UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ecological studies of reduced forest-fallow shifting cultivation of Karen people in Mae Chaem Watershed,… Wangpakapattanawong, Prasit 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-611965.pdf [ 11.38MB ]
Metadata
JSON: 831-1.0090492.json
JSON-LD: 831-1.0090492-ld.json
RDF/XML (Pretty): 831-1.0090492-rdf.xml
RDF/JSON: 831-1.0090492-rdf.json
Turtle: 831-1.0090492-turtle.txt
N-Triples: 831-1.0090492-rdf-ntriples.txt
Original Record: 831-1.0090492-source.json
Full Text
831-1.0090492-fulltext.txt
Citation
831-1.0090492.ris

Full Text

ECOLOGICAL STUDIES OF REDUCED FOREST-FALLOW SHIFTING CULTIVATION OF K A R E N PEOPLE IN M A E C H A E M WATERSHED, N O R T H E R N THAILAND, A N D IMPLICATIONS FOR SUSTAIN ABILITY by PRASIT W A N G P A K A P A T T A N A W O N G B.Sc, Chiang Mai University, 1993 M . S c , Iowa State University, 1996 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES FACULTY OF FORESTRY (Department of Forest Sciences) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A April 2001 © Prasit Wangpakapattanawong, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Fbreff Sci'l^Cf? The University of British Columbia Vancouver, Canada DE-6 (2/88) 11 Abstract The forest-fallow system of shifting cultivation of upland rice and other food plants practiced by the Karen people of Mae Hae Tai village, Chiang Mai, northern Thailand, is changing due to increasing population and a resulting decrease in per capita arable land-base. This has resulted in a reduction of the fallow period, which was 10 or more years in the past. The fallow is traditionally believed to act to restore and sustain soil fertility and control weed populations, but may also be important for the maintenance of upland rice productivity by maintaining soil structure and in other ways. Presently, this system involves one year of cropping followed by five years of crop-free fallow. The national Thai government is trying to change shifting cultivation to fixed-field agriculture, but some ecologists and social scientists oppose the idea using arguments about the ecological and cultural integrity of this traditional farming practice. There has been little empirical research to examine the advantages and disadvantages of the system. Ecological studies were conducted to examine nutritional aspects of the forest-fallow shifting cultivation using field experiments and a chronosequence of fields. The farmers were interviewed about their traditional knowledge of shifting cultivation system management. The yield of the upland rice crop under this system was found to be about 1 t/ha, but is variable within fields, between fields, and between years. The chronosequence study revealed that during the five years of fallow there was an increase in soil organic matter and total N attributed to the addition of litterfall from the fallow species, but a decline in pH, available P, and extractable K , Ca, and Mg. These decreases are attributed to nutrient uptake by the fallow vegetation and the decline in the effect of the burning at the end of the previous rotation. The largest changes in soil conditions took place when the 5-year fallow field was slashed, burned, and cropped. Standing-tree biomass Ill increased gradually during the fallow period. Chromolaena odorata dominated the first two years of the fallow period, and it accumulated about 7 t/ha of aboveground biomass within one year after rice harvesting. Fertilizer trials of the regular first-year and the experimental second-year upland rice crop showed that N was the most deficient nutrient in upland rice productivity, which support the nutritional role of the fallow. Growing rice for a second-year also revealed that soil pathogens may play an important role in decreasing upland rice productivity in consecutive-year cropping. Interviews with farmers were explainable in an ecological context; for example, the variability within and between crop fields which reflects variability of soil fertility, which in turn depends on localized topography and physical characteristics of the soils. The farmers responded that the fallow period could be reduced to a minimum of two to three years, and the data support that this might be associated with increased weed competition and less ash. The biogeochemical studies of the forest-fallow shifting cultivation system showed that nutrient losses via slash burning and harvested rice grain are important outputs of N . P was found to be lost the most via harvested rice grain, while losses in erosion and leaching may be important for K , Ca, and Mg. Quantitative assessment of other pathways of nutrient inputs (e.g. N fixation and soil weathering) and outputs (e.g. erosion and leaching) are needed for a complete bipgeochemistry of the ecosystem of the forest-fallow shifting cultivation in order to examine its sustainability. A series of carefully controlled and replicated field and pot experiments is needed to resolve the relative importance of the different contributions of fallow to the sustainability of upland rice. The following topics also deserve further research work: dynamics of N in the system, change in iv resource-allocation patterns between above- and belowground tree components, soil microbial activities and their effects on N cycling, and other roles of the fallow periods (e.g. maintaining good soil structure and providing useful plants and animals). The current fallow period of five years appears to be sustainable at the present landscape condition, but a further reduction in fallow length may pose a risk to the apparent sustainability of this forest-fallow shifting cultivation. Comparison of nutrient cycling between forest-fallow shifting cultivation and fixed-field farming by simple and/or computer models is needed to assess their sustainability. V Table of Contents Abstract i i Table of Contents ..v List of Tables xii List of Figures xxi Acknowledgements xxii Dedication xxiv Chapter 1 Introduction 1.1 Thailand and Deforestation 1 1.2 Northern Thailand and Its Land-Use 2 1.3 Shifting Cultivation 4 1.4 Shifting Cultivation in Northern Thailand 5 1.5 Shifting Cultivation Research 7 1.6 Shifting Cultivation Research in Thailand 8 1.7 Thesis Rationale 9 1.8 Thesis Approach 12 1.9 Major Hypothesis 14 1.10 Thesis Objectives and Strategies 14 1.11 Thesis Organization 15 Chapter 2 Research Site 2.1 Location • 17 2.2 Climate 19 2.3 Vegetation 21 vi 2.4 Population and Land-Use 21 2.5 Mae Hae Tai Karen Village 22 Chapter 3 Soils 3.1 Introduction 24 3.2 Objectives 25 3.3 Methods 3.3.1 Soil Sampling 25 3.3.2 Soil Chemical Analyses 28 3.3.3 Measurement of Mineralizable Nitrogen 29 3.3.4 Burning Effects on Mineralizable Nitrogen 30 3.3.5 Atmospheric Inputs of Nutrients 30 3.3.6 Remeasurement of Selected Soil Properties after One Year 31 3.4 Results and Discussion 3.4.1 Soil Texture 32 3.4.2 Soil Bulk Density 34 3.4.3 Soil pH 36 3.4.4 Effects of Burning on Soil pH 38 3.4.5 Soil Organic Matter 39 3.4.6 Soil Nitrogen 41 3.4.7 Mineralizable Nitrogen 42 3.4.8 Burning Effects on Mineralizable Nitrogen 44 3.4.9 C:N Ratio 44 3.4.10 Soil Available Phosphorus 45 3.4.11 Soil Extractable Potassium 47 3.4.12 Soil Extractable Calcium 49 3.4.13 Soil Extractable Magnesium 50 3.4.14 Soil Cation Exchange Capacity 51 3.4.15 Soil Extractable Iron 51 3.4.16 Correlations Amongst Soil Variables 52 3.4.17 Atmospheric Inputs of Nutrients 53 3.4.18 Soil Erosion (Literature Review) 55 3.5 Summary 57 Chapter 4 Upland Rice and Role of Nutrients in Rice Productivity 4.1 Introduction 60 4.2 Objectives and Rationale 63 4.3 Methods 4.3.1 Upland Rice Sampling 64 4.3.2 Chemical Analyses of Upland Rice Samples 65 4.3.3 The Second-Year Rice Experiment 65 4.3.4 Fertilizer Trial 66 4.4 Results and Discussion 4.4.1 Upland Rice Biomass 66 4.4.2 Upland Rice Nutrient Contents 67 4.4.3 Biomass of the First-Year (1999) Upland Rice Receiving Fertilizer Treatments 69 4.4.4 Nutrient Contents of the First-Year (1999) Upland Rice Receiving Fertilizer viii Treatments 70 4.4.5 The Second-Year Rice Experiment and Fertilizer Trial 74 4.4.6 Comparison of Upland Rice Yield Amongst Various Studies 77 4.4.7 Nitrogen Limitation in Upland Rice Productivity 79 4.4.8 Effect of Sulfur on Upland Rice Productivity 80 4.4.9 Effect of Fallow-Period Reduction on Upland Rice Yield 81 4.4.10 Yield:Seed Ratio of the Upland Rice 82 4.4.11 Effects of Soil Diseases and Pathogens on the Second-Year Rice Crop 83 4.5 Summary 84 Chapter 5 Trees 5.1 Introduction 86 5.2 Objective and Rationale 88 5.3 Methods 5.3.1 Estimation of Aboveground Tree Biomass and Slash Biomass 88 5.3.2 Tree Species Identification 89 5.3.3 Nutrient Contents of Aboveground Tree Biomass 90 5.3.4 Litterfall Sampling 90 5.4 Results and Discussion 5.4.1 Biomass Estimation 91 5.4.2 Comparison of Tree Biomass Amongst Various Studies 95 5.4.3 Tree Species 98 5.4.4 Nutrient Concentrations of Aboveground Tree Biomass 100 5.4.5 Litterfall 106 ix 5.5 Summary 109 Chapter 6 Shrub 6.1 Introduction 10 6.2 Objective and Rationale 112 6.3 Methods 113 6.4 Results and Discussion 6.4.1 Biomass ofC. odorata 114 6.4.2 Nutrient Contents of C. odorata 115 6.4.3 Nutrient Contents of C. odorata, Upland Rice, and Tree Components 117 6.5 Summary 118 Chapter 7 Synthesis of the Reduced Forest-Fallow Shifting Cultivation 7.1 Introduction 121 7.2 The Role of Fallow in Shifting Cultivation System: Alternative Hypotheses 122 7.3 Biogeochemistry of Forest-Fallow Shifting Cultivation 125 7.3.1 Data Used in Biogeochemistry of Forest-Fallow Shifting Cultivation 126 7.3.2 Discussion of Nutrient Budgets 130 7.4 Biogeochemistry of Shifting Cultivation with Different Crop-Fallow Combinations 142 7.5 Recycling of Human Waste in the Forest-Fallow Shifting Cultivation 152 7.6 Thesis Conclusions 7.6.1 Thesis Findings 153 7.6.2 Validity of the Alternative Hypotheses 160 7.6.3 Data Gaps 161 X 7.6.4 Future Research 162 References 164 Appendix 1 Weather Data .178 Appendix 2 Soil Moisture 180 Appendix 3 Soil Temperature 181 Appendix 4 Sub-Soil Properties 182 Appendix 5 Soil Profile Description 186 Appendix 6 Comparison of Soil Properties Amongst Various Studies 198 Appendix 7 Pot Experiment 204 Appendix 8 Wood Density 210 Appendix 9 DBH Distribution and Aboveground Biomass of Fallow Trees 211 Appendix 10 Farmer's Informal Interviews about Forest-Fallow Shifting Cultivation Management A . 10.1 Introduction 212 A.10.2 Objectives and Rationale 213 A . 10.3 Methods 214 A . 10.4 Results and Discussion A. 10.4.1 Farmers' Response 216 A.10.4.2 Summary of the Farmers' Common Responses 226 A . 10.5 Analysis of the Karen Knowledge on Shifting Cultivation Management A.10.5.1 Roles of the Fallow 227 A.10.5.2 Soil Fertility 228 A.10.5.3 Roles of Trees and Litterfall 229 xi A.10.5.4 Roles of C. odorata 231 A . 10.5.5 Farming Technology 232 A . 10.6 Summary 234 xii List of Tables Page Chapter 3 Soils 3.1 Characteristics of a chronosequence of shifting cultivation fields of Mae Hae Tai village 32 3.2 Soil texture at three depths in a chronosequence of shifting cultivation fields 33 3.3 Bulk density (g/cm3) of soils from shifting cultivation fields 34 3.4 Soil pH (1:1 in water) at three depths in a chronosequence of shifting cultivation fields 36 3.5 Soil pH (1:1 in KC1) at three depths in a chronosequence of shifting cultivation fields 37 3.6 Soil pH (1:1 in water) at three depths in shifting cultivation fields, measured in 1998 and 1999 37 3.7 Soil pH (1:1 in KC1) at three depths in shifting cultivation fields, measured in 1998 and 1999 38 3.8 Soil organic matter and organic carbon concentrations at three depths in a chronosequence of shifting cultivation fields 40 3.9 Soil organic matter concentrations (%) at three depths in shifting cultivation fields, measured in 1998 and 1999 40 3.10 Soil total N concentrations (%) at three depths in a chronosequence of shifting cultivation fields 41 3.11 Soil total N concentrations (%) at three depths in shifting cultivation fields, measured in 1998 and 1999 42 xiii Page 3.12 Soil N mineralization during a 30-day field incubation in a chronosequence of shifting cultivation fields 43 3.13 Soil C :N ratio at three depths in a chronosequence of shifting cultivation fields 45 3.14 Soil available P concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields 46 3.15 Soil available P concentrations (ppm) at three depths in shifting cultivation fields, measured in 1998 and 1999 47 3.16 Soil extractable K concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields 48 3.17 Soil extractable K concentrations (ppm) at three depths in shifting cultivation fields, measured in 1998 and 1999 48 3.18 Soil extractable Ca concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields 49 3.19 Soil extractable Ca concentrations (ppm) at three depths in shifting cultivation fields, measured in 1998 and 1999 50 3.20 Soil extractable M g concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields 50 3.21 Soil extractable Mg concentrations (ppm) at three depths in shifting cultivation fields, measured in 1998 and 1999 51 3.22 Soil CEC (cmol+/kg) at three depths in a chronosequence of shifting cultivation fields 51 3.23 Soil CEC (cmol+/kg) at three depths in shifting cultivation fields, measured in 1998 and 1999 ..52 3.24 Soil extractable Fe concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields 53 3.25 Correlation analyses of soil chemical properties the soils at three depths in a chronosequence of shifting cultivation fields 54 3.26 Nutrient concentrations and contents of rainfall and throughfall water 55 3.27 Summary of soil erosion studies in northern Thailand and northern Laos 57 3.28 Estimation of nutrient losses via soil erosion 58 Chapter 4 Upland Rice and Role of Nutrients in Rice Productivity 4.1 Upland rice biomass, Mae Hae Tai village, Chiang Mai, Thailand 67 4.2 Nutrient concentrations and contents of the upland rice crop of the 1998 and 1999 growing seasons 68 4.3 Biomass of the first-year upland rice components in plots receiving fertilizers (1999)..69 4.4 First-year upland rice biomass (1999 growing season) response to fertilizers. Mass response, and response per kg of added nutrients 71 4.5 Nutrient concentrations (%) of the first-year upland rice (1999) 72 4.6 Nutrient contents (kg/ha) of the first-year upland rice (1999) 73 4.7 Increased nutrient contents of the first-year upland rice (1999 growing season) resulting from added nutrients in fertilizers 74 4.8 Experimental second-year upland rice biomass response to fertilizers. Mass response, and response per kg of added nutrients 75 XV Page 4.9 Increased nutrient contents of the experimental second-year upland rice resulting from added nutrients in fertilizers 76 4.10 Nutrient concentrations and contents of the experimental second-year rice 76 4.11 Comparison of upland rice yields amongst various studies 78 4.12 Comparison of concentration of nutrients of upland rice grain in northern Thailand ...78 4.13 Comparison of nutrient contents of upland rice stems in northern Laos and northern Thailand 79 Chapter 5 Trees 5.1 Estimated tree biomass in shifting cultivation fields 91 5.2 Slash volume in the 5-year fallow field before burning 94 5.3 Comparison of some soil properties in the burial and secondary forests 96 5.4 Comparison of fallow-tree biomass (kg/0.04 ha) with other studies in northern Thailand 97 5.5 List of tree species (DBH > 5 cm in 0.04 ha) found in shifting cultivation fields and the forests 99 5.6 List of tree species in the 5-year fallow field, Mae Hae Tai village, and other sites in northern Thailand (species and family names are from Santisuk 1988) 101 5.7 Nutrient concentrations of different tree (10 years old, non-legume) components (Andriesse and Schelhaas 1987a) 102 5.8 Nutrient concentrations (%) of fallow-tree leaves (Pampasit 1998) 103 5.9 Nutrient concentrations (%) of fallow-tree wood (Pampasit 1998) 104 5.10 Nutrient concentrations (%) of fallow-tree bark (Pampasit 1998) 105 xvi Page 5.11 Aboveground litterfall (kg/ha) in shifting cultivation fields 107 5.12 Estimated nutrient contents (kg/ha) of litter 108 5.13 Comparison of nutrient contents (kg/ha) of upland rice, stemwood slash, and litterfall 108 Chapter 6 Shrub 6.1 Biomass (kg/ha; dried at 60° C) accumulation of C. odorata, Mae Hae Tai village, and other studies 114 6.2 Nutrient concentrations and contents of C. odorata 116 6.3 Nutrient concentrations (%) of C. odorata reported from various studies 117 6.4 Comparison of nutrient concentrations and contents of upland rice and C. odorata, and tree components 119 Chapter 7 Synthesis of the Reduced Forest-Fallow Shifting Cultivation 7.1 Nutrient pathways of total N (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers 131 7.2 Nutrient pathways of available P (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers 132 7.3 Nutrient pathways of extractable K (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers 133 7.4 Nutrient pathways of extractable Ca (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers 134 7.5 Nutrient pathways of extractable Mg (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers 135 xvii Page 7.6 Nutrient pathways in a 6-year rotation of one cropping year and five years of fallow 147 7.7 Nutrient pathways in a 6-year rotation of two cropping years and four years of fallow 148 7.8 Nutrient pathways in a 6-year rotation of four cropping years and two years of fallow 149 7.9 Nutrient pathways in a 6-year rotation of continuous cropping 150 7.10 Comparison of nutrient pathways among different crop-fallow combinations assuming a 6-year cycle 151 7.11 Nutrient contents (kg/ha) in soils, upland rice, shrub, trees, litter, and rainfall 155 Appendix 1 Weather Data A. 1.1 Temperature and rainfall data (1998 and 1999) at Mae Chaem Watershed Research Station, Royal Forest Department, Chiang Mai, Thailand (RFD 1999) 178 A . 1.2 Mean annual temperature and annual rainfall data from 1985 to 1999 at Mae Chaem Watershed Research Station, Royal Forestry Department, Chiang Mai, Thailand (RFD 1999) 179 Appendix 2 Soil Moisture A.2.1 Surface-soil (0-5 cm) monthly moisture content (%) 180 Appendix 3 Soil Temperature A.3.1 Surface-soil (0-5 cm) monthly temperature 181 xviii Page Appendix 4 Sub-Soil Properties A.4.1 Soil texture at four depths in a chronosequence of shifting cultivation fields ..182 A.4.2 Soil pH (1:1 in water) at four depths in a chronosequence of shifting cultivation fields 182 A.4.3 Soil pH (1:1 in KC1) at four depths in a chronosequence of shifting cultivation fields 183 A.4.4 Soil organic matter concentrations (%) at four depths in a chronosequence of shifting cultivation fields 183 A.4.5 Soil total N concentrations (%) at four depths in a chronosequence of shifting cultivation fields 183 A.4.6 Soil available P concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields 184 A.4.7 Soil extractable K concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields 184 A.4.8 Soil extractable Ca concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields 184 A.4.9 Soil extractable M g concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields 185 A.4.10 Soil CEC (cmoi/kg) at four depths in a chronosequence of shifting cultivation fields 185 A.4.11 Soil extractable Fe concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields 185 Page Appendix 6 Comparison of Soil Properties Amongst Various Studies A.6.1 Comparison of soil pH (1:1 in water) amongst various studies in northern Thailand .198 A.6.2 Comparison of S O M concentrations (%) amongst various studies in northern Thailand 199 A.6.3 Comparison of soil total N concentrations (%) amongst various studies in northern Thailand 200 A.6.4 Comparison of soil available P concentrations (ppm) amongst various studies in northern Thailand 201 A.6.5 Comparison of soil extractable K concentrations (ppm) amongst various studies in northern Thailand 202 A.6.6 Comparison of soil extractable Ca concentrations (ppm) of the secondary forest site with another study in northern Thailand 203 A.6.7 Comparison of soil extractable M g concentrations (ppm) of the secondary forest site with another study in northern Thailand 203 A.6.8 Comparison of soil CEC (cmol+/kg) of the secondary forest site with another study in northern Thailand 203 Appendix 7 Pot Experiment A.7.1 Comparison of biophysical characteristics of the 2-year and 4-year fallow fields 205 A.7.2 Upland rice biomass (g/pot) of the pot experiment 206 A.7.3 Upland rice biomass (g/pot) of the control treatment of the pot experiment 207 A.7.4 Upland rice biomass (g/pot) of the 21-0-0 fertilizer treatment of the pot experiment 207 XX Page A.7.5 Upland rice biomass (g/pot) of the 16-20-0 fertilizer treatment of the pot experiment 208 A.7.6 Upland rice biomass (g/pot) of the 15-15-15 fertilizer treatment of the pot experiment 208 A.7.7 Nutrient concentrations (%) of the upland rice of the pot experiment 209 Appendix 8 Wood Density A.8.1 Wood density of various tropical tree species 210 Appendix 9 DBH Distribution and Aboveground Biomass of Fallow Trees A.9.1 Numbers of trees in each D B H class in shifting cultivation fields (0.04 ha) 211 A.9.2 Fresh aboveground tree biomass (kg/0.04 ha) in each D B H class in shifting cultivation fields 211 List of Figures Pag Chapter 2 Research Site 2.1 Map of Thailand showing Mae Chaem Watershed. (Source: International Centre for Research in Agroforestry, ICRAF) 18 2.2 Mae Chaem Watershed showing the study site, Mae Hae Tai village, on the left. The series of small triangles represents soil sampling points. Rectangles indicate other ICRAF (International Centre for Research in Agroforestry) research sites. (Source: ICRAF) 20 Chapter 3 Soils 3.1 Mae Hae Tai village showing sampling sites (circles). The approximate location of the burial forest is shown (star). (Source: International Centre for Research in Agroforestry, ICRAF) 26 3.2 Three soil pits (small squares) in each shifting cultivation field 28 Chapter 5 Trees 5.1 D B H (cm) distribution of trees in shifting cultivation fields (0.04 ha) 93 5.2 Fresh aboveground biomass (kg/0.04 ha) of trees in shifting cultivation fields 94 X X l l Acknowledgements First of all, I graciously thank my supervisor, Dr. Hamish Kimmins of Forest Sciences at U B C , who led me to appreciate the importance of inductive research in ecological study and believed that conducting research in my own country would be beneficial to me as a future researcher and my country as a whole. His enthusiasm is greatly appreciated. I thank the following individuals and organizations: M y committee members, Drs. Art Bornke (Soil Science), Peter Jolliffee (Plant Science), Cindy Prescott (Forest Sciences), and Brad Seely (Forest Sciences), U B C . Dr. Soontorn Khamyong (Faculty of Agriculture, Chiang Mai University, Thailand) for his invaluable advice on administrative assistance, research methodology, laboratory spaces, and discussion. Drs. Pongsak Sahunalu (Faculty of Forestry, Kasetsart University, Bangkok, Thailand), Chusri Thisonthi (Biology Department, Chaing Mai University), Kanok Rerkasem (Faculty of Agriculture, Chiang Mai University) for routing and re-routing me to this research topic. Dr. David E. Thomas (Senior policy analyst of ICRAF-Chiang Mai) for leading my interest to this research and undeniable source of major funding for the work. Dr. Horst Weyerhaeuser for productive discussion People of Mae Hae Tai village for their smile and hospitality, especially Mr. and Mrs. Wongkaren for letting me taking over their home during my 20+ field trips to the village. Mr. Soontorn Saepun (ICRAF field assistant) who accompanied me to the site from the beginning to the end of the field work both on his motorcycle and a more comfortable 4x4 vehicle, shared many good stories (folklores, tales, and plain jokes), with his good work ethic and sense of humour, and said that "people become good friends when they go through hard times together and we are the case in point", and who finally educated me with the Karen dialect, of which Ta-Blur-Do-Ma (Thank you very much) is one of them. Ms. Pornwilai Saipongthong (ICRAF-Chiang Mai staff) for providing maps in this thesis. Jim Peters and Nipada Ruankaew (fellow graduate students) for fruitful discussion and good times. Ms. Rattanawan Mungkung (Tarn, a fellow Thai graduate student at UBC) , for her comments and proofreading. Mr. Cherdsak Liewlaksaneeyanawin (Liew, a fellow Thai student in U B C ' s Forest Sciences) for his helps on thesis completion and final submission. Staffs at ICRAF Chiang Mai office: Dr. Jureerat K . Thomas for discussion on experimental design, Khuns Pong, Jaeb, Saipim, Aoy Wattana, Aoy Umaporn, and Nikom for all their administrative helps and sharing good food and laughter. XX111 Mr. Seramethakul (Dr. Soontorn's assistant) for all his logistic helps. Mr. Anongrak of Soil Survey Division, Faculty of Agriculture, Chiang Mai, University, for his help with soil description. Students of the Forest Ecology group at Chiang Mai University (Tong, A , Um, Yon, Jom, and Fa) for their literally hard work carrying plant and soil samples back to the laboratory on the 5 t h floor of the building at midnights when the elevator already stopped working for the day. Staffs of CARE-Thailand in Mae Chaem for taking me through the rough roads of Mae Chaem in search of a perfect research site. Mr. Suchart Munmuang (Head of the Mae Chaem Watershed Research Station, Royal Forestry Department) for providing accommodation for both myself and my pot experiment. Also, with the weather data. Dr. Waree Chaitep (Sanpatong Rice Research Station, Chiang Mai) for providing data and information on upland rice research in northern Thailand. The DSPT (Development and Promotion of Science and Technology Talent Project), Thailand, for financial support of my education from my high-school through this highest degree. This must be one of the longest scholarships in the world, nothing less than 15 years in my case. I deeply appreciated this support. M y uncle and his family (the Tharnluangthongs of Chiang Mai) for providing a comfortable place to stay during my education and field work in Chiang Mai. M y high-school and college friends in Chiang Mai (Kook and Tie, Ann, Da, Pook, V i , Oar, and Jun) whose mental support is immense. Also, the Thai friends at U B C . M y best friend since our time at Iowa State University in Ames, U S A (Dr. Satjar Ravungsook). I cherish our friendship in the past and into the future. The major part of funding for my thesis research was from International Centre for Research in Agroforestry (ICRAF). The work was also partially funded by S E A M E O Regional Center for Graduate Study and Research in Agriculture (SEARCA) via the University Consortium program between Kasetsart University, Bangkok, Thailand, and U B C . I dedicate this work to my parents: M y Pa: who, with his own elementary-school education, secretly wanted to see his kids holding some university degree; who took me out to hunt game prizes every year on the Thai Children's Day (the second Saturday of January), among other occasions; who stood on his 6-hour train trip to visit me at school in Chiang Mai, and whose time on this world was too short. M y Mom: who has been working hard all her life; who says "being a good person is the only thing that she wants from her children". This is for you. THIS THESIS IS DEDICATED TO MY PARENTS: THE LA TE MR. SOMCHAI WANGPAKAPA TTANA WONG AND MRS. YOD (KAEWPIAH) WANGPAKAPATTANAWONG OF PHITSANULOK, THAILAND FOR THEIR LOVE AND DEVOTION 1 Chapter 1 Introduction 1.1 Thailand and Deforestation "Forest is the land area which no-one has authority to occupy or use" (The 1941 Forest Act of Thailand) Deforestation of tropical forests, which is believed to play an important. role in various environmental issues such as climate change, loss of biodiversity, and soil deterioration, has been gaining increasing attention in the past decades. Thailand is no exception; it has long been cited as one of many tropical countries with a high rate of deforestation. Debates on the causes and consequences of this forest loss have dominated conservation and environmental discussions for the last 30 years. The forested area in Thailand has declined dramatically over the past 30 years due to agricultural expansion and illegal forest encroachment and logging. In 1961, forest covered about 55% of the total land area of Thailand (Pragtong 1996). Forest cover was about 40% in 1973, about 26% in 1993 (Sangwanit 1995), and about 25% in 1998 (Royal Forestry Department 1998). In 1988, massive landslides, mud flows, and debris avalanches containing large volumes of downed trees and logs caused flooding in southern Thailand that destroyed villages and killed many people. These monsoon-related events were attributed to logging and deforestation of steep hillsides, and caused a strong and unprecedented conservation movement to force the government to manage Thai forests using more environmentally sound practices. As a consequence, the Thai government banned all logging concession in natural forests. However, the actual cause of the massive landslide is believed to be the planting of shallow-rooted rubber trees (Hevea brasiliansis) on steep slopes (Kunstadter 1990). This has been a common commercial practice in the south. The Thai government, via the Royal Forestry Department (RFD), has been trying to increase the forest cover of the country by several means. The target, according to the National Forest Policy in the National Economic and Social Development Plans, is to increase the forest cover to 40%. This is divided into 15% natural areas (including watersheds, national parks, wildlife sanctuaries, non-hunting areas, reserved parks, arboretums, botanical gardens, and reserved areas for specific studies), and 25% commercial forests (plantations, community forests, private tree farms, and timber concession areas) (Arbhabhirama et al. 1988). However, these target numbers were reversed in 1992 (i.e. 25% natural and 15% commercial) following the seventh (1992-1996) National Economic and Social Development Plan (Ganjanapan 1998). It was proposed to establish more national parks and reserved forest areas throughout the country, but especially in northern Thailand, where much of the forest is still intact: 14% of the total land area compared to 4.6, 4.1 and 2.4% in the central, northeast, and south, respectively (RFD 1998). 1.2 Northern Thailand and Its Land-Use Geographically, northern Thailand plays a crucial role in the regulation of downstream water supply to the Chao Phraya basin in the central region, where the extensive irrigated rice production lies. Currently, northern Thailand has approximately 400,000 and 650,000 ha of national parks and wildlife reserves, respectively, accounting for 10% of the region's total land area (Santisuk 1988; Rerkasem and Rerkasem 1994). The RFD has recently announced proposals for a number of new national parks in the north. The areas of northern Thailand are mostly inhabited by hill-tribe minorities and up-slope migrants. In most cases, the minorities settled in their present villages long before the 1960 Wildlife Preservation Act, the 1961 National Park Act, and the 1964 Reserved Forest Act were passed by the Thai Cabinet (Ganjanapan 1997). In fact, the Karen relocated from Myanmar (formerly Burma) more than 300 years ago, and the Lua are native to the area (Rerkasem and Rerkasem 1994; Ganjanapan 1998). In 1867, a Christian missionary visited northern Thailand and reported the use of swidden agriculture or shifting cultivation by the Karen and other hill-tribe minorities (Srisawas 1983). This has led to many conflicts regarding land-use policy. On the one hand, ecologists and environmentalists are delighted to have more protected areas; especially areas classified as class-1 watersheds. This watershed class is comprised of protected or conservation forests, and is divided into 2 subclasses: class 1A and IB. Both include areas of protected forest and headwater-source areas, usually at higher elevation with very steep slopes and permanent forest cover. The difference is that class IB is partially cleared for agriculture and settlement, and should have special attention on soil conservation measures (Arbhabhirama et al. 1988). On the other hand, social activists, anthropologists and sociologists oppose the proposal because those areas are mostly inhabited by hill-tribe minorities and up-slope migrants. They assert that this land-use change has political implications involving Thai nationality, prejudice against those minorities, and human rights violation. Confrontations between highland and lowland dwellers concerning water shortages and water pollution from fertilizers, herbicides, and pesticides have occurred almost annually in the north 4 (Forsyth 1996). In a series of demonstrations the highlanders were blamed for downstream dry-season water shortages and water contamination from horticultural chemicals. Ganjanapan (1998) recently discussed the political dimension of this issue at length. Thomas (Thomas, D.E., Senior Policy Analyst, ICRAF-Chiang Mai Office, per. comm.) has dubbed one of these confrontations as "the Chomtong Saga". Chomtong is one of the districts of Chiang Mai. It is the location of the Doi Inthanon national park to the east of Mae Chaem watershed, the study site for this thesis. There are a number of hill-tribe villages within this park. Such conflicts are not new. As recent as 1999, there was a confrontation between the government officials and the minorities and their supporters, including academics and social activists. The minorities demanded to stay in their current settlements and to be granted Thai citizenship. The confrontation between the government and the hill-tribe minorities ended with the enforcement of the government edicts. In northern Thailand and elsewhere (Warner 1991), highland shifting cultivation practices are often blamed for decreasing forest cover, dry-season water shortage, downstream sedimentation, and flooding. 1.3 Shifting Cultivation Conklin (1961) defined shifting cultivation or swidden agriculture (swidden is an old English word meaning "fire clearing" (Izikowitz 1951; Keen 1972 cited in Sutthi 1996) as an agricultural system in which fire is used to clear fields, and the period of cropping is normally shorter than the fallow period. Watters (1960) divided shifting cultivation practice into five major stages, which are 1) site selection, 2) clearing, which leaves certain tree stumps to sprout during the fallow period, 3) burning, which releases mineral nutrients to promote crop growth and reduce weed competition, 5 4) cropping, which lasts for one to three years, and 5) fallowing, which restores soil fertility. He recognized that there was variation in these stages from region to region (see also Nye and Greenland 1960; Ruthenberg 1971). Ketterings et al. (1999) listed some advantages of the burning stage, which are the clearance of unwanted weeds, the alteration of soil structure making planting easier, the addition of ash as a fertilizer, and the reduction of soil acidity. However, burning also poses a number of disadvantages such as alteration of soil structure (Ahn 1974; Uhl et al. 1981; Christanty 1986), losses of carbon, nitrogen and sulphur via volatilization depending on the temperature (Christanty 1986; Andriesse 1987 cited in Kleinman et al. 1995), and exposure of the soils to rain, wind erosion, and direct sunlight (Kleinman et al. 1995). 1.4 Shifting Cultivation in Northern Thailand In Thailand, shifting cultivation is practised by both ethnic Thai and hill tribes. In 1994, the estimated population of these shifting cultivators was one million, occupying five million ha of land in northern Thailand (Bass and Morrison 1994). Shifting cultivators are found mostly in the northern and western (bordered by Myanmar), and the northeastern (bordered by Laos) parts of the country. Shifting cultivation is known in Thailand as the farming method practiced by hill-tribes. It is widely thought that this involves clearing of primary or secondary forests followed by burning and cropping until the soils are exhausted; the hill-tribes then move to clear new patches of forest. The Thai term for shifting cultivation is "Rai Ruen Loi" , which literally means "movable farm". However, a number of anthropologists and ecologists recognize that shifting cultivation (or slash-and-burn agriculture) systems are more diverse than this. Kunstadter et al. (1978) classified shifting cultivation practice into three major groups; 1) short-cropping and short-fallow periods, 2) short-cropping and long-fallow periods, and 3) long-cropping and very long fallow periods. McGrath (1987) defined it more holistically as "a strategy of resource management in which fields are shifted in order to exploit the energy and nutrient capital of the natural vegetation-soil complex of the future site". Shifting cultivation is also classified into pioneer and rotational systems. The pioneer shifting cultivators, e.g. Hmong tribe, abandon their fields after several crops, whereas the rotational farmers, e.g. Karen and Lua tribes, return to their previously cropped fields after certain fallow periods, the length of which varies from place to place (Bass and Morrison 1994). Sanchez (1995) recognised shifting cultivation as a form of sequential agroforestry, in which crop and tree components occur at different phases. Fujisaka et al. (1996) recently classified shifting cultivation by using four main variables: initial vegetative cover, type of cultivator, final vegetative cover, and fallow length. Northern Thailand was once a prime area for opium production, and there was an assumption that shifting cultivation and opium cultivation were closely related (Schmidt-Vogt 1999). Massive enforcement to end opium production was initiated in association with the U N in the 1970s. Also, the government often views shifting cultivators as an underdeveloped segment of society that is in need of development, while the military sector has declared the region to be strategically important for national security (Forsyth 1996). The region is bordered by the politically unstable Myanmar, and is known internationally as the Golden Triangle of drug production and trafficking. As a result, a wide variety of government agencies, ranging from the Office of Narcotic Control Board, the Ministry of Education, and the Ministry of Defense, to the Ministry of Agriculture, are responsible for various aspects of shifting cultivation (Bass and Morrison 1994). Due to rapid changes in population, land use and development (road expansions), shifting cultivation practices are changing. They are either being changed into sedentary (permanent or fixed areas) farming of market-oriented crops, or into reduced-fallow shifting cultivation of partially subsistence crops. However, the division between the different groups of shifting cultivators is breaking down (Bass and Morrison 1994; Rerkasem and Rerkasem 1994) due to easier transportation, knowledge exchanges, and a greater influence of market. 