Natural teak-bearing forests in Myanmar have been managed under the systematic way so called Myanma Selection System (MSS, formerly named Brandis Selection System, BSS) over the centuries. Teak and hardwoods production were conducted under the control of girth limit, periodic limit, and annual allowable cut, AAC. Dr. Brandis, a German forester, initiated the development of yield formula of AAC for the assurance of future consistent supply of teak. This yield formula is still in use for the estimation of AAC, believing that it will provide the sustainable productivity of teak. However, nowadays, Myanmar has been experiencing degradation of natural teak forest, consequent affect of over-exploitation above the prescribed cut limit, and challenging of restoring its degraded forests. This study was aimed to reveal the real situation of teak forests in Bago Yoma Region, which was once famous as home of teak and birth place of MSS, and then proposed an alternative yield regulation for sustainable production of teak.

Chapter 1 of this study provides the important role of teak forests in Myanmar, current situation of those natural teak-bearing forests, current logging system in Myanmar, development of Brandis Selection System (BSS), attributes in developing yield estimation formula for AAC.

Chapter 2 provides a description of the study area in Pyu Kun reserved forest which is one of the reserved forests in Bago Yoma. In this chapter, first a general description of Bago Yoma, silviculture system and teak forest composition are mentioned. And then, the detail about data availability and data sources of selected study area was explained. The collected data were composed of number of compartments, number of felled or girdled trees, and number of remaining trees, and time intervals of harvest. The records of harvested and left trees are traced back from 1976 up to 2008. These data covers 64 compartments with different sizes in acres and different felling cycles as well even though felling cycle of thirty-year is set up for teak under MSS. Instead of diameter classes used in other countries, the number of teak trees is allocated under seven girth classes of Class:(4'00" ≤ g1 ≤ 4'11";5'00" ≤ g2 ≤5'05";5'06"≤ g3 ≤5'11";6'00"≤ g4≤6'05";6'06"≤ g5 ≤6'11";7'00"≤ g6≤7'05" and g7≥7'06"). The unit for girth class is feet and inches and the area of each compartment is converted into hectare. According to the observation data, the remaining trees are counted starting from 4 feet in girth while the trees with 6.6 feet and above are marked as to be cut. There is no information available for numbers of teak younger than 4 feet in girth.

Chapter 3 aimed to reveal the current teak production from natural teak forest and to make comparative analysis with prescribed and actual cut of teak trees. In this section, three types of analysis were conducted: comparative analysis on girth class distribution of teak over various felling cycles; analysis on actual and calculated AAC of teak; and analysis on actual and calculated stand table projection. In addition, this chapter addressed the disturbance tendency occurring in natural teak forest over time series and over the girth class as well. The enumeration data from study area were applied in this study. The original girth classes were redefined depending on three types of analysis. In the process of harvesting operation, number of teak trees is enumerated by girth class, at the time of (1) before felling, (2) at felling and (3) after felling. These enumeration data (1976-2009) were collected from 123 compartments. However, for the comparative analysis, 64 compartments which have been harvested twice during the period.

Through those enumeration data, first the condition of growing stock of teak over the various felling cycles (FC) was revealed. For the feasibility of data arrangement, felling cycles are classified into A (FC≤10yrs), B (11≤FC≤20yrs), and C (21≤FC≤30yrs) respectively. And then, growing stock conditions of teak over three types of FC were compared. For the second type of analysis, we applied the Brandis Yield formula of AAC. Yield is regulated by number of exploitable trees. Here, only trees left data from 64 compartments were applied that are necessary for future yield estimation by the Brandis formula. According to girth limit prescription by the forest department, the trees which have the minimum girth limit of 6'06" are supposed as harvestable or yield trees. Therefore, all of the trees which have 6'06" and above at gbh (girth at breast height) were considered as girth class one (CI). Another assumption is that thirty years are required to increase one foot in girth of trees which have 5'06" at gbh. Accordingly, the girth classes were reclassified into four groups: 4'00" ≤CIV≤4'11"; 5'00" ≤CIII≤5'05"; 5'06" ≤CII≤6'05"; CI≥6'06". As for the third analysis of this section, we used the recently proposed model for the Stand Table Projection. In that model, annual survival rate and periodic mortality rate were adopted and the mean girth increment of teak in Bago Yoma was considered as 0.79 inches per year. For this analysis, the three compartments which were harvested the first time in 1981 and second time in 2001 were chosen. To apply the data in the stand projection formula, the original set of seven girth classes are redefined into six groups with the equal interval of 6 inches in each girth class. The number of both harvested and remaining trees from enumeration data was used.

