influenced more by site effects than by GCA or SCA. Major effects on the proportion of trees with typical galls were site and GCA, but there was significant interaction between these effects in only one analysis. GCA x S interactions were significant in only 11 of the 90 analyses of proportion of trees with a particular gall type or location.

There were no significant GCA effects for thin, sunken or fat galls (mean or proportion) and only one significant GCA effect for the mean number of BTF galls. These results as well as those in Tables 12 and 13 confirm that there was no additive genetic variation for sunken and fat galls in this study. There was limited evidence of additive genetic variation in thin and BTF galls.

When the analyses were examined to compare significant effects for gall types with those for gall locations, it was found that gall types had more significant SCA effects than locations. GCA x S significant effects were more prevalent among gall types than gall locations. There were, however, several significant GCA x S effects for gall location. The impact of pathogenic variability and environmental variation on family performance was evident in these results. However, these tests were not designed to determine which of these aspects of site were most important.

The lack of consistency of significant differences found in this study was reflected in Blair's (1970) analysis of loblolly pine. He found significant differences among half-sib families in his 1963 plantings but not the 1964 plantings. Disease incidence in the 1963 plantings averaged 50% and in 1964, 41%. Although he










analyzed the number of galls per tree, he did not present the results of that analysis. He stated that the degree of additive genetic control (percentage incidence, galls per tree and severity index) ranged from weak to moderate and depended on the trait measured, site and year of planting. f Rockwood and Goddard (1973) found that in some progeny tests there were significant differences among half-sib families for percentage rust incidence while in others there were none. Differences among families were highly significant when progeny tests were combined. Barker (1973) also found significant and non-significant differences among families for his disease severity index depending on test site. When data from sites were combined there were significant differences among families.

Sohn (1977) found significant differences among half-sib

families of slash pine planted in several locations in two different years. Kinloch and Stonecypher (1969) found significant differences among control- and wind-pollinated families of loblolly pine at all sites they analyzed for both average number of galls per tree and percentage rust incidence.

Barker's (1973) combined analysis for severity index had a

highly significant interaction between sites and crossing sets, but a nonsignificant interaction between sites and families within crossing sets. He found little or no evidence of genotype x environment interaction.

Sohn (1977) found significant family by site interactions for percentage incidence with a binominal analysis of individual trees










and an analysis of plot means, but not with transformed plot means. In addition, the significant differences occurred only in tests planted in 1971, suggesting that planting years were important. Sohn (1977) found rust incidence varied at some sites from year to year, and genetic variances increased with increases in incidence. Conditions favorable to rust infection varied from year to year, resulting in variation in expression of genetic effects.

Results obtained from this study confirm the mixture of

previous results. Families were different in their resistance to fusiform rust infection. Some symptoms of the disease including thin, fat, BTF and sunken galls showed little or no additive genetic effects. RustL incidence was affected by site, year of planting and families planted. Witches' broom, typical, total number, RAM, RBMC, stem, limb and stern-to-limb galls manifested variation because of family and site. Analysis of transformed plot means reduced the number of GCA x S interactions just as Sohn (1977) found.

Individual tree heritabilities for rust incidence, RBMC and RAM as well as mean number of total, typical, BTF, fat, sunken, stem, limb and stem-to-limb galls were calculated (Table 16) using the model found in Table 6. Individual and family heritabilities for fat, thin and sunken galls were estimated as zero because variance components for males were zero or negative. This was expected based on the results of the diallel analysis. Estimates of h 2for percentage rust incidence and total number of galls per tree are comparable to those of Blair (1970) for loblolly pine. His estimates of h 2 for percentage rust incidence were 0.20 and
I










Table 16.


Individual (hi) and family heritabilities (hf) and their standard errors (s) for rust incidence, rust bush many cankers (RBMC), rust-associated mortality (RAM) and the mean number per tree of 10 fusiform rust gall types for slash pine, four- to six-years-old .


Heritability

Trait h2 s s
I- --s


Rust Incidencea/ RBMCa/ R AM./

Location of Galls

Total Stem Limb

Stem-Limb

Types of Galls

Typical

B TF/

Witches' Broom

Fat Thin

Sunken


0.196 0.346 0.240



0.227 0.088 0.055 0.157



0.046 0.004 0.128

0.0 0.0 0.0


0.681 0.809 0.751


0.695 0.519 0.397 0.680


0.306 0.048

0.643

0.0 0.0 0.0


A/Binomial analysis of variance. /Basally truncated fusiform gall.









0.12 for two planting years. Additionally his estimates ofh2 for total number of galls per tree were 0.29 and 0.09 for the same years.

Estimates of percentage rust incidence in slash pine are also similar to those obtained in this study. Rockwood and Goddard (1973) estimated h 2 to range from 0.035 to 0.277 in different II

incidence to range from 0.115 to 0.393, but her combined analysis estimates were 0.280 and 0.217 for two planting years. Sohn's combined estimate of family heritability for percentage rust incidence was 0.78, with a range of 0.389 to 0.752.

The highest estimates of heritabilities in this study, both

individual and family, were for rust incidence, RBMC, RAM and total galls per tree. These heritabilities suggested that selection against death and malformation because of Cqf is possible. Gall locations had moderate heritability values. Heritabilities of gall types, with the exception of witches' broom and typical, were very low or zero. Witches' broom and typical galls therefore, were the only gall types with evidence of additive genetic effects based on heritability estimates and the diallel analysis.

Genetic correlations from the model in Table 6 are found in Table 17. Total gall number is highly correlated with RAM (0.99) and RBMC (0.98). The correlation between total gall number and rust incidence was lower at 0.51. Rust incidence was correlated to a lesser degree with RAM and RBMC. The average number of typical galls was positively correlated with incidence (0.71) and negatively















Table 17.


Genetic correlations between various fusiform rust gall types and locations (mean number per tree), rust incidence and mortality (RAM and RBMC) for slash pine, four- to


six-years-old.


Stem- Witches Rust
Gall Types,/ Stem limb Limb Typical Thin BTF Sunken Broom Fat Incidence RAM RBMC


Total
Stem Limb Stem-Limb
Typical Thin BTF Sunken
Witches' Broom
Fat Rust Incidence RAM


0.60 -0.86 -0.43
0.32 0.25
0.87


-1 .05 0.28 0.96 0.97


_b/ -2.61
- -0.83
- 3.33
- 1.63
- 3.12


0.24 0.86 1.18 0.84
1.24


0.51
- 0.88
- 0.75
- 0.67
- 0.71


0.99
0.52
-0.95
-0.52
-1 .08


2.33 - 0.62 -2.96


- 0.94 -0.03


0.98
0.37
-0.96
-0.59
-1.11

-2.82

-0.12


0.25 0.19
1 .00


a/BTF = Basally truncated fusiform gall; RAM =Rust-associated mortality; RBMC =Rust bush, many
cankers.


b/Undefined, at least one variance component was zero.










correlated with RAM (-1.08) and RBMC Some genetic

correlations were larger than one or smaller than negative one. Although these values are conceptually impossible they are computationally feasible. Genetic correlations are usually subject to large sampling errors and are therefore, seldom very precise (Falconer 1960).

On the basis of genetic correlations, field counts of total number of galls or the number of typical galls would provide more information about potential RAM and malformation than simply recording presence or absence. Disease severity estimates such as Blair's (1970) severity index incorporate some of these features.

General combining ability estimates (gca) obtained from full-sib progeny tests are an indication of parental value. Table 18 contains gca estimates of parental values for several gall types. Parents 0047, 0048, 0098, 0287 and 0352 were the best general combiners for rust resistance and parents 0019, 0064 and 0350 were the worst on the basis of the traits in Table 18.

It is interesting to note that University of Florida (UF)

evaluations (Table 4) were high (greater than 1.0) for 0047, 0048 and 0098. However, the evaluation for 0352 was 0.68 and 0287 had a slightly below average (-0.06) evaluation. The UF evaluation for 0287 had a standard error of 0.35 representing a range of -1.28 to 1.54 for its individual test evaluations. Progeny performance in tests reported here indicate parental resistance levels to be higher than previously assessed and, particularly in the case of 0287, may reflect differences in crossing partners as suggested by Powers and Zobel (1978).










Table 18. General combining ability estimates of the proportion of
trees with fusiform rust galls for 25 slash pine parents
based on progeny performance in six locations in
southeast Georgia, aged four to six years.


Witches' Stem- Rust
Parent Typical Thin BTF.2/ Sunk Fat Broom Stem Limb Limb Incidence
----------------------------------- % ------------------------------------1 41.5 2.8 2.3 3.8 1.5 43.0 22.1 45.7 24.7 52.0
19 95.6 6.7 8.5 5.8 1.1 62.4 56.6 96.2 46.1 106.3 47 24.1 2.1 5.5 7.3 4.6 10.7 9.0 22.2 7.3 34.6
48 28.7 1.1 4.2 0.4 -0.3 1.5 -2.3 31.5 6.6 29.9
50 39.4 -1.4 3.3 0.6 5.2 30.6 19.5 32.6 17.6 47.3
60 51.2 6.2 6.6 7.3 0.7 12.8 32.3 46.4 18.3 62.9 64 99.4 1.7 22.2 12.3 16.1 56.9 35.9 89.9 68.7 109.2 05 47.2 0.6 1.3 1.3 0.7 36.8 18.9 48.5 25.0 64.3
71 52.6 -2.7 3.6 2.1 0.5 29.0 23.5 37.0 34.3 66.8
88 33.3 -0.5 3.5 0.1 1.2 16.3 3.6 39.9 5.2 35.7
96 57.0 2.8 5.9 1.6 3.1 24.6 21.9 57.1 20.7 70.4 98 25.9 5.1 4.4 1.6 4.8 -7.2 -2.0 16.0 2.9 31.3
141 72.1 -0.2 3.1 3.1 3.9 48.1 49.8 72.4 45.7 86.4
146 35.2 3.9 3.1 0.1 1.2 37.0 27.0 33.8 23.2 36.1 157 69.6 3,7 12.8 3.8 9.0 30.9 26.7 69.8 31.0 82.1 159 78.7 4.2 0.1 9.1 2.5 49.6 40.0 67.0 30.2 89.2 270 31.9 8.9 2.2 9.0 -4.8 27.6 37.6 38.7 22.1 59.3
284 38.5 1.0 3.8 6.8 -4.7 7.7 41.2 31.1 16.4 50.2
287 12.3 3.1 -1.9 0.2 3.0 8.5 1.2 6.9 13.3 24.0
288 61.4 8.9 6.6 7.5 2.2 50.6 45.1 64.9 36.9 78.0 295 67.8 3.0 4.2 5.8 9.9 46.4 28.7 64.7 53.1 82.5 350 96.1 4.1 5.6 0.8 6.8 57.9 50.3 92.4 53.0 100.8 352 26.9 0.3 7.9 0.9 2.5 15.9 13.4 26.0 11.3 33.8 354 82.4 -0.02 12.2 1.7 7.1 65.7 57.7 81.9 55.6 93.5
355 79.0 3.2 0.7 14.0 3.9 52.2 39.6 71.2 63.5 96.1


/Basally truncated fusiform gall.



Specific combining ability of a particular cross (sca) is

defined as the family deviation from the mean of the two parents.

Specific combining abilities of families for percentage rust

incidence and fat galls are presented in Table 29. Specific

combining ability values for the proportion of trees with fat galls

ranged from -8.00 (0284 x 0064) to 3.84 (0295 x 0064). Crosses with

0064 had the greatest positive and negative deviations from parental

means (sca) for the proportion of trees with fat galls. Parent 0064










had the highest estimate of gca for this trait as well as for proportion of trees with sunken galls and percentage rust incidence. Specific combining ability estimates for families in which 0064 was a parent varied greatly for these last two traits also.

Specific combining ability may be an influence in the

relatively good performance of parent 0287. Crosses between this parent and 0019 or 0060 have sca's of 5.31 and 2.25, respectively, for percentage rust incidence. When crossed with 0141, however, it has a sca of -7.63. Parent 0141 has a gca estimate midway between those of parents 0019 and 0060. Further crosses with 0287 and 0141 and other parents may disclose the reasons for this apparent good combination.

In several of the analyses for fat galls SCA effects were

significant. This fact, along with sca values for fat galls in some crosses, suggested that nonadditive genetic effects were important in this trait.


Artificial Inoculation

Experiment 2--RSC Procedure

Rust incidence (gall formation) following the RSC procedure ranged from 14.4% to 88.7% among the 20 seedling lots tested (17 full-sib families and 3 check lots). Proportion of galled seedlings was higher in all families than in the two resistant checks, and only two families had lower galled proportions approaching that of

-the resistant checks. Seven of the families had a higher properties of galled seedlings than the susceptible check. Means for all traits measured are indicated in Table 19 by seedling lot. A
















Family Pink Incidence Symptom Less than Resistance
Female Male Blush Healthy of Rust No Swelling Rough 25 mm Index
-------------------------- -------------------------3000 1/ - 15.8 24.4 66.2 9.4 11.5 25.9 209.4
40002/ - 5.0 41.7 31.7 29.2 5.0 11.7 195.2
5000-C/ - 1.8 57.6 14.4 31.7 3.3 9.4 144.8
0354 0350 1.6 18.0 63.9 18.0 11.7 36.1 159.3
0354 0088 2.7 15.3 48.9 38.0 10.9 30.2 122.2
0354 0048 4.2 15.5 62.9 28.2 0 25.9 140.4
0350 0088 8.3 20.0 56.0 30.0 8.0 26.0 102.3
0350 0048 19.4 33.8 53.0 15.9 4.6 20.5 171.2
0352 0001 7.3 38.8 34.9 27.2 5.3 9.3 197.1
0352 0159 3.2 34.2 34.9 30.9 11.1 5.3 163.1
0270 0050 41.0 42.6 38.5 13.9 7.8 14.6 183.3
0019 0060 6.5 8.9 88.7 2.4 17.9 23.2 61.3
0019 0287 16.9 18.3 70.7 15.5 6.4 37.9 132.9
0019 0047 44.3 38.0 49.4 17.3 4.7 23.3 157.5
0141 0287 13.3 28.3 58.3 16.7 1.7 33.3 166.6
0270 0355 31.9 16.0 79.8 5.1 17.2 25.7 128.4
0284 0295 12.2 19.6 77.6 4.3 7.2 23.6 108.4
0284 0064 13.4 20.5 70.4 9.1 8.8 25.2 112.5
0270 0295 28.6 13.6 82.2 4.2 23.8 22.7 142.1
0270 0064 39.1 8.9 88.7 1.8 22.7 16.7 149.7


RSC susceptible check lot, Georgia Slash

RSC resistant check lot, FA-2

RSC resistant cheCK lot, LA11


resistance index value was calculated for the seedling lots tested

based on the RSC index equation (Robert Anderson, personal

communication 1984) (Table 19). A high resistance index value

indicates a high occurrence of resistant reactions. The high index

value for the susceptible check was unexpected.

Families were significantly different for all traits analyzed

(Table 20) (see model, Table 7). Differences among families for


Table 19.


Percentage of seedlings in 20 slash pine families with various reaction types six months after artificial inoculation with fusiform rust and the resistance index of the families.









Table 20. Mean squares for the analyses of variance of the
proportion of of slash pine seedlings per family
expressing various disease symptoms resulting from
artificial inoculation with fusiform rust at the
Resistance Screening Center, Asheville, North Carolina.


Source of Variation
Disease Symptom Run Family


Healthy 0.165** 0.088**
SYMNOal 0.006 0.122**
Incidence of Rust 0.306** 0.400**
Rough Gall 0.001 0.029**
< 25 mm Gall 0.134** 0.lI0**
Thin Gall 0.I00** 0.042**
Fat Gall 0.164** 0.011*
BTF Gallb/ 0.002 0.001*
Adventitious Shoots 0.103** 0.137**
Sunken Areas 0.006 0.056*


-!/Symptom, no swelling.

P/Basally truncated fusiform gall.
*Significant at the 5% level.

**Significant at the 1% level.


mean number of fat and BTF galls and galls with sunken areas were significant at the 5% level while other traits were highly significant (1% level) (Table 20).

Carson (1984) analyzed five of these traits (galled, SYMNO,

rough, LT25M ana thin) on loblolly pine. With a composite inoculum source, there were significant differences among families for four of these but not for the fifth, LT25M.

Individual heritabilities were calculated using the model described on page 28 for the proportion of seedlings with the various artificial inoculation responses and the mean number of










responses per tree (Taole 21). Proportion of trees with SYMNO and galls with adventitious shoots had fairly high individual heritabilities, 0.28 and 0.33, respectively. Mean number of galls per seedling had the highest heritability with a value of 0.38. These results indicated that several resistant and susceptible characteristics are heritable traits.

Correlations between variables in the greenhouse produced

several significant observations (Table 22). The number of healthy seedlings in a family had significant negative correlations with mean number of galls, rough galls and galls with adventitious shoots, and was negatively related to other reactions except SYMNO. In general, it should be expected that symptoms suggesting resistance (such as SYMNO) should be positively correlated with other resistance symptoms and negatively correlated with symptoms suggesting susceptibility. Galls with adventitious shoots had a highly significant negative correlation with healthy and SYMNO and highly significant positive correlations with number of galls and rough galls. Galls with adventitious shoots appeared to be a susceptible reaction type. Sunken areas were negatively correlated with SYMNO and positively correlated with BTF, rough, and mean number of galls.









Table 21. Individual heritabilities (h2) and their standard
errors (s) for artificial inoculation (Al) symptoms
by both proportion of trees with the symptoms and
mean number of symptoms per tree.


Proportion Mean

Al Symptoms/ h2 s h2 s


Healthy 0.08 0.08 -

SYMNO 0.28 0.18 0.32 0.20

Incidence of Rust 0.23 0.16 0.38 0.23

Rough Gall 0.18 0.13 0.11 0.09

LT25M Gall 0.10 0.09 0.08 0.08

Thin Gall 0.06 0.07 0.06 0.07

Fat Gall 0.02 0.05 0.01 0.04

BTF Gall 0.05 0.06 0.05 0.06

Adventitious Shoots 0.33 0.21 0.32 0.20

Sunken Areas 0.11 0.09 0.09 0.08


A/SYMNO =symptom, no swelling; LT25M =gall less
BTF =basally truncated fusiform gall.


than 25 mm long;


The correlation between SYMNO and healthy was small (less than 0.4) and not significant. Healthy seedlings may be escapes or may have one of the two interactions Miller et al. (1976) described as either total incompatibility (i.e., no germination or penetration by basidiospore) or subliminal infection possibly due to an avirulent pathogen. SYMNO appeared to be a hypersensitive reaction.











Table 22. Pearson product-moment correlations between the mean number of artificial inoculation (AI) symptoms per tree of
slash pine seedlings inoculated with fusiform rust, based
on progeny of 20 families.


Al Symptoms
Al Symptom, Incidence aAdventitious Sunken
Symptoms no swelling of Rust Rough LT25M!/ Thin Fat BTF-/ Shoots Area
---------------------Galls ------------------------------


Healthy 0

Symptom, no swelling Incidence of Rust Rough Galls LT2 5r4a/ Galls Thin Galls Fat Galls BTF / Galls Adventitious Shoots


).35 -0.86** -0.52*

-0.70** -0.54*


-0.49

0.01


-0.54* 0.52*


-0.22

-0.41


-0.25

0.11


0.53* 0.14


-0.07 0.01

0.61*


-0.13

0.10

0.13


-0.33

-0.48


0.45 0.59*

0.08 0.28

-0.01


-0.65**

-0.7 9**


-0.37
-0.54"


0.78** 0.45*


0.67**

-0.03

0.13 0.09 0.42


0.70**

0.06 0.18

-0.39

0.63**


0.38


a/Gall less than 25 mm long. b/Basally truncated fusiform gall.

* Correlation significant at the 5% level.

**Correlation significant at the 1% level.










There were significant negative correlations between percentage rust incidence in the field and proportion of healthy and SYMNO seedlings in the greenhouse (Table 23). Field rust incidence had significant positive correlations with proportion of seedlings infected with any gall type, rough galls and galls with adventitious shoots. The proportion of trees with stem, limb, stem-to-limb and typical galls in the field were significantly negatively correlated with the proportion of healthy seedlings in the greenhouse. There were no significant correlations between RAM and RBMC and any of the artificial inoculation symptoms.

The positive correlations between the proportion of trees galled in the field and rough (significant) and LT25M (nonsignificant) galls in the greenhouse were surprising as both symptoms are considered resistant (Charles Walkinshaw, personal communication 1984). Walkinshaw and others (1980) found a negative correlation bdcween University of Florida rust evaluations and the proportion of seedlings with smooth galls. Therefore, they included this variable as well as proportions of trees with fat galls and SYMNO in their index. Rough galls, seemingly the opposite of smooth galls, were categorized by Carson (1984) and by RSC for their current index (Robert Anderson, personal communication 1985).

LT25M had a positive relationship with percentage rust

incidence and total number of galls per tree in the field study. Seedlings with LT25M% galls nine months after inoculation in the 1984 study of Griggs and others were free of gall symptoms and active












Table 23. Pearson product-moment correlations between field galls and artificial inoculation symptoms (proportions) and the
resistance index of 17 slash pine families and fusiform
rust.


Artificial Inoculation Symptomsl/

Incidence Adventitious Resistar
Field Galls Healthy SYMNO of Rust Rough LT25M Fat Shoots Inde,

Proportion
of Trees:
Typical -0.54* -0.27 0.50* 0.32 0.35 0.38 0.44 -0.44
Thin -0.32 0.64** 0.57* 0.63** 0.03 -0.19 0.52* -0.2c
BTF_/ -0.20 -0.01 0.14 -0.03 0.04 0.47 0.27 -0.1
Sunken -0.24 -0.70** 0.55* 0.64** -0.22 0.02 0.61** -0.1I
Fat -0.25 -0.25 0.01 0.02 0.12 0.57* 0.07 -0.05
Witches' Broom -0.29 -0.01 0.14 0.38 0.36 -0.24 -0.03 0.11
Stem -U0.52* -0.48 0.58* 0.46 0.42 -0.06 0.48 -0.3;
Limb -0.60* -0.18 0.48 0.34 0.39 0.30 0.36 -0.42
Stem-Limb -0.54* -0.48 0.60* 0.46 0.42 -0.12 0.42 -0.1.
Inciaence
of Rust -0.63** -0.51* 0.69** 0.54* 0.37 0.21 0.59* -0.4!
RHMf/ -0.16 -0.03 0.09 0.13 0.34 -0.11 0.00 0.0(
RBMCc/ -0.13 0.11 -0.01 0.06 0.41 -0.16 -0.13 0.0O

lean number
of galls:
Typical -U.34 -0.07 0.22 -0.01 0.16 0.68** 0.09 -0.3(
Thin -0.18 -0.61** 0.48 0.50 0.06 -0.28 0.45 -O.1;
BTFR/ -0.14 0.29 -0.03 -0.22 0.09 0.57* -0.08 -0.1,
Sunken -0.30 -0.75** 0.61** 0.76** -0.24 0.07 0.68* -0.1'
Witches' Broom -0.24 -0.05 0.08 0.34* 0.25 0.01 -0.06 0.1l
Fat -0.16 0.39 -0.11 0.13 0.19 0.56* -0.32 -0.0;
stem -0.20 0.02 0.08 0.11 0.44 -0.20 -0.05 -0.0
Limb -0.27 0.20 0.08 0.01 0.12 0.68** 0.05 -0.3(
Stem-Limb -0.53* -0.48* 0.62** 0.30 0.38 0.08 0.42 -0.1:
Total -0.29 -0.05 0.14 0.12 0.44 0.05 -0.02 -0.0�


a/SYMNO = symptom, no swelling, LT25M = gall less than 25 mm long.

0/Basally truncated fusiform gall.

S/RAM = rust-associated mortality, RBMC = rust bush many cankers.

*Correlation significant at the 5% level.

**Correlation significant at the 1% level.










mycelium 23 months later, and therefore, could be considered a resistant symptom in that study.

There were no significant correlations between field symptoms and the RSC resistance index in this study. Calculation of the correlations did not include the check lots. The resistance index did not successfully reflect family field performance in this study for any field symptom.

Griggs and Dinus (1977) found extremely low correlations

between artificial inoculation and field tests at young ages, but the correlations increased and were significant at age 12. They implied that the correlation was improved after incidence had increased over time. This suggested that artificial inoculation predicted performance better on sites with high incidence than on sites with moderate incidence. Because field tests in this study had moderate to high incidence levels it was expected that artificial inoculation would predict performance more accurately.

Pink blush was observed on the seedlings 10 days after inoculation. Families were significantly different in the percentage of seedlings expressing this character. Percentage of seedlings with a pink blush ranged from 1.6 to 44.3 (Table 19). In some cases the blush was a faint, small pink spot on the stem just above the cotyledons. In other cases the spot was larger, sometimes encircling the stem, or varied in color to a rich red. Pink blush was negatively correlated with SYMNO (r=-0.53) and essentially not correlated with percentage healthy trees (r=-0.04). It was










positively related to percentage of galled seedlings (r=0.30). Pink blush does not appear to be a resistant symptom. However, a more intensive study is needed to conclude this. Experiment 3--High Inoculum Concentration

Numbers of spots on each seedling were counted 10 and 48 days after inoculation (Table 24). Red spots observed 10 days after inoculation were generally small and confined to the stem. After 48 days, some spots were very large and numerous spots were present on the primary needles. Large spots in some cases appeared sunken. Color varied from red to purple. There were highly significant differences among families for spots both 10 and 48 days after inoculation. Lundquist and Luttrell (1982) stated that most families became pigmented within 48 days. Therefore, the results at 48 days for proportion of trees with spots were expected.

Average number of spots per seedling 10 and 48 days after

inoculation was correlated to field performance of the same families using Pearson correlations (SAS Institute Inc. 1982a). Significant negative correlations were found between proportion of trees with spots after 10 days and proportion of trees with typical galls, limb galls and rust incidence (r > -0.59).

Lundquist et al., (1982) found that under high inoculum

concentrations resistance was ineffective in family FA-2, a RSC resistant check lot. This family (4000) was included as a check lot in both this experiment and Experiment 2. Very little pigmentation occurred in the test, supporting the conclusion of Lundquist et al. (1982), that resistance of this family was not effective under high









Table 24. Results of artificial inoculation with high
concentrations of inoculum (2.0 x 106 spores/mi) on 17 slash pine families, Resistance Screening Center,
Asheville, North Carolina.


10 Days after Inoculation 48 Days after Inoculation
Family No. of % of Trees No. of % of Trees
Female Male Spots Mean With Spots Spots Mean with Spots


3000-a - 0 0 0 7 0.4 25.0
4000b. - 0 0 0 4 0.2 20.0
0354 0088 1 0.05 5.0 299 14.9 100.0
0098 0088 12 0.63 42.1 281 14.8 94.7
0354 0048 1 0.05 5.0 110 5.5 90.0
0352 0001 10 0.5 45.0 412 20.6 95.0
0288 0159 0 0 0 130 6.5 100.0
0019 0047 1 0.05 5.0 154 7.7 100.0
0350 0048 1 0.5 5.0 404 20.2 100.0
0270 0050 70 3.5 85.0 362 18.1 100.0
0141 0287 4 0.2 15.0 152 7.6 95.0
0350 0354 0 0 0 154 7.7 100.0
0019 0060 5 0.25 10.0 148 7.4 85.0
0096 0146 34 1.7 65.0 325 16.2 95.0
0019 0287 1 0.05 5.0 79 3.9 80.0
0352 0159 13 0.93 64.3 134 9.6 92.9
0270 0064 6 0.3 25.0 230 11.5 100.0


A/RSC susceptible check lot, Georgia Slash.

P./RSC resistant check lot, FA-2.



inoculum loads. Lundquist and Miller (1984) compared their results

with Mullick (1977) and other researchers. They reported that the

pigmentation seen in their study signified that a host had reacted

to prevent the spread of the fungus. They further stated that

pigmentation was not the complete mechanism of resistance, but was

symptomatic of physiological and anatomical resistance mechanisms.

Further research is needed to understand these mechanisms before


breeders can exploit the variation that is present.










Experiment 4--Inoculation of 1-0 Seedlings

Galls developing on 1-0 seedlings more closely resembled galls on trees in the field than those on the seedlings in Experiment 2. This suggested that host response varied because of maturation state as concluded by Day (1974). Families were significantly different for rust incidence, total number of galls per tree, fat galls, thin galls, and typical galls. Galls were located on the growth flush that had occurred just prior to inoculation. Many infected trees had stem-to-limb galls that probably would result in a RBMC if the trees were allowed to mature. Goddard and Schmidt (1971) also found differences among families in rust incidence when they inoculated 1-0 seedlings, and their results were correlated with the results of inoculation of the same seedlings at age six-to-ten weeks.