1.5 Shifting Cultivation Research Among the earliest studies of shifting cultivation were those of Conklin (1957), Watters (1960), and Nye and Greenland (1960, 1964). The Nye and Greenland (1960, 1964) study was ecological, and recognized the decline of soil fertility during cultivation followed by the abandonment of the cropped area into fallow fields. Since then, hundreds of studies of both scientific and social aspects relating to shifting cultivation have been undertaken in Africa, Central and South Americas, and Southeast Asia. The studies are somewhat site-specific due to high variability in both cultural and environmental factors. However, there are also a number of common features among these systems. For example, two main reasons for field abandonment are believed to be weed infestation and a decline in soil fertility, which is reflected in yield decline. Soil fertility usually continues to decline for the first few years of the fallow periods, and then increases toward the original precropping level (Sanchez 1976; Zinke et al. 1978). The shifting cultivation research done in Africa has focused on soil changes after land clearing (Nye and Greenland 1964), effects of clearing and cropping on soils and vegetation (Ayanaba et al. 1976), succession and soil fertility changes in fallow fields (Aweto 1981a, 1981b), effects of burning and ash fertilizer (Stromgaard 1984), changes in soil physical and chemical properties after land clearing (Hulugalle et al. 1984; Mueller-Harvey et al. 1985), and carbon dynamics in slash-and-burn agriculture (Kotto-Same et al. 1997). For Central and South Americas, the focus has been on impacts of slash-and-burn on tropical wet forest (Ewel et al. 1981), effects of clearing methods including slash-and-burn on soil properties (Alegre and Cassel 1986), biomass studies of fallow (Brubacher et al. 1989), fallow length and soil fertility (Wadsworth et al. 1990), fallow management (Unruh 1990; Szott and Palm 1996), nutrient budgets in shifting agriculture (Holscher et al. 1997), and fallow effect on crop yield (Silva-Forsberg and Fearnside 1997). These study areas either have different soil orders or forest types. Andriesse (1980) stated that natural variability of soil biological, chemical, and physical properties, within soil series may be higher that between crop and fallow periods. This can make comparion among different locations even more difficult (Andriesse and Schelhaas 1987a; Kleinman et el. 1995). 1.6 Shifting Cultivation Research in Thailand The first major observation of fallow vegetation in northern Thailand was reported by Credner in his 1928-1929 geographical survey of Siam (the former name of Thailand), which was written in German (Credner 1935a, 1935b cited in Schmidt-Vogt 1999). Schmidt-Vogt (1999) also cites a study of succession in shifting-cultivation fields by Loetsch (1958, 1962) as the first vegetation study in this region. However, the first comprehensive study of shifting cultivation in northern Thailand was undertaken by Kunstadter, Chapman, and Sabhasri in the late 1970s. Their review, 9 entitled "Farmers in the Forest", is considered the most exhaustive work on the subject done in Thailand to date (Kunstadter et al. 1978). Nakano (1978) also undertook a comprehensive study in northern Thailand. Other major studies of shifting cultivation in northeast Thailand are Kyuma and Pairintra (1983), Kyuma et al. (1985), and Tulaphitak et al. (1985a, 1985b). Recent studies in the north have examined adaptations of shifting cultivation systems (Hansen 1995), shifting cultivation in relation to perception of soil degradation (Forsyth 1996, 1999), changes in soil properties (Funakawa et al. 1997a, 1997b; Tanaka et al. 1997, 1998a, 1998b), and a comparison of fallow vegetation among hill-tribe groups practicing different kinds of shifting cultivation (Schmidt-Vogt 1999). Nevertheless, these studies still constitute only a small quantity of research in comparison to the need to understand the dynamics of shifting cultivation in northern Thailand. 1.7 Thesis Rat ionale The focus of my research is the short-cultivation-and-long-fallow type of shifting cultivation (Kunstadter et al. 1978) used by Karen. This is called forest-fallow shifting cultivation or rotational farming by the Karen themselves, and by some social activists. In recent years there has been a reduction of the fallow period, which is likely to result in degradation of soils. There were three main reasons for choosing to study the changing land-use system of this particular hill-tribe group: 1) The Karen tribe is the biggest highland-dwelling minority group in Thailand. They account for the high frequency of land under this type of shifting cultivation practice in northern and western 10 parts of Thailand. The total number of Karen villages was 1,813 in 1993 (about 60% of the total hill tribe population), with 732 villages in Chiang Mai province alone (NSC/NESDB 1993 cited in Rerkasem and Rerkasem 1994). In 1995, the total number of the Karen population was estimated at 321,900, which is about 46% of Thailand's total hill-tribe population (Tribal Research Institute 1995). 2) Even though there has been a steady increase in the efforts of the government and conservation groups in Thailand to halt shifting cultivation by changing it into fixed-field agriculture, there have rarely been quantitative studies comparing the advantages and disadvantages of these two systems. Many conservation groups have claimed that shifting cultivation is degrading the soils and highland ecological systems and that permanent agriculture is more "sustainable", without any empirical ecological evidence to support these claims. 3) This thesis work was financially supported by the International Centre for Research in Agroforestry (ICRAF), whose benchmark research site is the Mae Chaem watershed in Chiang Mai province, northern Thailand where the Karen is the majority. One of ICRAF's missions in northern Thailand is to compare the advantages and disadvantages of different land-use systems in agronomic sustainability, biodiversity, economic, and watershed-function aspects (Thomas 1995). The ICRAF study is intended to lead to adaptation and/or introduction of agroforestry land-use systems into this area that are both profitable and sustainable. M y research addresses the ecological sustainability of the system. 11 The Karen and the Lua, as stated earlier, were the first two groups of ethnic minorities to migrate into northern Thailand. These two tribes are culturally and socially different, but their farming practices (shifting cultivation) are somewhat similar and belong to the second group classified by Kunstadter et al. (1978). Previously, they subsistence farmed by growing upland rice for one year and then leaving the cropped fields to fallow for up to 10 years. In some cases, villages still maintain a long fallow-period of 10 or more years because difficult access by roads results in small or zero influence of market pressure and modern farming technology, and their remoteness from national protected areas has also protected them from a reduction in available land. However, in many villages, pressures from population increase and land limitation have reduced fallow periods to five years. This is the case in the village studied in this thesis. The shortest fallow period found in the Mae Chaem Watershed is three years. For villages closer to the city and with decent road conditions, farming has been changed from mainly subsistence to permanent commercial farming. Similarly, villages close to state protected-areas have been forced to change into fixed-field agriculture based on rotational crops of corn, rice, and soybean, with former fallow areas left to return to a forested condition for aesthetic and conservation reasons. This change has yielded many interesting research questions. For example, which type of farming, rotating fields or rotating within a fixed fields, is more sustainable in terms of soil fertility. The change into fixed-field farming encourages use of inorganic fertilizers, herbicides, and pesticides, which leads to environmental and health hazards. In forest-fallow shifting cultivation, such external chemical inputs and irrigation are very rare (Kleinman et al. 1995). However, a number of farmers in the study area have begun to use herbicides in their upland rice fields because it is less labour intensive than manual weeding. 12 1.8 Thesis Approach The research reported in this thesis deals with some ecological aspects of the sustainability of reduced-forest-fallow shifting cultivation. The concept of sustainability has been popular in the past few years, since the Rio Summit in 1992. Sustainable development is defined as "development that meets the needs of the present without compromising the ability of future generations to meet their own needs" (WCED 1987). However, it means different things to different people. F A O (1989) defined sustainable agriculture as involving "the successful management of resources for agriculture to satisfy human needs, while maintaining or enhancing the quality of the environment and conserving natural resources". Spencer and Swift (1992) suggested that a sustainable cropping system is "one in which the output trend is non-declining and resistant, in terms of yield stability, to normal fluctuations of stress and disturbance". In Thailand, such definitions have been commonly used by academics, environmentalists, activists, forestry and agricultural extensions, and farmers. However, finding a satisfactory definition can be as difficult as finding methods to measure it. Kimmins (1974) defined sustainability as "a non-declining pattern of change", which is similar to the definition of Spencer and Swift (1992). It wil l be used as the operational definition of sustainability throughout this thesis. In order to determine whether a system is sustainable, one must select measurable parameters, and monitor them over the long-term because the concept of sustainability implies a temporal dimension. Swift and Woomer (1993) suggested the use of soil organic matter (total soil organic matter, soil microbial biomass, labile fraction of organic matter), and nutrient mineralization capacity as indices of sustainability. Larson and Pierce (1991) presented a set of soil attributes, the minimum data set (MDS), which can be used to evaluate soil quality. The M D S includes 13 nutrient availability, total and labile organic carbon, soil particle size, plant-available water capacity, soil structure, soil strength, maximum root depth, pH, and electrical conductivity. These soil attributes have been suggested as indicators of soil quality because they generally relate to soil functions and productivity, hence sustainability. Soil quality is defined as "the capacity of a soil to function within ecosystem boundaries to sustain biological productivity, maintain environmental quality, and promote plant and animal health" (Doran and Parkin 1994). In the past, most ecological research on shifting cultivation and similar land-use systems focused either on soil processes (Forsyth 1996, 1999; Funakawa et al. 1997a, 1997b; Tanaka et al. 1997, 1998a, 1998b) or on vegetation (Schmidt-Vogt 1999). A few researchers combined both components by employing an ecosystem approach (Kyuma and Pairintra 1983; Christanty 1989; Christanty et al. 1996, 1997; Holscher et al. 1997; Mailly et al. 1997). This thesis is built on the ecosystem concept of "biogeochemistry", which is the nutrient pathways of a particular ecosystem. Three pathways are recognized; geochemical, biochemical, and biogeochemical (Kimmins 1997). Among these pathways, soil and vegetation are two major sinks and sources of the nutrients. The vegetation acts as a nutrient uptake, storage, recycling mechanism, and a vehicle for nutrient export. The soil serves as an anchor and supplier of nutrients for vegetation, a source of nutrient replenishment through weathering processes, and storage compartment for organic matter and nutrients that are recycled into inorganic forms via decomposition processes. In this study of shifting cultivation, nutrients bound in soils and plants are considered to play a key role in determining the productivity of the system. 14 1.9 Major Hypothesis The working hypothesis for thesis was that upland rice crop yields are nutrient limited and that the major function of the fallow period is to restore soil fertility following the period of cropping. A consequence of this hypothesis is that soil fertility and crop yields would decline i f the fallow were to be significantly reduced in length from five years. 1.10 Thesis Objectives and Strategies The working hypothesis predicts that there is a measurable decline in measures of soil fertility following the slash/burn and cropping period, and that these measures recover to initial levels over the fallow period. The thesis sought to document such a decline and recovery in soil characteristics over the six-year cycle by: 1) studying a chronosequence of fields covering the length of the cycle. 2) re-measuring the chronosequence after one year. 3) undertaking a bioassay of soil fertility (a pot trial) over the chronosequence of fields. 4) testing for declining crop yields and nutrient limitation on upland rice crop growth in a field fertilizer trial in a second-year upland rice field. The thesis sought to initiate the development of ecosystem nutrient budgets to help identify critical data gaps that will be used to address sustainability of this forest-fallow shifting cultivation system. 15 1.11 Thesis Organization The thesis work and presentation are divided into three major parts: 1) The first part represents biophysical aspects of the forest-fallow shifting cultivation system of the Karen of Mae Hae Tai village (Chapter 3-6). In order to document soil fertility changes within this system of shifting cultivation, it was essential to describe the characteristics of soils and vegetation at different times in the cropping/fallow cycle. There were no previous empirical data on these aspects of this type of shifting cultivation in this area, and, therefore, an initial gathering of basic data was necessary. This involved soil and vegetation sampling followed by chemical analyses for macro-nutrients (N, P, K , Ca, and Mg). The biomass of the crop (upland rice), shrub {Chromolaena odorata), trees, and litterfall was also measured. This aspect of the thesis is divided into sections on soils, crop, shrub, and trees, with measurement of litterfall. The results are summarized by approximate tabular models of nutrient flows in the system at different stages of shifting cultivation. This part of the thesis is mainly descriptive and involved inductive surveys of a variety of shifting cultivation fields in the study area. 2) The second part on the thesis involves field and pot experiments focusing on fertilizer Trials (Chapter 4 and Appendix 7). This hypothetic-deductive component of the thesis addressed two main hypotheses: 2.1 If nutrients are critical in maintaining upland rice productivity, the rice crop should respond positively to additional nutrient inputs, e.g. chemical fertilizers, and negatively to a second year of cropping without fertilizers. This was tested by fertilizer trials in the regular first-year rice cropping and the experimental second-year rice cropping. 16 2.2 If the fallow period plays a key role in maintaining upland rice productivity by replenishing soil fertility through the accumulation of organic matter and nutrients by the fallow vegetation, and i f nutrients and soil fertility are key determinants, a reduction in the fallow period should result in declining rice productivity. The pot experiment, using soils from young fallow-fields, e.g. 2- and 4-year fallow, was conducted to test the potential of these soils to support normal rice yields in comparison to soils from a 5-year fallow of Mae Hae Tai village. A fertilizer trial was also included in this experiment to test i f any difference in the rice growth potential of the soils from the different fallow ages could be attributed to nutritional factors. 3) The third part of the thesis consists of informal interviews with farmers to assess their knowledge of their shifting cultivation management (Appendix 10). In recent years, the indigenous and traditional knowledge of different groups of tribes or minorities concerning several aspects of plants and farming management has been recognized (i.e. ethnology). This part of the thesis was an attempt to gather the farmers' knowledge in relation to soil and vegetation aspects of the system, and to analyze the knowledge in an ecological context. This aspect of the thesis is by no means comprehensive, because the major portion of the thesis research is focused on biophysical aspects of the topic. Note that farmer interviews were also used in site selection and overall thesis design. 17 Chapter 2 Research Site 2.1 Location Thailand (7° 6* to 23° 3' N and 97.5° to 106° E) has a total land area of 51.4 million ha (Figure 2.1). The landscape of the northern region is mostly mountains of north-south orientation. The region covers about 10.8 million ha, which is 21% of the country land area (Santisuk 1988). The highest peak in the region and the country is the Inthanon Peak (2,565 m asl.). Most rivers in this region drain into the densely populated and extensively irrigated Chao Phraya River basin in the central plain, which passes through the capital city of Bangkok on its way to the Gulf of Thailand. Two rivers, one in the extreme east and the other west of the region, drain into the Mekong in Laos and the Salween in Myanmar, respectively. There are a number of major agriculturally productive intermountain valleys (150-380 m asl.) where more than 50% of the population is located (e.g. the cities of Chiang Mai and Chiang Rai). Chiang Mai is the largest city in the north, and the second largest in the country after Bangkok. The District of Mae (river) Chaem belongs to the province of Chiang Mai , and is about 120 km to the southwest of the Chiang Mai city. Mae Chaem River flows north-south through the center of the district and merges with the Ping River south of Chiang Mai. The Ping River originates north of Chiang Mai and provides water supply for the city and agricultural uses within and south of Chiang Mai valley. Mae Chaem is one of the major tributaries of the Ping River, which in turn serves as one of four major tributaries, along with Wang, Yom, and Nan, of the Chao Phraya River. Mae Chaem River provides approximately 40% of total flow of the Ping River (Thomas, D.E., Senior Policy Analyst, ICRAF-Chiang Mai Office, per. comm.). 18 Figure 2.1 Map of Thailand showing Mae Chaem Watershed. (Source: International Centre for Research in Agroforestry, ICRAF) 19 The Mae Chaem Watershed (Figure 2.2) is about 49 km wide (east-west) and 113 km long (north-south), with a total area of about 400,000 ha. The watershed has about 96% of its area (354,200 ha), in the Mae Chaem District, Chiang Mai province, and the rest is in Hod District. It ranges from 18° 10' to 19° 9' N , and from 98° 5' to 98° 30' E. The watershed can be geologically divided into three main parts; 1) the new sediment valley at 190-500 m asl., with 0-4% slopes, 2) the old sediment hills at 500-650 m asl., with 4-16% slopes up to 35% slope at some points, and 3) the highland at 700-2,000 m asl., with more than 35% slope. 2.2 Climate Overall, Thailand has a rainy tropical monsoon climatic system with warm southwest winds prevailing from mid May to October, and with dry and cool northeast winds from November to April. Average annual rainfall ranges from 1,000 to 1,700 mm, with heavy monsoon rain from July to September. Annual mean temperature is about 25 °C, with a range from 40 °C in late April in some areas to near freezing in many highland areas, especially in the north and northeast, during January and February (Na-Nagara 1990). The Mae Chaem Watershed has a monsoon climate type like the rest of the country. From 1985-1999, the mean monthly temperature was 21 °C (S.E. = 0.7), and mean annual rainfall was 1,214 mm (S.E. = 41) (RFD 1999). Figure 2.2 Mae Chaem Watershed showing the study site, Mae Hae Tai village, on the left. The series of small triangles represents soil-sampling points. Rectangles indicate other ICRAF (International Centre for Research in Agroforestry) research sites. (Source-ICRAF). 2 1 2.3 Vegetation The Mae Chaem Watershed is characterized by five major forest types; hill-evergreen (Castanopsis spp., Quercus spp.), hill-evergreen and pine (Pinus kesiya, P. merkusii), dry dipterocarp and pine, dry dipterocarp (Dipterocarpus spp., Shorea siamensis), and mixed deciduous forests (Tectona grandis or teak tree, Gmelina arbored). Generally, soils are slightly acidic (pH 4-6). Their texture is sandy clay loam in the hill evergreen forest, and sandy clay loam to sandy clay in the dry dipterocarp and mixed deciduous forests (Santisuk 1988). 2.4 Population and Land-Use More than 50% of the population in the north is ethnic or local Thai who are generally known as "Kon (people) Muang (local)" and who speak northern Thai dialect. Northern Thailand is also well known for hill-tribe minority groups, e.g. Karen, Lua, and Hmong. Geographically, the local Thai occupy the lowest portion of the landscape (300-700 m asl.), the Karen and Lua settled in the middle segment (700-1,000 m asl.), and the Hmong normally live above 1,000 m asl. They practice different kinds of farming, e.g. the Lua and the Karen usually farm rotationally with a combination of short cropping and longer fallow periods, while the Hmong normally practice long cropping followed by field abandonment. However, this division is, in fact, extremely dynamic, and has been changing in the last decades during rapid population increase and rural development, so one can no longer associate a certain hill-tribe with particular areas and farming practices (Bass and Morrison 1994; Rerkasem and Rerkasem 1994; Thomas 1995; Ganjanapan 1998). 22 One of the highly visible changes in land use is from subsistence shifting cultivation of rice and opium to permanent farming of market-oriented production of soybean and other subtropical vegetable and fruits. For those groups that still practice shifting cultivation, the fallow period has been significantly reduced (e.g. from more than 10 years to five or six years). 2.5 Mae Hae Tai Karen Village The study site is located in the Ban (village) Mae Hae Tai (south) Karen village (Figure 2.2) located about 40-km southwest of the Mae Chaem district center. About a quarter of the distance has decent paved road, another quarter has degraded paved roads with potholes, and the rest has unpaved roads. The unpaved roads get very muddy and slippery during rainy seasons making travel difficult. The average elevation of the village is 1,000 m asl. In 1997, Mae Hae Tai village had a total of 57 households (increased from 41 in 1986; Sujumnong 1980), and a total population of 343: 171 males and 172 females (Prungkaew 1998). Most village members are involved in shifting cultivation activities, and the total land under cultivation by the villagers is approximately 548 ha. They grow rain-fed upland rice for one year, leave the field to fallow for five years, and then return to crop again. On average, 91 ha of land are under upland-rice cultivation each year (Prungkaew 1998). Traditionally, the Karen grow both paddy and upland rice. The relative proportion of the two systems varies among the villages depending on geophysical attributes of each village. However, the Karen prefer to have as much paddy-rice areas as the physical characteristics allow because paddy rice commonly provides a better yield than the upland varieties, the paddy fields are 23 normally close to streams (Zinke et al. 1978; Rerkasem and Rerkasem 1994). Paddy rice is encouraged by the government policy of permanent settling of the villages, having permanent fields (in any forms) and of restricting the cutting of forests. The Thai Australia Highland Agricultural and Social Development Project (1990-91) reported that most villages under the project having land with 0-15% slopes already have fully developed paddy rice fields by their own initiatives (Rerkasem and Rerkasem 1994). However, at Mae Hae Tai village, the terrain is mainly on steep slopes ranging from 35% to 65% (Anongrak and Seramethakul 1999). As a result, the village only possesses 22 ha (or about 4% of their total cultivated area) of paddy-rice fields belonging to 18 households. Most fields are shifting cultivation on steep slopes. There is a government policy to grant land titles to persons with permanently settled fields even though it is likely that this will not happen in the near future (Rerkasem and Rerkasem 1994). A few villagers have begun to grow commercial crops, mostly cabbage, after upland rice or in separate fields. The produce is bought by the Hmong and lowland Thai from villages closer to the district center who have trucks for transporting the produce. Cabbage prices are influenced by the market demands in Chiang Mai and Bangkok and fluctuate greatly from year to year. 24 Chapter 3 Soils 3.1 Introduction Soils are an integral part of ecosystems. They serve as sink of organic and inorganic materials, which influence their fertility and subsequent crop yields. Consequently, knowledge of soil properties is essential to understanding flows of nutrients in shifting cultivation systems. This section describes physical and chemical characteristics of soils in shifting cultivation fields of Mae Hae Tai village. In shifting cultivation, once a secondary forest is cleared, burned, and cropped, the area is thought by some to be nutritionally exhausted due to nutrient export via harvested material (rice grain) and rapid decomposition of organic matter during the cropping period. Soil fertility has been thought to decline for the first few years of the fallow period, and then to increase toward the original precropping level (Sanchez 1976; Zinke et al. 1978). Based on this traditional perspective, it was hypothesized that organic matter levels in the soils of the shifting cultivation fields wil l decline after the harvest and then increase as a consequence of litterfall of fallow vegetation. Soil pH should then increase due to ash addition from burning and should decrease as a result of litterfall from fallow vegetation and subsequent decomposition. Soil nitrogen, which is mostly associated with organic matter, should follow the same trend as soil organic matter. If this hypothesis is correct reflection of soil change in response to the shifting cultivation practice of the Karen, there should be a critical length of fallow period below which soil properties wil l not have time to recover, and the productivity of the system should decline. 25 3.2 Objectives 1) To document physical and chemical properties of soils from shifting-cultivation fields and forests of Mae Hae Tai village, and changes across a chronosequence of fields. 2) To confirm any patterns revealed under objective # 1 by remeasuring all the chronosequence sites after one year. (This temporal replication was undertaken because it was not possible to conduct spatial replication). 3) To examine effects of burning on soil pH, mineralizable nitrogen, and phosphorus. 3.3 Methods 3.3.1 Soil Sampling Soils in the mountainous landscape of northern Thailand are often omitted from general soil survey studies. This is mainly because it is rather difficult to access such areas, and those soils tend to be taxonomically complex due to slope and micro-climatic effects. As a result, areas with slopes of more than 35% are classified as "slope complex" without further classification (Forsyth 1996), and this represents about 70% of the total land area in the northern region of Thailand (Hiranburana 1996). The chronosequence research strategy was employed because time was limited. The chronosequence approach assumes that time is the only difference between the sample sites; they are inherently similar and have similar management regimes. Thus, spatial sampling is used to replace sampling over time. Cropping and fallow fields, along with secondary forests, were selected to represent different stages of the shifting cultivation cycle. Five fields were selected: 1998 rice field, 2-year, 4-year, 5-year fallows, and secondary forest (Figure 3.1). Figure 3.1 Mae Hae Tai village showing sampling sites (circles). The approximate location of the burial forest is shown (star). (Source: International Centre for Reseach in Agroforestry, ICRAF) 27 Approximate horizontal distances between the village site and the sampling fields were 3 km (the 1998 rice field), 1.6 km (the 2-year fallow), 3.2 km (the 4-year fallow), 1.1 km (the 5-year fallow), and 0.5 km (the secondary forest). Prior to field selection, I had some informal conversations with the villagers regarding spatial variability of the fields to ensure that I had a representative chronosequence of fields. Most of the villagers thought that the 3-year fallow (in 1998) had more fertile soils than the other fields, which were said to have similar fertility, and therefore the 3-year fallow was not included in my chronosequence. More formal discussions with the farmers about their farming knowledge were also conducted during the course of the field study, and they are reported in Appendix 10. The selected fields, which had similar elevation, aspect, and slope, have been repeatedly used for shifting cultivation judging from multiple-stem fallow trees resulting from repeated cuttings. Therefore, field age should be the major factor influencing variation in soil and vegetation processes. However, the 2-year fallow field was terraced once about 30 years ago, which might cause slight variations of soil properties. In addition to the main chronosequence sites, I included a former shifting-cultivation field which the villagers had abandoned to allow it to return to permanent forest in 1994 following an agreement with CARE-Thailand (an NGO working in natural resource management in the area) and several governmental organizations (Prungkaew 1998). This field was located further away from the village than the other fields (approximately 4 km), but was included in the study because it was the same age as the 4-year fallow field but was located on a different aspect. This field is referred to as the 4-year-fallow-SW (as opposed to the 4-year-fallow-NW in the 28 chronosequence) because it has a southerly aspect. A secondary forest that has not been subject to shifting cultivation, was included in the soils study. The first soil sampling was done after rice harvesting in November and December, 1998. In each field, one 40x40 m plot was divided into lower, middle, and upper slope positions. One soil pit of 1 m was excavated at each slope position, approximately 20 m from each side (Figure 3.2). Soil samples were collected from each pit at 0-5, 5-15, 15-30, 30-40, 40-60, 60-80, and 80-100 cm depths for determination of chemical and physical properties. The soil profiles at mid-slope positions were described by soil-survey personel from the Faculty of Agriculture, Chiang Mai University. 40 m • • • Upper slope Middle slope Lower slope 40 m Figure 3.2 Three soil pits (small squares) in each shifting cultivation field. 3.3.2 Soil Chemical Analyses For analyses of major nutrients, soil samples were collected at 0-5, 5-15, and 15-30 cm depths from five positions at 2-m intervals to the left and five positions to the right of each soil pit, totalling 10 samples for each depth. The soil samples were air-dried and passed through a 2-mm sieve (note that the coarse fragment content was extremely low), and then combined into three 29 sets, namely points 1-3, 4-6, and 7-10, for chemical analyses. Analysis methods used were as followed: texture by the wet sieving and pipette method (Gee and Bauder 1986), pH (1:1 in water and KC1) (McLean 1982), organic matter (O.M.) by wet oxidation (Walkley and Black 1934), total N (micro Kjeldahl; Jackson 1958), available P by the Kurtz and Bray II method (Bray and Kurtz 1945), extractable K (extracted by 1 M ammonium acetate at pH 7.0, and read by flame photometry), extractable Ca and Mg (extracted by 1 M ammonium acetate at pH 7.0, and read by atomic absorption spectrophotometry), CEC (extracted by 1 M ammonium acetate at pH 7.0 and NaCl; Thomas 1982), and extractable Fe (extracted by 1 M ammonium acetate at pH 4.8, and read by atomic absorption spectrophotometry). A l l soil analyses were done at the Faculty of Agriculture's Central Laboratory, Chiang Mai University. It was initially intended to analyze for soil A l because it plays an important role in soil acidity and P availability. However, this was not done due to a lack of instrumentation for A l analysis at the Laboratory. 3.3.3 Measurement of Mineralizable Nitrogen The total N concentration of soil is only broadly indicative of soil nitrogen status. About 90% of soil N occurs in plant-unavailable organic forms (Foth and Ellis 1997). Consequently, measuring available inorganic forms of soil nitrogen (i.e. N d V - N and N H / - N ) provides more insight into its availability to plants than total N . The buried-bag incubation (Eno 1960) was conducted to measure soil N0 3 " -N and N H 4 -N . Soils were collected from the following fields: the rice fields, the 2-year fallow field, the 4-year fallow field, the 5-year fallow field, and the secondary forest. In each field, one set of five plastic pipes of 5 cm in diameter and 5 cm long was pushed into the soil to collect the soils (0-5 cm) from the middle slope position. Next, the pipes were put in double polythene plastic bags (a preliminary experiment using single-layered plastic bags failed 30 because of leaking, due to damage to the bags caused by soil animals). A l l 35 pipes were buried in their bags at the sampling sites. Small amounts of soils were also taken adjacent to the soil core collection areas to be analyzed for initial N C V - N and N H / - N concentrations. A l l soil samples were extracted with 2 M KC1, and analyzed for N0 3 " -N and N H 4 + - N by the hydrazine method (APHA, A W W A , and WPCF 1985). The final concentration of N 0 3 " - N + N H 4 + - N less the initial amounts divided by 30 (the days of incubation) were considered to be a measure of net nitrogen mineralization, and of net nitrification for the comparable N 0 3 " N data (Hart et al. 1994). 3.3.4 Burning Effects on Mineralizable Nitrogen Burning transforms some of the nutrients in slash biomass into plant-available forms, but can also reduce the total inventory of nutrients that have gaseous oxides, e.g. nitrogen and sulfur (Christanty 1986; Andriesse and Schelhaas 1987b; Kleinman et al. 1995). However, by removing litter and slash (energy sources for microbes), burning can reduce immobilization of the remaining mineral N . To evaluate this loss, ten soil samples were collected at 0-5 cm depth from each slope position seven days before and after burning and analyzed for N G Y - N and N H / - N concentrations. A l l soil samples were extracted with 2 M KC1, and analyzed for N 0 3 " - N and N H 4 + - N by the hydrazine method (APHA, A W W A , and WPCF 1985). 3.3.5 Atmospheric Inputs of Nutrients Deposition of nutrients via precipitation (rainfall in the tropics) is considered important for some nutrients (e.g. potassium and sulphur). For nitrogen and phosphorus, the amounts can be either significant (in natural ecosystems) or insignificant (in agricultural systems) in comparison with nutrients exported in harvested materials. Ewel (1986) reported inputs of N (ammonium and 31 nitrate) and phosphorus from rainfall for most places of 5-18 kg/ha/year, and less than 1 kg/ha/year, respectively. Rainfall in the open field (the rice field, after harvest), and throughfall in two partially open fields (the 2-year fallow and the 4-year fallow (SW)), and the secondary forest, was collected monthly using 5,000-cm3 plastic bottles in August and September, 1999. The bottles were attached with funnels and plastic screens to prevent litter and animals falling into the bottles. The rainfall and throughfall water was analysed for pH, and concentrations of M V - N and N H / - N (the hydrazine method), total P (digested by persulphate acid, developed by ascorbic acid, and read by spectrophotometry), K and Na (read by flame photometry), and Ca and M g (read by atomic absorption spectrophotometry). 3.3.6 Remeasurement of Selected Soil Properties after One Year Because of the high variability of slopes and aspects in the shifting cultivation fields of Mae Hae Tai village, it was difficult to find sites of the same age and site characteristics to provide a replicate chronosequence, which is crucial for reliable statistical analyses. As a result, I chose to re-measure some soil chemical properties (pH, N , P, K , Ca, Mg, and CEC) one year later (October, 1999) (temporal variation). This second soil sampling was done only at the mid-slope position because there were no differences among slope positions in the 1998 samples. Soils were re-measured in the following fields: 1-year fallow (previously 1998 rice field), 3-year fallow (previously 2-year fallow), 5-year fallow (previously 4-year fallow), and 1999 rice-field (previously 5-year fallow). In addition, I selected a different patch of secondary forest for this 1999 measurement. This forest is destined to be a burial (or cementary and sacred) forest for 32 Mae Hae Tai which is a Protestant village. The forest is relatively undisturbed. A l l data were analyzed by A N O V A and LSD (Least Significant Difference) using the SAS system for Windows Release 6.12 (SAS Institute 1996). 