THESlS LlBRABY Micéuiggam Qtate University This is to certify that the dissertation entitled A COMPREHENSIVE TREE IMPROVEMENT PLAN FOR TAIWAN presented by Nathan Yung-nai Yao has been accepted towards fulfillment of the requirements for Doctor of PhilosoPhy degree in Forestry Major professor ,/‘ MS U i: an Affirmatiw Action/Equal Opportunity Institution On 12771 MSU LIBRARIES -_.._- RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. A COMPREHENSIVE TREE IMPROVEMENT PLAN FOR TAIWAN BY Nathan Yung-nai Yao A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1982 ABSTRACT A COMPREHENS IVE TREE IMPROVEMENT PLAN FOR TAIWAN BY Nathan Yung-nai Yao Tree improvement work in Taiwan was started in the 19605. In the last 20 years, the results have emerged and shown positive aspects. Few were planned on a long-term basis. Some work plans have been tried. These were not comprehensive and never were put into action. So a compre- hensive work plan is critically needed. My work plan includes several chapters devoted to background information. Then is a review of research activ- ities and the capacities of the research institutions and a review of past tree improvement projects. Then come indiv- idual work plans for nine groups of species: seven conifers and two hardwoods. For each species or species group, much of the past work involves my own work. Provenance and progeny experiments are the main features of the work plan. They differ among species because of genetic natures, silvicul- tural factors and available experiments. Among them are five provenance tests, eight progeny tests, two clonal seed orchards and five seedling seed orchards converted from NATHAN YUNG-NAI YAO progeny tests. Others proposed are one hybridization trial and three arboreta. In one work plan, an attempt at workload and cost analyses is made to show that all the work plans proposed are workable and reasonable. Tree improvement work is tightly correlated with the silvicultural system in Taiwan. A chapter focusing on the gain through silviculture is provided. ACKNOWLEDGMENTS I would like to express my sincere thanks to my major professor, Dr. Jonathan W. Wright, for his enthusiasm, guidance, assistance, and encouragement throughout my doctoral program. Gratitude is also extended to the mem— bers of my guidance committee: Drs. James W. Hanover, Donald I. Dickmann, John H. Beaman and Wayne M. Adams. I am greatly indebted to my wife, Honlin Chen, for her full moral support, encouragement and love during these long years of my absence from home in Chushan, Taiwan, while she has taken care as a housewife, school teacher, and mother of three children. Financial support for my doctoral program was provided by the National Science Council, Republic of China for this six-month trip in 1982; and the Joint Commission on Rural ReconStruction, JCRR (now CAPD, the Council of Agricultural Planning and Development, ROC), AID PRogram in 1976-1978. Special mention should also be made of my employer, National Taiwan University (NTU) and the NTU Experimental Forest for allowing me time to com- plete the work. Without their permission and support, even the possibility of this research would not have been available. ii CHAPTER 5 TABLE OF CONTENTS List of Tables . . . . . . . List of Figures . . . . - . Introduction . . . . . . . Brief History of Taiwan Forestry and Present Status . . . . . . . Administration, Education, Research and Training in Forestry . . . . . Forestry Administrations . . . Universities and Colleges . . . Others . . . . . . . . Advanced Training . . . . . Review of Forest Genetics and Tree Improvement Work in Taiwan . . . . Early Period . . . . . . American Visitors and Their Contributions . . . . . . UNDP/FAO Project, 1966-1969 . . U.S.-China Seminar on Forestry, 1972, Taipei . . . . . . . . TFB Activities . . . . . . Forest Genetics Symposium, 1981, Taichung . . . . . . . NTU Experimental Forest Activities and My Background and Experience . Cryptomeria and Its Improvement . . Introduction . . . . . . Species Description, Silviculture and Growth . . . . . . . . iii vii 10 10 11 l4 14 15 20 22 23 23 25 34 34 35 CHAPTER Past Work on Genetics . . . . . 38 1941 Trial at Chitou 1967, 1968 Trials 1972 Trials Other Genetic Studies in Taiwan Work Plan . . . . . . . . 46 Follow-up of 1972 Provenance Trials Follow-up of 1967, 1968 Provenance Trials Management of AliShan Seed Production Area Improvement Seed for Immediate Use 200 Parent Half-Sib Progeny Test, Start 1990 Clonal Seed Orchard and Accompanying Progeny Test to be Established in 1986 Hybridization of Cryptomeria x Taiwania Start to Work in 1985 6 China-Fir and Luanta-Fir and Their Improvement 51 General Description and Distribution . 51 Previous Silvicultural and Genetic Work . 56 Work Plan for Further Genetic Improvement 59 China-fir Half-sib Progeny Test, Started in 1985 with 200 Parents Establishment of Luanti-fir Seed Production Areas Establishment of Luanta-fir Germplasm Bank 7 Taiwania and Its Improvement . . . . 64 General Description . . . . . . 64 Characteristics and Silvicultural Aspects 65 iv CHAPTER Past Work Related to Genetics . . . 70 NTU Provenance and Progeny Trials, Sowed 1972 A Plan for Saving the Species Flowering and Phenology Studies The Other Studies Work Plan . . . . . . . . 75 A. 400 Half-sib Progeny Test B. Improvements in Regeneration by Cuttings C. Follow-up 1972 NTU Experimental Forest Trial 8 Taiwan Red and Yellow Cypresses . . . . 80 General Description and Characteristics . 80 Silviculture . . . . . . . 81 Past Work Related to Genetics . . . 83 Work Plan for Taiwan Red Cypress Improvement . . . . . . . . 84 True Research Cost of a Large Experiment 84 Details of a 400-family Test 9 Taiwan Red Pine Improvement Plan . . . 91 General Description and Distribution . 91 Special Silvicultural Problems . . . 92 Previous Work on Genetics . . . . 94 Work Plan . . . . . . . . 99 Maintenance of NTU and TFRI Provenance-PrOgeny Tests 10 Luchu Pine and Its Improvement . . . . 101 General DeScription and Distribution . 101 Silviculture and Related Characteristics 102 v CHAPTER 11 12 13 14 15 Past Work Related to Genetics Work Plan Follow-up, Progeny Trial Sub-Tropical Pines Introduction Study Past Work Work Plan NTU Experimental Forest Trial Second Round Introduction Trial Paulownia and Its Improvement General Description and Distribution Previous Work on Genetics Work Plan Witch's Broom Disease Resistance Trial 1969 TFP Provenance- Kadam Introduction Study. General Description and Distribution Species Description and Distribution, Habitat Fruiting, Growth and Regeneration Planting Experiments in Taiwan Results Work Plan A. B. Tree Improvement Work Through Silviculture Conclusions Biblioqraphy Site Adaptability Trial Provenance Trial vi 104 109 112 112 116 117 117 119 120 122 122 125 127 129 133 137 142 TABLE 10 11 12 13 14 LIST OF TABLES Areas of Seed Production Areas of the Most Important Species Established by TFB from 1971 to 1978 (from Tsou, 1979) . . . . Current Tree Breeding Tests in Taiwan (from Hu 35 31., 1981) . . . . . . Summary of Test Plantations in Tree Breeding of NTU Experimental Forest (from NTU, 1981) . . . . . . . Height Growth of Cryptomeria on Three Site Classes in Taiwan (from Liu 23.21-r 1955) . Height Growth of Cryptomeria at Three to Five Locations in Taiwan and Japan (from Liu 9-12. .a_]..o' 1978) o o o o o o o 0 Relative Growth at Age 13 of Nine Seedlots of Cryptomeria Planted at Chitou in 1941 . Relative Growth at Age 10 of 13 Seedlots of Cryptomeria Planted in Chitou in 1967 . . Relative Growth at Age 9 of 8 Seedlots of Cryptomeria Planted in Chitou in 1968 . . Location, Size, and Survival of 7 Test Plantations in Cryptomeria Provenance Test in 1972 (from Chiang £3 11., 1979) . . . Relative Growth at Age 3 of 86 Seedlots of Cryptomeria Planted in 7 Test Plantations in 1972 (from Chiang §t_gl., 1979). . . . Comparison of Luanta-Fir and China-Fir . . Growth Rates of China-Fir and Luanta-Fir (from Wang and Kuo, 1960; Liu gt 21., 1976). Seed Set and Seedling Growth of Control Pollinated China-Fir and Luanta-Fir (from Chiang, 1975) O O O O O O O 0 Size of Taiwania 1 + 1 Seedlings in Relation to Seedbed Density (from Yao, 1970). . . vii 24 26 29 37 37 39 41 42 42 43 52 56 58 67 TABLE 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Size of Taiwania l + 0 Seedlings in Relation to Elevation (from Tai and Liu’ 1970) o o o o o o o o 0 Growth in an Even-Aged Taiwania Plantation at Chitou (from Hung, 1974) . . . . . Relation Between Spacing and Growth in a 15-year-old Taiwania Plantation (Yao, unpublished). . . . . . . . . Height at Age 7 of Taiwania Seedlings Grown from Seed Collected from 12 Natural Stands . Height at Age 7 of Taiwania Seedlings Grown from Seed Collected from 26 Individual Trees Comparison of Commercial and Test Plantings for Taiwan Red and Yellow Cypress . . . Detail of Improvement Thinnings in Taiwan Red and Yellow Cypress . . . . . . Height Growth of Natural Taiwan Red Pine in Different Sites and Elevations (from Hwang, 1979; Hwang, 1969) . . . . . Important Differences Among Taiwan Red, Luchu and Masso Pines . . . . . . Relative Height of Offspring of Eight Stands Planted in 1972 (from Chiang, 1977) . . . The Relative Values of Seed and Cone Characteristics of ZOO-Seedlot, 18-Region Study in Taiwan Red Pine . . . . . Relative Height at Age 5 of Offspring of 200 Trees Located in 18 Stands, TeSted in Three Plantations (from Yang and Yao, unpublished) Survival and Height of 2- to 4—year-old Luchu Pine Provenance Tests in Three Parts of Taiwan, When Measured in 1975 (from Tsou, 1978) o o o o o o o o o 0 Analysis of Variance in Growth of Pinus luchuensis . . . . . . . . . viii 67 69 69 72 73 88 89 93 93 96 97 98 106 107 TABLE 29 30 31 32 33 34 Relative Growth Rates of the Offspring of Eight Selected Plus Trees and Eight Selected Minus Trees Chosen as Controls . . Height Growth of Different Exotic Pines at Different Ages and Sites in Taiwan' . . . Relative Heights at Age 7 of Sub-Tropical Pine Species Tested at Four Locations (from Yao, 1981). . . . . . . . Growth of Kadam in Other Countries . . . The Growth of Kadam in Liukwei and Chungpu (from Liu £3 31., 1976) . . . . . . Size and Survival at Age 4 of Kadam and Three Other Species Tested at Hoshe and Shuili (from Yao and Yen, 1979). . . . . . ix 109 114 115 126 129 130 LIST OF FIGURES FIGURE 1 Hsien (County) Map of Taiwan . CHAPTER 1 INTRODUCTION Tree improvement plan is the planning of tree im- provement work dealing with the planting species involved. It is a package of work including two major aspects, forest genetics and tree breeding and silviculture. Also some other fields are related such as ecology, botany, meteor- ology, geography, etc. It is not a pure or theoretical study pgr fig, but it is a plan/plans with experimental data, practical experiences within given environmental factors, and also including financial factors, manpower and space availabilities, timing, etc., to meet certain specific conditions. In other words, it should be work- able with abundant and solid supporting reasons. An improvement plan could be a simple one but now we are dealing with a comprehensive tree improvement plan; it means it is well designed and inclusive. Trees are so- called long-term growing plants. Whenever it is mentioned, the time factor always should be emphasized. As a matter of fact, it should be time-schedule oriented. The main subject of the plan is the given species. Thus, the im- provement plan will be discussed by individual species. In Taiwan, more than 30 species are now generally consid— ered as economic for planting. Only a few are planted in large areas. Only these major species will be discussed in the text. The others would be benefited by the method- ologies of the discussed species or could be applied by following the pattern from the species which are under a similar situation or have close relationships. Taiwan is an offshore island of mainland China. It is bisected by the Tropic of Cancer. It has a tropical to sub-tropical c1imate--warm to hot and humid in general. The forest is exceedingly luxuriant. Forestry has been important due to such natural conditions. As its silvi- culture goes, because of a preponderance of old, overmature timber and difficult logging accessibility, forest manage- ment in Taiwan consists mainly of clear cutting followed by artificial reforestation (replanting). Gradually, tree im- provement works are involved in such planting-tree system. In recent years, the work attracted more interested people. So far, more than 44 research projects including 64 species have been studied and more works by individuals or by dif- ferent forestry institutes are going on. The whole situa- tion looks optimistic. Nevertheless, there is little gain or contribution to the whole silviculture or management system pg£_§§. Most of them are independent researchers, on a year-to-year basis. Some are poorly organized. For those reasons, a comprehensive tree improvement plan for Taiwan must be worked out in time. The purposes of my study can be generalized as follows: 1. To review the past work in tree improvement of Taiwan and to evaluate it. 2. To generate the comprehensive improvement plan for each species based on known information and data accumulated in the past and the needs. 3. To demonstrate a general scheme of a tree improvement plan and analyze the factors involved and discuss the alternatives and priorities when some factors vary. 4. To try to convince the forestry administrators that theoretically, it is not only feasible but also economical. CHAPTER 2 BRIEF HISTORY OF TAIWAN FORESTRY AND PRESENT STATUS Taiwan in Chinese literally means "terraced bay." It is also called Formosa which came from the "Ilha For- mosa" means beautiful island by Portuguese mariners in 16th century when they discovered the island. Back in 1206 A.D., it became a protectorate of the Chinese Empire. Then in 1624, the Dutch invaded and re- mained as colonists for 37 years, then in 1661 they were ousted by Chinese in Ming Dynasty. Almost in the same time in 1626, Spaniards occupied until 1641 but were dri- ven out by the Dutch. In 1887, Taiwan became a province of China. A few years later, in 1895 Japanese took the island until 1945. At last, at the end of the Second World War, Taiwan is back to China again. Taiwan's forestry started during the Japanese oc- cupation period (1895 to 1945). They introduced logging techniques as well as silvicultural systems. In the mean— time, they also brought their homeland favorite tree spe- cies "sugi" (Cryptomeria japonica, Taxodiaceae) in, with other species such as Japanese cypress (Chamaecyparis obtusa, Cupressaceae). "Sugi" was good, in some respects, even much better than in Japan. Later the "sugi" became the main planting species in the early Taiwan forest his- tory. Later on more species were planted. Some are na- tives of Taiwan but some had been introduced by early Chinese immigrants. Taiwania (Taiwania cryptomerioides, Taxodiaceae) and luanta-fir (Cunninghamia konishii, Taxo- diaceae) are native conifers with excellent growth rates, tree forms and superior wood qualities. Lately, small- scale plantings of those species began. The next one is China-fir (Chinese-fir, Cunninghamia lanceolata). It is a mainland China species, from south-central or south- eastern parts of China including Honan, Shensi and Yunnan, Kiangsu and Hainan. Today it is a widely cultivated spe- cies in mainland China too. No literature shows when or who brought in China-fir. At any rate it was one of the earliest to be introduced and most successful species in Taiwan's forestry history. Other species of conifers such as Taiwan red pine (Pinus taiwanensis), luchu pine (P. luchuensis), masso pine (P. massoniana), Taiwan red cypress (Chamaecyparis formosensis) are also common. In hardwoods, Taiwan acacia (Acacia confusa), Fabaceae), tung tree (Aleurites fordii, Euphorbiaceae), camphor tree (Cinnamomum camphora, Larra- ceae), Taiwan paulownia (Paulownia fortunei, Scrophulari- aceae), Taiwan zelkova (Zelkova formosana, Ulmaceae), etc., are generally common. In addition, bamboos must be men- tioned, the common ones including makino bamboo (Phyllos- tachys makinoi), mengtsungchu (P, edulis), machu (Sinoca- 1amus latiflorus) and green bamboo (Leleba oldhami), etc. Today, those species are still planted. However, there is much interest in fast growing species such as ex- otic subtropical and tropical pines, eucalypts, Laucaena spp., Kadam (Anthocephalus chinenses, Rubiaceae). Those are already in adaptability trials. It is easy to under- stand that the introduction of exotic species has played a key role in Taiwan's forestry. Reforestation is one of the main activities of Taiwan forestry. Based on Taiwan Forestry Bureau recent report (TFB, 1981), 30,571 ha (75,540 acres) of land were planted per year for the period 1971-1980. Finally, to understand the forestry background of Taiwan, some basic data and information must be briefly presented. The total area of Taiwan island is 35,980 square kilometers (13,892 square miles), which supports a population of 17,456,000 people. A central mountain range forms the backbone of the island. The eastern half is very steep and craggy, the western slope is flat, fer- tile and well-cultivated. Approximately 52% (1,870,000 ha) of the island is covered by forests. Of that forest land in the mountains, 75% is in national forests admin- istered for the Taiwan Provincial Government by TFB, 10% is in other public and private forests, and 15% is also national forest land but supervised by various other gov- ernment institutes such as universities or research insti- tutions. TFB is responsible for the large planting pro- grams. In the last ten years the TFB planted about 12,741 ha (31,448 acres) per year in average. The hsien (county) map of Taiwan is shown in Figure 1. 120° 121° - 25° T) £21335“ « \TAIPEI 2.. -' o O " DEADLI“ ...\‘ ..... f.“ .... . . 0' TAICHUNG s"“”?' “ §“", ..---'". ta * 24° Kr». ’ ’ 240-1 ! DE ! $ ! a? .... i o . $ ‘0 .' O I o g . #230 l 2‘ .... O -I N 23 SCALE 9 .29 - 4P .5910! l’220 ‘3 220- O 120 12‘1° Hg Figure 1. Hsien (county) map of Taiwan. 25 CHAPTER 3 ADMINISTRATION, EDUCATION, RESEARCH AND TRAINING IN FORESTRY For improvements in the quality of tree breeding research, the administrative system, manpower resources, education and training in Taiwan are all closely related. The following are the brief facts about them. Taiwan is a province of Republic of China, but also serves as a nation as well. So the forestry insti- tutions sometimes overlap, some being national and some provincial. Generally, they can be separated into three categories: (1) forestry administrations,(2) universities and colleges, and (3) others. Forestry Administrations Taiwan Forestry Bureau (TFB) and Taiwan Forestry Research Institute (TFRI) are included here. They are both provincial, mainly dealing with Taiwan's forestry in different aspects. As for tree improvement is concerned, the department of silviculture of TFRI and Reforestation Branch of TFB both are responsible for it. So far, in TFRI about 3 to 5 senior specialists are involved in the work. On the other hand, TFB has only 2 to 3 personnel in charge of the program. Generally TFB is an execution 10 unit, not supposed to do research except on its own prob- lems. However, it often subsidizes universities, TFRI, etc. Universities and Colleges There are three universities and two junior col- leges offering forestry courses. They are National Taiwan University (NTU), National Chunghsing University (NCHU), University of Chinese Culture (UCC), and Chiayi Agricul- tural Junior College (CYAJC). In addition some agricul- ture vocational schools have forestry. Tree improvement (including forest genetics and tree breeding) is a course required for silviculture major students in all universi- ties and colleges. Both NTU and NCHU offer graduate cour- ses at the masters and doctoral levels. Each year researchers in universities or colleges may get financial aids from supporting agencies. Both NTU and NCHU have experimental forests which are active and productive in doing research and management. Both have land and available research personnel. The personnel are willing to work jointly with professors in universities. Each university has two persons interested in tree improve- ment work. NTU experimental forest has four capable senior specialists working on such studies. Others The first one to be mentioned here is National Science Council (NSC) which is the top agency on science 11 and research administration and management. The majority of research funds and subsidies come from NSC. Next is the Council for Agriculture Planning and Development (CAPD, formerly JCRR--Joint Commission on Rural Reconstruction). Within it, forestry is a branch. The majority of employees in CAPD are specialists. So they serve as advisors in specific fields of agriculture. There are four silviculturists who are related to tree improvement programs. Third is Academia Sinica (AS). For pure research pg£_§g, it is the top-ranking agency in Taiwan. Unfortu- nately, forestry is excluded, but still one or two biolo- gists are doing basic researches related to forest genet- ics and tree breeding. Besides the main categories mentioned above, local governments of some hsiens (=counties) do tree improvement works. This is especially true for hsiens which have for- ests and wild lands, such as Nantou, Hwalian, Pingtung, Yilan. Those hsiens are actively involved and work close- ly in cooperation with TFB and CAPD people. Generally extension is their main job. Advanced Training Taiwan forestry authorities since 1960 recognized the importance of advanced training of their own foresters abroad in order to meet the need of expanding tree improve- ment programs. The training programs are sponsored by 12 separate individual institutions. From NTU Experimental Forest, F. 8. Kung studied a year (1963) at Oregon State University with Dr. K. K. Ching; C. H. Chiang (1967), C. F. Shih (1968), and B. J. Yen (1974) studied at University of Idaho with Dr. C. W. Wang (1974), and Mr. Chiang and Mr. Shih gained M.S. there; Y. N. Yao studied at Michigan State University with Dr. J. W. Wright and gained M.S. in 1971 and later (1976-77, 1982) he was back for his doctoral degree. TFB offered shorter termed training programs, gen- erally less than a year and emphasized in practical as- pects. C. F. Chen (1975) studied at North Carolina State University with Dr. B. J. Zobel, T. S. Lin (1966) at the University of Minnesota with Dr. Scott Pauley, E. K. Tsou (1967) at Michigan State University with Dr. J. W. Wright (1967). Other one-or-two month tours were offered for Japan, New Zealand and Australia for special projects. Of the TFRI staff, T. W. Hu studied at the Univer- sity of Idaho with Dr. C. W. Wang and gained an M.S. in 1966, B. Y. Yang (1965) worked with USDA in the southern United States for a year. Among others, P. C. Kuo of NTU studied a year at MSU in 1970 with Dr. J. W. Wright, Y. K. Fang of NCHU worked with Dr. H. Miyajima at Kyushu University in Japan in 1966. Such training programs are still going on. Today young foresters in Taiwan have more chances for advanced 13 training either in graduate schools in Taiwan or in for- eign countries. Generally Speaking, high-school level foresters in TFB, TFRI and NTU Experimental Forest are sent to NTU or NCHU for 6-12 months training on a non- degree policy. Those trainees follow programs individual- ly designed for each student. Generally the graduates of colleges seek advanced training in other countries. CHAPTER 4 REVIEW OF FOREST GENETICS AND TREE IMPROVEMENT WORK IN TAIWAN Early Period Forest genetics and tree improvement works were started early in Taiwan forestry history. The works are cumulative along with the forestry development and the activities. Since the 19005, there were activities re- lated to tree improvement. Some events had significant influences on later tree improvement practice. When in- terest in forest genetics increased in the 19605, tree improvement work became active. Inevitably some mistakes and controversies existed in that period of time. In the 19705, more intensive programs were started. Then some data and results became available. The important events related to this research will be discussed in detail, the contribution or flaws of them also will be the main con- cerns in this chapter. Forestry practice and cryptomeria (g. japonica) introduction in Taiwan began in the 19005, when Japanese occupied the island. When clear cutting became common, Japanese tried replanting with their familiar species. Then they took cryptomeria along. In 1903, the Experi- mental Forest of Tokyo University, Japan, was established 14 15 in Nantou Hsien, Taiwan. In 1911, cryptomeria was intro- duced and planted at Chitou, Nantou Hsien. It grew very well so the Japanese tested its adaptability by planting many small plantations on different sites and elevations. There were also larger experiments with different spacings to study the effects of spacing on growth rate. Some of these plantations are in Chitou today and NTU Experimental Forest keeps them for permanent growth studies. Another Japanese experiment at Chitou is also in- teresting. That is the cryptomeria race trial. Crypto- meria is a widely distributed native species of Japan, the geographic variations exist. It has been cultivated for hundreds of years and the Japanese have recognized differ- ent races. Different areas have been planted with differ- ent races. The race trial was established in 1941. It includes one 20-tree row of each of the most widely plant- ed races. There was no replication. This was the only cryptomeria race trial they left in Taiwan. The Japanese apparently took the record of this experiment with them. American Visitors and Their Contributions In the 19605, when forest genetics began booming in Taiwan, many activities gradually happened here and there. There must be some persons who made the progress possible, one of the key persons was Mr. K. Y. Tai. In the late 19505, he worked for NTU Experimental Forest. Later he transferred to JCRR a5 a specialist in charge of l6 forestry research. Most of the tree improvement events mentioned in this chapter relate to him. He always sup- ported the programs, and offered opportunities for young foresters for advanced training. He also recommended dis- tinguished foreign scientists to visit Taiwan. Most im- portant, he was unselfish in dealing the business. Now he is still an active person in forestry in Taiwan. In 1964, Dr. Kim K. Ching and his wife Dr. T. M. Ching (both of Oregon State University) came to Taiwan for a one-month stay. Their main purpose was to give an intensive training course in forest genetics and tree breeding for Taiwan foresters. The sponsors were JCRR, TFB, and NTU. Two main subjects were covered in the train- ing program. Dr. K. K. Ching gave introduction to forest genet- ics and demonstrations of practical techniques. Dr. T. M. Ching conducted a course in cytogenetics and related sub- jects including seed and pollen handling and storage tech- niques (Ching and Kung, 1965a, 1965b). That was the first time American forest geneticists were invited to Taiwan. The lectures were interesting and fresh. Dr. K. K. Ching especially had suggestions for research works. It was a long list which almost covered every aspect of forest gen- etics and general silviculture. It was true because tree breeding was just at beginning. The Chings' visit has opened a new era of tree breeding in Taiwan forestry history. 