The susceptible check had significantly greater rust incidence than the other families (Table 25). Family 0621 x 2005 had the next highest infection level. Many galls on these trees were short and fat and did not encompass the entire stem circumference. Orthogonal contrasts revealed that for total galls per tree the susceptible check was significantly different from 0621 x 2005 and 0298 x 2005 and that these families also differed. In addition, the contrasts disclosed differences in rust incidence between 0621 x 2005 and 0298 x 2005. Male 2005 was apparently susceptible.

This small sample size and lack of field data limit conclusions from this study. Field performance in Florida of these families should not be predicted from these results since the pathogen was not local. However, it can be seen that seedlings from the










Table 25.


Comparison of family means for several fusiform rust disease symptoms on artificially inoculated 1-0 slash pine seedlings.


Family Disease Symptoma/

Female Male % Rust Incidence Total Fat Thin Typical
--------galls per tree -------Susceptible Checkb/ 72.5A 1.95A 0.03B 0.2A 1.7A

0621 2005 52.5B 1.38A 0.18A 0.08 1.05B

0298 2005 22.5C 0.45B 0.03B O.0B 0.38C

Resistant Check'/ 28.9C 0.32B O.0B 0.0B 0.29C

0618 2007 17.5C 0.18B 0.OB O.0B 0.18C

0620 2002 17.4C 0.17B 0.OB 0.OB 0.13C

0298 2013 11.4C 0.17B O.OB 0.08 0.17C

0298 2002 10.5C 0.13B 0.05B O.OB 0.05C


.2/Means with the same letter are not significantly different. !/Checks were Container Corporation of America's resistant and
nonresistant orchard progeny.


resistant orchard check and some crosses with phenotypically resistant males had significantly less infection than the susceptible check. The study also demonstrates that older seedlings may be effectively artificially inoculated with Cqf at RSC. This may be useful in future studies with slash pine and Cqf.










Research Perspectives

Disease results from the interaction of a host, a pathogen, and their environments. Although fusiform rust disease of slash pine has been studied intensively for the last 20 years, many questions remain unanswered.

It is known that the fungus varies in its virulence. Carson (1984) found disease symptoms varied within loblolly pine families when different pathogen sources were used to inoculate seedlings. It can be concluded, therefore, that disease symptoms may vary because of the pathogen. However, additional research is necessary to corroborate these findings for slash pine and other species.

Environmental differences, such as site or year, cause

variation in rust incidence by affecting sporulation and infection. However, the impact of environment on variation in disease symptoms is unclear. Field disease symptoms in this study varied by site, but the fungal component was confounded with environment. No clear statement was possible of the effect of environment on disease symptoms. One method of addressing environmental effects may be to inoculate 1-0 seedlings as in Experiment 4 and then outplant them at different sites.

Future studies could include development of the relationship between microscopic processes described by Miller et al. (1976) and Mullick (1977) and gall types. Griggs et al. (1984) observed that LT25M galls disappeared over time. What of other gall types? Do they differ or change over time? Does the environment affect gall







69


development? Research aimed at answering these questions may add valuable information to the knowledge of disease resistance in slash pine.














CONCLUSIONS


Additive genetic variation of slasn pine in response to natural infection with fusiform rust exists for many symptoms: number of stem, limb, stem-to-limb, total, typical and witches' broom galls, the proportion of trees with these galls, incidence of rust, RAM and RBMC. However, this variation is not consistent. Little or no genetic variation is apparent for thin or sunken galls. Fat galls have evidence of specific combining ability, an expression of nonaduitive genetic effects. The existence of genetic variation appears to depend on site, planting year, inoculum and the sample of families in a test.

Individual tree and family heritabilities for the field traits--rust incidence, RAM, RBMiC, total galls per tree, stem-to-limD galls and witches' broom galls--were relatively high and, where comparisons were possible, similar to those found in other studies. Individual heritabilities were 0.196, 0.346, 0.240,

0.227, 0.157 ana 0.128, respectively. Family heritaoilities were

0.68, 0.81, 0.75, 0.70, 0.68 and 0.64, respectively.

KAM and RbMC were the most severe reactions to rust infection and were the most heritable in this study. Selection against families with high incidence of these traits is imperative in breeding for increased rust resistance.










Individual heritabilities of artificial inoculation symptoms were best for proportion of trees with SYMNO (0.28), rough galls (0.18), galled (any type) (0.23) and galls with adventitious shoots (0.33). Variation among full-sib families existed for all traits observed. Healthy, SYMNO and galled (any type) were correlated with several field symptoms but were more often significantly correlated with the proportion of trees with a field symptom than with the mean number of symptoms. Rough galls, considered to be a resistant symptom, had a significant positive correlation with percentage rust incidence in the field. The RSC resistance index did not successfully predict field performance of the full-sib families in this study.

The high inoculum concentration technique was not successful in predicting field performance. However, it may be useful in screening families whose resistance could be overcome in the field with high inoculum concentrations. Significant differences among families existed for their response to this inoculation procedure.

One-year-old progeny of phenotypically rust-resistant males

were successfully inoculated in this study. Families with three of the four males as a parent were as resistant as a rust-resistant check. The fourth male was apparently not resistant. Selection of phenotypically rust-resistant slash pines, therefore, continues to be a viable method of adding resistant selections to a breeding program.













LITERATURE CITED


Anderson, R. L., and T. A. Bancroft. 1952. Statistical Theory in
Research. McGraw-Hill, New York. 399 p.

Anderson, R. L., C. H. Young and J. D. Triplett. 1983. Resistant
screening center procedures manual: A step-by-step guide used
in the operational screening of southern pines for resistance
to fusiform rust. USDA Forest Serv Forest Pest Manag Rep
83-1-18. 55 p.

Blair, R. L. 1970. Quantitative inheritance of resistance to
fusiform rust in loblolly pine. Ph D Thesis, North Carolina
State Univ, Raleigh. 87 p.

Barker, J. A. 1973. Location effects on heritability and gain
predictions for ten-year-old loblolly pine. Ph D Thesis, North
Carolina State Univ, Raleigh. 105 p.

Carson, S. D. 1984. Indirect screening of loblolly pine for
fusiform rust resistance through controlled inoculation. Ph D
Thesis, North Carolina State Univ, Raleigh. 57 p.

Day, P. R. 1974. Genetics of Host Parasite Interaction. W. H.
Freeman and Company, San Francisco, CA. 288 p.

Dinus, R. J., G. A. Snow, A. G. Kais and C. H. Walkinshaw. 1975.
Variability of Cronartium fusiforme affects resistance breeding
strategies. In Proceedings 13th Southern Forest Tree Improvement Coference, p 193-196. National Technical
Information Service, Springfield, VA. 262 p.

Dwinell, L. D. 1974. Susceptibility of southern oaks to Cronartium
fusiforme and Cronartium quercuum. Phytopathology 64:400-403.

Dwinell, L. D. 1977. Biology of fusiform rust. In Management of
Fusiform Rust in Southern Pines (R. J. Dinus and R. A. Schmidt
eds) p 18-24. Symp Proc Univ Florida, Gainesville. 163 p.









Falconer, D. S. 1960. Introduction to Quantitative Genetics. The
Ronald Press Company, New York. 365 p.

Frampton, L. J., Jr., H. W. Amerson and R. J. Weir. 1983.
Potential of in vitro screening of loblolly pine for fusiform
rust. In Proceedings 17th Southern Forest Tree Improvement
Conference, p 325-331. National Technical Information Service,
Springfield, VA. 375 p.

Goddard, R. E., and J. T. Arnold. 1966. Screening select slash
pines for resistance to fusiform rust by artificial
inoculation. In Breeding Pest Resistant Trees (H. D. Gerhold,
E. J. Schreiner, R. E. McDermott, and J. A. Winieski, eds) p
431-435. Pergamon Press Ltd, London. 505 p.

Goddard, R. E., D. L. Rockwood and H. R. Kok. 1984. Cooperative
Forest Genetics Research Program. 26th Prog Rep, Univ Florida,
School Forest Resourc Conserv Rep No 35. 26 p.

Goddard, R. E., and R. A. Schmidt. 1971. Early identification of
fusiform rust resistant slash pine families through controlled
inoculation. In Proceedings l1th Southern Forest Tree
Improvement Conference, p 31-36. National Technical
Information Service, Springfield, VA. 284 p.

Goddard, R. E., R. A. Schmidt, R. K. Strickland and H. R. Kok.
1972. Cooperative Forest Genetics Research Program. 14th Prog
Rep Univ Florida, School Forest Resourc Conserv Rep No 20. 26
P.

Goddard, R. E., R. A. Schmidt and F. Vande Linde. 1975. Effect
of differential selection pressures on fusiform rust resistance
in phenotypic selection of slash pine. Phytopathology
65:336-338.

Goddard, R. E., and 0. 0. Wells. 1977. Susceptibility of
southern pines to fusiform rust. In Management of Fusiform Rust in Southern Pines (R. J. Dinus and R. A. Schmidt, eds),
p 52-58. Symp Proc Univ Florida, Gainesville. 163 p.

Gray, D. J., and H. W. Amerson. 1983. In vitro resistance of Pinus
taeda embryos to the fusiform fungus, Cronartium quercuum f.
sp. Tusiforme: Ultrastructure and histology. Phytopathology
73:1492-1493.

Griggs, M. M., and R. J. Dinus. 1977. Patterns of fusiform rust
increase and their implications for selection and breeding. In
Proceedings 14th Southern Forest Tree Improvement Conference, p 43-52. National Technical Information Service, Springfield,
VA. 352 p.










Griggs, M. M., R. J. Dinus and G. A. Snow. 1984. Inoculum source and density influence assessment of fusiform rust resistance in
slash pine. Plant Dis 68:770-774.

Griggs, M. M., and R. A. Schmidt. 1977. Increase and spread of fusiform rust. In Management of Fusiform Rust in Southern
Pines (R. J. Dinus and R. A. Schmidt, ed.), p 32-38. Symp Proc
Univ Florida, Gainesville. 163 p.

Griggs, M. M., and C. H. Walkinshaw. 1982. Diallel analysis of genetic resistance to Cronartium quercuum f. sp. fusiforme in
slash pine. Phytopathology 72:816-818.

Hirt, R. R. 1964. Cronartium ribicola: Its growth and
reproduction in the tissues of eastern white pine. St Univ Col
Forest Syracuse Univ, NY, Tech Pub 86. 59 p.

Hubbard, S. D., and R. L. Anderson. 1980. The resistance screening center. USDA Forest Serv Gen Rep SA-GR 16. 1 p.

Jewell, F. F. 1960. Inoculation of slash pine with Cronartium fusiforme. Phytopathology 50:48-51.

Jewell, F. F., D. C. Jewell and C. H. Walkinshaw. 1982.
Histopathology of anatomical mechanisms for resistance to
fusiform rust in slash pine. In Resistance to Diseases and
Pests in Forest Trees (H. M. Heybroek, B. R. Stephan and K. von
Weissenberg, eds.) p 110-118. Centre Agric Publ Document,
Wageningen, The Netherlands. 503 p.

Kinloch, B. B., Jr., and R. W. Stonecypher. 1969. Genetic variation in susceptibility to fusiform rust in seedlings from
a wild population of loblolly pine. Phytopathology
59:1246-1255.

Laird, P. P., and W. R. Phelps. 1975. A rapid method for mass
screening of loblolly and slash pine seedlings for resistance
to fusiform rust. Plant Dis Rep 59:238-242.

Lewis, R. A. 1973. Quantitative assessment and possible
biochemical indicators of variation in resistance to fusiform
rust in loblolly pine. Ph D Thesis, North Carolina State Univ,
Raleigh. 94 p.

Lundquist, J. E., and E. S. Luttrell. 1982. Early symptomatology
of fusiform rust on pine seedlings. Phytopathology 72:54-57.

Lundquist, J. E., and T. Miller. 1984. Development of stem lesions
on slash pine seedlings infected by Cronartium quercuum f. sp.
fusiforme. Phytopathology 74:514-518.










Lundquist, J. E., T. Miller and H. R. Powers, Jr. 1982. A rapid
technique for determining resistance of slash pine to fusiform
rust. Phytopathology 72:613-615.

Matthews, F. R., and S. J. Rowan. 1972. An improved method for
large-scale inoculations of pine and oak with Cronartium
fusiforme. Plant Dis Rep 56:931-934.

Miller, T., E. B. Cowling, H. R. Powers, Jr. and T. E. Blalock.
1976. Types of resistance and compatibility in slash pine
seedlings infected by Cronartium fusiforme. Phytopathology
66:1229-1235.

Mullick, D. B. 1977. The nonspecific nature of defense in bark
and wood during wounding, insect, and pathogen attack. In The
Structure, Biosynthesis, and Degradation of Wood. Vol 1TRecent Advances in Phytochemistry (F. A. Loewus and C. C.
Runeckles, eds) p 395-441. Plenum Press, New York. 521 p.

Powers, H. R., Jr. 1980. Pathogenic variation among singleaeciospore isolates of Cronartium quercuum f. sp. fusiforme.
Forest Sci 26:280-282.

Powers, H. R., Jr., and L. 0. Dwinell. 1978. Virulence of
Cronartium stable after 25 years. Plant Dis Rep 62:877-879.

Powers, H. R., Jr., F. R. Matthews and L. D. Dwinell. 1977.
Evaluation of pathogenic variability of Cronartium fusiforme on loblolly pine in the southern USA. Phytopathology 67:1403-1407.

Powers, H. R., Jr., F. R. Matthews and L. D. Dwinell. 1978. The
potential for increased virulence of Cronartium fusiforme on
resistant loblolly pine. Phytopathology 68:808-810.

Powers, H. R., Jr., R. A. Schmidt and G. A. Snow. 1981. Current
status and management of fusiform rust on southern pines. Ann
Rev Phytopathol 19:353-371.

Powers, H. R., Jr., and B. J. Zobel. 1978. Progeny of specific
loblolly clones vary in fusiform rust resistance according to
seed orchard of origin. Forest Sci 24:227-230.

Rockwood, D. L., and R. E. Goddard. 1973. Predicted gains for
fusiform rust resistance in slash pine. In Proceedings 12th
Southern Forest Tree Improvement Conference, p. 31-37. National
Technical Information Service, Springfield, VA. 352 p.

SAS Institute Inc. 1982a. SAS User's Guide: Basics, 1982
edition. SAS Institute, Inc, Cary, North Carolina. 923 p.









SAS Institute Inc. 1982b. SAS User's Guide: Statistics, 1982
edition. SAS Institute, Inc, Cary, North Carolina. 584 p.

Schaffer, H. E. and R. A. Usanis. 1969. General least squares
analysis of diallel experiments: A computer program - DIALL.
Genetics Dept Res Rep No 1, North Carolina State Univ,
Raleigh. 61 p.

Schmidt, R. A. 1972. A literature review of inoculation techniques
used in studies of fusiform rust. In Biology of Rust
Resistance in Forest Trees (R. T. BTngham, R. J. Hoff and G. I.
McDonald, eds). USDA Forest Serv Misc Publ 1221:341-356.

Schmidt, R. A., R. E. Goddard and C. A. Hollis. 1974. Incidence
and distribution of fusiform rust in slash pine plantations in Florida and Georgia. Instit Food and Agric Sci, Univ Florida
Bull 763. Gainesville. 21 p.

Schmidt, R. A., R. C. Holley and M. C. Klapproth. 1985. Results
from operational plantings of fusiform-rust-resistant slash and
loblolly pines in high-rust-incidence areas in Florida and Georgia. In Proceedings of the Rusts of Hard Pines Working Party Conference (J. Barrows-Broaddus and H. R. Powers, Jr.,
eds) p 33-41. The Georgia Center for Continuing Education,
Univ Georgia, Athens. 331 p.

Schmidt, R. A., H. R. Powers, Jr. and G. A. Snow. 1981.
Application of genetic disease resistance for the control of
fusiform rust in intensively managed southern pine.
Phytopathology 79:993-997.

Snow, G. A., R. J. Dinus, and A. G. Kais. 1975. Variation in
pathogenicity of diverse sources of Cronartium fusiforme on
selected slash pine families. Phytopathology 65:170-175.

Snow, G. A., R. J. Dinus, and C. H. Walkinshaw. 1976. Increase in
virulence of Cronartium fusiforme on a resistant slash pine.
Phytopathology 66:511-513.

Snow, G. A., and M. M. Griggs. 1980. Relative virulence of
Cronartium quercuum f. sp. fusiforme from seven resistant
families of slash pine. Phytopath Medit 19:13-16.

Snow, G. A., and A. G. Kais. 1970. Pathogenic variability in
isolates of Cronartium fusiforme from five southern states.
Phytopathology 60:1730-1731.

Snow, G. A., W. L. Nance and E. B. Snyder. 1982. Relative
virulence of Cronartium quercuum f. sp. fusiforme on loblolly
pine from Livingston parish. In Resistance to Diseases and
Pests in Forest Trees (H. M. Heybroek, B. R. Stephan and K. von
Weissenberg, eds), p 243-250. Centre Agric Publ Document,
Wageningen, The Netherlands. 503 p.










Sohn, S. 1977. Heritability of resistance to Cronartium fusiforme
in Pinus elliottii as affected by disease incidence and a
comparison of breeding procedures. Ph D Dissertation, Univ of
Florida, Gainesville. 92 p.

Sohn, S., and R. E. Goddard. 1979. Influence of infection percent
on improvement of fusiform rust resistance in slash pine.
Silvae Genetica 28:173-180.

Squillace, A. E., R. J. Dinus, C. A. Hollis and R. A. Schmidt.
1978. Relation of oak abundance, seed source, and temperature
to geographic patterns of fusiform rust incidence. USDA Forest
Serv Res Pap SE-186. 20 p.

Steel, R. G. D. and J. H. Torrie. 1960. Principles and Procedures
of Statistics. McGraw-Hill Book Company, Inc, New York. 481 p.

Walkinshaw, C. H. 1978. Cell necrosis and fungus content in
fusiform rust-infected loblolly, longleaf, and slash pine
seedlings. Phytopathology 68:1705-1710.

Walkinshaw, C. H., and R. L. Anderson. 1983. Fusiform rust:
Illustration of different symptoms in the greenhouse and
field. USDA Forest Serv Forest Pest Manag Rep 83-1-21. 13 p.

Walkinshaw, C. H., T. R. Dell and S. D. Hubbard. 1980. Predicting
field performance of slash pine families from inoculated
greenhouse seedlings. USDA Forest Serv Res Pap S0-160. 6 p.

Wells, 0. 0., and P. C. Wakeley. 1966. Geographic variation in
survival, growth, and fusiform rust infection of planted
loololly pine. Forest Sci Monogr 11. 40 p.

Wright, J. W. 1976. Introduction to Forest Genetics. Academic
Press, New York. 463 p.




















APPENDIX
















Gall Type Gall Location
Witches'
Family Typical Thin BTFA/ Sunken Broom Fat Stem Limb Stem-Limb Total
-------------------------------- mean number of galls per tree --------------------------------


Least squares estimates of the mean number of fusiform rust galls per tree by gall type or location for 48 slash pine families (four- to six-years-old) planted in six progeny tests in southeast Georgia.


Table 26.


0.368 0.522 0.261 0.297
0.514 0.376 0.197
0.246 0.215
0.348 0.041 0.179 0.270
0.283 0.267 0.273
0.436
-0.003 0.281 0.397 0.306 0.281 0.206 0.528
0.473


2.183 6.884 5.373 5.192
9.047 4.084 2.101
2.920
3.994 5.617
-0.461 1.092
4.720 3.640 3.785
5.042 4.409 4.892 4.477 4.359 4.714 5.210
4.030 5.382
4.447


-0.036 0.016 0.177 0.097
0.021
0.079 0.383 0.003 0.109 0.119 0.076 0.115
0.022 0.172
0.036 0.061
0.054 0.041 0.099
0.154 0.139
0.114 0.053 0.196 0.035


-0.008 0.087 0.067 0.069
0.045 0.059
0.074 0.034 0.029 0.027
0.020
0.001 0.052 0.079
0.014 0.011 0.018
-0.019 0.063 0.063 0.033
0.044 0.044 0.180 0.136


0.295 0.900
0.401 0.411 1.085 0.675 0.300 0.261
0.402 0.575
0.041 0.191
0.469 0.503
0.346 0.534 0.505
0.423 0.503
0.421 0.377
0.461 0.364 0.534 0.550


0.089 0.018
-0.005
-0.003 0.026 0.062 0.019 0.037
0.041 0.034 0.025 0.065 0.080
0.049 0.008
0.041 0.037 0.032 0.063 0.063 0.090 0.066 0.008
0.064 0.022


0.960 3.623 3.759 3.157
6.624 2.359 0.703 1.908 2.508 3.083
-1.100
-0.357
2.954 1 .733
2.994 3.325
2.554 3.589
2.564 2.575 2.537
3.443 3.008
3.148 3.084


0.855 2.739 1.353 1.738 1.909
1.349 1.201
0.765 1.271
2.186 0.598 1.270
1.496 1.624 0.523
1.444 1.418 1.306 1.632 1.387 1.873 1.396 0.816 1.806 0.890


10050 10159
190047 190060
190141 190287
600047 600287 960065
960146 980048 980088
1410047 1410060 1410287 1460065
1460071 1460096 1570065
1570071 1570096
1570146 2700050
2700064 2700295


0.945 2.535
1.221
1.789 1.769 1.156 0.817 0.780
1.149 2.141 0.457 1.106
1.441 1.534 0.603 1.223
1.492 0.975 1.523
1.420 1.781 1.382 0.776
1.706 0.913


0.037
0.034 0.096 0.105
0.046 0.024 0.215 0.097 0.025
0.154 0.037 0.039
0.045 0.031
0.043 0.063
0.012
0.058
0.074 0.024 0.055 0.029
0.034 0.069 0.071














Gall Type Gall Location
Witches'
Family Typical Thin BTFA/ Sunken Broom Fat Stem Limb Stem-Limb Total
-------------------------------- mean number of galls per tree ----------------------------------


a/Basally truncated fusiform gall.


Table 26--continued.


0.996 1 .596 1.152 0.767 0.822
1.915 1.367 2.008 1.583 2.505 0.696 1 .355 1.763 3.026 2.233
2.483 1.044 1.706 1.198 1.928 1.032 1.807 1.982


0.435 0.286
0.417 0.377
0.437 0.330
0.324 0.418 0.288 0.733
0.354 0.770
0.476 0.292 0.319 0.756 0.103 0.307
0.249 0.506 0.177
0.571 0.620


4.524 5.001 4.675
4.207
4.372 7.287
5.183 5.587 5.905 6.303
4.104 5.64 0" 4.047 6.883
4.454 8.451 4.444 2.898
3.014 5.081 5.069
3.534 11.747


2700355
2840050 2840064 2840295 2840355 2880001 2880050 2880159 2880352
2950064 2950270 2950355 3500048 3500088 3500098 3500354 3520001 3520050 3520159
3540048 3540088 3540098 3540350


0.935 1.606 1.263 1.027
1.013 1.523 1.091 1.960 1.310 2.115 0.655 1.629 1.820 2.705 2.085
2.674 0.817 1.601
0.964 1.972
0.901 1.911 1.798


0.084 0.029 0.015 0.007
0.040 0.075
0.024 0.123
0.046 0.038 0.161
0.046 0.048 0.015 0.079 0.010 0.015
0.041 0.011 U.021 0.006
0.041 0.046


0.022
0.075 0.131
0.054 0.049 0.075 0.067 0.035 0.105 0.160 0.077
0.052 0.091 0.256 0.132 0.156 0.132 0.081
0.040 0.262 0.090
0.184 0.157


0.121
0.068 0.068 0.098
0.120 0.077
0.090 0.090
0.040 0.084 0.056 0.118 0.015 0.019 0.003 0.011 0.058 0.013
0.043 0.009 0.010 0.056
0.021


0.509
0.497 0.441 0.265
0.408 0.830
0.614 0.648 0.693 1.129
0.413 0.632 0.382
0.434 0.314 0.902
0.462 0.395 0.595 0.395
0.450 0.266 0.867


0.022
0.073 0.036 0.028 0.008
0.048 0.043 0.043 0.029 0.205 0.050 0.067 0.133 0.187 0.076
0.132 0.010 0.085 0.062 0.079 0.063 0.188 0.110


3.094 3.119 3.106
3.064 3.114 5.041 3.492 3.162
4.034 3.066
3.054 3.521 1.808 3.565 1.901 5.212 3.298 0.885 1.567
2.647 3.860 1.155
9.145















Gall Type Gall Location
Witches'
Fami ly Typical Thin BTFA/ Sunken Broom Fat Stein L i mb Stem-Limb
---------------------------------------------- % ----------------------------------------------


Least squares estimates of the mean proportion of trees infected with fusiform rust galls by gall type or location for 48 slash pine families (four- to six-years-old) planted in six progeny tests in southeast Georgia.


Table 27.


10050 10159
190047 190060
190141 190287
600047 600287 960065
960146 980048 980088
1410047 1410060 1410287 1460065
1460071 1460096 1570065
1570071 1570096
1570146 2700050
2700064 2700295 2700355


30.57 69.67 55.46
74.94 84.71 55.87
34.16 36.11
50.44 52.53 19.36 34.77
55.91
59.24 35.93
42.01 47.00 37.30 59.23 57.96 67.20
50.74 40.69 77.12
47.19 49.16


1.13
2.94 5.57 7.32
3.93 2.29 1.52 7.90
0.42 1.33 1.92 2.72
2.45 1.50 0.88
4.16 0.22
5.21 1.52 0.91
4.76 2.47 1.03
4.80 5.49 7.33


0.06 0.71
8.37 7.72
2.73
4.74 5.24 0.06
4.86 9.28 2.26
4.99 3.76 7.69 1.38 2.29 1.32 0.50 5.70 10.27
7.35 9.19 2.55
14.63 1.36 0.25


0.02
7.33
6.54 5.09
3.41 5.60 7.57
2.45 1.71
1.51
1.40 0.32 5.00 7.78
0.34 0.51
1.12
0.01
2.47 2.89 2.53 2- 20
3.92 16.63 6.99 10.72


30.84 49.37 31 .27 28.95 66.84 37.78 19.02
14.91 29.06
33.43 3.75 16.97
27.41
27.53 21.69 35.19 30.98 33.59 37.20 31.96
24.08 32.34 28.22 36.07
42.81 43.74


3.044 1.494 0.297 0.635 2.502 4.885
2.247 2.251 1.687 2.213 0.710
3.591
7.154 2.598 0.259
2.121 1 .776
0.233 3.862 3.862 8.186
4.920 0.259
4.145 1.378 1.353


21.13 29.63 32.96
41.34 52.78 32.23 19.93 17.07
18.89 28.96 1.30
5.45 29.95
44.52 21.81 21.32 26.75 23.29
26.43 24.08 23.27
25.78 26.32
40.47 33.19 38.58


33.52 65.59 53.67 69.84 90.08
52.74 29.97 36.77
49.26 48.52 20.01
33.54 57.21 55.12 28.35
44.81 36.84
40.81 59.02 52.01 68.61 48.27 37.07 71.84
49.26 54.54


9.92 35.68 23.90
30.14 55.08 25.31
14.79 23.03 20. 1
26.27 2.99
11.84 27.32
24.87 26.61 27.08 28.15 18.01 27.80 33.22 28.28
24.38 19.74 48.96 39.28
42.08











Table 27--continued.


Gall Type Gall Location
Witches'
Family Typical. Thin BTFA/_ Sunken Broom Fat Stein Limb Stem-Limb
---- ----------------------------------------- % ---------------------------------------------


A/Basally truncated fusiform gall.