3.4 Results and Discussion A l l the shifting cultivation fields had comparable aspect, elevation, and drainage class, except the 4-year-fallow-SW that has a southern aspect (Table 3.1). Slopes ranged from 38 to 65%. The surface-soil pH values were similar among fallow fields, but significantly higher in the cultivated field and lower in the secondary forest. Table 3.1 Characteristics of a chronosequence of shifting cultivation fields of Mae Hae Tai village Field Elevation Aspect Slope Soil pH Drainage Class (in 1998) (m asl.) (%) (1:1 in Water) (0-5 cm) Rice Field 1,060 WNW 55 5.99a (0.14) Moderately well 2-year Fallow 1,080 N N E 38 5.02b (0.07) Moderately well 4-year Fallow-NW 1,020 N N W 50 5.07b (0.05) Moderately well 4-year Fallow-SW 940 ssw 65 5.81a (0.12) Well 5-year Fallow 1,060 ENE 39 4.92b (0.06) Moderately well Secondary Forest 1,100 N N W 45 4.58c (0.05) Imperfectly to moderately well 1) The pH data were calculated across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Different letters indicate significant differences (p<0.05) among fields. 3.4.1 Soil Texture Soil texture influences a number of important soil properties such as drainage, water holding capacity, and particle charges. Most soils of Mae Hae Tai village were clayey with clay contents of 35-47%), 43-58%, and 45-60% at 0-5, 5-15, and 15-30 cm depths, respectively (Table 3.2). The clay content increased with depth in the 0-30 cm layer, which is a characteristic of old tropical soils in which clay has been eluviated down from upper horizons to deeper horizons. The middle (B) horizons normally contain a higher amount of clay than the upper (A) and lower (C) horizons due to downward clay movement and deposition, and products of weathering processes (Troeh and Thompson 1993). Table 3.2 Soil texture at three depths in a chronosequence of shifting cultivation fields Field Depth Sand Silt Clay Textural Class (in 1998) (cm) (%) (%) (%) Rice Field (after harvest) 0-5 5-15 15-30 34(1.7) 25 (1.4) 22 (0.7) 29 (0.9) 32 (2.3) 25 (1.7) 37(1.0) 43 (1.4) 53 (1.6) Clay Loam Clay Clay 2-year Fallow 0-5 5-15 15-30 31 (2.7) 29 (2.4) 24(1.8) 24 (2.6) 25 (2.5) 26 (2.4) 45 (3.1) 46 (3.0) 50 (2.4) Clay Clay Clay 4-year Fallow-NW 0-5 5-15 15-30 24(1.6) 15 (0.7) 12(1.7) 29(1.1) 27(1.2) 27 (0.7) 47(1.1) 58 (1.2) 61 (1.7) Clay Clay Clay 4-year Fallow-SW 0-5 5-15 15-30 33 (2.4) 27 (2.3) 22(1.9) 31 (1.6) 29(1.6) 33 (0.8) 36 (2.0) 44 (2.3) 45 (2.6) Clay Loam Clay Clay 5-year Fallow 0-5 5-15 15-30 27(1.6) 22(1.9) 16(2.1) 33 (2.4) 35 (1.0) 33 (1.7) 40 (2.3) 43 (1.3) 51 (1.9) Clay Loam Clay Clay Secondary Forest 0-5 5-15 15-30 35 (1.2) 27 (0.8) 22(1.1) 26 (0.7) 26(1.1) 29 (2.3) 39(1.6) 47(1.7) 49 (2.4) Clay Loam Clay Clay Numbers in parentheses are standard errors (n=9). Some fields had clay-loam surface soils (0-5 cm) with slightly less clay content, (e.g. the 1998 rice field, 5-year fallow field, and the secondary forest), but clay still constituted more than 35% of the texture component. Funakawa et al. (1997a) reported a slightly lower clay content of the soils of Karen shifting-cultivation fields located west of the study site of this thesis. Their study area was at 18° 20' N and 98° E at elevations of 700-850 m asl., and the soils were in the same order, Ultisols. Their 7-year fallow field soils had about 25% clay content at the surface and the 34 clay contents also increased with depth. Clay contents of soils are closely related to their fertility because clay particles are responsible the water-holding capacity and the surface charge, which are key determinants of soil fertility and productivity. 3.4.2 Soil Bulk Density Bulk density and particle density are the determinants of total porosity of soils (Troeh and Thompson 1993), which affects compaction, drainage, and water holding capacity of the soils. However, only bulk density was determined in this thesis; it is assumed that it is a broad indicator of soil porosity and effects of cultivation thereon. The mean bulk density of the cropping and fallow fields (about 1.0 g/cm3) was slightly higher than that of the secondary forest (about 0.9 g/cm3; Table 3.3). At depths below 15 cm, the soil of the 1998 rice field had slightly higher bulk density than the other fields; this is likely due to the inherent property of the soil of the field and not a consequence of land uses. Cultivation practice should only affect the top layer of the soil, especially in shifting cultivation practice, in which hand tools are used to dig shallow holes for rice planting without major soil disturbance by mechanical tools. Table 3.3 Bulk density (g/cm ) of soils from shifting cultivation fields Field 0-5 5-15 15-30 30-40 40-60 60-80 80-100 (in 1998) cm cm cm cm cm cm cm Rice Field 1.04 1.13 1.26 1.39 1.28 1.24 1.40 (after harvest) 2-year Fallow 1.00 0.91 0.86 1.02 1.02 1.06 1.10 4-year Fallow-NW 1.02 1.09 1.04 1.10 1.05 1.17 1.16 4-year Fallow-SW 1.12 1.10 1.18 1.22 1.17 1.20 1.26 5-year Fallow 1.04 1.16 1.18 1.26 1.16 1.30 1.21 Secondary Forest 0.91 0.94 1.05 1.16 1.27 1.27 1.33 35 In general, cultivated soils tend to have a higher bulk density than non-cultivated soils (e.g. forest or grassland soils) because of loss of organic matter (Troeh and Thompson 1993). Cultivated soils, especially under continuous cropping, are affected by the direct impact of rain and, in some cases, by mechanical harvesting methods. In addition, the use of chemical fertilizers in continuous cropping may "harden" the soil (due to prolonged use of the fertilizers) resulting in higher bulk density, greater compaction, and reduced water infiltration. The soils in the cropping phase of shifting cultivation are also affected by direct impact of rain, but not by chemical fertilizers and mechanical harvesting, and the cropping period lasts only one year. From my data, it appears that the shifting cultivation practice of the Karen has had little or no impact on bulk density and soil compaction. Shifting cultivation has often been blamed for causing soil erosion in northern Thailand. However, it is shown here that soils of the shifting cultivation fields examined were comparable to the forest soils in terms of soil bulk density, which reflects soil compaction, and determines infiltration and erosion susceptibility. Ziegler and Giambelluca (1997) reported that roads might be more responsible for soil erosion than agricultural fields due to their compacted surfaces and because the networks of roads act as connectors for rapid downslope waterfiow. Visual evidence of erosion in Mae Hae Tai village appeared to be restricted to roads (with some slight rill and sheet erosion within the shifting cultivation fields; Anongrak and Seramethakul 1999), supporting this conclusion. 36 3.4.3 Soil pH The secondary forest in my study had the lowest pH throughout the top 30 cm, and the cropped field had the highest pH (Table 3.4). The latter is attributable to the ash effect from burning prior to rice planting. A l l of the fallow fields were of comparable pH. However, the pH at 0-5 cm depth apparently decreased slightly as the fallow aged. The addition of decomposing materials can render soils more acidic because of the release of organic acids, and root uptake of cations (e.g. Ca) also lowers pH (Troeh and Thompson 1993). Table 3.4 Soil pH (1:1 in water) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 5.99 (0.14)a 5.66 (0.1 l)a 5.79 (0.18)a 2-year Fallow 5.02 (0.07)b 4.78 (0.06)c 4.67 (0.06)c 4-year Fallow-NW 5.07 (0.05)b 4.65 (0.04)c 4.59 (0.04)cd 4-year Fallow-SW 5.81 (0.12)a 5.25 (0.10)b 4.93 (0.04)b 5-year Fallow Field 4.92 (0.06)b 4.70 (0.05)c 4.61 (0.04)cd Secondary Forest 4.58 (0.05)c 4.43 (0.02)d 4.43 (0.02)d 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significantly differences (P<0.05) among fields. Generally, old tropical soils (e.g. Ultisols and Oxisols) are acidic because a high proportion of their basic ions have moved down the profile via weathering and leaching processes. Uitlsols and oxisols, which are common in the tropics, are both relatively highly weathered. They are considered to be older and more weathered than their temperate counterparts. This has many important implications including lower cation exhange capacity (CEC) and phosphorus availability. In northern Thailand, highland soils generally have pH values between 5.3-5.6 (Hiranburana 1996). 37 The pH measured in 1:1 KC1 (Table 3.5) followed the same trend as pH measured in water. Differences between soil pH in KC1 and in water were all negative indicating net negative charges of the clay particles (Sanchez 1976). Permanent negative charge was found to be predominant in highland Ultisols soils with pH below 6 in northern Thailand (Wada and Wada 1985; Funakawa et al. 1997a). The ion exhange system is related to soil CEC. Table 3.5 Soil pH (1:1 in KC1) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 5.36(0.18)a 5.11 (0.19)a 5.03 (0.24)a 2-year Fallow 4.30 (0.06)b 4.04 (0.06)c 3.90 (0.07)bc 4-year Fallow-NW 4.30(0.10)b 3.91 (0.05)cd 4.07 (0.1l)b 4-year Fallow-SW 5.14(0.13)a 4.60 (0.12)b 4.25 (0.09)b 5-year Fallow 4.26 (0.07)b 4.01 (0.05)c 3.81 (0.03)bc Secondary Forest 3.90 (0.05)c 3.65 (0.02)d 3.58 (0.03)c 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significant differences (PO.05) among fields. The first measurement of pH in 1998 and the re-measurement in 1999 are shown in Tables 3.6 and 3.7. In general, soil pH changed very slightly after one year. However, the biggest change was from the 5-year fallow to the cropped field. This is interpreted to be the effect of burning and ash addition. Soil pH reported from other studies in northern Thailand (Nakano 1978; Sabhasri 1978; Andriesse and Schelhaas 1987a; Funakawa et al. 1997a) followed the same trend: the highest pH occurred in the cropped field, and pH declined gradually toward the secondary-forest stage. 38 Table 3.6 Soil pH (1:1 in water) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1998 5.74 (0.12) 5.49 (0.05) 5.89 (0.52) 1-year Fallow 1999 6.46 (0.27) 5.83 (0.21) 5.70 (0.03) 2-year Fallow 1998 5.07 (0.05) 4.86 (0.05) 4.74 (0.09) 3-year Fallow 1999 5.30 (0.06)* 5.21 (0.05)* 5.25 (0.05)* 4-year Fallow 1998 5.20 (0.01) 4.71 (0.07) 4.64 (0.07) 5-year Fallow 1999 5.16(0.14) 5.14(0.10)* 5.07 (0.08)* 5-year Fallow 1998 4.98 (0.11) 4.80 (0.09) 4.70 (0.07) Rice Field 1999 5.72 (0.17)* 5.55 (0.08)* 5.37 (0.02)* 1) The data were from middle slope position. 2) Numbers in parentheses are standard errors (n= =3). 3) Asterisks indicate a significant (PO.05) difference between 1998 and 1999 measures. Table 3.7 Soil p H ( l : l in KC1) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1998 5.26 (0.19) 5.16(0.12) 5.39 (0.58) 1-year Fallow 1999 5.97 (0.24) 5.23 (0.20) 4.84 (0.07) 2-year Fallow 1998 4.29 (0.06) 4.10(0.08) 3.95 (0.11) 3-year Fallow 1999 4.49 (0.05) 4.33 (0.05)* 4.33 (0.06)* 4-year Fallow 1998 4.39 (0.25) 3.94 (0.12) 4.12(0.22) 5-year Fallow 1999 4.68 (0.10) 4.37 (0.08)* 4.26 (0.05) 5-year Fallow 1998 4.24 (0.16) 4.02 (0.14) 3.85 (0.04) Rice Field 1999 4.92 (0.21) 4.59 (0.11)* 4.36 (0.01)* 1) The data were from middle slope position. 2) Numbers in parentheses are standard errors (n=3). 3) Asterisks indicate a significant (P<0.05) difference between 1998 and 1999 measures. 3.4.4 Effects of Burning on Soil pH In this study, burning increased soil pH at the surface (0-5 cm) from approximately 5.2 to 5.6, while the study by Sabhasri (1978) showed an increase from 6.2 to 6.4. Andriesse and Schelhaas (1987b) reported increases of pH after slash burning from 5.6, 5.4, and 5.3 to 6.0, 5.8, and 5.5 at 0-5, 5-10, and 10-25 cm soil depths, respectively. The increase in pH was largest in the surface soil where the liming effect of the ash was the greatest. 39 One of the benefits of burning is to increase soil pH, which is particularly important for acidic tropical soils. Other advantages of burning are insect eradication and weed reduction, and decreasing A l saturation. The latter is important for tropical acidic soils i f they are to be used for agriculture because high levels of A l can immobilize phosphorus in soils (Kleinman et al. 1995). Even though the slash is burned rather quickly, the temperature at the soil surface can reach 700 °C during an average slash burn (Andriesse and Schelhaas 1987b), which could affect chemical and physical properties of soils. 3.4.5 Soil Organic Matter Soil organic matter (SOM) is considered to be an indicator of soil fertility because it plays a critical role in supplying available plant nutrients and maintaining good soil structure. Especially in unfertilized soils, S O M supplies most of the available nitrogen and sulfur, most of the CEC especially in highly weathered acid soils, and half of the available phosphorus. S O M also helps to decrease P-fixation and improve soil structure and water retention capacity (Sanchez 1976). Hiranburana (1996) reported that highland soils in northern Thailand generally have about 3.5-5% organic matter in the surface horizon. Soil organic matter concentrations of the surface soils of shifting-cultivation fields of Mae Hae Tai village (5.8-6.7%, and almost 8% in the secondary forest) exhibited higher values than this (Table 3.8). Organic matter was highest in the secondary forest due to continuous addition of litterfall. SOM concentrations of the cropped and the fallow fields were comparable to each other. After one year, S O M changed slightly (Table 3.9). The largest decrease of SOM occurred when the 5-year fallow was converted to the cropping field, 40 but the change was not significant (P<0.05) due to the high variability of S O M in the cropped field, which was the result of patchy burning and variable mixing by cultivation. Table 3.8 Soil organic matter and organic carbon concentrations at three depths in a chronosequence of shifting cultivation fields Field (in 1998) Organic Matter (%) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) Total O.M. Organic Carbon 6.75 (0.44)ab 3.92 4.44 (0.20)ab 2.58 2.70 (0.13)c 1.57 2-year Fallow Total O.M. Organic Carbon 5.83 (0.32)b 3.38 5.25 (0.53)a 3.05 4.47 (0.22)a 2.59 4-year Fallow-NW Total O.M. Organic Carbon 6.70 (0.3l)b 3.89 4.42 (0.17)ab 2.56 3.58 (0.22)b 2.08 4-year Fallow-SW Total O.M. Organic Carbon 6.30 (0.53)b 3.65 3.96 (0.4l)b 2.30 2.94 (0.25)c 1.71 5-year Fallow Total O.M. Organic Carbon 6.46 (0.37)b 3.75 4.06 (0.16)b 2.35 3.07 (0.30)bc 1.78 Secondary Forest Total O.M. Organic Carbon 7.87 (0.42)a 4.56 4.79 (0.17)a 2.78 3.60 (0.14)b 2.09 1) Organic carbon was calculated as 58% of organic matter (Anderson and Ingram 1993). 2) The data were averaged across slope positions because there were no differences among slope positions. 3) Numbers in parentheses are standard errors (n=9). 4) Within each soil depth, different letters indicate significant differences (P<0.05) among fields. Table 3.9 Soil organic matter concentrations (%) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1998 6.43 (0.33) 4.55 (0.28) 2.76 (0.23) 1-year Fallow 1999 6.07(1.21) 4.03 (0.46) 2.44 (0.33) 2-year Fallow 1998 5.29 (0.16) 4.53 (0.10) 4.00 (0.19) 3-year Fallow 1999 6.02 (0.49) 4.74 (0.75) 4.27 (0.97) 4-year Fallow 1998 7.46 (0.60) 4.64 (0.51) 3.89 (0.69) 5-year Fallow 1999 6.29 (0.09) 4.48 (0.05) 3.10(0.29) 5-year Fallow 1998 7.22 (0.06) 4.00 (0.37) 2.61 (0.32) Rice Field 1999 5.78 (0.76) 4.07 (0.46) 3.15 (0.16) 1) The data were from middle slope-position. 2) Numbers in parentheses are standard errors (n=3). 3) Asterisks indicate a significant (P<0.05) difference between 1998 and 1999 measures. 41 3.4.6 Soil Nitrogen Nitrogen (N), which is regarded as one of the most important plant macro-nutrients, is added to plant-soil systems by way of rain, dust, and symbiotic and non-symbiotic fixation (Sanchez 1976). It is a constituent of chlorophyll, amino acids, and proteins, which results in a high demand for N by vegetation. Even though the atmosphere is about 78% N , many ecosystems are nitrogen-deficient because it occurs in the plant-unavailable form of N2 gas. Biological fixation constitutes the major input of N , accounting for 72% of the total nitrogen inputs (Foth and Ellis 1997). There has been no report on nitrogen-fixing plants (e.g. legumes) in shifting cultivation systems in Thailand. Total N concentrations in soils varied little among the chronosequence sites (Table 3.10). The differences between samples within the same field were also low. The secondary-forest soils contained higher concentration of N than the soils of the other fields. The changes in total N concentrations after one year showed no differences, except when the 5-year fallow field was cleared, burned, and cropped (Table 3.11). Table 3.10 Soil total N concentrations (%) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 0.31 (0.02)b 0.21 (0.01)a 0.13 (0.01)d 2-year Fallow 0.26 (0.02)b 0.25 (0.02)a 0.21 (0.01)a 4-year Fallow-NW 0.31 (0.01)b 0.21 (0.01)a 0.17 (0.01)bc 4-year Fallow-SW 0.30 (0.03)b 0.18(0.02)a 0.14(0.01)cbd 5-year fallow 0.31 (0.02)b 0.20 (0.01)a 0.14(0.01)cd Secondary Forest 0.37 (0.02)a 0.23 (0.01)a 0.17 (0.0l)b 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significant differences (PO.05) among fields. 42 Table 3.11 Soil total N concentrations (%) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1998 0.29 (0.03) 0.22 (0.01) 0.13 (0.01) 1-year Fallow 1999 0.33 (0.03) 0.23 (0.01) 0.19(0.02) 2-year Fallow 1998 0.24 (0.01) 0.22 (0.01) 0.20 (0.01) 3-year Fallow 1999 0.25 (0.02) 0.22 (0.03) 0.22 (0.04) 4-year Fallow 1998 0.34 (0.02) 0.23 (0.02) 0.18(0.03) 5-year Fallow 1999 0.32 (0.01) 0.28 (0.02) 0.20 (0.02) 5-year Fallow 1998 0.34 (0) 0.20 (0.02) 0.12(0.01) Rice Field 1999 0.27 (0.03)* 0.22 (0.01) 0.18(0.01) 1) The data were from middle slope position 2) Numbers in parentheses are standard errors (n=3). 3) Asterisks indicate a significant (P<0.05) difference between 1998 and 1999 measures. 3.4.7 Mineralizable Nitrogen Results from the soil incubation were highly variable, but the net changes of NCV-N and NFLi+-N suggested net immobilization of N (Table 3.12). One possible explanation is that the plastic bags used in this experiment were relatively thicker than the plastic bags used elsewhere, e.g. in British Columbia, and it might have limited gas exchange, which could have resulted in small mineralization of N . Factors controlling N mineralization rates include C:N ratio, soil pH, clay mineralogy, and soil moisture (Sanchez 1976). The rate of N mineralization from organic matter is highest at pH between 6 and 8 (Troeh and Thompson, 1993), while pH of the soils in this study was between 4.6-6. In the tropics, soil moisture has been reported to be the dominant factor influencing N mineralization (Sanchez 1976). The buried-bag technique does not take moisture fluctuations into account, unlike the more advanced resin-core method (Zou et al. 1992). The buried-bag technique was employed in this work because it was economically feasible. 43 Table 3.12 Soil N mineralization during a 30-day field incubation in a chronosequence of shifting cultivation fields Field (in 1998) Initial N0 3 " -N Final N0 3 " -N Net N e t N (Hg) (fig) Nitrification Mineralization (Hg/g/day) (Hg/g/day) Rice Field I* 161(102) 585 14 n.a. Rice Field II** 543 (102) 727 6.1 n.a. 2-year Fallow 1,190 (98) 259 -31 n.a. 4-year Fallow-NW 236(100) 203 -1.1 n.a. 4-year Fallow-SW 1,175 (110) 537 -21 n.a. 5-year Fallow*** 194 (102) 161 -1.1 n.a. Secondary Forest 931 (89) 627 -10.1 n.a. Initial N H 4 + - N Final N H 4 + - N Net N (NH 4 + ) fog) (Mg) Mineralization (Hg/g/day) Rice Field I* 1,502 (102) 1,075 -14 0 Rice Field II** 960 (102) 678 -9.4 -3.3 2-year Fallow 1,033 (98) 1,035 0 -31 4-year Fallow-NW 1,553 (100) 1,156 -13.2 -14.3 4-year Fallow-SW 1,454(110) 1,186 -9 -30 5-year Fallow*** 1,293 (102) 1,230 -2.1 -3.2 Secondary Forest 1,142 (89) 1,189 1.5 -8.6 *Rice field I was the 1998 rice field. **Rice field II was the second-year rice experiment plot (see Chapter 4). ***The 5-year fallow (in 1998) was the 1999 rice field, n.a. = not applicable. Numbers in parentheses are soil dry weight (g). The soils in the plastic pipes in my study had not had the organic debris removed, and this may have contributed to the N immobilization in the fallow field samples (Ross et al. 1985; Schimel and Parton 1986). N is associated with S O M and its mineralization depends on soil microclimate (moisture and temperature). The soils of the shifting cultivation fields, in this and other studies in northern Thailand, contain relatively large amounts of O.M. (Table 3.8) in comparison to a reported average of 3.5-5% of all northern-Thailand highland soils (Hiranburana 1996). Fresh organic matter can stimulate short-term N immobilization by microbes, but eventually results in increased mineralization. The effects of the removal of litter by burning on N immobilization, and the effect of rice fine root and stem litter should also be investigated. In short, therefore, the 44 role of organic matter as a source of mineral nitrogen in shifting cultivation warrants further investigation. Further studies involving consecutive measures of N mineralization during the course of incubation should also be undertaken, but this was not possible in this thesis. Other studies of N mineralization in tropical soils had different soil types and climates, e.g. sandy (about 50% sand) savanna soils in India (Singh et al. 1991) or lowland moist-forest soils in Costa Rica (Zou et al. 1992). Therefore, a comparison could not be made. 3.4.8 Burning Effects on Mineralizable Nitrogen Burning decreased M L V - N concentration by about 50% (from 12 ppm (S.E. = 1.4, n=30) to 6.5 ppm (S.E. = 1.1, n = 30)), but increased NH4 + -N concentration by about 50% (from 12 ppm (S.E. - 0.8, n = 30) to 25 ppm (S.E. = 2.4, n = 30)). Singh et al. (1991) reported an increase in inorganic nitrogen after burning of savanna soils, with the increase in NFLi+-N being larger than N O 3 - N . They attributed the increase in inorganic nitrogen to the reduction of nitrogen uptake by plants caused by burning. Vegetation took up most of the soil available inorganic nitrogen in the pre-burning stage. 3.4.9 C : N Ratio The soil in the secondary forest had the highest level of SOM, about 8% (Table 3.8), and also the highest N concentration, about 0.37% (Table 3.10), which resulted in a C:N ratio of about 12, which is similar to that of the other fields (Table 3.13). Soil C and N are normally associated because about 90% of N is found in organic forms (Foth and Ellis 1997). 45 Table 3.13 Soil C :N ratio at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 12.8 (0.2) 12.3 (0.1) 12.4 (0.1) 2-year Fallow 12.8 (0.2) 12.4 (0.2) 12.2 (0.1) 4-year Fallow-NW 12.7 (0.2) 12.1 (0.2) 12.5 (0.3) 4-year Fallow-SW 12.3 (0.2) 12.6 (0.1) 11.9 (0.3) 5-year Fallow 12.1 (0.1) 12.0 (0.2) 12.5 (0.1) Secondary Forest 12.3 (0.2) 11.9 (0.1) 12.3 (0.3) 1) The data were calculated across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). In other studies of shifting cultivation in northern Thailand, the C : N ratio of soils ranged from 13 (Nakano 1978; Funakawa et al. 1997a, 1997b) to 15 (Nakano 1978) in cropped fields. The secondary-forest soils studied by Andriesse and Schelhaas (1987a) had ratios between 15.5 and 19.5. Ultisols, the same soil order as in my study, in Brazil and Puerto Rico were reported to have C : N ratio of approximately 10 in the top 40 cm (Sombroek 1967; FAO-UNESCO 1971; Beinroth 1972 cited in Sanchez 1976). 3.4.10 Soil Available Phosphorus Tropical soils often have low phosphorus (P) availability for plants, which can be attributed to a variety of factors including soil acidity, and high soil aluminium and iron levels. The total P concentrations in soils can be high, but some forms are unavailable to plants because they are in organic forms or are bound with A l or Fe in acid soils, and Ca in basic soils (Foth and Ellis 1997). Hiranburana (1996) indicated that upland soils in northern Thailand have 10-15 ppm of available P (Bray II) in the surface layers. In this study, soil available P concentrations of the soils of the shifting cultivation fields were very low, about 1-6.4 ppm, except in the cropped field. In 46 Thailand, available P by Bray II is considered to be very low at < 10 ppm (Topark-ngarm 1983). The available P was higher in the harvested rice field than in the other fields (Table 3.14), but it was still considered to be low (a value between 10-30 ppm is considered low; Topark-ngarm 1983). The low availability of soil P should be attributed to a high P sorption capacity of the soils in the area. Hansen (1995) reported that, in a laboratory P sorption study, the 0-15 cm horizons from Ultisols in the highland of northern Thailand could adsorb up to 80% of added P, the highest P sorption capacity in his studies. The high P sorption capacity of the soils in northern Thailand can be mainly attributed to soil acidity. Table 3.14 Soil available P concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 22 (3.4)a 4.8 (0.8)a 3.3 (0.7)a 2-year Fallow 2.9 (0.4)b 1.7 (0.1 )bc 1.3 (0.2)b 4-year Fallow-NW 2.5 (0.2)b 1.7 (0.2)bc 1.6(0.2)b 4-year Fallow-SW 6.4(1.6)b 2.4 (0.2)b 1.9(0.3)b 5-year Fallow 2.9 (0.3)b 1.3 (0.1)c 1.1 (0.2)b Secondary Forest 2.2 (0.2)b 1.6 (0.2)bc 1.6(0.2)b 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significant differences (PO.05) among fields. 4) For Thailand, Bray II available P is high at > 90 ppm, low at 10-30 ppm, and very low at <10 ppm (Topark-ngarm 1983). The relatively higher available P in the cropped fields could be attributed to the burning (ash) effect and released from decomposing fine roots because the soils were sampled about one month after the rice harvest. In this study, burning increased soil available P in surface soil (0-5 cm) from 4.6 ppm (S.E. = 0.4, n = 15) to 9 ppm (S.E. = 0.6, n = 15). Andriesse and Schelhaas (1987b) also reported that burning increased soil available P (Bray II) from 18, 10, and 4 to 36, 27, and 8 ppm in 0-5, 5-10, and 10-25 cm soil depths, respectively. The repeated measurement of P after one year showed no differences (Table 3.15). 47 Table 3.15 Soil available P concentrations (ppm) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1998 24 (8.3) 3.8 (0.9) 3.2 (0.9) 1-year Fallow 1999 22 (10) 3.3 (1.0) 2.5 (0.5) 2-year Fallow 1998 3.3 (0.7) 1.5(0) 1.3 (0.3) 3-year Fallow 1999 2.2 (0.2) 1.3 (0.2) 1.2 (0.4) 4-year Fallow 1998 2.8 (0.2) 1.7(0.3) 1.5(0) 5-year Fallow 1999 3.3 (0.2) 2.5 (1.0) 1.8(0.3) 5-year Fallow 1998 3.3 (0.6) 1.3 (0.3) 0.8 (0.2) Rice Field 1999 3.5 (0.8) 2.0 (0.5) 1.7(0.3) 1) The data were from middle slope position. 2) Numbers in parentheses are standard errors (n=3). 3) Asterisks indicate a significant (PO.05) difference between 1998 and 1999 measures. Unlike N , P is not commonly found in gaseous forms, so loss of P via volatilization (as in burning) is rare. Losses of soil P are associated with losses of organic matter. Solubility and forms of P depend on soil pH (Troeh and Thompson 1993). As can be seen in the correlation analyses of all soil chemical properties later in this chapter, soil pH is highly correlated with available P (P<0.01). In the acid soils in this study (pH 4.6-6 at 0-5 cm), the dominant form of P in soil solution is FLvPOV, which is often bound with A l or Fe, rendering it unavailable to plants (Troeh and Thompson 1993). Soil available P was negatively correlated with extractable Fe (PO.01). 3.4.11 Soil Extractable Potassium Potassium (K) is often the third most limiting plant nutrient after N and P. A high concentration of K , > 200 ppm, is common in the topsoil of the highland areas of northern Thailand (Hiranburana 1996). Topark-ngarm (1983) indicated a level of 100 ppm of soluble plus exchangeable K to be low for Thailand soils. Therefore, it is generally uncommon to have soils with K deficiency in this part of the country. The cropped soils of Mae Hae Tai village contained 48 up to 360 ppm of extractable K in the surface horizon (Table 3.16), which is probably a residual effect of burning. Andriesse and Schelhaas (1987b) reported that burning slash increased extractable K about 18-fold. The gradual decrease in potassium concentrations as the fallow ages is thought to be a result of uptake by fallow vegetation (Tanaka et al. 1997). The repeated measurement showed no differences after one year (Table 3.17). Table 3.16 Soil extractable K concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 360 (17)a 292 (13)a 313 (20)a 2-year Fallow 141 (10)c 100 (12)d 93 (6)d 4-year Fallow-NW 195 (23)b 143 (19)bc 127 (20)cd 4-year Fallow-SW 347 (19)a 284 (18)a 248(18)b 5-year Fallow 209 (16)b 173 (13)b 150(14)c Secondary Forest 193 (14)b 117(5)cd 96 (5)d 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significant differences (P<0.05) among fields. Table 3.17 Soil extractable K concentrations (ppm) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1-year Fallow 1998 1999 315 (2) 321 (78) 263 (12) 261 (34) 353(19) 336 (35) 2- year Fallow 3- year Fallow 1998 1999 129 (4) 141(10) 76 (6) 101 (8)* 76 (10) 90 (8) 4- year Fallow 5- year Fallow 1998 1999 243 (25) 259 (86) 181(40) 173 (66) 160 (58) 156 (68) 5-year Fallow Rice Field 1998 1999 233 (22) 243 (23) 198 (21) 224 (23) 176 (28) 204(19) 1) The data were from middle slope position. 2) Numbers in parentheses were standard errors (n=3). 3) Asterisks indicate a significant (P<0.05) difference between 1998 and 1999 measures. 49 3.4.12 Soil Extractable Calcium Calcium (Ca) and magnesium (Mg) are normally sufficient in most soils, but Ca is occasionally added to highly acid soils to raise pH (Troeh and Thompson 1993). Burning transforms Ca in forest biomass into exchangeable soil Ca. Topark-ngarm (1983) indicated that, in Thailand, the combined concentration of soluble and exchangeable Ca and Mg is considered to be very low at 100 ppm. The soils of the cropped field of Mae Hae Tai village contained almost 1,900 ppm of extractable Ca, and the concentration gradually declined toward the secondary-forest stage with the exception of the high value in the 4-year fallow (SW) (Table 3.18). The reason for this high value, other than due to the warmer aspect, is not known. The declining soil Ca should reflect continuous uptake of Ca by growing vegetation in the fallow. Andriesse and Schelhaas (1987b) reported that burning increased exchangeable Ca about 15-fold. Soil extractable Ca decreased slightly after one year of fallow, and substantially increased when the 5-year fallow was cleared and burned (Table 3.19). Table 3.18 Soil extractable Ca concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 1,847 (149)a 1,297 (138)a 867(119)a 2-year Fallow 417(33)cd 246 (32)bc 175 (31)b 4-year Fallow-NW 665 (75)c 526 (279)b 178 (21)b 4-year Fallow-SW 1,122 (154)b 457 (79)bc 203 (35)b 5-year Fallow 308 (58)d 139 (26)c 161 (49)b Secondary Forest 182 (36)d 51 (6)c 42 (6)b 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significant differences (PO.05) among fields. 50 Table 3.19 Soil extractable Ca concentrations (ppm) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1998 1,633 (102) 1,138 (176) 771(170) 1-year Fallow 1999 1,625 1,050 663 2-year Fallow 1998 446 (83) 291(22) 192 (49) 3-year Fallow 1999 425 175 200 4-year Fallow 1998 813 (36) 250 (25) 175(13) 5-year Fallow 1999 500 188 113 5-year Fallow 1998 296(76) 150(29) 213 (144) Rice Field 1999 438 213 125 1) The data were from middle slope position. 2) Numbers in parentheses are standard errors (n=3). In 1999, n=l. 3.4.13 Soil Extractable Magnesium Magnesium (Mg), like Ca, is not normally deficient in most soils (Troeh and Thompson 1993). Burning also raises soil extractable Mg. The cropped-field soils contained significantly higher concentrations of extractable M g than the other fallow fields, and the concentrations gradually decreased as the fallow aged (Table 3.20). Andriesse and Schelhaas (1987b) reported that burning increased extractable Mg about 28-fold. The repeated measurement revealed marked reduction of soil M g after one year of fallow, and a substantial increase from the 5-year fallow to the cropped field (Table 3.21). Table 3.20 Soil extractable Mg concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 204(18)a 136 (6)a 124 (7)a 2-year Fallow 88 (6)bc 50 (7)b 34 (5)c 4-year Fallow-NW 113 (10)b 49 (4)b 36 (5)c 4-year Fallow-SW 168 (20)a 103 (13)b 60 (6)b 5-year Fallow 101 (13)bc 57 (8)b 42 (12)bc Secondary Forest 64 (10)c 19 (2)b 15 (2)d 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significant differences (P<0.05) among fields. 51 Table 3.21 Soil extractable M g concentrations (ppm) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1998 189(12) 141 (5) 108 (2) 1-year Fallow 1999 96 165 210 2-year Fallow 1998 94(11) 60 (9) 40 (8) 3-year Fallow 1999 35 35 38 4-year Fallow 1998 136(14) 49 (6) 35 (7) 5-year Fallow 1999 120 43 105 5-year Fallow 1998 107(16) 67(8) 68 (29) Rice Field 1999 165 93 58 1) The data were from middle slope position. 2) Numbers in parentheses are standard errors (n=3). In 1999, n=l. 3.4.14 Soil Cation Exchange Capacity Cation exchange capacity (CEC) was highest in the secondary forest (Table 3.22). The conversion of the 5-year fallow field to a cropped field appeared to result in a decrease in CEC (Table 3.23). Table 3.22 Soil CEC (cmol+/kg) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 23.8(1.5)b 21.2(1.4)b 22.6 (2.5)bc 2-year Fallow 26.7 (0.8)b 25.5 (2.1)ab 26.4 (0.8)ab 4-year Fallow-NW 24.1 (3.0)b 24.9 (0.9)ab 17.3 (1.5)c 4-year Fallow-SW 24.4 (2.2)b 21.4 (0.9)b 19.5 (0.6)c 5-year Fallow 24.9 (2.4)b 22.6(1.9)ab 21.5 (2.1)bc Secondary Forest 32.4(1.9)a 27.9 (4.0)a 29.1 (3.3)a 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significant differences (P<0.05) among fields. 3.4.15 Soil Extractable Iron Iron (Fe) is considered a micronutrient, but the amount of iron required by plants is the highest among the micronutrients (Troeh and Thompson 1993). Soil Fe was determined in this work 52 because it binds P rendering it unavailable to plants, especially in acid soils (Foth and Ellis 1997). Table 3.23 Soil CEC (cmol+/kg) at three depths in shifting cultivation fields, measured in 1998 and 1999 Field Year 0-5 cm 5-15 cm 15-30 cm Rice Field 1998 21.9 (2.5) 22.0 (2.4) 24.0 (3.4) 1-year Fallow 1999 27.9 24.1 21.9 2-year Fallow 1998 27.5 (0.5) 28.1 (0.2) 25.9(1.1) 3-year Fallow 1999 26.8 27.6 27.0 4-year Fallow 1998 20.4 (8.4) 27.2 (0.8) 16.2 (2.6) 5-year Fallow 1999 27.6 27.6 22.2 5-year Fallow 1998 29.5 (1.7) 20.7 (2.8) 19.6 (2.9) Rice Field 1999 18.7 17.5 18.6 1) The data were from middle slope position. 2) Numbers in parentheses are standard errors (n=3). In 1999, n=l. A l l shifting cultivation fields had high concentration of extractable Fe (extractable Fe at < 3 ppm is considered to be low in Thailand; Topark-ngarm 1983). The soils of the cropped field soils had significantly lower concentrations of extractable Fe than the other fallow fields, and the highest concentration was in the secondary-forest soils (Table 3.24). This may explain higher concentrations of available P in the cropped field (22 ppm) compared to the surface soils of the fallow field and the secondary forest (2-3 ppm) (Table 3.14). Slash burning increased available P by about 2-fold in the surface soils in this study. A reduction of extractable Fe (and possibly Al) , and an associated increase in available P may be one of the major benefits of the burning. 3.4.16 Correlations Amongst Soil Variables Correlation analyses were performed to construct statistical relationships among all measured soil variables (Table 3.25). There were three main trends in the analyses: 1) S O M and total N values were often correlated, but were not correlated with soil pH, 2) P, K , Ca, and M g were 53 highly correlated with each other (PO.01), and with soil pH (PO.01), and 3) soil Fe was negatively correlated with other variables such as P, cations and pH (P>0.01). Table 3.24 Soil extractable Fe concentrations (ppm) at three depths in a chronosequence of shifting cultivation fields Field (in 1998) 0-5 cm 5-15 cm 15-30 cm Rice Field (after harvest) 9.5 (1.4)d 16 (3.7)c 18 (3.2)c 2-year Fallow 36 (7.9)c 40 (6.6)b 37 (7.6)b 4-year Fallow-NW 55 (5.4)b 52 (6.7)b 40 (5.1)b 4-year Fallow-SW 19(5.0)d 30 (6.2)bc 33 (7.0)bc 5-year Fallow 56 (6.0)b 43 (3.5)b 32 (4.6)bc Secondary Forest 125 (8.6)a 78 (5.8)a 62 (4.4)a 1) The data were averaged across slope positions because there were no differences among slope positions. 2) Numbers in parentheses are standard errors (n=9). 3) Within each soil depth, different letters indicate significant differences (P<0.05) among fields. 3.4.17 Atmospheric Inputs of Nutrients The concentration of nutrients of rainfall in the rice field and the 2-year fallow field were comparable, except for K which was slightly higher in the 2-year fallow field (Table 3.26). The throughfall in the forests contained higher concentrations of nutrients than precipitation in the open (the rice field) and precipitation/throughfall in the partially open (the 2-year fallow) fields. The concentration of nutrients in precipitation may not be an accurate measure of precipitation inputs because the mouths of the water-collecting bottles were only about 30 cm above the ground level, and there was visual evidence of soil being splashed into funnels during heavy rain, especially in the rice field. However, the results should provide a broad indication of the potential nutritional contributions of rainfall and throughfall. 