17 In 1965, Dr. Chi-wu Wang (University of Idaho) visited Taiwan for a month, sponsored by JCRR, TFB, NTU, TFRI and Chunghsing Paper Company. There was a training program, the trainees (about 60) were from all over the different agencies in Taiwan and Dr. Wang was the sole in- structor. The program included lectures and field prac- tices. In the final week, a general report was made. In the report, there were tree breeding plan outlines for Taiwan. He placed emphasis on execution of tree breeding work and suggested establishment of a permanent agency on tree breeding. The main benefit to Taiwan forestry of this workshop was the involvement in tree breeding of a great number of interested foresters as well as various forestry agencies. The workshOp was very successful and stimulated Taiwan's interest in tree breeding. Dr. Wang returned to Taiwan for his second visit in 1968-69. This time he worked on a Taiwan red pine progeny-provenance test of 24 seedlots. Later three test plantations were established and 1-year results were pub- lished (Wang 35 31., 1972). In 1972, Dr. Wang spent most of the year in Taiwan. His involvements were as follows: (1) He participated as a representative of the Republic of China in U.S.-China Cooperative Science Program, Semi- nar on Forest Ecology and Genetics. (2) He arranged for seed from 86 stands of cryptomeria from Japan. With those seeds, a provenance test with eight test plantations was established in 1974 by NTU Experimental Forest. (3) He 18 organized another workshop, this time with other Chinese specialists. After the meeting, 19 short articles related to tree breeding were published as a proceedings (TFB, 1972b). Dr. Wang is now retired. He gave his personal enthusiasm for Taiwan's tree breeding work and trained several outstanding Chinese tree breeders from the Uni- versity of Idaho. Now, they are playing the main roles in Taiwan tree breeding field. Another visitor to Taiwan was Dr. Jonathan W. Wright of Michigan State University, U.S.A. He gave di- rect assistance to the work during his two visits to Tai- wan and tried to set up long-term tree improvement plans for Taiwan. In January and March, 1969, Dr. Wright was invited as visiting professor of National Taiwan University for a two-month visit. He taught a course in forest genetics at the university, gave some training classes for forestry pe0ple, and worked with forestry staffs of NTU and the planning and research people of TFB, JCRR, TFRI, and NCHU for a number of consultations. During his stay, several field trips, about two weeks in total, were arranged to let him see and understand the forestry of Taiwan. He ended the visit with a report on Taiwan tree breeding (Wright, 1970a). Some points made in that report are as follows: (1) Acacia confusa is a common low altitude planting species, but with poor trunk form. Improvement 19 is possible, either by silvicultural techniques or selec- tion of other promising species for replacing it. (2) The coordination of forestry and other programs should be strengthened. (3) Cooperation with foreign countries should be emphasized for the exchange of information. (4) In forestry research, there are deficiencies in orien- tation toward Taiwan's needs and the small size of most experiments. (5) For better professional training, "for- est genetics" should be developed at the graduate level. (6) The potential in tree breeding research for Taiwan forestry is very optimistic. (7) Research planning should be improved. A good work plan must consist of justifica- tion, objectives, literature review, methods, costs, time schedule and review. Suggested new projects include in- troduction and testing subtr0pica1 and tropical pines, introduction and testing Eucalyptus species, improvement of the Pinus taiwanensis-luchuensis-massoniana complex, combined provenance-progeny tests of Cryptomeria, Pinus elliottii, Cunninghamia konishii, g. lanceolata, Alnus formosana, Taiwania, Acacia confusa and Taiwania seed or- chard. In July, 1972, he was back again and stayed for one month. The purpose of his second visit was to assist and review the execution of the tree breeding projects. During his stay, he went and saw the test plantations in the field and discussed with the people in charge. Also, he gave some seminars. 20 After he left, he published three articles in Tai- wan with some Chinese co-authors. One was "Plan for the genetic improvement of Taiwan red pine" (Wright et_al., 1972), a plan to accomplish genetic research and estab- lish profitable plantations at the same time. UNDP/FAO Project, 1966-1969 During the late 19605, an important project deal- ing in part with forest genetics worked in Taiwan for 30 months. It is important because we learned good lessons. The project "Forest and Forest Industry Develop- ment Project" was sponsored jointly by the Chinese govern- ment and United Nations (UNDP/FAO Project). There was a team in charge of forest genetics and silviculture, con- sisting of E. H. Hinkle, B. Y. Yang, F. 8. Kung, J. H. Hsu and C. F. Chen. The purpose was to demonstrate acti— vities in forest genetics and tree breeding as the basis for an improved reforestation and planting program. They worked on Chamaecyparis formosensis, g. taiwanensis, Cun- ninghamia konishii, Taiwania cryptomerioides, Pinus tai- wanensis, P. luchuensis, and P. elliottii. The main tech- niques for genetic improvement were phenotypic selection and clonal seed orchard establishment. The team estab- lished Cunninghamia konishii, P. luchuensis and P. taiwan- ensis clonal seed orchards. At the same time, a Taiwania seed orchard was established by NTU Experimental Forest, JCRR and TFRI joint efforts. Unfortunately, there was no 21 progeny testing. The seed orchards continue to be main- tained and they produce seed which is harvested for com— mercial planting. With slash pine (E. elliottii), they applied the same pattern as mentioned above. Slash pine is native to the southern United States. Scions from 67+ trees were imported directly from Florida. A total of 2,518 grafts were made in March, 1969. Only 2% survived, and there was no further work on this species. The next project was the Eucalyptus introduction work, done by owning the UNDP/FAO Project with advice from L. D. Pryor, a eucalypt expert from Australian National University, Canberra, Australia. The eight species intro- duced were E. citriodora, E. grandis, E. saligna, E. 52f busta, E. tereticornis, E. decepta, E. globulus, and E. fastigiata. Before the introduction, Dr. Pryor visited Taiwan twice in 1967 in order to understand the local en- vironmental conditions and choose species with the best chances for success. In the spring of 1969, four low- elevation, private-lands, unreplicated plantations of 2.5 to 5.0 ha were planted. In 1970, TFB planted 250 ha in different places and on different elevations. Except for a brief 1972 visit to one plantation at Yangmei, there has been no follow-up. 22 U.S.-China Seminar on Forestry, 1972, Taipei In 1972, the U.S.-China Cooperative Science Pro- gram "Seminar on Forest Ecology and Genetics" (March 31 to April 8) was held in Taipei, Taiwan. U.S. participants included ten American forestry scientists: C. F. Cooper (San Diego College, EcolOQY), R. F. Fisher (University of Illinois, Ecology), G. Frankie (Texas A & M University, Genetics), J. F. Franklin (Forestry Science Laboratory, USDA, Oregon, Ecology), J. Hett (Division of Ecological Sciences, Oak Ridge National Laboratory, Tennessee, Ecolo- gy), W. J. Libby, Jr. (University of California, Genetics), D. Mueller-Dombois (University of Hawaii, Ecology), G. Schneider (Michigan State University, Ecology), T. L. Sharik (Oberlin College, Genetics), D. B. Zobel (Oregon State University, Genetics). The Chinese delegates with forest genetics interests included C. W. Wang (University of Idaho), P. C. Kuo (NTU), C. W. Yang and T. Kiang (JCRR), E. D. Tsou and C. F. Chen (TFB), S. C. WOo (Academia Sini- ca); F. S. Kung, T. W. Hu, B. Y. Yang, C. M. Lu (TFRI); Y. K. Fong (NCHU); C. M. Kung, C. F. Shi, Y. N. Yao, C. H. Chiang (NTU Experimental Forest). In the meeting, some suggestions were made (NSC, 1972) as follows: (1) Recommendations were made for re- search in Chamaecyparis, Juniperus squamata, and wide- ranging deciduous angiosperms common to both countries. (2) The exchange of scientific information materials and techniques was suggested. (3) The establishment of an 23 official agency in charge of forest genetics and tree im- provement activities in Taiwan was suggested. (4) The es- tablishment of a committee to work out plans for the es- tablishment of seed orchards, clonal banks and seed pro- duction areas. (5) Research in the ecology-and improve- ment of native hardwoods should be emphasized. All the American scientists were young, in their 305 to 405. It was hoped that all would take an interest in and continue to cooperate with Taiwan, but that has not happened. The suggested new agencies were not established. TFB Activities In 1979, TFB senior specialist E. K. Tsou (1979) reported current activities in tree improvement work of TFB. The main items are as follows: (1) A total of 6,231 ha of natural stands of 31 species were designated as seed production areas. Also, 23,138 phenotypically superior individual plus trees were located in "Table l. (2) Clo— nal banks of luchu pine (7 ha), Taiwan red pine (10 ha), Sassafras and Cinnamomum microarthum (1 ha), Taiwania (15 ha), luanta-fir (10 ha), Taiwan cedar (10 ha) were estab- lished. Taiwan red pine and those of luanta—fir have al- ready produced seed. Forest Genetics Symposium, 1981, Taichung In December, 1981, a symposium of forest genetics and breeding was held at Taichung, sponsored by NCHU. Six speakers presented papers. They were Y. N. Yao, C. H. 24 N me II A m mm II II MOHoomww mouhmmomm.mfiwmmeHOE mmcmummu .muomm taco Essoemssfiv .mmcsuHOM macsoasmm .mfimcomomvmo macamdb .msmmOEHOM umnEmpfiooflq .HossuHOM moufiuooad .flflnuwmmanm mssflxmum .Mmsmsoo maomofi "mmoosmumn,uo£uo ea mma m II H H II m mammOEHOH maamnoflz H mm m II II II m It mucuumm m>oxaom m II II II II ma II in mammofiuom mosa< II mm m II 0H oH mm II mmuomsm mfimnom m Hm m I: II II om Hm mmsoammsmu mMHMMmmmm mm moa m I! OH II mm on Es£DGMH0flE EsEoemscfiO II II Hm m H II II II osmmcomams mmstOGSomm .mGMHsommma .m .mflmcosm3wmu .m .Hflocmahm macaw .mflmsmsflno moons .moflnmamo wsxma “mummwcoo Hosuo ov mam mm m mva voH Hum mv mmmsomoeuom .Hm> unsuno .O mmm hmv II om mma voa Hmm om mflmsmmOEHom mwummmommEmsu m II II II In II II mm mcmfiwfl>mp mHHmMHouom com II m mma mha II hma II Hwnmwcox swamnmsflcssu m mm II II um I: m mNH mwmsoonosH moswm us an mma mud II mom II mmoflmmmmfioumhno chMBMwm mmm I: ma II II mowsommfl mauoaoummno wasw assomsmm wsxsz mcocsmz cmnmcm3x scams» cmnm was» smnmcmx tonnage masons mucosa Hasm Iago manhcmq uoauumfio amouom an Amoucuoosv mou< mofloomm “anma .some segue mama 09 Head 20mm mes mm ommmanmaamm mmHommm azeemomsH Amos was no mange oneopoomm ammm mo mammm H mqmflfi 25 Chiang (NTU Experimental Forest), P. J. Wang (Academia Sinica), F. H. Huang (NCHU), J. C. Yang, T. W. Hu (TFRI). The scope of the research in the meeting included genetic engineering, isozyme implicatins tissue culture, tree species introduction and current improvement work. Hu, T. W., of TFRI reported current activities on tree breeding in Taiwan (Hu EE.El°I 1981). The paper 44 research projects, 64 species have been worked on since 1967. Details of the activities are in Table 2. The re— port gave no results except for a species introduction trial of leucaena (Leucaena leucocephala). On a good site at Tulan, the annual volume growth of a fast growing 3 type from El Salvador was 40.3 m per ha per year at age 3. NTU Experimental Forest Activities and My Background and Experience My thesis is "A Comprehensive Tree Improvement Plan for Taiwan." In large part, it is based on a review of the literature. However, I have had personal experi- ence with many of the experiments cited. Therefore, in this section, I give a brief account of the NTU Experimen- tal Forest, where I work, and its relationship to Taiwan tree improvement activities, and of my personal participa- tion in tree improvement studies. The experimental forest of National Taiwan Univer- sity (NTU Experimental Forest) was established in 1949. It is managed under the supervision of College of Agricul- 26 TABLE 2 CURRENT TREE BREEDING TESTS IN TAIWAN (from Hu_§§ al., 1981) NO. . . No. Planting Spec1es Planta- Sponsor Seedlots 'tions Date Provenance Tests: Araucaria cunninghamii 9 3 1969 TFRI Chamaecypgris (7 sppf) 18 1 1975 TFRI Cryptomeria japonica 86 9 1973,74 TFRI,NTU Cunninghamia lanceolata 4 1 1970 TFRI Dendrocalamus giganteus 4 3 1974 TFRI Pseudotsuga menziesii l4 2 1977 TFRI Taiwania cryptomerioides 6 2 1972 TFRI Progeny Test: Alnus formosana 11 2 1970 NTU Camellia oleifera 100 2 1980 TFRI Cinnamomum osmophloeum 147 1 1980 TFRI Cunninghamia konishii’ 56 2 1977 TFRI Daemonorops margaritas 80 3 1981 TFRI Melia azedarach 130 3 1981 TFRI Sapium sebiferum 120 3 1981 TFRI Schima superba 12 1 1977 NTU Pinus luCBuensis 214 2 1970,72 TFRI Egacia confusa 156 3 ‘ 1972 TFRI Broussonetia papyrifera 108 3 1977 TFRI Calocedrus—formosana 104 2 1973 TFRI Chamaecyparis formosensis 224 3 1975,76 TFRI Cunninghamia lanceolata 97 3 1972,73 TFRI Einus taiwanensis 225 4 1971 TFRI TaIWania cryptomerioides 90 2 1980 TFRI 27 ture of NTU. NTU Experimental Forest is located in the center part of Taiwan. It is 33,522 ha in size, which is about one-hundredth of Taiwan land area. Of the total area, 6,380 ha (19%) is operable forest land including 5,280 ha of manmade forest (1978 data). The rest is natu- ral forest, protection forest, bamboo forest or non-forest land, etc. The objectives of the forest management are as follows: (1) To provide facilities and field practice for University students, (2) To undertake forestry research, and (3) To demonstrate advanced forest management tech- niques. There are about 100 members on the staff. Of them, 39% are college graduates and one-third have masters degrees. NTU Experimental Forest research activities are basically the responsibility of the Department of Forest of the University. However, some research is undertaken by the Experimental Forest staff. Due to the heavy loads of administrative work, most staff members only work part- time in research. Research funding comes from several sources. The forest itself has a small regular appropriation for re- search. Individuals often apply for research funds to NSC (National Science Council) or CAPD (Council for Agri- cultural Planning and DevelOpment). The forest publishes its own research publication, "NTU Experimental Forest Bulletin." Both faculty and forest staff members can pub- lish in it. Twice a year, both the faculty and forest 28 staff members gather to review and discuss the research projects and related matters. The most common species in the NTU Forest are cryptomeria, China-fir, luanta-fir and taiwania. Those four species constitute 80% of the plantations. There are also plantations of hardwoods (paulownia, tung tree, zel- kova). Bamboo, cultivated by bamboo farmers, occupies large areas. NTU Experimental Forest contains 93 test planta— tions or research stands, totally 115 ha in size. Among them are some cryptomeria growth study test plantations established in 1921 which are the oldest record-kept test plantations in Taiwan forestry history. The test planta- tions related to tree breeding are listed in Table 3. Following are the research works in which I am personally involved. They are discussed species by spe- cies, and trial by trial. A. 1941 cryptomeria provenance trial. This 9- seedlot experiment was in Chitou and was established by the Japanese. In 1966, I was in charge as a forest ranger there and was responsible for this experiment. Later, I found that record file of this experiment was missing, so I managed it as a common plantation. When Dr. Sakaguchi of Japan visited Chitou in 1972, we asked him for help if he could find the information of this experiment in Japan. He published in 1973 in Japan some growth data. In 1979, when I returned to measure the growth, I found the planta- 29 TABLE 3 SUMMARY OF TEST PLANTATIONS IN TREE BREEDING OF NTU EXPERIMENTAL FOREST (from NTU, 1981) . Area . No. Title (ha) Established Seed- lots Alnus formosana progeny test 1.0 1970 11 Alnus formosana progeny test .3 1974 1 Schima superba progeny test 1.0 1970 12 Kadam introduction trial 1.2 1974 5 Kadam introduction trial 1.8 1980 5 Cryptomeria provenance test 15.1 1974 83 Luanta-fir seed production area 4.0 1971 1 Luanta-fir x China-fir .2 1978 6 full-sib progeny test Luanta-fir provenance test .5 1980 l China-fir provenance test .5 1980 l Taiwania provenance test 3.0 1973 12 Taiwania progeny test 3.0 1973 26 Taiwan red pine progeny test 1.5 1972 24 Taiwania and Luanta-fir 1.2 1966 20 clone banks Exotic pine introduction trial 4.0 1973 25 Sequoia sempervirens and .l 1975 2 Sequoiadendron giganteum Introduction trial 30 tion was gone, except for a few trees. B. 1967, 1968 cryptomeria provenance trials. These 13-seedlot experiments were started by a rice breed- er, Prof. Dr. C. J. Yu of the NTU Department of Botany, who obtained seeds in 1968 and 1969. As Division Chief of Silviculture of NTU Experimental Forest, I became in- volved in 1969. In 1972, when the plantations were 4 and 5 years old, I made the record files, supervised mainten- ance of the plantations. I measured them five times from 1972 to 1982. There has been no publication. C. 1973 cryptomeria provenance trial. A team was assigned responsibility for this 86-seedlot experiment es- tablishment. I was a member of this team. The seeds were obtained by Dr. C. W. Wang from Japan. The team wrote a work plan for a nursery test in Chitou in 1972, and start- ed a project funded by JCRR (Joint Commission on Rural Re- construction) and established seven test plantations in 1973. Before outplanting, the seedlings were measured in Chitou in 1974. During planting season, I was in charge of the Chitou nursery, and establishment of the Salishien plantations, including the mapping and labeling. In Feb- ruary, 1976, I and three others measured the plantations. And in 1979, we jointly published the first report on this trial (Chiang gE_§l., 1979). In addition to helping on the measurements, I did the ANOVAs. D. 1972 Taiwania provenance and progeny trials. 31 The seeds from TFRI, the 12-seedlot provenance test and 26-seedlot progeny trials were started as separate experi- ments. I am totally involved. I started the nursery test and supervised establishment of the test plantations in Chitou and Sanpin in 1972 and 1973. I measured the growth at age 1, 2, 4, 5, and 7. A complete report of the growth analyses was published in 1981. E. 1966 Taiwania density study. I established the test plantation in 1966 in Chitou of NTU Experimental Forest. The total area was 2.0 ha. In 1972, the first preliminary results were published (Yao, 1972). In 1979, I measured again including diameter, height, crown size and mortality: This is the first spacing study in Tai- wania. I include detailed data in my thesis. F. Taiwan pines nursery growth studies. In the past several years, I compared the growth in nurseries of Taiwan red pine, masso pine and luchu pine. The growth of them was uneven within the same species, sometime even strongly contrasted in the same nursery beds. Such phe- nomena have never been found in cryptomeria or China-fir or Taiwania. It is suspected to be due to mycorrhizae effects in microsite differences. A further study is underway. G. 1973 TFRI Taiwan red pine progeny trial. This was established by TFRI. I became involved when it was tested in nursery since 1972. I worked with TFRI members to collect the growth data in the nursery of this 200- 32 seedlot trial. In 1979, I worked on the calculation of ANOVA of the measurements including nineteen traits in cone, seed, nursery and test plantation growth. In 1979, we wrote a paper "Geographic variation in Taiwan red pine," and now it is in review. H. 1971 TFB luchu pine trial. Mrs. E. K. Tsou of TFB established the 214-seedlot progeny test. From 1979 to 1981, I had a grant from TFB to work with Mrs. Tsou on this trial. My major work is to assist in the measurements, the report writing. I. 1973 NTU Experimental Forest exotic pine trial. In 1972, P. C. Kuo and I jointly published a study on "Selection of exotic pine species for testing in Taiwan" (1972). In the meantime, I contacted IUFRO (International Union of Forestry Research Organization), and asked for seed from Dr. J. Burley of Oxford, Great Britain. Then a 25-seedlot trial was started in 1974 in Chitou and Chushan and outplanted in four test plantations in 1975. I meas- ured the seed characteristics and year-old seedling height. I also measured the test plantations at ages 1, 5 and 7. The results have been reported (Yao, 1974, 1979a, and 1981). J. 1975 NTU Experimental Forest Kadam trial. I cooperated with my associate Mr. B. J. Yen to establish test plantations at Hoshe and Hsueli. The seed was from the Philippines. It was supplied by Dr. S. C. Wu of NTU when he had a trip there in 1974. We measured the test plantations in 1978, and thinned the Hsueli plantation. 