To find out of disturbance tendency occurring in natural teak forest over time series and over the girth class, first, time series are defined as t1, t2, t3 and t4 representing before first harvest, after first harvest, before second harvest, after second harvest. The total number of trees cut and left was allocated under the time series of t1 and t3 while only remaining trees were considered at t2 and t4.Those trees were distributed under the rearranged four groups of girth class with the equal interval of one foot in girth.

By the result of growing stock condition of teak over three types of felling cycles (A, B, C), the remaining trees for the next cutting cycle are apparently losing in C-type of Felling Cycle compared with A and B types. But, as an effect of long time interval, trees in larger girth class were found increase though trees in smaller girth class in the next cutting cycle of C-type. It is suggested that for keep applying felling cycle of thirty-year which could give enough time to restore the forest after first harvest, intensive care of maintaining the trees left for next cycle is required. By the results of using AAC yield formula in the enumeration data of teak trees, over-exploitation of teak under A-type occurs seriously compared with other types of felling cycle. As a result of applying stand table projection model, actual stand table after 20 years is apparently lower than the projected one. According to results of examining disturbance tendency, disturbance on teak growing stock occurs at the time series t3. Apparently, 85% and 76.5% of CIV and CIII trees are losing at that time series. The harvest tendency of CI trees over two felling cycles (t2 and t4) is decreasing from about 12% to 4%. Consequently, significant loss of CII trees at t4 is probability due to over-exploitation of teak. It can be concluded that silvicultural operation for younger teak generation is seriously necessary such as gap planting and maintenance against disturbances likely occurred at the younger trees.

By checking the applicability of currently using yield formula and proposed one for stand table projection with the current conditions of teak forests, the author proposed an alternative yield regulation method for teak production. The concept of the new model was based on maturity of forests by considering the utility of maturity as an index for sustainable growing stock of teak. Here, the maturity has the unit of number of trees by years. By calculating maturity of each stand, we could decide which stand should be harvested or not. As there is no information for age from uneven-aged natural teak forest, the required year for tree increment between each girth class is calculated by adopting Mitscherlich equation. In this formula, parameters are estimated by applying in the yield table of Indonesia teak plantation which seemed similar to growth pattern of Myanma natural teak. And then, the maturity of teak in total gbh distribution of all compartments was calculated through the formula. By using the enumeration data of 64 compartments from research area, first, the total girth distributions over the measurement of time series t1, t2, t3 and t4 were examined. The result suggested that the loss of number of remaining trees between t2 and t3 were attributed to illegal loggings because the time interval between t2 and t3 was the rest-time for next legal cutting. Based on the girth distribution and substituted parameters for the required years for the increment of teak in each girth class.

The maturities in total girth distribution of teak were calculated According to the result of estimating maturity in total girth distribution, the maturity decreased drastically through t1, t2, t3 and t4. Decreasing maturities between t2 and t4 were because of first and second legal harvesting operations. Therefore, the reason of decreasing maturity at t3 was probability due to human disturbance and natural disaster. Through these results, unsustainable situation of current teak forests could be clarified. We also introduced three patterns of estimated maturities in some specific teak stands. As for pattern 1, the illegal cutting in younger stand was the main factor of decreasing maturity of the stand through all of measurement times. By the results of pattern 2, the decrease in maturity of stand between t2 and t3 is relatively small but, after second harvesting, maturity is dramatically decrease and as a result, second harvesting should not be conducted for maintaining maturity of that stand. For the pattern 3, the maturity of measurement time 3 and 4 are found higher than that of time 1 and 2, and as a result, sustainability of that teak stand is relatively good by comparing pattern 1 and 2 results.

Chapter 5 provides a general discussion and conclusion about the disadvantages of existing yield model, lack of silvicultural operations of MSS and advantages of the proposed one with the consideration of current growing stock of teak. It was said that, in natural forests where teak occur scattered in mixtures with other hardwood species, teak was found with the amount of 3 or 5 trees per acre. According to current enumeration data of study site, even one tree could not find in one acre. In that situation, the sustain yield of teak could not rely on only number of larger trees. In the new concept of maturity, the required years for growing up to next higher girth class of teak were taken into account. As long as the number of trees by years is found increasing in each girth class, the maturity of that forest is stable and that stand could be chosen for harvesting operation. While Brandis' method was applicable only for the forests with the excess amount of mature trees, the author proposed that the new one would be feasible to apply in the current situation of natural forests which has been facing the problems of degradation and overexploitation of timber in Myanmar. As for further study to improve this new model, conducting research on growth rate of teak in each girth class from the natural forests by installing permanent sample plots is recommended.