36.13 37.58
34.63 35.81 30.10 28.37
49.09 30.01
29.48 31.64 38.78 20.75
26.94 20.10
54.76 22.25 16.35 21 .57 27.75 26.29 23.99 61.81


41.36 53.19 50.98
45.91 54.95 43.07 62.88
54.54 77.21
45.59 73.56 59.53 64.83 56.11 85.76 32.59 29.63
40.34 62.87 56.67
45.25 90.42


28.22 37.93 31.97 36.16 32.59
26.40 29.98
26.64 62.02 33.17 62.75
31.47 26.83 22.51
49.16 19.16 15.37 16.02 31.16
24.83 28.70
65.48


6.10
10.64 3.89 2.28 5.80 5.76 3.31 5.21 13.21
4.11 3.64 4.15 7.15
4.22 6.92 6.97 5.19
4.55 11.01
4.29 10.14 9.88


6.51
4.35 9.65
9.40 5.27
4.09 8.85
3.94 8.34 3.53 11.63 0.72
1.33 0.38 0.86 3.91 0.81 3.53 0.57 0.59 2.60
1.49


32.35 29.26 17.68 29.05
49.43 38.42 46.30 36.53 60.77
34.29 46.36 26.97
29.46 16.64 67.49 29.66 18.96 33.52 29.72 36.25 18.88 75.18


2.650
3.405 1 .780 0.259 3.554 2.181 1 .629 2.797
16.843 3.259
4.352 5.125 3.745 5.258 10.228 1.059
3.054 3.723 3.139 3.792 7.534
2.664


48.24 61.40 56.28 53.92 51.30
44.37 69.48 50.97 79.64
42.27
84.49 64.12 65.25
61.47 86.42 34.75 34.78 43.89 61.76 52.09 56.43 89.28


1.47
0.88 0.55
2.42 6.27 2.26 8.32 3.89 3.37 8.55 1.58 3.25 0.82 5.67
0.14 1.23
1.64 1.06 1.10 0.27
2.20
3.17


2840050 2840064 2840295 2840355 2880001 2880050 2880159 2880352
2950064 2950270 2950355 3500048 3500088 3500098 3500354 3520001 3520050 3520159
3540048 3540088 3540098 3540350












Variance components and their standard errors replication, general combining ability (GCA), combining ability (SCA) and error effects for number of galls per tree for various fusiform types or locations as computed by the FORTRAN DIALL, for each crossing set and test.


for
specific the mean rust gall program,


X-ing Variance Component
Set Test Gall Replication GCA SCA Error


34 Total
Stem Limb
Stem-Limb
Typical
Witches' Broom
B TFA/
Fat Thin
Sunken

35 Total
Stem Limb
Stem-Limb
Typical
Witches' Broom
BTF/
Fat
Thin
Sunken

36 Total
Stem Limb
Stem-Limb
Typical
Witches' Broom
B TFI/
Fat
Thin
Sunken

37 Total
Stem Limb
Stem-Limb
Typical
Witches' Broom
8TFI/
Fat
Thin
Sunken


0.53 + 0.81
-0.001 + 0.001
0.29 + 0.55
0.03 + 0.03 0.09 + 0.38
0.002-+ 0.006
0.009 ; 0.01
0.01 +-0.01
0.001 + 0.001
-0.0002-+ 0.00006

0.11 + 0.16
0.0001 + 0.001
0.1 + 0.13
0.00-3 + 0.004
0. 11 +70.13 0.007 *; 0.009 0.0037+ 0.002
-0.001 ; 0.0003 0.0 + 0.00002
0.0005 + 0.0003

-0.01 + 0.06
-0.00073 + 0.002
-0.001 +-0.03
0.0002-+ 0.013
-0.02 + 0.01
-0.01 + 0.004
0.001 + 0.0008*
-O.OoooT + 0.007 0.0003 + 0.0005
-0.0001 T 0.0002

0.004 + 0.07 0.008 + 0.015
-0.04 +0.01
-0.006 + 0.01
-0.045 + 0.02
-0.007 + 0.016 0.003 + 0.004
-0.0001+ 0.002
-0.0001 + 0.0004
-U.001 +0.001


-1.07 + 1.31
0.04 + 0.02**
-0.99 + 1.05 0.01 + 0.04
-0.94 + 0.85
0.2 + 0.12**
-0.01 + 0.01
-0.01 + 0.01
-0.0002 + 0.0004
-0.0001 ; 0.0004

2.11 + 1.25* 0.02 + 0.01*
1.26 + 0.75**
0.04 0.03*
1.02 + 0.63**
0.09 + 0.06
0.01 + 0.007*
-0.0007 + 0.0004
0.00001-+ 0.00003
0.001 + U.001

0.45 + 0.37 0.01 + 0.01 0.17 + 0.13 0.02 + 0.02 0.11 + 0.09 0.02 + 0.02
0.0009 + 0.0006*
0.004 + 0.003
-0.00002 + 0.0002
-0.0001 +0.0002

0.85 + 0.56*
0.03 + 0.03
0.18 + 0.12* 0.11 4 0 08* 0.35 + 0.23*
0.09 + 0.07 O.O00T + 0.005 0.U003 0 0.002 0.0008 + 0.0006
0.001 +-0.002


5.28 + 3.92*
-0.0057+ 0.003 4.46 + 3.22* 0.11 - 0.08* 3.59 + 2.69* 0.004 + 0.014 0.05 +-0.04* 0.05 + 0.04* O.O00T + 0.001 0.001 + 0.001

-0.17 + 0.16
0.003-+ 0.003
-0.16 + 0.11 0.01 + 0.01
-0.03 + 0.14 0.05 + 0.04*
-0.0027+ 0.004
-0.001 + 0.002
-0.0001 + 0.0001
0.003 + J.002*

0.29 + 0.27
-0.003 + 0.004
0.09 + 0.09 0.006 + 0.03 0.05 + 0.06
-0.003 + u.02
-0.0006-+ 0.0004
-0.02 + 7.007
-0.0003 + 0.0006
-0.0001 ; 0.0005

0.16 + 0.2
0.004 + 0.02 0.003 + 0.06 0.012 + 0.03 0.014 + 0.08
0.04 + 0.06 0.012 + 0.01
-0.0009 + 0.003
-0.0006 + 0.0006
-0.003 +-0.002


5.87 + 1.87 0.02 ; 0.007
4.58 + 1.45 0.12 + 0.04 4.16 + 1.32 0.06 + 0.02 0.07 � 0.02 0.06 + 0.02
0.006-+ 0.002 0.002 ; 0.0006

1.1 + 0.35
0.01-+ 0.003 0.83 + 0.26 0.03 + 0.01 0.72 + 0.23 0.06 + 0.02 0.02 + 0.007 0.009 + 0.003
0.0003-+ 0.0001
0.003 +-0.001

0.86 + 0.21
0.04 + 0.009 0.34 + 0.08 0.18 + 0.04 0.27 + 0.07 0.14 + 0.03
0.004-+ 0.001
0.1 + U.02
0.005 + 0.001 0.004 T 0.001

1.0 + 0.23 0.17-+ 0.04 0.51 + 0.12
0.21 +0.5
0.64 + 0.14 0.31 0 0.07 0.04 + 0.009 0.033 + 0.007 0.008 + 0.002
0.03 +-0.007


Table 28.












Table 28--continued.


X-ing Variance Component
Set Test Gall Replication GCA SCA Error


2 36 Total
Stem
Limb
Stem-Limb
Typical
Witches' B
BTFa/
Fat Thin
Sunken


37 Total
Stem Limo
Stem-Limb
Typical
Witches' Broom
BTFa/
Fat Thin
Sunken


-0.003 + 0.02 -0.002 T 0.001 -0.004 - 0.01
-0.0002 + 0.001 -0.012 +-0.006 Room 0.006 T 0.007
0.0 + 0.002
-O.U ; 0.0001
-0.0053 + 0.0002 -0.0004 T 0.0001


0.02 + 0.05
0.002-+ 0.005
-0.008 + 0.02
-0.01 + 0.005 0.008+ 0.03 0.002 ; 0.01
0.0001-+ 0.001
-0.00004+ 0.0008
-0.00005 T 0.0002
0.0005 + 0.001


40 Total 0.03 + 0.04
Stem 0.001 + 0.001
Limb 0.03 +0.04
Stem-Limb -0.0007 + O.Ou
Typical 0.07 + 0.06*
Witches' Broom -0.005 + 0.005 BTFa/ 0.0004-+ 0.001
Fat -0.0 + 000005
Thin -O.OOT + 0.004
Sunken -0.0000S + 0.0001


41 Total
Stem Limb
Stem-Limb
Typical
Witches' Broom
BTFa/
Fat Thin
Sunken


-0.09 + 0.05
-0.001 + 0.0007
-0.06 +-0.05 0.0 + 0.004
-0.07 + 0.04
0.3 + 0.3 O.OoT + 0.001 0.001 T70.001 0.002 ; 0.005
0.0002-+ 0.0002


0.008 + 0.03
-0.001 0.005
0.005 T 0.01
0.0005 + 0.0007
0.003 +0.01
-0.003 T 0.01
0.0006-+ 0.0007 0.0001 T 0.0001 0.0002 + 0.0002 0.0006 + 0.0003

0.07 + 0.06
0.007 + 0.007
0.04 + 0.03
-0.002 + 0.007
0.04 +-0.04
0.002 + 0.001 0.002 ; 0.002
0.0006--+ 0.0004 0.0002 + 0.0002
0.004 +-0.003

0.04 + 0.09
-0.0004 + 0.001
0.05 + 0.08
0.0003 + 0.002
-0.002 +-0.08 0.001 ; 0.002 0.0008-+ 0.001
0.0001 7 0.0002
-0.00005-+ 0.005 0.0001 +-0.0004

0.05 + 0.15
0.0004 + 0.003
-0.02 + U.06 0.03 + 0.03 0.03 + 0.08 0.005-+ 0.01
-0.001 ; 0.001
-0.0001-+ 0.0004
0.001 + 0.004
0.0001+ 0.0001


0.01 + 0.05 0.006 + 0.8
-0.006 " 0.02
-0.002 7 0.001
-0.006 T 0.019
0.017 7 0.02
0.0003-+ 0.0007
-0.0003 ; 0.0002
-0.0005 T 0.0003
-0.0007 T 0.0002

-0.045 + 0.05
-0.0003-+ 0.006
-0.04 + 0.02
-0.004 + 0.02
-0.02 +-0.03
-0.03 + 0.008
-0.0002 + 0.002
-0.002 +-0.0006
-0.0004 + 0.0002
-0.0003 T 0.002

0.1 + 0.1 O.oo + 0.002
0.075 + 0.1
0.Ou0 + 0.004 0.16 +-0.13*
-0.02 ; 0.007
0.0005 + 0.002 0.0001 7 0.0002
-0.0005 ; 0.01
0.0004 ; 0.0006

0.09 + 0.2
0.003 + 0.005
-0.04 + 0.1
0.02 - 0.02 0.002 + 0.1
-0.01 +0.02
0.0037+ 0.004 0.0002-+ 0.001
-0.001 + 0.007
-0.0001-+ 0.0001


0.18 + 0.06
0.016 + 0.005
0.12 + 0.03 0.01 + 0.003
0.1 + 0.03 0.03+ 0.01
0.002-+ 0.0007
0.002 + 0.0005 0.003 + 0.0008 0.003 + 0.0008

0.47 + 0. I
0.04 + 0.01 0.3 + 0.08 0.12 + 0.03 0.25 + 0.07 0.13 + 0.04 0.01 + 0.003 0.009 + 0.003 0.003 + 0.0007
0.01 + 0.003

0.3 + 0.1
0.007 + 0.002
0.32 +-0.09
0.02 + 0.006 0.22 " 0.06 0.09 + 0.03 0.01 + 0.003
0.001 + 0.0003
0.06 + 0.02
0.002 + 0.001

0.99 + 0.3
0.02 T 0.005
0.9 + 0.3
0.04-+ 0.01
0.8 +0.2
0.16 + 0.05
0.01 + 0.003 0.004 + 0.001
0.04 + 0.01
0.001 + 0.0002












Table 28--continued.


X-ing Variance Component
Set Test Gall Replication GCA SCA Error


4 34 Total 0.44 + 0.9
Stem -0.001-+ 0.001
Limb 0.29 + 0.67
Stem-Limb 0.0057+ 0.02
Typical 0.27 +-0.63
Witches' Broom 0.02 ; 0.005
BTFa/ -0.0027+ 0.002
Fat -0.0004 + 0.003
Thin -0.0003 0.0003
Sunken 0.0005 + 0.002

35 Total -0.06 + 0.06
Stem 0.002-+ 0.003
Limb -0.03 +70.08
Stem-Limb -0.006-+ 0.003
Typical 0.02 +70.07
Witches' Broom -0.02 T 0.01
BTFa/ 0.001-+ 0.002
Fat -0.0000o + 0.001
Thin 0.0001 +-0.0002
Sunken -0.0003 _ 0.0002

40 Total 0.07 + 0.1
Stem -0.0007 + 0.001
Limb 0.07 + 0.1
Stem-Limb -0.0007 + 0.001
Typical 0.01 + v.05
Witches' Broom 0.03 + 0.03
BTFa/ 0.001-+ 0.001
Fat 0.001 ; 0.0009*
Thin -0.00057+ 0.0006
Sunken -0.0001 7 0.0002

41 Total -0.02 + 0.02
Stem -0.000E + 0.0007
Limb -0.002 +70.02
Stem-Limb -0.004 + 0.005
Typical 0.02 +-0.03
Witches' Broom -0.005-+ 0.003 BTFa/ 0.001 ; 0.001
Fat -0.0002 + 0.0006
Thin 0.009 +-0.008*
Sunken -0.0003 + 0.0004


2.28 + 1.98
0.022 + 0.016
1.2 + T.2 0.8 + 0.6 0.89 + 1.3
0.08 ; 0.05 O.O00T + 0.005
0.001 --0.002
0.00047+ 0.004 0.005 +-0.004

0.07 + 0.3
0.01 + 0.01 0.07 + 0.15
-0.01 + 0.01
0.006-+ O.I 0.01 + 0.03
0.000-3 + 0.0009
-0.0005 ; 0.0004
0.00003-+ 0.0003 0.0005 +-0.0004

0.1 + 0.1
0.ooT + 0.001
0.1 + 0.1
0.0OT + 0.001
0.06 +-0.05 0.01 + 0.04
-0.001 + 0.002
-0.001 + 0.001
0.0002+ 0.0006
0.0004 ; 0.001

0.5 + 0.3
-0.01 + 0.02 0.4 + 0.3*
-o.ooT+ 0.009
0.3 + 0.2
0.04 + 0.02*
-0.0067+ 0.02
0.001 + 0.0007
0.002 ;0.003
0.00005 + 0.002


0.19 + 1.6
0.006-+ 0.007
0.08 +-1.2
-0.02 T 0.02
0.08 ; 1.6
-0.01 ; 0.02
0.004-+ 0.008
-0.005 � 0.006
0.008 + 0.006**
-0.002 + 0.004

0.57 + 0.56 0.01 0.01 0.02 " 0.2 0.04 � 0.04
0.2 +-0.2
0.02+ 0.06
-0.001-+ 0.002
-0.001 ; 0.001
0.0004-+ 0.0007
-0.0008 + 0.0006

-0.1 + 0.1
-0.0004 + 0.001
-0.1 + 071
-0.00-5 + 0.001
-0.08 +-0.06 0.05 + 0.05
0.003-+ 0.004
0.004 + 0.003**
-0.001 + 0.001 0.001 � 0.001

0.02 + 0.08 0.03 T 0.02*
-0.04 + 0.02
-0.002-+ 0.01
-0.04 +-0.02
-0.01 � 0.005 0.03 + 0.02
-0.001+ 0.0008
-0.002 4 0.004 0.001 + 0.002


6.2 + 2.19
0.01T + 0.005
4.9 + T.7
0.16-+ 0.06
4.6 + 1.6
0.12-+ 0.04 0.02 + 0.009
0.03 + 0.01
0.005-+ 0.001 0.02 +-0.007

1.2 + 0.4
0.02 + 0.005
1.2 +-0.4
0.08-+ 0.03
0.6 + 0.2 0.2 + 0.07
0.01 + 0.004 0.009 + 0.003 0.002 + 0.001 0.004 + 0.001

1.1 + 0.3
0.01-+ 0.002
1.0 + 0.3
0.02 - 0.005
0.5 + 0.2
0. 1 + 0.05
0.01 + 0.003 0.003 + 0.001 0.009 " 0.003 0.003 T 0.001

0.2 + 0.09
0.007 + 0.003
0.13 +70.05 0.05 � 0.02 0.15 + 0.06 0.04 + 0.02
0.005-+ 0.001 0.005 ; 0.002 0.02 +7-0.007 0.004-+ 0.002












Table 28--continued.

X-ing Variance Component
Set Test Gall Replication GCA SCA Error


5 34 Total
Stem Limb
Stem-Limb
Typical Witches'
BTFa/
Fat
Thin
Sunken


-0.3 + 0.3
0.00-6 + 0.01
-0.34 +70.3
0.005-+ 0.01
-0.2 + 0.2 Broom 0.01 + 0.02
0.01 + 0.04
-0.003 + 0.002
-0.02 +70.03
-0.000 + 0.0005


35 Total 0.07 + 0.1
Stem -0.01 0.003
Limb 0.08 + 0.1
Stem-Limb 0.0003 + 0.006
Typical 0.13 + 0.11*
Witches' Broom 0.02 ; 0.04 BTFa/ -0.001-+ 0.01
Fat 0.0001 + 0.0008
Thin 0.0006 + 0.0007
Sunken 0.0006 + 0.001

40 Total 0.02 + 0.04
Stem -0.002 + 0.001
Limb 0.02 +-0.04
Stem-Limb -0.001 + 0.001
Typical 0.006 ; 0.02
Witches' Broom -0.0005 + 0.003
BTFa/ -0.0005 0.0004
Fat 0.0004 + 0.0004
Thin 0.0007 7 0.001
Sunken 0.0001 + 0.0004

41 Total -0.1 + 0.9
Stem -0.02 + 0.006
Limb -0.05 ; 0.07
Stem-Limb 0.o00ooT + 0.003
Typical -0.06 + 0.06
Witches' Broom 0.001 + 0.01
BTFa/ -0.001 ; 0.0007
Fat -0.0003 + 0.0004
Thin 0.002 + 0.002
Sunken 0.0006 + 0.0005


-0.1 + 0.5
-0.04-+ 0.06
0.4 + 0.3
0.05-+ 0.03
-0.02 7 0.3 0.1 + 0.08 0.01 + 0.05
0.001 + 0.001 0.001 + 0.05 0.002 + 0.001

0.4 + 0.4
0.02 + 0.02 0.17 + 0.2
-0.01 T 0.04
0.3 + 0.3
0.01 + 0.06
-0.002-+ 0.02
0.0002 + 0.0006 0.0003 + 0.0008
0.001 + 0.001

0.23 + 0.2
0.001 + 0.002
0.1 + 0.1
0.00T + 0.004
0.05 + 0.08 0.02 + 0.02
0.0003 + 0.0004
-0.0003 + 0.0003
-0.0003 7 0.0006 0.0003 + 0.0004

0.3 + 0.9
0.008 + 0.009
0.1 + 0.7
0.01 + 0.005
0.3 +-0.4
-0.03 + 0.08
0.003-+ 0.002
-0.0006 + 0.0009
0.002 + 0.001
0.0001 + 0.0002


-0.4 + 0.8
0.08 + 0.08**
-1.0 + 0.5
-0.02 + 0.01
-0.2 + 0.5
-0.01 + 0.01
-0.02 + 0.07
-0.006 + 0.003
-0.02 + 0.08
-0.002 + 0.001

0.5 + 0.4*
-0.0001 + 0.02
0.2 + 0.2
0.1 + 0.08**
0.2 + 0.2 0.06 + 0.1
0.01 + 0.03
-0.001 + 0.001
0.0008 + 0.001
-0.002 + 0.001

0.17 + 0.16
0.002 + 0.004
0.1 + 0.1 O.0oT + 0.004
0.12 + 0.1* 0.03 + 0.02*
-0.001 + 0.001
0.0009 + 0.0008
-0.001 + 0.002
-0.0002 + 0.0006

1.04 + 0.98*
-0.002-+ 0.009
0.9 + 0.8*
-0.006 + 0.003
0.3 + 0.4
0.12 + 0.1*
-0.003 + 0.001 0.0006 + 0.001
-0.001 + 0.001 0.0 + 0.0003


3.7 + 1.5
0.08 + 0.03
3.7 + 1.5
0.07 + 0.03
2.1 + 0.9
0.08 + 0.03
0.3 + 0.1
0.02 + 0.009
0.3 + 0.1
0.005 + 0.002

0.5 + 0.2 0.08 + 0.3
0.4 + 0.1
0.06 + 0.02 0.2 + 0.09
0.3 + 0.1
0.1 + 0.04
0.007O + 0.002 0.002 + 0.0009
0.007 + 0.003

0.4 + 0.1
0.02 + 0.04
0.3 + 0.1
0.02 + 0.006
0.2 + 0.07 0.05 + 0.01
0.01 ; 0.003
0.002 + 0.0005
0.01 + 0.004 0.005 + 0.001

0.4 + 0.2 0.03 -+ 0.1 0.3 + 0.1
0.02 + 0.007
0.4 + -0.2
0.05 + 0.02
0.007-+ 0.003 0.003 + 0.001 0.003 + 0.001
0.001 + 0.0004












Table 28--continued.


X-ing Variance Component
Set Test Gall Replication GCA SCA Error

9 40 Total 0.09 + 0.07* 0.04 + 0.06 0.07 + 0.09 0.3 + 0.07
Stem -0.001 + 0.0004 0.0007 + 0.0009 -0.002 + 0.002 0.02-+ 0.04
Limb 0.05 + 0.05 0.03 + 0.04 0.02 + 0.06 0.3 + 0.07
Stem-Limb 0.005 + 0.004* 0.009-+ 0.005* -0.004 + 0.002 0.02+ 0.006
Typical 0.04 +-0.03** 0.04 +-0.03 0.01 + 0.03 0.1 + 0.03
Witches' Broom 0.003 + 0.005 0.006-+ 0.008 0.004+ 0.01 0.05-+ 0.01
BTFa/ 0.0006 + 0.001 0.002 + 0.002 -0.001 + 0.002 0.016 + 0.004
Fat 0.0001 + 0.0003 0.00002 + 0.0003 0.0 + U.001 0.005 ; 0.001
Thin -0.001 +70.0004 -0.0007 +70.0008 -0.0003 + 0.002 0.01 + 0.004
Sunken 0.0004+ 0.0007 0.0004 70.0009 0.0008 -0.001 0.007-+ 0.002
41 Total 0.07 + 0.08 0.7 + 0.4* 0.12 + 0.16 0.6 + 0.1
Stem 0.0002 + 0.003 0.0004 + 0.001 -0.007 + 0.003 0.04-+ 0.01
Limb 0.01 + J.04 0.42 + 0.27* 0.11 + 0.13 0.4 +0.1
Stem-Limb 0.005+ 0.008 0.04 + 0.02* -0.002 + 0.01 0.07-+ 0.02
Typical -0.01 + 0.02 0.23 + 0.16 0.07 + 0.1 0.4 + 0.1
Witches' Broom 0.04 + 0.05 0.08 + 0.07 -0.008 + 0.07 0.4 + 0.1
BTFa/ 0.000T + 0.0009 0.001-+ 0.0007 -0.003 + 0.001 0.01-+ 0.003
Fat 0.002 +70.002 0.005 TO.004 0.0007 + 0.003 0.01 + 0.003
Thin 0.003 + 0.002* 0.001 + 0.0006* -0.002 +-0.0006 0.009-+ 0.002
Sunken 0.002 T 0.002 -0.0007+ 0.002 0.002 T 0.004 0.014 + 0.004

a/Basally truncated fusiform gall.

*F test for the effect significant at the 5% level.

**F test for the effect significant at the 1% level.



















Table 29. Specific combining ability for fusiform rust incidence and fat galls based on the proportion of trees diseased or having fat galls. Values reflect performance of 48
families (four- to six-years-old) in six progeny tests in southeast Georgia.




Specific Combining Ability
----------------------------------------------------------- Males ----------------------------------------------------------Female UuO] 0047 0U48 0050 0060 0064 0065 0071 0088 0096 0098 0141 0146 0159 0270 0287 0295 0350 0352 0354 0355


Percentage of Trees Infected With Rust


UU01 UO 19
uobu UO90 U098
U141 U 14b

U27U
u 2d4
u268


U35U 03b
03b4


-3.13
-2.8


b.88


5.8/


-1.46


-4.51


1.38
6.97
-4.92


-1.7



2.15


1.15
-4.04


-7.32


-2.18


4.20


-6.47 2.01 1.06 -2.06


-3.19


-7.65
3.24


1.83

-6.14


6.57



-2.19


-1.24

2.55


5.31
2.25


-7.63


0.78
-1.13


-3.93


-6.57


4.02 0.58


b.97


-1 .94


-5.47
-1.89

7.32


4.82


U.58




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EZMUI0Y7V_NLHITU INGEST_TIME 2017-07-20T21:06:15Z PACKAGE UF00102787_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

GENETIC VARIATION IN SYMPTOMOLOGY OF SLASH PINE IN RESPONSE TO FUSIFORM RUST BY PATRICIA ADLENE LAYTON A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1985

PAGE 2

This research is dedicated to my parents, James B. and Rebecca Layton, who always ''think I can." Their love and faith in me have always provided a stable base in my life. Their help was even more concrete in this study in that they gallantly volunteered to help in data collection at the Rust Screening Center.

PAGE 3

ACKNOWLEDGEMENTS I am deeply grateful to W. Paul Shelley, Jr., Erin C. Shelley and other members of the Shelley family who established the W. P. Shelley, Sr., Graduate Forestry Fellowship which I received during my last two years of graduate study. I would like to thank the members of my supervisory committee for their guidance and inspiration: Ors. Ray E. Goddard, Ramon C. Littell, Donald L. Rockwood, Robert A. Schmidt, Anthony E. Squillace and Charles J. Wilcox. Additionally I would like to thank Ors. Tom Miller, George Blakeslee, Mitch Flinchum, Ron Dinus, Nancy Pywell, Arnett Mace, Nick Comerford, Everett Hall and Bro Kinloch for their advice and encouragement. This project would not have been possible without the help and support of the University of Florida Cooperative Forest Genetics Research Program. I would like to thank the staff at the University, which includes Greg Powell, Harm Kok, B. J. Rabe, Charles Akins, and James Hayes, as well as the industrial members. In addition, SFRC students Penny Parnell and Robert Richens were a great help in collecting the data. Two of the industries, Brunswick i i i

PAGE 4

Pulp Land Company and Container Corporation of America, provided materials for this study. I am especially grateful to James Hodges and the research staff at Brunswick for their assistance in collecting the field data. Portions of the project were conducted at the US Forest Service's Rust Screening Center in Asheville, N orth Carolina. I appreciate the advice and help of the staff, especially Robert Anderson and Carol Young. I wish to thank Martin Marietta Energy Systems for helping in the completion of this project. I would like to thank my fellow graduate students for their encoura~ement and support throughout this project. I would especially like to thank Tim Meyer for his friendship and help. Lastly, I want to thank Rick Cantrell, my husband, for his love, patience and understanding. iv

PAGE 5

ACKNOWLEDGEMENTS LIST OF TABLES LIST OF FIGURES ABSTRACT . INTRODUCTION LITERATURE REVIEW TABLE OF CONTENTS i i i vi X xi 4 MATERIALS ANO METHODS 14 Field Tests--Experiment . . . . . . 14 Artificial Inoculations . . . . . 19 Experiment 2--RSC Procedure . . . 19 Experiment 3--High Inoculum Concentration . . 21 Experiment 4--Inoculation of 1-0 Seedlings . . 23 An a 1 ys is . . . 2 3 RESULTS AND DISCUSSION 31 Field Tests--Experiment . . . . . . 31 Artificial Inoculation . . . . 54 E x periment 2--RSC Procedure . . . 55 Experiment 3--High Inoculum Concentration . . 64 Experiment 4--Inoculation of 1-0 Seedlings 66 Research Perspectives . 68 CONCLUSIONS LITERATURE CITED. APPENDIX .. BIOGRAPHICAL SKETCH V 70 72 79 90

PAGE 6

LIST OF TABLES Table 1. Location, year of establishment, site index, fifth year height and rust incidence of Brunswick Pulp Land Company's diallel progeny tests . . . 15 2. Sets of slash pine crosses used in field and artificial inoculation experiments . . . . . 16 3. Percentage of trees infected with fusiform rust for 48 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia 17 4. 5. 6. University of Florida rust evaluations, standard errors and number of tests used for evaluation of slash pine parents Analysis of variance and expected mean squares for a modified half-diallel as computed by DIALL Analysis of variance and expected mean squares for a four-stage nested design using individual tree data . . ....... 7. Analysis of variance and expected mean squares for Experiment 2 using mean number of symptoms per tray . . . . . . . . . 8. Analysis of variance and expected mean squares for 1 8 24 26 2 9 Experiment 4 using individual tree data 30 9. Percentage of rust-associated mortality for 48 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia. . . 32 10. Percentage of trees dead and classified as 'rust bush many cankers' for 48 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia . 33 vi