54 Table 3.25 Correlation analyses of chemical properties of the soils at three depths in a chronosequence of shifting cultivation fields Depth (cm) pH (H 2 0) pH (KC1) SOM N P K Ca M g Fe CEC 0-5 cm p H ( H 2 0 ) -pH (KC1) ** -S O M NS NS -N NS NS ** -P ** ** NS NS -K ** ** NS NS ** -Ca ** ** NS NS ** ** -M g ** ** NS NS ** ** ** -Fe ** ** * ** ** ** ** ** -CEC NS * * ** NS NS NS NS * 5-15 cm pH (H 2 0) -pH (KC1) ** -S O M NS NS -N NS NS NS -P ** ** NS ** -K ** ** NS NS ** -Ca ** ** NS * ** ** -M g ** ** NS ** ** ** ** -Fe ** ** NS NS ** ** ** ** -CEC NS * NS NS NS * NS NS NS -15-30 cm pH (H 2 0) pH (KC1) ** -S O M * NS -N * * ** -P ** ** NS NS -K ** ** ** ** ** -Ca ** ** NS NS ** ** -M g ** ** * ** ** ** ** -Fe ** ** ** ** NS ** ** ** -CEC NS NS NS NS NS NS NS NS NS -* P < 0.05, ** PO.01 , and NS is Non-Significant. Underlined indicates negative correlation. 55 Table 3.26 Nutrient concentrations and contents of rainfall and throughfall water Field Month (1999) pH N 0 3 " - N N H 4 + - N Total P K Ca M g Na Rainfall Rice Field (after harvest) August (mg/1) (kg/ha) 6.2 0.002 0.005 0 0 0.02 0.06 0.2 0.5 0.02 0.05 0.03 0.07 2.0 4.6 September (mg/1) (kg/ha) 6.2 0.03 0.07 0.1 0.2 0.07 0.14 0.2 0.4 0.05 0.11 0.02 0.04 1.0 2.2 Throughfall 2-year Fallow August (mg/1) (kg/ha) 6.5 0.001 0.002 0 0 0.03 0.06 0.8 1.9 0.02 0.05 0.03 0.07 2.6 6.1 September (mg/1) (kg/ha) 6.3 0.001 0.002 0 0 0.04 0.08 0.3 0.6 0 0 0.02 0.04 1.6 3.5 4-year Fallow-SW August (mg/1) (kg/ha) 6.9 0.05 0.12 0.3 0.7 0.1 0.2 6.4 15 1.0 2.3 0.5 1.1 12 27 September (mg/1) (kg/ha) 6.5 1.1 2.4 0.03 0.06 0.3 0.6 3.5 7.5 1.2 2.6 0.4 0.9 7.3 16 Secondary Forest August (mg/1) (kg/ha) 6.2 0.06 0.13 0.1 0.2 0.09 0.21 0.7 1.6 0.2 0.5 0.02 0.05 2.0 4.6 September (mg/1) (kg/ha) 5.8 0.3 0.7 0.3 0.5 0.2 0.4 2.5 5.4 0.2 0.4 0.05 0.11 5.0 11 The amounts of rainfall were 232 mm for August, and 215 mm for September. 3.4.18 Soil Erosion (Literature Review) Shifting cultivation practices in northern Thailand occurs exclusively on sloping hills having various degrees of steepness; Mae Hae Tai village has its shifting-cultivation fields on hills with slopes ranging from 38 to 55%. The Thai Forestry Department had declared that areas with more than 35% slopes are unsuitable for agricultural activities due to possible erosion hazard. 56 Soil-erosion studies in Thailand (Hurni 1983; Kyuma and Pairintra 1983; Sajjapongse 1997) and in northern Laos (Roder et al. 1995) have yielded highly variable results concerning potential and actual soil erosion depending on the type of erosion (e.g. gully or sheet erosion) and on the measurement method. Sanchez (1976) pointed out that traditional shifting cultivation normally results in low erosion rates because the period of bare soils is rather short. Recent studies indicate that the perception of soil erosion caused by upland farming is often exaggerated, and much of the gully erosion that can be seen in fields predates the start of upland farming activities (Forsyth 1996). Unpaved roads contribute more to erosion than agricultural lands because road surfaces are more compact with low infiltration rates (Ziegler and Giambelluca 1997). In addition, upland farmers understand the risks of soil erosion, and historically have avoided farming on the steepest slopes (Forsyth 1996). I believe that the farmers are well aware of soil erosion. Several of the farmers whom I interviewed stated that some surface soil is moved down the slope when it rains, and that it is deposited at the valley bottom where they have paddy rice fields, and the sediment acts as an excellent fertilizer (see also Kleinman et al. 1995). This is similar to the situation in Borneo where the Dayaks people plant swamp rice at the slope base (Dove 1985). In my study, I did not collect data on soil erosion. Table 3.27 presents some estimations of soil erosion from various studies in northern Thailand and Laos. Estimated nutrient losses via erosion are shown in Table 3.28. 57 Table 3.27 Summary of soil erosion studies in northern Thailand and northern Laos Location Land Use Field and Farming Description Soil Erosion (t/ha/year) Source Tung Chao, Northern Thailand Shifting Cultivation (Fallow Fields) 50% slope 30-40 m slope length 54% slope 24 m slope length 40 m slope length 108-152 89 120 Hurni (1983) Northern Thailand Various Upland and various farming practices 30-125 Boonchee and Anecksamphant (1993) Pa Dua, Northern Thailand Upland Farming 0-9.9% slope > 40% slope (estimated for 1963-1991) 24 64 Forsyth (1996) Chiang Mai , Chiang Rai, Northern Thailand Farmers' Practice Alley Cropping Planting crops up and down the slope Leucaena and pigeon peas 124 37 Sajjapongse (1997) Hillside Ditches A l l have 20-50% slope 8 Northern Laos Shifting Cultivation Cropping period 0.3-30 Roder et al. (1995) 3.5 Summary The soils (0-5 cm) of the shifting cultivation fields of Mae Hae Tai village were rather clayey (about 40% clay), and were slightly acid (pH between 5-6 in water). Soil bulk density was approximately 1 g/cm , with 6.0-6.8%) O.M., and about 0.3% total N . Soil available P was very low (2-3 ppm), except in the 1998 cropped field where available P was somewhat higher due to the ash released when the fallow was slashed and burned and fine roots of rice and trees decomposed. The soils were believed to be sufficient for rice production in K , Ca, and Mg, and the ash effect appeared to enhance their availability. When soil properties were remeasured after one year, changes were not significant except when the 5-year fallow field was converted to a cropping field by field clearing and burning, and cropping. Table 3.28 Estimation of nutrient losses via soil erosion Field (in 1998) Soil Nutrient The lowest erosion rate, The highest erosion rate, 0.30 t/ha/year 152 t/ha/year (Roderetal. 1995b) (Hurni 1983) Rice Field Total O.M. 10.5 5,301 (after harvest) Total N 0.5 244 Available P 0.3 169 Extractable K 5.6 2,828 Extractable Ca 29 14,507 Extractable M g 3.2 1,600 2-year Fallow Field Total O.M. 8.7 4,431 Total N 0.4 198 Available P 0.04 22 Extractable K 2.1 1,068 Extractable Ca 6.3 3,167 Extractable Mg 1.3 666 4-year Fallow Field- Total O.M. 10 5,211 N W Total N 0.5 241 Available P 0.04 19 Extractable K 3.0 1,518 Extractable Ca 10 5,174 Extractable M g 1.7 881 4-year Fallow Field- Total O.M. 10.6 5,363 SW Total N 0.5 255 Available P 0.1 54 Extractable K 5.8 2,951 Extractable Ca 18.9 9,552 Extractable Mg 2.8 1,431 5-year Fallow Field Total O.M. 10 5,106 Total N 0.5 245 Available P 0.05 23 Extractable K 3.3 1,651 Extractable Ca 4.8 2,437 Extractable Mg 1.6 797 Secondary Total O.M. 10.8 5,463 Forest Total N 0.5 257 Available P 0.03 15 Extractable K 2.6 1,341 Extractable Ca 2.5 1,263 Extractable Mg 0.9 447 59 It was initially hypothesized that SOM, as an indicator of soil fertility, would decline after nutrients were exported in rice grain and rapid decomposition of organic matter during the cropping period, and would have increased as a result of litter addition from fallow vegetation towards the end of the fallow period. However, there was no clear trend of S O M or N changes, as the fallow aged. The change of SOM in the chronosequence of the shifting cultivation fields was rather small in a crop-fallow rotation of six years. Changes in nutrient contents of different compartments (e.g. upland rice, trees and litterfall, and shrub) of the system are reported in the following chapters. 60 Chapter 4 Upland Rice and Role of Nutrients in Rice Productivity 4.1 Introduction Rice is the main food crop in Asia, including Thailand. Two major rice varieties planted in Thailand are glutinous (Oryza glutinosa) and non-glutinous rice (O. sativa). The glutinous species is mainly consumed by Thai people in the northern and northeasten parts of the country. However, the Karen people in northern Thailand prefer the non-glutinous variety, but grow a smaller portion of glutinous rice for use in ceremonies. A large number of rice varieties are used by different villages. This chapter describes the importance of rice as the main food staple grown in Thailand. It provides a brief introduction to paddy rice and dry or upland rice. Planting methods of the upland rice by the Karen shifting cultivators are described, and the 1998 and 1999 yields and nutrient contents of upland rice are reported. In Thailand, rice is planted in four conditions: irrigated, rainfed lowland, rainfed dryland, and deepwater. Setboonsarng (1996) reported that, in 1992, the area of rice growing in the rain-fed dryland system in Thailand was about 6.7 million ha, which accounted for 70% of the total rice-growing area of the country, but only about 60% of total rice production. This is because rainfed-dryland rice normally provides the lowest average yield of about 1.8 t/ha, compared to 2.1, 2.5, and 4 t/ha for deepwater, rainfed lowland and irrigated rice, respectively (Setboonsarng 1996). This value compares with the range in yield of upland rice in Asian countries of 0.5-2 t/ha. It is marginally more than that of Africa (0.5-1.5 t/ha) but less than that of Latin America (1-3 t/ ha; Pande 1994). Physiologically, rice is more adapted to water-logged or anaerobic than to dry or 61 aerobic soil conditions, so it can grow well and provides better yield under wet than under dry soil conditions (Ponnamperuma 1975). The Karen farmers start to plant their upland rice in the middle or late April before the rain starts in May. Their method of planting is rather simple. They usually plant in pairs. One person randomly digs small holes (about 14 holes/m2; Zinke et al. 1978) using an iron dibble-stick, and then the other person throws approximately 5-20 rice seeds into each hole (Sutthi 1996). The seeds are not covered with soil, and some of the seeds land around rather than in the hole. This results in an approximate rice clump spacing of 25-35 cm (Sutthi 1996). The farmers mention that all the seeds wil l run into the holes when the rain comes. Sutthi (1996) stated that the farmers realize that soil eroded from up-slope areas will fill the holes once it rains. The average amount of seeds used is 250 kg/ha. This estimate is based on the Karen's two traditional methods of measuring area. The first uses the armlength, of which one armlength equals one meter, which is similar to that of the traditional Thai measurement. The second method is based on the amount of rice seeds used to plant the fields; they normally use 40 kg of rice seeds to plant an area of 0.16 ha when seeding is done before the rainy season starts. The upland rice is ready to be harvested in October and November for short-lived and long-lived varieties, respectively. The harvest is done by cutting rice stems with ripe grain 20 cm above the ground level using a curved knife. The cut portion is left to be sun-dried and then threshed within the field, resulting in within-field variability in rice production due to the uneven distribution of decomposing materials because the locations of the threshing actitvity tends to be distributed 62 randomly from year to year. The Karen leave all rice plant components except the grain to decompose in the field, whereas in many lowland regions rice straw is used as substrate for cultivating mushrooms or is burned. Leaving the straw to decompose in the field, as opposed to burning it, should have a significant effect on soil organic matter and nutrient balance, especially N , in the system. Normally, the Karen shifting cultivators of Mae Hae Tai village grow rain-fed upland rice for one year, and then leave the cultivated field to fallow for five years. One assumption that is commonly made in studies of such shifting cultivation systems is that the export of nutrients in harvested materials can lower site productivity unless they are replenished by fertilizer. In shifting cultivation systems the nutrients are supplied to the crops via several pathways; e.g. release from burned slash and soil organic matter decomposition processes, and smaller inputs from soil weathering and precipitation. However, two other explanations of why they only grow rice for only one year are weed and insects/disease problems. The farmers usually do manual weeding, which is labor intensive. It is possible that their choice not to grow rice for the second year is because weeds become more prominent in the second year. One of the causes of the weed problems in shifting cultivation could be a relatively large proportion of bare soils due to the large planting space of about 30 cm, which leaves spaces for weed establishment (Bomke, A. , Faculty of Agricultural Sciences, U B C , per. comm.). This should have important implications in planting techniques and weed management. Problems with insects and diseases were not mentioned by the farmers, but have been suggested elsewhere as a problem with continuous cropping. 63 However, there is little empirical evidence with respect to either weed competition, or soil fertility, or insects/disease aspects of a second year of cropping because the farmers of Mae Hae Tai village fallow their fields after one year of cropping. Consequently, there is little actual evidence of yield decline in second-year cropping, or why it occurs i f it does. The two main reasons given by the farmers for not sustaining cropping are the increasing weed competition, which undoubtedly exacerbates the effects of any decline in soil fertility on rice growth; and the reduction in soil fertility. Only the latter was addressed in this study. 4.2 Objectives and Rationale 1) To document the biomass and nutrient (N, P, K , Ca, and Mg) contents of upland rice growing in the forest-fallow shifting cultivation system. Biomass of upland rice accounts for the nutrient content of the vegetation component in the cropping phase of the shifting cultivation system. The biomass and nutrient contents of upland rice were examined to determine nutrients exported (lost) out of the system in grain, and nutrients retained in the system as leaf, stem, and root components of the rice crop. Data on rice biomass and nutrient contents are put into tabular models of nutrient flows of the system in Chapter 7 to illustrate the relative magnitude of nutrient loss via this pathway. The loss of nutrients in exported rice grain is a component of an assessment of whether there will be a decline of soil fertility i f the fallow period is further reduced. Nutrient contents of shrubs, trees, and litter are reported in Chapters 5 and 6. 2) To test for nutrient limitations in the shifting cultivation system by bioassay, based on crop growth and nutrient contents in a fertilizer trial in a first-year rice field and an experimental second-year rice field, also with a fertilizer trial. 64 4.3 Methods 4.3.1 Upland Rice Sampling I started my field work in June 1998. The rice crop had already been planted in April of the same year, so there was no soil sampling of the 1998 rice field prior to the cropping. The 1998 rice sampling was done at the end of October. A plot of 20x40 m was selected within the rice field, and twenty-five l x l m quadrats were systematically established at 2-m intervals to form five rows along the contour and five columns down the slope within the plot. Common shifting cultivation practices such as patchy burning (slash is located randomly in the field, and incompletely burned slash is piled randomly and re-burned) and random piling of cut weeds should result in patchiness of soil fertility within a field. Stratified systematic sampling of the rice was used to detect this patchiness, which might have not been found i f stratified random sampling was used. The rice was harvested from each 1 m 2 subplot. As noted above, farmers normally use a curved knife to cut the rice stem at about 20 cm above the ground. However, I instructed the farmers to cut the stem as close to the ground as possible in this study. In 1999, the 1998 five-year fallow field was cleared in February, burned in March, and planted in April. Soil properties of this 5-year fallow field were reported in Chapter 3. The rice planting of this cropping involved a fertilizer trial described in section 4.3.4 in this chapter. Biomass of upland rice in 1999 was estimated from the control treatment of the fertilizer trial. Root biomass was measured in 1999 but not in 1998 (shootroot ratios from the 1999 control plot were used to estimate root biomass of the 1998 rice). 65 4.3.2 Chemical Analyses of Upland Rice Samples The above-ground upland rice parts were then air-dried, separated into different components (grain, stem and leaves) and chemically analyzed for N by micro Kjeldahl (Jackson 1958). For P, K , Ca, and M g analyses, the samples were prepared by wet digestion (HN03 :HC104 =6:1), and then determined by: the vanado-Molybdate method and read by spectrophotometry for P; flame photometry for K ; and atomic absorption spectrophotometry for Ca and Mg. The grain was separated into good and bad grain using a seed-cleaning machine at the Agronomy Department, Chiang Mai University. 4.3.3 The Second-Year Rice Experiment A second-year rice experiment was conducted to determine the importance of nutrition in the farmers' decision to grow rice for only one year. I asked a farmer to grow the same variety of rice in 1999 in the same plot from which I harvested the rice in October 1998. The rice was planted on April 25 t h, 1999. Unfortunately, the experiment was disturbed on June 13 t h by cattle that broke down the fence around the plot. However, the farmer claimed that until the cattle got into the plot, the rice had performed rather poorly. It was yellowish green instead of bright green, and the height growth was poor. Rice at the age at which the experiment was disturbed is normally close to 30 cm in height, but it appeared to be less than 20 cm in height judging from uneaten (by the cattle) rice stems left in the field on June 13 t h. The rice was re-planted on June 14 , but this resulted in the rice being six weeks younger than intended at the time of the harvest. This second-year experiment crop also involved a fertilization experiment. The fertilizer treatments (see section 4.3.4 for details on the treatments) were applied once in July 1999 when the rice crop was about one month old. 66 4.3.4 Fertilizer Trial In order to examine the role of nutrients in reduced forest-fallow shifting cultivation, two fertilizer trials were conducted in 1999. The first examined fertilizer responses in a first-year rice field, and the second examined similar responses in the second-year rice field. The trials used a 5x4 Latin Square design to account for potential response variability in the sloping plot. The total number of subplots was 20, and each subplot was 2x2 m in size. There were five replicates for each of the following treatments: 1) control; and fertilized with 2) 21-0-0, 3) 16-20-0, and 4) 15-15-15 N P K . Commercial chemical fertilizers were used in the fertilizer trial, and the actual amounts of each nutrient in the fertilizers applied (8 g for each 2x2 m subplot or 20 kg/ha) in the trial are as follows: fertilizer 21-0-0 (N) contained 4.2 kg of N/ha; fertilizer 16-20-0 (NP) contained 3.2 kg of N/ha, and 1.8 kg of P/ha; fertilizer 15-15-15 (NPK) contained 3 kg of N/ha, 1.3 kg of P/ha, and 2.5 kg of K/ha. The three fertilizer treatments wil l be referred to as the N , NP, and N P K treatments from hereon. The treatments were applied in July 1999 when the first-year rice was two months old. The fertilizers were applied on the soil surface surrounding the rice plants. In October, 1999, rice was harvested, separated into grain, other aboveground parts, and roots; the samples were then air-dried, weighed, and analyzed for N , P, K , Ca, and M g (see section 4.3.2 for the analysis methods). 4.4 Results and Discussion 4.4.1 Upland Rice Biomass In 1998, most rice-growing areas in Thailand suffered from an aphid epidemic that resulted in below-average yield all over the country. The amount of rainfall was also relatively low (856 mm while the 15-year average was 1,214 mm; RFD 1999), possibly due to the E l Nino effect. As 67 a result, the estimated grain yield of Mae Hae Tai village was merely 519 kg/ha, with a harvest index of only 22% (Table 4.1). In contrast, the grain yield in 1999 was 902 kg/ha (a 73% increase over 1998). The 1999 annual rainfall was 1,342 mm. Interestingly, the total aboveground biomass excluding grain was about 58% higher in 1998 than in 1999. Table 4.1 Upland Rice Biomass, Mae Hae Tai village, Chiang Mai, Thailand Rice Component 1998 Dry Weight (kg/ha) 1999 Dry Weight (kg/ha) Good Grain 519 902 Bad Grain 77 40 Stem 1,528 Not measured Leaf 330 Not measured Stem+Leaf 1,858 1,225 Root 1,609* 1,061 ShootRoot 1.15:1* 1.15:1 Harvest Index (%) 22 42 * Estimated from shoot:root ratio of the 1999 rice. 4.4.2 Upland Rice Nutrient Contents Harvesting represents the largest loss of nutrients from many agricultural systems. For this particular system of shifting cultivation, about 30% of the N , 25% of the P, and 20% of the Ca and M g in the aboveground biomass was exported via the 1998 upland rice grain; for K , the export was only 9% (Table 4.2). The quantities of nutrients exported in the 1998 harvested grain were 7.3, 1.6, 2.1, 0.5, and 0.4 kg/ha of N , P, K , Ca, and Mg, respectively. Stems accounted for the largest amount of N in the rice plants (11 kg/ha) even though the N concentration in stems was less than half that of grain: 0.7% in stem, and 1.4% in the grain. This was because the stem biomass (1,528 kg/ha) was almost triple the grain biomass (519 kg/ha) (Table 4.1). Similarly, the stem accounted for more of the other nutrients than the grain. This 68 lends support to the idea that the shifting cultivation system I studied can be sustainable given sufficient periods of fallow. The nutrient export is relatively low and nutrients are replenished via several processes (e.g. soil weathering and precipitation input), which can compensate for the nutrients removed as long as the fallow period is sufficiently long. Table 4.2 Nutrient concentrations and contents of the upland rice crop of the 1998 and 1999 growing seasons Rice Part Year N P K Ca M g % kg % kg % kg % kg % kg /ha /ha /ha /ha /ha Good 1998 1.4 7.3 0.3 1.6 0.4 2.1 0.1 0.5 0.1 0.4 Grain 1999 1.2 11 0.2 1.8 0.3 2.7 0.03 0.3 0.1 0.9 Bad Grain 1998 1.2 0.9 0.2 0.2 0.7 0.5 0.1 0.1 0.1 0.05 1999 1.2 0.5 0.2 0.1 0.6 0.2 0.1 0.04 0.1 0.04 Stem 1998 0.7 11 0.3 4.6 1.2 18 0.1 1.5 0.1 1.5 1999* 1.2 15 0.04 0.5 1.7 21 0.2 2.5 0.1 1.2 Leaf 1998 1.3 4.3 0.1 0.3 1.0 3.3 0.2 0.7 0.1 0.3 1999* 1.2 - 0.04 - 1.7 - 0.2 - 0.1 -Exported 1998 31 7.3 24 1.6 9 2.1 19 0.5 16 0.4 1999 40 11 75 1.8 11 2.7 9 0.3 40 0.9 Retained 1998 69 16 76 5.1 91 22 81 2.3 84 1.9 1999 60 16 25 0.6 89 23 91 2.7 60 1.4 A The 1999 data were from the control plots of the field fertilizer experiment. *Leaf and stem components in 1999 were composited before chemical analyses. The nutrients removed in the harvested rice was similar in the 1998 and 1999 growing seasons (Table 4.2), except for P, of which the ratio of exported to retained was reversed between the two growing seasons. In 1998, the amount of P exported was about 25% of the aboveground total, whereas it was 75% in 1999. However, the between cropping-year differences were smaller in the cases of N and K , in which grain export increased from 31 to 40% (N), and from 9 to 11% (K) in the 1998 and 1999 growing seasons, respectively. 69 There should be a lower risk of future deficiencies of K , Ca, and M g than of N and P because of the practice of leaving crop residue to decompose in the fields. In addition, the soil contents of K , Ca, and M g were higher in comparison to N and P (see Chapter 3). There could be nutrient deficiencies of some micronutrients, but this was not studied. 4.4.3 Biomass of the First-Year (1999) Upland Rice Receiving Fertilizer Treatments Overall performance of the upland rice was improved by all fertilizer treatments (Table 4.3). The highest grain biomass was obtained in the N fertilizer treatment (1,354 kg/ha), but it was not significantly (P < 0.05) different from the other two fertilizer treatments (1,218 kg/ha for the NP treatment and 1,066 kg/ha for the N P K fertilizer treatment). The unfertilized control resulted in significantly lower grain yield than that of the N and the NP treatments, but was similar to that of the N P K treatment. The larger grain biomass of the N treatment indicated N deficiency of the soils. The lower biomass of the NP and N P K (than the N treatment) treatments reflected smaller addition of N in the NP and N P K treatments. Table 4.3 Biomass of the first-year upland rice components in plots receiving fertilizers (1999) Treatment Good Bad Grain Leaf and Root ShoofcRoot* Harvest Grain (kg/ha) Stem (kg/ha) Index (kg/ha) (kg/ha) (%) Control 902 40 1,225 1,061 1.1 42 (150)b (ll)bc (314)b (202)b (0.1) Fertilizer 1,354 54 1,906 1,571 1.2 41 21-0-0 (152)a (7)a (228)a (182)a (0.02) Fertilizer 1,218 52 1,672 1,434 1.2 41 16-20-0 (192)a (4)ab (212)a (224)a (0.1) Fertilizer 1,066 50 1,746 1,272 1.4 37 15-15-15 (163)ab (3)ab (266)a (179)ab (0.04) 1) Numbers in parentheses are standard errors (n=5). 2) Within a column, values followed by different letters are significantly (p<0.05) different. * Statistical comparison was not done. 70 The highest aboveground (leaf and stem, excluding grain) biomass was obtained in the N fertilizer treatment (1,906 kg/ha). The NP and N P K fertilizer treatments gave a similar aboveground biomass (1,672 kg/ha and 1,746 kg/ha for the NP and the N P K treatments, respectively). The trend of belowground (root) biomass response was the same as the aboveground (leaf, stem, and grain) biomass. The shootroot ratios (1.1-1.4) and the harvest indices (37—43) were comparable among the treatments. The responses of the rice crop per kg of fertilizer nutrient added are presented in Table 4.4. The grain biomass response to added N was equal when N was applied in the fertilizer alone or in combination with P; 108 and 99 kg/ha of increased grain biomass per kg of added N in the N and the NP fertilizer treatments, respectively. However, the increased grain biomass per kg of added N was less in the N P K treatment (55 kg/ha). Effects of N on the grain biomass were decreased when used in conjunction with K in the N P K treatment. This may indicate that only N is important for grain production, and the addition of K will only improve leaf and stem biomass. The increased shoot response per kg of added N in the N fertilizer treatment (162 kg/kg N) was slightly higher than in the NP treatment (140 kg/kg N). The increased root biomass of the N (121 kg/kg N) and the NP (117 kg/kg N) fertilizer treatments were similar. 4.4.4 Nutrient Contents of the First-Year (1999) Upland Rice Receiving Fertilizer Treatments For nutritional analyses of rice tissues, all five replications for each treatment were pooled before the chemical analyses and therefore statistical analyses could not be done (Table 4.5). The grain component contained the largest concentration of N (1.2% in the unfertilized control plot), while the stem contained the largest concentration of K (1.7% in the unfertilized control plot). 71 Interestingly, the bad-grain component contained similar concentrations of all nutrients to those of good grains, except for K , which was higher in the bad grains (0.6%) than in the good grains (0.3%). However, due to the small proportion of the bad-grain component (about 4% of the total aboveground biomass excluding grain) the total nutrient contents of this component was low (Table 4.6). On the other hand, the large biomass of the stem made it the largest accumulator of all nutrients, except P. Table 4.4 First-year upland rice biomass (1999 growing season) response to fertilizers. Mass response, and response per kg of added nutrients Treatment N , P, K in Good Increased Leaf Increased Root Increased fertilizer Grain Grain and Shoot (kg/ha) Root (kg/ha) (kg/ha) (kg/kg added nutrient) Stem (kg/ha) (kg/kg added nutrient) (kg/kg added nutrient) Control 0 902 (150)b 0 1,225 (314)b 0 1,061 (202)b 0 Fertilizer N , 4.2 1,354 N , 108 1,906 N , 162 1,571 N , 121 21-0-0 P ,0 K , 0 (152)a (228)a (182)a Fertilizer N , 3.2 1,218 N , 99 1,672 N , 140 1,434 N , 117 16-20-0 P, 1.8 K , 0 (192)a (212)a (224)a Fertilizer N , 3 1,066 N , 55 1,746 N , 174 1,272 N , 70 15-15-15 P, 1.3 (163)ab (266)a (179)ab K , 2.5 1) Numbers in parentheses are standard errors (n=5). 2) Within a column, values followed by different letters are significantly (p<0.05) different. Fertilization resulted in a greater export of N , P, and K out of the system via harvested grains (Table 4.6). N exports were increased by about 27% in the NP and N P K treatments, and about 45% in the N treatment compared with the unfertilized control plots. Increases in P exports were similar (about 18%) in the N and NP treatments, and about 47% in the N P K treatment. 72 Table 4.5 Nutrient concentrations (%) of the first-year upland rice (1999) Treatment Nutrient Good Grain Bad Grain Leaf+Stem Root Control N 1.2 1.2 .1.2 0.6 P 0.2 0.2 0.04 0.05 K 0.3 0.6 1.7 1.1 Ca 0.03 0.05 0.2 0.01 M g 0.05 0.06 0.08 0.06 Fertilizer N 1.2 1.1 0.8 0.6 21-0-0 P 0.2 0.2 0.04 0.06 K 0.3 0.8 1.7 0.9 Ca 0.01 0.04 0.2 0.01 Mg 0.03 0.07 0.1 0.06 Fertilizer N 1.1 1.2 0.7 0.6 16-20-0 P 0.2 0.2 0.05 0.06 K 0.3 0.7 1.8 0.9 Ca 0.01 0.03 0.09 0.003 M g 0.06 0.06 0.08 0.06 Fertilizer N 1.3 1.1 0.7 0.6 15-15-15 P 0.2 0.2 0.05 0.06 K 0.3 0.6 1.8 0.9 Ca 0.01 0.03 0.08 0.003 M g 0.05 0.06 0.09 0.06 The total N content (41 kg/ha) in all rice components (grains, leaves and stem, and roots) in the N fertilizer treatment was 8 kg/ha more than the N contents (33 kg/ha) in the unfertilized control plot (Table 4.7). The NP and the N P K treatments resulted in 1 kg/ha of extra N in all rice components when compared to the unfertilized control plot. Total rice P uptake in all of the fertilizer treatments was 1.0-1.6 kg/ha of P more than the control, whereas the increased K contents were 15, 11, and 10 kg/ha in the N , NP, and N P K fertilizer treatments, respectively. The N fertilizer treatment added 4.2 kg/ha of N , but the increased N content in the rice biomass (from the unfertilized control) was 8 kg/ha (Table 4.7). This indicated that fertilization caused the rice crop to extract an extra 3.8 kg of N/ha from the soils, which was thought to be a result of an increase in the soil volume occupied by an increased root mass due to the improved rice nutrition. Table 4.6 Nutrient contents (kg/ha) of the first-year upland rice (1999) Treatment Nutrient Good Bad Grain Leaf+ Root Total Grain Stem Control N 11 0.5 15 6.4 33 P 1.8 0.1 0.5 0.5 2.9 K 2.7 0.2 21 12 36 Ca 0.3 0.02 2.5 0.1 2.9 M g 0.4 0.02 1.0 0.6 2 Fertilizer N 16 0.6 15 9.4 41 21-0-0 P 2.7 0.1 0.8 0.9 4.5 K 4.1 0.4 32 14 51 Ca 0.1 0.02 3.8 0.2 4.1 M g 0.4 0.04 1.9 0.9 3.2 Fertilizer N 13 0.6 12 8.6 34 16-20-0 P 2.4 0.1 0.8 0.9 4.2 K 3.7 0.4 30 13 47 Ca 0.1 0.02 1.5 0.04 1.7 M g 0.7 0.03 1.3 0.9 2.9 Fertilizer N 14 0.6 12 7.6 34 15-15-15 P 2.1 0.1 0.9 0.8 3.9 K 3.2 0.3 31 11 46 Ca 0.1 0.02 1.4 0.04 1.6 M g 0.5 0.03 1.6 0.8 2.9 However, when N was applied in combination with P and K , the recovery rates of N from the fertilizers were decreased. Only 1 kg of N/ha appeared in the increased N content of the rice when 3.2 kg of N/ha was applied in the NP fertilizer treatment, and 1 kg of N/ha was found in the N P K treatment, in which 3 kg of N/ha was applied. The increased P contents of the rice crop was 1.3 and 1.0 kg/ha in the NP and the N P K fertilizer treatments, respectively, whereas the actual added amounts of P were 1.8 and 1.3 kg/ha, respectively. Thus, the recovery rate of P fertilizer appears to have been lower than that of N fertilizer. This may reflect the high P fixation capacity of the clay-rich, acidic soils. 74 Table 4.7 Increased nutrient contents of the first-year upland rice (1999 growing season) resulting from added nutrients in fertilizers Treatment Nutrient Added Nutrient Total Nutrient Increased in Fertilizer* Content of Nutrient (kg/ha) Upland Rice Content (kg/ha) (kg/ha) Control N 0 33 -P 0 2.9 -K 0 36 -Fertilizer N 4.2 41 8 21-0-0 P 0 4.5 1.6 K 0 51 15 Fertilizer N 3.2 34 1 16-20-0 P 1.8 4.2 1.3 K 0 47 11 Fertilizer N 3 34 1 15-15-15 P 1.3 3.9 1 K 2.5 46 10 'Fertilizer was added at 20 kg/ha. 4.4.5 The Second-Year Rice Experiment and Fertilizer Trial As noted earlier, the second-year rice experiment was re-planted due to the crop damage by the cattle, which made the second-year rice six weeks younger at the time of harvest than the first-year rice. The first-year and the second-year rice crops were harvested at the same time of year (October). At the harvest time, the second-year rice was brown and wilted with only a small amount of grain. This might possibly have been caused partially by changes in temperature and photoperiod relative to rice phenology due to late planting (Kimmins, J.P., Forest Sciences Department, U B C , per. comm.), but the second-year rice appeared to under-perform (in comparison to the regular first-year rice) from the start of the experiment, and, as stated earlier, the initial second-year crop was stunted and yellowish before it was damaged by cattle. A comparison of the rice yield performance of the 1999 first-year regular crop and the experimental second-year planting is, therefore, complicated because of the confounding effects of the delayed 75 second-year planting. However, the fertilizer trial within the second-year rice study still permits a bioassay of nutritional constraints on rice growth and yield. The unfertilized second-year rice yielded a mere 70 kg/ha of aboveground biomass with no measurable amount of grain (Table 4.8). The second-year fertilized rice produced more biomass than the unfertilized control but also performed very poorly, also with no measurable amount of grain. The largest aboveground biomass (leaf and stem) was in the N fertilizer treatment (199 kg/ha), but was similar to the 115 kg/ha in the NP treatment due to high variability of biomass. The recovery rates of added N , NP, and N P K fertilizers are shown in Table 4.9. Nutritional analyses of rice components are shown in Table 4.10. Nutrient concentrations of the rice crop in both control and fertilized treatments were generally comparable. Table 4.8 Experimental second-year upland rice biomass response to fertilizers. Mass response, and response per kg of added nutrients Treatment N , P, K in Shoot Increased Root Increased ShootRoot Fertilizer (kg/ha) Shoot (kg/ha) Root (kg/ha) (kg/kg added nutrient) (kg/kg added nutrient) Control 0 70 (32)b 0 4.7 (2)b 0 14 Fertilizer N , 4.2 199 N , 31 10 N , 1 19 21-0-0 P ,0 K , 0 (68)a (1.6)a Fertilizer N , 3.2 115 N , 14 10 N , 2 13 16-20-0 P, 1.8 K , 0 (15)ab (2)ab Fertilizer N , 3 98 N , 9 6.6 N , l 15 15-15-15 P, 1.3 K , 2.5 (19)b (1.7)ab 1) Numbers in parentheses are standard errors (n=4). 2) Within a column, values followed by different letters are significantly different (P<0.05) for leaf and stem, and P>0.10 for root. Table 4.9 Increased nutrient contents of the experimental second-year upland rice resulting from added nutrients in fertilizers Treatment Nutrient Added Nutrient Total Nutrient Increased in Fertilizer* Content of Nutrient (kg/ha) Upland Rice Content (kg/ha) (kg/ha) Control N 0 0.8 -P 0 0.1 -K 0 1.5 -Fertilizer N 4.2 2.3 1.5 21-0-0 P 0 0.2 0.1 K 0 3.1 1.6 Fertilizer N 3.2 1.1 0.3 16-20-0 P 1.8 0.1 0 K 0 1.5 0 Fertilizer N 3 1.2 0.4 15-15-15 P 1.3 0.1 0 K 2.5 1.6 0.1 *Fertilizer was added at 20 kg/ha. Table 4.10 Nutrient concentrations and contents of the experimental second-year rice* Treatment N P K Ca M g % kg/ha % kg/ha % kg/ha % kg/ha % kg/ha Leaf+Stem Control 1.0 0.7 0.1 0.1 2.0 1.4 0.2 0.1 0.1 0.1 Fertilizer 1.1 2.2 0.1 0.2 1.5 3.0 0.2 0.4 0.1 0.2 21-0-0 Fertilizer 0.9 1.0 0.1 0.1 1.2 1.4 0.2 0.2 0.1 0.1 16-20-0 Fertilizer 1.1 1.1 0.1 0.1 1.5 1.5 0.1 0.1 0.1 0.1 15-15-15 Root Control L I H I 0 l 0.005 L3 O l 0.01 0.0005 O l O005 Fertilizer 1.1 0.1 0.1 0.01 1.1 0.1 0.01 0.001 0.1 0.01 21-0-0 Fertilizer 1.0 0.1 0.1 0.01 1.1 0.1 0.03 0.003 0.1 0.01 16-20-0 Fertilizer 1.0 0.1 0.1 0.01 1.2 0.1 0.04 0.003 0.1 0.01 15-15-15 *Each treatment had 5 replications, but they were pooled before chemical analyses, so statistical analysis could not be done. 77 Presently, poor yield in consecutive-year cropping is not a key issue in the forest-fallow shifting cultivation of the Karen of Mae Hae Tai and other Karen villages in Mae Chaem Watershed because the rice crop is grown for only one year. It will become more critical i f the farmers choose to prolong their cropping period rather than reduce their fallow period i f the overall crop-fallow period was to be reduced. However, this is unlikely because most of the farmers commented that they recognize the disadvantages of continuous cropping. They would rather choose to reduce the fallow period than to grow the rice crop for consecutive years (see Appendix 10 on Farmers' Interviews). 