ミャンマーはチークが天然に生育する数少ない国のひとつである。同国の中部に位置するバゴ山系の天然林は特に高品質なチーク材を産することで有名で、19世紀半ばにはドイツ人林学者ブランディスが本数管理を基本に伐採量を決定する方式を導入し、比較的最近まで資源を維持してきた。しかし近年では国情が不安定なこともあって管理が行き届かず、天然林資源の劣化が進行している。本研究は、ミャンマーの天然林において特に価値の高いチークを持続的に生産するため、データに基づく現状把握を行い、ブランディス方式を改善した収穫規整の方法を提案したものである。本論文は全5章からなる。

第1章には、研究の背景と目的が記されている。ブランディス方式は、ミャンマーのチーク資源を原始の状態から開発するために作られたもので、過剰な蓄積を切り下げながら何十年もかけて定常状態に移行させてきた経緯が記されており、また近年の統計からチーク資源が質量とも減少している様子が示されている。チーク資源の劣化の原因を解明するとともに、ブランディス方式に替わる収穫規整の方法の開発を目的として掲げている。

第2章では、研究対象地とデータの説明が述べられている。バゴ地域の1つの行政区を選び、国有林経営の原簿を営林署で詳細に調べた結果、110の林班(伐採の単位)のうち64の林班で1980年以降に2回の択伐が行われていた。これらについて各回の択伐前後のチーク立木本数を径級別に数え上げ、収穫量や成長量、残存量の原データとして用いた。

第3章では、チーク択伐林経営と資源に関する最近30年間の現状がデータに基づいて分析された。観測期間中に2回択伐された64の林班では、その択伐の間隔が長いものでは30年、短いものでは僅か8年のものが含まれていた。しかしブランディス方式では収穫可能な成熟木の本数のみで伐採量を決めるので、短い間隔で択伐された林班では成熟木が著しく減少していることが判明した。また択伐間隔の長い林班においては、小径木からの進界が期待されるところであるが、実際には未熟木の本数が減少しており、その原因は主に周辺住民らによる違法伐採と推察された。この結果、ミャンマーの天然林におけるチーク資源の減少は、成熟木の過剰伐採と未熟木の違法伐採という異なる原因によって複合的に引き起こされていることがデータから実証的に示された。

第4章では、ここまでの分析の結果を踏まえた上で、ブランディス方式を改善する代替的方法が提案されている。ここでは新たに成熟度という概念が導入され、成熟木・未熟木を含めて一定径級以上のすべてのチーク立木について、その径級に到達するに要する年数が推定され、チーク資源量が径級別の本数と要する年数の積和として表現された。成熟度は[本数×年数]という単純な次元を有し、単に次の択伐までの期間年数を現存本数に乗じることでその間の成熟度の増加分を推定することができるという特徴をもつ。これをベースに、違法伐採による減少や、進界による増加を同じ基準で考慮することが可能となった。これは成熟木の本数のみで伐採量を決定するブランディス方式よりも優れた収穫規整となりうるものである。

第5章は、全体の考察と結論である。もとよりミャンマーのチーク材生産は、広大な天然林に点在するチークを対象として本数で管理するという簡便な収穫規整であり、また近年の顕著な傾向として択伐の間隔が短く不規則になっており、そこに違法伐採も加わって、保続生産するには大きな障害となっていた。択伐間隔の短縮や違法伐採など社会経済的な要因は容易に取り除けないものの、本研究で提案した成熟度の概念は、森林管理を技術的に改善することが述べられている。

以上の通り、本研究はミャンマーの天然林経営において、チーク資源量が過剰伐採と違法伐採という異なる2つの原因で損耗していることを、国有林職員として自らがデータを収集して実証的に示し、成熟度という新たな概念を導入してこれまでの収穫規整方式を技術的に改善した。この成果は現地での森林管理に貢献するのみならず、学術上の意義も大きい。よって審査委員一同は本論文が博士(農学)の学位論文として価値あるものと認めた。