PAGE 7

Table 11. Results of analyses of variance of the proportion of trees with different fusiform rust symptoms (arcsin square root transformation) and the mean number of these symptoms per tree for family and test differences . . . . 39 12. Significant replication (Reps), general combining ability (GCA) and specific combining ability (SCA) F values (5% level of significance) obtained from analysis of variance using DIALL for number of fusiform rust galls of several types/locations on slash pine (18 analyses for each type) 13. Significant replication (Reps), general combining ability (GCA) and specific combining ability (SCA) F values (5% level of significance) obtained from analysis of variance using DIALL for the proportion (arcsin square root transformation) of slash pine trees with various fusiform rust symptoms (18 analyses for 41 each symptom) . . 42 14. Significant site (S), replication (Reps), general combining ability (GCA), specific combining ability (SCA),and interaction (GCA x S) F values (5% level of significance) obtained from analysis of variance using DIALL for number of fusiform rust galls of various types/locations on slash pine, planted at several sites (nine analyses for each type) 45 15. Significant site (S), replication (Reps), general combining ability (GCA), specific combining ability (SCA),and interaction (GCA x S) F values (5% level of significance) obtained from analysis of variance using DIALL for the proportion (arcsin square root transformation) of slash pine trees with various rust symptoms planted at several sites (nine analyses for each symptom) . 46 16. Inaividual (ht) and family heritibilities (h~) and their standard errors (s) for rust incidence, rust bush many cankers (RBMC), rust-associated mortality (RAM) and the mean number per tree of 10 fusiform rust gall types for slash pine, fourto six-years-old 50 Vi i

PAGE 8

Table 17. Genetic correlations between various fusiform 18. rust gall types and locations (mean number per tree), rust incidence and mortality (RAM and RBMC) for slash pine fourto six-years-old General combining ability estimates of the proportion of trees with fusiform rust galls for 25 slash pine parents based on progeny performance in six locations in southeast Georgia, aged four to six years 19. Percentage of seedlings in 20 slash pine families with various reaction types six months after artificial inoculation with fusiform rust and 52 54 the resistance index of the families . . . 56 20. Mean squares for the analyses of variance of the proportion of slash pine seedlings per family expressing various disease symptoms resulting from artificial inoculation with fusiform rust at the Resistance Screening Center, Asheville, North Carolina 57 21. Individual heritabilities (h2) and their standard errors (s) for artificial inoculation (Al) symptoms by both proportion of trees with the symptoms and mean number of symptoms per tree . 59 22. Pearson product-moment correlations between the mean number of artificial inoculation (Al) symptoms per tree of slash pine seedlings inoculated with fusiform rust, based on the progeny of 20 families 60 23. Pearson product-moment correlations between field galls and artificial inoculation symptoms (proportions) and the resistance index of 17 slash pine families and fusiform rust Vii i 62

PAGE 9

Table 24. Results of artificial inoculation with high concentrations of inoculum (2.0 x 106 spores/ml) on 17 slash pine families, Resistance Screening Center, Asheville, North Carolina . . . . . 65 25. Comparison of family means for several fusiform rust disease symptoms on artificially inoculated 1-0 slash pine seedlings . . . . . . 67 26. Least squares estimates of the mean number of fusiform rust galls per tree by gall type or location for 48 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia 79 27. Least squares estimates of the mean proportion of trees infected with fusiform rust galls by gall type or location for 48 slash pine families (four to six-years-old) planted in six progeny tests in southeast Georgia 28. Variance components and their standard errors for replication, general combining ability (GCA), specific combining ability (SCA) and error effects for the mean number of galls per tree for various fusiform rust gall types or locations as computed by the FORTRAN program, DIALL for each crossing 81 set and test. . . . 83 29. Specific combining ability for fusiform rust incidence and fat galls based on the proportion of trees diseased or having fat galls. Values reflect performance of 48 families (four-to six-yearsold) in six progeny tests in southeast Georgia. . 88 ix

PAGE 10

LIST OF FIGURES Figure 1. 2. Gall types commonly found on field-grown slash pine trees infected with fusiform rust ... Galls found on slash pine seedlings six months after artificial inoculation with fusiform rust 3. Frequency distribution for slash pine trees 4. by the total number of fusiform rust galls, in six progeny tests in southeast Georgia, at four-, five-, and six-years-old ............ Frequency distribution for two slash pine families, 0098 x 0088 (resistant) and 0350 x 0354 (susceptiole), for the total number of fusiform rust galls at age five years, planted in a progeny test in Camden County, Georgia X 20 22 35 38

PAGE 11

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy GENETIC VARIATION IN SYMPTOMOLOGY OF SLASH PINE IN RESPONSE TO FUSIFORM RUST By Patricia Adlene Layton December, 1985 Chairman: R. E. Goddard Major Department: Forest Resources and Conservation Genetic variation in the reaction of slash pine to fusiform rust was investigated in field and greenhouse situations. In the field, fourto six-year-old plantings of five crossing sets, each involving five or six parents, were studied. Each crossing set was planted at two or four sites. Rust incidence in the tests ranged from 46% to 89%. Survival and stem form of each tree was determined. Galls on infected trees were counted and classified as to location and type. Variances in traits measured were analyzed by crossing set, crossing sets among sites and all crossing sets at all sites. Significant differences among families were found for some gall types, gall locations, total number of galls per tree, incidence, xi

PAGE 12

1 11ortality and bush stem form. Additive genetic effects 1\lere found in many traits, but the number of fat galls had non-additive effects. Total number of galls was the best field measure of susceptibility, having high genetic correlations with mortality and bush stem form. Seedlings from 17 of the field-tested families were artificially inoculated. After six months, there were differences among families for the proportion of seedlings and number per tree with seven resistance and susceptibility indicators. Although heritabilities were often low, those for proportion of trees with galls, purple spots and adventitious shoots were 0.23, 0.28 and 0.33, respectively. Following artificial inoculation with high concentrations of rust inoculum, families varied in the number of red spots at 10 and 48 days. Correlations with family field performance were low. Galls developing on artificially inoculated one-year-old seedlings resembled field gall types more than those on younger seedlings. Families differed in percentage of trees with rust and the mean number of fat galls. Three of four phenotypically resistant males increased family resistance to a level equal to or better than a rust-resistant orchard check. These studies indicated genetic variation for specific disease reactions in the field and confirmed such variation in artificial inoculation studies. Proportion of trees infected, total galls per tree, mortality, and bushstem form had the highest heritabilieies and should be considered in screening for rust resistance. Xii

PAGE 13

INTRODUCTION The major pest of southern commercial forests is fusiform rust (Cronartium guercuum (Berk.) Miyabe ex Shirai f. sp. fusiforme) (Cqf). In a 1979 survey of five states (South Carolina, North Carolina, Virginia, Georgia and Florida), more than one-third of the acreage of slash (Pinus elliottii Engelm. var. elliottii) and loblolly (f.:__ taeda L.) pine had more than 10 of the trees infected (Robert Anderson, personal communication 1985), representing an annual loss of $32,894,000. Losses throughout the South are expected to be much greater. Disease incidence has been increasing over time (Schmidt et al. 1974; Griggs and Schmidt 1977). Management practices, such as mismatching species and site, fertilization and intensive site preparation (Powers et al. 1981), are implicated in this increase. Another reason for the increase may be the abundance of rust-susceptible Quercus spp. (alternate host of the fungus) in southern forests (Squillace et al. 1978). Modifying these management practices as well as utilizing more effective silvicultural techniques for suppressing Quercus spp. may help

PAGE 14

2 reduce the epidemic. Planting rust-resistant pines is an effective way of significantly decreasing disease losses (Schmidt et al. 1981; Schmidt et al. 1985). The disease is caused by a heteroecious, macrocyclic rust fungus. Pycnia and aecia occur on pine while uredia and telia occur on oak. Basidiospores produced from telia on oak infect pine during favorable environmental conditions. Most southern red oaks, especially water oak (Q. nigra L), are susceptible to the fungus (Dwinell 1974). Although much is known about the life cycle (Dwinell 1977), knowledge is lacking about the genetics of the fungus. There is a wide range of variation in the ability of Cqf to infect pine (Snow and Kais 1970; Snow et al. 1975; Powers et al. 1977). Spores from galls on resistant families evidenced increased virulence in comparison to other spore collections when used to reinoculate the resistant families (Griggs and Walkinshaw 1982; Powers et al. 1978; Snow and Griggs 1980). Various review articles have summarized f~nily resistance in slash and loblolly pine (Goddard and Wells 1977; Powers et al. 1981; Schmidt et al. 1981). Certain loblolly pine geographic seed sources also are resistant to infection by Cqf (Wells and Wakely 1966) In the past, family resistance to Cqf has been evaluated by percentage incidence in both field and greenhouse studies. This method is effective in screening families for resistance, but does little to identify resistance mechanisms or their inheritance. More recent methods have evaluated disease resistance by examining

PAGE 15

3 s~nptoms resulting from artificial inoculation ( C arson 1984, Framptom et al. 1983; Walkinshaw et al. 1980). The objectives of this study were as follows: l. To examine and describe qualitative characteristics (symptoms) of the response of slash pine to Cqf, both in the field under natural conditions and in the greenhouse using artificial inoculation techniques; 2. To determine heritability and genetic correlations of the traits; and 3. To correlate field and greenhouse observations and suggest symptoms that may lead to a more precise prediction of field response by greenhouse testing.

PAGE 16

LITERATURE RcVIEW As fusiform rust oecame epidemic, tree improvement programs incorporated disease resistance into their selection and breeding programs. Although not a total solution, resistance is the primary pest management strategy in intensively managed forests (Schmidt et al. 1981; Schmidt et al. 1985). Original mass selections were essentially random for rust resistance, out approximately 7.5% had some degree of resistance when evaluated using artificial inoculation techniques (Goddard and Arnold 1966). Goddard and Arnola (1966) reported 0.199 as an estimate of heritability from artificial inoculation. Estimates of heritability for percentage rust incidence in field tests of slash pine varied greatly. Sohn and Goddard (1979) stated that they varied with the incidence levels in the tests examined. Indiviaual heritabilities varied from 0.054 to 0.238 (Rockwood and Goddard 1973) in several progeny tests. Sohn (1977) estimated individual heritabilities for incidence ranging from 0.067 to 0.285 in other progeny tests. The relationship oetween incidence and heritability in these studies supported Sohn and Goddard's (1979) conclusion. 4

PAGE 17

5 Various estimates of rust resistance are reported for loblolly pine. Kinloch and Stonecypher (1969) reported family heritabilities for incidence of 0.65 to 0.85 over four different environments. Blair (1970) estimated heritabilities for number of galls per tree, severity index and percentage incidence. In one field test, these estimates were 0.29, 0. 2 2 and 0.20, and in another they were 0.09, 0.04 and 0.12, respectively. In the same tests as Blair, but at an older age, Barker (1973) estimated the heritabilities of the severity index to be 0.11 and 0.35. These heritabilities supported the concept that substantial gains in rust resistance were possible. Screening selections for rust resistance became a priority. Inoculation techniques were developed to screen families for incorporation in breeding programs (Jewell 1960; Goddard and Schmidt 1971). Schmidt (1972) describes early techniques and their deficiencies were identified by Laird and Phelps (1975) as 1. temperature and/or humidity not adequately controlled; 2. inoculation time not controlled; 3. inoculum concentration not controlled; 4. techniques not applicable to large numbers of seedlings; 5. results not repeatable; and 6. results not quickly obtainable. Laird and Phelps (1975) tested three methods of artificial inoculation and determined that the concentrated basidiospore spray system (CBS) (Matthews and Rowan 1972) was best for minimizing these problems.

PAGE 18

6 In 1973, the United States Forest Service established a screening center (RSC) for rust resistance in loblolly and slash pines from throughout the South at Asheville, North Carolina, utilizing the CBS system. There, in a controlled environment, seedlings could be inoculated with various pathogen sources at optimum inoculum levels (Hubbard and Anderson 1980). The percentage of trees galled at six to nine months of age was used to rank family resistance. Resistant and susceptible checks were used in every test. Screening by artificial inoculation (all methods) was plagued by lack of adequate correlation with family field performance. The University of Florida Cooperative Forest Genetics Research Program began field testing families rather than relying solely on artificial inoculation procedures. Short-term field progeny tests were established to mass screen selections in their tree improvement program (Goddard et al. 1972). Later, other tests were established to retest the most resistant families in areas of high rust incidence (Goddard et al. 1984). Many families with high degrees of resistance were found with these methods. However, open-pollinated families varied tremendously in their reactions. Powers and Zobel (1978) found significant differences in infection occurred on open-pollinated seedlings of the same clones produced in different orchards. These results illustrate the strong influence of the pollen parent in resistance. To understand the genetics of resistance better, testing of full-sib families (artificial inoculations and field tests) must be performed.

PAGE 19

7 Artificial inoculation methods were useful in exploring host variation in response to infection. Families were found that were resistant to Cqf from some regions, but not to spores originating in other areas (Snow and Kais 1970). Gall form (ratio of length to width) varied by host family and inoculum source (Snow et al. 1982). Resistance was overcome in some families by particular isolates of the pathogen (Griggs and Walkinshaw 1982; Powers and Dwinell 1978). Selection for increased virulence in the pathogen has been found in slash pine families (Dinus et al. 1975; Griggs and Walkinshaw 1982; Snow and Griggs 1980). Snow et al. (1976) stated that the greater virulence of inocula from resistant trees was evidence of pathogenic specialization. Virulence varies greatly in the fungus. Powers (1980) found as much variation in 10 single aeciospore inoculations from one gall as he did in standor county-wide collections of aecia. Powers and 0winell (1978) reported that pathogen virulence had not increased with time when comparing isolates from galls originating in 1945 and 1970; however, the 1970 isolates were more variable in their response. The fungus infects pine in a haploid (lN) state, while the aeciospore, collected from pine for testing, is a dikaryon (N+N). The biological mechanism for this change is not known. This was studied in the related species f. ribicola by Hirt (1964), but he found no conclusive evidence of sexual recombination. Until the sexual processes of Cqf are understood, the exact nature of pathogenic variation will be unknown.

PAGE 20

8 Increasing the accuracy of assessment of host variability is a continuing goal. Walkinshaw et al. (1980) developed a scoring procedure for open-pollinated slash pine families tested by the RSC to better predict resistance. They examined twelve traits. Correlations between RSC performance and field evaluations were increased by utilizing the proportion of trees in a family with 1) smooth galls, 2) symptoms without swelling and 3) fat galls. These variables accounted for 62% of the variation in field performance evaluations. Predicting field performance with similar methods for loblolly pine has not been reported. Walkinshaw and the staff at the RSC continue to test methods for this species (C.H. Walkinshaw, personal communication 1984). Walkinshaw and Anderson (1983) described seven fusiform rust symptoms found on seedlings in the greenhouse and eight on trees in the field. In the greenhouse, the most resistant symptom found was a purplish discoloration on the stem and/or a needle base, called SYMNO. Fungal tissue was not found in the area of discoloration. A short gall of less than 25 millimeters and galls with a rough textured, slightly discolored surface were also indicative of resistant reactions. More susceptible reactions included fat galls (gall diameter is twice that of the stem below the gall), typical galls (gall is fusiform shape and not abnormally large or small), smooth galls (gall surface is smooth and generally green on six-months-old slash pine seedlings) and baseball bat galls (base of the gall is constricted, but there is a typical fusiform shape at the top). Baseball bat, fat and typical galls were also found on

PAGE 21

9 trees in the field. I n addition, field symptoms included thin galls (only slightly swollen), twisted galls (stem twi s ts i n the area of the gall), sunken galls (a depression in the gall surface) and witches broom galls (a loss of apical dominance ) The authors proposed using these symptoms to increase the prediction c apabilities. Prediction methods rely on the premise that resistant or susceptiole reactions in seedling progeny of a particular open-pollinated family can predict response of other family members grown in the field. Day (1974) advised that possible limitations of this premise include the following: 1. The observed resistance or susceptibility may be a feature of the juvenile tissues (cotyledons and unsuberized stems). 2. Observed reactions may reflect test envir~nment as opposed to host genotype. Other problems with relating field and artificial inoculation tests are pathogen related. Composite inoculum from the area where host families are to be outplanted eliminates (or complicates) many of the problems of pathogenic variability. Inoculum concentrations are extremely variable in the field, but are fairly constant in artificial inoculations. Lundquist et al. (1982), utilizing extremely high inoculum concentrations, found breakdowns in resistance for some slash pine families--a product of both inoculum concen tr ation and host genotype. In their tests, families were ranked by observations of pigmented spots on seedlings and t he

PAGE 22

10 rankings were similar to those achieved by the CBS method (percentage galled). The authors suggested that two types of resistance may exist: one that persists under high inoculum loads and one that does not. They did not consider that high inoculum concentrations significantly increase the probability of a virulent spore's landing on a seedling. Further testing of this method is needed. Lundquist and Luttrell (1982) reported that pigmentation patterns found in response to rust inoculation were similar to those observed when seedlings were wounded or exposed to ultraviolet light. They suggested this may have been a specious effect or a reaction by-product. Timing of pigmentation varied with host-parasite interaction and did not vary with other stress-related reactions, suggesting it was related to a resistance mechanism. Miller et al. (1976), utilizing different methods of inoculation from Lundquist and Luttrell (1982), suggested there might be four general types of host-parasite interactions at work. The first interaction was the inability of germinated basidiospores to penetrate and/or infect host epidermal cells. This was considered total incompatibility of host and pathogen. The second interaction, subliminal infection, resulted in infection and a sparse spreading of mycelium with a few scattered haustoria. Host cells did not show the distortion typically found in infected trees. Miller et al. (1976) suggested that the pathogen in this interaction was avirulent. The authors described three hypersensitive reactions as a third interaction. Two of these

PAGE 23

11 hypersensitive reactions resulted from stem infections and produced darkly stained reaction zones. Purple stem lesions were visible for both reactions. They differed primarily in size and time of development. The third hypersensitive stern reaction resulted only from primary needle infection. Internal reaction zones developed and were stabilized from enlargement within three months. The fungus was confined to the reaction zone and the seedlings recovered from infection. There was no evidence of resistance in the needle tissue; resistance occurred only after the fungus reached the cambial area and began to spread. The fourth host-parasite interaction was typical gall aevelopment commonly found with susceptible pines and virulent Cqf. Resistant zones similar to those identified by Miller et al. (1976) were described by Jewell et al. (1982). Fungal tissues were effectively isolated in resistant pines. In 1982, Jewell et al. reported that wounding trees produced reactions similar to those produced by fusiform rust infection. Resistant seedlings differentiated tissues incompatible to rust thereby limiting the spread of the fungus. Susceptible seedlings produced initial resistant reactions, but were unable to produce incompatible tissues. Walkinshaw (1978) found seedlings that had necrotic areas surrounded by a periderm. His discussion did not point out differences between resistant and susceptible seedlings, but centered mainly on the number of necrotic, tannin-filled areas. Lundquist and Miller (1984) described a relationship between macroscopic pigmentation patterns and microscopic events in resistant lesions. These lesions resulted from artificial

PAGE 24

12 inoculation with high concentrations of inoculum. Developing phellem cells formed a boundary that effectively checked the spread of the disease and distorted infected cells, pushing them toward the epidermis. Prior to or in conjunction with periderm formation, an impermeable layer was formed. Mullick (1977) discussed three non-specific processes that are initiated during pathogen attack: 1) phellogen restoration, 2) vascular cambium restoration and 3) sapwood conduction blockage. The process triggered was dependent on depth of attack, with processes one, two and three occurring with increasing depth of injury. They may act independently or collectively depending on the attack. The author used microscopic techniques not previously used in observation of Cqf infections in pine tissues. His processes appeared to describe the observations of Frampton et al. (1983), Gray and Anerson (1983), Jewell et al. (1982), Lundquist and Miller (1984), Miller et al. (1976), and Walkinshaw (1978). Mullick (1977) proposed that host resistance or susceptibility is a function of the successfulness of completing the periderm under the pathogen's influence. During restoration, the pathogen is exposed to a host of chemicals that might influence its ability to successfully survive (e.g. phytoalexins). Virulence results from the pathogen's l) not triggering the host response, 2) blocking the triggered response or 3) growing and spreading faster than phellogen formation can respond. The microscopic differences in host response as discussed by the authors above should be represented by corresponding differences

PAGE 25

13 in macroscopic symptoms. Carson (1984) found family differences in macroscopic symptoms on artificially inoculated loblolly pine. Walkinshaw and Anderson (1983) described different macroscopic symptoms on slash pine in field studies. However, they did not present evidence of family variation for these symptoms.

PAGE 26

MATERIALS A N D METHODS Four experiments were conducted to examine fusiform rust resistance variation in slash pine. Three experiments contained f ull-sib families in five sets of crosses. The crossing sets, belonging to Brunswick Pulp Land Company (BPL), were designed as fouror five-parent half-diallels with an additional cross between the females and a male from another geographic area (i.e. a wide cross). Field tests of the crossing sets ranged from fourto six-years-old. Seed from these sets were sent to the RSC for a standard artificial inoculation study and a high inoculum concentration technique study (Lundquist et al. 1982), Experiments 2 and 3, respectively. The fourth experiment involved artificially inoculating a group of 1-0 nursery grown seedlings at RSC. These seedlings were donated by Container Corporation of America (CCA) and are part of a rust-resistant factorial mating design. Field Tests--Experiment 1 Six progeny tests containing four diallel crossing sets each were planted by BPL near Brunswick, Georgia. Progeny tests were randomized complete block plantings with three to five blocks and sevento ten-tree row plots for each family. Crossing sets one, two, four, five and nine were examined. Test sites were cleared, windrowed and bedded prior to machine planting. Tests 34 and 35 1 4

PAGE 27

l 5 Table l. Location, year of establishment, site index, fifth height and rust incidence of Brunswick Pulp Land Company's diallel progeny tests. Crossing 5th Year Rust Test Location/Count,i'. Planted Set Mean Ht. Incidence (Georgia) (yr) (feet) (%) 34 Wayne 1978 1,3,4,5 14. 5 89 35 Wayne 1978 1,3,4,5 14.8 61 36 Camden 1979 1,2,6, 10 10.4 46 37 Camden 1979 1,2,6, 10 12. l 58 40 Wayne 1980 2,4,5,9 _b/ 54 41 Camden 1980 2,4,5,9 _::/ 66 ~/Average height of dominant and codominant trees at age 25. ~/Fifth year heights were unavailable. year Site I nde x2_/ 60 55 65 60 65 65 were disked before bedding. All tests sites were in the coastal plain. Tests 34, 35 and parts of 37 were planted on well-drained ultisols. Tests 36 and 40 were planted on somewhat poorly to poorly drained spodosols. Test 41 and portions of Test 37 were planted on somewhat poorly drained entisols. Site index at age 25 and fifth year heights (Tests 34 to 37) are presented in Table l, which contains details of the tests and their locations. The parents and crosses made are shown in Table 2. Not all crosses were represented at all possiole locations. A listing of crosses in each test may De found in Table 3. The 25 parents in the crosses were evaluated previously for rust resistance in open-pollinated progeny tests by the University

PAGE 28

16 of Florida Cooperative Forest Genetics Research Program. Parental evaluations (Table 4) are expressed as weighted mean standard deviations away from progeny test mean incidence levels. T h e weighting factors were for incidence level (Sohn and Goddard 1979) Table 2. Sets of slash pine crosses used in field and artificial inoculation experiments. Female Males Set 1 0350 0098 0088 0354 1,2,3~/ 1, 2 1, 3 0350 1 1 2 0098 1 1, 3 Set 2 0146 0096 0 065 0157 1 1 1 0146 1 1 0096 1 3 1 S et 4 0352 0001 0159 02 88 1 l 1 3 0352 1,2,3 1,2,3 0001 1 1 Set 5 0141 0060 0287 0019 1 1,2,3 1,2,3 0141 1 1 2, 3 0060 1 S e t 9 0050 0295 0355 02 8 4 1 1, 2 1 0270 1 2, 3 1,2 1, 2 0295 l .. ~/l=Cross included in Experiment 1; 2=Cross included in Experiment 2; 3=Cross included in Experiment 3. 0048 1,2,3 1 2, 3 1 0071 1 1 0050 1 1 1 0047 1,2,3 1 1 0064 1,2 1,2,3 1

PAGE 29

17 Table 3. Percentage of trees infected with f usiform rust for 4 8 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia. Least F amil t Progent Tests Squares Female M ale 34 35 36 37 40 41 Mean --------------------------%-------------------------0354 0350 100.0 100.0 97. l 95.8 97.7 0 098 %.7 55.8 4 8 .2 58.9 64.9 00 88 93.3 65.2 38.2 43.7 58.5 0048 90.0 63.3 49. l 69.6 66.5 0350 OU98 100.0 52.6 47.6 57.3 64. 8 0 088 100.0 68.5 48. l 50.2 70. l 004 8 89.6 56.7 37.9 70.5 66. l 0354 78.6 97. l 95.2 0098 0088 76.7 31. l 31. 7 23.3 37. 8 0048 63.3 13.3 17.3 22.3 25.0 0 157 0146 42.9 51.9 59.4 38.8 56.9 0 096 52.3 82.4 70.4 78.8 7 9 .5 0065 56.2 7 5. l 68.3 61.l 74.3 0071 24.8 83. l 62.5 82.2 72.4 0146 0096 42.9 45.6 0065 37. 1 46.8 25.0 50.8 51.4 0071 27.8 45.4 40.3 62.5 53.5 0096 0065 46.2 56.2 56.8 68.2 65.2 0146 46.4 61.7 59. 8 0 288 0352 80.4 86.7 40.9 58.7 62.9 0001 96. 7 66.7 56.9 53.8 63.7 0159 97.0 79.3 71. 5 96. l 82.9 0050 100.0 45.8 48.7 30.0 57.7 0 352 0001 73.4 40.7 41. l 30.4 41.4 0159 81. 7 51. l 64.6 42 .8 54.9 0050 8 3.3 27.4 31.0 41. 7 0001 0159 96. 7 73.3 73.6 77. 9 0050 46.3 45. l 0019 0141 100.0 100.0 91.9 92 .8 95.8 0060 96. 7 89.6 66.9 85.5 82.9 02 8 7 93.0 71. 7 66.9 65.5 70.5 0047 88.9 45.0 8 0.0 67.3 0141 0060 95.2 40.8 76. 8 0 2 8 7 43.5 36.7 47.6 0047 91. 7 50.0 66.4

PAGE 30

Table 3--continued. Family Female Male 34 35 18 Progeny Tests 36 37 40 41 Least Squares Mean --------------------------%-------------------------0060 0287 48.6 27.5 52.5 45.8 0047 85.0 46.7 25.6 45.9 0284 0050 43.3 51.4 55.7 0 295 57.4 61.3 65.2 0355 59.2 70.4 71.3 0064 70.8 83.7 75.7 0270 0050 36.9 59.4 54.7 0295 45.0 85.5 71. 7 0355 66.4 62.5 72.2 0064 70.8 97.2 91.4 0295 0355 78.2 100.0 96.6 0 064 82.3 88.5 92.7 0270 41.9 58.6 67. 0 Tab 1 e 4. University of Florida rust evaluatioAs, standard errors and number of tests used for evaluation of slash pine parents. No. of Rus~/ Standard.Q./ No. of Rust Standard Parent Eval. Error Parent Tests Eval. Error 0l 57 9 l 9 12 0287 7 -.06 .35 0146 19 1.20 14 0047 18 1. 10 17 0096 10 .86 16 0284 13 .75 18 0065 4 .82 11 0270 7 .39 .14 0071 9 .64 10 0295 8 10 18 0288 11 -.03 18 0355 i4 .31 13 0352 10 .68 18 0064 4 .53 .26 0001 10 .41 13 0354 11 12 13 0159 14 .45 .20 0350 8 .49 .31 0050 12 1.07 .25 0098 12 1. 15 .24 0019 10 -.41 .21 0088 9 .29 14 0141 11 -.08 .09 0048 21 l.53 17 0060 7 .07 .36 0621 14 .35 .18 0298 12 .31 17 0618 18 .98 14 0620 12 .60 .21 .~./Rust Evaluation= L Infection Wei~ht x Lot Weight x Progeny Mean)/(Error Mean Square) 12). ((Test Mean ,2.lstandard trror = (Variance of a family rating/number of tests) l/2 .

PAGE 31

19 and number of families in a test (Lewis 1973). The standard error (Table 4) of an evaluation was based on the unweighted standard deviations. Data were collected from the tests in either March or June, 1984. Gall development was assumed not to change dramatically during this time. However, each test was measured within a three-day period and four of the six tests were measured during a two-week period in June. Tests 36 and 37 were measured during a one-week period in March. Data collected from each tree included survival, stem form, and the number of 1) stem, 2) limb, 3) stem to-limb, 4) fat, 5) thin, 6) typical, 7) basally truncated fusiform (BTF) (referred to as baseball bat by Walkinshaw and Anderson (1983)), 8) witches' broom, 9) twisted, and (10) sunken galls. Gall types are illustrated in Figure 1. Artificial Inoculations Experiment 2--RSC Procedures Seventeen families (Table 2) were tested at RSC using routine inoculation procedures described by Anderson et al. (1983). The inoculum originated in the Nassau County, Florida, area which is approximately 30-75 miles south of the field test sites. This same inoculum source was used by Walkinshaw et al. (1980) in developing their rust resistance prediction system. Two runs (replications) of 24-60 seedlings from each family were inoculated with a concentration of 25,000 spores per milliliter of solution. The variable number of seedlings was due to germination and survival differences. Three standard check lots were included with this

PAGE 32

Typical BTF Twisted 20 Fat Witches 1 s Broom Rust Bush Many Cankers Thin Sunken Figure 1. Gall types commonly found on field-grown slash pine trees infected with fusiform rust.