4.4.6 Comparison of Upland Rice Yield Amongst Various Studies In comparison of the rice yield to other studies, Roder et al. (1997b) reported 1,100 kg/ha of upland rice grain and 1,200 kg/ha of rice stem in northern Laos in 1991, which is similar to this study (Table 4.11). Generally, rain-fed upland rice provides relatively low yields compared with paddy rice. Mae Chaem District Office (1991) reported average annual yields of 3.5 and 1.9 t/ha for lowland paddy and upland rice, respectively. In a yield trial at the International Rice Research Institute (IRRI) in 1972 in the Phillippines, highland varieties of upland rice provided grain yield from 2.3-3.4 t/ha, and while lowland varieties gave between 2.2-4.1 t/ha of grain (Chang and Vergara 1975). In comparison with another study in northern Thailand (Zinke et al. 1978), the exported nutrients in upland rice grain as a percentage of nutrients in the total aboveground biomass followed the same trend; N had the highest proportional export, followed by P and K (Table 4.12). The variation between locations should be related to different levels of soil fertility because they are 78 all from the same climate area. In general, nutrient concentrations from these two studies were very comparable. Zinke et al. (1978) studied the Lua shifting cultivation system with a 10-year rotation in northern Thailand. Roder et al. (1997b) showed highly different nutrient contents of rice stems in northern Laos compared with this study (Table 4.13). Table 4.11 Comparison of upland rice yields amongst various studies Location Year Rice Grain Yield (kg/ha) Source Northern Thailand 1967/68 Paddy 2,193 Kunstadter et al. (Lua Tribe) Upland 1,032 (1978) Northern Thailand 1967/68 Paddy 1,483 Kunstadter et al. (Karen Tribe) Upland 955 (1978) Northern Laos 1991 Upland 1,100 Roder et al. (1997b) Northern Thailand 1992/93 Paddy 2,148 Hansen (1995) (Various Tribes) Upland 993 Northern Thailand 1998 Upland 519 This Study (Karen Tribe) 1999 Upland 902 (1998 and 1999) Northern Thailand 1998 Upland 1,126 This Study (1998) (Karen Tribe)* *I obtained rice biomass from another Karen village (Ban Tun) about 30 km east of the study site, which was intended to compare with the data of my study village, but it was later discarded due to inadequate resources for chemical analyses. The rice crop of Ban Tun village was not affected by the 1998 aphid infestation possibly because it was a different variety of rice. Table 4.12 Comparison of concentrations of nutrients of upland rice grain in northern Thailand Source N (%) P (%) K (%) Ca (%) M g (%) Zinke et al. (1978)* 1.3 (16) 0.4 (5) 0.3 (4) 0.1 (1) 0.1 (1) This study (1998 yield) 1.4 (20) 0.3 (5) 0.4 (5) 0.1 (1) 0.1 (1) This study (1999 yield) 1.2 (40) 0.2 (6) 0.3 (11) 0.03 (1) 0.05 (2) * = Yield from 1967. Numbers in parentheses are approximate ratios of nutrients expressed as a percentage of the nutrient in lowest abundance. 79 Table 4.13 Comparison of nutrient contents of upland rice stems in northern Laos and northern Thailand Source Year Stem N P K Ca (kg/ha) (kg/ha) (kg/ha) (kg/ha) (kg/ha) Roder et al. 1991 1,200 3.5 1.0 2.5 4.0 (1997b) This Study- 1998 1,528 11 4.6 18 1.5 4.4.7 Nitrogen Limitation in Upland Rice Productivity The upland rice crop responded positively to N , and the increase in grain biomass in the N fertilizer treatments was significant, which indicated N limitation of the soils. There are two possible explanations for this N deficiency. The buried-bag experiment in this study resulted in net immobilization of N (with some limitations on the credibility of the results; see section 3.4.7 in Chapter 3), and slash burning decreased litter and forest floor N , and N C V - N in the soil (see section 3.4.8 in Chapter 3). Changing from slash-and-burn to slash-and-mulch would reduce burning loss of N , but would make upland rice production more labor intensive. A second alternative would be to use N-fixing species during the cropping and/or the fallow period. There has been no report of natural N -fixing species in shifting cultivation fields in northern Thailand. Also, N-fixing species have long been thought to require a higher amount of P than non N-fixing species due to their relatively higher requirement for energy (as in ATP, of which P is a part of). Therefore, the use of introduced N-fixing species into the system was thought to be unsuccessful in the soils with the very low available P. However, Sprent (1999) challenged the view that N-fixing legumes require a higher amount of P. As a result, the relationship between N-fixing species and their P requirement in the shifting cultivation system warrants further investigation. 80 4.4.8 Effect of Sulfur on Upland Rice Productivity Another nutrient that is less frequently studied (than N , P, and K) in the aspect of soil fertility is sulfur (S), but this nutrient may be important in shifting cultivation in northern Thailand for four reasons. The first reason is that plants generally require S in a similar amount to P (Sanchez 1976; Blair and Lefroy 1998). The second reason is that there was a report of possible S deficiencies in soils in about 30-40% of northern Thailand (Ffoult et al. 1983). Chanakorn (1969) and Hoult et al. (1978) reported that the concentrations of phosphate-extractable SO42" in the 0-15 cm horizons from Ultisols of northern Thailand were between 5 and 21 ppm. Soils with concentrations of phosphate-extractable SO42" below 12 ppm and/or less than 1% of organic carbon concentrations are considered to be S deficient (Hoult et al. 1983). The concentrations of organic carbon of the soils (at 0-5 cm) of the shifting cultivation fields of Mae Hae Tai village were about 3.4-4.6%, but S concentration was not determined, so it was inconclusive whether the soils are S deficient. There may be a small tendency for S sufficiency in the soils of Mae Hae Tai village because the soil parent material was shale and the texture was mostly clay (30-47% clay, which increased with depths). Sandy soils (e.g. less than 15% clay in northeast Thailand) developed from consolidated sediments are more likely to be S deficient (Hoult et al. 1983). Nevertheless, Hoult et al. (1977) reported a large increase (actual numbers were not available) in grain yield of upland rice applied with S in Ultisols in northern Thailand. The third reason is that burning in shifting cultivation systems results in volatilization of nutrients that have gaseous oxides, e.g. S and N (Christanty 1986; Andriesse and Schelhaas 1987b; Kleinman et al. 1995). The last reason is that the plant available form of S (SO42") is subject to leaching, like NO3VN. 81 Besides N and P, there were also 4.8 and 2.6 kg of S/ha in each of 20 kg/ha of the N and NP fertilizers, respectively (24% of S in the N fertilizer and 13% of S in the NP fertilizer; Hagstrom 1986). The N P K fertilizer contained a very small amount of S. There were no chemical analyses for S in the soils and the rice crop of Mae Hae Tai village, so it could not be concluded whether the two largest increases in rice biomass in the N and NP fertilizer treatments were related to the presence of S in the fertilizers. However, S, unlike N and P, is at a lower risk of loss out of the system via harvested material because most S is in the rice straw (Blair and Lefroy 1998). The practice of leaving rice straw to decompose, as opposed to burning, in the cropped field conserves S within the system because between 40-60% of S in rice straw can be lost via burning (Lefroy et al. 1994). Beaton and White (1997) reported a loss of S between 40-60% from the burning of rice straw in Thailand. Aspects of S availability and crop requirement in shifting cultivation systems should receive further investigation. 4.4.9 Effect of Fallow-Period Reduction on Upland Rice Yield Interestingly, the yields reported for the shifting cultivation of the Lua (Kunstadter et al. 1978), which were similar to the 1999 yield in this study, were associated with a 10-year rotation, in comparison with the 6-year rotation of the Karen in this study. Although there were some differences in soil and vegetation between these two locations, they are climatically comparable. The major difference between these two sites was the larger fallow biomass in the 10-year rotation, which presumably provided a larger amount of ash (contributed by branches and stems of small trees and leaves and branches of large trees) following burning than the current 6-year rotation. It could be inferred from this that the upland rice, although varying in the varieties used 82 in the two studies, generally requires limited amounts of major nutrients, other than N , and the additional amount of nutrients provided in the ash after 10 years of fallow exceeded what the crop required. It is also possible that the larger biomass (i.e. bigger trees) provided by the 10-year rotation was not completely burned (the slash is burned twice, but it is still not thoroughly burned after the second burning), and as a consequence the net ash fertilizer produced may have been similar to that given by the current 6-year rotation. Therefore, fallow length beyond five years may not be nutritionally important, but there are no quantitative studies on this issue. The differences between these two studies (Kunstadter et al. 1978 and this study 1999), and the many questions that remain, limit the interpretation of comparisons between them: no firm conclusions can be drawn. However, the comparisons raise questions concerning the length of fallow needed to sustain this system, and point out the need for further, more critical studies. 4.4.10 Yield: Seed Ratio of the Upland Rice The approximate amount of rice seeds the Karen used in this study was 250 kg/ha, while it was only about 55 kg/ha in another part of northern Thailand (Hansen 1995). As a result, a yield:seed ratio (a ratio of grain yield and amount of planted seeds) of the rice crop in this study was very low (2.1 in 1998, and 3.6 in 1999) in comparison to about 19 (with a comparable rice yield of 993 kg/ha; Table 4.11) in the study by Hansen (1995). Despite the similar shifting cultivation systems, there were some management differences (e.g. rice varieties and crop rotations), which could contribute to this substantial difference. Hansen (1995) indicated that occasionally the farmers in his study area plant maize or soybean following the upland rice. In addition, it was not stated whether fertilizers were normally used in his study area, while fertilizer is not commonly 83 used in my study area. Besides the above management differences, there were also differences in soil properties, and micro-climatic conditions. The very low yield: seed ratio suggests that the yield return could be increased by using different planting techniques to decrease the amount of seeds used (Bomke, A. , Faculty of Agricultural Sciences, U B C , per. comm.). 4.4.11 Effects of Soil Diseases and Pathogens on the Second-Year Puce Crop The shoot:root ratios of the second-year rice (12-19) were much higher than those of the regular first-year rice (1.1-1.4). This suggested a dramatic decrease in the root biomass, which might have been caused by an attack of root aphids, which is reported to be common in continuous cropping of upland rice in the highlands of northern Thailand (Hansen 1995). The second-year rice also had stems that were easy to break at their base, showing that lower stems and roots had been damaged by pathogens or plant parasitic animals (e.g. nematodes). The failure of root/shoot ratios to support the conclusion that the poor performance of the second-year rice was solely or even largely due to poor nutrition demands further considerations. Hansen (1995) also indicated that the larvae of several chafer beetles (white grubs; Coleoptera: Scarabaedaes) feed on upland-rice roots, especially in fields that are cultivated for three or more years. Root nematodes and termites are also a common problem (Hansen 1995). Ventura and Watanabe (1978) attributed yield decline in the continuous planting of dryland rice in the Philippines to infectious micro-organism in soils and roots (left in the soil after harvesting). There have been several other studies that suggested "soil sickness" (e.g. fungus infection and nematodes) as a primary cause of poor upland rice in continuous cropping system (Nishizawa 84 and Ohshima 1972; IRRI 1976; Tanaka 1976; IRPJ 1977, 1978; Nishio and Kusano 1979; Aung and Prot 1990). Therefore, there are two possible factors contributing to poor rice yield in the consecutive-year cropping: nutritional and soil-disease factors, and it may be that the latter is more important than the former. The role of the fallow periods may be as much a function of pathogen control as of maintaining soil fertility. However, Hansen (1995) suggested that declining soil fertility in the continuous-cropping fields made the rice crop more susceptible to disease and pest infestation, so declining fertility may still be a major factor even though its action may be much indirect as direct. 4.5 Summary For the Karen of Mae Hae Tai village, upland rice provided about 0.9 t/ha of grain in 1999. The low value in 1998 (519 kg of grain/ha) is attributed to the aphid infestation and drought, which complicated comparisons between the 1998 and the 1999 yields. It was shown that the amounts of nutrients exported via harvested rice grain are small in relation to the nutrients stored in the soils and fallow vegetation, and do not appear to constitute a threat to the sustainability of the present system. Tabular models of nutrient flows are shown later in Chapter 7. It was shown in the fertilizer trial that both the first-year and the second-year upland rice treated with chemical fertilizers provided larger amounts of biomass than the untreated control. Therefore, nutritional factors, especially N , appeared to play a critical role in improving first-year rice productivity. A clear comparison between the first-year and the second-year rice crops could not be made because the second-year rice crop was six weeks than the first-year rice as a 85 consequence of crop damaged by cattle, but it appeared to be very unhealthy from the start of the experiment. Stem break-off and decreased root biomass in the second-year rice suggested that soil disease and parasites are a major factor in poor rice yield in continuous cropping. This conclusion supports the results of previous studies. Experiments to examine role of soil pests and pathogens and their interaction with declining soil fertility in regulating rice productivity in my study area are needed to confirm this preliminary finding. 86 Chapter 5 Trees 5.1 Introduction Trees are an integral part of the lives of the Karen, from birth to death. In the past, when a Karen child was born, his/her navel cord was cut by a bamboo blade, and tied to a certain tree so that the child was connected to the tree and nature as a whole. Upon their death, the bodies are buried within the sacred forest. As a culture, the Karen are well-known as animal hunters and plant gatherers. Many wild and domesticated plant species are utilized as food, medicine, firewood, and construction materials. They are a people of forested landscapes. Within the vicinity of a particular village, there are three main types of forest; namely conserved (including burial or cemetary forest), utilized, and watershed forests. The Karen, many of whom are Christain, bury dead bodies, and they do this within a patch of sacred forest close to their villages (an area of about 3.2 ha of the burial forest, that was sampled for soils in Chapter 3, for Mae Hae Tai village; CARE-Thailand 1997). The utilized forest in Mae Hae Tai village was the secondary forest (4.8 ha; CARE-Thailand 1997) where soils were sampled in Chapter 3. Mae Hae Tai village also has a 24 ha patch of conserved forest next to the burial forest (CARE-Thailand 1997). The villagers are allowed to extract non-timber forest products from the conserved and utilized forests, but timber extraction for construction requires permission of the village committee. The watershed forest is the area in which no major disturbance activities are allowed (non-timber forest products can be extracted for food and medicinal purposes, but timber extraction is not 87 allowed), and is usually located on ridge-tops. The Mae Hae Tai village shares a watershed forest with a nearby Karen village. In the past, the villagers were afraid of spirits that they believe were in this forest, and rarely encroached upon it. However, CARE-Thailand (1997) indicated that the Karen's conversion from animism to Christianity has resulted in less belief in untouchable spirits of the forest, and the villagers have begun to extract some timber from the watershed forest. This is unlike the conserved forest and the utilized forest where some regulations on timber extraction (by permission) were imposed (CARE-Thailand 1997). The term "forest-fallow shifting cultivation" clearly implies the presence of trees in this farming system. When the forest fallow reaches its designated age, the villagers slash it manually using hand tools. Typically, the villagers begin to clear 5-year old fallow land in February. Most trees larger than 15 cm in diameter are left uncut, but their branches are trimmed to reduce shading effect during the rice growing season. Smaller trees are cut at about 50 cm above the ground (Sutthi 1996). Many small trees have multiple stems as a result of multiple past cuttings and sprouting. The slash is left on the ground to be sun-dried for about a month before burning. The farmers normally burn the slash twice; first there is a broadcast burn of the slash, and then incompletely-burned wood is piled and the piles are burned. This results in a non-uniform distribution of the ash. Although the ash is considered to be a good source of fertilizers by the farmers, but there is no attempt to re-distribute it uniformly over the field. Ketterings et al. (1999) observed that even though farmers in Indonesia understand the fertilizer effect of ash, they also do not re-distribute the ash before planting. The second burning is also incomplete, but there is normally no third burning. 88 5.2 Objective and Rationale The objective of this part of the thesis was to estimate tree biomass and determine total litterfall in a chronosequence of shifting cultivation fields. Trees and other fallow vegetation capture nutrients from soils and store them in their biomass during the fallow period. The nutrients are circulated back to the soils via litterfall and decomposition processes. Trees are cut and burned at the end of the fallow period. Burning of the slash releases stored mineral nutrients in tree biomass to the soils to be taken up by the rice crop. It also causes a loss of N to the atmosphere. Measurement of tree biomass and estimation of its nutrient content is needed to determine the role of the trees in nutrient storage and release in the shifting cultivation system, and permit an estimate to be make of possible nutritional consequences of further reductions in the fallow period. 5.3 Methods 5.3.1 Estimation of Aboveground Tree Biomass and Slash Biomass The non-destructive sampling protocol for live tree biomass of Hairiah et al. (1999) was used to estimate the standing aboveground biomass in 2-year fallow, 4-year fallow (NW), 5-year fallow, and burial and secondary forest fields. The method is based on allometric relationship between D B H and whole-tree biomass. The live aboveground tree biomass sampling method used was as follows. Two 5 x 40 m plots were established randomly along the contour within the field. A l l trees with D B H (diameter at 1.3 m above the ground) more than 5 cm within the plots were measured. If trees had branches below 1.3 m, branches with basal diameter more than 5 cm were measured at the branch bases. A n allometric equation was then used to estimate aboveground 89 fresh biomass (Brown 1997; Hairiah et al. 1999). Because the study area had an annual rainfall below 1,500 mm (the area received about 1,200 mm; RFD 1999), the following equation was used: biomass (kg/tree) = 0.139DBH 2 3 2 , (DBH in cm, R 2 = 0.89). The line intersect sampling method (Shiver and Borders 1996) was used to estimate the biomass of slash before burning of the 5-year fallow. Nine lines (three at each slope position) of different lengths were laid as three triangles within the slashed field. The diameter in cm of all slash that intersected the lines was measured. Volume of the slash was calculated by the following equation: slash volume (m /ha) = 1.2337(sum of D )/L, (D = slash diameter in cm, and L = line length in m). Published wood density values (kg/m3; Kayama 1978; Martawijaya and Sumarni 1978) were used to convert the measured volume to fresh biomass. The fresh biomass of the standing tree and the slash was converted to dry biomass using moisture content reported by Sabhasri (1978). 5.3.2 Tree Species Identification Because one of the objectives of this work was to document nutrient content changes in soil and vegetation over the chronosequence of shifting cultivation, detailed identification of tree species was not performed. It was recognized that the nutrient concentrations and therefore the nutrient content per unit of biomass vary between species, but to have conducted this biomass/nutrient study at the species level would have been a separate study (e.g. Pampasit 1998). However, while biomass and nutrient values were not documented at the species level, a number of the tree species found in the shifting cultivation system (i.e. in the fallow areas and secondary forests) were identified and compared with species-specific data published in the literature (Kunstadter 90 1978; Nakano 1978; Sabhasri 1978; Santisuk 1988; Rerkasem and Rerkasem 1994; Schmidt-Vogt 1999). In the 5-year fallow field, the tree species (DBH > 5 cm) were directly identified in an area of 0.16 ha by Dr. Soontorn Khamyong (my Thai supervisor) of the Faculty of Agriculture, Chiang Mai University. For the other fallow fields and the forests, trees were first identified by their Karen names by an assistant from the village. The Karen names were then compared with scientific/common names lists of CARE-Thailand (1999) and Schmidt-Vogt (1999). As a result, there were some tree species with only Karen names. 5.3.3 Nutrient Contents of Aboveground Tree Biomass Shifting cultivators rely on the ash from the burned fallow biomass as fertilizer for their rice crop in the cropping period; no external inputs of fertilizer are used in the traditional system. Nutrient concentrations of trees were not determined in this study. Therefore, data from literature (Andriesse and Schelhaas 1987a; Pampasit 1998) that reported the chemical concentrations of similar tree species in combination with my biomass data were used to estimate the nutritional contribution of trees in this shifting cultivation system. 5.3.4 Litterfall Sampling One of the major roles of trees in the forest-fallow shifting cultivation system is returning nutrients back to the soil via litterfall, which is a major energy source for soil microbes and soil animals. In order to quantify this aboveground contribution of trees, the litter-trap method was employed. Twenty litter traps were installed at each of four different sites: the 2-year and 4-year 91 fallow (NW and SW), and the secondary forest, totalling 80 litter traps. The 5-year fallow field was not included because it was to be cleared and cropped before the end of the sampling period. The traps were cone-shaped with top opening of 60 cm. They were made from 2 mm nylon mesh with an aluminum frame. The other end of the trap was loosely closed with a small aluminum string for easy opening during monthly litter collection. The top height of the traps varied due to variability in tree height in the fallows and the secondary forest. The litter traps were set in December, 1998, and litter was collected monthly for one year, oven-dried at 60 °C for two days, and then weighed. 5.4 Results and Discussion 5.4.1 Biomass Estimation The number of trees in the fallow fields increased from the young to the older fallows, but was lower in the secondary forest than in the oldest fallow field. Tree biomass was greatest in the secondary and burial forests (Table 5.1). Table 5.1 Estimated tree biomass in shifting cultivation fields Field Number of Number of Estimated Estimated (in 1998) Trees Trees Standing-Tree Standing-Tree (/0.04 ha) (DBH > 5 Fresh Biomass Fresh Biomass cm) (/0.04 ha) (kg/0.04ha) (kg/m2) 2-year Fallow 52 7 57(18) 0.14 4-year Fallow-NW 103 23 267 (85) 0.67 5-year Fallow 143 48 2,151 (688) 5.38 Secondary Forest 101 55 3,162 (1,265) 7.91 Burial Forest 117 62 4,347 (1,739) 10.9 Numbers in parentheses are estimated dry biomass, based on moisture content from Sabhasri (1978). 92 A l l trees in the 2-year and 4-year fallow field were smaller than 15 cm (DBH) in the sampling area of 0.04 ha (Figure 5.1, see also Table A.9.1 in Appendix 9). A major difference between the 4-year and 5-year fallow fields was a two-time greater number of trees with D B H between 5-9 cm in the 5-year fallow than the 4-year fallow. The 5-year fallow also had seven trees between 10 to 39 cm in diameter, whereas there was none in the sampling area of 0.04 ha in the 2-year and 4-year fallow. There were some large trees in the 2-year and 4-year fallow, but they were found outside the sampling area. The number of small trees (DBH < 10 cm) was lower in the secondary and burial forests, and there were more large trees (DBH > 10 cm) in the forests than in the fallow fields. About 80% of standing-tree biomass in the 5-year falllow, the secondary forest and the burial forest was accounted for by the trees larger than 10 cm in diameter (Figure 5.2, see also Table A.9.2 in Appendix 9). The large variation in the abundance of large trees between and within individual fields represents a problem for this study and a source of error in the analysis conducted in Chapter 7. The absence of large trees in my sampling areas will have resulted in underestimation of standing biomass of trees of the entire fallow fields. However, forest-fallow shifting cultivators normally cut small trees (< 15 cm) at the end of fallow period but leave larger trees, only topping them and branch pruning to reduce shading effect on the rice. As a consequence, slash biomass (estimated in my study using the line intersect method) that is burned usually consists mainly of trunks and branches/foliage of small trees and branches/foliage of large trees, and this biomass wil l vary much less between and within fields than the large tree stem component. The main factor causing variation in the biomass compartment is field age, which was accounted for in this 93 study. Because they remain uncut, stem biomass of large trees (DBH > 15 cm) may not be nutritionally important in the shifting cultivation cycle. 100 90 80 -70 -60 50 J 1 o t-40 -1 a z 30 -20 -10 -0 • 3 to 5 cm Q 5 to 9 cm 10 to 14 cm g 15 to 19 cm 20 to 24 cm • 25 to 29 cm B 30 to 34 cm • 35 to 39 cm 2-year faDow 4-year fallow (NW) 5-year fallow Secondary forest Figure 5.1 DBH (cm) distribution of trees in shifting cultivation fields (0.04 ha) Burial forest At the end of the fallow period, there was approximately 2,150 kg/0.04 ha of standing-tree biomass estimated by the allometric equation (Table 5.1). The estimated volume of the slash in the 5-year fallow field measured by the line intersect method was 15.2 m3/ha, which was calculated to be 9.76 t/ha (or about 390 kg/0.04 ha) of dry slash biomass (Table 5.2). This indicated that 18% (390 kg out of 2,150 kg in 0.04 ha) of standing tree biomass was slashed at the end of the fallow. 94 2000 7 2-year fallow 4-year fallow (NW) 5-year fallow Secondary forest Burial forest Figure 5.2 Fresh aboveground biomass (kg/0.04 ha) of trees in shifting cultivation fields Table 5.2 Slash volume in the 5-year fallow field before burning Slope position Line length Volume (m) (m3/ha) Upper 13.5 18.2 15.7 23.4 16.0 12.5 Middle 19.7 19.1 18.5 6.7 24.1 21.0 Lower 15.3 10.7 17.6 11.8 14.0 11.4 Average 17.2 15.2 (S.E. = 0.6) Wood density (kg/W) 642* Estimated dry biomass (t/ha) 9.76 *See appendix 8 for details of wood species. 95 The burial forest had the largest biomass at about 4,300 kg per 0.04 ha. The tree biomass is greater in the burial forest (4,347 kg/0.04 ha) than in the secondary (3,162 kg/0.04 ha) forest. The burial spots were located on the lower slope of the burial forest, away from the main village road, but their location on the lower slope was random. The effect of centuries of burying bodies on the nutritional status of the soil, and its possible effects on tree biomass and nutrient content, was not investigated, although a comparison of a set of soil properties was made between the burial and the secondary forests (in October, 1998 and 1999 for the secondary forest, and the burial forest, respectively). The soil sampling in the burial forest deliberately avoided recent burial spots. Overall, there were only slightly significant differences in pH and total N in the surface soils (0-5 cm), and in pH and SOM in the 5-15 cm depth between the two forests (Table 5.3). It is not known i f older burial sites existed in the area of the burial forest that was sampled. The burial forest had a greater aboveground biomass than the secondary forest, but the surface soils were not significantly different in organic matter, available P, and extractable K . There were some significant differences in pH and total N as noted above. The differences in biomass were attributed to the timber harvesting in the secondary forest rather than any nutritional effects of the burials. 5.4.2 Comparison of Tree Biomass Amongst Various Studies Estimated aboveground tree biomass from this study was compared with other studies in northern Thailand (Table 5.4). There was a similar increase in tree biomass in the shifting cultivation practiced by the Lua (Sabhasri 1978) to that observed in this study. The increase of standing-tree biomass from the 4-year to the 5-year fallow in this study was about 8-fold (due to 96 the high proportion of large standing trees in the sampling plot, Figure 5.1), but there was only small increase of biomass from the 4-year to the 7-year fallow in the study of the Lua system (Sabhasri 1978). Table 5.3 Comparison of some soil properties in the burial and secondary forests Soil property Forest Year 0-5 cm 5-15 cm pH (1:1 in water) Secondary 1998 4.58 (0.05) 4.43 (0.02) Burial 1999 4.95 (0.04)* 4.94(0.03)* p H ( l : l i n K C l ) Secondary 1998 3.90 (0.05) 3.65 (0.02) Burial 1999 4.16(0.02)* 4.13 (0.02)* Organic Matter (%) Secondary 1998 7.88 (0.42) 4.79 (0.17) Burial 1999 7.23 (0.56) 4.02 (0.14)* Total N(%) Secondary 1998 0.37 (0.02) 0.23 (0.01) Burial 1999 0.31 (0.01)* 0.26 (0.01) Available P (ppm) Secondary 1998 2.2 (0.2) 1.6 (0.2) Burial 1999 2.7 (0.4) 2.1 (0.3) Exchageable K (ppm) Secondary 1998 193 (14) 117(5) Burial 1999 163 (11) 130(10) 1) The data were from middle slope position. Numbers in parentheses are standard errors (n=3). 2) Statistic comparisons were made within the same soil depth between the two years. 3) Asterisks indicate significant (P<0.05) differences between the secondary forest and the burial forest. The 4-year fallow field of Sabhasri (1978) had about six times (1,522 kg/0.04 ha of fresh weight) greater aboveground biomass than that of my study (267 kg/0.04 ha of fresh weight). The aboveground standing-tree biomass of my 4-year fallow field was entirely attributed to trees with D B H < 15 cm. There were no data on the size distribution of the 4-year fallow field of Sabhasri (1978), i.e. it was not known whether his 4-year fallow field consisted of trees bigger than 15 cm or not. The much larger aboveground biomass may be attributed to a few big trees (DBH > 15 cm) like my 5-year fallow field, in which 80% of the biomass was attributed to trees larger than 15 cm. 97 Table 5.4 Comparison of fallow-tree biomass (kg/0.04 ha) with other studies in northern Thailand Field Total Total Stemwood Total Stemwood Standing Standing Biomass**b Standing Biomass**0 Biomass*3 Biomass* * A b Biomass**A° 2-year Fallow 57 (18)*** n.a. n.a. n.a. n.a. 4-year Fallow 267 (85)*** 1,522 (1,035) 752 (526) n.a. n.a. 5-year Fallow 2151 (688)*** n.a. n.a. n.a. n.a. 7-year Fallow n.a. 1,577 (1036) 875 (612) n.a. n.a. 10-year Fallow n.a. 3,194 2,030 (2,092) (1,421) (2,068) (1,584) Secondary ForestA A 3,162 (1,265)*** n.a. n.a. n.a. n.a. Burial ForestA A 4,347 (1,739)*** n.a. n.a. n.a. n.a. Old Forest n.a. 13,014 n.a. n.a. n.a. (> 80 years old) (7,738) Numbers in parentheses are oven dried weight, n.a. = not available. * non-destructive method (allometric equation). **destructive method. *** Based on moisture contents from Sabhasri (1978). A Total standing-tree biomass included stems, branches, and leaves. A A age unavailable. "This study (1998),bSabhasri (1978), and cAndriesse and Schelhaas (1987a). Andriesse and Schelhaas (1987a) employed the destructive method to assess aboveground biomass of a 10-year fallow forest (2,068 kg/0.04 ha oven-dried weight of all aboveground biomass and 1,584 kg/0.04 ha of oven dried weight of stems) that was a part of the shifting cultivation system. They broke the whole-tree biomass down into branches, leaves, and trunk, and twig. The trunk accounted for the largest portion of the aboveground biomass (77%) while branches, twigs, and leaves accounted for 9, 8, and 6%, respectively. Their size distribution of trees was not described. The biomass of mature tropical forests is reported to vary from 200 to 400 t/ha, of which 75% is in branches and trunk, 15 to 20% is in surface roots, 4 to 6% in is leaves, and 1 to 2% in litter (Sanchez 1976). 98 5.4.3 Tree Species Santisuk (1988) classified forest types in northern Thailand according to ecological (climate and soils, which are influenced by elevation) and anthropogenic factors. This reflects the fact that human influences on vegetation in this area have been rather pronounced. The four main forest types are evergreen, lower montane, upper montane, and deciduous forests (Santisuk 1988). Mae Hae Tai village is located at approximately 1,000 m asl., which is an average elevation for Karen and Lua settlements. This elevation is a vegetation transitional zone from lower-elevational seasonal rain forest or tropical mixed deciduous forest to higher-elevational forests (lower montane rain forest). Some representative botanical families of the lower-elevational forests are Dipterocarpaceae (Dipterocarpus spp., and Shorea spp.) and Verbenaceae (Gmelina arborea, and Tectona grandis or teak tree). Some common families found in higher-elevational forests are Fagaceae {Castanopsis spp., and Quercus spp.) and Theaceae (Schima wallichii) (Santisuk 1988). Some common species found in fallow fields are Eugenia spp., Castanopsis spp., and S. wallichii. A number of these tree species sprout rather rapidly after the rice harvesting; the large trees sprout leaves and branches, while the small trees resprout new stems and many of them have multiple stems resulting from periodic past cuttings (up to 10 stems in Eugenia spp.). For Mae Hae Tai village, there were 9, 17, 19, 34, and 36 species, within 0.04 ha plots, in the 2-year, 4-year, 5-year fallow fields, secondary, and burial forests, respectively (Table 5.5). The total number of tree species. (DBH > 5 cm) gradually increased as the fallow ages because many tree species in the young fallow fields are saplings (i.e. D B H < 5 cm). Some common tree species found in both fallow fields and the forest are: Eugenia spp. (Myrtaceae), Castanopsis 99 spp. and Lithocarpus spp. (Fagaceae), Anneslea fragans and Schima wallichii (Theaceae), and Olea salicifolia (Oleaceae). Table 5.5 List of tree species (DBH > 5 cm in 0.04 ha) found in shifting cultivation fields and the forests Tree Species 2-year 4-year 5-year Secondary Burial Fallow Fallow Fallow Forest Forest Anneslea fragans X X X Castanopsis acuminatissima X X X X * Eugenia spp. X X X X X Gardenia sootepensis X X * Hedyotis tenelliflora X * Ilex umbellulata X Leea indica X * Lithocarpus spp. X X X X Melanorrhoea usitata X X X X Melastoma affine X Olea salicifolia X X X X Quercus sp. X * X * Schima wallichii X X X X S. roxberghii X Termimalia chebula X Total Number of Identifiable Species 4 7 11 9 7 Number of Un-identified Species with Karen Names 5 10 8 25 29 Total Number of Species 9 17 19 34 36 * Scientific names obtained by comparing Karen names with Schmidt-Vogt (1999). Sabhasri (1978) reported higher species numbers in the Lua's fallow fields. He found 55, 77, and 60 species in 0.02 ha plots in 2-year, 6-year, and 9-year fallow fields. This was reasonable because I only identified standing trees with D B H > 5 cm, while all plant species were identified in his study. Andriesse and Schelhaas (1987b) reported a total of all plant species in an 0.02 ha plot of a 10-year fallow field to be 24. Hansen (1995) identified tree species (DBH > 5 cm) in the 100 same type of forest as in this study in another part of northern Thailand. He found between 8 and 13 species of trees in sample plots of 0.02-0.03 ha. The types of tree species found in these two studies and this study were similar. The species identified in the 5-year fallow field (0.16 ha) of Mae Hae Tai village were also compared with those found in fallow fields of various ages and secondary forests in other studies in northern Thailand (Table 5.6). Many of the tree species identified are used as food, medicine, firewood, animal feed, dyes, construction materials, and for decoration and worship, fencing, and insect repellent (Kunstadter 1978; Rerkasem and Rerkasem 1994). Therefore, one disadvantage of fallow reduction is a possible loss of fallow species. The species that may be lost are those of late successional stages (i.e. species found in older secondary forests), but there are no published studies that discuss this. 5.4.4 Nutrient Concentrations of Aboveground Tree Biomass As noted earlier, the reduction of the fallow period results in a substantial reduction in accumulated aboveground biomass. Andriesse and Schelhaas (1987a) analyzed nutrient concentrations in a mixed-species tree community from a 10-year old shifting cultivation fallow forest (Table 5.7). Pampasit (1998) reported a more detailed study of nutrient compositions of different tree components (e.g. leaf, wood, and bark) of hill-evergreen-forest trees in northern Thailand, which was a similar type of forest to that found in Mae Hae Tai village (Tables 5.8-5.10). Some of the tree species found in his study were also present in this study plots, but some of the trees in the two studies were only similar at the genus level. 