33 In 1981, we made the second measurement, in order to make second thinning in 1983. Now NTU Experimental Forest is cooperating with TFRI on kadam wood prOperty studies and further adaptability testing. K. 1970 Taiwan alder and guger trial. NTU Experi- mental Forest started hardwood studies quite early. Mr. C. M. Kung established the alder (Alnus formosana) and guger (Schima superba) progeny trials in 1970. Two years later he died without record files. I then became involved. First I labeled the trees and established the record sys- tem, including the plantation maps. Then I measured and supervised the test plantations. And I also distributed the location maps to those who might be interested. In 1975 and 1977, two research papers of these two trials were published in NTU Experimental Forest technique bulletin. CHAPTER 5 CRYPTOMERIA AND ITS IMPROVEMENT Introduction Cryptomeria japonica D. Don of Taxodiaceae, also called cryptomeria or sugi, is one of the major economic coniferous species in Taiwan. It is not native, but was introduced from Japan in the 19005. Because of its good timber qualities, fast growth, ease of planting, and high survival, it used to be commonly planted. At the end of 1965, it had been planted on 21,820 ha. Between 1966 and 1975, an average of 1,487 ha per year (TFB, 1976) were planted. In recent years, due to heavy loss by squirrel damage, the planting has been reduced. But as it is a tra- ditional species, it cannot stop being planted like this. If the damage would be under control, a high planting rate would be back very soon. Cryptomeria has been planted on most parts of the island. It is especially abundant and excellent in growth in Chitou, Alishan, Chutung, Puli, and Chilanshan. It grows best where it is warm and humid with fertile and deep soils. When it is planted on a good site, it can reach 50 to 60 m in height. The optimum elevational range is from 1,200 to 2,400 m in elevation, but sometimes it can grow as low as 900m. The optimum climate condition is about 10 to 34 35 17°C in average temperature, 300+ mm rainfall per year. The humidity is often over 80%. The growth rate generally is 33% faster than in Japan. It cannot grow well on dry sites, or in a windward slope facing west or southwest. The wood is mainly used for construction, bridges, shipbuilding, poles, wooden ware and barrels. The wood is light and soft, but stiff. The specific gravity is 0.30 to 0.33. The heartwood varies from russet brown to dull brown in color. This character is said to relate to vari- ety seed source and the lighter color trees are the best for furniture or panels. Cryptomeria is a native in Japan. It occurs in al- most all the main islands except Hokkaido. It covers a wide geographic range from 30°15' to 40°42' N in latitude and 129°50' to 142°03' E in longitude. Cryptomeria grows from sea level to 2,800 m elevation in the high mountains. Cryptomeria is also a common native species in the southern part of mainland China, being especially abundant in the southeastern provinces of Kiangsi, Anhwei, Chekiang, and Fukien. Species Description, Silviculture and Growth Cryptomeria is a large tree, to 60 m in height. The bark is reddish brown, peeling off in long strips. Leaves are spirally arranged in 5 vertical ranks, green, awl-shaped, l to 2 cm long, incurved, tapering or obtuse. 36 Staminate strobili are in cylindrical clusters. The pol— len sacs are 3-5. Cones are brown, subglobose, 1 to 1.5 cm long with 20-30 scales. Each scale has 2 to 5 seeds with narrow wings. There are about 93,000 to 118,000 seeds per liter (110,000 to 136,000 seeds per pound). Cryptomeria in Taiwan does not bear as much seed or seed with as good germination as in Japan. Therefore, Taiwan- ese foresters have relied completely on Japanese seed sup- plies. Taiwan formerly imported 2,000 liters per year, but in recent years has imported only half as much. The cones mature in October and November. If the seed is collected and stored at room temperature, it lasts about 8 months. When sown in soil, it takes 2 to 3 weeks for germination. The juvenile growth is much faster than luanta-fir and China-fir, and also than taiwania. One-year-old seedlings are good for outplanting. Cryptomeria is rather a selective species as to site. The height growth on different site classes and dif- ferent parts of Taiwan and Japan are shown in Table 4 (Liu gg_gi., 1955) and Table 5 (Liu gE_§l., 1978). The expected rotation is 20-30 years. The estima- 3 ted mean annual increment is 10 to 12 m per ha per year, 3 per ha per year. with a maximum of 19 m The pests and diseases of cryptomeria are usually not serious except for squirrel damage. In the nursery, needle blight (Fungi, Imperfecti) used to cause serious 37 TABLE 4 HEIGHT GROWTH OF CRYPTOMERIA ON THREE SITE CLASSES IN TAIWAN (from Liu_g§ al., 1955) Age (years) Height on Site Class (meters) I II III 10 11.1 8.9 6.8 15 13.9 11.2 8.1 20 16.3 13.2 10.0 25 18.5 14.9 11.4 30 21.4 18.1 13.8 40 24.2 19.6 14.9 TABLE 5 HEIGHT GROWTH OF CRYPTOMERIA AT THREE TO FIVE LOCATIONS IN TAIWAN AND JAPAN (from Liu_§§ al., 1978) Height (meterST (Yggis) Taiwan Japan Chitou Alishan Wulai Akita Kyushu 10 8.5 7.1 8.6 5.2 15 12.7 10.8 12.0 6.3 7.2 20 16.0 14.1 14.5 8.7 9.0 25 18.5 16.9 16.2 11.1 10.8 30 20.7 19.3 17.6 13.4 12.4 35 22.5 21.2 18.6 15.5 13.9 40 24.1 22.5 19.8 17.4 15.3 45 25.5 23.4 21.2 19.1 16.7 damage. Now, however, chemical controls have already been put into nursery management systems, and have been very effective. The other nursery disease is damping off caused by Eythium, Pellicularia, Fusarium, and other fungi, and 38 it is also common. It happens during the lignifications stage of young seedlings. Infected seedlings rot at the bottom. Sterilization of seed and soil generally results in good control. Squirrel damage on cryptomeria is serious. After age 10 the damage happens. The squirrels bite the tree bark and tear it off in long strips. Sometimes they peel the bark and girdle the trunk. The damage causes growth stagnation. At the beginning, the tree crown becomes brown and dried out, then the wounded tree may survive a while without the tip, but the growth would never be re- covered. In Luikuei, 55% of cryptomeria trees of age 45 were damaged (Hwang EE.E£°' 1979). Wang and Kuo (1980) surveyed the damage island-wide. Between 52 and 66% in different stands were damaged. Sometimes, Taiwan forest- ers are so discouraged that the disaster seems out of con- trol. Therefore, many have stopped planting these species. As guns are not legal, the only other control method is poisoning, as is practiced on the NTU Experimental Forest. Past Work on Genetics Many research works on cryptomeria have been done in Taiwan and Japan. In Taiwan, most are growth and man- agement. In recent years, NTU Forest especially has been interested in provenance tests, and four different trials have already been conducted. A brief summary of each tri- al is given as follows. 39 1941 Trial at Chitou In May 1941, a 9-seedlot provenance test was es- tablished at Chitou by the Tokyo Imperial University of Japan. This trial, which covered 3.47 ha, was unreplica- ted. The nine seedlots were from three different parts of Japan (Table 6). TABLE 6 RELATIVE GROWTH AT AGE 13 OF NINE SEEDLOTS OF CRYPTOMERIA PLANTED AT CHITOU IN 1941 Location, . . Seedlot Island and Helg?§ of m Elameter Locality e n 3 Nor. Honshu, Akida 120 110 1 Cent. Honshu, Yoshino 122 120 2 Cent. Honshu, Chito 144 119 4 Kyushu, Etaya A 89 100 5 Kyushu, Obi 100 104 6 Kyushu, Etaya B 86 102 7 Kyushu, Aya 72 77 8 Kyushu, Yakutoshi 88 ‘93 9 Kyushu, Satzema 73 74 Actual Mean 710.0 m, 15.5 cm This experiment was measured in 1946 by Japanese foresters, who returned to Japan with the data and the plantation map. In 1973, these data were published in Japan by Sakaguchi. The experiment was re-measured in 1954 by Young and Chiao, who published their report in Taiwan in 1955. At that time, they did not have maps or any written data on the experiment, and so they relied on the memories of 40 the foremen who had planted and cared for the experiment. I calculated the correlations between age 5 (1946) and age 13 (1954) height, using seedlot means as items. This correlation was r = .92. This indicated two things: (a) the age-age correlation was high, and (b) the fore- men's memories were probably correct. The results are summarized in Table 6. The three seedlots from Honshu grew much faster than the six from Kyushu. There was a very strong correlation (r = .97) be- tween height and diameter. This plantation was cut in 1975. 1967, 1968 Trials In 1967, a 13-seedlot provenance test was estab— lished in Chitou. The seed, except one seedlot from Ali- shan, Taiwan, were given by Dr. K. Sakaguchi, Government Forest Experiment Station of Japan, and with the help of the late Prof. C. J. Yu, Dept. of Botany, NTU. The test plantation is at 1,300 m in elevation, with 10-tree row plots and 10 replications. The spacing is 2 x 2 m. In 1977, I measured the height at age 10 and analyzed the data. The growth was good, and competition had begun. Thinning is needed. The average height of the test plantation was 8.38 m. The results are as in Table 7. The differences were significant. The Alishan seed- lot grew well but those from Shikoku grew best. In 1968, the second test plantation with 8 seed- 41 TABLE 7 RELATIVE GROWTH AT AGE 10 OF 13 SEEDLOTS OF CRYPTOMERIA PLANTED IN CHITOU IN 1967 Seedlot Location (%H:;g£:an) 1 Nor. Honshu, Ajigasawa 83 2 East Honshu, Fukushima 106 11 East Honshu, Tomioka 98 12 East Honshu, Daishi 109 8 East Honshu, Takuhagi 98 3 East Honshu, Jiyama 90 4 South Honshu, Miyama 77 9 Shikoku, None 110 10 Shikoku, Uwajima 109 7 Shikoku, Ehima 105 6 Kyushu, Fukuyama 103 5 Taiwan, Alishan 106 Actual mean 8.38 m lots were established in the same area. With same experi- mental design, the test plantation was measured in 1977 at age 9. The difference was significant. The growth is better than in the 1967 trial and the average height was 9.06 m. The results are as in Table 8. 1972 Trials In 1972, a provenance trial of 86 seedlots was established by NTU Experimental Forest. There are seven test plantations located all over Taiwan (Table 9). Each test plantation consists of 2-tree plots and 10 replica- tions. Spacing was 2 x 3 m. Due to uneven germination, the total seedlots per test plantations vary. In 1977 at age 3, the height, diameter, branch length, and branch 42 angle were measured. The height data are summarized in Table 10 (Chiang EE.El'I 1979). TABLE 8 RELATIVE GROWTH AT AGE 9 OF 8 SEEDLOTS OF CRYPTOMERIA PLANTED IN CHITOU IN 1968 Seedlot Location (%H:;g::an) 4 East Honshu, Takahagi 98 5 East Honshu, Tomiaka 96 8 East Honshu, Daishi 97 9 East Honshu, Nikko 99 10 East Honshu, Nakatokawa 104 12 South Honshu, Shingo 102 7 Shidoku, Ehima 97 13 Kyushu, Obi 106 Actual mean 9.06 m TABLE 9 LOCATION, SIZE, AND SURVIVAL OF 7 TEST PLANTATIONS IN CRYPTOMERIA PROVENANCE TEST IN 1972 (from Chiang g3 al., 1979) Test Location Elevation Survival Plantation in Hsien (m) Seedlots J (%) Chitou Nantou 820 86 93 Salishien Nantou 1980 86 95 Chilanshan A Ilan 1800 81 60 Chilanshan B Ilan 1100 82 77 Yuli Hualien 950 71 82 Alishan Chiayi 2160 79 95 Kuanshan Taitung 1700 82 91 In each plantation there were significant differ- ences among seedlots. Also, there was consistency from 43 TABLE 10 RELATIVE GROWTH AT AGE 3 OF 86 SEEDLOTS OF CRYPTOMERIA PLANTED IN 7 TEST PLANTATIONS IN 1972 (from Chiang eE_a1., 1979) Seedsource in Region Chi Sali Chilanshan . Ali Kuan tou shien A B Yull shan shan Mean Toho, Honshu 85 93 89 - 91 ;83 .87 88 88 Chubu, Honshu 97 96 95 95 92 92 91 94 Kanto, Honshu 99 104 102 103 102 102 110 103 Chyukoku, Honshu 105 103 102 102 113 112 99 105 Shikoku 111 103 114 107 113 102 109 108 Kyushu 102 100 97 102 96 101 102 100 Actual mean 1.99 1.28 0.80 1.62 1.26 1.02 0.81 -- F value** 3.0 3.3 3.9 3.5 4.5 3.5 2.7 -- **All F values are significant at 1% level. plantation to plantation. As each of the seven test plan- tations, seedlots from Shikoku and southern Honshu grew fastest and those from northern Honshu grew slowest. One seedlot, No. 49 from Kyushu, grew 17 to 32% faster than average in each of the seven test plantations, averaging 22% faster than average at all places. In this test, except for 9 seedlots, most are from manmade stands. So the results are not good to study geo- graphic variations. But it gives help for Taiwan crypto- meria reforestation program, because we learned where is the right site for the right seed source. The results of these 4 tests, some general con- clusions, can be summarized as follows: (a) Northern Honshu seed sources are slow growing. 44 (b) Seed sources from Kyushu and Shikoku are better than average. (c) Variation among seedlots and among test plantations is large. Other Genetic Studies in Taiwan Other studies in Taiwan can be mentioned as fol- lows. Kung (1970b) studied the flowering induction in conifers by hormones. He applied gibberellin to 8 coni- fers including cryptomeria. It was effective on produc- ing flowers on 2-year-old cryptomeria seedlings, but not on other species. Ku (1967) described the silviculture practice of cryptomeria in Taiwan, including its races, introduction history, pr0posed seed production areas, total planting areas, and some suggestions on seed procurement and its production improvement strategy. Chang (1969) compiled annotated bibliography on Taiwan cryptomeria. Yao (1977) (unpublished) finished a special problem study on the sil- vicultural system of cryptomeria in Taiwan. Now, some interesting reports will be cited. These reports do give some fine ideas or insights into cryptomeria improvement. Rin (1972) and Lee (1972) re- ported the visit of Dr. K. Sakaguchi to Taiwan in June and July 1972. Dr. Sakaguchi made several points concern- ing the seed source and its adaptability: (l) The growth of individual trees differs within stands. (2) Poor stem 45 form is either genetic reason or due to weed control. (3) Light or dark heartwood color is genetic or due to water stress in the environment. Rin (1973) also report- ed the visit of Dr. H. Miyajima, Professor of Kyushu Uni- versity, in July and August 1973. Dr. Miyajima responded to questions with almost the same answers that Dr. Saka- guchi had given a year before. They both agreed that the provenance trial would clean the seed source complication and give the better seed source for Taiwan's planting. Wright (1970b) wrote a report on Japanese tree breeding experiments for Taiwan foresters when he visited Japan in 1969. He stressed the need for provenance trials. In Japan, the publications on cryptomeria are nu- merous. Here I will cite only those relating to its gen- etics. Hashizume (1973) conducted an intensive study on flowering control in cryptomeria. Toda (1980) studied the karyotype of 3 cultivars in Kyushu. Some differences were found on length of certain chromosomes, but not in the arm ratio. Miyazaki and Sakai (1969) studied the ap- plication of zymography for identification of clones. They concluded the technique is useful for clone identi- fication. Goo (1971) studied flowering of cuttings. Year- old cuttings produced more flowers than twigs on parent trees. Ohba 23.2l3 (1971) studied controlled pollination in greenhouse with year-old seedlings induced to flower by gibberellin. They successfully made F1 and backcross F2 generation within 3 years, and studied a lethal reces- 46 sive gene. Ohba and Marai (1971) continue studying some recessive albino gene by controlled pollination. I found no references to provenance tests on progeny tests. Work Plan Follow—up of 1972 Provenance Trials 1. Maintain all seven plantations. 2. At age 10-15, thin to 1 tree per plot (to a 3 x 4 m spacing) and at age 15—20 remove the 50% slowest growing seedlots (to give 4 x 6 m Spacing). 3. Measure height, diameter, height-diameter ra- tio, and squirrel damage per tree, flowering and fruiting at 5-year intervals. 4. Perform ANOVA for each plantation and for all plantations combined. Compute variousness among and be- tween parental regions. Complete age-age correlations. 5. Publish after each set of measurements, mak— ing each publication 5—6 pages long. Follow-up of 1967, 1968 Provenance Trials 1. Proceed as in "A" above. Management of AliShan Seed Production Area The AliShan Seed Production Area is a 30+ year-old plantation on AliShan from which seed has been collected for the last 15 years. It covers an area of 15 ha and has been thinned to 100 trees per ha, or 1,500 trees total: the trees that are 25+ m tall and 50+ cm in diameter. 47 This area produces 1,200 to 1,500 liters of seed per year, which is sufficient for current needs. The gen- etic quality of the seed is good (see Table 7), but not so good as some stands in Japan. The seed is collected by climbing. This area will be maintained as at present. Improving Seed for Immediate Use Contact Japanese foresters to determine if it is possible to bring seed from stand no. 49 and the other stands producing the best offspring in the 1972 provenance trial. If it is possible, arrange for the purchase of seed from such stands. This may require 1-3 man-months of effort. 200 Parent Half-Sib Progennyest, Start 1990 1. The AliShan seed production area produces seedlings capable of growing 6% faster than average (Table 7). Presumably, stand no. 49 and some other Japanese stands in Kyushu and Shikoka produce seedlings capable of growing 22% faster than average. Therefore, a major new improvement program should be based on those stands. 2. In 1988, remeasure 1973 NTU Experimental For- est provenance test. Determine the 15 stands giving the fastest growing progeny. 3. Go to Japan with a 4-man team, locate those 15 stands, mostly in Kyushu and Shikoku, in fall of 1988. 4. Select best (tallest) 20 parents in each stand, 48 label the trees, collect 50 cones per tree, and keep cones separate by tree. Return in 1989 to re~collect from each tree. As controls, collect seed from 3 average trees per stand. 5. For each stand, record numbers of trees, aver- age height and diameter, age, site characteristics. For each selected tree, record height, diameter, crown size. 6. Store the seeds in a freezer and sow seeds in the nursery in spring of 1990. 7. In 1991, establish five test plantations, each to include 4-tree plots and eight replications. Plant ex- perimental trees on 3 x 4 m spacing, filling every other row with non-experimental cryptomeria. That is, actual spacing at planting time will be 3 x 2 m. Remove the fuller ones as soon as they are merchantable for small poles (probably at age 10-15). 8. Manage all five plantations as potential seed orchards. To do this, remove the two poorest trees per plot at age 10-15 and thin by removing the poorest fami- lies starting at age 15. Plan on Spacing of 6 x 4 m at age 20 and 6 x 8 m at age 25. Wide spacing is essential for heavy seed production. Clonal Seed Orchard and Accompagying Progeny Test to be Established in 1986 1. In 1983 (age 10), re-measure 1973 NTU Experi- mental Forest provenance test. 49 2. Determine the 10 families having the best growth and the least squirrel damage. Within each family, select the five best parents. Select the same number (50) of parents from the best commercial stands in Taiwan. 3. Collect scions from the 100 selected parents and establish two clonal seed orchards, one at the AliShan and one at Kuanwu. Each orchard should contain 100 clones x 25 ramets==2,500 ramets. Spacing should be 3 x 4 m. The space required will be 2,500 x 12 m = 30,000 m2 = 3.0 ha per orchard. 4. At the same time, establish two half-sib pro- geny test plantations of the 100 selected parents, each to contain 100 families x 4-tree plots x 8 replications = 3,200 trees. At a spacing of 3 x 2 m, each plantation will occupy 3,200 x 6 = 19,200 m2 = 1.92 ha. 5. At age 10, measure progeny test and remove from the seed orchards the 25 clones producing the slow- est growing families. 6. To determine the permanent value of these seed orchards, collect the seed orchard seed in 1996 and estab- lish small progeny test including seed orchard seed and a bulked sample of all the Japanese seed collected for the 1990 progeny test-seed orchard program. 50 Hybridization of Cryptomeria x Taiwania Start to Work in 1985 To improve the squirrel damage resistance of cryp- tomeria, taiwania is immune from the damage in general. 1. Raise 20 cuttings collected from aged crypto— meria and plant them in Chushan in 1984. Also, select 2 taiwania aged trees in Chitou as parents. 2. Treat the cryptomeria clones with GA and ex- pect flowering in 1985 to 1986. 3. Do controlled pollinations at Chushan with fe- male cryptomeria x male taiwania and reciprocal. Also, make crosses within species and make unpollinated control crosses. 4. Because seed production of cryptomeria is al- ways better at Alishan than at Chitou, repeat the experi- ment at Alishan. CHAPTER 6 CHINA-FIR AND LUANTA-FIR AND THEIR IMPROVEMENT General Description and Distribution Cunninghamia lanceolata Hook. of Taxodiaceae is called China-fir (Taiwan) or Chinese fir (Mainland China). It was discovered by J. Cunningham in 1701 in Chowshan archipelago of Chekiang Province, Mainland China. It is distributed near the southern part of the Yangtze River, in the southwestern provinces of Kweichow, Yunnan, Szech- uan, Chekiang, Fukien, and Hainan. In Taiwan, it was in- troduced from Fukien in early times, so commonly is called Fukien-fir. Cunninghamia konishii Hayata of Taxodiaceae, also called luanta-fir, is native to Taiwan. It is similar to China-fir in general appearance, but different in some de- tails. The main differences are shown in Table 11. It occurs naturally in the northern and central parts of the island at elevations of 1,300 to 2,000 m, mostly scattered in forests of Taiwan red cypress (Chamaecyparis formosen- gig). It occasionally forms pure stands. According to the UNDP/FAO project, 1969 data showed that luanta-fir is most abundant in the forest district office areas of Chu- tung, Tachia and Luanta. Luanta-fir was discovered in 1907 and named in 51 52 1909. Then the argument began as to whether it should be named E. konishii or E. lanceolata. a famous British botanist, visited Taiwan. In 1912, H. G. Wells, He visited LuantaShan (the mountain of luanta-fir) where the species was discovered. After he saw the beautiful natural for- ests, he recommended acceptance of the name E. konishii. COMPARISON TABLE 11 OF LUANTA-FIR AND CHINA-FIR Characteristic Luanta-fir China-fir Natural range Elevational range in Taiwan, m Seed length, mm Seed width, mm Seed weight, gm/liter Seeds per liter, no. Leaf length, mm Leaf width, mm Cone size, mm Staminate strobili per cluster, no. Rotation, years Wood Specific gravity Taiwan 1,300-2,000 3-5 2-4 214 90,400 8-25 15-25 14-25x16-27 14-16 60 .38-.40 Mainland 500-1,500 4-8 3-5 330 43,000 24-88 25-50 27-42x27-42 16-31 15-25. .35-.36 China-fir is a very commonly planted species. It is planted all over the island in the low mountain areas from 500 to 1,500 m elevation. Thus it is planted in warm, humid regions of 170 to 22°C average temperature and 2,000- 53 3,500 mm annual rainfall. Each year more than 1,000 ha are planted. From 1967 to 1976, according to TFB, a total of 11,100 ha were planted (TFB, 1977). China-fir trees are generally fast-growing. The expected rotation age is 15-25 years, but plantations are sometimes cut at age 10. The maximum size is 45 m height and 1.8 m diameter. This species generally has low wind resistance. Wind damage is common in regions subject to typhoons. Luanta-fir is also an important species for plant- ing in Taiwan. Before 1960, the total plantation area was less than 100 ha, but it has been increasing rapidly. From 1967 to 1976, an average of 118 ha per year were planted and the trend is still increasing. The trees are usually large, the maximum size being 50 m height and 2.5 diameter. Besides being fast-growing and pest resistant, it has high growth vigor, straight stems and beautiful crown form. It can be planted below its natural distri- bution, growing well as low as 700 m elevation. The wood of China-fir is light, soft and easy to work. It has a strong, resinous odor. The specific gra- vity is .50 to .36. It is used for various purposes such as posts, beams, stringers, staircases, ceilings, siding, sub-floors, window sashes, doors, bridges, boat building, furniture, farm implements, coffins, boxes and crates, and slack cooperage. The wood of luanta-fir is relatively light and 54 low in shrinkage. It is very stiff and strong, and is of better quality than is that of China-fir. The specific gravity is .38 to .40. The uses are multiple. China-fir is generally about 20 m tall. The trunk is straight and the bark is brownish outside, red inside. The branches are short, whorled and horizontally spreading. China-fir cones are globose, 3-4 cm long. Seeds are three per scale, narrowly winged. Trees begin to flower at age 5. The flowers appear in April and the seeds ripen in October and November. There are about 43,000 seed (330 grams) per liter (59,000 seeds per 1b.). The seed is considered of best quality when collected from trees more than 20 years old. The germination rate is usually 25-30%. Seeds require 4-5 weeks for germination. When the seed is stored at room temperature, it lasts 6 months. For planting, l-year-old seedlings are commonly used. Cunninghamia is one of the few conifer genera which can sprout from the base if out. Coppice regenera- tion is often used after clear cutting. Pr0pagation by cuttings is also feasible but is seldom used. The esti- mated mean annual increment at rotation age is 13-14 1113 per ha per year. Luanta-fir has reddish inner bark and brownish outer bark, fissured into scales. As in China-fir, the branches are whorled and spreading. The leaves are much shorter than those of China-fir. 55 Luanta-fir cones mature in November. Seed quality is best if the cones are collected from old native stands. In plantations, coning starts at age 20. The seed germi- nation rate is 20-27%. Silvicultural practice is similar to that for China-fir. The growth rate of China-fir on three different site classes has been summarized by Wang and Kuo (1960). Liu EE.