PAGE 33

21 test. Data collected from each seedling included the number of 1) symptoms without swelling (SYMN0), 2) galls present, 3) rough galls, 4) fat galls, 5) galls less than 25 millimeters long (LT25M), 7) BTF galls and 8) sunken areas on the gall and the occurrence of adventitious shoots in the galled area. Gall types are illustrated in Figure 2. Data were collected with the assistance of an RSC staff member who routinely collects data on seedlots tested there. Presence of a pink blush on seedlings 10 days after inoculation was also recorded. This symptom may be related to the red spots reported by Lundquist and Luttrell (1982). It consisted of a faint, irregular spot on the hypocotyl one to three weeks after inoculation. The relationship of this symptom to resistance is unknown. Experiment 3--High Inoculum Concentration Fifteen full-sib families from the BPL crosses (Table 2) were inoculated following techniques of Lundquist et al. (1982). The inoculum concentration of 2.0 x 10 6 spores per milliliter was extremely high, allowing only 20 seedlings per family to be inoculated with the limited quantity of available basidiospores. Eleven of the families were used in Experiment 2 (Table 2). Numbers and types of spots after 10 and 48 days were tallied for each seedling in the test. Personnel at RSC transferred the seedlings to different trays between the two measurement dates, preventing ooservation of the increase in spots per tree over time.

PAGE 34

~ SYMNO l_~ BTF ? Thin 22 Rough Gall with Adventitious Shoots I LT25M ,~V Fat Gall with Sunken Area Figure 2. Galls found on slash pine seedlings six months after artificial inoculation with fusiform rust.

PAGE 35

23 Experi1nent 4--Inoculation of 1-0 Seedlings This experiment contained six full-sib families and two check lots grown in the CCA nursery near Archer, Florida. The seedlings were lifted in January 1984, transplanted in one gallon pots filled with sterilized medium and transported to the RSC. There were two seedlings per pot and 20 pots per family or check lot. The seedlings were inoculated after their first flush of growth with the CBS system using 35,000 spores per milliliter. The pathogen source was from Louisiana as the trees flushed earlier than expected and the Nassau County pathogen source was unavailable. The six families were l) 0298 x 2002, 2) 0298 x 2005, 3) 0298 x 2013, 4) 0620 x 2002, 5) 0618 x 2007 and 6) 0621 x 2005. The four male parents (2002, 2005, 2007 and 2013) were phenotypically rust resistant, selected in stands exposed to very high levels of natural inoculum. Most trees in these stands were severely infected. The male parents are presumed likely to have genetic resistance (Goddard et al. 1975). The female orchard selections and their rust evaluations are shown in Table 4. The check lots were groups of seedlings from a rust-resistant orchard and an orchard where there was no selection for rust resistance. Observations on these seedlings were the same as those in Experiment l. Analysis To determine least squares estimates of family performance across tests, the GLM procedure in SAS (SAS Institute Inc. 1982b) was used. The model for the analysis was Yij = u +Ti+ Fj + eij' where Ti was the effect of the i th test, Fj was the

PAGE 36

24 effect of the }h full-sib family and e .. was the error lJ representing the family x test interaction. The analysis was performed on the means of gall types and locations, and the proportion of trees (arcsin square root transformation) having a particular gall type or location. Each diallel crossing set was analyzed using DIALL, a FORTRAN program (Schaffer and Usanis 1969). This program allows computation of general (GCA) and specific (SCA) combining ability, site, replication and GCA by site effects (Table 5) by a least squares approach to the analysis of variance. The computation method used an abbreviated forward Doolittle solution rather than direct matrix inversion. Estimates of the variance components were calculated by inverting the matrix of coefficients of the expected mean squares Table 5. Analysis of variance and expected mean squares for a modified half-diallel as computed by DIALL. Source df Expected Mean Squares Site~/ 1.!2/ Ve + c9VGcAxS + c10VR + c11Qs Replication within Site 4 Ve + c7VGCAxS + cgVR General Combining Ability (GCA) 4 Ve + c4VGCAxS + c5VscA + c6VGCA Specific Combining Ability 4 Ve + czVGCAxS + c3VscA (SCA) GCA x Site 4 Ve + Cl VGCAxS Error 36 Ve ~/Analysis for one site follows a similar model except for site and GCA x site effects. WThe degrees of freedom here are an example of those generated by the program.

PAGE 37

25 and multiplying the computed mean squares for each variable by this inverse. The standard error of a variance component was computed by the equation (Anderson and Bancroft 1952) std err = ( ~. (2a.MS 2 /DF .+2)) 1 1 2 1 1 1 1 where a = the coefficients of the linear combination of 1 t h e mean squares used t o estimate the variance component, MSi= the i th mean square, DFi= the degrees of freedom for the i th effect. In calculation of the F test, when the denominator mean square was smaller than a lower order effect containing that variance component in its expected mean square, the lower order effect was used as the denominator. For example, in an analysis of one site, if the SCA mean square was smaller than the error mean square, the error mean square was used as the denominator in the F test computation for GCA. DIALL did not adjust variance component computations to reflect this problem. All VGCA were computed by subtracting the mean square for SCA and dividing by the coefficient of VGCA' Traits analyzed included total number of galls, number of galls by location on the tree, and all six gall types found on trees in the field. Plot means were used in this analysis. Additionally the proportion of trees in a family having a certain gall type or gall location was analyzed. Proportion data were transformed by the arcsin square root transformation (Steel and Torrie 1960). GCA by site interaction was tested by pairing only those diallel crossing sets planted in the same year to avoid confounding year effects.

PAGE 38

26 Estimates of heritability and variance components were computed oy pooling all families at all sites as if they were in a nested design (Table 6). This method made use of all data available to compute the estimates. The procedure, while computationally feasible, would oe e x pected to produce biased estimates of variance components. The analysis accounted for the 18 crossing set-site groups (CSS), males within CSS, females within males within CSS and replications within females within males within CSS (Table 6). The NESTED procedure in SAS (SAS Institute Inc. 1982b) was used to analyze individual tree data. Incidence, rust-associated mortality and rust bush many cankers were analyzed as binary data. Table 6. Analysis of variance and expected mean squares for a four-stage nested design using individual tree data. Source Crossing Set-Site (CSS) Males (CSS) Female (Males, CSS) Replication (Females, Males, CSS) Error df Expected Mean Squares 17 58 Ve+ 9.ZOVR(FMCSS) + 31.73VF(MCSS)+ 59.23VM(CSS) 75 Ve+ 9.05VR(FMCSS) + 30.80VF(MCSS) 405 Ve+ 8.28VR(FMCSS) 4179 Ve

PAGE 39

27 Individual tree heritabilities were calculated by the equation ( 1 .-Jright 1976) h2 =-,-,--____ 4 _vM_(~C_SS~) ____ I Ve+ VR(FNCSS) + VF(MCSS) + VM(CSS) ~ vhere Ve= variance due to trees within plots, VR(FMCSS) = variance due to replications nested within females, males and CSS, VF(MCSS) = variance due to females nested within males and CSS, VM(CSS) = variance due to males nested within CSS. Family heritabilities were calculated similarly by the equation VM(CSS) Ve/59.33 + VR(FMCSS)/31.73 + VF(MCSS)/9.20 + VM(CSS) Standard errors(s) for individual heritabilities were calculated by the equation (Wright 1976) (l -hi/4) (l +Khi/4) s = --------[K(M l)/2] l/Z and for family heritabilities by the equation (Wright 1976) s = (l -t)(l +Kt) [K(M l)/2]1/2 where t = hf/4 h~= individual heritability, K = number of trees in a male half-sib family, M = number of male half-sib families. The NESTED procedure calculates a covariance analysis. Utilizing the covariance and variance estimates of the male source

PAGE 40

28 of variation, an estimate of 1/4 of the additive genetic variance, genetic correlations were calculated by the equation COVXY r =------:-,=-g (V x V ) l/ 2 where X y covxy = the covariance between two traits for males nested within CSS, Vx = VM(CSS) as described above for a trait, Vy= VM(CSS) as described above for another trait. General combining ability estimates for traits were computed using the SAS MATRIX procedure (SAS Institute Inc. 1982b). The computer program for this was developed by J. P. van Buijtenen (personal communication 1983). Data in Experiment 2 were analyzed using the GLM procedure in SAS. Factors in the random model were runs, family, runs x family and trays within runs x family was the error component (Table 7). This model was used because the results from it would be similar to results obtained by RSC. The RSC does not account for male and female effects in their screening procedure analysis. Tray means for the traits observed were used in the calculation. Pink blush was analyzed with a one-way analysis of variance on three to six tray means for each of 20 families. Variance components were determined using the VARCOMP procedure (MIVQUEO method) (SAS Institute Inc. 1982b) for females (VF), males (VM), run x males (VRM), run x females (VRM), females x males (VMF), trays within runs x females (VT(RF)), trays within runs x males (VT(RM)) and error (VE). This model accounted for

PAGE 41

29 male and female effects and was a more complete model than the previous one (Table 7). Individual tree heritabilities were calculated by the equation h 2 = I Table 7. Analysis of variance and expected mean squares for Experiment 2 using mean number of symptoms per tray. Source df Expected M ean Squaresi/ Run Ve + c4 VR x F + c5 VR F ami 1~/ 19 Ve + c2 VR x F + c3 VF Run* Family 19 Ve + c, VR x F Trays(Run Family) 112 Ve ~/Due to the unbalanced design coeffic,ents (c) in the true expected mean squares were unequal. Q/Includes three checklots. and standard errors were calculated similarly. Correlations between variaoles were calculated by the SAS CORR procedure (SAS Institute Inc. 1982a). Data from Experiment 3 were analyzed to determine within and between family differences for the number of spots found 10 and 48 days after inoculation using the model for a one-way analysis of variance for a completely randomized design. There were 15 families with approximately 20 trees per family

PAGE 42

30 Data in Experiment 4 also were analyzed using the SAS GL M procedure. Independent variables in the model were family and pots within family (Table 8). Table 8. Analysis of variance and expected mean squares for Experiment 4 using individual tree data. Source Family Pots(Family) Error df 7 146 140 Expected Mean Squares Ve + l .8Vp(F) + 35.5VF V e + l .9Vp(F) Ve

PAGE 43

RESULTS AND 01SCUSSION Field Tests--Experiment l Rust incidence in the six field tests ranged from 46% to 89% (Table l). Test 34 had the highest percentage incidence and rust-associated mortality (RAM) (Tables 3 and 9). All 30 trees of family 0354 x 0350 were dead from rust and each had a rust bush form. More than 30 cankers were observed on eacn of these trees with a maximum of 75 on some trees. To facilitate measurements, data were not collected from dead trees with a rust bush form and many galls (RBMC), but a classification was established to account for these trees. Total gall number was assumed to be 31 and the number of distorted, stem, limb and stem-to-limb galls was set at one. All other trees dead due to rust were classified as having one stem gall; this was true on most tree remains found. Although several of the other tests contained ~~MC trees, Test 34 had the highest percentage (Table 10). A comparison of Tables 9 and 10 indicates that as ~AM increased, there was an associated increase in the number of trees having a RBMC form. The frequency distribution of number of galls per tree is found in Figure 3. Distribution of galls per tree for individual families did not always follow this same pattern. 31

PAGE 44

32 Table 9. Percentage of rust-associated mortality for 48 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia. Least F ami l.z'. Progen.l'. Tests Squares Female Male 34 35 36 37 40 41 Mean __ -__ ------%-----__ --_ --_ 0354 0350 100.0 30.0 17.2 9. l 27.7 0098 32.l o.o 16.0 3.3 2.8 0088 53.3 10.3 6.7 0.0 5.2 0048 40.0 10.0 17.6 o.o 5. l 0350 0098 36.7 7. l 5.9 4.2 3.0 0088 41.4 20.7 8.0 3.7 7. 1 0048 34.5 3.3 3. l 8.7 2.3 0354 15. 2 14.3 10.2 0098 0088 10.3 o.o 13. 3 0.0 0.3 0048 3.8 0.0 3.3 0.0 0. l 0157 0146 14. 3 10.3 0.0 0.0 7.3 0096 3. l 0.0 0.0 o.o 2.8 0065 10.0 0.0 0.0 0.0 4.5 0071 0.0 0.0 0.0 0.0 2.2 0146 0096 5.3 5.9 0065 5.7 9. l o.o 0.0 5.8 0071 5.0 12. 5 o.o 0.0 4.7 0096 0065 0.0 o.o 2.7 0.0 2.8 0146 0.0 2.6 6.5 0288 0352 45.8 20.0 0.0 o.o 6.2 0001 56.7 23. 3 2.6 0.0 11. l 0159 50.0 10.3 2.9 3.0 8.4 0050 54.5 8.3 2.7 0.0 7.5 0352 0001 50.0 10.3 0.0 0.0 5.7 0159 28.6 10.3 3.2 0.0 4.6 0050 30.0 6.9 0.0 3.3 0001 0159 48.3 20.0 0.0 9.8 0050 11. 5 6.7 0019 0141 86.7 11. l 2.7 2.9 16.3 0060 53.3 9. l 2.7 0.0 7.3 0287 48.3 7. l 0.0 0.0 5.0 0047 37.9 26.7 0.0 o.o 7.7 0141 0060 0.0 0.0 1.3 0287 0.0 0.0 4.8 0047 45.8 3.8 0.0 4.4 0060 0287 0.0 o.o o.o 2.2 0047 14.3 20.0 0.0 l. 7 0284 0050 0.0 3.2 6.5 0295 2.6 0.0 7. l 0355 0.0 2.7 6.5 0064 o.o o.o 4.8 0270 0050 2.8 0.0 6.5 0295 0.0 0.0 4.8 0355 0.0 o.o 4.8 0064 0.0 5.6 8.4 0295 0355 0.0 5.9 8.4 0064 0.0 0.0 4.8 0270 0.0 0.0 4.8

PAGE 45

33 Table 10. Percentage of trees dead and classified as 'rust bush many cankers' for 48 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia Least Fami lt Pro9ent Tests Squares Female Male 34 35 36 37 40 41 Mean -__ -__ ----~---__ ----------0354 0350 100.0 30.0 6.9 4.6 24.3 0098 32. l 0.0 0.0 0.0 7.7 0 08 8 53.3 1 0 .3 6.7 0.0 5.9 004 8 40.0 10. 0 2 .9 0.0 3.7 0350 0098 36.7 5. l o.o o.o 2.0 00 88 41.4 20.7 o.o 3.7 5.7 0048 31.0 3.3 3. l 4.3 l. 9 0354 9. l 10.7 8 .3 0098 0088 6.9 0.0 6.7 0.0 0. l 0048 3.7 0.0 0.0 0.0 O. l 0157 0146 5.7 6.9 0.0 0.0 4.8 0096 0.0 0 0 0.0 0.0 2.5 0065 0.0 o.o 0.0 0.0 2.5 0071 0.0 0.0 o.o 0 0 2.5 0146 0096 5.3 6.2 0065 5.7 6. l 0.0 0.0 5.2 0071 o.o o.o 0.0 0.0 2.5 0096 0065 0.0 0.0 0.0 0 0 2.5 0146 0.0 0.0 3.6 0288 0352 45.8 20.0 0.0 0.0 5.0 0001 56.7 23.3 0.0 o.o 8.8 0159 40.0 10.3 0.0 0.0 4.0 0050 50.0 8 .3 0.0 0 0 4.9 0352 0001 46.2 10.3 o.o 0.0 4.2 0159 21.4 10.3 0.0 0.0 2.4 0050 23.3 6.9 0.0 l. 7 0 001 0159 37.9 20.0 0.0 6.7 0050 3.8 0.6 0019 0141 86 7 7.4 2.7 2.9 37. l 0060 46.7 4.5 0.0 0.0 3.5 0287 34.5 7. l o.o 0.0 2.6 0047 34.5 26.7 0.0 0.0 5.3 0141 0060 0.0 0.0 0.7 0287 0.0 0.0 3.6 0047 45.8 3.9 0.0 3.5 0060 0287 0.0 0.0 0.0 1.4 0047 14.3 13.3 0.0 0.4 02 8 4 0050 0.0 0.0 3 6 0295 0.0 0.0 3.6 0355 0.0 0.0 3.6 0064 0.0 0.0 3.6 0270 0050 0 0 0.0 3.6 0295 0.0 0.0 3.6 0355 0.0 0.0 3.6 0064 0.0 0.0 3.6 0 2 95 0355 0 .0 2 .9 5. l 0064 0.0 0.0 3.6 0270 0.0 0.0 3.6

PAGE 46

Figure 3. Frequency distribution for slash pine trees by the total number of fusiform rust galls, in six progeny tests in southeast Georgia, at four-, five-, and six-years-old.

PAGE 47

2 1.9 1.8 1.7 1.6 1.5 1.4 (/) 1.3 w w_... 1.2 0::: en I-0 1. 1 IJ..c 0~ 1 0::: :, 0.9 WO mt= 0.8 :::E:) 0.7 z 0.6 0.5 0.4 0.3 0.2 0 1 0 0 '\1 2 3 4 5 6 7 GALLS PER TREE 8 9 10 1 1 31 w U1

PAGE 48

36 Families relatively resistant to the disease had a skewed distribution with few trees with multiple galls, whereas more susceptible families often had many trees with many galls per tree. Differences in distribution patterns are illustrated by two families planted in Test 37 (Figure 4). Family 0098 x 0088 was fairly resistant (average percentage rust incidence of 37.8) and exhibited a strongly skewed pattern, whereas family 0350 x 0354 (average percentage rust incidence of 95.3) followed a different pattern (Figure 4). Griggs and Dinus (1977) found that distribution patterns reflected resistance. Results of the analyses of full-sib family means are presented in Table 11. This analysis provided least squared means for families for all traits. Twisted galls were infrequent and were deleted from this and all further analyses. All variables analyzed, with the exception of average number of sunken galls per tree and proportion of trees having sunken galls, differed significantly among tests. Sunken galls were rare with only 201 galls found. Significant differences among families were found for proportional data for gall location, the gall types witches broom and typical, and rust incidence. (In this and further discussion the term incidence refers to the presence of any galls.) However, mean gall count per tree varied among families only for witches' broom galls and total galls per tree (Table 11). Families did not vary significantly for most gall types. Except for witches broom galls, there did not appear to be any unique pattern of gall development associated with families across all planting sites. It is notable

PAGE 49

Figure 4. Frequency distribution for two slash pine families, 0098 x 0088 (resistant) and 0350 x 0354 (susceptible), for the total number of fusiform rust galls at age five years, planted in a progeny test in Camden County, Georgia.

PAGE 50

14 13 12 1 1 10 Cl) w 9 w n= I8 LL 0 7 n= w 6 m '.:? ::) 5 z 4 3 2 1 0 0 1 2 3 4 5 6 7 8 GALLS PER TREE 021 0098x0088 lSS) 0350x0354 9 10 31 w OJ

PAGE 51

39 Table 11. Results of analyses of variance of the proportion of trees \'lith different fusiform rust symptoms (arcsin square root transformation) and the mean number of these symptoms per tree for family and test differences. M ean Squares Galls Family Test Proportion of trees: Locations Any 0. 117** 0.723** Stem 0.061** 0.980** Limb 0.093** 0.799** Stem-Limb 0.061** 1.370** Types Typica 1 0.088** 0.868** Witches' Broom 0 .077** 0.752** BTFi/ 0.019 0.139** Fat 0.014 0. 183** Thin 0.014 0.038* Sunken 0.017 0. 136 M ean number per tree: Locations Total 13. 166* 561.985** Stem 9.216 401.512** Limb 0.862 13.933** Stem-Limb 0.073 0.792** Types Typi ca 1 0.823 13.696** Witches' Broom 0.016* 0.755** BTFil 0.016 o. 128** Fat 0.006 0.075** Thin 0.005 0.014* Sunken 0.003 0.004 ~/Basally truncated fusiform ga 11. Significant at 0.05 level. * Significant at 0.01 level. that there was a significant tendency for some families to have a higher incidence of stem or stem-to-limb galls than others, although the number of galls at a specific location on the tree did not vary

PAGE 52

40 significantly by family. Some bias in these results was possible because the analysis used did not allow consideration of genetic relationships among half-sib families. An analysis considering all genetic relationships was conducted using DIALL (Table 5). Crossing sets were analyzed individually at each planting site. As each crossing set involved only five parent trees, restricted and considerable differences in ranges of genetic expression were to be expected. Results of the average number of galls by type or location as well as the percentage of trees having that characteristic are presented in Tables 12 ana 13. As each of the five crossing sets were planted at two or four sites, there were 18 individual analyses for each gall type or location. For all variables, one or more major sources of variation (replication, general combining ability or specific combining ability) in at least one of the analyses was significant (Tables 12 and 13). No gall type or location had a consistent pattern of differences across all sets and planting sites. Significant differences among replications were infrequent as expected because the tests were designed with replications creating blocks to reduce variation in growth resulting from soil-site variation and not to block for differences in fusiform rust infection.

PAGE 53

41 Table 12. Significant replication (Reps) general combining ability (GCA) and specific combining ability (SCA) F values (5% level of significance) obtained from analysis of variance using DIALL for number of fusiform rust galls of several types/locations on slash pine (18 analyses for each type). Source of Variation Reps GCA SCA Galls ---number of significant F values-Location Total Stem Limb Stem-Limb Types Typical Witches Broom BTFil Fat Thin Sunken l 0 0 l 3 0 l l 2 0 ~/Basally truncated fusiform galls. 3 3 4 4 2 2 l 0 0 0 3 l 2 2 3 3 2 2 l l Average number of galls per tree of each type more often had significant SCA effects than GCA (12 versus 5). Fat, thin and sunken galls per tree had no significant GCA effects, hence no expression of additive genetic variance. However, their SCA effects were sometimes significant suggesting the presence of nonadditive genetic effects. The average number of typical, witches broom and BTF galls per tree had both significant GCA and SCA effects. The mean number of galls per location had the opposite pattern from gall types as there were more than twice as many significant GCA effects (11) than SCA effects (five). Total number of galls per tree had an equal number of significant GCA and SCA effects.

PAGE 54

42 Taole 13. Significant replication (Reps) general combining abili t y (GCA) and specific combining ability (SCA) F values (5 % level of significance) obtained from analysis of variance using DIALL for proportion (arcsin square root transformation) of slash pine trees with various fusiform rust symptoms (1 8 analyses for each symptom). Source of Variation Reps GCA SCA Galls ---number of significant F values-Locations A 11 l 8 0 Stem 0 3 2 Limb 0 3 l Stem-Limb 0 4 0 Types Typical 0 4 l Witches' Broom 2 2 l BTF~/ 0 l 3 Fat 0 0 3 Thin l l 0 Sunken 0 0 l M ortality RAM~} l 4 l RBMCf./ 0 2 2 ~/Basally truncated fusiform ga 11. ~/ R ust-associated mortality ./Rust bush with many cankers. The proportion of trees with each gall type had three more significant GCA effects than the mean number of gall types. Two of these were for the proportion of trees with typical galls and the other for thin galls. Typical and witches' broom galls had more significant GCA effects than SCA effects. There were nine significant SCA effects for the proportion of gall types, and each

PAGE 55

43 type had at least one except for thin galls. There was no evidence of additive genetic effects for the proportion of trees with fat and sunken galls The proportion of trees with each gall location (stem, limb and stem-to-limb) had more than three times the number of significant GCA effects than SCA effects. The proportion of trees with stem-to-limb galls had no significant SCA effects. The proportion of trees galled (rust incidence) also had no significant SCA effects, but did have eight significant GCA effects. RAM had four significant GCA effects while RBMC had two. These results suggested that gall locations (mean or proportion) were more controlled by additive genes than gall types were, with the exception of typical and witches' broom galls. There appeared to be more additive genetic effects for the incidence of a gall type than for the average number of gall types per tree. In crossing set two, there were significant GCA or SCA effects for only four analyses, fewer than any other crossing set. In contrast, crossing set one had significant GCA or SCA effects in 41 analyses. Such widely different results possibly are because of differences in the extent of genetic variation within crossing sets as only five parent trees were involved in each crossing set. Substantial differences among sets were expected. Lack of consistency for significant F values for the trait's variance components implies that the variances are different among the eighteen subpopulations. Each analysis can be considered a

PAGE 56

44 subpopulation because at a test there are two to four crossing sets. Also, at each test although the host genotype component is controlled, the fungal component is not. The host-parasite interaction is implicitly different, thereby producing subpopulations. Variance components and their standard errors for each crossing set in each test as computed by OIALL (Table 5) are presented in Table 28. There are discrepancies in the significance of a variance component when judged by the F test or the standard error. Large amounts of variation are inherent in the estimation of both variance components and their standard errors. This variation, in addition to the small sample size and unbalanced data, may have contributed to the discrepancies. As an example, in crossing set l, in Test 34 a highly significant F value was calculated for the GCA effect of the mean number of stem galls per tree; however, the large standard error of VGCA makes this value appear non-significant. The GCA mean square (0.431) when divided by the SCA mean square (0.00796) equaled 54.l, a highly significant F value (4 and 5 degrees of freedom). Subtracting the SCA mean square from the GCA mean square and dividing by the coefficient for VGCA (10.05) yielded 0.0421 as an estimate of VGcA The standard error of the estimate calculated by the equation on page 25 was [(2(1/10.05) 2 (0.431) 2 / (4+2)) + (2(1/10.05) 2 (0.00796) 2 / (5+2))] 112 = 0.0245 which is slightly more than half of VGcA

PAGE 57

45 Each crossing set was planted at two sites in a single year. Year effects were not estimable as tests were not planted at the same site in different years. Site effects, without confounding age or year differences, were estimated by analyzing each crossing set at two sites planted in the same year. In general, sites were different for mean gall number per tree by type or location (Table 14). GCA by site interaction (GCA x S) was significantly different in 21 of 90 (10 traits in each of nine crossing set-site Table 14. Significant site (S), replication (Reps), general combining ability (GCA), specific combining ability Galls Location Total Stem Limb (SCA) and interaction (GCA x S) F values (5% level of significance) obtained from analysis of variance using DIALL for number of fusiform rust galls of various types/ locations on slash pine, planted at several sites (nine analyses for each type). Source of Variation Site Rees GCA SCA GCAxS :-=:----number of significant F values------8 0 1 4 3 5 0 2 0 3 5 0 1 2 2 Stem-Limb 9 0 1 0 2 Types Typical 9 0 l 3 3 Witches Broom 5 0 1 0 2 tiTF~/ 3 0 1 1 1 Fat 5 0 0 1 1 Thin 1 2 0 0 1 Sunken 2 1 0 0 4 ~/Basally truncated fusiform gall. combinations) different analyses for mean number of galls per tree (all types and locations). Crossing set four had significant F-values for

PAGE 58

46 GCA x S for every gall type or location in at least one analysis, which is approximately half of the significant GCA x S effects. Confounding pathogenic variation with environmental variation in the site effect obscured the cause of the significant GCA x S effect. Site effects were significant in 40 of the 90 analyses of the proportion of trees having the various gall types or locations (Table 15). Two-thirds of the analyses for proportion of trees with Table 15. Significant site (S), replication (Reps), general combining ability (GCA), specific combining ability (SCA) and interaction (GCA x S) F values (5% level of significance) obtained from analysis of variance using DIALL for the proportion (arcsin square root transformation) of slash pine trees with various rust symptoms planted at several sites (nine analyses for each symptom). Source of Variation Site Reps GCA SCA GCAxS Galls ~---number of significant F values------Location A 11 Stem Limb Stem-Limb Types Typical Witches I Broom BTFil Fat Thin Sunken Mortal it.Y. RAM RBMC~/ 7 6 6 8 6 7 2 4 0 l 5 3 .~/Basally truncated fusiform gall. ~/Rust-associated mortality. ~/Rust bush with many cankers. l 0 l 0 0 2 0 0 l 0 0 0 4 0 2 l 0 2 4 l l 2 0 l 4 l l l 2 2 0 l l 0 l l 0 l l 0 0 3 l l 2 l 0 2