101 Table 5.6 List of tree species in the 5-year fallow field, Mae Hae Tai village, and other sites in northern Thailand (species and family names are from Santisuk 1988) Family Tree Species Also found in Anacardiaceae Mangifera indica M. caloneura in old forest" Rhus chinensis 4-, 7-, and 10-yrfallowb old forest0 Annonaceae Polyalthia sp. -Combretaceae Terminalia chebula T. bellerica in 7-yr fallow6 Dilleniaceae Dillenia obovata Hoogl. D. aurea in 7-and 10-yr fallow6 Dipterocarpaceae Shorea roxberghii G.Don. S. obtusa in 7-yr fallow S. talura in 4-and 7-yr fallow and old forestb Elaeocarpaceae Elaeocarpus stipularis B l . E. robustus in old forest6 Equifoliaceae Ilex umbellulata Loes 7-yr fallow6 Ericaceae Vaccinium sprenggelli Sleumer old forest0 Euphorbiaceae Antidesma sootepense Craib. Aporusa sp. 4- and 7-yr fallow0 7-yr fallow and old forest6 old forest0 Glochidion sphaerogynum 7-yr fallowb, old forest0 Kruz. Phyllantus indica P. emblica in fallowb and old forest0 Fagaceae Castanopsis acuminatissima C tribuloides in fallow and old forest6 C. indica, C. diversiflora, and C. armata in old forest0 Flacourtiaceae Flacourtia sp. 4- and 7-yr fallow0 Hypericaceae Cratoxylum spp. old forest0 Juglandaceae Engelhardtia sp. -Lauraceae Phoebe paniculata Nees old forest6 Lythraceae Lagerstroemia sp. 10-yr fallow0 Melastomataceae Melastoma qffine D.Don M. malabathricum in fallow6 M. normale in old forest0 Memecylon sp. old forestb Myrtaceae Eugenia albiflora Duth.ex Kurz old forest0 E. cumini Druce old forest0 Tristania rufescens Hance Oleaceae Olea salicifolia Wall.ex old forest0 G.Don 0. maritima in 4- and 7-yr fallowb 0. rosea in old forest0 102 Table 5.6 (Cont.) List of tree species in the 5-year fallow field, Mae Hae Tai village, and other sites in northern Thailand (species and family names are from Santisuk 1988) Family Scientific Name Also found in Rhamnaceae Zizyphus sp. Z. brunoniana and Z incurva in fallowb Z. oenoplia in old forestc Rubiaceae Gardenia sootepensis Morinda sp. Randia canthium Wendlandia tinctoria DC. R. tomentosa in 7-yr fallow*5 W. paniculata in 4- and 7-yr fallowb and old forest0 Styracaceae Styrax benzoides Craib 7-yr fallowb old forest0 Theaceae Anneslea fragrans Wall. Schima wallichii Korth. 7-yr fallow6 old forest0 4- and 7-yr fallow old forest0 Tiliaceae Colona sp. C. floribunda in old forest0 aNakano (1978). "Sabhasri (1978). cSchmidt-Vogt (1999). Table 5.7 Nutrient concentrations of different tree (10 years old, non-legume) components (Andriesse and Schelhaas 1987a) Tree Component N (%) P (% ) K ( % ) Ca(%) Mg(%) Trunk 0.2 0.04 0.3 0.5 0.06 Branch 0.3 0.04 0.4 0.9 0.1 Leaves 1.3 0.1 0.9 0.9 0.25 Fresh Leaves (This study, 1998) 1.3 0.1 0.9 0.6 0.2 (0.1) (0.01) (0.1) (0.1) (0.04) Numbers in parentheses are standard errors (n=8). Table 5.8 Nutrient concentrations (%) of fallow-tree leaves (Pampasit 1998) 103 Species No. of C N P K Ca M g Samples (%) (%) (%) (%) (%) (%) Aporosa sp.A A. villosa* 9 41 2.0 0.7 0.9 0.5 0.3 (1.4) (0.05) (0.1) (0.1) (0.04) (0.01) Castanopsis 29 47 1.8 0.6 0.8 0.4 0.2 acuminatissima^* (0.5) (0.05) (0.04) (0.03) (0.04) (0.01) Eugelhardtia sp.A E. spicata* 1 45 3.8 0.9 2.2 1.1 0.3 Eugenia albiflora Duth.ex Kurz A E. cumini DruceA E. oblata* 12 46 1.4 0.5 1.0 0.6 0.2 (0.7) (0.1) (0.05) (0.1) (0.05) (0.01) Memecylon sp.A M. celastrinum* 1 39 1.4 0.6 0.4 1.0 0.2 Phoebe paniculata NeesA P. lanceolata* 3 45 1.9 0.4 0.9 0.6 0.3 (1.3) (0.1) (0.01) (0.4) (0.1) (0.01) Schima wallichii Korth.A* 5 48 1.4 0.5 0.8 0.4 0.2 (1.7) (0.1) (0.1) (0.1) (0.08) (0.02) Styrax benzoides CraibA S. apricus* 10 47 1.8 0.4 0.9 0.8 0.2 (1.0) (0.1) (0.1) (0.1) (0.2) (0.01) Vaccinium sprenggelli 20 48 1.4 0.4 0.4 0.4 0.2 SleumerA* (0.8) (0.04) (0.04) (0.05) (0.03) (0.01) Wendlandia tinctoria D C . A * 8 47 1.8 0.6 1.7 0.4 0.2 (1.2) (0.1) (0.1) (0.1) (0.03) (0.02) Average 45.4 1.9 0.6 1.0 0.6 0.2 (1.0) (0.2) (0.04) (0.2) (0.1) (0.01) N/P N / K N/Ca N/Mg 3.2 1.9 3.2 9.5 Mixed Species 8 n.a. 1.3 0.1 0.9 0.6 0.2 (This study 1998) (0.1) (0.01) (0.1) (0.1) (0.04) N/P N / K N/Ca N / M g 13 1.4 2.2 6.5 A species found in my study. * species found in Pampasit (1998). Numbers in parentheses are standard errors, n.a = not available. Table 5.9 Nutrient concentrations (%) of fallow-tree wood (Pampasit 1998) 104 Tree Species No. of C N K Ca M g Samples (%) (%) (%) (%) (%) Aporosa sp.A A. villosa* 13 49 0.2 0.5 0.1 0.04 (0.4) (0.02) (0.03) (0.02) (0.004) Castanopsis acuminatissimaA* 36 50 0.2 0.2 0.06 0.03 (0.2) (0.01) (0.01) (0.01) (0.002) Eugelhardtia sp.A E. spicata* 2 49 0.4 0.2 0.1 0.12 (0.3) (0.01) (0.07) (0.01) (0.04) Eugenia albiflora Duth.ex Kurz A E. cumini DruceA E. oblata* 19 50 0.1 0.3 0.1 0.1 (0.3) (0.01) (0.02) (0.02) (0.01) Memecylon sp.A M. celastrinum* 1 48 0.3 0.2 0.1 0.02 Phoebe paniculata NeesA P. lanceolata* 4 49 0.1 0.1 0.03 0.03 (0.4) (0.02) (0.04) (0.002) (0.004) Schima wallichii Korth.A* 10 50 0.1 0.2 0.1 0.02 (0.3) (0.02) (0.02) (0.03) (0.003) Styrax benzoides CraibA S. apricus* 9 50 0.2 0.3 0.1 0.04 (0.1) (0.05) (0.05) (0.02) (0.01) Vaccinium sprenggelli 15 50 0.2 0.2 0.03 0.04 SleumerA* (0.1) (0.01) (0.01) (0.01) (0.01) Wendlandia tinctoria D C . A * 9 50 0.2 0.4 0.02 0.01 (0.4) (0.01) (0.03) (0.01) (0.002) Average 50 0.2 0.3 0.1 0.04 (0.2) (0.03) (0.04) (0.01) (0.01) N / K N/Ca N/Mg 0.7 2 5 A species found in my study. * species found in Pampasit (1998). Numbers in parentheses are standard errors. Table 5.10 Nutrient concentrations (%) of fallow-tree bark (Pampasit 1998) 105 Tree Species No. of C N K Ca M g Samples (%) (%) (%) (%) (%) Aporosa sp.A A. villosa* 12 47 0.6 0.4 0.3 0.1 (0.6) (0.2) (0.05) (0.1) (0.02) Castanopsis acuminatissimaA* 38 48 0.6 0.3 0.7 0.1 (0.2) (0.02) (0.02) (0.05) (0.01) Eugelhardtia sp.A E. spicata* 3 47 1.0 0.5 0.8 0.2 (1.3) (0.1) (0.3) (0.2) (0.02) Eugenia albiflora Duth.ex Kurz A E. cumini DruceA E. oblata* 21 47 0.5 0.6 0.5 0.2 (0.4) (0.04) (0.05) (0.1) (0.02) Phoebe paniculata NeesA P. lanceolata* 4 48 0.6 0.7 0.3 0.1 (0.5) (0.1) (0.04) (0.02) (0.01) Schima wallichii Korth.A* 10 48 0.6 0.6 0.4 0.1 (0.6) (0.1) (0.05) (0.1) (0.02) Styrax benzoides CraibA S. apricus* 8 48 (0.4) 0.3 (0.02) 0.4 (0.03) 0.8 (0.2) 0.1 (0.01) Vaccinium sprenggelli 14 48 0.5 0.2 0.1 0.1 SleumerA* (0.3) (0.04) (0.01) (0.01) (0.01) Wendlandia tinctoria D C . A * 8 48 0.8 0.7 0.2 0.1 (0.7) (0.1) (0.2) (0.1) (0.02) Average 47 0.6 0.5 0.5 0.1 (0.2) (0.06) (0.06) (0.09) (0.02) N / K N/Ca N/Mg 1.2 1.2 6 A species found in my study. * species found in Pampasit (1998). Numbers in parentheses are standard errors. 106 The amounts of nutrients contained in the stemwood slash at the end of the fallow in this study area were calculated by combining the estimated fresh slash biomass (9.76 t/ha, Table 5.4) of the fallow trees in the 5-year fallow field with the reported data on stemwood nutrient concentrations (0.2% N , 0.3% K , 0.1% Ca, and 0.04% Mg; Pampasit 1998, and 0.04% P; Andriesse and Schelhaas 1987a). Prior to burning, the stemwood slash of the 5-year fallow was estimated to contain approximately 20, 3.9, 29, 9.8, and 3.9 kg/ha of N , P, K , Ca, and Mg, respectively. Andriesse and Schelhaas (1987b) reported that burning resulted in a 25% reduction in N concentration in slash as it was converted to ash, but increased P, K , Ca, and M g by about 29-, 18-, 15-, and 28-fold, respectively. 5.4.5 Litterfall In Mae Hae Tai village, the amount of the aboveground litterfall in the secondary forest was about 9 t/ha in 1999, whereas it was about 5.9 t/ha for the 4-year fallow field, or about 65% of the amount of the secondary forest (Table 5.11). The 2-year fallow had about 44% of that of the secondary forest._The differences were primarily due to the differences in tree numbers and total biomass (Table 5.1). In the early stage of the fallow, leaves constituted a major part of the litterfall, e.g. 95% and 83% in the 2-year, and 4-year fallow fields, respectively. The composition of the litter was about 66% leaves in the secondary forest, with the rest being branches, bark, flowers, and fruits. Nutrient concentrations of litter were not determined in this study because of the limited funding for chemical analyses, so data from Tsutsumi et al. (1983) were used to estimate nutrient contribution of litter (Table 5.12). Their study was done in a similar type of forest to that of my 107 study, but it was in northeast Thailand. Total annual litterfall in the forest in their study was about 7.4 t/ha, compared to 9 t/ha in my study. After one year, the 5.9 t/ha of litter in the 4-year fallow re-circulated approximately 67, 6, 47, 136, 18 kg/ha of N , P, K , Ca, and Mg, respectively, to the soil. Nutrient contents of upland rice components, stemwood slash (at the end of 5 years fallow period), and litterfall (of the 4-year fallow, NW) are compared and reported in Table 5.13. Table 5.11 Aboveground litterfall (kg/ha) in shifting cultivation fields Month 2-year Fallow 4-year Fallow 4-year Fallow Secondary (NW) (SW) Forest December, 1998 413(97) 383 (90) 264 (85) 395(78) January, 1999 310(98) 609 (95) 613 (88) 982(86) February 396 (97) 843 (94) 894 (92) 962 (86) . March 641 (96) 925 (90) 684 (85) 685 (83) April 406 (94) 469 (87) 671 (77) 970 (80) May 332 (93) 443 (61) 459 (63) 1,183 (61) June 402 (97) 438(67) 356 (77) 680 (67) July 168 (86) 293 (76) 195 (72) 469 (53) August 150 (95) 339 (79) 242 (65) 532 (38) September 223 (96) 263 (88) 313(67) 439 (61) October 212 (93) 320 (79) 330(66) 373 (54) November 384 (87) 577(66) 477 (65) 1,449 (43) Total 4,037 (95) 5,902 (83) 5,498 (78) 9,119(66) Numbers in parentheses are percentage of leaf litter. Annual litterfall in the 2-year fallow re-circulated 46, 4, 32, 93, and 12 kg/ha of N , P, K , Ca, and Mg, respectively, The high cation contents of litter reflects the relatively high cation contents of the soils of Mae Hae Tai shifting cultivation fields (see Chapter 3). Burning transfers basic cations from litter and slash to the mineral soils (Andriesse and Schelhaas 1987b). The stemwood slash also re-circulated substantial amounts of nutrients to the soils, in addition to the litterfall contribution. 108 Table 5.12 Estimated nutrient contents (kg/ha) of litter Field Annual Litter N P K Ca M g (in 1998) (kg/ha) Total Litter 2-year Fallow 4,037 46 4.0 32 93 12 4-year Fallow 5,902 67 5.9 47 136 18 Secondary 9,119 104 9.1 73 210 27 Forest Leaf Litter 2-year Fallow 3,835 25 1.9 17 21 6.1 4-year Fallow 4,899 32 2.4 22 26 7.8 Secondary 6,019 39 3.0 27 33 9.6 Forest 1) Nutrient concentrations (1.14% N , 0.1% P, 0.8% K , 2.3 %Ca, and 0.3% Mg) of total litter were from Tsutsumi et al. (1983). 2) Leaf-litter nutrient concentrations of N (0.65%), P (0.05%), K (0.45%), Ca (0.54%), and M g (0.16%) were estimated to be 50, 50, 50, 90, and 80% of the nutrient concentrations of fresh leaves of this study (Table 5.7). Table 5.13 Comparison of nutrient contents (kg/ha) of upland rice, stemwood slash, and litterfall Plant component Biomass N P K Ca M g (kg/ha) Upland Rice 1998* Good grain 519 7.3 1.6 2.1 0.5 0.4 Leaves 330 4.3 4.6 18 1.5 1.2 Stems 1,528 11 0.3 3.3 0.7 0.3 1999 Good Grain 902 11 1.8 2.7 0.3 0.9 Stems + Leaves 1,225 15 0.5 21 2.5 1.2 Stemwood Slash 5-year Fallow 9,760 20 3.9 29 9.8 3.9 Total Litterfall 2-year Fallow 4,037 46 4.0 32 93 12 4-year Fallow-NW 5,902 67 5.9 47 136 18 * The grain yield was relatively low due to the aphid infestation. The aphids were grain feeders and therefore did not affect leaves and stems; the reduction of grain production may have contributed to greater leaf and stem production. 109 5.5 Summary The reduction of the fallow period has resulted in a substantial decrease in fallow biomass, which the farmers believe to be beneficial as a source of ash fertilizer for the rice crop following burning. However, slash burning causes a loss in N , which was found to be limiting in upland rice production in this study, and any additional accumulation of N in tree biomass may be lost during the burning. Therefore, the fallow reduction from 10 to five years may not be critically important for soil fertility. Shortening of the fallow period threatens the loss of some tree and other plant species that are found in later successional stages. These species are thought to invade fallow fields after five to six years (based on differences in tree species composition reported for 5-year fallow field and 10 year-fallow field in nearby area; Sabhasri 1978). 110 Chapter 6 Shrub 6.1 Introduction Weed infestation is a common phenomenon in shifting cultivation systems. Increasing competition from weeds and the human resource demands of weeding in continuously cropped fields is one reason that farmers fallow their fields. However, a number of positive functions of weeds have been recognized: erosion reduction (Toky and Ramakrishnan 1981), soil fertility maintenance (Mishra and Ramakrishnan 1984), and nutrient recycling (Swamy and Ramakrishnan 1988). This chapter describes the biomass and nutrient contents of the dominant weed species, Chromolaena odorata, in the Karen shifting cultivation system. Weed is defined as "a plant in the wrong place, being one that occurs opportunistically on land or in water that has been disturbed by human activity or on cultivated land, where it competes for nutrients, water, sunlight, or other resources with cultivated plants. Under different circumstances the plant may itself be cultivated (e.g. it may grow from seed or propagate vegetatively from the residue of a previous crop)." (Allaby 1994). C. odorata is the most dominant weed species in forest-fallow shifting cultivation fields in Mae Hae Tai village. It is also common in northeastern India (Saxena and Ramakrishnan 1983), and northern Laos (Roder et al. 1997a). It is a member of the Asteraceae (Compositae) family, and was introduced into southeastern Asia from tropical America. However, it is also known by its common name Siam weed after the former name of Thailand (Obatolu and Agboola 1993). I l l Credner (1935a cited in Schmidt-Vogt 1999) noted the wide distribution of C. odorata in northern Thailand during his geographical survey of the area. The Mae Hae Tai villagers, as well as the rural people elsewhere in Thailand, use crushed leaves of this species to stop bleeding of cut wounds. It is relatively fast growing. In the later part of the dry season (late March), following rice harvesting (in early November), C. odorata normally begins to grow rapidly and soon dominates the early stage of the fallow. In Mae Hae Tai village, this species can grow up to 2 m in height within one year after rice harvesting, and is dominant in 1- and 2-year fallow fields. The farmers of Mae Hae Tai village normally hand weed two to three times during each growing season. Roder et al. (1997a) reported that farmers in northern Laos have increased weeding frequency per growing season from an average of 1.9 times in the 1950s to 3.9 times in 1992, and attributed this to the shortening of fallow periods. Weeding is done by cutting the weed plants at ground level using curved knives, and the cut weeds are left in randomly dispersed piles in the field. Although weed control is more effective i f the weeds are manually pulled out of the ground, the farmers claim that weed pulling is more time consuming. One important implication of cutting rather than pulling is that weeds have viable rootstocks or underground stems following cutting, which results in rapid re-growth after rice harvesting. 112 MacDicken (1994) showed that C. odorata is important because of its role as re-circulator of P. P was one of the limiting nutrients in his study area and is generally limiting in the tropics. His study suggested that C. odorata appears to be an effective accumulator of P, and recommended that the use of this species as a short-rotation fallow might be a useful means of increasing P availability. In comparison, Roder et al. (1995) reported that a number of fallow species in northern Laos could accumulate significant amounts of Ca, Mg, and K within their biomass, but insignificant amounts of N and P. In the system they studied, they found that a large portion of ecosystem N and P was in the soils, especially in organic forms. C. odorata is regarded as an excellent fallow species due to its fast growth after crop harvest, high biomass production, suppression of other weed species, and rapid decomposition rate (Roder et al. 1995). Soil structure is better under C. odorata than under bamboo cover according to farmers who were interviewed in northern Laos; Roder et al. (1995) recommended further investigation of the effects of this species on nutrient mobilization and soil biological, chemical, and physical properties. The farmers in northern Thailand and northern Laos prefer C. odorata because it flowers once a year, makes good organic fertilizer when composted, and is easy to manage (Rerkasem and Rerkasem 1994; Roder et al. 1997a). Obatolu and Agboola (1993) showed that it decomposed faster than other weeds, which is probably due to its low C/N and lignin/N ratios. 6.2 Objective and Rationale The objective of this component of the study was to measure biomass accumulation and nutrient contents of C. odorata in the first year of fallow. 113 Fallow vegetation captures nutrients from soils and stores them in their biomass. Quantification of this storage is necessary in order to understand the overall biogeochemistry of this farming system. Fast growing early serai species can capture and hold nutrients that might otherwise be lost by leaching, denitrification, or P sorption by mineral soils before slower growing woody fallow vegetation can recover. The farmers in northern Thailand and northern Laos believe that C. odorata helps to improve soil fertility (Rerkasem and Rerkasem 1994; Roder et al. 1997a). The analysis of the nutrient contents and contribution of C. odorata to soil fertility was conducted in part to evaluate this traditional knowledge. 6.3 Methods Biomass sampling of C. odorata was carried out three times in 1999 following the harvest of the 1998 upland rice field of the study chronosequence: in May, August, and October. During the dry season after the harvest (November to April) its growth was small. The field was divided into three slope positions: lower, middle, and upper. Five 50x50 cm quadrats were placed along the contour at each slope position, for a total of 15 quadrats in the field for each sampling time. For the first two samplings, the weeds within the quadrats were entirely pulled from the ground to obtain whole-plant biomass. Roots were not sampled at the last sampling in October. Data for root biomass at 12-months after rice harvest were calculated based on shootroot ratio of C. odorata at the same age from Saxena and Ramakrishnan (1983). The samples were air-dried, and a 20% sub-sample was oven-dried at 60 °C for moisture correction. The samples were then separated into leaf, stem and root components for weighing and chemical analyses of N , P, K , Ca, and M g (see 4.3.2 in Chapter 4 for the analysis methods). 114 6.4 Results and Discussion 6.4.1 Biomass of C. odorata About seven months after the rice harvest, C. odorata reached about 1.6 t/ha of total aboveground biomass with relatively equal proportions of leaf and stem components, and about 300 kg/ha of roots (Table 6.1). At 10-months, the total aboveground biomass was about double that of seven months after the rice harvest. At the end of the first growing season, the aboveground biomass reached 7.1 t/ha with almost 80% as stem biomass. Table 6.1 Biomass (kg/ha; dried at 60 °C) accumulation of C. odorata, Mae Hae Tai village, Thailand, and other studies Sampling Time Leaf Stem Total Root Shoot:Root Aboveground Ratio 7-month after rice harvest 842 722 1,564 295 5:1 (May 1999) (62) (99) (160) (40) Moisture Content (%) - 4.9 - 3.1 10-month after rice harvest 1,192 1,843 3,035 464 6.5:1 (August 1999) (66) (109) (148) (22) Moisture Content (%) 3.0 5.0 - 3.1 12-month after rice harvest 1,500 5,643 7,143 1,553* 4.6:1* (October 1999) (62) (196) (136) Moisture Content (%) 8.4 8.9 - -Other studies Northeastern 239 a D 617a D 856aD - -Thailand 691 a c l,674 a c 2,365ac -(Tsutsumi et al. 1983) 850 c d 5,950cd 6,800cd -12-month old - - 8,752 1,920 4.6:1 Northeastern India (Saxena and Ramakrishnan 1983) Numbers in parentheses are standard errors (n=15). * Root biomass was estimated based on shootroot ratio of Saxena and Ramakrishnan (1983). a 10-month old. b N o burning. 0 Burning. d 22-month old. 115 Saxena and Ramakrishnan (1983) reported a slightly larger total aboveground biomass (8.7 t/ha) of C. odorata after one growing season in northeastern India. In northeastern Thailand, the species reached a total aboveground biomass of 6.8 t/ha after two growing seasons (Tsutsumi et al. 1983). Tsutsumi et al. (1983) also found that experimental burning can increase total aboveground biomass of C. odorata after burning. 6.4.2 Nutrient Contents of C. odorata At the end of one growing season, leaf concentrations of N , P, K , Ca, and M g were 2.5, 0.4, 1.8, 0.8, and 0.7%, respectively (Table 6.2). These concentrations were higher than those of its stem component. However, the total amounts of nutrients, except N and Mg, were greater in the stem component because it accounted for about 80% of the total aboveground biomass. N concentration in the stem gradually declined while the other nutrients in the stem remained rather constant throughout the growing season. If it is assumed that no nutrients are re-translocated before leaves drop, the nutrients re-circulated to the soils each year by C. odorata leaf litter were 38, 6, 27, 12, and 11 kg/ha of N , P, K , Ca, and Mg, respectively. In comparison, Hairiah et al. (1996) reported that, in Nigeria, C. odorata litter returned 65-80, 5-10, and 16-24 kg/ha of N , P, and K , respectively; the amount of leaf litter they reported (4-4.5 t/ha) was higher than the 1.5 t/ha measured in my study. However, their site was in the humid zone, in which vegetation growth rate is higher than in my study area. In Indonesia, where annual rainfall is about 2,500 mm, C. odorata was reported to reach a total aboveground biomass of 12.5 t/ha after one year (Hairiah et al. 1996). A comparison of nutrient concentrations of C. odorata components grown in different soil orders is presented in Table 6.3. 116 Table 6.2 Nutrient concentrations and contents of C. odorata Age Nutrient Leaf Stem Root Total (%) (kg/ha) (%) (kg/ha) (%) (kg/ha) (kg/ha) Seven Months N 2.7 23 1.1 7.9 1.0 3.0 34 (after rice (0.2) (0.2) (0.1) harvest, May P 0.3 2.5 0.2 1.4 0.1 0.3 4.2 1999) (0.03) (0.01) (0.01) K 2.0 17 2.8 20 1.5 4.4 41 (0.03) (0.05) (0.05) Ca 0.8 6.7 0.4 2.9 0.03 0.1 9.7 (0.01) (0.02) (0) M g 0.5 4.2 0.2 1.4 0.2 0.6 6.2 (0.03) (0.01) (0.01) 10 Months N 2.4 29 0.7 13 0.9 4.2 46 (after rice (0.03) (0.04) (0.02) harvest, P 0.4 4.8 0.2 3.7 0.1 0.5 9 August 1999) (0) (0) (0) K 1.7 20 2 37 1.4 6.5 64 (0.1) (0.1) (0.07) Ca 0.9 11 0.3 5.5 0.02 0.1 17 (0.1) (0.02) (0) M g 0.6 7.2 0.1 1.8 0.1 0.5 9.5 (0.05) (0.01) (0.01) 12 Months N 2.5 38 0.6 34 0.9* 14** 86 (after rice (0.04) (0.01) (0.02) harvest, P 0.4 6.0 0.2 11 0.1* 1.6** 19 October 1999) (0.02) (0) (0) K 1.8 27 1.6 90 1.4* 22** 139 (0.03) (0.06) (0.07) Ca 0.8 12 0.3 17 0.02* 0.3** 29 (0.09) (0.04) (0) M g 0.7 11 0.1 5.6 0.1* 1.6** 18 (0.02) (0.01) (0.01) Numbers in parentheses are standard errors (n=3). * Nutrient concentrations were from the 10-month old C. odorata. ** Root biomass was estimated based on shootroot ratio of Saxena and Ramakrishnan (1983). Table 6.3 Nutrient concentrations (%) of C. odorata reported from various studies 117 Component N P K Ca Mg Soil Order Source Leaf 2.5 0.4 1.8 0.8 0.7 Ultisols This study (1999) Stem 0.6 0.2 1.6 0.3 0.1 (Northern Thailand) Leaf+Stem 2.2 1.0 2.5 0.5 0.4 Alfisols Obatolu and Agboola (Nigeria) (1993) Leaf 3.7 0.5 2.6 0.1 0.2 Entisols and MacDicken (1994) Stem 2.7 0.2 2.9 0.1 0.1 Inceptisols Whole Plant 2.4 0.3 2.4 0.1 0.2 (Philippines) Leaf 2.1 0.1 2.0 0.6 n.a. Northern Laos* Roderetal. (1995) Stem 0.5 0.04 1.0 0.3 n.a. * Soil order not available, n.a. = not available. The weeding practice of the Karen shifting cultivators normally leaves the belowground component alive throughout the cropping season, and as a result weeds can establish themselves rapidly, thereby benefiting from the flush of nutrients releasing from decomposing non-harvested materials (leaves, stems, and roots) in the soil following the rice harvest. However, C. odorata can also establish itself by sprouting rather rapidly in areas of a cropped field that have been nutrient depleted. 6.4.3 Nutrient Contents of C. odorata. Upland Rice, and Tree Components Nutrient concentrations of C. odorata and upland rice were compared to evaluate the ability of C. odorata to utilize soil nutrients in the soil after the rice harvest. Leaves of the upland rice had nutrient concentrations of 1.3, 0.1, 1.0, 0.2, and 0.1% of N , P, K , Ca, and Mg, respectively, which are lower compared with concentrations for C. odorata leaves (2.5, 0.4, 1.8, 0.8, and 0.7%, respectively) (Table 6.4). The N concentration of C. odorata leaves was almost double that of upland rice. The concentrations of nutrients in stems were comparable between the two plants, and K constituted the highest concentration in the stem. At the end of one growing 118 season, C. odorata reached approximately 7.1 t/ha of aboveground biomass, which was about four and six times higher than that of the upland-rice aboveground biomass, excluding grain, in 1998 and 1999, respectively. The larger aboveground biomass of C. odorata resulted in the greater nutrient contents after one year. C. odorata leaves contained the highest concentration of N (2.5%) following by the tree leaves (1.9% Pampasit 1998; and 1.3% this study), and the upland rice (1.3%) (Table 6.4). The P concentration of C. odorata leaves (0.4%) was second to that of the trees (0.6%). MacDicken (1994) reported that C. odorata is a good accumulator of P. This perennial species also appeared to be a good accumulator of K , Ca, and M g with concentrations of 1.8, 0.8, and 0.7%, respectively, in its leaves. Saxena and Ramakrishnan (1983) concluded that C. odorata has a great ability to accumulate a large amount of biomass and nutrients, especially in its stem, in comparison to sprouting perennial weed species such as the tropical grass Imperata cylindrica, which accumulates more in the leaf component. 6.5 Summary C. odorata can achieve a total aboveground biomass of about 7 t/ha within one year of fallow. Its leaves are approximately 2.5% N . At the end of one fallow-year, C. odorata aboveground biomass contained about 72, 17, 117, 29, and 17 kg/ha of N , P, K , Ca, and Mg, respectively. The nutrient contribution of this species diminishes over the 5-year fallow period because it gets shaded by re-sprouting trees. 119 Table 6.4 Comparison of nutrient concentrations and contents of upland rice, C. odorata, and tree components Plant Nutrient Good Grain Leaf Stem Leaf+ (%) (kg/ha) (%) (kg/ha) (%) (kg/ha) Stem (kg/ha) Upland Rice Biomass - 519 - 330 - 1,528 1,858 (1998) (kg/ha) N 1.4 7.3 1.3 4.3 0.7 11 15 P 0.3 1.6 0.1 0.3 0.3 4.6 4.9 K 0.4 2.1 1.0 3.3 1.2 18 21 Ca 0.1 0.5 0.2 0.7 0.1 1.5 2.2 M g 0.1 0.4 0.1 0.3 0.1 1.5 1.8 Upland Rice Biomass - 902 - - - 1,225* 1,225 (1999) (kg/ha) N 1.2 11 - - 1.2* 15* 15 P 0.2 1.8 - - 0.04* 0.5* 0.5 K 0.3 2.7 - - 1.7* 21* 21 Ca 0.03 0.3 - - 0.2* 2.5* 2.5 Mg 0.1 0.9 - - 0.1* 1.2* 1.2 C. odorata Biomass - - - 1,500 - 5,643 7,143 (12 months (kg/ha) old) N - - 2.5 38 0.6 34 72 P - - 0.4 6.0 0.2 11 17 K - - 1.8 27 1.6 90 117 Ca - - 0.8 12 0.3 17 29 M g - - 0.7 11 0.1 5.6 17 Tree Biomass - - - - - - -(This Study) (kg/ha) N - - 1.3 - - - -P - - 0.1 - - - -K - - 0.9 - - - -Ca - - 0.6 - - - -M g - - 0.2 - - - -Tree Biomass - - - - - - -(Pampasit (kg/ha) 1998) N - - 1.9 - 0.2 - -P - - 0.6 - - - -K - - 1.0 - 0.3 - -Ca - - 0.6 - 0.1 - -M g - - 0.2 - 0.04 - -* Leaf and stem components of the 1999 upland rice were combined before weighing and chemical analyses. 120 C. odorata appears to have a great ability to extract large amounts of soil nutrients after rice harvesting. Therefore, an important role played by C. odorata appears to be its ability to reduce leaching losses from cropping fields by rapid nutrient uptake and biomass accumulation at the early stage of the fallow. These nutrients are assumed to be passed on to the trees at the later stage of the fallow as the C. odorata is shaded out, and to the next rice crop as the fallow is slashed and burned. Much of the N recovered by the vegetation is assumed to be returned to the mineral soil via litterfall incorporation into the mineral layers by soil animals, and fine root turnover, where it is protected from loss at the time of the burning. 121 Chapter 7 Synthesis of the Reduced Forest-Fallow Shifting Cultivation 7.1 Introduction Shifting cultivation fields appear devastating to those not familiar with this agricultural system, especially during the stages of slashing and burning. Similarly, a freshly ploughed field may look devastated to those who have never seen ploughing before. The visual appearance has led to the perception by the public that shifting cultivation is very degrading of local and landscape ecosystems. However, as stated earlier there is more than one type of shifting cultivation system based on periods of cropping and fallowing: 1) short-cropping and short-fallow periods, 2) short-cropping and long-fallow periods, and 3) long-cropping and very long fallow periods (Kunstadter etal. 1978). Although shifting cultivation systems have often been labeled as a destructive form of agriculture, some ecologists believe that the short-crop-and-long-fallow shifting cultivation system can be sustainable, provided the length of the fallow provides sufficient time for the soil to recover. However, relatively little has been said about other ecological roles the fallow might play that may not be fulfilled i f the fallow is too short. The short-crop-and-long-fallow shifting cultivation system, which is also called rotational forest-fallow shifting (swidden) cultivation, was recently classified as a sequential agroforestry system (Sanchez 1995; Thomas 1995). Forest-fallow shifting cultivation of the Karen was the focus of this thesis. 122 The shifting cultivation practices used by certain groups of minorities living in mountainous areas of Thailand are changing (Bass and Morrison 1994; Ganjanapan 1994; Rerkasem and Rerkasem 1994). This is partly due to increasing accessibility by roads, which has resulted in the introduction of modern farming technology and improved access to markets. These changes, in combination with increasing population and land limitation, have resulted in shorter fallow periods. In some cases, the fallow period has disappeared, and has been replaced by continuous cropping for commercial rather than subsistence purposes; field abandonment, which has universally been associated with shifting cultivation, no longer occurs. In remote villages where the crop-fallow shifting cultivation system still exists, the fallow period is becoming shorter because of population increase and there is growing concern that this is threatening the productivity and sustainability of the system. 7.2 The Role of Fallow in Shifting Cultivation System: Alternative Hypotheses The term "forest-fallow shifting cultivation" clearly implies the presence of trees in this farming system. The fallow period is believed to play a number of important roles. The nutritional role of the fallow period is the major focus of this thesis work, but in addition to the important role of fallow in soil fertility maintenance, there are other possible roles as suggested in the following alternative hypotheses: 123 1) Maintaining good soil structure A n alternative hypothesis to the dominant and traditional nutrient hypothesis with regard to the role of fallow trees is the role of fallow as the maintainer of soil structure. This hypothesis suggests that soil physical properties (e.g. bulk density, aeration, hydraulic conductivity, and water-holding capacity) should be superior in shifting cultivation systems than in continuous cropping systems, and that this is as important as, or more important than, the nutritional considerations of fallow. When most fallow trees are slashed, their belowground component is still viable and re-sprouts rapidly during the cropping stage and early fallow period. Trees are normally deep rooted, so one of the roles of their belowground component should be to maintain soil structure by the production and turnover of roots, and provision of litterfall and dead roots as an energy source for soil microbes and animals that play an important role in creating and maintaining soil structure. The farmers of Mae Hae Tai village cited the presence of earthworms as an indicator of good soils. 2) Lessening soil erosion effects Soil erosion hazard has always been regarded as one of the major disadvantages in sloping-land agriculture. Various studies on soil erosion potential and actual erosion problems in northern Thailand, northern Laos, and elsewhere, yielded highly variable results (Table 3.27, Chapter 3) due to landscape variability and different measurement methods. Soil erosion problems within the shifting cultivation fields of Mae Hae Tai village were slight to moderate (Anongrak and 124 Seramethakul 1999). One of the arguments is that the crop-fallow shifting cultivation practice of the Karen provides an almost continuous canopy cover, which undoubtedly lessens soil erosion hazards. The only chance of significant soil erosion is when the soil is bare between the burning (March) and rice establishment (April) stages, which is normally a dry season (March and April). The average (1985-1999) monthly rainfall in March and April was 11 mm (S.E. = 8, n = 15, eight years with zero rainfall) and 50 mm (S.E. = 13, n = 15, five years with zero rainfall), respectively (RFD 1999). Therefore, soil erosion potential caused by rain should be rather small, which leaves only the erosion potential caused by wind, which has not been studied. Thus, a second alternative hypothesis with regard to the role of the fallow is its role in lessening soil erosion hazards due to continuous canopy cover: soil erosion in shifting cultivation should be lower than that of a continuous cropping system 3) Providing usable plants and animals for consumption and medicinal purposes The Karen, and other hill minority tribes, are known for their utilization of wild plants and animals for various purposes such as food and medicines, and fallow fields are one of the sources for these plants and animals. If the fallow period has to be further reduced, some changes in the species composition of plant communities in the fallow areas would be expected; some species found in older-fallow fields would disappear from the landscape. In a continuous-cropping, fixed-field system there would be a significant loss of species unless there was an adequate system of forest reserves. Thus, an alternative hypothesis concerning the importance of the 125 fallow to the Karen people is that it provides useful plants and animals. The shifting cultivation system with fallow period should be more beneficial than further-reduced-fallow shifting cultivation and fixed-field agriculture. 4) Control of weeds After five years, most of the herb and shrub weeds have been reduced by tree shading. Weed problems are thus much less following a period of four or five years of fallow, than with no fallow or a very short (one to two year) fallow. 5) Control of pests and diseases of crop species Monoculture of non-native crops may require insecticides and fungicides i f they are to be grown in sequential years. Fallow eliminates or greatly reduces these crop pests. However, alternating food crop (e.g. corn and rice) may eliminate or reduce this problem. 7.3 Biogeochemistry of Forest-Fallow Shifting Cultivation This thesis is built on the ecosystem concept of "biogeochemistry", which includes the nutrient pathways of a particular ecosystem. Three pathways are recognized; geochemical, biochemical, and biogeochemical (Kimmins 1997). In this study of shifting cultivation, nutrients bound in soils and plants are considered to play a key role in determining the productivity of the system. 126 The main components of the biogeochemistry of forest-fallow shifting cultivation are ecosystem compartments (soil and plants), inputs (precipitation and soil weathering), outputs (crop harvesting, erosion, firewood, and leaching), and re-circulation (crop and plant uptake and return by above and belowground litterfall and foliar leaching). Nutrient pathways of five macronutrients (N, P, K , Ca, and Mg) are presented here in order to analyze system stability, i.e. an ecosystem will be likely to decline in productivity i f nutrient outputs exceed nutrient inputs, and i f this leads to a biologically significant decline in soil fertility. Sulfur, which may be important in the biogeochemistry of this system, was not measured or included in the analyis. 7.3.1 Data Used in Biogeochemistry of Forest-Fallow Shifting Cultivation Data used in the nutrient pathways were from this study, published literature, and unpublished reports and theses as follow: 1) Ecosystem 1.1 Soil (Chapter 3). 1.2 Tree, biomass (Table 5.1, Chapter 5) and nutrient concentrations (Tables 5.7 and 5.9, Chapter 5). 2) Inputs 2.1 Rainfall. The measured rainfall data were from the months of August and September 1999, whose rainfall represented about 33% of the total annual rainfall in 1999. These data were extrapolated to estimate nutrient contribution from the annual 127 rainfall. 2.2 Weathering. There were no published data on nutrient input via weathering processes in Thailand and surrounding areas; therefore, a reasonable estimate could not be made. 3) Within-Ecosystem Re-circulation 3.1 Rice biomass retained (leaves, stems and roots) after harvest (Table 4.2, Chapter 4). 3.2 Above- and belowground biomass and nutrient contribution of C. odorata, assuming it is shaded out and dies at the end of the second year (Tables 6.1 and 6.2, Chapter 6). 3.3 Litterfall mass and nutrient contents (Table 5.12, Chapter 5). 3.4 Fine root biomass from a similar secondary forest (Sundarapandian and Swamy 1996), and nutrient concentrations (0.65% N , 0.05% P, 0.45% K , 0.54% Ca, and 0.16% Mg) that were estimated to be 50% of nutrient concentrations of fresh leaves (from this study) for N , P, and K , but they were estimated to be 90 and 80% for Ca, and Mg, respectively, because these two nutrients are re-absorbed (internal cycling) at lower rates before plant tissue senescense than are N , P, and K(Kimmins 1997). 4) Outputs 4.1 Harvested material (good grain) of upland rice (Table 4.2, Chapter 4). 