§l° (1976), using stem analysis, studied growth rate of luanta-fir. Results for both species are summar— ized in Table 12. The table shows that luanta-fir grows as well or better than China-fir. Estimated mean annual increment of luanta-fir at rotation age is 13-14 m3 per ha per year. Spacing requirements of luanta-fir have been stu- died by Hung (1977). He concluded that thinning should start by age 8 and that the first thinning should not be delayed beyond age 12 for any reason. Disease and insect damage in this species are minimal. Damping off in the nursery can be a problem, but can easily be controlled by chemicals and good nursery management. In plantations, squirrel damage is severe. According to Hwang SE El. (1979), 82 and 91% of the China-fir and luanta-fir trees in a 21-year-old plantation at Liukwei were damaged. The damage does not usually cause mortality, but often permits the entry of decay organisms. For example, in NTU Experi- mental Forest, a China-fir plantation at Hsiengen was given a final cut. We found that only 57% of the wood 56 was merchantable. Chemical control of squirrels is now generally used in NTU Experimental Forest. This is report- ed to be effective, but dead bodies are seldom recovered. Previous Silvicultural and Genetic Work China-fir is an important Species. ”Considerable amount of research has been performed on it and luanta-fir. Wang and Kuo (1960) published a report on the silvics of both species. They covered many details and included a large bibliography. Liu (1968, 1971, 1973, 1974) and Liu gE_§l. (1964, 1979) studied China-fir varieties, growth, and controlled pollination. They concluded that China-fir and luanta-fir should be considered varieties of a single species. TABLE 12 GROWTH RATES OF CHINA-FIR AND LUANTA-FIR (from Wang and Kuo, 1960; Liu 3; al., 1976) Height of Chinaefir Luanta-fir Age on-site quality I II III Years meters 5 7 5 4 5 10 12 10 8 13 15 16 13 10 18 20 18 14 12 20 25 21 17 13 21 30 22 Liu £3,3l. also started a small provenance test, including four seedlots representing both taxa. They 57 planted one plantation with ZOO-tree plots, five replica- tions and a 1.5 x 2 m spacing. They published diameter and volume data. At age 12, after two thinnings, luanta- fir grew best. Kung (1970a) studied pollen storage of luanta-fir. After 3 years, some pollen remained 15-26% viable. Chiang (1975) studied controlled pollination, us- ing four parent trees of China-fir and two of luanta-fir. The experiments were conducted at Chitou on the NTU Experi- mental Forest. The crosses were made in the tops of the tree crowns of trees 20-35 m tall. The seed yields and germination rates were generally high (Table 13). Due to its wide distribution on the mainland, China-fir also is a major planting species in southern mainland China today. Yu gE_3E. (1981) studied flower bud development in Nanpin, Fukien province. Differentiation of female flowers is about 70-80 days later than that of male flowers. Both microscpores and ovules are formed in January and February. Han EE.El° (1980) worked on the karyotypes of trees from Hunan and Fukien. They found varying numbers of B-chromosomes. Chueh (1980) reported on tissue culture, finding that each of four different formulas was equally effective. A half-sib progeny test of 35 seedlots was report- ed in Nanking (Anon., 1977). According to the abstract, the 7-year data showed that one seedlot was almost twice as tall as another. There were also differences in stem 58 form, crown form and leaf color. No details of design were given in the abstract. TABLE 13 SEED SET AND SEEDLING GROWTH OF CONTROL POLLINATED CHINA-FIR AND LUANTA—FIR (from Chiang, 1975) Seeds/ Germination Height Type Of cross cone rate age 2 no. % cm Selfing China-fir 76 43 57 Luanta-fir 63 12 31 Open pollinated China-fir 93 34 56 Luanta-fir 83 26 55 Crossed Luanta-fir x China-fir 83 50 62 China-fir x Luanta-fir 58 1 47 China-fir x China-fir 81 56 51 Productivity in relation to site and stand density were the topics of papers by Chang 25.21' (1980), Anon. (1981), Sheng (1980), and Lun EE.El° (1980). Hwang (1978) reported 18-month data for a luanta- fir provenance test including seedlings from six stands grown at three elevations in Taiwan. Both seed source and elevation had significant effects on growth rate. Hwang and Yang (1978) selected phenotypically taller and average luanta-fir seedlings in a nursery and planted them. They measured them five years later. The 59 "plus" seedlings, which started out as 50% taller than the average ones, were 13% taller than average after 5 years. Work Plan for Further Genetic Improvement China-fir Half-sib Progeny Test, Started in 1985 with 200 Parents This trial is needed because each year, NTU Ex- perimental Forest supplies a large amount of China-fir seed from Hoshe Tract to neighboring forest institutions including the Luanta and Nannon District Offices of TFB. The Hoshe plantations are believed to give superior seed and the seed supply is scarce. Few seeds are produced on commercial plantations, probably because they are too crowded. We do not know where seeds are better than those from Hoshe. This test will serve a dual purpose: (1) to find out where the seeds are better, and (2) to establiSh a seed supply system and offer better seed for commercial planting. Details of the proposed study are as follows. 1. This trial will include seed from Mainland China and from stands in Taiwan. It will include seed from five parents from each of 40 stands, all seed to be kept separate by parent. The trial will include four test plantations, with 4-tree plots and ten replications per plantation. 2. Ask Dr. C. W. Wang of the University of Idaho and Dr. J. W. Wright of Michigan State University for help 60 in obtaining seeds from four to six stands of China-fir in Mainland China. From each stand, 500 viable seeds are de- sired. 3. Contact hsien governments of Yilan, Nantou, Taichung, Chiayi, Maioli, Hsinchu and Tainan. Ask for in- formation about China-fir plantations, including area, lo- cation, age, growth rate, condition. 4. Contact TFB forestry offices of Wenshan, Chu- tung, Tachia, Puli, Luanta, Yushan, Nannon, Kwanshan, Yuli, Mukua, Lanyuan, Tahsueshan. Ask for the same information as obtained from the hsien governments. 5. In 1983-84, select 40 plantations having the best growth for their ages and site conditions. Include the Hoshe plantation and plantations on NCHU Experimental Forest. 6. Select four better than average and one poor- formed tree in each stand and collect 50-100 cones from each parent. 7. In 1985, start the nursery test in the Hoshe Nursery of NTU Experimental Forest. Grow 100 seedlings for field planting. Replicate the experiment four times in the nursery and measure the seedlings at the end of the year in the nursery. 8. In spring 1986, establish four test planta- tions in Hsinchu, Nantou, Tainan, and Taichung Hsiens. Each plantation will contain as many as possible of the 200 families, will have 10 replications and 4 trees per 61 plot. Spacing will be 2 x 3 m. There will be 1,670 trees per ha. Each plantation will occupy 4.8 ha. Additionally, three border rows will surround each plantation. 9. Be good on weed control in test plantations, weeding 3-4 times per year for the first 3 years and 1-2 times per year for the next 2 years. 10. Check survival after one month and measure at 4-year intervals. 11. Do selective thinnings at ages 8, 12, and 16. After age 16, leave the best 20 trees (600 total) in each of the best 30 (=15%) families in each seed orchard. The intermediate age 8 and age 12 thinnings will be light- er, removing about half the trees at each time. The age 16 thinning will have about 120 trees per ha, at an aver— age spacing of 9 x 9 m. 12. The seed orchards should begin to produce improved seed about 1996. By the age-16 thinning, each of the 600 trees per seed orchard should produce about 0.5 liters per year, or about 300 liters per year per seed orchard. 13. If the experiment shows that certain Taiwan stands produce superior seed, those stands can be thinned and managed to produce additional seed. Establishment of Luanta-fir Seed Production Areas Natural stands of luanta-fir are expensive and difficult to collect seed from. Therefore, special seed 62 production areas are needed even though the seed might not be improved genetically. Details are as follows: 1. Select five conveniently located plantations at 1,000-1,800 m elevation and 20+ years of age. 2- In each plantation, thin 3 ha to leave about 100 trees per ha. That has been shown to be a good spac- ing for maximum seed production per ha in red pine and a number of other American pine species. Thinning to leave 800 trees per ha would increase growth per tree but would probably not be effective in promoting heavy seed produc- tion. Establishment of Luanta-fir Germplasm Bank Hinkle and others working for the UNDP/FAO Project listed 95 compartments located in 12 working circles in which there were native stands of luanta-fir in 1970. Many of those stands may be cut and their germplasm lost. To prevent such loss, a germplasm bank is proposed. De- tailed procedures are as follows: 1. In 1984-85, collect scions and/or seeds from each of five trees in each of two stands in each of the 12 working circles. 2. If enough seeds are collected, establish a combined progeny-provenance test. This should contain the offspring of 5 parents per stand x stands x 5 replica- tions x lO-tree plots x 3 plantations = 750 seedlings per stand. The plantations would contain 18,000 trees if all 63 24 stands were represented and would occupy 18,000 x 6 = 108,000 m2 = 10.8 ha at a spacing of 3 x 2 m. 3. If there is no seed, collect scions and es- tablish a clone bank. Use a design similar to that des— cribed in step 2 above, except that 2-tree plots would be used, necessitating only 20% of the space required for the seedling experiment. 4. Because germplasm conservation is the goal, manage the above plantations as arboreta. Be good in weed control, keep good spacing for good growth and a moderate amount of fruiting. When thinning, maintain the maximum diversity of parents and parental stands. Do not attempt to convert into a seed orchard for maximum seed production. However, use as a breeding arboretum should be encouraged. 5. If it is possible to collect sufficient seed from a sufficient number of trees, establish an additional combined provenance-progeny test similar to that described in step 2 above. Manage in the same manner as outlined for the China-fir provenance-progeny test, for maximum seed production per ha. CHAPTER 7 TAIWANIA AND ITS IMPROVEMENT General Description Taiwania cgyptomerioides Hayata of the Taxodiaceae has the common names of taiwania and Taiwan-fir. It is a Tertiary relic species in Taiwan. It is an interesting species, not only for its rarity and geobotanical aspects, but also for its valuable wood. Taiwania is a native conifer, growing in the cen- tral mountain ranges at 1,800 to 2,600 m elevation, in humid valleys with adequate rainfall and deep, fertile soils. It is usually scattered through the forests of Taiwan red cypress or other conifers and hardwoods. Pure stands are rare. It was first discovered in 1904 in wu- shoonkeng of Nantou Hsien, a part of the NTU Experimental Forest. Until 1916, Taiwania was also found in the south- western part of Mainland China, about 2,500 km away in Yunnan. Later, it was reported and found in Hupeh, Hunan, and Szechuan (Li and Keng, 1954; Wang, 1961). Taiwania was planted little in the past. But in recent years, it has become one of the most promising spe- cies for planting. From 1960 to 1970, TFB planted about 165 ha, or 330,000 taiwania seedlings. From 1970 to 1980, 1,100 ha of pure or mixed taiwania plantations were estab- 64 65 lished. In NTU Experimental Forest, there were 148 ha of plantations prior to 1970 and 222 ha after that. Taiwania is considered a potentially valuable species because it grows as rapidly as other commonly planted species and is free of squirrel damage. The wood of taiwania is moderately light, soft, with a specific gravity of .37 - .40. However, it is stiff and strong in bending. The wood is also very resist- ant to termites and marine borers. It is easy to work with tools and machines and is very stable when seasoned. It is used for general construction, furniture, boat building, wharf pilings, coffins, veneers and plywoods. Characteristics and Silvicultural Aspects As mentioned before, taiwania is a native of Main- land China as well as Taiwan. The Mainland China form was named E. floussiana Gaussen in 1939. Taiwania is a large tree, 60 m tall, 2-3 m in dia- meter. The trunk is straight, of pyramidal shape with pendulous branches when young. The crown of old trees is dome-shaped with few branches. The leaves are dimorphic, linear, sharply pointed on young trees, and scale-like on old trees. The staminate strobili are 5-7 per cluster, densely arranged at the tips of branchlets. Mature cones are small, globose, 12-20 scales per cone, and two seeds per scale. The seeds mature in November. There are about 142,000 seeds or 96 g of seed per liter (671,000 seeds per 66 1b.). Generally, the germination rate is about 17%. The age of first flowering is not exactly known. A single tree planted in the open in Liukwei at 1,098 m elevation produced seed at age 26, but nearby 40-year-old plantations produced no seeds. At Chitou, I harvested some seeds from widely spaced trees 32-46 years old. Possibly, on favorable sites, widely spaced trees might start to flower at age 20-25. Seed production varies from year to year, generally being on a 3-year cycle. Vegetative production by cuttings or grafting is possible. Cuttings made from 2-4-year-old seedlings root easily and can be propagated en masse. Grafting is easily done. However, grafted trees do not fruit well. No fe- male flowers were found on trees grafted 10 years previ- ously. In the nursery stage, taiwanias grow much more slowly than cryptomeria, planting Stock is usually 2 years old. Among the factors affecting growth in the nursery is seedling density. Yao (1970) studied this, with re- sults as shown in Table 14. The greater the space per tree, even up to 20 x 20 cm spacings, the greater the seedling growth. With increased space, the top/root ratio increased slightly. The elevation at which seedlings are grown also affects their growth (Table 15). Top growth is greatest at low elevations in warm areas. The top/root ratio is also highest in low elevation nurseries. As a high top/ 67 TABLE 14 SIZE OF TAIWANIA 1 + 1 SEEDLINGS IN RELATION TO SEEDBED DENSITY (from Yao, 1970) Space per seedling Seedling weight weigfiérggtio cmz g no. 100 3.3 1.70 110 3.6 1.71 123 3.9 1.80 139 3.9 1.74 156 4-0 1.91 178 4.0 1.93 204 4.4 1.89 238 5.0 1.92 278 4.8 1.81 333 6.4 1.92 40° 6-8 2.12 TABLE 15 SIZE OF TAIWANIA 1 + 0 SEEDLINGS IN RELATION TO ELEVATION (from Tai and Liu, 1970) Weight of Top Root Elevation of Nursery TOp/root weight ratio m g no. 150 11.6 1.9 6.0 650 5.7 7.8 2.8 1,050 5.2 7.5 2.5 root ratio is not desired for subsequent growth, seedlings should be grown at elevations of at least 650 m. Fertilization affects the nursery growth of seed- lings but not the top/root ratio (Wang and Huang, 1971). Light intensity is also important. The higher the inten- 68 sity, the better the growth. Therefore, according to Lin (1974) and Lin and Chang (1975), taiwania is classified as an intolerant species. Survival in taiwania plantations is generally low, often less than 50%. The main reasons for the mortality are the scanty root systems and transplanting shock. Re— searchers have sought improvement by the use of container- ized stock and by heavy branch pruning before outplanting (Liu and Ho, 1973; Ho, 1976). Pruning is generally help- ful for survival. The first recorded taiwania plantation was estab- lished at Chitou, starting in 1909. Of that plantation, 30 trees remain. Of a plantation established at Toulin on Alishan, 23 trees remain. In 1980, I measured the 30 old trees at Chitou. They ranged from 42 to 71 years old (average 50 years) and averaged 29.7 m tall by 77 cm dia- meter. The tallest was 35 m tall. At Chitou, until age 15, taiwania grows slower than cryptomeria or China-fir. After that, however, the growth rate increases, and by age 30, taiwania can grow faster than the other two commonly planted conifers. Growth rates in a Chitou plantation are summarized in Table 16 (from Hung, 1974). In plantations, growth is affected by stand den- sity. I studied the spacing at Chitou since 1966. At the age of 15, results varied greatly with spacing (Table 17). Spacing affected most aSpects of growth--diameter, crown 69 TABLE 16 GROWTH IN AN EVEN-AGED TAIWANIA PLANTATION AT CHITOU (from Hung, 1974) . Diameter Age Height breast high years m cm 5 3.5 2.6 10 9.3 10.5 15 14.5 16.4 20 19.0 22.2 25 23.3 26.2 30 25.5 29.4 TABLE 17 RELATION BETWEEN SPACING AND GROWTH IN A lS-YEAR-OLD TAIWANIA PLANTATION (Yao, unpublished) Diameter Height Spage per breast Height Dead Crown diameter ree high area ratio m2 cm m % m2 no. 1.8 12 12.7 20 1.9 120 2.2 12 14.8 22 3.5 119 3.1 14 14.7 5 3.7 104 3.7 14 14.4 1 4.2 101 4.2 16 14.5 3 6.1 93 6.3 18 15.4 1 6.9 84 6.8 18 15.2 1 10.7 86 7.8 18 14.5 0 8.5 79 8.0 19 15.7 0 9.6 85 11.1 19 15.3 0 11.4 79 size, form ratio, and mortality. tree, increased space had relatively little effect. With more than 6 m2 per There- fore, at age 15, a space of 2 x 3 m can be considered as desirable, giving rapid growth per tree and a large number 70 of trees per ha. Some people do not accept E. floussiana and con- sider the genus Taiwania to contain only a single species, that is to be monotypic. Most monotypic genera are pest- free. Taiwania is no exception and is one of the most pest-resistant conifers in Taiwan. Few serious damages have been reported. In nurseries, damping off can be a problem, but it is easily controlled by good nursery man- agement. In plantations, some minor insect damage was found in Chitou. It was caused by yellow-spot bat moth (Phassua signifer Walker). The moths attacked young (age 3) trees by girdling and caused mortality (Wang, 1968). Squirrel damage is generally slight. I checked mixed cryptomeria-taiwania plantations at Chitou for squirrel damage. The results were that 42% of the cryptomeria and 0% of the taiwania trees were injured. Past Work Related to Genetics NTU Provenance and Progeny Trials, Sowed 1972 In 1971, TFRI collected seeds from 12 natural stands, the seeds from each stand being bulked. At the same time, seeds were collected from 26 single trees loca- ted in four different hsiens. The seeds were sown at NTU Experimental Forest Nursery in 1972 and test plantations were established in 1973. There were two separate experi- ments, one being the provenance test involving the stand collections and the other being the half-sib progeny test 71 involving the individual-tree collections. Test planta- tions for each experiment were established at Chitou and Sanpin in 1973. Each of the four test plantations follows a randomized complete block design with 4-tree plots, 10 replications and a 2 x 2.5 m spacing. The trees were mea- sured periodically, the last time at age 7 in 1979. Those measurements are summarized in Tables 18 and 19. In all plantations there was a strong correlation between height and diameter, the seedlots which were tall- est also being the largest in diameter. Hence, Tables 18 and 19 contain only height data. For the provenance tests, both height and diameter differences were highly signifi- cant statistically at both Chitou and Sanpin, but not for the two plantations combined. For the progeny test, there were highly significant differences among stands of origin and among families within stands for the Chitou plantation, but only among families for the Sanpin plantation. A com- bined ANOVA of the two progeny test plantations showed significant differences among families but not among stands of origin. The preliminary information from these tests can be summarized by saying that there are differ- ences among individual trees but not among stands. For practical reforestation work, seeds may be collected from stand. However, improvement can be obtained by selecting individual trees. 72 TABLE 18 HEIGHT AT AGE 7 OF TAIWANIA SEEDLINGS GROWN FROM SEED COLLECTED FROM 12 NATURAL STANDS Relative héight at age 7 at Chitou Sanpin (% of mean) Location of parental stand, locality and hsien Kwanwu, Hsienshu No. l 95 98 Kwanwu, Hsienshu No. 2 101 93 Tabachienshan, Hsienchu No. 1 98 106 Tabachienshan, Hsienchu No. 2 97 95 Tahsueshan, Taichung No. 1 99 95 Tahsueshan, Taichung No. 2 97 90 Tahsueshan, Taichung No. 3 97 112 Tahsueshan, Taichung No. 4 105 95 Tachia, Taichung 105 107 Chitou, Nantou 110 121 Tanta, Nantou 88 95 Kwanshan, Taitung 107 100 Actual mean height, m 7.3 4.4 Actual mean diameter, cm 10 6 A Plan for Saving the Species Wright EE.El° (1973) discussed the problem of saving this possibly endangered species and providing for its genetic improvement at the same time. Preservation of stands, not individuals, should be the goal. Also, planting large areas at wide spacings would help. For saving the space, interplanting of another species is an idea. Flowering and Phenology Studies Taiwania flowering phenology, physiology and pol- len storage has been intensively studied at NTU Experimen- tal Forest (Kung, 1965; Kung and Kiang, 1969; Wang, Kung 73 TABLE 19 HEIGHT AT AGE 7 OF TAIWANIA SEEDLINGS GROWN FROM SEED COLLECTED FROM 26 INDIVIDUAL TREES Relative height at age 7 at No. and location of parent Chitou Sanpin (% of mean) 122 Kwanwu, Hsienchu 88 -- 130 Kwanwu, Hsienchu 99 -- 131 Kwanwu, Hsienchu 86 -- 141 Tabachienshan, Hsienchu 96 102 142 Tabachienshan, Hsienchu 97 98 143 Tabachienshan, Hsienchu 94 90 144 Tabachienshan, Hsienchu 98 106 145 Tabachienshan, Hsienchu -- 95 146 Tabachienshan, Hsienchu 105 100 147 Tabachienshan, Hsienchu 104 100 148 Tabachienshan, Hsienchu 96 102 149 Tabachienshan, Hsienchu 92 99 103 Tahsueshan, Taichung -- 90 106 Tahsueshan, Taichung -- 95 107 Tahsueshan, Taichung 98 92 108 Tahsueshan, Taichung 90 90 110 Tahsueshan, Taichung 109 107 111 Tahsueshan, Taichung 107 108 112 Tahsueshan, Taichung 102 103 151 Chitou, Nantou 111 108 152 Chitou, Nantou 105 113 154 Chitou, Nantou 99 105 155 Chitou, Nantou 102 83 156 Chitou, Nantou 109 107 157 Chitou, Nantou 110 112 158 Chitou, Nantou 106 95 Actual mean height, m 7.7 4.6 Actual mean diameter, cm 11 6 and Kung, 1969; Kung, Kiang and Kung, 1969). Female stro- bili are generally found in the top crown of the trees and male strobili lower. Seldom do they occur on the same branches. In Chitou, pollen sheds in February and March, 74 while female strobili mature in March, remaining open for pollination for 5-7 days. Grafting to induce flowering and to conserve germ- plasm has been tried at NTU Experimental Forest since 1965 (Kung and Kung, 1966; Kung, 1974b; Kung, 1973). Male flowering occurred many years ago but nothing happened on female. Gibberellins were tried on the grafted trees, but without results. Some grafts died out gradually. The Other Studies Kuo 22.21' (1972) made a karyotype analysis from a single tree in Yilan. So far, no aneuploidy or poly- ploidy has been found. Chiang (1974) studied the germi- nating seed physiogenetic characteristics and related to its seedsources by means of manometric technique; differ- ences were found only among individuals. Kung (1974a) re- ported one-year-old results on 6-seedlot provenance test; seed from Tahsueshan was best in growth. Lu (1973, 1975) reported 5-seedlot provenance results in seed character- istics and 2-year-old seedling height; seed from Tahsue- shan was poorest in growth. Hwang and Yang (1978) reported S-year-old results on plus-seedling selections on one-year-old seedlings. The plus seedlings decreased the superiority from 150 to 107%, while the check group increased from 50 to 95%. The average height at age 1 and 5 were 30 and 253 cm, respect- ively. Survivals were 80% on average, not significant be- 75 tween groups. Wang g£_§E. (1976) and Wang (1976) reported that no differences were found among seedsources in their iso- zyme study and neither in nutrient absorption in 2-year- old seedlings, respectively. Work Plan A. 400 Half-sib Progeny Test In the current situation, about 1,500 ha would be planted annually, but one limiting factor is how much seed would be available in that year. Generally about 1,200 liters of seed are needed for a single year's demand. TFB recently paid much attention to this species. They not only have selected 1,137 parent trees in natural stands within eight regions for seed production, but also established cutting and grafting clone bank at three areas totally 15 ha in size. In NTU Experimental Forest, they have long been involved in Taiwania research, not only in tree improve- ment but also in growth studies as well. The results certainly are helpful for further improvement. For Taiwania testing, in order to increase the improvement potential, the Mainland China seed source should be included. It serves the purpose of preserving the species for both Taiwan and Mainland China. 1. Ask Dr. C. W. Wang of the University of Idaho and Dr. J. W. Wright of Michigan State University to help 76 obtain 4 to 6 seedlots of seed from Mainland China in 1983. Each seedlot shall contain 500 viable seeds. 2. Ask for 400 seedlots from TFB selected par- ents in the eight regions. Each shall contain 200 cones per parent. 3. Based on the results of NTU Experimental For- est trial, collect the seed from superior parents such as Nos. 151, 157, 156 in Chitou; 110 and 111 in Tahsueshan. 4. Start nursery test in Chitou in 1986. Each seedlot must have at least 60 x 1.2 = 72 seedlings for planting and extra 2,400 for arboretum establishment. The young seedlings will be containerized in the second year and then outplanted in spring of 1988. 