PAGE 59

47 limb galls had significant GCA effects. Proportion of trees with stem-to-limb galls was influenced more by site effects than by GCA or SC A M ajor effects on t h e prop o rtion of trees with typical galls were site and GCA, but there was significant interaction between these effects in on l y one analysis. GCA x S interactions were significant in only 11 of the 90 analyses of proportion of trees with a particular gall type or location. There were no significant GCA effects for thin, sunken or fat galls (mean or proportion) and only one significant GCA effect for the mean number of BTF galls. These results as well as those in Taoles 12 and 13 confirm that there was no additive genetic variation for sunken and fat galls in this study. There was limited evidence of additive genetic variation in thin and BTF galls. When the analyses were examined to compare significant effects for gall types with those for gall locations, it was found that gall types had more significant SCA effects than locations. GCA x S significant effects were more prevalent among gall types than gall locations. There were, however, several significant GCA x S effects for gall location. The impact of pathogenic variability and environmental variation on family performance was evident in these results. However, these tests were not designed to determine which of these aspects of site were most important. The lack of consistency of significant differences found in this study was reflected in Blair's (1970) analysis of loblolly pine. He found significant differences among half-sib families in his 1963 plantings but not the 1964 plantings. Disease incidence in the 1963 plantings averaged 50 and in 1964, 41%. Although he

PAGE 60

48 analyzed the number of galls per tree, he did not present the results of that analysis. He stated that the degree of additive genetic control (percentage incidence, galls per tree and severity index) ranged from weak to moderate and depended on the trait measured, site and year of planting. J Rockwood and Goddard (1973) found that in some progeny tests there were significant differences among half-sib families for percentage rust incidence while in others there were none. Differences among families were highly significant when progeny tests were combined. Barker (1973) also found significant and non-significant differences among families for his disease severity index depenaing on test site. When data from sites were combined there were significant differences among families. Sohn (1977) found significant differences among half-sib families of slash pine planted in several locations in two different years. Kinloch and Stonecypher (1969) found significant differences among controland wind-pollinated families of loblolly pine at all sites they analyzed for both average number of galls per tree and percentage rust incidence. Barker's (1973) combined analysis for severity index had a highly significant interaction between sites and crossing sets, but a nonsignificant interaction between sites and families within crossing sets. He found little or no evidence of genotype x environment interaction. Sohn (1977) found significant family by site interactions for percentage incidence with a binominal analysis of individual trees

PAGE 61

49 and an analysis of plot means, but not with transformed plot means. In addition, the significant differences occurred only in tests planted in 1971, suggesting that planting years were important. Sohn (1977) found rust incidence varied at some sites from year to year, and genetic variances increased with increases in incidence. Conditions favorable to rust infection varied from year to year, resulting in variation in expression of genetic effects. Results obtained from this study confirm the mixture of previous results. Families were different in their resistance to fusiform rust infection. Some symptoms of the disease including thin, fat, BTF and sunken galls showed little or no additive genetic effects. Rust incidence was affected by site, year of planting and families planted. Witches' broom, typical, total number, RAM, RBMC, stern, 1 imb and stem-to-1 imb ga 11 s manifested variation because of family and site. Analysis of transformed plot means reduced the number of GCA x S interactions just as Sohn (1977) found. Individual tree heritabilities for rust incidence, RBMC and RAM as well as mean number of total, typical, BTF, fat, sunken, stem, limb and stem-to-limb galls were calculated (Table 16) using the model found in Taole 6. Individual and family heritabilities for fat, thin and sunken galls were estimated as zero because variance components for males were zero or negative. This was expected based on the results of the diallel analysis. Estimates of h 2 for percentage rust incidence and total number of galls per tree are comparable to those of Blair (1970) for loblolly pine. His estimates of h~ for percentage rust incidence were 0.20 and

PAGE 62

50 Table 16. Individual (hf) and family heritabilities (hf) and their standard errors (s) for rust incidence, rust bush many cankers (RBMC), rust-associated mortality (RAM) and the mean number per tree of 10 fusiform rust gall types for slash pine, fourto six-years-old Trait h2 --IRust Incidence~/ o. 196 RBMC~/ 0.346 RAM~/ 0.240 Location of Ga 11 s Total 0.227 Stem 0.088 Limb 0.055 Stem-Limb o. 157 Types of Galls Typica 1 0.046 BTF-~/ 0.004 Witches' Broom o. 128 Fat 0.0 Thin 0.0 Sunken o.o .~/Binomial analysis of variance. ~/Basally truncated fusiform gall. Heritabi 1 it)'.'. s h2 _!.!f.08 0.681 12 0.809 .09 0.751 .09 0.695 .05 0.519 .04 0.397 .07 0.680 .04 0.306 .02 0.048 .06 0.643 0.0 0.0 0.0 s 31 51 .37 .34 1 5 10 .25 .09 .02 21

PAGE 63

51 O. 12 for two planting years. Additionally his estimates of hi for total number of galls per tree were 0.29 and 0.09 for the same years. Estimates of percentage rust incidence in slash pine are also similar to those obtained in this study. Rockwood and Goddard (1973) estimated hi to range from 0.035 to 0.277 in different progeny tests. Sohn (1977) estimated hi for percentage rust incidence to range from 0. 115 to 0.393, but her combined analysis estimates were 0.280 and 0.217 for two planting years. Sohn's combined estimate of family heritability for percentage rust incidence was 0.78, with a range of 0.389 to 0.752. The highest estimates of heritabilities in this study, both individual and family, were for rust incidence, RBMC, RAM and total galls per tree. These heritabilities suggested that selection against death and malformation because of Cqf is possible. Gall locations had moderate heritability values. Heritabilities of gall types, with the exception of witches' broom and typical, were very low or zero. Witches' broom and typical galls therefore, were the only gall types with evidence of additive genetic effects based on heritability estimates and the diallel analysis. Genetic correlations from the model in Table 6 are found in Table 17. Total gall number is highly correlated with RAM (0.99) and RBMC (0.98). The correlation between total gall number and rust incidence was lower at 0.51. Rust incidence was correlated to a lesser degree with RAM and RBMC. The average number of typical galls was positively correlated with incidence (0.71) and negatively

PAGE 64

Table 17. Genetic correlations between various fusiform rust gall types and locations (mean number per tree), rust incidence and mortality (RAM and R8MC) for slash pine, fourto six-years-old. StemWitches Rust Ga 11 Types~/ Stem limb Limb Typical Thin BTF Sunken Broom Fat Incidence RAM RBMC Total Stem Limb StemLimb Typical Thin IHF Sunken Witches' Broom Fat Rust Incidence RAM 0.60 -0.86 -0.43 0.32 0.25 0.87 -1 .05 -!2.1 -2.61 0.24 0.28 -0.83 0.86 0.96 3.33 1. 18 0.97 1.63 0.84 3. 12 1.24 2.33 !IBTF = Basally truncated fusiform gall; RAM =Rust-associated mortality; RBMC cankers. ~/Undefined, at least one variance component was zero. 0. 51 0.99 0.98 0.88 0.52 0.37 0.75 -0.95 -0.96 0.67 -0.52 -0.59 0.71 -1 .08 l. 11 0.62 -2.96 -2 .82 0.94 -0.03 -0. 12 0.25 0. 19 1.00 =Rust bush, many U7 N

PAGE 65

53 correlated with RA M (-1 .08) and R BMC (-1.11). Some genetic correlations were larger than one or smaller than negative one. Although these values are conceptually impossible they are computationally feasible. Genetic correlations are usually subject to large sampling errors and are therefore, seldom very precise (Falconer 1960). On the basis of genetic correlations, field counts of total number of galls or the number of typical galls would provide more information about potential RAM and malformation than simply recording presence or absence. Disease severity estimates such as Blair's (1970) severity index incorporate some of these features. General combining ability estimates (gca) obtained from full-sib progeny tests are an indication of parental value. Table 18 contains gca estimates of parental values for several gall types. Parents 0047, 0048, 0098, 0287 and 0352 were the best general combiners for rust resistance and parents 0019, 0064 and 0350 were the worst on the basis of the traits in Table 18. It is interesting to note that University of Florida (UF) evaluations (Table 4) were high (greater than 1.0) for 0047, 0048 and 0098. However, the evaluation for 0352 was 0.68 and 0287 had a slightly below average (-0.06) evaluation. The UF evaluation for 0287 had a standard error of 0.35 representing a range of -1.28 to 1.54 for its individual test evaluations. Progeny performance in tests reported here indicate parental resistance levels to be higher than previously assessed and, particu l arly in the case of 0287, may reflect differences in crossing partners as suggested by Powers and Zobel (1978).

PAGE 66

54 Table 18. General combining ability estimates of the proportion of trees with fusiform rust galls for 25 slash pine parents based on progeny performance in six locations in southeast Georgia, aged four to six years. Witches' StemRust Parent Typical Thin BTF~/ Sunk Fat Broom Stem Limb Limb Incidence -----------------------------------%------------------------------------l l'3 47 4 1 3 50 60 64 65 71 08 96 9 8 141 146 157 159 270 2 8 4 287 288 295 350 352 354 355 41.5 95.6 24. l 2 8 .7 39.4 51.2 99.4 47.2 52.6 33.3 57.0 25.9 7 2. l 35.2 69.6 7 8 .7 31.9 3 8 .5 12.3 61.4 67. 8 96. l 26.9 8 2.4 79.0 2.8 6.7 2. l l. l -1.4 6.2 l. 7 0.6 -2.7 -0.5 2. 8 5. l -0.2 3.9 3 ~ 7 4.2 8.9 1.0 3. l 8 .9 3.0 4. 1 0.3 -0.02 3.2 2.3 8 .5 5.5 4.2 3.3 6.6 22.2 l. 3 3.6 3.5 5.9 4.4 3. l 3. l 12.8 0. l 2.2 3.8 -1.9 6.6 4. 2 5.6 7.9 12.2 0.7 3.8 5.8 7.3 0.4 0.6 7.3 12. 3 l. 3 2. l 0. l 1.6 1. 6 3. l 0. l 3.8 9. l 9.0 6.8 0.2 7.5 5.8 0.8 0.9 1. 7 14.0 ~/Basally truncated fusiform gall. 1.5 l. l 4.6 -0.3 5.2 0.7 16. l 0.7 0.5 l. 2 3. l 4.8 3.9 l. 2 9.0 2.5 -4.8 -4.7 3.0 2.2 9.9 6.8 2.5 7. l 3.9 43.0 62.4 10.7 1.5 30.6 12.8 56.9 36.8 29.0 16.3 24.6 -7.2 48. l 37.0 30.9 49.6 27.6 7.7 8.5 50.6 46.4 57.9 15.9 65.7 52.2 22. l 56.6 9.0 -2.3 19.5 32.3 35.9 18.9 23.5 3.6 21. 9 -2.0 49.8 27.0 26.7 40.0 37.6 41.2 l. 2 45. l 28.7 50. 8 13.4 57.7 39.6 45.7 96.2 22.2 31. 5 32.6 46.4 89.9 4 8 .5 37.0 39.9 5 7. l 16.0 72.4 33.8 69.8 67.0 38.7 31. l 6.9 64.9 64.7 92.4 26.0 81. 9 71. 2 24.7 46. l 7 3 6.6 17.6 18.3 68.7 25.0 34.3 5.2 20.7 2.9 45.7 23.2 31.0 30.2 22. l 16.4 13.3 36.9 53. l 53.0 11. 3 55.6 63.5 52.0 106.3 34.6 29.9 47.3 62.9 109.2 64.3 66. 8 35.7 70.4 31. 3 86.4 36. l 82. l 89.2 59.3 50.2 24.0 78.0 82.5 100.8 33. 8 93.5 96. 1 Specific combining ability of a particular cross (sea) is defined as the family deviation from the mean of the two parents. Specific combining abilities of families for percentage rust incidence and fat galls are presented i~ Table 29. Specific combining ability values for the proportion of trees with fat galls ranged from -8.00 (0284 x 0064) to 3.84 (0295 x 0064). Crosses with 0064 had the greatest positive and negative deviations from parental means (sea) for the proportion of trees with fat galls. Parent 0064

PAGE 67

55 had the highest estimate of gca for this trait as well as for proportion of trees with sunken galls and percentage rust incidence. Specific combining ability estimates for families in which 0064 was a parent varied greatly for these last two traits also. Specific combining ability may be an influence in the relatively good performance of parent 0287. Crosses between this parent and 0019 or 0060 have sea's of 5.31 and 2.25, respectively, for percentage rust incidence. When crossed with 0141, however, it has a sea of -7.63. Parent 0141 has a gca estimate midway between those of parents 0019 and 0060. Further crosses with 0287 and 0141 and other parents may disclose the reasons for this apparent good combination. In several of the analyses for fat galls SCA effects were significant. This fact, along with sea values for fat galls in some crosses, suggested that nonadditive genetic effects were important in this trait. Artificial Inoculation Experiment 2--RSC Procedure Rust incidence (gall formation) following the RSC procedure ranyed from 14.4 % to 88.7% among the 20 seedling lots tested (17 full-sib families and 3 check lots). Proportion of galled seedlings was higher in all families than in the two resistant checks, and only two families had lower galled proportions approaching that of the resistant checks. Seven of the families had a higher properties of galled seedlings than the susceptible check. Means for all traits measured are indicated in Table 19 by seedling lot. A

PAGE 68

56 Table 19. Percentage of seedlings in 20 slash pine families with various reaction types six months after artificial inoculation with fusiform rust and the resistance index of the families. Fam i l ,l Pink Incidence Symptom Less than Resistance Female Male Slush 25 mm Healthy of Rust No Swelling Rough -----------------------%-------------------------3000~1 15.8 24.4 66.2 9.4 11. 5 25.9 4oooe/ 5.0 41. 7 31. 7 29.2 5.0 11. 7 500Cf.l 1.8 57.6 14.4 31. 7 3.3 9.4 0354 0350 1.6 18.0 63.9 18.0 11. 7 36. l 0354 0088 2.7 15.3 48.9 38.0 10.9 30.2 0354 0048 4.2 15.5 62.9 28.2 0 25.9 0350 0088 8.3 20.0 56.0 30.0 8.0 26.0 0350 0048 19.4 33.8 53.0 15.9 4.6 20.5 0352 0001 7.3 38.8 34.9 27.2 5.3 9.3 0352 0159 3.2 34.2 34.9 30.9 11. l 5.3 0270 0050 41.0 42.6 38.5 13.9 7.8 14.6 0019 0060 6.5 8.9 88.7 2.4 17.9 23. 2 0019 0287 16.9 18.3 70.7 15.5 6.4 37.9 0019 0047 44.3 38.0 49.4 17.3 4.7 23.3 0141 0287 13.3 28.3 58.3 16.7 l. 7 33.3 0270 0355 31. 9 16.0 79.8 5. l 17. 2 25.7 0 284 0295 12.2 19.6 77 .6 4.3 7.2 23.6 0284 0064 13.4 20.5 70.4 9. l 8.8 25.2 0270 0295 28.6 13.6 82.2 4.2 23.8 22.7 0270 0064 39. l 8.9 88.7 1.8 22.7 16.7 E/ RSC susceptible check lot, Georgia Slash QI RSC resistant check lot, FA-2 :=.I RSC resistant checK lot, LA11 resistance index value was calculated for the seedling lots tested based on the RSC index equation (Robert Anderson, personal comnunication 1984) (Table 19). A high resistance index value indicates a high occurrence of resistant reactions. The high index value for the susceptible check was unexpected. Families were significantly different for all traits analyzed (Table 20) (see model, Table 7). Differences among families for Index 209.4 195. 2 144.8 159.3 122.2 140.4 102.3 171 2 197. l 163. l 183.3 61.3 132.9 157.5 166.6 128.4 108 4 112. 5 142. l 149.7

PAGE 69

57 Table 20. Mean squares for the analyses of variance of the proportion of of slash pine seedlings per family expressing various disease symptoms resulting from artificial inoculation with fusiform rust at the Resistance Screening Center, Asheville, North Carolina. Disease Symptom Healthy SYMN~7 Incidence of Rust Rough Gall < 25 mm Gall Thin Ga 11 Fat Ga 11 BTF Ga 11.Q./ Adventitious Shoots Sunken Areas ~/symptom, no swelling . ~./Basally truncated fusiform gall. *Significant at the 5% level. **Significant at the 1% level. Source of Variation Run Family o. 165** 0.088** 0.006 0.122** 0.306** 0.400** 0.001 0.029** 0. 134** 0.110** o. 100** 0.042** 0.164** 0.011* 0.002 0.001* 0. 103** 0. 13 7** 0.006 0.056* mean number of fat and BTF galls and galls with sunken areas were significant at the 5% level while other traits were highly significant (1% level) (Table 20). Carson (1984) analyzed five of these traits (galled, SYMNO, rough, LT25M ana thin) on loblolly pine. With a composite inoculum source, there were significant differences among families for four of these but not for the fifth, LT25M. Individual heritabilities were calculated using the model described on page 28 for the proportion of seedlings with the various artificial inoculation responses and the mean number of

PAGE 70

58 responses per tree (Taole 21). Proportion of trees with SYMN0 and galls with adventitious shoots had fairly high individual heritabilities, 0.28 and 0.33, respectively. Mean number of galls per seedling had the highest heritability with a value of 0.38. These results indicated that several resistant and susceptible characteristics are heritable traits. Correlations between variables in the greenhouse produced several significant observations (Table 22). The number of healthy seedlings in a family had significant negative correlations with mean number of galls, rough galls and galls with adventitious shoots, and was negatively related to other reactions except SYMN0. In general, it should be expected that symptoms suggesting resistance (such as SYMN0) should be positively correlated with other resistance symptoms and negatively correlated with symptoms suggesting susceptibility. Galls with adventitious shoots had a highly significant negative correlation with healthy and SYMN0 and highly significant positive correlations with number of galls and rough galls. Galls with adventitious shoots appeared to be a susceptible reaction type. Sunken areas were negatively correlated with SYMN0 and positively correlated with BTF, rough, and mean number of galls.

PAGE 71

59 Table 21. Individual heritabilities (h2) and their standard errors (s) for artificial inoculation (AI) symptoms by both proportion of trees with the symptoms and mean number of symptoms per tree. Proportion Mean A I Symptoms~/ h2 s h2 s Healthy 0.08 0.08 SYMNO 0.28 0. 18 0.32 0.20 Incidence of Rust 0.23 0. 16 0.38 0.23 Rough Gall 0. 18 0. 13 0. l l 0.09 L T25M Gall 0. l 0 0.09 0.08 0.08 Thin Ga 11 0.06 0.07 0.06 0.07 Fat Gall 0.02 0.05 0.01 0.04 BTF Gall 0.05 0.06 0.05 0.06 Adventitious Shoots 0.33 0.21 0.32 0.20 Sunken Areas o. 11 0.09 0.09 0.08 ~ISYMNO =symptom, no swelling; LT25M =gall less than 25 mm long; BTF =basally truncated fusiform gall. The correlation between SYMNO and healthy was small (less than 0.4) and not significant. Healthy seedlings may be escapes or may have one of the two interactions Miller et al. (1976) described as either total incompatibility (i.e., no germination or penetration by basidiospore) or subliminal infection possibly due to an avirulent pathogen. SYMNO appeared to be a hypersensitive reaction.

PAGE 72

60 Table 22. Pearson product-moment correlations between the mean number of artificial inoculation (AI) symptoms per tree of slash pine seedlings i noculated w ith fusiform rust, base d on progeny of 20 families. A I S mp t o ms AI Sym p t o m Inc i den c e a b A d v entiti o us Su n ke n Sym pto m s n o s w e lling of l
PAGE 73

61 There were significant negative correlations between percentage rust incidence in the field and proportion of healthy and SYMNO seedlings in the greenhouse (Table 23). field rust incidence had significant positive correlations with proportion of seedlings infected with any gall type, rough galls and galls with adventitious shoots. The proportion of trees with stem, limb, stem-to-limb and typical galls in the field were significantly negatively correlated with the proportion of healthy seedlings in the greenhouse. There were no significant correlations between RAM and RBMC and any of the I, JJ, ~ ( artificial inoculation symptoms. r'r Ut c,;i The positive correlations between the proportion of trees galled in the field and rough (significant) and LT25M (nonsignificant) galls in the greenhouse were surprising as both symptoms are considered resistant (Charles Walkinshaw, personal communication 1984). Walkinshaw and others (1980) found a negative correlation bccween University of Florida rust evaluations and the proportion of seedlings with smooth galls. Therefore, they included this variable as well as proportions of trees with fat galls and SYMNO in their index. Rough galls, seemingly the opposite of smooth galls, were categorized by Carson (1984) and by RSC for their current index (Robert Anderson, personal communication 1985). LT25M had a positive relationship with percentage rust incidence and total number of galls per tree in the field study. Seedlings with LT25~ galls nine months after inoculation in the 1984 study of Griggs and others were free of gall symptoms and active

PAGE 74

62 Table 23. Pearson product-moment correlations between field galls and artificial inoculation symptoms (proportions) and the resistance index of 17 slash pine families and fusiform rust. Artificia l Inoculation Sym ptoms4/ Field Ga lls Proportion o f Trees: Typical Thin B TFE.I S un ken Fat Witches' Broom S tem Lim b S tem-Li m b Incidence of Ru st KA:-f.l RBM CS / M ean number of gal ls: Ty pical Thin B TFE.I S unken witc hes' B room Fa t ::i tem Lim b S tem-Limb Tota 1 Healthy 0 .54* -0 32 -0. 20 0 2 4 0 25 -0. 29 0.52 -0.60* 0 .54* 0 .6 3 ** ll 16 -0. 1 3 -U.34 -0. 1 8 -0. 14 -0.30 0 24 0 16 0 20 0 27 -0.53* -0. 29 SYMNO 0 27 0.64** 0 .01 -0.70** -0. 2 5 -0. 0 1 -0.4 8 -0. 1 8 -0.4 8 0 51* 0 03 o 11 0 .07 -0.61** 0.29 -0.75** 0.05 0.39 0 02 0.20 0.48 -0.05 Incidence of Rust 0 .5 0 0 57 0 1 4 0.55* 0.01 0 .14 0 .5 8* 0 48 0.60* 0.69** 0.09 -0.01 0.22 0.4 8 -0.03 0.61** 0.08 0 11 0.08 0.0 8 0.62** 0. 14 Rough 0.32 0 63** 0 03 0 64 ** 0 .02 0.38 0 46 0.34 0.46 0 .54* o 13 0 .06 0 .01 0.50 -0. 22 0 76"* 0.34* 0 13 0 11 0.01 0.30 0. 12 L T 25M U .3 5 0 0 3 0 04 -0.22 0 1 2 0. 3 6 0 .42 0 39 0 42 0 .37 0 34 0 4 1 0 .16 0 0 6 0 09 -0. 2 4 0 .25 0 19 0 44 o 1 2 0 38 0 44 ~ / 5 YMN0 = symptom no swe 11 i ng, LT25M = ga 11 less than 2 5 mm 1 ong. ~ /Ba sally truncated fusiform gall. /R AM = rust-associated mortality, RBMC = rust bush many cankers. *Correlation significant at the 5% level. ** C orrelation significant at the 1 % level. Fat 0 .3 8 0. 19 0 .47 0.02 0.57* 0 24 0.06 0 .30 0 12 0 2 1 0 11 -0. 16 0.68** -0. 28 0 .57* 0. 0 7 0 01 0.56* 0 20 0 68* 0.08 0 0 5 Adve ntitious Shoots 0 .44 0 .5 2 0 27 0.61** 0.07 -0. 0 3 0.48 0 .36 0 .4 2 0 .59* 0 00 0 13 0.09 0 45 0 08 0 .6 8* 0 0 6 -0. 32 -0. 05 0 .05 0.42 0 02 R esistar lnde> 0 41 0. 2: 0 l ; -0. 1 : -0 .0! 0. 1 1 0 3; 0 4 : 0. 1 : 0 4! 0 0f 0 0~ 0 3~ 0 I; 0 l 0. ] I 0 11 o o; o o: 0 .3 ( 0 1 : 0 0t

PAGE 75

63 mycelium 23 months later, and therefore, could be considered a resistant symptom in that study. There were no significant correlations between field symptoms and the RSC resistance index in this study. Calculation of the correlations did not include the check lots. The resistance index did not successfully reflect family field performance in this study for any field symptom. Griggs and Dinus (1977) found extremely low correlations between artificial inoculation and field tests at young ages, but the correlations increased and were significant at age 12. They implied that the correlation was improved after incidence had increased over time. This suggested that artificial inoculation predicted performance better on sites with high incidence than on sites with moderate incidence. Because field tests in this study had moderate to high incidence levels it was expected that artificial inoculation would predict performance more accurately. Pink blush was observed on the seedlings 10 days after inoculation. Families were significantly different in the percentage of seedlings expressing this character. Percentage of seedlings with a pink blush ranged from 1.6 to 44.3 (Table 19). In some cases the blush was a faint, small pink spot on the stem just above the cotyledons. In other cases the spot was larger, sometimes encircling the stem, or varied in color to a rich red. Pink blush was negatively correlated with SYMNO (r=-0.53) and essentially not correlated with percentage healthy trees (r=-0.04). It was

PAGE 76

64 positively related to percentage of galled seedlings (r=0.30). Pink blush does not appear to be a resistant symptom. However, a more intensive study is needed to conclude this. Experiment 3--High Inoculum Concentration Numbers of spots on each seedling were counted 10 and 48 days after inoculation (Table 24). Red spots observed 10 days after inoculation were generally small and confined to the stem. After 48 days, some spots were very large and numerous spots were present on the primary needles. Large spots in some cases appeared sunken. Color varied from red to purple. There were highly significant differences among families for spots both 10 and 48 days after inoculation. Lundquist and Luttrell (1982) stated that most families became pigmented within 48 days. Therefore, the results at 48 days for proportion of trees with spots were expected. Average number of spots per seedling 10 and 48 days after inoculation was correlated to field performance of the same families using Pearson correlations (SAS Institute Inc. 1982a). Significant negative correlations were found between proportion of trees with spots after 10 days and proportion of trees with typical galls, limb galls and rust incidence (r > -0.59). Lundquist et al., (1982) found that under high inoculum concentrations resistance was ineffective in family FA-2, a RSC resistant check lot. This family (4000) was included as a check lot in both this experiment and Experiment 2. Very little pigmentation occurred in the test, supporting the conclusion of Lundquist et al. (1982), that resistance of this family was not effective under high

PAGE 77

65 Table 24. Results of artificial inoculation with high concentrations of inoculum (2.0 x 106 spores/ml) on 17 slash pine families, Resistance Screening Center, Asheville, North Carolina. 10 Oats after Inoculation 48 Oats after Inoculation Family N o. of % of Trees No. of % of Trees Female Male S~ots Mean With Spots Spots Mean with Spots 300~ 0 0 0 7 0.4 25.0 400@ 0 0 0 4 0.2 20.0 0354 0088 1 0.05 5.0 299 14. 9 100.0 0098 0088 12 0.63 42. l 281 14.8 94.7 0354 0048 1 0.05 5.0 110 5.5 90.0 0352 0001 10 0.5 45.0 412 20.6 95.0 02 8 8 0159 0 0 0 130 6.5 100.0 0019 0047 l 0.05 5.0 154 7.7 100.0 0350 0048 1 0.5 5.0 404 20.2 100.0 0270 0050 70 3.5 85.0 362 18. 1 100.0 0141 0287 4 0.2 15.0 152 7.6 95.0 0 350 0354 0 0 0 154 7.7 100.0 0019 0060 5 0.25 10.0 148 7.4 85.0 0096 0141:i 34 1. 7 65.0 325 16.2 95.0 0019 02 8 7 l 0.05 5.0 79 3.9 80.0 0352 0159 13 0.93 64.3 134 9.6 92.9 0270 0064 6 0.3 25.0 230 11. 5 100.0 ~!RSC susceptible check lot, Georgia Slash. !~/RSC resistant check lot, FA-2. inoculum loads. Lundquist and Miller (1984) compared their results with Mullick (1977) and other researchers. They reported that the pigmentation seen in their study signified that a host had reacted to prevent the spread of the fungus. They further stated that pigmentation was not the complete mechanism of resistance, but was symptomatic of physiological and anatomical resistance mechanisms. Further research is needed to understand these mechanisms before breeders can exploit the variation that is present.