4.2 Firewood, approximately 75 kg/ha/year (Zinke et al. 1978), and concentrations of N , 128 K , Ca, and M g were 0.2, 0.3, 0.1, and 0.04%, respectively (Table 5.9, Chapter 5; Pampasit 1998). Concentration of P was 0.04% (Andriesse and Schelhaas 1987a). 4.3 Burning. Andriesse and Schelhaas (1987b) reported that burning resulted in a 25% reduction in N concentration in slash as it was converted to ash, but increased P, K , Ca, and M g by about 29-, 18-, 15-, and 28-fold, respectively. The transfer of P, K , Ca, and Mg could be considered as a part of nutrient re-circulation, and it was not accounted for in this study. Toky and Ramakrishnan (1981) reported that about 28%) of ash could be lost by strong winds during the dry season (March-April) in a shifting cultivation system in northeastern India; hence, some of the ash nutrients may be lost. Therefore, the gain of P, K , Ca, and M g could be partially reduced. Only the direct loss of N via burning was considered in the nutrient budget sheet. Prior to the burning in this study, the stemwood slash and the leaf litter of the 5-year fallow were estimated to contain approximately 20 and 35 kg of N/ha, respectively. As a result, about 14 kg of N/ha could be lost by burning. 4.4 Leaching. There were no published data on nutrient output via leaching in Thailand, so a reasonable approximation could not be made. As a result, I estimated amounts of nutrients leached based on values reported in a study of run-off and infiltration losses in shifting cultivation system (Jhum) in northeastern India (Toky and Ramakrishnan 1981). Their study site was located on sloping hills at 129 26° N , while my study site was located at 19° N . Their annual rainfall was 1,420 mm in 1978 in comparison to 1,340 mm in 1999 at my study site. Their reported nutrient leaching values in the cropped field was 9.2, 0.07, 14, 4.6, and 2.3 kg/ha for N (as NO3VN), P, K , Ca, and Mg, respectively. The reported leaching values for the 5-year fallow field were 1.1, 0.02, 0.5, 2.7, and 0.9 kg/ha of N (as N 0 3 ~ -N), P, K , Ca, and Mg, respectively. The reported leaching values for the 10-year fallow field were 0.5, 0.01, 0.2, 1.6, and 0.5 kg/ha of N (as N O 3 -N) , P, K , Ca, and Mg, respectively, and these values were used for the secondary forest. I estimated the leaching values for the 2-year and 4-year fallow field to be between those of the cropped field and the 5-year fallow. 4.5 Erosion. Loss of sediment from soil erosion was taken from a study of run-off and infiltration losses in forest-fallow shifting cultivation on sloping hills in northeastern India (Toky and Ramakrishnan 1981). The losses of sediment in their study were 30 and 1 t/ha in crop field and 5-year fallow field, respectively. Based on these reported data, I estimated sediment losses in the 2-year and 4-year fallow fields, and the secondary forest to be 20, 10, and 0.5 t/ha, respectively. Potential amounts of lost nutrients were then calculated from my own soil nutrient contents (Chapter 3). For N , it was assumed that only N C V - N could be lost in surface run-off sediment. Concentrations of N C V - N measured in the rainy season in this study were 3.5, 12, 2.4, 1.9, and 10 ppm for the cropped field, the 2-year fallow, 4-year 130 fallow, 5-year fallow, and the secondary forest, respectively (Table 3.12, Chapter 3). 7.3.2 Discussion of Nutrient Budgets Nutrient pathways of N , P, K , Ca and M g are presented in Tables 7.1-7.5. The ecosystem of forest-fallow shifting cultivation had about 6.4, 6.3-6.6, and 6.9 t/ha of total N in soil and tree compartments combined in the cropped field, fallow fields, and the secondary forest, respectively (Table 7.1). Almost all (95-99%) of the total ecosystem N in shifting cultivation fields was found in the soils, rather than in the tree component. The total ecosystem N content was several-fold higher than the contents of available P, extractable K , Ca, and M g (Tables 7.2-7.5) because all forms of N in the soils (e.g. organic and inorganic nitrogen) were determined. The ecosystem content of P was relatively low because the values presented are available P (Table 7.2). P tends to be relatively unavailable in tropical soils because low soil pH and high contents of soil Fe and A l typically result in soil sorption of P. Hansen (1995) reported that, in a laboratory P sorption study, the 0-15 cm horizons from Ultisols in the highland of northern Thailand could adsorb up to 80% of added P, the highest P sorption capacity in his studies. The ecosystem contents of extractable K , Ca, and M g were highest in the cropped field due to ash input to the soils (Tables 7.3-7.5). 131 Table 7.1 Nutrient pathways of total N (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers. Ecosystem Compartment Cropped 2-year 4-year 5-year Secondary (1998) Field Fallow Fallow Fallow Forest Ecosvstem Inventories Soil 0-5 cm 1,602 1,300 1,586 1,612 1,690 5-15 cm 2,380 2,267 2,289 2,300 2,154 15-30 cm 2,457 2,709 2,661 2,485 2,669 Tree (aboveground) - 1.8 8.5 69 127 Tree (belowground) - 0.3 1.3 10 19 Input (annual) Rainfall 0.9 0.9 0.9 0.9 0.9 Re-circulation (annual) Upland Rice Retained 16 - - - -C. odorata Aboveground - 72 - - -Belowground - 14 - - -Litter (aboveground) Leaf - 25 32 35 39 Total - 46 67 86 104 Tree (fine roots) 1.3 2.6 9.1 11 12 Output (annual) Upland Rice Harvested 11 - - - -Firewood - - 0.15 0.15 0.15 Leaching 9.2 6.1 3.1 1.1 0.5 Erosion 0.1 0.2 0.02 0.002 0.01 Burning 14 - - - -Total Ecosystem Inventories 6,439 6,278 6,546 6,476 6,659 Input (annual) 0.9 0.9 0.9 0.9 0.9 Re-circulation (annual) 17 135 76 97 116 Output (annual) 34 6.3 3.3 1.3 0.7 Output-Input (annual) 33 5.4 2.4 0.4 -0.2 Output-Input (6-Year) 41 132 Table 7.2 Nutrient pathways of available P (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers Ecosystem Compartment (1998) Cropped Field 2-year Fallow 4-year Fallow 5-year Fallow Secondary Forest Ecosvstem Inventories Soil 0-5 cm 11 1.5 1.3 1.5 1.0 5-15 cm 2.7 0.8 0.9 0.8 0.7 15-30 cm 2.1 0.6 0.8 0.6 0.8 Tree (aboveground) Tree (belowground) _ 0.2 0.03 0.9 0.1 6.9 1.0 13 1.9 Input (annual) Rainfall 0.6 0.6 0.6 0.6 0.6 Re-circulation (annual) Upland Rice Retained 0.6 _ _ — C. odorata Aboveground Belowground - 17 1.6 - - -Litter (aboveground) Leaf Total - 1.9 4.0 2.4 5.9 2.0 8.0 3.0 9.1 Tree (fine roots) 0.1 0.2 0.7 0.9 0.9 Output (annual) Upland Rice Harvested 1.8 _ _ _ Firewood - - 0.03 0.03 0.03 Leaching 0.07 0.05 0.02 0.02 0.01 Erosion 0.7 0.1 0.03 0.003 0.001 Total Ecosystem Inventories 16 3.1 4.0 11 17 Input (annual) 0.6 0.6 0.6 0.6 0.6 Re-circulation (annual) 0.7 23 6.6 8.9 10 Output (annual) 2.6 0.15 0.08 0.05 0.04 Output-Input (annual) 2.0 -0.5 -0.5 -0.6 -0.6 Output-Input (6-Year) -0.1 133 Table 7.3 Nutrient pathways of extractable K (kg/ha) of shifting cultivation fields. Ecosystem Inventories and annual transfers Ecosystem Compartment (1998) Cropped Field 2-year Fallow 4-year Fallow 5-year Fallow Secondary Forest Ecosvstem Inventories Soil 0-5 cm 186 70 100 109 88 5-15 cm 165 45 78 99 55 15-30 cm 197 40 66 89 50 Tree (aboveground) Tree (belowground) _ 1.7 0.3 8.3 1.2 67 10 123 18 Input (annual) Rainfall 2.7 2.7 2.7 2.7 2.7 Re-circulation (annual) Upland Rice Retained 23 _ _ _ C. odorata Aboveground Belowground - 117 22 - - -Litter (aboveground) Leaf Total - 17 32 22 47 24 59 27 73 Tree (fine roots) 0.9 1.7 6.1 7.7 8.1 Output (annual) Upland Rice Harvested 2.7 _ _ _ Firewood - - 0.23 0.23 0.23 Leaching 14 9.3 4.7 0.5 0.2 Erosion 11 2.8 2 0.2 0.1 Total Ecosystem Inventories 548 157 254 374 334 Input (annual) 2.7 2.7 2.7 2.7 2.7 Re-circulation (annual) 24 173 53 67 81 Output (annual) 28 12 6.9 0.9 0.5 Output-Input (annual) 25 9.3 4.2 -1.8 -2.2 Output-Input (6-Year) 35 134 Table 7.4 Nutrient pathways of extractable Ca (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers Ecosystem Compartment (1998) Cropped Field 2-year Fallow 4-year Fallow 5-year Fallow Secondary Forest Ecosystem Inventories Soil 0-5 cm 954 208 340 160 83 5-15 cm 735 111 287 80 24 15-30 cm 546 75 93 95 22 Tree (aboveground) Tree (belowground) _ 1.2 0.2 5.6 0.8 46 6.8 84 13 Input (annual) Rainfall 0.5 0.5 0.5 0.5 0.5 Re-circulation (annual) Upland Rice Retained 2.7 _ _ _ _ C. odorata Aboveground Belowground - 29 0.3 - - -Litter (aboveground) Leaf Total - 21 93 26 136 31 171 33 210 Tree (fine roots) 1.1 2.2 7.6 9.2 9.7 Output (annual) Upland Rice Harvested 0.3 _ _ _ _ Firewood - - 0.08 0.08 0.08 Leaching 4.6 3.1 2.9 2.7 1.6 Erosion 55 8.3 6.7 0.3 0.1 Total Ecosystem Inventories 2,235 395 726 388 226 Input (annual) 0.5 0.5 0.5 0.5 0.5 Re-circulation (annual) 3.8 125 144 180 220 Output (annual) 60 11 9.7 3.1 1.8 Output-Input (annual) 59.5 10.5 9.2 2.6 1.3 Output-Input (6-Year) 83 135 Table 7.5 Nutrient pathways of extractable M g (kg/ha) of shifting cultivation fields. Ecosystem inventories and annual transfers Ecosystem Compartment (1998) Cropped Field 2-year Fallow 4-year Fallow 5-year Fallow Secondary Forest Ecosystem Inventories Soil 0-5 cm 105 44 58 52 29 5-15 cm 130 23 27 33 8.8 15-30 cm 78 15 19 25 7.6 Tree (aboveground) Tree (belowground) _ 0.3 0.05 1.6 0.2 13 1.9 23 3.5 Input (annual) Rainfall 0.3 0.3 0.3 0.3 0.3 Re-circulation (annual) Upland Rice Retained 1.4 _ _ _ C. odorata Aboveground Belowground - 17 1.6 - - -Litter (aboveground) Leaf Total - 6.1 12 7.8 18 8.0 21 9.6 27 Tree (fine roots) 0.3 0.6 2.2 2.7 2.9 Output (annual) Upland Rice Harvested 0.9 _ _ _ _ Firewood - - 0.03 0.03 0.03 Leaching 2.3 1.5 1.4 0.9 0.5 Erosion 6 1.8 1.1 0.1 0.03 Total Ecosystem Inventories 313 82 106 125 72 Input (annual) 0.3 0.3 0.3 0.3 0.3 Re-circulation (annual) 1.7 31 20 24 30 Output (annual) 9.2 3.3 2.5 1.0 0.6 Output-Input (annual) 8.9 3.0 2.2 0.7 0.3 Output-Input (6-Year) 15 136 Nutrient contents in the tree compartment gradually increased as the fallow aged, which reflected gradual increases in nutrient uptake by the fallow trees. Portions of the nutrients in the fallow trees are transferred back to the soils when the fallow is slashed and burned. Nutrients that have gaseous oxides are likely to be lost from the system (e.g. N and S). It is unlikely that some of these volatilized nutrients are transferred back to the ecosystem in rainwater because the burning takes place in March, which is the dry season. Proportions of nutrients lost via burning were not studied. Nutrient input (geochemistry) to these ecosystems consists mainly of nutrients contributed by rainfall and soil weathering. In this study, annual rainfall contributed about 0.9, 0.6, 2.7, 0.5, 0.3 kg/ha of N , P, K , Ca, and Mg, respectively. The amounts of N , K , Ca, and M g in the rainwater appear to be trivial when they are compared with the inventories of these nutrients in the soils and the trees. The amount of P in the rainwater, on the other hand, could be significant considering that the soils of the fallow fields and the secondary forests had only about 1-1.5 kg of available P/ha at 0-5 cm depth. The time scale of soil weathering as a part of the geochemical cycle is generally longer than for the other two nutrient pathways (i.e. biogeochemical and biochemical) (Kimmins 1997). Therefore, nutrient contributions via soil weathering may not play a significant role in short-rotation shifting cultivation system. In addition, the soils of the study area appeared to be 137 sufficient in K , Ca, and Mg (which are contributed more by soil weathering than N and P in general), and these nutrients were not particularly crucial for good grain production of the rice. Nutrient re-circulation represents the biogeochemical cycle, which is nutrient cycling within a particular ecosystem. The practice of the Karen farmers of leaving unusable components of the rice crop within the cropped field after harvesting and threshing resulted in 16, 0.6, 23, 2.7, and 1.4 kg/ha of rice plant N , P, K , Ca, and M g remaining within the system. If the rice straw was to be exported out of the system (e.g. for mushroom cultivation), 23 kg of K/ha (about eight times that of the K that is estimated to be contributed in the rainwater) could be lost from the system. If the rice straw was burned, up to 16 kg of N/ha (almost 18 times that of the N that is estimated to be contributed in the rainwater) could be lost from the system. C. odorata re-circulates about 86, 19, 139, 29, and 19 kg/ha of N , P, K , Ca, and Mg, respectively, when it is shaded out by the re-growing trees. This fallow shrub species occupied the harvested field rather rapidly after rice harvesting, and it appeared to be able to absorb large amounts of nutrients. This is highly beneficial because the soil nutrients at the early stage of fallow have a high potential to be lost via soil erosion and leaching, and roots of the re-establishing fallow trees are assumed to occupy only a relatively small soil volume initially. The amounts of nutrients re-circulated by C. odorata in the second year fallow field were greater than the amounts circulated in the litter of the fallow trees, except for Ca. For P and K , the second 138 year re-circulation by C. odorata as it is shaded out is greater than tree litter circulation in any year (Tables 7.2 and 7.3). Nutrients are lost from the shifting cultivation system in harvested material (rice grain and firewood), by leaching and erosion, and in the fire. Nutrient losses via export of the rice grain amounted to 11, 1.8, 2.7, 0.3, and 0.9 kg/ha of N , P, K , Ca, and Mg, respectively. N loss via firewood extraction (0.15 kg/ha/year) in the 4- and 5-year fallows was substantial smaller than that of the rice-grain export. Loss of K in the firewood material (0.23 kg/ha/year in the 4- and 5-year fallows) was also substantially lower than that of the rice-grain export. There was some slight rill and sheet erosion within the shifting cultivation and adjacent fields (Anongrak and Seramethakul 1999) of Mae Hae Tai village. Although the measured soil bulk density, which reflects soil compaction and determines infiltration and erosion susceptibility, varies very slightly among shifting cultivation fields (1.0-1.1 g/cm3) and the forest (0.9 g/cm3), the potential losses of soils nutrients via erosion should be rather low. The erosion losses of N , P, K , Ca, and M g are estimated to be 0.1, 0.7, 11, 55, and 6 kg/ha, respectively, in the cropping period. The potential of nutrient losses would have decreased as the fallow aged due to uptake by vegetation and soil protection by canopy cover. 139 Only loss of N as M V - N in run-off sediment was considered here. The loss of N via run-off sediment was slightly higher in the 2-year fallow than in the cropped field because the concentration of N03_-N was higher in the former field. The loss of N03~-N was also higher in the secondary forest than in the 5-year fallow because of the same reason. The first monsoon rain (late May or early June) after rice planting can cause substantial losses of K , Ca, and M g via run-off sediment because slash burning transferred these nutrients to ash, which is susceptible to be lost from the system by water erosion, and also by wind (Toky and Ramakrishnan 1981). Loss of nutrients by wind erosion was not quantified in this study. Loss of N via leaching is normally in a form of M V - N . It was suspected that leaching loss of N in the shifting cultivation system of Mae Hae Tai village was rather small for three reasons. Firstly, about 90% of soil N occurs in plant-unavailable organic forms (Foth and Ellis 1997). The organic matter concentrations of the soils were rather high (6-8%) in this study, and organic N is rather stable, i.e. recalcitrant, making it less susceptible to be leached. Secondly, the buried-bag experiment suggests net N immobilization in a 1-month period in the rainy season, which indicates that soil microbial activities may have rendered N unavailable for leaching. Lastly, slash burning reduced M V - N concentration of the soils by about 50% (section 3.4.8, Chapter 3), so there is less N available for leaching. The acidic soils and the presence of Fe (and possibly Al) result in occluded forms of P rendering it less susceptible to leaching. Potassium is more susceptible to be lost via leaching than Ca and M g because of its single valency, which results in 140 a weaker binding to clay particles of the soils than those of Ca and Mg. Nutrient losses in soil erosion and leaching, like many site-specific soil processes, vary greatly between different locations and according to different land-use history (see Table 3.27 for the summary of soil erosion studies). Sanchez (1976) reported that rubber-tree covered Ultisol soils in Ivory Coast could lose up to 79, 2.9, 63, 31, and 40 kg/ha of N , P, K , Ca, and Mg, respectively, via leaching. Christanty (1989) studied nutrient cycling in a bamboo-based agroforestry system (Talun-Kebun) in volcanic-derived soils in West Java, Indonesia, and found that leaching appeared to be an unimportant pathway of nutrient loss. She found that the losses of N , P, K , Ca, and Mg, in the first-year cropping field in were 0.3, 0.3, 2.8, 7.7, and 6.1 kg/ha, respectively, and the values for the mature fallow (with bamboo as a dominant species) were 0.04, 0.04, 0.4, 2, 1.2 kg/ha, respectively. However, the question remains about the importance of this loss pathway, and studies of nutrient losses via erosion and leaching in sloping-land shifting cultivation are needed to assess their magnitude in the nutrient balance of the system. The sulfur content of soils and vegetation was not studied in this thesis work. However, it is possible that it can be important in shifting cultivation system in northern Thailand (see section 4.4.8, Chapter 4, for discussion). Sulfur, like N (especially NCV-N), can be lost from the system by burning and leaching. 141 Different pathways of nutrient output play significant roles in budgets of different nutrients. The measured loss of N in rice grain (11 kg/ha) was slightly more than the estimated leaching loss (9.2 kg/ha), but was several fold higher than the estimated erosion loss (0.1 kg/ha). Burning resulted in an estimated loss of about 14 kg of N/ha, which was greatest among N outputs. The loss of P in rice grain (1.8 kg/ha) was greater than that of erosion (0.7 kg/ha) and leaching (0.07 kg/ha). Even though the soil available P (Bray II) was very low, it may be of less concern than N because P is mostly found in organic and occluded forms that are less susceptible to erosion and leaching. The upland rice crop also requires a lesser amount of P than N . For the three cations (K, Ca, and Mg), the losses in erosion and leaching dominate the losses in the rice grain. The soils of the cropped and fallow fields of Mae Hae Tai have higher levels of CEC (24-27 cmofTkg) than those of another study in northern Thailand (20-23 cmofTkg in a 10-year fallow; Andriesse and Schelhaas (1987a)). Consequently, the actual leaching loss of the cations of the Mae Hae Tai soils may be somewhat lower than the levels cited (Toky and Ramakrishnan 1981) in this study. There are a number of implications regarding advantages of forest-fallow shifting cultivation. Firstly, i f the forest-fallow shifting cultivation system was to be converted to fixed-field farming, in which the fallow trees were likely to be completely cleared, a higher rate of N loss in leaching would be expected because deep-rooted fallow trees lessen nutrient leaching to below crop-142 rooting zones. Secondly, continuous farming can cause a building-up of insect pests and diseases. Unless the farmers can afford to use fungicides and insecticides, field burning is one of the feasible options, and frequent burning can have adverse effects on the N budget. N deserves closer attention because the rice crop needs N in relatively high amounts. The only N input included in the balance sheet was from rainwater, whereas there are certainly other pathways of N input (e.g. symbiotic N-fixation). There are no reports on N-fixing plant species in the shifting cultivation system. This input pathway should be studied. The Karen gather wild plants for consumption, and some of the vegetation are legumes. These have not been studied for their N-fixing capacity. Total N was found in the soils in large amounts resulting from a long-time accumulation of organic matter. However, N mineralization appeared to be inadequate for rice production as suggested by the lower rice yield in the unfertilized control plots (than the N treated rice) of the fertilizer trials (N was available in a soluble form), and the buried-bag incubation suggested net N immobilization. It appears that a high proportion of N may be recalcitrant rendering it less available to herbaceous crop plants like upland rice. 7.4 Biogeochemistry of Shifting Cultivation with Different Crop-Fallow Combinations The population of the Karen in the mountainous areas of northern Thailand is steadily increasing. This wil l lead to a decrease in arable land per capita. The arable land-base is also under pressure from an increase in national park areas and pressure from the environmental movement to set 143 areas of forest aside for conservation purpose. In contrast, it had been observed that emigration of the hill-tribe minorities from their villages into the cities for wage work is also on the rise. However, the rural-urban migration was slowed by the 1997 economic crisis in Thailand and other Asian countries. The movement of population between the mountainous areas and the cities is currently un-documented. Therefore, it is rather difficult to predict the present and future dynamics of land-use in this area. In any event, the subsistence forest-fallow shifting cultivation system of the Karen is changing. M y soil data and the data on harvest export of materials suggest that the current one-crop-and-five-years-of-fallow combination is sustainable in terms of soil fertility and rice productivity, albeit at low levels of production. However, it is possible that the overall length of the shifting cultivation cycle wil l continue to be reduced. The farmers will either choose to prolong their cropping period or reduce their fallow period to provide for the growing population. M y interviews (Chapter 7) indicated that the further reduction of the fallow period was the preferred choice of the farmers as they believed that an extended cropping period would result in low rice productivity and more problems with weeds. Based on the biogeochemical budgets of the forest-fallow shifting cultivation presented in Tables 7.1-7.5, I prepared budget sheets for N , P, K , Ca, and M g in which I varied crop and fallow combinations. These combinations were: two crop years and four fallow years, four crop years 144 and two fallow years, and six years of continuous cropping (fixed-field farming). These nutrient budget sheets were prepared to address two main objectives: the first was to assess the relative priority of each particular nutrient pathway for further study; the second was to prepare a preliminary comparison of the biogeochemical implications of a change from shifting cultivation to fixed-field farming. The starting conditions for these tabular analyses were those of the 5-year fallow field. If the cropping period was to be prolonged, it would be expected that the rice yield would decline due to decreasing soil fertility and build-up of soil diseases and weeds as indicated in my second-year rice experiment and fertilizer trial (Chapter 4). There are two management options to maintain rice productivity in the shortening of the fallow period in shifting cultivation. The first one is to use external inputs such as chemical fertilizers (the use of human feces is also an option; see section 7.5 in this chapter). The second method is to control soil diseases and weeds by burning the cropped field and/or using herbicides and insecticides. In this exercise, it was assumed that chemical fertilizer N is used in the three combinations of shifting cultivation system: two crop years and four fallow years, four crop years and two fallow years, or six years of continuous cropping (fixed-field farming). For comparison purposes, I assumed that in the absence of fertilizer application the grain yield would gradually decline by 40, 60, and 80% when the cropping period was increased to be two, four, and six years, 145 respectively based on the poor performance of the second year upland rice. From my fertilizer trial, it was found that 1 kg of added N/ha could increase the grain yield of the rice crop about 100 kg/ha (Table 4.4, Chapter 4). Therefore, the extra amount of N required to maintain the rice grain at 1 t/ha (average yield from this and other studies; Table 4.11, Chapter 4) would be 4, 6, and 8 kg/ha in the crop-and-fallow combinations of two crops and four fallow years, four crops and two fallow years, or six years of continuous cropping, respectively. This calculation ignored the effects of pests and diseases on the ability of the rice to respond to fertilizer addition, and of the effects of fertilizer on the ability of the rice to resist these pathogens. Chemical inputs of the nutrients P, K , Ca, and Mg, were not considered in this tabular evaluation. The rice crop did not appear to respond to added P and K in the fertilizer trials. The rice varieties used may be well adapted to low soil P availability (Garrity et al. 1990), even though the soil P availability was very low, and the very low soil P availability levels given in the soils analysis might indicate that the Bray II method was not the best soil testing method for this particular area (Bomke, A. , Soil Science Department, U B C , per. comm.). The available soil K levels appear to be sufficient, with high extractable concentrations. The levels of Ca and M g also appear to be sufficient (and the concentrations are increased by burning), and the rice crop requires the relatively lower amounts of these two nutrients than of N , P, and K . 146 Burning and use of herbicides and insecticides to control soil diseases and weeds can have a number of consequences. Firstly, burning can reduce N concentrations in the slash by 25% (based on a study in northern Thailand; Andriesse and Schelhaas 1987b), or up to 70% (in the burning of forest floor and bamboo leaves in a study in Indonesia; Christanty 1989). Frequent burning can reduce the vigor of stumps of the fallow trees. Secondly, the use of broadcast herbicides can have a harmful effect on beneficial weeds such as C. odorata (a good biomass and nutrient accumulator, Chapter 6) and some edible and medicinal weeds. The nutrient pathways of the one-crop-and-five-fallow years (the current system) are shown in Table 7.6, which is simply a summation of Tables 7.1-7.5. The budget sheets for the alternate crop/fallow combinations are presented in Tables 7.7-7.9. A comparison of nutrient pathways (input, re-circulation, and output) of the different theoretical systems is presented in Table 7.10. As expected, prolonging the cropping period and shortening the fallow period results in increasing nutrient output and decreasing nutrient re-circulation. Nutrients are continuously taken out of the system via grain harvest when the cropping period was extended. The second-year rice experiment suggested a decline in rice productivity in consecutive-year cropping due to soil disease and nematodes, but this was not taken into account in this tabular analysis as it was assumed that insects and pathogens would be controlled by insecticides and other biocides. 147 Table 7.6 Nutrient pathways in a 6-year rotation of one cropping year and five years of fallow. Component Total N Available Extractable Extractable Extractable (kg/ha) P (kg/ha) K (kg/ha) Ca (kg/ha) M g (kg/ha) Ecosvstem Inventories* 6,476 11 374 388 125 Cropping Period Input Precipitation 0.9 0.6 2.7 0.5 0.3 Re-circulation Upland Rice 16 0.6 23 2.7 1.4 Fine Roots 1.3 0.1 0.9 1.1 0.3 Total 17 0.7 24 3.8 1.7 Output Upland Rice 11 1.8 2.7 0.3 0.9 Leaching 9.2 0.07 14 4.6 2.3 Erosion 0.1 0.7 11 55 6.0 Burning 14 - - - -Total 34 2.6 28 60 9.2 Fallow Period Input Precipitation 4.5 3.0 14 2.5 1.5 Re-circulation C. odorata 86 19 139 29 19 Litter 199 18 138 400 51 Fine Roots 23 1.8 16 19 5.5 Total 308 39 293 448 76 Output Firewood 0.3 0.1 0.5 0.2 0.1 Leaching 10 0.1 15 8.7 3.8 Erosion 0.2 0.1 5 15 3.0 Total 11 0.3 20 24 6.9 Total Ecosystem Inventories 6,476 11 374 388 125 Input 5.4 3.6 17 3.0 1.8 Re-circulation 325 40 317 452 78 Output 45 2.9 48 84 16 Output-Input (6-Year) 40 -0.7 31 81 14 * Starting inventories at beginning of the rotation. 148 Table 7.7 Nutrient pathways in a 6-year rotation of two cropping years and four years of fallow Component Total N Available Extractable Extractable Extractable (kg/ha) P (kg/ha) K (kg/ha) Ca (kg/ha) M g (kg/ha) Ecosystem Inventories* 6,476 11 374 388 125 Cropping Period Input Precipitation 1.8 1.2 5.4 1.0 0.6 Fertilizer 4.0 - - - -Re-circulation Upland Rice 32 1.2 46 5.4 2.8 Fine Roots 2.6 0.2 1.8 2.2 0.6 Total 35 1.4 48 7.6 3.4 Output Upland Rice 22 3.6 5.4 0.6 1.8 Leaching 18 0.14 28 9.2 4.6 Erosion 0.2 1.4 22 110 12 Burning 28 - - - -Total 68 5.1 55 120 18 Fallow Period Input Precipitation 3.6 2.4 11 2.0 1.2 Re-circulation C. odorata 86 19 139 29 19 Litter 113 9.9 79 229 30 Fine Roots 12 0.9 7.8 10 2.8 Total 211 30 226 268 52 Output Firewood 0.2 0.03 0.23 0.08 0.03 Leaching 9.2 0.07 14 6 2.9 Erosion 0.2 0.13 4.8 15 2.9 Total 9.6 0.23 19 21 5.8 Total Ecosystem Inventories 6,476 11 374 388 125 Input 9.4 3.6 16 3.0 1.8 Re-circulation 246 31 274 276 55 Output 78 5.3 74 141 24 Output-Input (6-Year) 69 1.7 58 138 22 * Starting inventories at beginning of the rotation. 149 Table 7.8 Nutrient pathways in a 6-year rotation of four cropping years and two years of fallow Component Total N Available Extractable Extractable Extractable (kg/ha) P (kg/ha) K (kg/ha) Ca (kg/ha) M g (kg/ha) Ecosystem Inventories* 6,476 11 374 388 125 Cropping Period Input Precipitation 3.6 2.4 11 2.0 1.2 Fertilizer 6.0 - - - -Re-circulation Upland Rice 64 2.4 92 11 5.6 Fine Roots 5 0.4 3.6 4.4 1.2 Total 69 2.8 96 1.5 6.8 Output Upland Rice 44 7.2 11 1.2 3.6 Leaching 37 0.3 56 18 9.2 Erosion 0.4 2.8 44 220 24 Total 81 10 111 240 37 Fallow Period Input Precipitation 1.8 1.2 5.4 1.0 0.6 Re-circulation C. odorata 86 19 139 29 19 Litter 46 4.0 32 93 12 Fine Roots 2.6 0.2 1.7 2.2 0.6 Total 135 23 173 124 32 Output Leaching 6.1 0.1 9.3 3.1 1.5 Erosion 0.2 0.1 2.8 8.0 1.8 Total 6.3 0.2 12 11 3.3 Total Ecosystem Inventories 6,476 11 374 388 125 Input 11 3.6 16 3.0 1.8 Re-circulation 204 26 269 139 39 Output 87 10 123 251 40 Output-Input (6-Year) 76 6.4 107 248 38 * Starting inventories at beginning of the rotation. 150 Table 7.9 Nutrient pathways in a 6-year rotation of continuous cropping Component Total N Available Extractable Extractable Extractable (kg/ha) P (kg/ha) K (kg/ha) Ca (kg/ha) M g (kg/ha) Ecosystem Inventories* 6,476 11 374 388 125 Cropping Period Input Precipitation 5.4 3.6 16 3.0 1.8 Fertilizer 8.0 - - - -Re-circulation Upland Rice 96 3.6 138 16 8.4 Total 96 3.6 138 16 8.4 Output Upland Rice 66 11 16 1.8 5.4 Firewood - - - - -Leaching 55 0.4 84 28 14 Erosion 0.6 4.2 66 330 36 Total 122 16 166 360 55 Total Ecosystem Inventories 6,476 11 374 388 125 Input 13 3.6 16 3.0 1.8 Re-circulation 96 3.6 138 16 8.4 Output 122 16 166 360 55 Output-Input (6-Year) 109 12 150 357 53 * Starting inventories at beginning of the rotation. Nutrient output via soil erosion and leaching is substantially increased when the cropping period is lengthened because the shorter fallow period causes a reduction in the role of the fallow vegetation in minimizing erosion and leaching. The roles of the fallow vegetation in biogeochemical pathways, e.g. litterfall and fine-root turnover, are also reduced when the fallow period is further shortened. Reduced fallow period results in more frequent clearing of fallow 151 vegetation, which might lead to a loss of the vigor of the fallow vegetation. Table 7.10 Comparison of nutrient pathways among different crop-fallow combinations assuming a 6-year cycle Component Total N Available P Extractable Extractable Extractable (kg/ha) (kg/ha) K (kg/ha) Ca (kg/ha) M g (kg/ha) Ecosystem 6,476 11 374 388 125 Inventories Input 1 crop + 5 fallow 5.4 3.6 17 3.0 1.8 2 crop + 4 fallow 9.4 3.6 16 3.0 1.8 4 crop + 2 fallow 11 3.6 16 3.0 1.8 6 crop 13 3.6 16 3.0 1.8 Re- circulation 1 crop + 5 fallow 325 40 317 452 78 2 crop + 4 fallow 246 31 274 276 55 4 crop + 2 fallow 204 26 269 139 39 6 crop 96 3.6 138 16 8.4 Output 1 crop + 5 fallow 45 2.9 48 84 16 2 crop + 4 fallow 78 5.3 74 141 24 4 crop + 2 fallow 87 10 123 251 40 6 crop 122 16 166 360 55 Output-Input 1 crop + 5 fallow 40 -0.7 31 81 14 2 crop + 4 fallow 69 1.7 58 138 22 4 crop + 2 fallow 76 6.4 107 248 38 6 crop 109 12 150 357 53 152 If shifting cultivation were to be converted to fixed-field farming, there would be no nutrient re-circulation by C. odorata, and by litter and fine roots of the fallow vegetation. Using chemical fertilizers and fallow enrichment (e.g. planting N-fixing species) are methods of accelerated nutrient replenishment. The use of fertilizers is too expensive for most of the farmers (as was expressed in their interviews), and it would very likely lead to environmental hazards such as downstream water contamination. The second option, involving the use of N-fixing herbs, shrubs, or trees, is still in need of some fundamental research on benefits of the N-fixing species and their management in this particular area of northern Thailand. 7.5 Recycling of Human Waste in the Forest-Fallow Shifting Cultivation Nutrient losses out of the shifting cultivation system via rice grain could potentially be returned to the system by recycling of human waste. Arnaud and Sanchez (1990) reported that 30-50% of Ca, and 70-80% of P are absorbed during digestion; 50-70%o of M g is absorbed from foods (Shils 1990). Alfin-Slater and Mirenda (1980) indicated that the human body can utilize about 70% of N in mixed dietary protein. The remaining nutrients are not absorbed, and a portion of the absorbed nutrients is subsequently excreted. Thus, human feces contain considerable quantities of nutrients. The use of human feces as organic fertilizers has been long practiced by people in many parts of the world (e.g. China). However, this has not been a traditional practice in the shifting cultivation in northern Thailand because the cultivated fields are rather far from the village sites, e.g. 1-3 km in case of Mae Hae Tai village. Other reasons may be cultural, religious, or social. 153 In 1999, the quantities of nutrients exported in the harvested grain were 11, 1.8, 2.7, 0.3, and 0.9 kg/ha of N , P, K , Ca, and Mg, respectively. If human feces were to be recycled back to the shifting cultivation fields, the amounts of nutrients would be 3.3 kg of N/ha (30% excreted), 0.45 kg of P/ha (25% excreted), 0.18 kg of Ca/ha (60% excreted), and 0.36 kg of Mg/ha (40% excreted). A comparable value was not available for K , but it may be insignificant because only about 10% of K in the aboveground biomass of the rice crop was exported via the rice grain. The potential for nutrient conservation by recycling of human waste could have important implications for the nutrient balance of the system. This is especially true for N because it appears to be the most limiting nutrient and the one most likely to be lost out of the system via burning and leaching. 7.6 Thesis Conclusions 7.6.1 Thesis Findings The main objective of the research project of which this thesis is a part is to examine the dynamics of nutrients in the reduced forest-fallow shifting cultivation cycle. The thesis work is divided into three major parts. 154 1) Some biophysical aspects of the forest-fallow shifting cultivation system of the Karen of Mae Hae Tai village (Chapter 3-6) This work attempted to understand some of the major nutrient flows within this system of shifting cultivation, so it was essential to describe the characteristics of soils and vegetation at different times in the cropping/fallow cycle (Table 7.11). There were no previous empirical data on these aspects of this type of shifting cultivation in this area, and, therefore, an initial gathering of basic descriptive data was required. 1.1 Soils Soils of shifting cultivation fields of Mae Hae Tai village were mainly clayey with more than 35% of clay content in the surface and higher levels at greater depths. The bulk density of the soils (0-5 cm) ranged from 0.9 g/cm3 in the secondary forest to 1.04 g/cm3 in the cropped field. The soils were medium to slightly acid (pH 5-6 in water). Soil organic matter was highest in the secondary forest (7.9%), and was 6.8% and 5.8-6.7% in the cropped and the fallow fields, respectively. Total N concentration of soils ranged from 0.26% in the fallow fields to 0.37% in the secondary forest. Available P, and extractable K , Ca, and M g were highest in the cropped field, and were inversely correlated with soil Fe content. 155 Table 7.11 Nutrient contents (kg/ha) in soils, upland rice, shrub, trees, litter, and rainfall Ecosystem Mass Total Avail. P Extractable Extractable Extractable Compartment (kg/ha) N K Ca M g Soil (0-5 cm) Rice field (1998) - 1,602 11 186 954 105 (after harvest) Upland Rice 1998* Grain 519 7.3 1.6 2.1 0.5 0.4 Leaf 330 4.3 0.3 3.3 0.7 0.3 Stem 1,528 11 4.6 18 1.5 1.5 1999 Grain 902 11 1.8 2.7 0.3 0.9 Leaf + Stem 1,225 15 0.5 21 2.5 1.2 Root 1,061 6.6 0.5 11 0.1 0.6 C. odorata (1999) Leaf 1,500 38 6.0 27 11 11 Stem 5,643 34 11 90 15 5.6 Root 1,553 14 1.6 22 0.3 1.6 Trees** 5-year Fallow (1998) 17,200 69 6.9 67 46 13 Secondary Forest (1998) 31,625 127 13 123 84 23 Litter 4-year Fallow (1998) TotalA 5,902 67 5.9 47 136 18 Leaf 4,899 32 2.4 22 26 7.8 Secondary Forest (1998) TotalA 9,119 104 9.1 73 210 27 Leaf 6,019 39 3.0 27 33 9.6 Rainfall Rice field (1998) - 0.9 0.6 2.7 0.5 0.3 *Note that rice grain yield was below normal in 1998 because of drought and aphid attack. **Aboveground component, excluding leaves. A Aboveground litter. 