5. Establish test plantations in three places including Tahsueshan, Moukuashan and Chitou. The Chitou one will be treated as an arboretum. The other two will be converted to seed orchards after several selective thinnings. The test plantations will be 6 trees per plot, 5 replications with 2.02 x 2.02 m spacing. There will be 400 x 6 x 5 = 12,000 trees in each test plantation. Thus, each plantation will occupy 12,000 x 2.02 x 2.02 = 48,600 m2 = 4.86 ha. For Chitou plantation, there will be 200 x 4 x 3 = 2,400 trees which will occupy 2,400 x 4 x 4 = 38,400 m2 = 3.84 ha. If more seedlots are available, the extra ones could be included in the Chitou test plantation. 6. Thinnings will take place at ages 12, 16 and 20. (a) Thin first at year 1988 + 12 = 2000 A.D. Remove 77 25% (=600) families, and 33.33% within family. Thus, there will be 12,000 trees remaining. (b) Thin next at year 1988 + 16 = 2004 A.D. Remove 50% (=150) families, leaving 150 families. In each remaining family, leave half (=20) best trees. Thus, there will be 3,000 trees remaining. (c) Thin third at year 1988 + 20 = 2008 A.D. Remove 70 of the 150 families, leaving 80 families. Leave 15 best trees (of an original 60) per family. Thus, the seed orchard will contain 80 x 15 = 1,200 trees. 7. For Chitou test plantation, thinning is sched- uled at age 15, 25. Remove 20% of trees regardless of family. 8. The seed orchard presumably can produce im- proved seeds in year 2008 A.D. The total among them will be 1,200 trees x 1 liter per tree = 1,200 liters per year. For summarizing, we are doing combined family and within-family selection, removing 80% of the families and 75% of the trees within family; we will leave 20% of the families and 25% of the trees within family. Thus, we must start with 1,200 x 1/.20 x 1/.25 = 24,000 trees. The number of families and trees per family can be as ' follows: Families Trees per family 100 240 200 120 300 80 400 60 The final spacing should be 9 x 9 m, so the plantations 78 will occupy 1,200 x 9 x 9 = 97,200 m2 = 9.72 ha. Because only 1 of 20 trees will remain after the final thinning, the intial space per tree should be 81/20 = 4.05 m2 = 2.02 x 2.02 m per tree. If 400 - family progeny test is taken, the details of the thinnings are as follows: m2 Age Year Families Tree/family Total trees Area/tree 0 1988 400 60 24,000 4.05 12 2000 300 40 12,000 8.10 16 2004 150 20 3,000 27 20 2008 80 15 1,200 81 B. Improvements in Regeneration by Cuttings One difficulty encountered in the improvement of Taiwania is the delay in flowering and fruiting until the trees are more than 20 years of age. There is a method which could be used to mass produce genetically improved cuttings without waiting for seed. Details of this method follow. 1. Collect seeds from those parents (Nos. 151, 157 and 156 at Chitou and N05. 110 and 111 at Tahsueshan) which produced the best offspring in the NTU Experimental Forest 1972 trials. 2. Raise the seedlings in a greenhouse, raising 5,000 at a time if there is enough room. 3. When the seedlings are a year old, cut them back and make cuttings. There can be two cuttings per tree the first year and three or four cuttings per tree the second year. 4. Root the cuttings in the proper medium and 79 transfer them to containers. 5. It should be possible to produce 5,000 x 2 = 10,000 rooted cuttings the first year and 15,000-20,000 per year the second year. C. Follow-up 1972 NTU Experimental Forest Trial 1. Measure them periodically. 2. The initial spacing was 2 x 2.5 m (2,000 trees per ha). Now thinning is urgent for both test plan- tations in Chitou. 3. Measure and do first thinning in 1983 at age 12. Remove every other row of the plantations (50% thin- ning). Thin across plots to leave 2 trees per plot. The density will be 1,000 trees per ha. 4. Measure and do second silvicultural thinning in 1987 at age 16. Another 50% thinning, same as in first thinning. Now only one tree remains per plot. There are 500 trees per ha. 5. Convert the test plantations to commercial plantations and manage them as commercial plantations. Or give wider spacing in the third thinning for seed pro- duction, but the seed quality is just as commercial seeds. CHAPTER 8 TAIWAN RED AND YELLOW CYPRESSES General Description and Characteristics Taiwan red cypress (Chamaecyparis formosensis Mats.) and Taiwan yellow cypress (E. obtusa var. formosana Rehd. = E. taiwanensis Mas. 32 Suzuki) are both valuable conifers in Taiwan. They are native, happening only in the central mountain ranges of Taiwan at elevations of 1,000 to 2,800 m. Generally they are mixed with other conifers or hard- woods. Few are found in pure stands. Taiwan yellow cypress differs from Taiwan red cy- press taxonomically. In reforestation programs, they are seldom treated separately. The plantations are all called Taiwan red cypress plantations. Taiwan red cypresses are long—lived trees, often hundreds or thousands of years old. A giant red cypress tree in Chitou is estimated as 2,800 years old. The growth of red cypress is moderate to slow but the wood is super. TFB plants about 340 ha per year. Up to 1975, 8,800 ha had been planted. In recent years, more have been planted. For instance, in the single year 1975, 1,795 ha were planted. So today red cypress is as import- ant as any other species for planting in Taiwan (Chu, 1975). 80 81 The wood of Taiwan red cypress is light and mod- erately soft but very stiff and strong in bending. The specific gravity is .33 to .35. The wood has a character- istic odor. It is exceedingly durable against decay and termites. The wood is also straight and even-grained with an attractive figure. It is used mainly for general con- struction, furniture, woodenware and carving. Japanese and Chinese are typically fond of red cypress furniture and buildings. The huge logs are especially valuable when they are used for temples and shrines. Depending on the taxonomic status of Taiwan yellow cypress, there are six or seven species of Chamaecyparis in the world. The others are E. obtusa S. and Z., and E. pisifera S. and Z. of Japan, and E. lawsoniana Par1., E. nootkatensis Spash, and E. thyoides BSP of the United States or Canada. Silviculture Taiwan red cypress is a tall tree, sometimes reaching 65 m height and 6.5 m diameter. The bark is more or less smooth, thin and reddish brown. The leaves are opposite and scale-like. The staminate strobili are ovoid or subglobose. The cones are suberect, 11 x 8 mm in size, with 12 scales per cone and two seeds per scale. The seeds are slightly winged. The seed for reforestation is from natural stands. It matures in November and December. There are about 82 240,000 seeds (280 grams) per liter (379,000 seeds per 1b.). The germination rate is generally 8-10%. It takes 3-4 weeks for germination. There is no definite information to show when this species starts to flower. One experiment showed fruiting of 6-8-year-old trees treated with gibberellin (Chu, 1975). Propagation by cuttings is also possible with up to 84% rooting. Natural regeneration for red cypress is feasible. Some good naturally regenerated stands happen in Taipinshan of Ilan Hsien. But it is not found in other areas. The question of presence or absence of natural regeneration is important. The growth in manmade forests varies with age, spacing and other factors. The available data are not consistent. For instance, the growth of Chitou is not merely moderate but is almost as fast as for cryptomeria. Yang and Wu (1956) compared stand growth of cryptomeria and red cypress in Chitou, with results as follows: Age (years) 10 15 ‘ 20 25 30 Height, red cypress, m 8.3 11.9 14.5 16.8 18.5 Height, cryptomeria, m 9.7 13.9 17.2 19.5 20.8 Hwang (1977) compared growth of 12 stands at dif- ferent elevations and ages. The results were diverse. At 1,300 m elevation in Laonon, height at age 10 was 7.0 m. At 1,350 m elevation at Chitou, height at age 45-50 was 23 m. He concluded that Taiwan red cypress grew best at elevations of 1,000 to 2,000 m. Furthermore, the young stands grew best in the open, poorly as understories. 83 For many years, TFB has planted Taiwan red cypress at a very wide spacing of 4 x 6 m, leaving a space between the rows unweeded. The practice reduces damage from frost. NTU Experimental Forest has practiced planting of Taiwan red cypress at 2,000 m elevation, with severe mortality from frost damage. In recent years, the forest has adopt- ed the TFB practice of wide planting with partial weeding to reduce frost damage. Although squirrels attack Taiwan red cypress, dam— age is much lighter than on cryptomeria or China-fir (Hwang gE_EE., 1979; Kuo, 1957). On the other hand, squirrels damaged E. obtusa from Japan severely at Chitou. Heartwood rot caused by Stereum salcutum Brut. is often found in aged trees. The majority of trunks become hollowed when more than 100 years old. In plantations, the damage is slight because the manmade forests are still too young. Past WOrk Related to Genetics A symposium sponsored by TFB was held for discus- sion of Taiwan red cypress silviculture techniques in 1974. Eight papers were presented; of them, the ones on natural regeneration and interspecific variation were topics of interest to geneticists (Chu, 1975). Liu EE.El' (1975) reported on the seed and char- acteristics of month-old seedlings of a trial including 22 seedlots from seven species. 84 TFB in recent years seems especially interested in this species. Since 1974, they established a clonal seed orchard. They also located and labeled 4,598 parent trees for seed production (Chu, 1975). Work Plan for Taiwan Red Cypress Improvement Taiwan red cypress wood is valuable, but the area of manmade plantations is much less than for cryptomeria or China-fir. The main reason for that is the generally slow growth and the expected long rotations (80 years for NTU Experimental Forest). In recent years, TFB has plant- ed more than ever. Taiwan red cypress has such characteristics that no other species can replace it, but the long rotations are a problem. Therefore, improvement in growth rate is urgent. The main experiment to accomplish this is a half- sib progeny test. True Research Cost of a Large Experiment TFB has started a mass planting program since 1975, each year more than 1,000 ha are planted. Moreover, TFB has also selected 4,597 plus trees for seed production. This indicates that TFB is interested in this species now, and obviously they are very concerned about the seed sup- ply in quantity and quality. On the other hand, TFB subsidizes the research in- stitutions every year to work on certain species in which TFB is interested and the species are commonly economically 85 important. No doubt Taiwan red and yellow cypress fit the case. Most forestry research institutions are capable of carrying out their researches independently. But gen- erally, they are short of land and funds, especially when a long-term project which involves large amounts of money and land is dealt with. Tree improvement works are almost all in this category. To consider the need and demand of both sides of the institutions, the Taiwan red and yellow cypress im- provement work can be carried out jointly by TFB and other research institutions, either TFRI or NTU Experimental Forest. It is possible for TFB to handle the experiment alone, but it is better to let the research institution handle part of the work. Because the quality research is always Specialized, researchers are helpful to deal with specialized problems. In doing this work, some adjustments must be added to the routine silvicultural system of TFB. Some of these adjustments mean more intensive management. Some do cost more, but the results or gains of the improved seed will pay for them. Others involve doing things differently, but may not cost more. Generally, TFB will take care of most silvicultural part of work such as seed collection, seedling raising, plantation establishment and management. Counterpart in research institutions should have a detailed work plan, work closely with the TFB in the field, and all 86 in the test period. In this work plan, no detail steps will be given, but the general considerations for both sides will be mentioned and especially the workload and cost for a com- mercial or test plantation will be discussed. 1. The need of seed supply per year in TFB. Assuming TFB will plant 2,000 ha per year, 3,000 seedlings per ha, then 2,000 x 3,000 = 6,000,000 seedlings per year. There are 4,000 seedlings generally produced by one liter of seed. So, 6,000,000 x 1/4,000 = 1,500 liters of seeds are needed per year. 2. The need of parent trees for seed production. Assuming each tree produces .5 liter seed per year, so we need 1,500 x 1/.5 = 3,000 trees to supply the seed for TFB mass planting program, which means that we must retain 3,000 superior parents in the seed orchard to supply im- proved seed after several selective thinnings. So these 3,000 superior parents are our goal in this half-sib pro- geny test. 3. To proceed from the initial to the final seed orchard, we should remove 80% of the families and 75% of the trees in each family. Thus, for every tree in the fi- nal orchard, we must plant l/(1-.80)(1-.75) = 20 trees. That means the seed orchards should start with 3,000 x 20 = 60,000 trees. The goal of 60,000 trees could be met with one of the following combinations: 87 Families Trees/family 100 600 200 300 300 200 400 150 500 120 600 100 4. The size and spacing of the test plantation and seed orchard. In the seed orchard, we prefer the fi- nal Spacing of 9 x 9 m, so the size is 3,000 x (9 x 9) = 243,000 m2, in other words it is 24,30 ha of seed orchards. And the initial spacing in test plantation is 243,000 x l/60,000 = 4.05 m2 per tree or 2.01 x 2.01 m. 5. Workload and cost analysis. To understand the test plantation and commercial plantings, a comparison based on the work items is made in Table 20. It is clear that the workload of 400-family progeny test is only a small portion in the TFB routine work. Com- paratively speaking, testing seed collection would be more tedious than other items. And certainly it costs much more, if even one more parent tree or more area is added. But it seems not in this case, because individual trees producing the seed are definite (must exceed 3.75 liters per tree). And it is no question that TFB would like to get the seed from their selected parents and in most of the stands. Other items, such as mapping, measuring and calculating the ANOVA, research fellows are responsible for, expenses are little, and they will be in the joint research project. 88 6. Use of wood from the tests. 7. Thinning schedule. Taiwan red and yellow cypress grow slower than cryptomeria or China-fir, so the intermediate thinnings would begin at age 20, then 30 and 40. The detail of the thinnings is as in Table 21. TABLE 20 COMPARISON OF COMMERCIAL AND TEST PLANTINGS FOR TAIWAN RED AND YELLOW CYPRESS Commercial 400-fami1y Worhoad Extra Item planting progeny ratio work or per year test (%) included Seed Supply seed amt. (liters) 1,500 1.4 .09 included seed treatment bulk separate -- extra seed packaging bulk separate -- extra no. of parents 3,000 400 13.33 included NurseryyMgt. nursery bed area (m2) 25,005a 400b 1.60 included no. seedlings 6,000,000 60,000 1.00 included growing in mix replicate -- extra seedling packaging and shipping mlx repllcate -- extra Plantation Mgt. site preparation (ha) 2,000 24.3 1.22 included planting in mix replicate --. extra thinning (times) 2 3 -- extra thinning (ha) 4,000 48.6 2.43 extra thinning simple complicated -- extra mapping yes yes -- extra border row planting no yes -— extra measurement (year) 10 5 -- extra 3.06 liter per m2, 16.67 m2 per one liter seed. b 2 One family per row, 10 rows per m , 10 replicates. 89 TABLE 21 DETAIL OF IMPROVEMENT THINNINGS IN TAIWAN RED AND YELLOW CYPRESS . . Trees/ Total Space/ Age Year Families Family Trees Tree 0 1988 400 150 60,000 5.05 20 2008 200 100 20,000 12.15 30 2018 100 50 5,000 48.60 40 2028 80 37.5 3,000 81.00 Details of a 400-family Test In the last section, I attempted to show what parts of the cost of a progeny test should be assigned to research and to action agencies. Here I consider detailed steps of the progeny test in chronological order. 1, Start seed collection in 1983. Ask TFB to §E2£g_the seed they will collect from their selected parents. 2. Ask for 100 cones (about 500 viable seeds) from each parent from their 12 natural areas. The re- search institution should help in the work. 3. If it is a poor year, additional seed will be collected in 1984 and 1985, and the already-collected seed will be stored in a freezer. 4. Cooperate with TFRI, and ask Mr. T. Liu or T. W. Hu for 8 to 10 seedlots of exotic cypress and join the testing. 5. Start the nursery trial in 1986 at Lanyang 90 District Forestry Office (Lanyang DFO) which is special- ized in Taiwan red and yellow cypress breeding program, according to TFB's tree improvement plan. 6. According to the statement in the last sec- tion, if the family size is 400, then the trees within family will be 150 to meet the requirement of 80 and 75%, respectively, on selective thinnings among families and within family. In other words, for this test, we need to raise 60,000 seedlings in total. 7. To give preliminary results in the first one or two years of the test, the replicated randomized de- sign in nursery is needed. The lay-out work is the re- sponsibility of the researchers. And the nursery manage- ment is as routine as commercial nursery management. 8. The seedlings will be ready for outplanting in spring of 1988. The three test plantations will be in Lanyang DFO. Based on the seed orchard requirement of 9 x 9 m spacing and 3,000 superior parents, the size of the plantations, as we have mentioned, is 24.03 ha, and the initial spacing is 2.01 x 2.01 m. 9. The selective thinnings will be done at age 20 (2008 A.D.), in 30 (2018) and 40 (2028). The thinnings will follow the schedule shown in Table 21. 10. After the third cutting in 2028, it becomes a seed orchard and will produce the improved seed of 3,000 x .5 liter = 1,500 liters per year. 11. At maturity, the trees will be harvested commercially. CHAPTER 9 TAIWAN RED PINE IMPROVEMENT PLAN General Description and Distribution Pinus taiwanensis Hayata, also called Taiwan red pine or Taiwan two—needled pine, is a native of Taiwan, distributed in the central mountain ranges of the island. It has a straight trunk, is adaptable to poor soils, and grows rapidly. It began to be planted in the early 19205 in Alishan. In the 19505, it was planted at the rate of 400 ha per year. Since then it has been widely planted. In 1971, it reached the peak of 4,800 ha per year. But in recent years, due to seedsource and silviculture diffi- culties, the planting rate dropped to about 500 ha per year. It still plays an important role in TFB's reforesta- tion from 650 m at Julianchi to 3,000 m in Pashienshan. Pure, large, natural stands can be found. It is most abun- dant along the Tachia River Valley in Taichung Hsien be- tween the elevations of 800 and 1,800 m. Today, stands under 1,000 m are rare. Its geographic distribution is limited to a 1° strip along 121° E between 22°30' and 24° 30'N. The wood is yellowish white in color and moderate- ly heavy. The specific gravity is between .47 and .51. It has high nail-holding ability and is easy to dry. It 91 92 is used for construction and building, bridges, railroad ties and poles. Also, it is the raw material of pulp. Taiwan red pine is a large tree growing up to 35 m in height and 80 cm in diameter. Its bark is fissured into small scales. Cones are oblong to ovoid, 6 x 4 cm in size, about 100 scales per cone. Seeds are winged, 5 x 2 mm in size. There are about 37,000 seeds (500 gms) per liter (34,000 seeds per 1b.). Flowering starts at age 3. Flowering times vary with elevation, the flowers appearing from February to April. Individual male catkins shed pollen for 3-5 days and female flowers are receptive for 7-10 days (Wang, 1971; Wang and Lin, 1974). Variations in growth rate with site quality and elevation are given in Table 22, taken from publications by Hwang (1969, 1970). Taiwan red pine is one of three hard pines planted in Taiwan, the other two being luchu pine from Okinawa and masso pine from Mainland China. Some of the most import- ant characters differentiating the three are shown in Table 23. Special Silvicultural Problems As mentioned before, in recent years, TFB has had some bad experiences in planting Taiwan red pine. Many hectares of plantations have crooked stems, stagnant growth and shrubby form after 10 years growth. However, near 93 TABLE 22 HEIGHT GROWTH OF NATURAL TAIWAN RED PINE IN DIFFERENT SITES AND ELEVATIONS (from Hwang, 1979; Hwang, 1969) Natural Stand Height on . . site class Height at elevation Age I II 1000 1500 2000 (yrs.) m m m m m 10 12.6 10.5 7.5 7.6 5.8 15 16.1 13.5 11.7 12.0 8.8 20 19.2 16.2 15.1 16.1 11.8 25 22.2 18.9 17.9 19.8 14.5 30 24.7 21.4 20.2 22.1 16.6 35 -— -- 22.0 23.6 18.3 TABLE 23 IMPORTANT DIFFERENCES AMONG TAIWAN RED, LUCHU AND MASSO PINES Item Taiwan red pine luchu pine Masso pine (E. taiwanensis)(E. 1uchensis)(E. massoniana) 500-1300 m in northern Below 1000 m in northern 800-2800 m in cent. mtn. Elevation range in Taiwan range forming pure forest Growth rate fast Trunk form straight Bark color grey-brown Needle length 8-11 cm and central Taiwan fast crooked dark brown 12-18 cm and coastal hills slow crooked reddish to greyish brown 12-20 cm such plantations may be others 0118 trees . consisting of tall, vigor- Wang EE.§l' (1980) investigated more than four 94 thousand ha of lO-year-old young red pine plantations in Ilan, Chutung, Tachia and Luanta. They tried to find out the causes of its failures in reforestation. They measur- ed the diameters, took soil samples, collected local cli- matic data and other site factor information. In the mean- time, they inquired the questions concerning the background of the seedlings, including where the seed comes from, nursery practices and methods, planting and management, etc. Their data indicate that roots competition to un- favorable sites, seed sources, and planting of more than one seedling in a hole were major factors responsible for poor growth. Lanner and Hinkle (1970) reported on one specific cause of stem defects. Disease problems are generally minor; the most serious disease is a needle cast caused by Lophodermium pinastri. Previous Work on Genetics Kuo et 31. (1972) made karyotype analyses of these three species. They reported no differences in karyotypes. Hu and Yang (1972) took increment cores and stu- died wood specific gravity and fiber length of trees in light stands near the Tachia Valley. The stands ranged in elevation from 700 to 2,700 m elevation. Some were the same as sampled in 1971. Specific gravity and fiber length varied considerably, individual tree values ranging from .39 to .56, and 2.78 to 4.41 mm, respectively. Stand means varied from .44 to .54, and 3.17 to 4.01, respect- 95 ively. Specific gravity decreased with increasing eleva- tion, but fiber length did not vary with elevation. A combination progeny-provenance test was started in 1972 by NTU Experimental Forest, using seed from dif- ferent trees than those mentioned in the two previously mentioned experiments. The 1972 test includes the off- spring of 24 trees growing in 8 different stands, ranging in elevation from 900 to 2,450 m. The test was planted in three separate areas. This 1972 experiment was measured by Li (1976), Wang et_§l. (1972), and Chiang (1977). At the low and medium elevation test plantations, there were strong cor- relations (r=.93 and .95, respectively) between 16 months and 5 years height, and there were significant differences among the offspring of different stands, low elevation seedlings growing the fastest (Table 24). In the high- elevation plantation, with its high mortality, the differ- ences among stand—provenance were not significant. In the fall of 1971, B. Y. Yang and others from TFRI collected cones from 200 native trees growing in 15 localities of Taiwan. Wang and Lin (1974) and Yang and I (unpublished) measured the cone and seed characteristics of these 200 trees, with results as shown in Table 25. Seed and cone size differences were statistically signifi- cant among trees within locality and among localities, but did not vary consistently with elevation. Cone and seed shapes (length/width ratio) were relatively constant 96 but cone-scale shape was variable. The longest cones had the longest but not the widest scales. There was a nega- tive correlation (r=.59) between elevation and seed weight, low elevation trees producing the heaviest seeds. TABLE 24 RELATIVE HEIGHT OF OFFSPRING OF EIGHT STANDS PLANTED IN 1972 (from Chiang, 1977) Place of Origin Percent of mean, age 5 at: Locality Elev. Wulai Chitou Trefong Tahsueshan 900 118 120 116 Tachia 930 122 142 90 Puli 1180 126 134 91 Yuli 1200 111 98 108 Chutung 2000 94 90 -- Luanta 2250 82 75 110 Lanyang 2250 83 82 94 Mukua 2300 80 59 96 Actual mean, cm. 140 440 140 Survival, % 48 58 23 The seeds collected by Yang and others from TFRI in 1971 were used to establish another progeny-provenance test. The seeds from 200 parent trees located in 24 stands were sown in 1972 and three test plantations were established in 1973. In 1977-78, Yang and I analyzed the nursery and 5-year data. The results are shown in Table 26. The two Zuyunshan plantations are located close to each other but were planted one year apart. In both, 97 come no w« ca mm on mm n ma NH m m AEEV some Hmsuom voa mm mm cm mm mm mm mm Hm vmmm BB oma NHH moa voa mm moa moa HHH woa mmmm me oaa mm NHH mm hm vm vm hm MOH bmmm Um mHH «OH vaa Goa mva voa boa moa moa mmmm MB mv mad mHH mad mva moa mm mm vm oovm Ad mm om boa mm Aha mm mm moa mm comm Mm mm mm NHH «OH mma mm vm mm mm OOHN mm mHH mm moa voa mm HHH vm HHH moa omom 3M moa mm om hm vm mm ooH Hoa Hoa mama m3 moa mm mm mm mm Hoa vm om mm mesa Sm odd mod mm boa mm boa HOH hm ooa mvma MO HHH voa moa HOH mm moa mm mm Hoa Hmma Mm NHH moa mm moa mm mm mad OOH ooa mwma 93 boa wm mm ooa mm moa HHH voa mod OMNH m0 boa moa mm vm mm mm mHH hm moa mNHH BM mHH em vm HOH mm moa mma moa mm Hmm MM hm mHH mm mm mm moa NNH moa Hod mam NS mm mm mm Hm om mm mm moa NOH ooh mm .UHZ .cmq .QH3 .amq .Uflz .cwa .uz .Ufiz .cmq .