PAGE 78

66 Experiment 4--Inoculation of 1-0 Seedlings Galls developing on 1-0 seedlings more closely resembled galls on trees in the field than those on the seedlings in Experiment 2. This suggested that host response varied because of maturation state as concluded by Day (1974). Families were significantly different for rust incidence, total number of galls per tree, fat galls, thin galls, and typical galls. Galls were located on the growth flush that had occurred just prior to inoculation. Many infected trees had stem-to-limb galls that probably would result in a RBMC if the trees were allowed to mature. Goddard and Schmidt (1971) also found differences among families in rust incidence when they inoculated 1-0 seedlings, and their results were correlated with the results of inoculation of the same seedlings at age six-to-ten weeks. The susceptible check had significantly greater rust incidence than the other famities (Table 25). Family 0621 x 2005 had the next highest infection level. Many galls on these trees were short and fat and did not encompass the entire stem circumference. Orthogonal contrasts revealed that for total galls per tree the susceptible check was significantly different from 0621 x 2005 and 0298 x 2005 and that these families also differed. In addition, the contrasts disclosed differences in rust incidence between 0621 x 2005 and 0298 x 2005. Male 2005 was apparently susceptible. This small sample size and lack of field data limit conclusions from this study. Field performance in Florida of these families should not be predicted from these results since the pathogen was not local. However, it can be seen that seedlings from the

PAGE 79

67 Table 2 5. Comparison of family means for several fusiform rust d i sease sym p toms on a rtificia l l y inoculated 1-0 slash pine seedlings. Family D isea s e Symptom~/ Female Male % Rust Incidence Total Fat Thin Typical --------g a lls per tree-------Susceptible Checi
PAGE 80

68 Research Perspectives Disease results from the interaction of a host, a pathogen, and tt1eir environments. Although fusiform rust disease of slash pine has been studied intensively for the last 20 years, many questions remain unanswered. It is known that the fungus varies in its virulence. Carson (1984) found disease symptoms varied within loblolly pine families when different pathogen sources were used to inoculate seedlings. It can be concluded, therefore, that disease symptoms may vary because of the pathogen. However, additional research is necessary to corroborate these findings for slash pine and other species. Environmental differences, such as site or year, cause variation in rust incidence by affecting sporulation and infection. However, the impact of environment on variation in disease symptoms is unclear. Field disease symptoms in this study varied by site, but the fungal component was confounded with environment. No clear statement was possible of the effect of environment on disease symptoms. One method of addressing environmental effects may be to inoculate l-0 seedlings as in Experiment 4 and then outplant them at different sites. Future studies could include development of the relationship between microscopic processes described by Miller et al. (1976) and Mullick (1977) and gall types. Griggs et al. (1984) observed that LT25M galls disappeared over time. What of other gall types? Do they differ or change over time? Does the environment affect gall

PAGE 81

69 development? Research aimed at answering these questions may add valuable information to the knowledge of disease resistance in slash pine.

PAGE 82

CONCLUSIONS Additive genetic variation of slasn pine in response to natural infection with fusiform rust exists for many symptoms: number of stem, limb, stem-to-limb, total, typical and witches' br o om galls, the proportion of trees with these galls, incidence of rust, ~AM and RB M C. However, this variation is not consistent. Little or no genetic variation is apparent for thin or sunken galls. Fat galls have evidence of specific combining ability, an expression of nonadaitive genetic effects. The existence of genetic variation appears to depend on site, planting year, inoculum and the sample of families in a test. Individual tree and family heritabilities for the field traits--rust incidence, RAM, RBMC, total galls per tree, stem-to-limo galls and witches' broom galls--were relatively high and, where comparisons were possible, similar to those found in other studies. Individual heritabilities were 0. 196, 0.346, 0.240, 0.227, O. 157 ana 0. 128, respectively. Family heritabilities were 0.68, 0.81, 0.75, 0.70, 0.68 and 0.64, respectively. KAM and R~MC were the most severe reactions to rust infection and were the most heritable in this study. Selection against families with high incidence of these traits is imperative in breeding for increased rust resistance. 70

PAGE 83

71 Individual heritabilities of artificial inoculation symptoms were best for proportion of trees with SYMNO (0.28), rough galls (0.18), galled (any type) (0.23) and galls with adventitious shoots (0.33). Variation among full-sib families existed for all traits observed. Healthy, SYMNO and galled (any type) were correlated with several field symptoms but were more often significantly correlated with the proportion of trees with a field symptom than with the mean number of symptoms. Rough galls, considered to be a resistant symptom, had a significant positive correlation with percentage rust incidence in the field. The RSC resistance index did not successfully predict field performance of the full-sib families in this study. The high inoculum concentration technique was not successful in predicting field performance. However, it may be useful in screening families whose resistance could be overcome in the field with high inoculum concentrations. Significant differences among families existed for their response to this inoculation procedure. One-year-old progeny of phenotypically rust-resistant males were successfully inoculated in this study. Families with three of the four males as a parent were as resistant as a rust-resistant check. The fourth male was apparently not resistant. Selection of phenotypically rust-resistant slash pines, therefore, continues to De a viable method of adding resistant selections to a breeding program.

PAGE 84

LITERATURE CITED Anderson, R. L., and T. A. Bancroft. Research. McGraw-Hill, New York. 1952. Statistical Theory in 399 p. Anderson, R. L., C.H. Young and J. D. Triplett. 1983. Resistant screening center procedures manual: A step-by-step guide used in the operational screening of southern pines for resistance to fusiform rust. USDA Forest Serv Forest Pest Manag Rep 83-1-18. 55 p. Blair, R. L. 1970. Quantitative inheritance of resistance to fusiform rust in loblolly pine. PhD Thesis, North Carolina State Univ, Raleigh. 87 p. Barker, J. A. 1973. Location effects on heritability and gain predictions for ten-year-old lob1olly pine. Ph O Thesis, North Carolina State Univ, Raleigh. 105 p. Carson, S. 0. 1984. Indirect screening of loblolly pine for fusiform rust resistance through controlled inoculation. PhD Thesis, North CarJlina State Univ, Raleigh. 57 p. Day, P.R. 1974. Genetics of Host Parasite Interaction. W. H. Freeman and Company, San Francisco, CA. 288 p. Dinus, R. J., G. A. Snow, A.G. Kais and C.H. Walkinshaw. Variability of Cronartium fusiforme affects resistance strategies. In Proceedings 13th Southern Forest Tree Improvement Conference, p 193-196. National Technical Information Service, Springfield, VA. 262 p. 1975. breeding Dwinell, L. D. 1974. Susceptibility of southern oaks to Cronartium fusiforme and Cronartium guercuum. Phytopathology 64:400-403. Dwinell, L. 0. 1977. Biology of fusiform rust. In Management of Fusiform Rust in Southern Pines (R. J. Oinus and R. A. Schmidt eds) p 18-24. Symp Proc Univ Florida, Gainesville. 163 p. 72

PAGE 85

73 Falconer, D. S. 1960. Introduction to Quantitative Genetics. The Ronald Press Company, New York. 365 p. Frampton, L. J., Jr., H. W. Amerson and R. J. Weir. 1983. Potential of in vitro screening of loblolly pine for fusiform rust. In Proceedings 17th Southern Forest Tree Improvement Conference, p 325-331. National Technical Information Service, Springfield, VA. 375 p. Goddard, R. E., and J. T. Arnold. 1966. Screening select slash pines for resistance to fusiform rust by artificial inoculation. In Breeding Pest Resistant Trees (H. D. Gerhold, E. J. Schreiner"-;R. E. McDermott, and J. A. Winieski, eds) p 431-435. Pergamon Press Ltd, London. 505 p. Goddard, R. E., D. L. Rockwood and H. R. Kok. 1984. Cooperative Forest Genetics Research Program. 26th Prag Rep, Univ Florida, School Forest Resourc Conserv Rep No 35. 26 p. Goddard, R. E., and R. A. Schmidt. 1971. Early identification of fusiform rust resistant slash pine families through controlled inoculation. In Proceedings 11th Southern Forest Tree Improvement Conference, p 31-36. National Technical Information Service, Springfield, VA. 284 p. Goddard, R. E., R. A. Schmidt, R. K. Strickland and H. R. Kok. 1972. Cooperative Forest Genetics Research Program. 14th Prag Rep Univ Florida, School Forest Resourc Conserv Rep No 20. 26 p. Goddard, R. E., R. A. Schmidt and F. Vande Linde. 1975. Effect of differential selection pressures on fusiform rust resistance in phenotypic selection of slash pine. Phytopathology 65:336-338. Goddard, R. E., and 0. 0. Wells. 1977. Susceptibility of southern pines to fusiform rust. In Management of Fusiform Rust in Southern Pines (R. J. Dinusand R. A. Schmidt, eds), p 52-58. Symp Proc Univ Florida, Gainesville. 163 p. Gray, D. J., and H. W. Amerson. 1983. In vitro resistance of Pinus taeda embryos to the fusiform fungus, Cron art i um guercuum f. sp. fusiforme: Ultrastructure and histology. Phytopathology 73:1492-1493. Griggs, M. M., ana R. J. Dinus. 1977. Patterns of fusiform rust increase and their implications for selection and breeding. In Proceedings 14th Southern Forest Tree Improvement Conference, p 43-52. National Technical Information Service, Springfield, VA. 352 p.

PAGE 86

74 Griggs, M. M., R. J. Oinus and G. A. Snow. 1984. Inoculum source and density influence assessment of fusiform rust resistance in slash pine. Plant Dis 68:770-774. Griggs, M. M., and R. A. Schmidt. 1977. Increase and spread of fusiform rust. In Management of Fusiform Rust in Southern Pines (R. J. Dinus and R. A. Schmidt, ed.), p 32-38. Symp Proc Univ Florida, Gainesville. 163 p. Griggs, M. M., and C.H. Walkinshaw. 1982. Diallel analysis of genetic resistance to Cronartium guercuum f. sp. fusiforme in slash pine. Phytopathology 72:816-818. / Hirt, R.R. 1964. Cronartium ribicola: Its growth and reproduction in the tissues of eastern white pine. St Univ Col Forest Syracuse Univ, NY, Tech Pub 86. 59 p. Hubbard, S. 0., and R. L. Anderson. 1980. The resistance screening center. USDA Forest Serv Gen Rep SA-GR 16. l p. Jewell, F. F. 1960. Inoculation of slash pine with Cronartium fusiforme. Phytopathology 50:48-51. ( Jewell, F. F., D. C. Jewell and C.H. Walkinshaw. 1982. Histopathology of anatomical mechanisms for resistance to fusiform rust in slash pine. In Resistance to Diseases and Pests in Forest Trees (H. M. Heybroek, B. R. Stephan and K. von Weissenberg, eds.) p 110-118. Centre Agric Publ Document, Wageningen, The Netherlands. 503 p. Kinloch, B. B., Jr., and R. W. Stonecypher. 1969. Genetic variation in susceptibility to fusiform rust in seedlings from a wild population of loblolly pine. Phytopathology 59:1246-1255. Laird, P. P., and W.R. Phelps. 1975. A rapid method for mass screening of loblolly and slash pine seedlings for resistance to fusiform rust. Plant Dis Rep 59:238-242. Lewis, R. A. 1973. Quantitative assessment and possible biochemical indicators of variation in resistance to fusiform rust in loblolly pine. PhD Thesis, North Carolina State Univ, Raleigh. 94 p. Lundquist, J.E., and E. S. Luttrell. 1982. Early symptomatology of fusiform rust on pine seedlings. Phytopathology 72:54-57. Lundquist, J.E., and T. Miller. 1984. Development of stem lesions on slash pine seedlings infected by Cronartium guercuum f. sp. fusiforme. Phytopathology 74:514-518.

PAGE 87

75 Lundquist, J.E., T. Miller and H. R. Powers, Jr. 1982. A rapid technique for determining resistance of slash pine to fusiform rust. Phytopathology 72:613-615. Matthews, F. R., and S. J. Rowan. 1972. An improved method for large-scale inoculations of pine and oak with Cronartium fusiforme. Plant Dis Rep 56:931-934. Miller, T., E. B. Cowling, H. R. Powers, Jr. and T. E. Blalock. 1976. Types of resistance and compatibility in slash pine seedlings infected by Cronartium fusiforme. Phytopathology 66:1229-1235. Mullick, D. B. 1977. The nonspecific nature of defense in bark and wood during wounding, insect, and pathogen attack. In The Structure, Biosynthesis, and Degradation of Wood. Vol 11 Recent Advances in Phytochemistry (F. A. Loewus and C. C. Runeckles, eds) p 395-441. Plenum Press, New York. 521 p. Powers, H. R., Jr. 1980. Pathogenic variation among single aeciospore isolates of Cronartium guercuum f. sp. fusiforme. Forest Sci 26:280-282. Powers, H. R., Jr., and L. D. Dwinell. 1978. Virulence of Cronartium stable after 25 years. Plant Dis Rep 62:877-879. Powers, H. R., Jr., F. R. Matthews and L. D. Dwinell. 1977. Evaluation of pathogenic variability of Cronartium fusiforme on loblolly pine in the southern USA. Phytopathology 67:1403-1407. Powers, H. R., Jr., F. R. Matthews and L. D. Dwinell. 1978. The potential for increased virulence of Cronartium fusiforme on resistant loblolly pine. Phytopathology 68:808-810. Powers, H. R., Jr., R. A. Schmidt and G. A. Snow. 1981. Current status and management of fusiform rust on southern pines. Ann Rev Phytopathol 19:353-371. Powers, H. R., Jr., and B. J. Zobel. 1978. Progeny of specific loblolly clones vary in fusiform rust resistance according to seed orchard of origin. Forest Sci 24:227-230. Rockwood, D. L., and R. E. Goddard. 1973. Predicted gains for fusiform rust resistance in slash pine. In Proceedings 12th Southern Forest Tree Improvement Conference, p. 31-37. National Technical Information Service, Springfield, VA. 352 p. SAS Institute Inc. 1982a. SAS User's Guide: Basics, 1982 edition. SAS Institute, Inc, Cary, North Carolina. 923 p.

PAGE 88

76 SAS Institute Inc. 1982b. SAS User 1 s Guide: Statistics, 1982 edition. SAS Institute, Inc, Cary, North Carolina. 584 p. Schaffer, H. E. and R. A. Usanis. 1969. General least squares analysis of diallel experiments: A computer program DIALL. Genetics Dept Res Rep Nol, North Carolina State Univ, Raleigh. 61 p. Schmidt, R. A. 1972. A literature review of inoculation techniques used in studies of fusiform rust. In Biology of Rust Resistance in Forest Trees (R. T. BTngham, R. J. Hoff and G. I. McDonald, eds). USDA Forest Serv Misc Publ 1221:341-356. Schmidt, R. A., R. E. Goddard and C. A. Hollis. 1974. Incidence and distribution of fusiform rust in slash pine plantations in Florida and Georgia. Instit Food and Agric Sci, Univ Florida Bull 763. Gainesville. 21 p. Schmidt, R. A., R. C. Holley and M. C. Klapproth. 1985. Results from operational plantings of fusiform-rust-resistant slash and loblolly pines in high-rust-incidence areas in Florida and Georgia. In Proceedings of the Rusts of Hard Pines Working Party Conference (J. Barrows-Broaddus and H. R. Powers, Jr., eds) p 33-41. The Georgia Center for Continuing Education, Univ Georgia, Athens. 331 p. Schmidt, R. A., H. R. Powers, Jr. and G. A. Snow. 1981. Application of genetic disease resistance for the control of fusiform rust in intensively managed southern pine. Phytopathology 79:993-997. Snow, G. A., R. J. Dinus, and A.G. Kais. 1975. Variation in pathogenicity of diverse sources of Cronartium fusiforme on selected slash pine families. Phytopathology 65:170-175. Snow, G. A., R. J. Dinus, and C.H. Walkinshaw. 1976. Increase in virulence of Cronartium fusiforme on a resistant slash pine. Phytopathology 66:511-513. Snow, G. A., and M. M. Griggs. 1980. Relative virulence of Cronartium guercuum f. sp. fusiforme from seven resistant families of slash pine. Phytopath Medit 19:13-16. Snow, G. A., and A.G. Kais. 1970. Pathogenic variability in isolates of Cronartium fusiforme from five southern states. Phytopathology 60:1730-1731. Snow, G. A., W. L. Nance and E. B. Snyder. 1982. Relative virulence of Cronartium guercuum f. sp. fusiforme on loblolly pine from Livingston parish. In Resistance to Diseases and Pests in Forest Trees (H. M. Heybroek, B. R. Stephan and K. von Weissenberg, eds), p 243-250. Centre Agric Publ Document, Wageningen, The Netherlands. 503 p.

PAGE 89

77 Sohn, S. 1977. Heritability of resistance to Cronartium fusiforme in Pinus elliottii as affected by disease incidence and a comparison of breeding procedures. PhD Dissertation, Univ of Florida, Gainesville. 92 p. Sohn, S., and R. E. Goddard. 1979. Influence of infection percent on improvement of fusiform rust resistance in slash pine. Silvae Genetica 28:173-180. Squillace, A. E., R. J. Dinus, C. A. Hollis and R. A. Schmidt. 1978. Relation of oak abundance, seed source, and temperature to geographic patterns of fusiform rust incidence. USDA Forest Serv Res Pap SE-186. 20 p. Steel, R. G.D. and J. H. Torrie. 1960. Principles and Procedures of Statistics. McGraw-Hill Book Company, Inc, New York. 481 p. Walkinshaw, C. H. 1978. Cell necrosis and fungus content in fusiform rust-infected loblolly, longleaf, and slash pine seedlings. Phytopathology 68:1705-1710. ~alkinshaw, C.H., and R. L. Anderson. 1983. Fusiform rust: Illustration of different symptoms in the greenhouse and field. USDA Forest Serv Forest Pest Manag Rep 83-1-21. 13 p. Walkinshaw, C.H., T. R. Dell and S. D. Hubbard. 1980. Predicting field performance of slash pine families from inoculated greenhouse seedlings. USDA Forest Serv Res Pap S0-160. 6 p. Wells, 0. 0., and P. C. Wakeley. 1966. Geographic variation in survival, growth, and fusiform rust infection of planted loololly pine. Forest Sci Monogr 11. 40 p. Wright, J. W. 1976. Press, New York. Introduction to Forest Genetics. 463 p. Academic

PAGE 90

APPENOIX

PAGE 91

Table 26. Least squares estimates of the mean number of fusiform rust galls per tree by gall type or location for 48 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia. Ga 11 TyQe Gall Location Witches 1 Family Typical Thin BTF.~/ Sunken Broom Fat Stem Limb Stem-Limb Total --------------------------------mean number of galls per tree-------------------------------10050 0.945 0.037 -0.036 -0.008 0.295 0.089 0.960 0.855 0.368 2. 183 10159 2.535 0.034 0.016 0.087 0.900 0.018 3.623 2.739 0.522 6.8 8 4 190047 l 221 0.096 0. 177 0.067 0.401 -0.005 3.759 l. 353 0.261 5.373 190060 l 789 0.105 0.097 0.069 0.411 -0.003 3. 157 l. 738 0.297 5. 192 1 9 0141 l. 769 0.046 0.021 0.045 1.085 0.026 6.624 1.909 0.514 9.047 190287 1 156 0.024 0.079 0.059 0.675 0.06 2 2.359 1. 349 0.376 4.0 8 4 600047 0.817 0.215 0.383 0.074 0.300 0.019 0.703 l 20 l 0. 197 2. l 01 600 2 87 0.780 0.097 0.003 0.034 0.261 0.037 1. 908 0.765 0.246 2. 920 '--l \Cl 960065 1. 149 0.025 0. 109 0.029 0.402 0.041 2.508 1 271 0.215 3.994 960146 2. 141 0. 154 0.119 0.027 0.575 0.034 3.083 2. 186 0.34 8 5.617 980048 0.457 0.037 0.076 0.020 0.041 0.025 1. 100 0.598 0.041 -0.461 9800 8 8 1. 106 0.039 0. 115 0.001 o. 191 0.065 -0.357 1. 270 0. 179 1 .09 2 1410047 1 441 0.045 0.022 0.052 0.469 0.0 8 0 2.954 1 .496 0.270 4. 720 1410060 1.534 0.031 0. 172 0.079 0.503 0.049 1 733 1.624 0.283 3.640 1410287 0.603 0.043 0.036 0.014 0.346 0.008 2.994 0.523 0.267 3.7 8 5 1460065 1.223 0.063 0.061 0 011 0.5 3 4 0 041 3.325 1.444 0.273 5.042 1460071 1 .492 0.012 0.054 0.018 0 505 0.037 2.554 1.418 0.436 4.409 1460096 0.975 0.058 0.041 -0.019 0.423 0.032 3.5 8 9 1. 3 06 -0.003 4. 8 92 1 5 70065 1. 523 0.074 0.099 0.063 0.503 0.063 2.564 1. 632 0.281 4.477 1570071 1.420 0.024 0.154 0.063 0.421 0 063 2.575 1.3 8 7 0.397 4.359 1570096 1 781 0.055 0.139 0.033 0.377 0.090 2.537 l. 873 0.306 4.714 1570146 1.382 0.029 0. 114 0.044 0.461 0.066 3.443 1.396 0.281 5.210 2 700050 o. 776 0.034 0.053 0.044 0.364 0.00 8 3.008 0.816 0.206 4.030 2700064 1. 706 0.069 0. 196 0. 180 0.534 0.064 3. 148 1.806 0.52 8 5.3 8 2 2700295 0.913 0.071 0.035 0.136 0.550 0.022 3.084 0.890 0.473 4.447

PAGE 92

Table 26--continued. Gall Ttee Ga 11 Location Witches' F amilt Tteical Thin BTF~/ Sunken Broom Fat Stem Li rnb Stem-Limb Total --------------------------------n~an number of galls per tree--------------------------------2700355 0.935 0.084 0.022 0. 121 0.509 0.022 3.094 0.996 0.435 4.524 2840050 1 .606 0.029 0.075 0.068 0.497 0.073 3. 119 1. 596 0.286 5.001 2840064 1.263 0.015 o. 131 0.068 0.441 0.036 3. 106 1 15 2 0.417 4.675 2840295 1.027 0.007 0.054 0.098 0.265 0.028 3.064 0.767 0.377 4.207 2840355 1.013 0.040 0.049 o. 120 0.408 0.008 3. 114 0.822 0.437 4.372 288000 1 1. 523 0.075 0.075 0.077 0.830 0.048 5. 041 1 915 0.330 7.287 2880050 1 091 0.024 0.067 0.090 0.614 0.043 3.492 1. 367 0.324 5. 1 83 2880159 1. 960 0. 123 0.035 0.090 0.648 0.043 3. 162 2.008 0.418 5.587 l'.'.880352 1.310 0.046 o. 105 0.040 0.693 0.029 4.034 1. 583 0.288 5.905 w 0 2950064 2. 115 0.038 0. 160 0.084 1 129 0.205 3.066 2 .505 0. 733 6. 303 2950270 0.655 O. 161 0.077 0.056 0.413 0.050 3.054 0.696 0.354 4. 104 2950355 1 .629 0.046 0.052 o. 118 0.632 0.067 3.521 1. 355 o. 770 5.646 3500048 1.820 0.048 0.091 0.015 0.382 0. 133 1. 808 1. 763 0.476 4.047 3500088 2.705 0.015 0.256 0.019 0.434 0. 187 3.565 3.026 0.292 6.883 3500098 2.085 0.079 o. 132 0.003 0.314 0.076 1. 901 2.233 0.319 4.454 3500354 2.674 0.010 o. 156 0.011 0.902 0. 132 5.212 2.483 0. 756 8.451 3520001 0.817 0.015 0. 132 0.058 0.462 0.010 3.298 1.044 0. 103 4.444 3520050 1 .601 0.041 0.081 0.013 0.395 0.085 0.885 1. 706 0.307 2 .898 3520159 0. 964 0.011 0.040 0.043 0.595 0.062 1. 567 1 198 0.249 3.014 3540048 1. 972 U.021 0.262 0.009 0.395 0.079 2.647 1.928 0.506 5.081 3540088 0.901 0.006 0.090 0.010 U.450 0.063 3.860 1.032 0. 177 5.069 3540098 1. 911 0.041 0. 184 0.056 0.266 0. 1 88 1 155 1 .807 0.571 3.534 3540350 1.798 0.046 o. 157 0.021 0.867 0. 110 9. 145 1.9 82 0.620 11. 747 ~/Basally truncated fusiform gall.

PAGE 93

Table 27. Least squares estimates of the mean proportion of trees infected with fusiform rust galls by gall type or location for 48 slash pine families (fourto six-years-old) planted in six progeny tests in southeast Georgia. Family 10050 l 0159 190047 190060 190141 190287 600047 600287 960065 960146 980048 980088 1410047 1410060 1410287 1460065 1460071 1460096 1570065 1570071 1570096 1570146 2700050 2700064 2700295 2700355 Ga 11 T e Gall Location Witches' Typical Thin BTF~I Sunken Broom Fat Stem Limb Stem-Limb -------------------------------------------------------------------------------------------30.57 69.67 55.46 74.94 84.71 55.87 34. 16 36. 11 50.44 52.53 19.36 34. 77 55.91 59.24 35.93 42.01 47.00 37.30 59.23 57. 96 67.20 50.74 40.69 77. 12 47. 19 49.16 l. 13 2.94 5.57 7.32 3.93 2.29 1.52 7.90 0.42 1.33 1.92 2.72 2.45 1.50 0.88 4. 16 0.22 5.21 1.52 0.91 4.76 2.47 1.03 4.80 5.49 7.33 0.06 0.71 8.37 7. 72 2.73 4.74 5.24 0.06 4.86 9.28 2.26 4.99 3.76 7.69 1.38 2.29 1.32 0.50 5.70 10.27 7.35 9. 19 2.55 14.63 1.36 0.25 0.02 7.33 6.54 5.09 3.41 5.60 7.57 2.45 l. 71 l. 51 1.40 0.32 5.00 7.78 0.34 0.51 l. 12 0.01 2.47 2.89 2.53 2.20 3.92 16.63 6.99 10. 72 30.84 49.37 31. 27 28.95 66.84 37.78 19.02 14.91 29.06 33.43 3.75 16.97 27.41 27.53 21 .69 35. 19 30.98 33.59 37.20 31. 96 24.08 32.34 28.22 36.07 42 .81 43. 74 3.044 1.494 0.297 0.635 2.502 4.885 2.247 2.251 l .687 2.213 0.710 3.591 7. 154 2.598 0.259 2. 121 1. 776 0.233 3.862 3.862 8. 186 4.920 0.259 4. 145 l. 378 1.353 21. 13 29.63 32. 96 41.34 52.78 32.23 19.93 17.07 18.89 28. 96 1.30 5.45 29.95 44.52 21 81 21.32 26.75 23.29 26.43 24.08 23.27 25.78 26.32 40.47 33. 19 38.58 33.52 65.59 53.67 69.84 90.08 52.74 29.97 36. 77 49.26 48.52 20.01 33.54 57.21 55. 12 28.35 44.81 36.84 40.81 59.02 52.01 68.61 48.27 37 .07 71. 84 49.26 54.54 9.92 35.68 23.90 30. 14 55.08 25.31 14. 79 23.03 20. l~ 26.27 2.99 11 84 27.32 24.87 26.61 27.08 28. 15 18.01 27 .80 33.22 28.2 8 24.38 19. 7 4 48.96 39.28 42.08 O:.