>mHm :oflmmm «mamomlmcou *mcoo amcfiz w comm hEmmm mZHm 0mm 243H¢B ZH MQDBm ZOmeMImH mUHBmHmmBU¢m¢mU MZOU 02d Dmmm m0 mMDQ¢> M>Hfidflmm mmfi mm mqmdfi .Bonnmmmloom m0 98 low elevation trees grew fastest. Also, there were signi- ficant correlations between seed weight-height at age 5 (r=.84, .95) and between heights at age 1 and 5 (r=.68, .66, respectively). TABLE 26 RELATIVE HEIGHT AT AGE 5 OF OFFSPRING OF 200 TREES LOCATED IN 18 STANDS, TESTED IN THREE PLANTATIONS (from Yang and Yao, unpublished) Height at: Place of Origin Zuyunshan Tahsueshan Locality Elevation 1500 m 1500 m 2100 m (m) (% of mean) (% of mean) Songho 700 133 145 108 Maiyuan 919 121 123 90 Kukwan 981 146 140 101 Kuantochi 1125 142 137 108 Chinshan 1230 120 125 100 Wanta 1263 118 130 97 Sunkwan 1621 84 95 99 Chiayang 1643 106 97 105 Songmao 1745 91 103 102 Wushe 1803 122 140 100 Kwanwu 2050 70 90 71 Kwanshan 2100 -- 90 -- Shianyang 2300 -- 75 -- Alishan 2400 -- 61 -- Tayulin 2523 75 83 110 Songchuankian 2537 71 68 104‘ Tahsueshan 2623 77 76 95 Tanta 2654 75 88 111 Actual mean ht., cm 250 180 140 Survival, % 61 59 65 99 Work Plan Maintenance of NTU and TFRI Provenance-Progeny Tests These are basically good experiments, sufficiently latge to solve the main questions about geographic varia- tion. They should be maintained, but need not be repeated. The main procedures are as follows: A. Maintenance. Weed the plantations once a year. Cut vines. Maintain firebreaks. B. Measurement. Re—measure height, diameter, stem form, fruit production, pest damage at 5-year inter- vals. C. At age 10, thin to the two best trees in each plot. Shortly after crown closure, remove the 50% of families having the poorest growth. D. Thin again (by family and within family) at age 15 to 20. E. Analysis and publication. Do ANOVAs on the data from each plantation separately and all plantations combined. Do age-age correlations. Publish at 5-year in- tervals. Include data from both experiments in the same publication. F. Maintain these plantations as experiments and demonstration areas. Do not convert them into seed or- chards. To do so, we would remove 15 or 18 seedlots or leave Kukwan, Kwantochi and Sangho. That is, we would leave 25 families for the seed orchard. With that strategy, 100 there would be no gain in seed orchard, because we could as easily go back to the parent stands to collect the seed there. Furthermore, it is very expensive for running a seed orchard. If there are some missing plots, then the families would be less than 25. If there were an attempt at improve- ment by removing the poorest 5 of the 25 families, only 20 families would remain. Because mortality rate averaged 40%, those 20 families would contain only 192 trees, hard— ly enough to be considered as a seed orchard. CHAPTER 10 LUCHU PINE AND ITS IMPROVEMENT General Description and Distribution Luchu pine (Pinus luchuensis Mayr.) is a native pine in Ryukyu Islands. According to Liu gt_al. (1967), it was introduced to Taiwan in 1903 by Tashiro Antey, a Japanese taxonomist. Then, in 1905 and 1906, more seeds were taken in. Since then, it is widespread in low eleva- tion areas all over Taiwan. The Ryukyu Islands are located between Japan and Taiwan, at latitudes from 240 to 29°N. In most of the areas there, from seashore to mountain slopes, the land used to be covered by luchu pines. Due to the Second World War, most forests were destroyed, but some are left in the southern islands of the group (Mirov, 1967). Except for the original introduction, all seeds used in Taiwan are collected in Taiwan. Luchu pine is very adaptable to the sites of the northern part of Taiwan. The sites which are good for masso pine (P. massoniana) or acacia (Acacia confusa) are also good for luchu pines. It is a sub-tropical species, favors warm and wet to humid climate, and prefers sandy loam and red clay soils. It is not suitable for peat or hard clay. Generally, it is planted in areas below 1,200 101 102 m in elevation. It grows well in mountain or hilly areas, but very poorly near the sea coast. Before 1960, 2,000 to 3,000 ha were planted annually. From 1966 to 1970, the planting areas were 4,377, 1,959, 2,121, 2,075, and 1,750 ha, respectively, per year. But in recent years, the plant- ing areas dropped sharply, to totals of 263, 48, 58, 0, 158 and 36 ha per year for the years 1971 to 1976, respectively (TFB, 1972a, 1977). For its future plantings, it is expect- ed that about 300 ha per year would be planted. The wood of luchu pine is heavy and hard. The sap- wood is light yellow and the heartwood is yellowish brown in color. The fiber length and width are 4.04 mm and 44.3 u, respectively. The specific gravity is .44 to .49. The major uses are for construction, mining poles, pulp, ship- building and oleoresin. Silviculture and Related Characteristics Luchu pine is a hard pine. It is in Subsection Sylvestres. Its close relatives are Japanese red pine (P. densiflora), Japanese black pine (P. thunbergii), masso pine (P. massoniana), Taiwan red pine (P. taiwanensis), etc. It is a medium-size tree of 20 m and 50 cm in dia- meter. It flowers in March and the cone matures in Octo- ber and November (Liu, 1977; Chang, 1950; Ho, 1968). There are about 42,000 seeds for 459 grams per liter (41,000 seeds per 1b.), and the germination rate is generally high, up to 45%. It bears seeds every three years, and begins 103 to flower when four to five years old. Luchu pines grow faster than other native pines in the young stage, up to 20 years of age. After 30 years old, the growth rate declines. The general expected rota- tion ages are 20 to 25 years, with the first thinning at age 10. One private stand in northern Taiwan, 19 years old with 1,290 trees per ha, averaged 16.4 cm in diameter, 9.7 m height, and 130 m3 3 per ha volume. In other words, it grew about 6 m per ha per year (JCRR, 1966, unpublished). Lin (1974), Liu and Lin (1970) studied the effects of site quality and spacing on luchu pine growth. The concluded that this species grew best on fine to medium sandy loam soils 20 to 50 cm deep, and 200 to 600 m in elevation. Stand yield at age 30 varied from 275 to 447 m3 per ha, averaging 350 m3. Other studies on growth of luchu pines on different sites included Chiang (1956), Hwang (1952, 1968), Hung gt_al. (1976), etc. Most of their data were collected from individual trees or small groups of trees, and the growth per ha was often over-estimated. Fang (1968, 1969) studied grafting on luchu pine. He found that after 10 months, graft survival rates were 39, 28 and 23%, respectively, for scions taken from trees 10, 20 and 30 years old. Scions from young trees had the fastest growth as well as the highest survival. In another study in spring of 1967, he tested effects of cold storage and date of grafting. Graft survival ranged from 1 to 38%. Grafts stored at 70 to 80% relative humidity were more 104 successful than fresh scions, and grafts made in February survived better. There are no data on flowering of graft- ed trees or the rooting of luchu pine cuttings. Two major diseases of luchu pine exist: pine leaf blight (Cercospora pinidensiflorae) and blister rust (Egg: nartium quercum). Others are damping off of young seedlings in nurseries and pine canker (Diplodia spp.). As for in- sects, the most important is the pine caterpillar (Dendro- limus punctatus W.), doing the most damage on trees less than 10 years old. In 1956, it destroyed about 1,000 ha luchu pines in northern Taiwan. In 1974, it almost de- stroyed all the luchu pines in NTU Experimental Forest arboretum in Chushan. Its biological control studies have been tried by cytoplasmic polyhedrosis virus (CPV), which seems effective (Ying, 1970; Yie gt al., 1967; Koyama, Past Work Related to Genetics In pine root tips, there are 2n = 24 chromosomes. For luchu pine, Shidei and Moromizato (1971) made a com— plete karyotype analysis in Japan. The ten pairs were metacentric and the other two were submetacentric. Second- ary constrictions were observed on the short arms of the third, sixth and tenth chromosomes. The chromosomes were 8 to 13 um long and 1.5 to 2 um thick at meiosis. No literature has been known on luchu pine's hy- bridization work so far. Wright (1976) mentioned that most 105 crosses of pines in subsection Sylvestres can be made be- tween Asiatic Species, but luchu pine was excluded. Yang and Yang (personal contact) of TFRI have tried the crosses between luchu pine and Taiwan red pine and F1 hybrids have been outplanted, but no further information is available. In 1969, an island-wide provenance and progeny test of luchu pine was started by Tsou (1976, 1978) of TFB. She collected seeds from 214 trees located in 27 stands in 18 regions over a 3—year period. She included the offspring of 214 different parent trees into 17 test plantations in northern, central, and southern Taiwan from 1971 to 1973. The total area of test plantation is about 35 ha. In the winter of 1975, first measurements were made of the 2-4- year-old tree. The average survivals and height are sum- marized in Table 27. In luchu pines, it is often mentioned that northern Taiwan is more favorable for its growing, but it is not always true according to these results. For the performance of the seedlots in each test plantation, there are some very consistent results for cer- tain good seedlots. For instance, seedlot No. 50--Chiaochi (northeastern Taiwan), No. 3--Shiaokotou (northern Taiwan) were both very good in most test plantations. (On the other hand, some seedlots grew poorly in all test plantations. Interestingly, seedlots from Yilan and Chiaochi grew at very different rates even though the parent stands are neighbors. The results of the analysis of variance for Mrs. .321!" 106 Tsou's experiment are shown in Table 28. She made her cal- culations in terms of deviations from plantation means, so did not calculate plantation mean squares (which undoubt~ edly would have been significant). There were large and significant differences among the offspring of different trees in the same stand. TABLE 27 SURVIVAL AND HEIGHT OF 2- TO 4-YEAR—OLD LUCHU PINE PROVENANCE TESTS IN THREE PARTS OF TAIWAN, WHEN MEASURED IN 1975 (from Tsou, 1978) Location of Age when Number of Height Survival Plantation measured plantations (cm) (%) Northern 2 1 207 69 Taiwan 3 3 272 84 4 l 381 80 Central 2 3 248 -- Taiwan 3 3 350 90 4 1 488 95 Southern 2 1 --- 70 Taiwan 3 3 296 83 4 1 404 92 This experiment of Mrs. Tsou's is important because it gives needed information on luchu pine. Also, it is a model of a combined progeny-provenance test which can be followed in other Taiwan tree improvement projects. With this model, it is possible to locate specific small areas having the best families. To calculate the ANOVA of such an experiment is sometimes complicated, especially when the 107 TABLE 28 ANALYSIS OF VARIANCE IN GROWTH OF PINUS LUCHUENSIS (from Tsou, 1978) Source of Degree of Mean F variation freedom square value Seedlot 96 254.9 Seedlot between stands 14 1059.3 9.01** Seedlot within stand 82 117.6 6.92** Seedlot x plantation 452 41.7 2.45** Replication within plantation 27 117.8 Error 610 17.0 **Statistically significant at 1%‘Ievel. experiment consists of more than 200 seedlots, 20 replica- tions per plantation and 17 plantations with 10-25% mortal- ity. In such test plantations, irregularities and missing plots are inevitable. To calculate the ANOVA for combined progeny-prove- nance experiment, we generally use "nested analysis of variance." That is a technique new to Taiwan but not to the world. I (Yao, 1975) discussed the characteristics of nested analysis of variance and its applications to tree breeding work, and gave examples to show how to calculate the ANOVA including the expected mean squares (EMS). For missing plots or irregular experiments, Wright (1978) suggested working with deviations from plantation means. I (Yao, 1979b) also followed the method to calcu- late ANOVA with some examples in Taiwan. This approach is not only to handle the missing plots per se, but it is ef- ficient to calculate the ANOVA, especially when several 108 test plantations are combined for analysis. Mrs. Tsou's paper provides a detailed model for the application of this method to practical problems. Mrs. Tsou selected most of her parent trees for their superiority in growth rate, form, or both. However, in several stands she chose some parents with poor form or poor growth rate, to serve as controls. The results for eight such pairs of plus and minus trees are presented in Table 29. In five of the eight pairs, the offspring of the excellent parent (plus tree) grew faster than the offspring of the poor parent (minus tree). In the other three pairs, the plus family was shorter than or the same as the minus family. These results indicate that the plus tree selec- tion may have resulted in slight genetic improvement in growth rate but that progeny testing can result in even more. For example, the five best families, as measured by the results of the progeny test, grew 22, 29, 31, 38 and 41% faster than average, as opposed to a 2% gain from the plus tree selection alone. This was a combined provenance and progeny test. Provenance are often undertaken to obtain data on natural selection and natural variation patterns. That was not possible for luchu pine, a recently introduced species. However, the experiment showed that trees from certain plantations grew very well and that others grew very poor 1y. Therefore, the provenance test was a success in show- ing which stands yield good seed. 109 TABLE 29 RELATIVE GROWTH RATES OF THE OFFSPRING OF EIGHT SELECTED PLUS TREES AND EIGHT SELECTED MINUS TREES CHOSEN AS CONTROLS Relative height of offSpring of Selected plus trees Selected minus trees Parental stand (% of mean) Shiaokotou 112 102 Chiaochi 101 103 Kaoshifu 94 93 Pinglin 98 98 Yangminshan 110 102 Yangminshan 105 104 Chushan 104 101 Chushan 101 101 Average 102.9 100.5 Work Plan Luchu pine used to be a very important economic species; it is generally fast-growing, non-selective as to site requirements, and easy to establish. These qualities made it planted in huge lower hilly stands all over the land. The main use of it is for pulp and mining logs, etc. Today, the planting areas are much less than before, so the tree improvement potentials seem dismal. But the fact is that the major use of the wood does not change, and the amount of wood needed for pulp is dramatically increasing. And in Taiwan, so far, no other species can take the place of the luchu pine. So, in the long run, unless we are to import more wood for pulp, much more luchu pine plantings 110 are expected, because the land is available and so is the market. Therefore, this luchu pine genetic improvement plan is justified. Furthermore, since the progeny testing by E. K. Tsou has already started, more improvement works can be done fast just following the progress of the test- ing. In the next 10 years, more than 3,000 ha can be ex- pected to be planted with luchu pine. The traits to be investigated will be growth rate, stem form and fiber length. The plan will be as follows. Follow-up, 1969 TFB Provenance-Progeny Trial Tsou's progeny tests have shown positive results and should be maintained as part of an improvement program. And further considerations are as follows: 1. Remeasure in 1983 the height, stem form, and count cones per tree to measure stem form in cm, measure departure from a vertical line, or in number of crooks per stem. 2. Publish the results in 1984. The reports should include seed source information, differences in data of measured characteristics, ANOVA tables, compari- sons between controls and selected trees. The test plan- tations site data should show by location map and by brief descriptions. Growth data table must be concise and read- able. Table 4 in Tsou's 1978 report (results in height) could be much shortened, and the rest of the tables except Table 5 (ANOVA results) could be taken off or changed to 111 simpler forms. 3. In plantations Sanwen-C, Sanwen-S, Yuchi—N, Yuchi-C, Liuchi-N, and Liuchi-C, select 15 ha areas with good survival and convert to seed orchards. To do this, thin by family and then cut the poorest trees in the best families. (a) For the next 10 years of 3,000 ha plantings, 60 liters of seeds are needed each year. (b) The seed or- chard will be 15 ha in total, each having 100 trees per ha at an average spacing of 10 x 10 m, and approximately pro- ducing 5 liters seed per ha per year, then it will meet the need. (c) Improved seeds from seed orchard will be enough for future plantings. So establishment of seed production areas of the parent-tree stands will not be necessary. 4. First thinning should take place in winter of 1983, and one-half of the poor families will be thinned out. 5. Remeasure the trees in 1987 and then do the second thinning, and thin the remaining half. In 1990, thin again and leave 10 families. 6. For immediate plantings, use seed for Chiaochi, Shiaokotou, and Yangminshan. Also, use the seed from thin- ned seed orchards. 7. Start second generation improvement by con- trolled breeding in the seed orchards in 1990. CHAPTER 11 SUB-TROPICAL PINES INTRODUCTION STUDY Sub-trOpical pines are an important group of trees. They have been intensively studied in many countries of the world. They are generally fast-growing, genetically flexible, and adaptable to various sites. The wood can be used as pulp, furniture, power line poles, fuel wood, construction, etc. In Taiwan since the 19005, more than 300 tree spe- cies have been introduced (Liu et al., 1967). The list in- cludes several exotic pines (Kiang, 1970). Today, more than 80% of the coniferous plantations are of exotic spe- cies. So, introducing new species for planting in Taiwan is one of the important steps for improving Taiwan refores- tation programs. The "Sub-tropical Pines Introduction Study" is therefore of interest in tree improvement work. The main purposes of this study were as follows: (a) To introduce some potentially useful exotic pines for future planting in Taiwan, and (b) To supply breeding stock which could be used in future attempts at species hybridization. Past Work As stated before, introduction activities in the past were numerous. A few exotic pines are still left 112 113 growing in many parts of Taiwan. Almost all such plantings are in small patches and are without records. The success- es but not the failures can be demonstrated. Kuo and Yao (1972) discussed the selection of ex- otic pine species for testing in Taiwan. They surveyed commonly planted pine species in other sub-tropical parts of the world and compared the performance of different spe- cies in various countries. In addition, the environmental factors of the given areas were examined in detail. Such data should be the main consideration for the selection decision for Taiwan. Other factors to be considered are wood usage, pests and diseases. Lin (1959, 1965), Hsu (1966), and Hung 22 El. (1976) have reported on pine species adaptability trials in Tai- wan. Their results are summarized in Table 30. Other stu- dues include those of Liang (1969) on turpentine quality, of Hung (1955, 1958) on thinning of slash pine, and of Hung (1952) on luchu pine. In 1973, 25 seedlots representing seven sub-tropi- cal pines from other countries and four pines from Main- land China or Taiwan were planted in Nantou Hsien in the NTU Experimental Forest (Yao, 1974, 1979a, 1981). They were planted in four test plantations, each with eight replications, 2.4 x 3 m spacing and l6-tree plots. The test sites varied in elevation from 820 to 1,200 m, in rainfall from 2,052 to 2,888 mm per year, in mean annual temperature from 170 to 23°C. The test sites vary in loca- 114 TABLE 30 HEIGHT GROWTH OF DIFFERENT EXOTIC PINES AT DIFFERENT AGES AND SITES IN TAIWAN Height (m) at: Species of pine LifikWEi Liukwei Taichung Several Places Age 5 Age 8 Age 12 Age 22 kesiya 4.7 -- -- -- caribaea 4.3 -- -- -- radiata 2.2 -- —- -- pinaster 1.7 -- -- -- rigida .6 -- -- -- elliottii -- 6.1 8.1 15.9 taeda -— 5.4 8.5 -- echinata -- 4.6 -- -- luchuensis 4.3 6.1 9.0 11.8 massoniana 3.2 4.8 -- 17.6 taiwanensis -- 5.4 -- —- morrisonicola -- 4.1 -- -- Authority Lin Lin Hsu Hung 33 31. 1959 1965 1966 1976 tion from 120°4o' to 120°5o' E and from 23°35' to 23°48' N. Generally, the test sites are subétropical, warm and humid, with rainfall mostly in the summer. The test plantations were measured at age 7. None of the Taiwan or Chinese pines grew faster than average. Pinus kesiya from the Philippines but not from Thailand grew 10% or more faster than average. Pinus caribaea var. hondurensis grew well, much better than the other two va- rieties. Pinus oocarpa seems promising, but most of the Pinus merkusii died out. Slash pine (Pinus elliottii), which at one time was considered as an excellent possibili- ty for Taiwan, seems to have been overrated, as it grew 115 less well than several others, including 3. taiwanensis. The results are given in Table 31. TABLE 31 RELATIVE HEIGHTS AT AGE 7 OF SUB-TROPICAL PINE SPECIES TESTED AT FOUR LOCATIONS (from Yao, 1981) Pinus species Relative height at Place of origin Chi-Hsien-Pin— tou jgin lin Hoshe Mean kesiya kesiya kesiya kesiya kesiya kesiya kesiya kesiya kesiya caribaea var. hondurensis bahamensis caribaea oocarpa merkusii patula elliottii elliottii taiwanensis massoniana armandii morrisonicola (% of mean) Zambales, Phil. 113 121 106 102 111 Abra, Phil. 117 116 107 102 111 Bontoc, Phil. 113 124 108 109 114 Pangasinan, Phi1.117 120 110 103 113 Abra, Phil. 125 125 110 114 119 Benguet, Phil. 122 127 110 105 116 Copperbelt, Zam. 114 111 104 105 109 Chiengmai, Thai. 90 101 91 103 96 Petchabeen, Thai. 92 103 93 86 94 Alamicamba 116 99 '104 104 106 Nicaragua 96 93 106 108 100 Bahamas 84 89 92 98 89 Boneto, Nicarag. 116 95 102 106 105 Chiengmai, Thai. -- 36 -- -- -- Malawi 128 100 97 109 109 Queensland 89 89 90 90 90 Rep. S. Africa 85 88 87 93 88 Taiwan v 100 107 94 83 96 Taiwan 77 -- -- -- -- Taiwan 87 100 90 77 89 Taiwan 61 57 -- -- 59 Actual mean, meters 5.7 4.0 4.9 5.6 5.1 116 Work Plan NTU Experimental Forest Trial 1. Manage them as arboreta. Give them good main- tenance on weed control and spacing. Remeasure at 5-year intervals. 2. Start control pollination and hybridization trials among species of subsection Sylvestres, such as P. massoniana, P. taiwanensis, P. luchuensis, P, merkusii, P. kesiya as soon as they are flowering. Prepare to harvest the seed whenever they begin to fruit. Second Round Introduction Trial 1. Cooperate with IUFRO for getting the seed. Make an introduction plan of 50 seedlots, based on the re- sults of NTU Experimental Forest trial in "A," emphasizing the species or varieties doing best in that trial. 2. Emphasize on provenance trial on P. kesiya, P. oocarpa var. echeternal, P, caribaea var. hondurensis. Include also g. cubensis, P. tabulaeformis, P, yunannensis. Give 3 x 4 m in spacing in favor of flowering. Give 20 trees per plot to promote the pollen exchange within the plot. Do better in weed control, maintain good survival on each seedlot. Establish four test plantations in dif- ferent sites for comparisons. Collect most subtropical and tropical pine species in the world and establish pine Species arboretum in Chushan. CHAPTER 12 PAULOWNIA AND ITS IMPROVEMENT General Description and Distribution Taiwan paulownia (Paulownia taiwaniana Hu et Chang) of the Scrophulariaceae is a very fast-growing hardwood native to low elevations. It is also a typical cash tree species. It used to be called the green gold of Taiwan. It used to be planted on private forest land and land of the hsiens (counties). Because its wood is capable of various uses in Japan, the price is much higher than for other tree species. Therefore, Taiwan became the main source of paulownia exports to Japan. The rotation age is relatively very short, 6-8 years. There are about 1,900 ha of paulownia plantations in Taiwan. About 65% of them are in eastern Taiwan, including Hwalien and Taitung. Taiwan paulownia is native of Taiwan. Because it is accessible and the wood is so valuable, most of the natural stands are gone. In addition to g. taiwaniana, Taiwan has two other native species, 3. kawakamii and P. fortunei. All are dis- tributed along the central mountain range between 500 and 1,500 m in elevation. There are also 6-7 species native to Japan and 4-7 native to Mainland China and Korea. 117 118 The wood of Taiwan paulownia is light in weight and color, with little difference between heartwood and sap- wood. It is resistant to moisture, water and fire. The specific gravity is .28 to .32. Because it is light, soft and easily worked, it has a wide range of uses such as musical instruments, furniture, wooden boxes, clogs, medi- cine containers, and gift containers. Especially in Japan, it is considered a high quality raw material for furniture. Paulownia is a large deciduous tree, up to 20 m tall. The leaves are large, 20 x 20 cm, heart-shaped, and slightly hairy. The fruits are egg-shaped capsules, 4 cm long, containing a large number of very small, winged seeds. The fruits mature in October and November. There are about 47 grams or 240,000 seeds per liter (1.5 million seeds per 1b.). The fresh seeds have 50-60% germination rate, which decreases very rapidly with storage at room temperature. When stored at 5°C, it will retain germination capacity for 3 years. In planting programs, generally root cuttings are used. Cuttings made from the roots of l-2-year-old seed- lings are best. They should be 10-15 cm long and 1.5+ cm in diameter. The cuttings are buried directly in the plant- ing site in the spring. Average survival is 55-80%. The trees will be 1.5-2 m tall after one year's growth. In the Spring of the second year, the tops Should be cut off to permit the root system to send up a sprout. This pro- duces a stronger root system than if not cut back. 119 With intensive silviculture, fast growth is expect- ed. So the density is not less than 1,000 trees per ha. The best density is 500 trees per ha. With the denser spacing, thinning will be done in 3-4 years, when up to 50% of the trees should be removed. The final cutting gen- erally should not exceed age 15, but this depends on the market demands and on the management purposes. According to growth data supplied by Rin (1979), 4-year-old trees reach 36 cm diameter. In a 6-year-old plantation, trees were 14.1 m tall and 23 cm in diameter. In an 8-year-old plantation having 800 stems per ha, the average size was 13.4 m in height and 21 cm diameter. The above grown dry weight was 28.3 tons per ha and the volume 3 3 was 129.1 m per ha, for an average growth rate of 16 m per ha per year. Today, Taiwan paulownias face a severe disease prob- lem. That is the witch's broom disease caused by a myco- plasma. The disease began about 1970 and spread rapidly until 1975, when it occurred in almost all parts of Taiwan. So far, there is no way to control the disease. The dis- ease does not cause death, but it seriously decreases growth. As soon as a plantation is infected, the general practice is to cut, burn and switch to another crop. Previous Work on Genetics Taiwan paulownia is a very intensively cultivated Species. Therefore, most studies emphasized silvicultural 0“. 120 aspects. Some also relate to its genetics. Rin (1979) made a comprehensive Silvicultural study. He gave detailed descriptions on all Species in the world and a complete bibliography of Taiwan and Japan. Perng (1978) studied its success and failure in plantations. He mentioned that "nearly all the planting materials came from Hualien area. Perhaps they cannot grow well in other sites." Lee (1977) listed the research activities on Taiwan paulownia in TFRI. He gave only brief descriptions of each research project without details, results or references. In the genetics and tree improvement aspects, they are as follows: the selection and breeding for paulownia improvement started in 1973 to 1976 in cooperation with TFB. Tissue culture study in 1974 in cooperation with Department of Botany of NTU. Exotic species introduction trial in 1974 in coopera- tion with TFB. Hybridization trials on g. taiwaniana x P. fortunei started in 1974, in cooperation with TFB. Other experiments started were a provenance trial (1975), clone bank (1977), isozymes of three species (1975), and nematode resistance. Work Plan Witch's Broom Disease Resistance Trial Lots of efforts have been done on paulownia in silvicultural aspects, such as improving the growth rate or wood quality by fertilization or pruning. Little has been done on witch's broom disease control. As a matter 121 of fact, this disease is a limiting factor for paulownia culture. If it is out of control, there will be no more paulownia plantings. A. Progeny test 1. Select 100 parents of healthy paulownia both in the wild and in plantations, especially make selections on infested populations. 2. Start progeny tests at Hualien, Taitung, and Nantou Hsien. 3. Establish test plantations where the witch's brooms are. Each plantation should include 10 replications and 4-tree plots. 4. Measure the growth periodically with emphasis on witch's broom disease. 5. Select families with high resistance. B. Paulownia species trial 1. Collect the seed by species (about 10) from their original regions, including Japan and Mainland China (cooperate with TFRI). 2. Raise the seedlings in infested areas in Taitung and Hualien. 3. Establish replicated test plantations. 4. Grow them in the same regions and measure the growth periodically and emphasis on the disease. CHAPTER 13 KADAM INTRODUCTION STUDY General Description and Distribution Kadam (Anthocephalus chinensis) is a non-commercial waste tree of tropical Asia. It has rapid growth, good wood quality, multiple possible uses, and an ability to coppice. Thus, it has become a very promising planting Species in recent years. In Taiwan, it was introduced in 1962, 1964 (Sheng, 1964; Yang, 1964) and again in 1973 (Yao and Yen, 1979). This chapter is a comprehensive discussion on this species including its geographic distribution, ecology, Species description, and growth data both in Tai- wan and in its natural ranges. Finally, a work plan of kadam is made for further testing in Taiwan. Species Description and Distribution, Habitat Anthocephalus chinensis (Lank.) Rich. ex Walp. of the Rubiaceae (=A. cadamba of Troup and others), widely known as kadam, kaatoan, bangkal, kelempayan, or laran, is a rapidly growing evergreen timber tree up to 30 m tall and 87 cm in diameter with a straight axis and widely spreading nearly horizontal branches. Its bark is gray, smooth to slightly fissured. The inner bark is light yellow with a thin green outer layer, fibrous and bitter. Twigs are 122 123 stout, hairless, with ringed nodes, green when young, and become brown later. Leaves are simple, opposite. Flowers are numerous, fragrant, consisting of a tubular base. Fruits are compound with numerous fruits pressed together like a fleshy ball. It occurs in Nepal, India, Burma, Thailand, Malay- sia, the southern Philippines, and New Guinea (Whitmore, 1975: Troup, 1921; Fox, 1971; Monsalud and Lopez, 1967). In India, kadam is found in the sub-Himalayan tract from Nepal eastward in Bengal and Assam. It occurs in the low country of Ceylon (now Sri Lanka) up to 2,000 feet alti- tude and in all parts of Borneo, Sumatra and Java. In the Philippines, it is found in Bukidnon, Cotabato, Davas, Basilan, Zamboanya, and Mindanao. Kadam is a light-demanding tree which colonizes bare areas. It is a tree of moist, warm regions, often occurring on alluvial ground along rivers, and also in swamps. On stiff, badly-drained ground, growth is poor. In its natural habitat in India, the absolute maximum Shade temperature varies from 360 to 43°C, the minimum from 30 to 6°C, and the normal rainfall from 1,524 to 5,080 mm or more (Troup, 1921). In Sarawak and Brunei, on the island of Borneo, it occurs mostly at low elevations along river banks (Ohtani et al., 1962). According to Ashton (1964), it is also common on day lithosal soils on hill ridges. In Sabah, it is abundant in the logged-over areas of the low land depterocarp forest, and is one of the few species 124 able to thrive on soils disturbed by heavy machinery. Of- ten it dominates such areas later. It is also found in the virgin forest and forms a constituent of freshwater swamp forests. In secondary forest at higher elevations, it is confined usually to streamside sites up to about 305 m elevation. For the most part, the soils on which it is found in Sabah are c1ayscn:clay-loams (Fox, 1971). It is generally distributed below 1,000 in elevation in primary forests. Wyatt-Smith (1965) gives to Anthocephalus/Eugenia densiflora var. angustifolia riverside association, the rank of a sub-seral community-~"Riparian Fringe." This community has a secondary or invasive status, found on ri- ver banks subject to flooding and on newly formed accreting soil on the inside of river bends. Also, it is reported (Gyekis, 1966) that kadam is present in unproductive swamp forest. Kadam wood is light in color, of fine and even tex- ture and workable. The timber which has no apparent heart- wood is soft and light. According to different authors, the specific gravity varies from .38 to .53, and fiber length from 1.50 to 1.67 mm (Liu 95 §_1_., 1976). The wood is generally free of odor or taste. It is not suitable for structure use because of its softness and low durability in the open when untreated. It is ra- pidly attacked by termites, lasting only 8 months in the Philippines. The timber saws and splits well and seasons rapidly with little structural degradation, despite high 125 shrinkage. It has long been known as potential pulp wood. Among other promising commercial uses are plywood, wooden clogs, matchsticks, pencils, and wood carving. Fruiting, Growth and Regeneration Kadam fruits early. It flowers after 4 or 5 years, or when the trees are 12-28 m tall. In the Philippines, annual flowering occurs after age 7. Some trees flowered at age 2 in Taiwan. It flowers from April to June, the fruits ripening from September through February. As soon as ripe, they fall to the ground. In Taiwan, flowers ap- pear in April and May, the fruits ripen in November and December, and its fruits annual after age 10. Seed quali- ty is very good. Seed dispersal is accomplished by bats and birds which feed on the fruits, and also by cattle and other animals, and by water (Troup, 1921; Fox, 1971; Whit- more, 1975). Fox (1971) believes that on exposed soil, it arises from bat drOppings subsequently slightly dispersed over the surface by heavy rain. The seeds are minute, about .63 x .47 mm in size. There are approximately 17,000 seeds per gram, about 10,000 seeds per multiple fruit. The germination rate generally is high. In the Philippines, it is reported 95 percent after two months refrigerator storage, compared with 5 percent with fresh seed. Fox (1971) reports in Sabah it germinates most readily with seed which has been stored for 6-12 months. The air-dry seeds, if kept in airtight 126 bottles and stored in a refrigerator, can retain validity for 2 years (Monsalud and Lopez, 1967). Germination takes about 3 weeks. The young seed- lings are tiny, and growth at first is Slow. After it grows to 5 to 10 inches after the first 3 months, it reach- es 2 m in the next season. Extensive plantings began in 1933 in Java by the Dutch, then more in east Borneo later by the taungya system (Whitmore, 1975). It has been suc- cessful in trail plots in Puerto Rico, and was recommended for west Africa (Puerto Rico, 1963; Lamb, 1966). It is highly recommended in the Philippines. In Taiwan, small- scale-trials have been started. The growth in different areas is summarized as in Table 32. TABLE 32 GROWTH OF KADAM IN OTHER COUNTRIES Country Age H7;?ht ?Zg) Author India 4 9 19 Troup, 1921 India 22 37 61 Troup, 1921 Philippines 9 18 25 Masalud & Lopez, 1967' Philippines 15 23 41 Masalud & Lopez, 1967 Sabah 4 -- 13 Fox, 1968 Sabah 8 -- 14 Fox, 1968 Sabah ll -- 18 Fox, 1968 Sabah 30 38 65 Whitmore, 1975 Puerto Rico 2 4 -- Puerto Rico, 1963 Puerto Rico 4 6 -- Puerto Rico, 1963 Puerto Rico 30 20 65 Puerto Rico, 1963 h~3‘-EL‘ “ 127 It grows very fast in height until 6 to 8 years old. After that, growth slows down and becomes particular- ly slow after age 20. It is naturally self-pruning. Small crowned trees usually die off. For volume growth, Whitmore (1975) predicts in Sabah volumes of 373 m3 per ha at age 30, with the trees 38 m tall and 65 cm in diameter. Kadam can be reproduced by sprouting. In Philip- pines, some young sprout shots 16 months old had 6 m in height, 9 cm in diameter (Monsalud and Lopez, 1967). In Taiwan, it is also very common after the thinnings. So copping regeneration is very promising. Kadam seedlings are sensitive to drought and excess moisture also. Seedlings are much subject to damage by insects, especially during the first few weeks. In dense stands, they are eaten by caterpillars (Arthosclista hi: laralis) (Mastan, 1969). Fox (1971) noted damages from three other insects. Deer may occasionally browse young trees (Hellinga, 1950). Planting Experiments in Taiwan In 1963, 1966, TFRI established two test plantation with seeds sent from Puerto Rico. The first test planta- tion is in Liukwei, located at 120°38' E, 23° 00' N, 250 m above sea level. Annual rainfall is 3,121 mm; raining sea- son is from May to October. The soil is alluvial, a mix- ture of sandy-loam, and gravel. The next one is at Chung- pu, 60 km northwest at an elevation of 180 m above sea 128 level. Annual rainfall is 2,790 mm, heaviest from June through September. The soil is a loam, and deeper and better than at Liukwei. In 1975, two test plantations were established with kadam seeds from the Philippines in NTU Forest. The one Hoshe Tract, 120°53' E, 23°35' N, at an elevation of 830 m. It is on an east-facing 15o slope, formerly in tung trees. The soil is well-drained, sandy loam and gravel mixture. This site receives 2,045 mm rainfall per year, mostly from March to September. Average July and January temperatures are 250 and 16°C, respectively. The planta— tion is a mixture of kadam and three other species. It follows a randomized complete block design with 20 repli- cations. There are 20 trees per plot, each species plant- ed in 4 rows. The spacing is 2.5 x 2 m. The second one is in Shuli about 40 km north of Hoshe, at an elevation of 250 m. Annual rainfall is 2,438 mm. The soil is sandy- 1oam, deep and fertile. This plantation is pure, the fac- ing is 1.8 x 1.8 m totally 0.17 ha and 435 kadam trees. Results In 1975, three trees were cut from each of the Liukwei and Chungpu plantation for growth analysis (Liu gt‘al., 1976). Each of these trees was representative of the dominant, co-dominant and suppressed crown classes. The Liukwei and Chungpu plantations were 11 and 8 years old, respectively, at the time. They remeasured the plan- 129 tations in 1979 at ages 14 and 11. Height, not diameter growth, seemed to Slow down after age 8. The results are Shown in Table 33. TABLE 33 THE GROWTH OF KADAM IN LIUKWEI AND CHUNGPU (from Liu gt al., 1976) Plantation 4 6 8 11 14 Liukwei height (m) 4.3 7.0 9.6 11.3 11.3 diameter (cm) 4 6 9 14 23 Chungpu height (m) 6.4 9.8 13.1 12.5 diameter (cm) 7 12 17 24 The growth in Chungpu was much faster than it is in Liukwei because of the better site conditions. The NTU Forest plantations had different growth rates on different Sites, too. The growth at age 4 in Hoshe and Shuili is shown in Table 34 (Yao and Yen, 1979) . Work Plan In Taiwan, most commercial plantations are coni- fers. Hardwoods are confined to limited Species. This is not because of lack of hardwood species, but because they are generally slow-growing. So introduction of ex- otic, fast-growing hardwood planting in the low elevation areas of Taiwan is an urgent need. Kadam as discussed in 130 the text fits the case. The objective of kadam introduc- tion is to use the wood as light wood industry raw materi— als except the pulps, veneer, medications-use, and gift boxes, furniture, matchsticks, etc. TABLE 34 SIZE AND SURVIVAL AT AGE 4 OF KADAM AND THREE OTHER SPECIES TESTED AT HOSHE AND SHUILI (from Yao and Yen, 1979) Species Height (cm) Diameter (cm) Survival % Kadam 234 5 46 Taiwan alder 580 7 75 Albizia falcata 512 7 42 Tung tree 219 4 62 Shuili Plantation kadam 1151 18 -- Today, Taiwan's lower elevation areas mostly are so-called fringe lands, the land usage is diversified. If the tree is the objective, mostly they are Paulownia, fruit trees, Taiwan acacia or bamboos. Kadam would be a very potential competitor for replacing Paulownia or fruit trees, because the disease problems of them almost become hopeless. Such lands are all over the island, and the area would be very large. For instance, according to 1977 sur- vey, the Paulownia plantations were 19,046 ha (Rin, 1979). So for the Kadam, planting rate should be considerable. But when the experimental results showed the Site was very selective. So, for the beginning of its introduction, the 131 rate would be expected to be 500 ha for the next 5 years. According to the Taiwan test plantation's results, the difference between sites was substantial; therefore, the seed sources importance seems diminishing. Neverthe- less, the wood prOperty shown before indicated that some differences existed, and maybe they are due to seed sour- ces. If so, we have to select the better ones, because the major usage of the wood is for more specific purposes. Two major approaches will follow: A. Site Adaptability Trial 1. Cooperate with hsien governments and coordinate with Council for Agriculture Planning and Development (CAPD) and TFB. 2. Select plantation sites based on site differ- ences, the elevation range will be between 250-800 m. 3. Collect seed from Taiwan kadam trees and raise the seedlings. 4. Plant Paulownia and Taiwan acacia for compari- ' '1 (”.41 son. 5. Measure the growth periodically. 6. Establish the plantation with 3 x 4 m or 3 x 3 m Spacing. 7. The plantation size not less than .5 ha; 1.0 ha each is suggested. 8. When Paulownia and Acacia included 3 replica- tions are needed. 9. Follow-up, tending should be emphasized. 132 B. Provenance Trial 1. Take 20 seedlots and seeds from its natural range, 3-5 grams each seedlot is enough. 2. Establish 5 test plantations (or more) on dif- ferent sites, eastern Taiwan must have one. 3. The traits investigated Should include growth rate and wood quality (Specific gravity, fiber length, etc.). CHAPTER 14 TREE IMPROVEMENT WORK THROUGH SILVICULTURE Tree improvement work in Taiwan is built on a mass— planting base. Intensive silviculture system will make the tree improvement significant and important. Poor silvicul- tural practice could lessen the effect. In fact, the im- provement practice should be considered as the products of well-develOped and managed silvicultural system. Each year, more than 90 million (30,000 x 3,000) seedlings are planted. That gives the potential of the tree improvement. In Taiwan, the relationship between silviculture and tree improvement is much correlated, mainly because of several following reasons: (1) environmental factors, (2) replanting systems, (3) social factors, and (4) forestry policies. Due to Taiwan's geographic nature and its location, the environmental impacts are tremendous. Rugged topogra- phy and fragile, thin soil structure are the major features which form a poor combination to affect the environment. The climatic factors are even more fierce. Abundant and unevenly distributed rainfall, high humidity, and generally warm temperatures prevail. These climatic factors help the trees grow, but also hasten the grass growth. The weeds 133 wnnv 4. B tharad. I. If. m 28 ' «a 134 become strong competitors. So survival rates are generally low and there seems no room for tree improvement to func- tion. Even if the survival were improved by some Silvi- cultural techniques, tree growth still will be suppressed by heavy weeds in the plantations. If so, the superior genetic quality may be suppressed. It is common in a large plantation before the first weed control is finished, the second weed control must be started. For such conditions, tree improvement can function only when plantations are well managed. In Taiwan, labor used to be cheap. But in recent years, labor sources for forestry are scarce. The wage is getting higher. On the other hand, the planting season is short. When the season is due, there are not enough labor- ers around. This is also true in weeding and other acti- vities. Silviculture and tree improvement work are both long-range businesses. Credits and gains are cumulative in terms of the time factor. Therefore, a precise, clear and stable forestry policy is needed to ensure that every effort is worthwhile. Changes in policy which lead to changes in Species being planted mean that everything must start again. In Taiwan, the planting Species are numerous. To face a multi-choice Situation, it is hard for tree im- provement work to follow the policy change. For instance, seed orchards generally need 20 to 30 years to produce im- proved seed and may become out-of-date if a species is dis- : .. 135 carded. Most Taiwan forest lands are public. Profit is not the main interest. Thus, policy changes without strong reasons are possible. For the most part, Silviculture is a leading force, of which tree improvement work is a part. Occasionally, tree improvement becomes the leading factor as with suc- cessful tree introduction. Also, strong tree breeding re- search programs also make the reverse trend possible. That I is true when some hybrids are made, or publishing abroad ‘ makes scientific fame. Due to Taiwan's variable sites, the planting areas for any one Species are relatively limited. Under such circumstances, very progressive and intensive tree improve ment programs for few Species seem not possible. However, when certain basic principles in tree improvement work can be accomplished through the silviculture in any Species we are dealing with: 1. Better survival means more gains. 10% less mortality = 10% more gain. 2. Seedling quality control is the first step in tree improvement work. Poor, unhealthy seedlings always cause stagnant growth. 3. Seed sources must be controlled. And this must start from the person who is in charge of the seed procurement or seed collection. 4. Seed sources must be labeled whenever seed- lings from them are planted. 136 5. Tree improvement starts from improved seed. 6. In Taiwan, denser plantings are favored. In- tensive thinning schedules should be followed. Maximum growth rate by individual trees can be reached only through appropriate Spacing. 7. Site selection is a prerequisite for a tree improvement program for every species. Even a superior strain will grow poorly unless a suitable site is met. 8. Whenever possible, a long-term improvement plan should be included in the management plan. 9. Silvicultural improvements such as tending, pruning, vine cutting, thinning, weeding must be adequate to make genetic improvement worthwhile. 10. In NTU Experimental Forest, tree improvement experiment plantations always are better in survival or growth and management than commercial ones. This is also true in other institutions such as TFB or TFRI. It proves two things. Tree improvement works help growth; second, experimental plantations are in good Shape because of more careful management. Expand the idea and treat every plan- tation as an experiment. 