PAGE 94

Table 27--continued. Ga 11 T e Gall Location Witches 1 Family Typical Thin BTF~I_ Sunken Broom Fat Stem Limb Stem-Limb ----------------------------------------------3/.--------------------------------------------2840050 2840064 2840295 2840355 2880001 2880050 2880159 2880352 2950064 2950270 2950355 3500048 3500088 3500098 3500354 3520001 3520050 3520159 3540048 3540088 3540098 3540350 48.24 61.40 56.28 53.92 51.30 44.37 69.48 50.97 79.64 42.27 84.49 64. 12 65.25 61 .47 86.42 34.75 34. 78 43.89 61. 76 52.09 56.43 89.28 l.47 0.88 0.55 2.42 6.27 2.26 8.32 3.89 3.37 8.55 l.58 3.25 0.82 5.67 0. 14 l. 23 l.64 l.06 l. 10 0.27 2.20 3. 17 ~/Basally truncated fusiform gall. 6. l 0 10.64 3.89 2.28 5.80 5.76 3.31 5.21 13.21 4. 11 3.64 4. 15 7. 15 4.22 6.92 6.97 5. 19 4.55 11 .01 4.29 l 0. 14 9.88 6.51 4.35 9.65 9.40 5.27 4.09 8.85 3.94 8.34 3.53 11 63 0.72 l.33 0.38 0.86 3.91 0.81 3.53 0.57 0.59 2.60 l.49 32.35 29.26 17.68 29.05 49.43 38.42 46.30 36.53 60. 77 34.29 46.36 26.97 29.46 16.64 67.49 29.66 18. 96 33.52 29.72 36.25 18.88 75. 18 2.650 3.405 l. 780 0.259 3.554 2. 181 l 629 2.797 16.843 3.259 4.352 5. 125 3. 745 5.258 10.228 l .059 3.054 3. 723 3. 139 3.792 7.534 2.664 36. 13 37.58 34.63 35.81 30. l 0 28.37 49.09 30.01 29.48 31.64 38.78 20.75 26.94 20. 10 54.76 22.25 16.35 21 5 7 27. 75 26.29 23.99 61. 81 41.36 53. 19 50.98 45. 91 54.95 43.07 62.88 54.54 77. 21 45.59 73.56 59.53 64.83 56. 11 85.76 32.59 29 .63 40.34 62.87 56.67 45.25 90.42 28.22 37.93 31. 9 7 36. 16 32.59 26.40 29.98 26.64 62 .02 33. 17 62. 7 5 31 .47 26.83 22.51 49. 16 19. 16 15. 37 16.02 31. 16 24.83 28. 70 65.48 (X) N

PAGE 95

83 Table 28. Variance components and their standard errors for replication, general combining ability (GCA), specific combining ability (SCA) and error effects for the mean number of galls per tree for various fusiform rust gall types or locations as computed by the FORTRAN program, DIALL, for each crossing set and test. X-in9 Test ___ Ga _l_l ____ R"""e.__p l _i"""c-'--a t_i"""o_n 34 Total S tem Lim b Stem -Limb Typical Witches' Broom B TFi/ Fat Thin S unken 35 Total Stem Lim b Stem -Limb Typica 1 Witches' B room BTFil Fat Thin Sunken 36 Total S tem Lim b Stem Limb Typical witches' Broom BTFil Fat Thin Su n ke n 3 7 Total Stem L imb Stem -Limb T ypica l Witches' B room B TFi / Fat Thin S unken 0 .53 + 0. 8 1 -0. 00 1-+ 0 0 01 0 2 9 +-0.55 0.03 + 0.03 0 09 + 0.38 o o oZ-+ 0 006 0 009 + 0.01 0 .01 +-0.01 0.00 1-+ 0.001 0 .0002-+ 0.00006 0 11 + 0. 16 0.0003 + 0 .001 0. 1 + 0-:-13 0 003 + 0.004 0. 11 +-0.13 0 .0 0 7 + 0 009 o o or+ 0.002 0 00 1 + 0.0003 o o + 0.00002 0.0005 + 0 0003 0 0 1 + 0.06 0 0003 + 0.002 0 .001 +-0. 03 0.0002-+ 0 .0 13 0.02 + 0.01 -0.01 + 0 .004 0.001-+ 0.0008* -o.o oooT + 0 007 0 00 03 +-0.0005 0 .0001 + 0 .0 002 0 004 + 0.07 0 008 + 0 .015 0 l.) 4 +-0.01 0 .006-+ 0 0 1 0.045 + 0.02 0 00 7 + 0 .01 6 0 0 03 + 0 004 0 0 001 + 0 002 0 00 01 + 0 000 4 0 00 1 +0 001 Variance C omponent GCA SCA 1.07 + 1. 3 1 0 04 + 0 02 ** U 99 + 1 05 0.01 + 0 04 -0. 94 + 0. 8 5 0.2 +-0.1 2 ** 0 01 -+ 0 01 0 0 1 + 0 0 1 0 .0002 + 0 0004 0 000 1 + 0.0004 2 .1 1 + 1.25* 0 02 + 0 01* 1. 26 + 0.75** 0.04 + 0.03* 1. 0 2 + 0.63** 0 09 + 0 06 0 01 + 0 007* -0.0002 + 0 .0004 0.0000 1-+ 0 00003 0 00 1 + 0 00 1 0.45 + 0.37 0.01 + 0 0 1 0.17 + 0 13 0.0 2 + 0 02 0 11 + 0 0 9 0 02 + 0.02 0. 0 009 + 0 0006* 0 .004 +0 .003 -0. 0 0002 + 0 00 02 0 0 001 +0 0002 0. 8 5 + 0 56* 0 03 + 0 03 0 1 8 + 0 1 2 0 11 + 0 08* 0 35 + 0 23 0 09 + 0 07 o. o ouT + 0 005 U U003 + 0 00 2 0 0008 + 0 0006 0 00 1 +0.002 5 28 + 3.92* 0 00 5-+ 0 003 4.46 +3 22 o 11 + 0.08* 3.59 + 2 .69* 0.004-+ 0 .014 0 0 5 + -0. 04 0 05 + 0 04* o o ooT + 0 .001 0 001 +-0.001 0 .17+0.16 0 00 3-+ 0.003 0 .16 +-0.ll 0 0 1 + 0 .01 0 03 + 0 .14 0 05 + 0 04* 0 002 -+ 0.004 0 00 1 + 0 002 o ooooT + 0 0001 0 003 + 0.002* 0. 29 + 0 27 o oo r+ 0 .004 0 0 9 +-0.09 0 00 6-+ 0 .03 0 0 5 +0 .06 o .oor+ 0 02 0 .0006-+ 0 0004 0 02 + 0 007 0 0 003 + 0 000 6 -0.0001 + 0 0005 0.16+0.2 o ooc + 0 02 0 003 + 0 0 6 0 0 1 2 + 0.03 0 0 14 + 0.08 0.04 +0 06 0 012 -+ 0 01 0 0009 -+ 0 00 3 0.00 06 + 0 000 6 0 003 +-0 002 Error 5 8 7 + 1. 8 7 0 02 + 0 007 4.5 8 + 1 .4 5 o. 12 + 0.04 4.16 + 1.3 2 0.06 + 0 02 0. 07 + 0.02 0 .06 + 0.0 2 0.006 -+ 0.002 0 .U0 2 + 0 .000 6 1.1 + 0 .35 0 01 -+ 0 .00 3 0.83 + 0.26 0 .03 + 0.01 0.72 + 0.23 0 06 + 0 02 0.02 + 0.007 0 009 -+ 0.003 0.0003-+ 0.0001 0 .0 03 +-0.001 0 8 6 + 0.21 0.04 + 0.009 0.34 + 0.08 0.18 + 0.04 0.27 + 0.07 0 .14 + 0 .0 3 0.004-+ 0.001 0 l + 0.02 0.005 + 0.001 0 .004 + 0.001 l .0 +0. 23 0.17 -+ 0 .04 0 .51 + 0 12 0 2 1 + 0.5 0. 64 + 0. 14 U.3 1 + 0 07 0 0 4 + 0.009 o o3r + 0 0 0 1 0 008 -+ 0 .002 0.03 +0 .007

PAGE 96

Table 28--continued. X i n g S _e_t __ Te_s_t ___ Ga ;_l_ l ____ R -'-" ep c.... l _i-"ca ;_t'--i..;;.. on __ 2 36 Total S tem L irnb Stem -Li m b Typical W itche s B room BTFa/ FatThin Sunk en 37 Total Stem 0 003 + 0 02 0. 002 + 0 00 1 0 004 + 0 .01 0 0002 -+ 0 00 1 0.012 +-0.006 0 006 + 0 .0 07 0 0 + 0 002 0 0 + 0 0001 0 0003 + 0.0002 0.0004 + 0 000 1 0 02 + 0.05 0 0 02-+ 0 005 0.008 + 0 02 0 0 1 +-0.005 L irno Stern Lim b Typical Wit ches' BTFa/ Fat 0 008 -+ 0 03 8 roorn 0 002 + 0 01 0 00 01-+ 0.001 0.00004 -+ 0.0008 T hin Su n ken 40 Tota 1 S t em L irn b Stem Lim b Ty pical W itches' Broo m B TFa/ FatThin S un ken 41 Total Stem L irnb S te m L i m b Typical W itch es ll r oom B TFa / Fat Thin Su nken 0 00005 + 0 0002 0 00 05 +-0.001 0 .03 + 0 04 0 00 1-+ 0 00 1 0 0 3 +0 04 0.00 02 + 0 00 1 0 0 7 + 0 0 6* 0 005 -+ 0.005 0.0004 -+ 0.001 0. 0 + 0-:-00005 o .ooT + 0 00 4 0 .0 0005" + 0 000 1 0 09 + 0 05 0 00 1+ 0 0007 0 06 +0 05 0 0 +-0. 004 0 07 -+ 0 0 4 0 3 +0 3 o oo T + 0 00 1 0.001 + 0 00 1 0 002 + 0 .005 0 0002 -+ 0 0002 84 Var ian c e Co mponent GCA SCA 0.008 + 0.03 0 001 + 0.005 0 0 0 5 + 0 0 1 0 0005 -+ 0 0007 0 .003 +-0. 0 1 -0.003 + 0 .01 0 0006 -+ 0 000 7 0 000 1 + 0 00 0 1 0.0002 + 0 0002 0.0006 + 0 00 03 0 07 + 0.06 0 007 -+ 0 00 7 0 .04 +-0. 03 0 002 -+ 0 007 0 04 + 0 0 4 0 002 -+ 0 00 1 0 002 + 0. 002 0 000 6-+ 0.0004 0 .0002 + 0 0002 0.004 +0 00 3 0.04 + 0 09 0.0004 + 0.00 1 0 .05 + 0. 08 o oooJ + o 002 0 002 +-0. 08 0.00 1 + 0 002 0 .000 8 -+ 0 0 01 0 0 001 + 0 0002 0 00005 -+ 0 0 05 0.000 1 +0 0004 0 05 + 0. 1 5 0 0004 + 0 003 U.02 + 0.06 0.03 + 0 03 0 03 + 0 08 U 005 -+ 0 0 1 0.00 1 + 0 00 1 0 000 1-+ 0 0004 0 001 +0 004 0 000 1-+ 0 000 1 0 0 1 + 0 05 0 00 6-+ 0 . 8 0 006 + 0 02 0 002 + 0.00 1 0 00 6 + 0 019 0 0 1 7 + 0 02 o ooo r+ 0 .0001 0 0003 + 0 0002 0 0 005 + 0 0003 0 0007 + 0 0002 0 0 45 + 0 05 o o oor+ 0 0 06 0.04 + 0. 02 0 004 -+ 0 02 0 02 + 0 03 0 03 + 0.008 0 0 002 + 0.002 0 0 0 2 +-0. 0 0 0 6 -0. 0004 -+ 0 0002 -0. 00 03 + 0 0 02 0 l + 0 l 0 002 + 0 .002 0 075 + 0. l 0 0 0 2 + 0.0 04 0. 16 + 0 13* 0.02 + 0 .007 0 000 5 + 0 002 0 0 00 l + 0 0002 0 0 0 05 + 0 0 1 0.0004 + 0 0006 0 09 + 0 2 0 003 -+ 0 005 0 04 +0. l 0 02 + 0 02 0 002 -+ 0 l 0 0 l +0 02 0. 003-+ 0 .004 O U002 -+ O OUl 0 00 1 +0 007 0 000 1-+ 0 000 1 Error 0 .1 8 + 0.06 0 016 -+ 0 005 0 .12 +0 03 0 .01 + 0 003 0.1 +0.0 3 o or + 0 0 1 o o or+ 0.0001 0 002 + 0 0005 0 00 3 + 0 0008 0 003 + U 0008 0 .47+ 0.I 0 04 + 0 01 0 3 +0 08 0 .1 2 -+ 0.03 0 25 + 0 07 0 .13 + 0 04 0 0 1 + 0 00 3 0 009 -+ 0.003 0. 003 + 0 0007 0 01 +-0.003 0 3 + 0 l o oo i + 0.002 0 32 +0 09 0.02 + 0.006 0.22 + 0 06 0.09 + 0.03 0 0 1 + 0 003 0 0 01-+ 0.0003 0 0 6 +0 0 2 0 002 -+ U .001 0 99 + U 3 0 02 + 0 00 5 0.9 +0 3 0 04-+ 0 01 0 8 +0 2 0 .1 6-+ 0 05 0.0 1 + 0 003 0 00 4-+ 0.00 1 0 0 4 +0 0 1 0 001 -+ 0 0002

PAGE 97

35 Table 28 --continued. X -ing Variance Component Test Gall Replication GCA SCA Error 4 34 Total 0 44 + 0 9 2 2 8 + 1.98 0 .19 + 1.6 6 2 + 2.19 S tem 0.00 1 + 0.001 0.022-+ 0 0 1 6 0.006 + 0 007 0 0 14 + 0 005 Limb 0.29 +-0.67 1.2 + T.2 0 08 +-1 2 4.9+1.7 S tem-Limb 0.005-+ 0.02 0. 8 + 0.6 0 02 + 0 02 0 .1 6-+ 0 06 Typical 0.27 + 0 63 0 89 -+ 1.3 0.0 8 + 1. 6 4.6 +-1 .6 Witches' Broom 0.02 + 0.005 0.08 + 0.05 0.01 + 0.02 0 .1 2 -+ 0.04 BTFa/ -0.002-+ 0 002 0.0001 + 0 005 0 004-+ 0 008 0.02 + 0 009 Fat 0.0004 + 0.003 0.001 +-0.002 -0.005 + 0.006 0 03 + 0 01 Thin -0.0003 + 0.0003 0.0004-+ 0.004 0 .008 + 0 006** 0 005 + 0 001 Su n ken 0.0005 + 0.002 0.005 +-0.004 0 002 + 0 004 0.02 + 0.007 35 Total 0.06 + 0.06 0 07 + 0 3 0 57 + 0.56 1. 2 + 0 4 Stem 0.002-+ 0 .0 03 0.0 1 + 0 01 0.01 + 0.01 0.02-+ 0.005 Limb 0 03 +-0.08 0 07 + 0 15 0.02 + 0.2 1.2 +-0.4 Stem-Limb 0.006-+ 0 003 -0.01 + 0.01 0 04 + 0.04 0.08-+ 0 03 Typical 0.02 +-0.07 0 006 -+ 0 l 0 2 +-0.2 0.6 +-0.2 Wit ches' Broom 0 02 + 0 0 1 0 0 1 +-0.03 0.02 -+ 0 06 0.2 + 0.07 BTFa/ 0.001 + 0 002 0.0003 + 0.0009 0.001 + 0.002 0 01 + 0.004 Fat -0.00005 + 0.00 1 0.0005 + 0 0004 -0.001 + 0 .0 0 1 0.009 + 0.003 Thi n 0.0001 +-0.0002 0 00003 -+ 0.0003 0.0004 -+ 0.0007 0 002 + 0 001 Sunke n -0.0003 + 0.0002 0.0005 +-0.0004 -0.0008 + 0 0006 0 004 + 0 00 1 40 Total 0 07 + 0. l 0. l + 0. l -0. l + 0 l l. l + 0 3 Stem 0.0003 + 0.001 o.ooT + 0.001 -0.0004 + 0 00 1 0 01-+ 0 002 Limb 0.07 + O. l 0 .1 + o 1 0. l + 0 -:-1 1.0 +-0.3 Stem -Limb 0 0003 + 0.00 1 o.oo T + 0.001 0 00"5 + 0 00 1 0.02 + 0.005 Typica l 0 0 1 + 0.05 0 06 +-0.05 0.08 +-0.06 0.5 +-0.2 Wit ches Broom 0 03 + 0 03 0 0 1 + 0.04 0 05 + 0 05 0.1 + 0 05 BTFa/ 0 00 1 + 0.001 -0.001 + 0 002 o oor+ 0.004 0 01 + 0 003 Fat 0 001 + 0.0009* -0.001 + 0 00 1 0 004 + 0.003** o.oor+ 0.00 1 Thin 0 0005-+ 0.0006 0.0002-+ 0.0006 0 00 1 + 0.00 1 0.009 + 0.003 Sunken -0.0001 + 0.0002 0 0004 + 0.00 1 0.001 + 0 00 1 0.003 + 0.001 41 Total -0.02 + 0.02 0.5 + 0 3 0.02 + 0.08 0 2 + 0 09 Stem 0.0003 + 0.0007 -0.01 + 0. 0 2 0 03 + 0.02* o ooi + 0 003 Limb -0.002 +-0.02 0 .4 +-0.3* 0 .0 4 + 0 02 0 1 3 +-0.05 Stem-Limb 0 004 + 0 005 o.ooT+ 0.009 -0.002-+ 0.01 0 05 + 0 02 Typical 0.02 +-0.03 0.3 +-0.2 0.04 +-0.02 0 15 + 0.06 Witches' Broom 0.005-+ 0 003 0.04-+ 0.02* -0.01 + 0 005 0.04 + 0 02 BTFal 0.00 1 + 0.001 -0.006-+ 0.02 0 03 + 0 02 0.005-+ 0.001 Fat 0.0002-+ 0.0006 0 001 + 0 0007 -0.001-+ 0 0008 0.005 + 0.002 Thin 0 009 + 0.008* 0 002 +0.003 0 002 + 0 004 0.02 +-0.007 Su n ken -0.0003-+ 0.0004 0 00005 + 0.002 0 001 + 0.002 0.004 + 0 002

PAGE 98

Table 28--continued. X -in9 S _e_t __ T e_s_t __ ..:..Ga;;;._l;_l ____ R ..c.ep.__l--'i..c. c a.c...t_i..:..on __ 5 34 Total S tem Limb S tem-Limb Typical Witches B ffd/ Fat T hin Su n ke n 35 Tota 1 S tem Limb 0.3 + 0 3 0 006 + 0 .01 0 3 4 +0.3 0 .005-+ 0.01 0 .2 + 0.2 B room 0.0 1 + 0.02 0 0 1 + 0 0 4 0 .003-+ 0 002 0.02 +-0.03 0 000 5 + 0 0005 0 .07 + 0.1 Stem-Limb T y pical Witches' Broom B TFa / 0.0 l + 0.003 0 .0 8 + 0. 1 0. 0003 + 0.006 0 .13+0.ll* 0 02 + 0.04 0 0 01-+ 0.01 0 000 1... Q.0 00 8 0 .0006 + 0.0007 0 .0006 + 0.001 Fat Thin Su nken 40 Total St em Limb S tem-Limb Typ ica 1 W itches' B TFa/ Fat Thin S unken 41 Total S tem Limb 0.02 + 0.04 0 0 0 2 -+ 0 .001 0 02 +0 .04 0 .001-+ 0.001 0 .006 + 0 .0 2 8 room 0 0005 -+ 0.003 -0.0005 + 0.0004 0 000 4 + 0.0004 0.00 07 + 0 00 1 0.0001 + 0.0004 0. l + 0 .9 0 02 -+ 0 00 6 0.05 + 0 .07 S tem-Limb Typical Witches' Broom BTFa/ o o ooT + 0 003 0.06 + 0.06 0 001 -+ 0.01 0.00 1 + 0.0007 Fat Thin S un k en -0.0003-+ 0 0004 0 .00 2 +0 002 0.0006... 0. 000 5 86 Varia nce Co mponent GCA SCA 0 l + 0 5 0 04 -+ 0 06 0 .4 ... 0 .3 0.05-+ 0.03 0 .0 2 + 0.3 0. l +-0.08 0.0 1 + 0.05 0 001 + 0 00 1 0.001 + 0.05 0 002 + 0 00 1 0 4 + 0 .4 0 02 -+ 0 .0 2 0.17+0.2 -0.01 + 0.04 0.3 +-0.3 0 0 1-+ 0.06 -0.00Z-+ 0.02 0 0002 -+ 0.0006 0.0003 + 0 0008 0.001 ... 0 .001 0.23 + 0.2 0 .0 0 1-+ 0.002 0 l + O. l 0.004 + 0.004 0.05 ... 0 08 0 02 + 0 02 0.0003 + 0.0004 -0.0 003 + 0 .00 03 -0. 0003 + 0.0006 0.0003 + 0.0004 0.3 + 0 .9 0.008 + 0 009 0.1 + o. 7 0.01 ... 0 .0 0 5 0.3 ... 0 .4 0.03 -+ 0.08 0 003 -+ 0 002 0 0006 -+ 0.0009 0 002 +0 00 1 0.0001 + 0.0002 0 .4 + 0.8 0 08 -+ 0 .0 8** 1 0 +0 .5 -o.oZ-+ 0 0 1 0 2 +0 .5 -0.01 + 0 0 1 0 02 + 0 07 0 00 6-+ 0 003 0.02 +-0.08 0 002 -+ 0 001 0.5 + 0.4* 0 00 01 + 0 02 0 2 + 0-:-2 0.1 +0.0 8 ** 0.2 + 0.2 0.06-+ 0.1 0.01 + 0.03 0 00 1-+ 0 00 1 0 .000 8 -+ 0.001 0 .002 _:-0.001 0 .17+0.16 0.002 -+ 0.004 0 .1 + 0.1 o.ooT + 0 .0 04 0 .1 2 +0. 1* 0 .03 + 0.02* -0.001 + 0.001 0.0009 -+ 0.0008 -0. 00 1 +0 .0 02 -0.0002-+ 0.0006 1. 0 4 + 0 98* 0 002 -+ 0.009 0 .9 + 0.8* 0 00 6 + 0.003 0.3 + 0 .4 0 12 -+ 0.1 o .oor ... 0 .0 0 1 0.00 06-+ 0 0 01 0.00 1 ... -0. 00 1 0 0 + 0. 0003 Error 3. 7 + l .5 0 08 -+ 0 .03 3 7 ... -1 5 a .or+ o.o3 2 .1 ... 0 9 0 .08-+ 0 .03 0 .3 +-0. l 0 .02-+ 0.009 0.3 +0. l 0 00 5 + 0 .00 2 0 .5 + 0.2 0 .08-+ 0.3 0.4 +0 l 0.06-+ 0 .02 0.2 ... -0.09 0 .3 + 0 1 0. l + 0.04 0 .007 + U.002 0.002 + 0.0009 0.007 + 0. 00 3 0 .4 + 0. l 0 .02-+ 0 .04 0 .3 +-0. l 0.02-+ 0.006 0 .2 ... -0. 0 7 0 .05-+ 0.0 1 0.01 + 0 .003 o.ooZ-+ 0.0005 0.01 ... -0. 00 4 0.005-+ 0.0 01 0 4 + 0 2 0 .03... 0.1 0.3 +0 l 0.02-+ 0 .007 0 .4 ... -0. 2 0 .05-+ 0 02 0 0 07-+ 0 003 0.003 + 0 00 1 0 003 + 0.001 0.00 1 + 0 0004

PAGE 99

8 7 T ab l e 28 --c o n t in u ed. X ing Set Test Ga l l Re p lication --------~---Variance Compone n t GCA SCA Error 9 40 Total 0.09 + 0.07* 0.04 + 0.06 0 07 + 0 09 0.3 + 0 0 7 Stem -0 00 1+ 0 0004 0.0007 + 0 0009 0 .002-+ 0 002 0.02-+ 0.04 Limb 0.05 + 0.05 0.03 + 0.04 0.02 +-0.06 0 3 +-0 07 Stem-Limb 0 005 + 0 004 0.009 + 0 005* 0.004-+ 0 002 0 .02+ 0 006 Typical 0 04 + 0 03 * 0.04 + 0 03 0.01 +-0.03 0.1 + -0.03 ',litches' Broom o oar+ o oos 0 006 + 0 008 0 004 + 0.01 0 05 + 0 0 1 B TF ~ / 0 000 6-+ 0 00 1 0.002 + 0 002 0 001 + 0 002 0 016-+ 0 004 F a t 0 000 1 + 0 0003 0.00002 + 0 0003 0.0 + 0.001 0 005 + 0 001 Thin 0.001 + 0 0004 0 0007 + 0.0008 -0 0003 + 0 002 0 0 1 + 0 004 Sunken 0 0004 + 0 0007 0 0004 + 0 0009 0.0008 + 0 00 1 0 007-:: 0 002 41 Total 0 07 + 0 08 0.7 + 0.4* 0 12+0. 1 6 0.6 + 0 l Stem 0.0002 + 0 003 0 0004 + 0 001 0 007-+ 0 003 0.04-+ 0 01 Limb 0 0 1 + 0 04 0 42 + 0 27* 0 11 +-0 13 0 4 +-0 1 Stem-Limb 0 005 + 0.008 0.04 + 0.02* 0.002-+ 0.01 0 07-+ 0 02 Typical 0.0 1 + 0 02 0 23 + 0. 1 6 0.07 +-0 1 0 4 +-0 l Witches' Broom Q 04 + 0 05 0 08 + 0.07 0 008 + 0.07 0.4 + 0 1 B TF~/ o.ooo T + 0.0009 0.00 1-+ 0 0007 0.003 + 0 00 1 0.0 1+ 0.003 Fat 0.002 + 0 002 0 005 +0 004 0 0007-+ 0 003 0 0 1 + 0 003 Thin 0 003 + 0 002* 0.001 + 0 0006* 0 002 +-0 0006 0.009-+ 0 002 Sunken 0 002 + 0.002 0 0007 + 0 002 0 002 + 0 004 0 014 + 0.004 a/ sasa lly t ru n e a ted fusiform gall *F t est for the effect significant at the 5% level. test for the effect s i gnifica n t at the 1% l evel.

PAGE 100

T ab l e 2 9. S p e ci f ic c ombining ability for fusifo rm rust in c id e nce an d f a t g a lls ba sed o n the proportion o f tr ee s d is e as e d or h a ving fat galls. Value s refl e ct p er f o r ma n c e of 4 8 f amili e s (fourto s ix-year s -old) in six prog en y t e s ts in s o utheast G eorg ia. S p e cif ic Co m b ining A bi l ity -------------------------------------Ma l es-------------------------------------------fe 111 a l e UU0 l 0047 0 U48 0050 0060 00 64 0 0 65 00 71 0088 0096 0098 0141 0 14 6 0 1 5 9 02 7 0 0287 0295 0350 0352 0354 0355 Pe r centage o f Trees In fec t ed W i t h R u s t UU0l -4 5 1 7 3 2 uU l ':l 3 1 3 1. 7 0.5 5 3 1 U 0bU 2 8 2 25 U0':lo 2. 18 6 57 U U9 8 '.J 'J 7 4 2 u Ul41 ::,.88 2 15 7. 63 U 1 4b 6.47 2.0 1 -7.65 u I 'Jl l. U6 2 06 3 24 -2 1 9 unu l. 38 7 l ::, 0 78 5 .4 7 u2d4 6.97 4 04 -1. 1 3 1. 8 9 u 2o8 ::i.u7 -4.92 0. 7 6 97 Ud::, J 1 9 3 93 7.3 2 uJ::,U U.7 1. 83 -1. 2 4 1 94 uJ::.2 l 46 l. l 3 6.5 7 0354 4.82 6 14 2.55 0 5 8 0) OJ

PAGE 101

Table 29--continued. Specific Combining Ability ------------------------------------------------------------Males------------------------------------------------------------F e mal e OUOl 0047 OU4 8 0050 0060 0064 0065 0071 OUBB 0096 009 8 0141 0146 0159 0270 0 2 87 0 2 95 0350 035 2 0354 0355 Proportion of Trees With Fat Galls OUUl -0.31 -0.51 UU19 2 .55 -0. 2 6 0.0 2 .84 UUbO -0.4 0.4 00% -0. 2 1 0.06 UU9!:l -1 .!:>4 0.5 9 0141 2 .9 0.3 -3. 19 Ul4b l. 17 0.93 -1.n (X) Ul!:>7 -0.99 -0.89 2 14 -0. 18 \.0 UL?U 0.06 -7. lb l. 17 l. 8 u2 !:l 4 2 .4 -8.0 -0.8 2 0.66 u 2 8!:l l. 7 -l.!J2 -U.72 -0.4 5 U 2 ':J5 3 84 o. 71 2 .55 UJ!:>U l. 75 -0.25 -0.54 3.28 UJ!:>2 -U.94 -0.8 1.22 U354 -U. 2 6 0.36 1.58 -4.29

PAGE 102

BIOGRAPHICAL SKETCH Patricia Adlene Layton was born July 21, 1954, in Columbia, South Carolina. She graduated from Airport High School, West Columbia, South Carolina, in 1972. In 1976, she received her B.S. degree in forest management from Clemson University, Clemson, South Carolina. She completed the requirements for the Master of Science degree at Texds A & M University in December, 1978, majoring in forest science, with a specialization in genetics. Pat was employed as a forest geneticist with Owens-Illinois, Inc., Western Woodlands for three and one-half years in Jasper, Texas. In 1982, she returned to graduate school to pursue her Ph.D. in forest genetics at the University of Florida: She expects to receive her degree in December 1985. Currently, she is a research associate at Oak Ridge National Laboratory. She is a member of Xi Sigma Pi, Alpha Lambda Delta, Alpha Zeta, Phi Sigma, Gamma Sigma Delta, Sigma Xi and the Society of American Foresters. She was honored in 1984 and 1985 by being selected as a recipient of the University of Florida Presidential Recognition Award. In 1984, she was also chosen as the School of Forest Resources Student of the Year as well as being chosen the Forestry Department Student of the Year. 90

PAGE 103

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Goddard, Chairman Pr essor of Forest Resources and Conservation I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Donald L. Rockwood Associate Professor of Forest Resources and Conservation I certify that I have read this study and that in my op1n1on it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Robert A. Schmidt Professor of Forest Resources and Conservation

PAGE 104

I certify that I have read this study and that in my op1n1on it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Anthony E. Squillace Professor of Forest Resources and Conservation I certify that I have read this study and that in my op1n1on it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. cifbt:{! Wdc o/Professor of Dairy Science This dissertation was submitted to the Graduate Faculty of the School of Forest Resources and Conservation in the College of Agriculture and to the Graduate School, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. / December, 1985 Dean, Graduate School

PAGE 105

U N IVER S IT Y O F F LO RIDA II I II IIIIII Ill Ill lllll lllll II IIIIII IIII II IIIIII II 111111111111111 3 1262 08553 4344