11. Large-scale, good test plantations are few but there are many small-scale experiments in different in- stitutions. Coordination among them lags. Teamwork must be emphasized. It is a powerful tool, more efficient and cost-saving and the outcome would be more fruitful. CHAPTER 15 CONCLUSIONS Taiwan is an offshore island of Mainland China. Two-thirds of the land is in mountain ranges and covered by forests. It is located in a sub-tropical region with heavy rainfall. Strong typhoons are generally common. Forests become not only economically but also ecologically important. Every year, more than 30,000 ha of forest land are replanted to economic species. Intensive planting pro- grams and clear cutting--estab1ishment of even-aged manmade forest--is the main Silvicultural system in which we are engaged. Tree improvement and tree breeding works are playing a main role in it. Only strong and intensive se- lection makes the progress and the gain. The effects we are talking about are improvements in growth rate, pest resistance and wood quality. Mass planting programs are essential. Tree im- provement work would give little gain if only small, limit- ed areas or a few ha each year are planted. Thus, in this research, only potential and/or large-area planting species are mentioned. Tree improvement work in Taiwan since the 1960s has been well activated. For more than 20 years, a variety 137 138 of experiments have been established, and the accomplish- ments are substantial. These facts are generally included in the text. But when we look carefully at the details, generally we found the works not adequate to meet the need in terms of practical sense. It is quite evident that most of the projects lacked solid goals, with ambiguous objectives and limited gain. A real and true tree improvement plan should be comprehensive and also it must be on a long-term base. Some would argue and say that may not be so because basic research itself is important. No one can deny that. But here come the key questions: Who are the sponsors of the research projects, and who will be willing to support the research? In other words, several factors are involved in the decision. Among the key parts of research planning are in justification, objectives, literature review, meth- ods, costs, time schedule and review. To meet those re- quirements, each plan Should have clear facts. In Taiwan, the planting species which I mentioned before are numerous and it is difficult to decide species -' 2.5 ‘ ‘v 43.14'13 priority. Even cryptomeria faces controversy on its im— portance. In fact, criteria on measuring the importance L is not the objective of my research. More important here is to emphasize the methodology for a given species. Gen- erally, I take for granted some important Species by their planting areas. For other species this is not true, but 139 I would rather sacrifice some importance than the method- ology. In this research, I presented nine work plans for nine species or groups of species--seven in conifers and two in hardwoods. They are cryptomeria (Cryptomeria jg: ponica), China-fir and luanta-fir (Cunninghamia lanceolata, g. konishii), taiwania (Taiwania cryptomerioides), Taiwan red and yellow cypress (Chamaecyparis formosensis, 9. ob: tusa var. formosana), Taiwan red pine (Pinus taiwanensis), l luchu pine (P. luchuensis), sub-tropical pines (Pinus tai- wanensis, luchu pine (P. luchuensis), sub-tropical pines (P. kesiya, P. caribaea, g. oocarpa, g. merkusii, P. elli- ottii, P. pinaster, etc.), paulownia (Paulownia fortunei, P. kawakamii, P. taiwaniana), and kadam (Anthocephalus chinensis). In each individual plan, general description and past work related to genetics are the first part. That is followed by the work plan itself. The work plans follow a chronological order, and are organized by independent pro- jects. The construction of each plan is based on the sup- porting information in the first part. For the next 10- ll’,J\-* oh 1‘: .._.—._‘__ year period, each work plan includes a provision for good quality commercial seed to be used before actually improved seed is available. Progeny testing is generally a feature of each work plan, especially half-sib progeny testing. That is not meant to be the only method used, but is practical for the 140 first generation. Moderate selection and screening pro- cess (200 to 400 families per best on the average) are the general strategies. Seed orchards converted from progeny tesns are the final goal of many work plans. The size of the seed orchard is well demonstrated by careful calcula- tion from the seed demand in sequential procedures. For endangered species such as taiwania and luanta-fir, some conservation ideas are also included. These include germ- plasm banks and arboreta. But in this case, more emphasis is on "planting more" concept rather than a mere garden per ES! so some silvicultural approaches are involved. Intensive as progeny-provenance tests are already underway in Taiwan red pine and luchu pine. For these two species, the work is well done. So in this case, no more new testings are suggested. Only follow-up is given. This policy of follow—up is also applied to some other species and in other established experiments which are worthy of being continued. More important, I put estab- lished experiments into the work plan mainly because I considered that this package of plans is for Taiwan in "H .‘d. I :R" “Isl-ful whole, so it must be comprehensive. Cost iS hardly predictable and almost impossible #- to be well worked out in detail in the work plan. But I tried to be reasonable whenever dealing with a species. In the case of Taiwan red cypress, I did try to demonstrate the workload and cost analyses for TFB (Taiwan Forestry Bureau) in the work plan. The purpose is to Show adminis- 141 trators that tree improvement work is actually only a small part of this workload, with small costs and small reasons being no reason for them to neglect or reject them. Besides the species-by-species work plans, there are several chapters dealing with background information such as forestry history, administration and school system,‘ training programs, funding, NTU Experimental Forest where I work, and my own personal involvement. In the last chapter, I tried to compare the rela- tionship between Silviculture and tree improvement work. As a matter of fact, they are two facets of forest manage- ment and are woven together. Some improvements in forest growth may be made by genetic means, others by pruning, thinning, etc. The work plan of tree improvement for Taiwan which I prOpose here involves foresters in Taiwan and foreign countries. It should be treated as a team work plan, justifiable and also feasible. B IBLIOGRAPHY ~15 BIBLIOGRAPHY Anon. 1977. Progeny test of plus trees of Cunninghamia lanceolata and heritability estimates. Acta Gen- etica Sinicae 4(2): 152-158 (Chin.), also Forestry Abst. 1980, 41(3): 1532, 128-129. . 1981. Studies on the classification for plant- ing area of Chinese-fir. Scientia Silvae Sinicae 17(1): 37-45 (Chin.). Ashton, P. S. 1964. Ecological studies in the mixed dip- terocarp forests of Brunei State. Oxford Forestry Memo. No. 25. Chang, L. M. 1950. Tree life cycle observations in botani- cal garden, Taipei. Taiwan Forestry Res. Inst. Notes 53: 389-392 (Chin.). . 1969. Abstracts of cryptomeria in Taiwan. Quar. J. Chinese Forestry 2(4): 239—251 (Chin.). Chang, S. S., K. S. wu and S. G. He. 1980. Studies on the productivity of Chinese-fir plantations in Jiangxi. Scientia Silvae Sinicae 16, Supplement: 65-76 (Chin.). Cheng, T. L. 1978. Study on the reforestation of Taiwan red pine and causes of its success and failure. Masters Thesis, Dept. of Forestry, NTU, 107 pp. (Chin.). Chiang, C. H. 1974. Oxygen consumption in seed germina- tion and early growth of different provenances of Taiwania; J. Chin. Agr. Assn. 88: 23-33 (Chin.). . 1975. Controlled pollination of Cunninghamia species. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 116: 329-342 (Chin.). . 1977. Progeny-provenance test of Pinus taiwan- ensis. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 120: 1-20 (Chino). O ' B. J. Yen, Y. N. Yao, C. F. Shih, G. T. Chu' T. Kiang and Y. N. Wang. 1979. Provenance test 142 4w¢mm—z 143 of cryptomeria. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 123: 1-22 (Chin.). Chiang, S. S. 1956. Growth comparisons on three pine spe- cies in Chiaochi (P. luchuensis, P. massoniana, E. thunbergii). Taiwan Forests 2(1): 22-34 (Chin.). Ching, K. K. and F. S. Kung. 1965a. Tree breeding. Nat'l Taiwan Univ. Exp. Forest Serial Publ. 32, 43 pp. (Chin.). Chu, J. S. (ed.). 1975. Symposium on Taiwan red cypress-- Proceedings. Taiwan Forestry J. 1(13): 3-37 (Chin.). Chueh, K. Y. 1980. The preliminary tissue study on Chin- ese-fir. Scientia Silvae Sinicae 16, Supplement: 137-146 (Chin.). Fang, Y. K. 1968. Effects of grafting on the growth and flowering habits of luchu pine with scions of dif- ferent tree ages. Quar. J. Chin. Forestry 1(1): 174-183 (Chin.). . 1969. Experiment on the influence of cold stor- age of scion and the proper time of grafting for luchu pines. Quar. J. Chin. Forestry 2(4): 134- 147 (Chin.). Fox, J. E. D. 1971. Anthocephalus chinensis, the laran tree of Sabah. Econ. Bot. 25: 221-233. Goo, M. 1971. On flowering cuttings of Cr tomeria japo- nica. J. Jap. Forestry Soc. 53(6): 165-I69. Gyekis, K. D. 1966. Anthocephalus cadamba, a fast-growing industrial tree species for the tropics. Turrialba 17: 321-329. Hellinga, G. 1950. Aan plant scorten 0p bedrijesgrootte. Tectona 40: 196-199. Ho, C. P. 1976. The taiwania planting experiments with branch-pruned, bare-rooted seedlings. Quar. J. Chin. Forestry 9(1): 23-28 (Chin.). Ho, F. C. 1968. Investigation on flowering and fruiting period and the flower and fruit colors in Kentin Park, Henchien. Annual Sci. Bull., Taiwan Provin- cial Museum 11: 84-107 (Chin.). Hsu, H. Y. 1966. Growth comparison of Pinus taeda, P. elliottii, and P. luchuensis in Tatung District. 144 Taiwan Forestry Res. Inst. Notes 222: 2022-2023 (Chin.). Hu, T. W., C. M. Lu, and Y. L. Chung. 1981. Forest genet- Huang, Hung, ic improvement work in Taiwan. Proc. Symp. Forest Genetics and Breeding, Dec. 17, 1981, Taichung: 82-96. ., and B. Y. Yang. 1972. Variation in wood spe- cific gravity and fiber length of Pinus taiwanen- sis, U.S.-China c00perative science program semi- nar on forest ecology and genetics, Mar. 31-Apr. 8, Nat'l Sci. Coun. 1972, 5 pp. C. W. 1952. Study on the growth and the management of the standing crops of luchu pines. Taiwan For- estry Res. Inst. Reference 11, 50 pp. (Chin.). L. P. 1955. Study on slash pine young plantation growth. Taiwan Forestry Res. Inst. Bull. 45, 2 pp. (Chin.). . 1968. Study on thinning and tending effects of slash pine. Taiwan Forestry Res. Inst. Bull. 160, 62 pp. (Chin.). . 1974. Study on the stand growth of taiwania plantations. Taiwan Forestry Res. Inst. Bull 236, 26 pp. (Chin.). . 1977. Study on the effect of thinning experi- ments on luanta-fir plantations in Lianhwachih Forest District (I). Quar. J. Chin. Forestry 10 (1): 61-94 (Chin.). ., H. H. Lo and L. N. Lo-cho. 1976. Study on the growth condition of plantations of three exotic pines planted on different districts in Taiwan (P. elliottii, P. taeda, P. luchuensis). Quar. J. Chin. Forestry 9(1): 29-42 (Chin.). Hwang, C. W. 1952. Studies on the growth and the manage- ment of the standing crops of luchu pines. Taiwan Forestry Res. Inst. Reference 11, 50 pp. (Chin.). . 1968. Construction study of yield table of Hwang, luchu pine in Taiwan. Taiwan Forestry Res. Inst. Bull. 173, 173 pp. (Chin.). K. K. 1969. Studies on the growth of Taiwan red pine stand of various habitats. Taiwan Forestry Res. Inst. Bull. 177, 36 pp. (Chin.). 145 . 1970. The compilation of the yield and volume table of Taiwan red pine stands. Taiwan Forestry Res. Inst. Bull. 193, 43 pp. (Chin.). . 1977. Studies on plantation growth of Taiwan red cypress. Quar. J. Chin. Forestry 10(2): 95- 109 (Chin.). Hwang, S. G. 1978. Effects of different elevations on seedling growth of luanta-fir provenances. Taiwan Forestry Res. Inst. Bull. 303, 12 pp. (Chin.). ., Z. Y. Kung and D. C. Tsai. 1979. Investigation control of squirrel damage in Liukwei. Taiwan Forestry Res. Inst. Bull. 318, 17 pp. (Chin.). E ., J. S. Yang. 1978. The effect of phenotype selection on the survival and growth of the seed- lings of Cunninghamia lanceolata, Taiwania or to- merioides. Taiwan Forestry Res. Inst. Bull. £08, 9 pp. (Chin.). a Kiang, T. 1970. Introduction and contribution of exotic E species in Taiwan. Quar. J. Chin. Forestry 3(3): Koyama, R. 1967. An application of nuclear and cytoplasmic polyhedrosis virus against Dendrolimus punctatus in Taiwan. Quar. J. Chin. Forestry 1(1): 25-30. Ku, E. T. 1967. Silviculture practice of cryptomeria. Quar. J. Chin. Forestry 3(2): 149-154 (Chin.). Kung, C. M. 1970a. Effects of different storing agents of the pollen grains of China-fir and luanta-fir. Quar. J. Chin. Forestry 3(3): 101-111 (Chin.). ‘ h“. . 1970b. Study on conifer flowering inducing by GAS application. Nat'l Taiwan Univ. Exp. Forest Notes 33: 6-10 (Chin.). ., and T. Kiang. 1969. Morphological study on the strobilus of taiwania. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 69, 14 pp. (Chin.). ., T. Kiang and F. S. Kung. 1969. Studies on the storage and the viability of taiwania pollen grains. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 70, 9 pp. Kung, F. S. 1965. Preliminary studies on the development of the strobilus and the storage, viability of the pollen grains of Taiwania. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 45, 5 pp. (Chin.). 146 Kung, F. S. 1973. The establishment of Taiwania clone bank (I)--Affinity Study. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 111, 8 pp. (Chin.). . 1974a. Genetic improvement of Taiwania (I)-- Provenance Study. Taiwan Forestry Res. Inst. Bull. 254, 13 pp. (Chin.). . 1974b. Study on the flowering effect of gibber- ellic acid to the grafts of Taiwania. Taiwan For- estry Bureau Planting Notes 1974: 54-61 (Chin.). ., and C. M. Kung. 1966. Report of the establish- ment of Taiwania seed orchard. Quar. J. Taiwan Forestry 2(4): 6-28 (Chin.). Kuo, P. C. 1957. A preliminary survey of squirrel damage to forest trees in Taiwan. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 12, 20 pp. (Chin.). ., and Y. N. Yao. 1972. Selection of exotic pine 5 Species for testing in Taiwan. Quar J. Chin. For- ’ estry 2(5): 1-8 (Chin.). Kuo, S. R., T. T. Wang, and T. C. Huang. 1972. Karyotype analysis of some Formosan gymnosperms. Taiwania, Nat'l Taiwan Univ. 17(1): 66-80. Lamb, A. F. A. 1966. Report on a visit to Nigeria. Mimeographed, Ministry of Overseas Development, London. Lanner, R. M. 1969. Observations of stem form defects in Taiwan red pine. Appendix II: 1-8. In "Final re- port on silviculture and forest genetics in Taiwan" by E. H. Hinkle, UNDP/FAO Project, 1966-69, Taiwan. ., and E. H. Hinkle. 1970. Some shoot and cone characteristics of Taiwan red pine. Pacific Sci. 14(3): 414-416. Lee, H. T. 1972. Silviculture of cryptomeria and cypress. Quar. J. Chin. Forestry 5(3): 51-53 (Chin.). L. Lee, S. P. 1977. Summary of the studies and experiments of Taiwan paulownia. Taiwan Forestry J. 3(9): 50-53 (Chin.). Li, H. L., and H. Keng. 1954. Icones Gymnospermum Formo- sanarum. Taiwania, Nat'l Taiwan Univ. l(5):25-84. Li, S. J. 1976. Seedling variation of Pinus taiwanensis as related to geographic seed origins. Taiwania, 147 Nat'l Taiwan Univ. 21(2): 138-149. Liang, J. H. 1965. Investigation and study on the yield of turpentine and others of exotic pines in Taiwan. Taiwan Forestry Res. Inst. Bull. 107, 20 pp. (Chin.). Lin, J. F., and S. Z. Tong. 1980. Studies on the stand density control diagram for Cunninghamia lanceo- lata. Scientia Silvae Sinicae 16(4): 241- 251 (Chin.). Lin, T. Y. 1974. Estimation of productivity and yield of luchu pine woodland. Quar. J. Chin. Forestry 7(1): 55-83 (Chin.). Lin, W. C. 1959. Study on the adaptability of exotic pines planted in southern Taiwan (I). Taiwan Forestry Res. Inst. Bull. 62, 34 pp. (Chin.). . 1965. Study on the growth of various species of pine in relation to elevations and climatic factors. Taiwan Forestry Res. Inst. 119, 38 pp. (Chin.). ., and T. Y. Chang. 1975. Effect of shade-screen on the survival and growth of Taiwania seedlings. Taiwan Forestry Res. Inst. Bull. 275, 10 pp. (Chin.). Lin, W. F. 1974. Effect of altitudes and light intensi- ties on growth of seedlings of Taiwania. Quar. J. Chin. Forestry 7(4): 3-12. Liu, J. Y. 1977. The observation of plant phenology. "Forests," Nat'l Taiwan Univ. Forestry Dept. 10: 64-80 (Chin.). Liu, S. C., and F. W. Horng. 1978. Growth and yield of cryptomeria plantation in Taiwan (II). Taiwan Forestry Res. Inst. Bull. 309, 47 pp. (Chin.). ., F. F. Tai and P. T. Wang. 1976. The growth of 1 planted luanta-fir in relation to site conditions. % Taiwan Forestry Res. Inst. Bull. 280, 43 pp. (Chin.). ., ., C. C. Chu, and S. C. Lin. 1976. Growth and wood properties of fast growing timber trees--Anthcephalus cadamba. Quar. J. Chin. Forestry 9(1): 9-16 (Chin.). Liu, S. H., and T. Y. Lin. 1970. Yield tables and volume tables of luchu pine in north Taiwan. Joint Bull. Nat'l Chunghsin Univ. and Taiwan Forestry Bureau, Liu, S. Liu, T., Liu, T. Liu, Y. Lu, C. M. 1973. Variations in seedling characters of Mastan, 148 1970, 50 pp. (Chin.). H., P. L. Yang, H. Y. Shu, and S. F. Chen. 1955. Yield tables of Japanese cryptomeria in Taiwan. Taiwan Forestry Res. Inst. Bull. l, 39 pp. (Chin.). C. M. Lu and T. W. Hu. 1975. Interspecific vari- ation and provenance test in Genus Chamaec aris. Taiwan Forestry Res. Inst. 279, 37 pp. (Chin.). S., J. L. Wang, and J. C. Liao. 1967. The exotic tree species in Taiwan. Quar. J. Taiwan Forestry 3(1): 114-163 (Chin.). C. 1968. Studies on varieties of China-fir: experiments on the transplanting and planting. Quar. J. Chin. Forestry 1(2): 106-125 (Chin.). '- -10... “Jun . 1971. The clone bank and controlled pollina- tion of luanta-fir. Quar. J. Chin. Forestry 5(1): 1-17 (Chin.). . 1973. The management of the clone bank of luanta-fir. Quar. J. Chin. Forestry 6(3): 1-16 (Chin.). . 1974. Studies on varieties of China-fir (2). Quar. J. Chin. Forestry 7(4): 25-40 (Chin.). ., and C. P. Ho. 1973. The Taiwania seedling root pruning study for outplantings. Today's Silvicul- ture, Taiwan Forestry Bureau, 48: 14-17 (Chin.). ., F. Y. Lu, C. C. Tsai, and C. H. Ou. 1979. Provenance test and study of luanta-fir. Quar. J. Chin. Forestry 12(2): 1-9. ., W. C. Shieh, C. P. Ho, and C. S. Hsiao. 1964. Investigation and studies on varieties of China-fir. Joint Commission of Rural Reconstruction and Chung- hsing Univ. Joint Bull. 1964, 64 pp. (Chin.). Taiwania from various provenances (1). Taiwan Forestry Res. Inst. Bull. 246, 18 pp. (Chin.). ., 1975. Variation in seedling characters of Tai- wania from various provenance (II). Taiwan For- estry Res. Inst. 261, 16 pp. (Chin.). K. M. 1969. Prospects of Anthocephalus cadamba as a plantation species. Special subject theSis, CFl, Oxford. 149 Mirov, N. T. 1967. "P. luchuensis" in the Genus Pinus. The Ronald Press, N. Y., p. 283. Miyazaki, Y., and K. I. Sakai. 1969. Use of zymography for identification of a clone in Cr tomeria ja- ponica. J. Jap. Forestry Soc. 51(91: 235-239. Monsalud, K. P., and F. R. Lopez. 1967. Kaatoan Bangkal-- "A wonderful tree." Philippines Lumberman 13(1): 60-64. NSC. 1972. U.S.-China cooperative science program semi- nar on forest ecology and genetics, 1972: 1-61. Mar. 31-Apr. 8, Nat'l Sci. Counc. NTU Forest. 1981. The research plots of the Experimental Forest. Nat'l Taiwan Univ. Exp. Forest, 1981, 10 PP- Ohba, K., and M. Murai. 1971. Recessive genes producing albino and light green seedlings in sugi, Cr to- meria japonica. J. Jap. Forestry Soc. 53(6): 177-180. ., Y. Okada, and M. Murai. 1971. Advancing the generation of sugi, Cryptomeria japonica, by green- house cultivation. J. Jap. Forestry Soc. 53(5): 133-137. Ohtani, S., and R. Siepono. 1962. Initial forest regenera- tion. Survey report of East Kalimantan, Indonesia. Perng, K. D. 1978. Study on the causes of success and failure of Taiwan paulownia plantations. Masters thesis, Nat'l Taiwan Univ., Dept. Forestry, 145 pp. (Chin.). Puerto Rico. 1963. Annual report 1963. Institute of TrOpical Forestry, Rio Piedras, Puerto Rico, For- est Serv., USDA. Rin, U. C. 1972. Silvics on cryptomeria and cypress. Today's Silviculture, Taiwan Forestry Bureau, 44: 73-44 (Chino). . 1973. Discussion on efficient silviculture. Today's Silviculture, Taiwan Forestry Bureau, 51: 11-34 (Chino). . 1979. Silvicultural studies on Taiwan paulownia (Paulownia taiwaniana Hu et Chang). Nat'l Chung- hsing Univ. Exp. Forest Res. Bull. 178, 228 pp. (Chin. and English summary); 150 Sheng, T. S. 1964. Notes of Kaatoan Banghal's introduc- tion. Taiwan Forestry Res. Inst. Notes 189: 1659- 1660 (Chin.). Sheng, W. T., L. Wang, and H. Y. Chang. 1981. A prelimi- nary study on the climatic regions of Chinese-fir growth areas. Scientia Silvae Sinicae 17(1): 50- 57 (Chin.). Shidei, T., and S. Moromizato. 1971. Karyotype analysis of luchu pine. J. Jap. Forestry Soc. 53(1): 13-18. Tai, K. Y., and C. S. Liu. 1965. Effects of altitude on growth of seedlings of cryptomeria, China-fir and taiwania. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 42, 24 pp. (Chin.). TFB. 1972a. "Pinus luchuensis" in Forestry in Taiwan. Taiwan Forestry Bureau 1972, pp. 25-30 (Chin.). . 1972b. The tree breeding workshop. Today's Silviculture 40: 1-75 (Chin.). . 1976. Forest statistics of Taiwan. Taiwan Forestry Bureau, 1976, p. 24 (Chin.). . 1977. Forest statistics of Taiwan. Taiwan Forestry Bureau, 1977, 159 pp. (Chin.). Toda, Y. 1980. On the karyotype of Cryptomeria 'a onica (V)--in Kyushu. J. Jap. Forestry Soc. 6257;: 264- 269. Troup, R. S. 1921. The Silviculture of Indian Trees. Clarendon Press, Oxford: 614-616. Tsou, E. R. 1976. Luchu pine progeny test and the selec- tion for the seed orchard establishment. Taiwan Forestry J. 2(2): 25-27 (Chin.). . 1978. Selection of the juvenile luchu pines before the establishment of progeny test seed or- chard. Quar. J. Chin. Forestry 11(2): 31-56 (Chin.). . 1979. Current status on tree breeding of Tai- wan Forestry Bureau--(I), (II). Taiwan Forestry J. 5(6): 17-24; 5(7): 15-20 (Chin.). Wang, C. W. 1961. The Forest of China. Harvard Univ., . Cambridge, Mass. 313 pp. ., T. T. Wang, E. R. Kuo, C. H. Chiang, S. J. Li, Wang, Wang, Wang, Wang, J. K. K. T. 151 T. W. Hu, Y. F. Suen, and P. Y. Yang. 1972. Pro- geny-provenance test of Pinus taiwanensis Hay. Station Paper 11, Forestry and Range Exp. Stn., Univ. Idaho, 11 pp. L. 1971. Differentiation and development of the pollen of Pinus taiwanensis. Taiwan Forestry Res. Inst. Bull. 206, 18 pp. (Chin.). ., and W. Y. Lin. 1974. The formation of female strobili and the development of cones in Pinus tai- wanensis. Quar. J. Chin. Forestry 7(1): 43-54 (Chin.). C. 1968. The control of Phassus signifer of taiwania plantation at Chitou. Nat'l Taiwan Univ. Exp. Forest Monthly Notes 12: 4-6 (Chin.). R. 1975. Discussion on failure of Taiwan red pine plantings. Taiwan Forestry J. 1(7): 9-12 (Chin.). T. 1976. Studies on the growth and ability of nutrient absorption of taiwania seedlings from different provenances. Memoirs of Coll. Agr., Nat'l Taiwan Univ. 16(2): 54-73 (Chin.). ., and Y. T. Huang. 1971. Effects of N-P-K fer- tilizers on the growth of Taiwania nursery seed- lings. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 89, 40 pp. (Chin.). ., C. Kao, J. C. Yang, and J. G. Shyu. 1976. Iso- enzyme variation of peroxidase in Taiwania. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 118-l: l-16 (Chin.). ., F. S. Kung, and C. M. Kung. 1969. Studies on structure and development of the strobilus of Tai- wania. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 68, 20 pp. (Chin.). ., and P. C. Kuo. 1960. Taiwan economic trees (I). Nat'l Taiwan Univ. Exp. Forest, Forestry Series I, 216 pp. (Chin.). ., and . 1980. Squirrel damage to eco- nomic forests in Taiwan. Nat'l Sci. Counc. Month- ly 8(6): 527-550 (Chin.). ., C. W. Lin, and T. L. Cheng. 1980. Reforesta- tion of Taiwan red pine and causes of its success 152 and failure. Taiwan Forestry Bureau, 1980. 58 pp. Whitmore, T. C. 1975. "Kadam" in TrOpical Rain Forests Wright, of the Far East. Clarendon Press. Oxford, pp. 84-85. J. W. 1970a. Report on Taiwan tree breeding. Nat'l Taiwan Univ. Exp. Forest Serial Publ. 47. 37 pp. . 1970b. Some recent Japanese tree breeding ex- periments. Nat'l Taiwan Univ. Exp. Forest Serial Publ. 48, 12 pp. . 1976. "Hybridization" in Introduction to For- est Genetics. Academic Press, pp. 324-328. ., Y. N. Yao, and T. W. Hu. 1972. Plans for the genetic improvement of Taiwan red pine. Nat'l Taiwan Univ. Exp. Forest Serial Publ. 58, 28 pp. ., F. S. Kung, T. W. Hu, and T. Liu. 1973 Profitable approaches to saving and improv- ing Taiwania. Quar. J. Chin. Forestry 6(4): 7-12. . 1978. Analysis of variance calculations for ir- regular experiments. Lake States Forest Tree Im- provement Conf. Proc. 13, 1978, pp. 147-157. Wyatt-Smith, J. 1965. Manual of Malayan silviculture. Yang, B. Part III. Malaysia Forestry Rec. No. 23. Y. 1964. Introducing cadamba. Taiwan Forestry Res. Inst. Notes 189: 1660-1661 (Chin.). Yang, Y. C., and C. Y. wu. 1956. Comparison of the growth Yao, Y. of Taiwan red cypress and cryptomeria at Chitou. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 7, 20 pp. (Chin.). N. 1970. Effects of transplanting density on the growth of seedlings of cryptomeria, China-fir and taiwania. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 81, 20 pp. (Chin.). . 1974. Introduction of exotic pine Species for testing in Taiwan. Nat'l Taiwan Univ. Exp. Forest Tech. Bull. 114: 57-68 (Chin.). . 1975. The characteristics of nested analysis of variance and its applications to tree breeding work. Quar. J. Chin. Forestry 8(3): 59-75 (Chin.). Yao, Yie, Y. S. 153 N. 1979a. Introduction of tropical and subtrOpi- cal pine species for testing in Taiwan--4-year, 9- month results. Memoirs of Coll. Agr., Nat'l Taiwan Univ. 19(1): 53-65 (Chin.). . 1979b. Analysis of variance for an experiment with missing plots. Quar. J. Chin. Forestry 12 (2): 79-90 (Chin.). ., and B. J. Yen. 1979. The fast-growing species: kadam. Its silvicultural techniques, wood proper- ties, usage, introduction status in Taiwan, and its growth comparisons. Quar. J. Chin. Forestry 12(1): 101-109 (Chin.). . 1981. Introducing tropical and subtropical pine species for testing in Taiwan, 7-year results. Proc. Symp. Forest Genetics and Breeding, Nat'l Chunghsin Univ., Taichung, Taiwan, pp. 30-51. T., P. P. Lee, S. T. Hsu, and M. Y. Tang. 1967. Biological study on the more important insect-pest attacking Genus Pinus introduced from the U.S-A. Quar. J. Chin. Forestry 1(1): 1-24 (Chin.). Yu, S. T., C. J. Chen, Y. L. Bai, and M. L. Huang. 1981. A preliminary observation on the flower-bud initi- ation of Chinese-fir in Nanpin, Fujian. Scientia Silvae Sinicae 17(1): 46-49 (Chin. ”TlT('ITIHjfi)flfllj]lljJMM’I‘EI'IMITMIYI‘s