156 1.2 Upland rice Upland rice produces about 1 t/ha of grain, and there is a wide range of yields between years (due to variation in climate, diseases, and insects), between fields (differences in soils), and within fields (soil and topographic variation, and nutrient redistribution due to past farming practices). A l l aboveground rice components except the grain, and all belowground biomass, are left to decompose in the harvested fields, which become fallowed after one year of cropping. About 40% of aboveground rice biomass in 1999 was exported in the form of grain, which accounted for 40, 75, 11, 9, and 40% of the total N , P, K , Ca, and Mg, in the aboveground upland rice biomass, respectively. The amount of harvested grain was lower in 1998 due to drought and aphid infestation. The quantities of grain harvested according to the farmers' interviews were higher than in the experimental plots, possibly due to variability in soil fertility and rice varieties. 1.3 Trees The role of trees as perceived by the farmers is to provide ash fertilizer for upland rice production without external nutrient inputs, and to sustain good soil structure. Fallow trees also contribute a large amount of litterfall (about 6 t/ha annually in the 4-year fallow field in comparison to 9 t/ha in the secondary forest, compared with about 1.2 t/ha of aboveground upland rice biomass left in the field after harvest). This litterfall is incorporated into soils via decomposition processes or is burned following slashing of the 5-year fallow. 157 1.4 Shrub Shrubs (and herbs) are considered weeds only during the rice-growing period because at that time they interfere with rice growth. However, they are not considered as weeds during the fallow period because some of them are useful as food or medicines. C. odorata is the dominant shrub species during the cropping period and the first two years of fallow. It can provide a large aboveground biomass of approximately 7 t/ha one year after rice harvest. 2) Second-year rice experiment and fertilizer trials (Chapter 4) This hypothetic-deductive component of the thesis addressed two main hypotheses: 2.1 If nutrients are critical in maintaining rice productivity, the rice crop should respond positively to additional nutrient inputs as chemical fertilizers, and negatively to a second year of cropping without fertilizers. This was tested by fertilizer trials in the regular first-year rice cropping and the experimental second-year rice cropping. The second-year rice was six weeks younger than the first-year rice because it was damaged by cows, and was replanted. Therefore, a direct comparison between the two rice crops could not be made. The second-year rice was also attacked by soil pathogens resulting in root and lower stem damage and subsequently much lower total biomass than the regular first-year rice and almost no grain. However, the fertilizer trials both in the first-year and second-year rice showed that rice 158 growth was improved by adding fertilizers. Therefore, soil nutrients are clearly a critical part of maintaining and improving rice productivity. 2.2 If the fallow period plays a key role in maintaining rice productivity by replenishing soil fertility through the accumulation of organic matter and nutrients by the fallow vegetation, and i f nutrients and soil fertility are key determinants, a reduction in the fallow period should result in declining rice productivity. A pot experiment using soils from young fallow-fields (2- and 4-year fallow) was conducted to test the potential of these soils to support normal rice yields in comparison to soils from a 5-year fallow of Mae Hae Tai village. A fertilizer trial was also included in this experiment to test i f any difference in the rice growth potential of the soils from the different fallow ages could be attributed to nutritional factors. The results from the pot experiment were inconclusive due to a pot drainage problem, and could not answer the research question on the potential of young-fallow soils in maintaining rice production. The pot experiment is reported in Appendix 7. The soil data do not indicate any directional changes in the levels of soil nutrients during the fallow period; therefore, it is inconclusive whether fallow reduction would result in a decline in soil fertility. The anecdotal results of these studies support the hypothesis that one of the major roles of the fallow is control of insect pests and diseases of the crop species. 159 3) Farmers' informal-interviews for their knowledge about shifting cultivation management (Appendix 10) This part of the thesis was an attempt to assess the farmers' knowledge in relation to soil and vegetation aspects of the system, and to consider the knowledge in an ecological context. This aspect of the thesis was by no means comprehensive and was not a formal survey, because the major portion of the thesis research was focused on the biophysical aspects of the topic. However, this subproject contributed to the inductive phase of the project in which hypotheses and questions were formulated for testing based on the existing knowledge of the topic and the study area, and the selection of fields for the chronosequence was based on the farmers' opinions. Most of the farmers had a similar level of knowledge of the shifting cultivation system. There was general agreement that: 1) the main reasons for growing rice for only one year are increasing weeds (leading to a higher labor demand for weeding) and lower rice productivity; 2) C. odorata is regarded as beneficial for rice production; and 3) ash fertilizer is necessary for rice production. However, some young farmers (20-30 years old) only know about the "what" of their farming practices, and not about the "why" in comparison with older farmers. 160 7.6.2 Validity of the Alternative Hypotheses The nutritional role (i.e. soil fertility) of fallow was the main focus of this thesis, but this was only one part of a bigger picture. Alternative hypotheses of the other roles of the fallow are listed in section 7.2. 1) Maintaining good soil structure This hypothesis was not directly tested in this study. However, the similar values of soil bulk density of the soils of the shifting cultivation fields and the secondary forest suggest that the fallow maintains good soil structure, with no evidence of change in bulk density over the management cycle in the current system. 2) Providing usable plants and animals for consumption and medicinal purposes The farmers cited many plant species as edible or usable as medicines or remedial substances during the data-gathering period. These plants can be found both in the forests and in the fallow. Other studies also report this aspect of the fallow (i.e. Kunstadter 1978; Rerkasem and Rerkasem 1994). 3) Control of pests and diseases of crop species The second-year experimental rice was attacked by soil pathogens resulting in root and lower stem damage and subsequently much lower total biomass than the regular first-year rice and 161 almost no grain. This supports an important role of the fallow as an insect/disease break (i.e. Hansen 1995). 7.6.3 Data Gaps In Thailand, there has been a continuing effort to change shifting cultivation to fixed-field farming, and this trend is very likely to continue into the foreseeable future. Therefore, amongst other things such as soil structure, pests and diseases, nutrient balances of these two farming systems should be compared to assess their advantages and disadvantages, so that land-use and policy changes can be made accordingly. The biogeochemical budget approach is potentially useful for the assessment of the nutrient balance of an ecosystem by identifying the major nutrient transfer pathways and by identifying data gaps that require further research. Loss of N from the shifting cultivation system is dominated by rice harvest and burning, while loss of P is dominated by rice harvest. Leaching causes the largest loss of K , while soil erosion dominates losses of Ca, and M g from the system. The simple, static, tabular analysis used in this thesis identified several key processes that need to be quantified (e.g. N fixation, leaching, and erosion) in order to fully describe the biogeochemistry of this agricultural system, and to answer questions about the biogeochemical aspects of the sustainability of the current forest-fallow shifting cultivation and alternative farming systems. 7.6.4 Future Research Aspects of this land use system which require further investigations include: 162 1) What are the total losses of nutrients (e.g. via burning, leaching, and erosion) and total input (e.g. via N fixation, weathering, and etc.), especially for N . Nitrogen was revealed as the most limiting nutrient, but the mechanisms of nitrogen input and loss are not well known. If chemical fertilizers are added to maintain crop productivity, losses of N via leaching and denitrification will also be of particular concern. 2) A series of carefully controlled and replicated field and pot experiments is needed to resolve the relative importance of the different contributions of fallow (i.e. fallow as a control of weeds, insects, and diseases, and of maintaining soil fertility, soil physical properties, and other forest values). There is also a need to determine interactions of fertilizers, insecticides, fungicides, and herbicides on crop yields i f these chemicals are to be used to maintain upland rice productivity in forest-fallow shifting cultivation with a shorter fallow period and continuous cropping. 3) Accurate biomass determination and nutrient analyses of fallow trees. To evaluate the nutritional role of the fallow vegetation more accurately, better data are needed for biomass and nutrient contents. 163 4) Changing patterns of resource allocation between above- and belowground in the shifting cultivation system. This study focused only on aboveground process, yet it is known that much of the production of vegetation under conditions of nutrient limitation is expected in belowground biomass. The turnover of fine roots following the slashing of fallow may be a major source of nutrient for the rice, but there is no knowledge about this. 5) The effect of removing the readily available carbon source in the slash by fire on the immobilization of soil nitrogen by soil microbes requires investigation. As answers to these questions and data on these processes become available, there wil l be a need to synthesize the accumulated knowledge into a more dynamic model than the tabular model that was analyzed in this thesis. The original goal of using a dynamic ecosystem management model was frustrated by lack of data, but such models should be used as the vehicle by which to synthesize the products of the research suggested above. There should be an interplay between the model that is to be used for this purpose and the design of data gathering and future experiment. Empirical data collection, experimentation and modeling should be intricately linked activities when studying systems such as the swidden agricultural system of these Thai ecosystem. 164 References Ahn, P .M. 1974. Some observation on basic and applied research in shifting cultivation. F A O Soils Bulletin 24:123-154. Alegre, J . C , and Cassel, D.K. 1986. Effect of land-clearing methods and postclearing management on aggregate stability and organic carbon content of a soil in the humid Tropics. Soil Science 142:289-295. Alfm-Slater, R.B., and Mirenda, R. 1980. Nutrient requirement: what they are and bases for recommendations. In Nutrition and the Adults: Macronutrients, ed. Alfin-Slater, R.B., and Kritchevsky, D., 1-48. New York: Plenum Press. Allaby, M . 1994. The Concise Oxford Dictionary of Ecology. Oxford: Oxford University Press. Anderson, J.M., and Ingram, J.S.I. 1993. Tropical Soil Biology and Fertility: A Handbook of Methods, 2d ed. Wallingford: C A B International. Andriesse, J.P. 1980. Nutrient level changes during a 20-year shifting cultivation cycles in Sarawak (Malaysia). In Proceedings of Conference on Classification and Management of Tropical Soils 1977, edited by Joseph, K.T., 479-491. Kuala Lumpur: Malaysia Society of Soil Science. . 1987. Monitoring Project of Nutrient Cycling in Soils Used for Shifting Cultivation under Various Climatic Conditions in Asia. Part I Text. Amsterdam: Royal Tropical Institute. Andriesse, J.P., and Schelhaas, R . M . 1987a. A monitoring study on nutrient cycles in soils used for shifting cultivation under various climatic conditions in tropical Asia. II Nutrient stores in biomass and soil-results of baseline studies. Agriculture, Ecosystems, and Environment 19: 285-310. 1987b. A monitoring study on nutrient cycles in soils used for shifting cultivation under various climatic conditions in tropical Asia. I l l The effects of land clearing through burning on fertility level. Agriculture, Ecosystems, and Environment 19:311-332. Anongrak, N . , and Seramethakul, D. 1999. Soil Survey of Mae Hae Tai village. Chiang Mai: Chiang Mai University. Photocopied. A P H A , A W W A , and WPCF. 1985. Standard Methods for Examination of Water and Wastewater, 16d ed. Washington, DC: American Public Health Association, American Wastewater Works Association, and Water Pollution Control Federation. Arbhabhirama, A. , Phantumvanit, D., Elkington, J., and Ingkasuwan, P. 1988. Thailand Natural Resources Profile. Singapore: Oxford University Press. 165 Arnaud, C D . , and Sanchez, S.D. 1990. Calcium and phosphorus. In Present Knowledge in Nutrition, 6d ed. Brown, M.L . , 212-223. Washington, DC: International Life Sciences Institute-Nutrition Foundation. Aung, T. and Prot, J.C. 1990. Effects of crop rotations on Pratylenchus zeae and on yield of rice Cultivar U P L Ri-5. Revue de Nematologie 13:445-447. Ayanaba, A. , Tuckwell, S.B., and Jenkinson, D.S. 1976. The effects of clearing and cropping on the organic reserves and biomass of tropical forest soils. Soil Biology and Biochemistry 8:519-525. Aweto, A.O. 1981a. Secondary succession and soil fertility restoration in south-western Nigeria. I Succession. Journal of Ecology 69:601-607. . 1981b. Secondary succession and soil fertility restoration in south-western Nigeria. II Soil fertility restoration. Journal of Ecology 69:609-614. Bass, S., and Morrison, E. 1994. Shifting Cultivation in Thailand, Laos, and Vietnam: Regional Overview and Policy Recommendations. London: Forestry and Land Use Programme, International Institute for Environment and Development. Beaton, J.D., and White, M . 1997. Occurrence and correction of S deficiencies in the Asian and Pacific region: A review and update. Sulfur in Agriculture 20:31-46. Beinroth, F.H. 1972. The General Pattern of the Soils of Puerto Rico. Transactions of the Fifth Caribbean Geological Conference. Geological Bulletin 5:225-229. Blair, G., and Lefroy, R. 1998. Sulfur and carbon research in rice production systems. Field Crops Research 56:177-181. Boonchee, S., and Anecksamphant, C. 1993. Sustaining soil organic matter for upland rice production in northern Thailand. In Soil Organic Matter Dynamics and Sustainability of Tropical Agriculture, ed. Mulongoy, K. , and Merckx, R., 155-161. Chichester: John Wiley. Bray, R.H., and Kurtz, L.T. 1945. Determination of total, organic, and available forms of phosphorus in soils. Soil Science 59:39-45. Brown, S. 1997. Estimating Biomass and Biomass Change of Tropical Forests, a Primer. Rome: F A O . Brubacher, D., Arnason, J.T., and Lambert, J.D.H. 1989. Woody species and nutrient accumulation during the fallow period of milpa farming in Belize, C.A. Plant and Soil 114:165-172. 166 CARE-Thailand. 1997. Community forest management plan of Mae Hae Tai village. Chiang Mai: Mae Hae Tai Community Organization, Pang Hin Fon Watershed Management Unit (Royal Forestry Department), and CARE-Thailand. Photocopied. . 1999. A survey of changes of evergreen forests, Mae Hae Tai village. Chiang Mai: Mae Hae Tai Community Organization, Faculty of Forestry, Kasetsart University, and C A R E -Thailand. Photocopied. Chanakorn, O. 1969. The status of sulfur in important upland soils in Thailand and methods of evaluating it. M.S. Thesis, Kasetsart University. Chang, T., and Vergara, B.S. 1975. Variety diversity and morpho-agronomic characteristics of upland rice. In Major Research in Upland Rice, ed. International Rice Research Institute, 71-90. Manila: International Rice Research Institute. Christanty, L . 1986. Shifting cultivation and tropical soils: Patterns, problems, and possible Improvement. In Traditional Agriculture in Southeast Asia, ed. Marten, G.G., 226-240. Boulder: Westview Press. . 1989. Analysis of the sustainability and management of the Talun-Kebun system of West Java, Indonesia. Ph.D. Thesis, The University of British Columbia. Christanty, L. , Mailly, D., and Kimmins, J.P. 1996. "Without bamboo, the land dies": Biomass, litterfall, and soil organic matter dynamics of a Javanese bamboo talun-kebun system. Forest Ecology and Management 87:75-88. Christanty, L. , Kimmins, J.P., and Mailly, D. 1997. "Without bamboo, the land dies": A conceptual model of the biogeochemical role of bamboo in an Indonesian agroforestry. Forest Ecology and Management 91:83-91. Conklin, H.C. 1961. A study of shifting cultivation. Current Anthropology 2: 27-61. Credner, W. 1935a. Siam: das Land der Tai. Stuttgart. . 1935b. Volkerschichtung und Volkerbewegungen im mitten Hinterindien. In Sven-Hedin-Festschrift. Geografiska Annaler, 269-283. Stockholm. Doran, J.W., and Parkin, T.B. 1994. Defining and assessing soil quality. In Defining Soil Quality for a Sustainable Environment, ed. Doran, J.W., Coleman, D.C., Bezdicek, D.F., and Stewart, B .A. , 3-21. Madison: Soil Science Society of America. Dove, M.R. 1985. Swidden Agriculture in Indonesia: The Subsistence Strategies of the Kalimantan Kantu. Berlin: Mounton. 167 Eno, C.F. 1960. Nitrate production in the field by incubation the soil in polyethylene bags. Soil Science Society of America Proceedings 24:277-279. Ewel, J., Berish, C , Brown, B., Price, N . , and Raich, J. 1981. Slash and burn impacts on a Costa Rican wet forest site. Ecology 62:816-829. Ewel, J.L. 1986. Designing agricultural ecosystems for the humid tropics. Annual Review of Ecology and Systematics 17:245-271. FAO. 1989. Sustainable Agricultural Production: Implication for International Agricultural Research. Rome: Technical Advisory Committee, CGIAR, FAO. FAO-UNESCO. 1971. Soil Map of the World. Vol . IV: South America. Paris: United Nations Educational, Scientific, and Cultural Organization. Forsyth, T. 1996. Science, myth, and knowledge: Testing Himalayan environmental degradation in Thailand. Geoforum 27:375-392. . 1999. Questioning the impacts of shifting cultivation. Watershed 5:23-29. Foth, H.D., and Ellis, B .G. 1997. Soil Fertility, 2d ed. Boca Raton: CRC Press. Fujisaka, S., Hurtado, L. , and Uribe, R. 1996. A working classification of slash-and-burn agricultural systems. Agroforestry Systems 34:151-169. Funakawa, S., Tanaka, S., Kaewkhongkha, T., Hattori, T., and Yonebayashi, K . 1997a. Physicochemical properties of the soils associated with shifting cultivation in Northern Thailand with special references to factors determining soil fertility. Soil Science and Plant Nutrition 43:665-679. Funakawa, S., Tanaka, S., Shinjyo, H. , Kaewkhongkha, T., Hattori, T., and Yonebayashi, K . 1997b. Ecological study on the dynamics of soil organic matter and its related properties in shifting cultivation systems of Northern Thailand. Soil Science and Plant Nutrition 43:681-693. Ganjanapan, A . 1998. The politics of conservation and the complexity of local control of forests in the northern Thai highlands. Mountain Research and Development 18:71-82. Ganjanapan, S. 1997. A comparative study of indigenous and scientific concepts in land and forest classification in northern Thailand. In Seeing Forests for Trees: Environment and Environmentalism in Thailand, ed. Hirsch, P., 247-267. Chiang Mai: Silkworm Books. 168 Garrity, D.P., Mamaril, C.P., Soepardi, G. 1990. Phosphorus requirements in upland rice-based cropping systems. In Phosphorus requirements for sustainable agriculture in Asia and Oceania, ed. International Rice Research Institute, 333-347. Manila: International Rice Research Institute. Gee, G.W., and Bauder, J.W. 1986. Particle-size analysis. In Methods of Soil Analysis. Part 1 Physical and Mineralogical Methods. 2d ed. Page, A . L . , 383-411. Madison: Soil Science Society of America. Hagstrom, G.R. 1986. Fertilizer sources of sulfur and their use. In Sulfur in Agriculture, ed. Tabatabai, M.A . , 567-581. Madison: American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. Hairiah, K. , Kasniari, D.N., van Noordwijk, M . , de Foresta, H. , and Syekhfani. 1996. Litterfall, above- and belowground biomass and soil properties during the first year of a Chromolaena odorata fallow. Agrivita 19:184-192. Hairiah, K . , van Noordwijk, M . , and Palm, C. 1999. Methods for sampling above- and below-ground organic pools. In Modelling Global Change Impacts on the Soil Environment, ed. Murdiyarso, D., van Noordwijk, M . , and Suyamto, D.A., 46-77. Bogor: BIOTROP-GCTE/Impacts for Southeast Asia (IC-SEA). Hansen, P.K. 1995. Shifting cultivation adaptations and environment in a mountainous watershed in Northern Thailand. Ph.D. Thesis, Copenhagen: Royal Veterinary and Agricultural University. Hart, S.C., Stark, J .M., Davidson, E.A., and Firestone, M . K . 1994. Nitrogen mineralization, immobilization, and nitrification. In Methods of Soil Analysis. Part 2 Microbiological and Biochemical Properties, ed. Soil Science Society of America, 985-1018. Madison: Soil Science Society of America. Hiranburana, N . 1996. Assessing and solving nutrient deficiencies in highlands. In Highland farming: Soil and the future?: Proceedings held in Chiang Mai 21-22 December 1995, edited by Turkelboom, F., van Look-Rothchild, K , and van Keer, K. , 97. Chiang Mai: Soil Fertility Conservation Project, Maejo University and Catholic University of Leuven. Holscher, D., Moller, R.F., Denich, M . , and Folster, H . 1997. Nutrient input-output budget of shifting agriculture in eastern Amazonia. Nutrient Cycling in Agroecosystems 47:49-57. Hoult, E.H. , Tonkla, K. , and Singhatat, V . 1977. Field Agronomy and Soils Research Programmes Results 2520, Thai-Australia Land Development Project. Canberra: The Australian Development Assistance Bureau. 169 . 1978. Field Agronomy and Soils Research Results 2521, Part 2 Soil Fertility Research, Thai-Australia Land Development Project. Canberra: The Australian Development Assistance Bureau. Hoult, E.H., Andrews, A. , and Keerakit-Kasikorn, P. 1983. Sulfur deficiencies in the agriculture of Thailand, Kampuchea, Laos, and Vietnam. In Sulfur in Southeast Asia and South Pacific Agriculture, ed. Blair, G.J., and Til l , A.R., 115-146. Armidale: The University of New England. Hulugalle, N.R., Lai, R., and Ter Kuile, C.H.H. 1984. Soil physical changes and crop root growth following different methods of land clearing in western Nigeria. Soil Science 138:172-179. Hurni, H . 1983. Soil erosion and soil formation in agricultural esosystems: Ethiopia and northern Thailand. Mountain Research and Development 3:131-142. IRRI. 1976. 1975 Annual Report. Manila: International Rice Research Institute. . 1977. 1976 Annual Report. Manila: International Rice Research Institute. . 1978. 1977 Annual Report. Manila: International Rice Research Institute. Izikowitz, K . G . 1951. Lamet-hill Peasants in French Indochina. Gothenburg: Elanders Boktryckeri Aktiebolag. Jackson, M.I. 1958. The Soil Chemical Analysis London: Constable. Kayama, T. 1978. Pulping and papermaking properties and wood properties of tropical hardwoods. In IUFRO Conference on Wood Quality and the Utilization of Tropical Species, ed. IUFRO, 197-209. Laguna: International Union of Forestry Research Organizations. Keen, F.G.B. 1972. Upland Tenure and Land Use in Northern Thailand. Bangkok: SEATO. Ketterings, Q.M., Wibowo, T.T., van Noordwijk, M . , and Penot, E. 1999. Farmers' perspectives on slash-and-burn rubber producers in Sepunggur, Jambi Province, Sumatra, Indonesia. Forest Ecology and Management 120:157-169. Kimmins, J.P. 1974. Sustained yield, timber mining, and the concept of ecological rotation; a British Columbia view. Forestry Chronicle 50:27-31. . 1997. Forest Ecology: A Foundation for Sustainable Development. 2d ed. New Jersey: Prentice-Hall. 170 Kleinman, P.J.A., Pimentel, D., and Bryant, R.B. 1995. The ecological sustainability of slash-and-burn agriculture. Agriculture, Ecosystems, and Environment 52:235-249. Kotto-Same, J., Woomer, P.L., Appolinaire, M . , and Louis, Z. 1997. Carbon dynamics in slash-and-burn agriculture and land use alternatives of the humid forest zone in Cameroon. Agriculture, Ecosystems and Environment 65:245-256. Kunstadter, P. 1990. Impacts of economic development and population change on Thailand's forests. In Tropical Resources: Ecology and Development, ed. Furtado, J.I., Morgan, W.B., Pfaffin, J.R., and Ruddle, K. , 171-190. New York: Harwood Academic Publishers. Kunstadter, P., Chapman, E.C., and Sabhasri, S., ed. 1978. Farmers in the Forest: Economic Development and Marginal Agriculture in Northern Thailand. Honolulu: The University Press of Hawaii. Kyuma, K. , and Pairintra, C , ed. 1983. Shifting Cultivation: A n experiment at Nam Phrom, Northeast Thailand, and Its Implications for Upland Farming in the Monsoon Tropics. Kyoto: Faculty of Agriculture, Kyoto University. Photocopied. Kyuma, K. , Tulaphitak, T., and Pairintra, C. 1985. Changes in soil fertility and tilth under shifting cultivation. I General description of soil and effect of burning on the soil characteristics. Soil Science and Plant Nutrition 31:227-238. Larson, W.E., and Pierce, F.J. 1991. Conservation and enhancement of soil quality. In Evaluation for Sustainable Land Management in the Developing World. Vol . 2: IB S R A M Proceedings 12(2). Bangkok: International Board for Soil Research and Management. Lefroy, R.D.B., Chaitep, W., and Blair, G.J. 1994. Release of sulfur from rice residues under flooded and non-flooded soil conditions. Australian Journal of Agricultural Research 45:657-667. Loetsch, F. 1958. Der Einflub des Brandrodungsbaues auf das Gefuge des Tropenwaldes und die Wasserfuhrung der Strome, untersucht am Beispiel Nordthailands. Erdkunde 12:182-205. . 1962. Der Einflub des Brandrodungsbaues auf das Gefuge des Tropenwaldes und die Wasserfuhrung des Mae Nam Chao Phraya. Thailand-Studien, Bd. 15. Schriften des Instituts fur Asienkunde in Hamburg :61-80. Mae Chaem District Office. 1991. Annual Report. Chiang Mai. Photocopied. MacDicken, K . G . 1994. Sustainability of Leucaena leucocephala fallows in shifting cultivation on the Island of Mindoro, Philippines. Ph.D. Thesis, The University of British Columbia. Martawijaya, A. , and Sumarni, G. 1978. Resistance of a number of Indonesian wood species against Cryptotermes cyanocephalus light. In ITJFRO Conference on Wood Quality and the Utilization of Tropical Species, ed. IUFRO, 371-383. Laguna: International Union of Forestry Research Organizations. Mailly, D., Christanty, L. , and Kimmins, J.P. 1997. "Without bamboo, the land dies": Nutrient cycling and biogeochemistry of a Javanese bamboo talun-kebun system. Forest Ecology and Management 91:155-173. McGrath, D.G. 1987. The role of biomass in shifting cultivation. Human Ecology 15:221-242. McLean, E.O. 1982. Soil pH. In Methods of Soil Analysis. Part 2 Chemical and Microbiological Properties, 2d ed. Page, A . L . , 199-224. Madison: Soil Science Society of America. Mischung, R. 1990. Geschichte, Gesellschaft und Umwelt: eine kulturokologische Fallstudie uber zwei Bergvolker Sudostasiens. Habilitation thesis. Frankfurt. Mishra, B.K. , and Ramakrishnan, P.S. 1984. Nitrogen budget under rotational bush fallow Agriculture (Jhum) in north-eastern India. Plant and Soil 80:237-246. Mueller-Harvey, I., Juo, A.S.R., Wild, A . 1985. Soil organic C, N , S, and P after forest clearance in Nigeria: mineralization rates and spatial variability. Journal of Soil Science 36:585-591. Na-nagara, T. 1990. Soil constraints on sustainable plant production in Thailand. In Soil Constraints on Sustainable Plant Production in the Tropics, ed. Tropical Agriculture Research Center, 40-59. Ibaraki: Tropical Agriculture Research Center. Nakano, K . 1978. An ecological study of swidden agriculture at a village in northern Thailand. South East Asian Studies 16:411-446. Nishio, M . , and Kusano, S. 1979. Pyrenochaeta sp. as one of the causal agents of upland rice soil sickness. Japan Agricultural Research Quarterly 13:37-42. Nishizawa, T., and Ohshima, Y . 1972. Annual changes in the nematode fauna in experimental fields with continuous or rotational croppings with 4 crops: 16 th Annual meeting of the Japanese Society of Applied Entomology and Zoology held in Japan April 1972. Shizuoka. NSC/NESDB. 1993. A Directory of Highland Communities and Population 1993. Bangkok: National Security Council and National Economic and Social Development Board. Nye, P.H., and Greenland, D.J. 1960. The Soil under Shifting Cultivation. Tech. Comm. No. 51, Comm. Bur. of Soils, Farnham Royal. 172 . 1964. Changes in soil after clearing of tropical forest. Plant and Soil 21:101-112. Obatolu, C.R., and Agboola, A . A . 1993. The potential of Siam weed (Chromolaena odorata) as a source of organic matter for soils in the humid tropics. In Soil Organic Matter Dynamics and Sustainability of Tropical Agriculture, ed. Mulongoy, K. , and Merckx, R., 89-99. Chichester: John Wiley. Pampasit, S. 1998. Studies on concentration and accumulation of chemical elements in trees and soils of tropical hill evergreen forest, northern Thailand. Ph.D. Thesis, Ehime University. Pande, H . K . 1994. Improved Upland Rice Farming Systems. Rome: FAO. Ponnamperuma, F.N. 1975. Growth-limiting factors of aerobic soils. In Major Research in Upland Rice, ed. International Rice Research Institute, 40-43. Manila: International Rice Research Institute. Pragtong, K . 1996. Community forestry activities in Thailand. In Group Training Course on Community Forestry Development Techniques, ed. Royal Forestry Department, 17-32. Bangkok: Royal Forestry Department. Photocopied. Prungkaew, J. 1998. Report on Land Use in the Integrated Natural Resource Management in Watersheds of Mae Chaem District, Chiang Mai Province, and Sob Moey District, Mae Hong Son Province. Chiang Mai: CARE-Thailand. Photocopied. Rerkasem, K. , and Rerkasem, B . 1994. Shifting Cultivation in Thailand: Its Current Situation and Dynamics in the Context of Highland Development. London: Forestry and Land Use, International Institute for Environment and Development. Royal Forestry Department (RFD). 1998. Forest Area in Thailand (1961-1998). Bangkok: Forest Resources Division, Forest Resource Office. Royal Forest Department. Photocopied. . 1999. Weather data (1985-1999) at the Mae Chaem Watershed Research Station. Chiang Mai: Royal Forest Department. Photocopied. Roder, W., Phouaravanh, B. , Phengchanh, S., and Keoboulapha, B . 1995. Relationships between soil, fallow period, weeds, and rice yield in slash-and-burn systems of Laos. Plant and Soil 176:27-36. Roder, W., Phengchanh, S., and Keobulapha, B . 1997a. Weeds in slash-and-burn rice fields in northern Laos. Weed Research 37:111-119. Roder, W., Phengchanh, S., and Maniphone, S. 1997b. Dynamics of soil and vegetation during crop and fallow period in slash-and-burn fields of northern Laos. Geoderma 76:131-144. Ross, J.D., Speir, T.W., Tate, K.R., and Orchard, V . A . 1985. Effects of sieving on estimations of 173 microbial biomass, and carbon and nitrogen mineralization in soil under pasture. Australian Journal of Soil Research 23:319-324. Ruthenberg, H . 1971. Farming Systems in the Tropics. Oxford: Clarendon Press. Sabhasri, S. 1978. Effect of forest fallow cultivation on forest production and soil. In Farmers in the Forest: Economic Development and Marginal Agriculture in Northern Thailand, ed. Kunstadter, P., Chapman, E.C., and Sabhasri, S., 160-184. Honolulu: The University Press of Hawaii. Sajjapongse, A . 1997. The sloping lands network and the opportunity for technology transfer. In Soil Conservation Extension: From Concepts to Adoption. Sombatpanit, S., Zobisch, M . A . , Sanders, D.W., and Cook, M.G. , 325-335. Bangkok: Soil and Water Conservation Society of Thailand. Sanchez, P. A . 1976. Properties and Management of Soils in the Tropics. New York: John Wiley. . 1995. Science in agroforestry. Agroforestry Systems 30:5-55. Sangwanit, U . 1995. Reforestation research in Thailand. In Caring for the Forest: Research in a Changing World. Congress Report: JUFRO X X World Congress held in Finland 6-12 August 1995, ed. International Union of Forestry Research Organizations, 525-531. Tampere, iTJFRO. Santisuk, T. 1988. A n account of the vegetation of northern Thailand, Geoecological Research Vol . 5, ed. Schweinfurth, U . Stuttgart: Franz Steiner Verlag Wiesbadens Gmth. SAS Institute. 1996. SAS system for Windows, Release 6.12. Cary: SAS Institute. Saxena, K . G . , and Ramakrishnan, P.S. 1983. Growth and allocation strategies of some perennial weeds of slash and burn agriculture (Jhum) in northeast India. Canadian Journal of Botany 61:1300-1306. Schimel, D.S., and Parton, W.J. 1986. Microclimatic controls of nitrogen mineralization and nitrification in shortgrass steppe soils. Plant and Soil 93:347-357. Schmidt-Vogt, D. 1999. Swidden farming and fallow vegetation in northern Thailand, Geoecological Research Vol . 8, ed. Schweinfurth, U . Stuttgart: Franz Steiner Verlag Wiesbadens Gmth. Setboonsarng, S. 1996. Rice research priorities in Thailand. In Rice Research in Asia: Progress and Priorities, ed. Evensen, R.E., Herdt, R.W., and Hossein, M . , 217-238. Wallingford: IRRI and C A B International. Shils, M.E . 1990. Magnesium. In Present Knowledge in Nutrition, 6d ed. Brown, M . L . , 224-232. Washington, DC: International Life Sciences Institute-Nutrition Foundation. 174 Shiver, B.D., and Borders, B.E. 1996. Sampling Techniques for Forest Resource Inventory. New York: John Wiley and Sons. Silva-Forsberg, M.C. , and Fearnside, P .M. 1997. Brazilian Amazonian caboclo agriculture: effect of fallow period on maize yield. Forest Ecology and Management 97:283-291. Singh, R.S., Raghubanshi, A.S., and Singh, J.S. 1991. Nitrogen-mineralization in dry tropical savanna: effects of burning and grazing. Soil Biology and Biochemistry 23:269-273. Sombroek, W.G. 1967. Amazon Soils. Wageningen: Center for Agricultural Publications and Documentation. Spencer, D.S.C., and Swift, M.J. 1992. Sustainable agriculture: Definition and measurement. In Biological Nitrogen Fixation and Sustainability of Tropical Agriculture, ed. Mulongoy, K. , Gueye, M . , and Spencer, D.S.C. Chichester: John Wiley. Sprent, J.I. 1999. Not all nitrogen fixing legumes have a high requirement for phosphorus. Agroforestry Forum 9:17-20. Srisawas, N . 1983. Thailand. In Swidden Cultivation in Asia. Volume I Content Analysis of the Existing Literature: A Stocktaking Exercise, ed. UNESCO, 289-330. Bangkok: Office for the Regional Adviser for Social Science in Asia and the Pacific, UNESCO Regional Office for Education in Asia and the Pacific. Stromgaard, P. 1984. The immediate effect of burning and ash-fertilization. Plant and Soil 80:307-320. Sujumnong, P. 1980. Risk Factors of Morbidity and Mortality of Karen Infant and Children. Chiang Mai: Social Research Institute, Chiang Mai University. Sundarapandian, S.M., and Swamy, P.S. 1996. Fine root biomass distribution and productivity patterns under open and closed canopies of tropical forest ecosystem at Kodayar in Western Ghats, South India. Forest Ecology and Management 86:181-192. Sutthi, C. 1996. Highland Swidden Agriculture: Indigenous Knowledge. Chiang Mai: Nuntakan Graphic Printing. Swamy, P.S., and Ramakrishnan, P.S. 1988. Ecological implications of traditional weeding and other imposed weeding regimes under slash-and-burn agriculture (Jhum) in northeastern India. Weed Research 28:127-136. Swift, M.J . , and Woomer, P. 1993. Organic matter and the sustainability of agricultural systems: 175 Definition and measurement. In Organic Matter Dynamics and Sustainability of Tropical Agriculture, ed. Mulongoy, K. , and Merckx, R., 3-18. Chichester: John Wiley. Szott, L.T., and Palm, C.A. 1996. Nutrient stocks in managed and natural humid tropical fallows. Plant and Soil 186:293-309. Tanaka, I. 1976. Agro-ecological roles of irrigation in rice culture. In Symposium on water management in rice field: Proceedings of a symposium on tropical agriculture researches August 1975 edited by Tropical Agriculture Research Center. Ibaraki: Tropical Agriculture Research Center. Tanaka, S., Funakawa, S., Kaewkhongkha, T., Hattori, T., and Yonebayashi, K . 1997. Soil ecological study on dynamics of K , Mg, and Ca, and soil acidity in shifting cultivation in northern Thailand. Soil Science and Plant Nutrition 43:695-708. Tanaka, S., Funakawa, S., Kaewkhongkha, T., and Yonebayashi, K . 1998a. Labile pools of organic matter and microbial biomass in the surface soils under shifting cultivation in Northern Thailand. Soil Science and Plant Nutrition 44:527-537. . 1998b. N Mineralization process of the surface soils under shifting cultivation in Northern Thailand. Soil Science and Plant Nutrition 44:539-549. Thomas, G.W. 1982. Exchangeable cations. In Methods of Soil Analysis. Part 2 Chemical and Microbiological Properties, 2d ed. Page, A . L . , 159-165. Madison, Soil Science Society of America. Thomas, D.E. 1995. Opportunities and limitations for agroforestry systems in the highlands of North Thailand. In Highland farming: Soil and the future?: Proceedings held in Chiang Mai 21-22 December 1995, edited by Turkelboom, F., van Look-Rothchild, K , and van Keer, K . Chiang Mai: Soil Fertility Conservation Project, Maejo University and Catholic University of Leuven. Toky, O.P., and Ramakrishnan, P.S. 1981. Run-off and infiltration losses related to shifting agriculture (Jhum) in north-eastern India. Environmental Conservation 8:313-321. Topark-ngarm, B . 1983. Laboratory manual for soil fertility. Khon Kean: Faculty of Agriculture, Khon Kean University. Photocopied. Tribal Research Institute. 1995. The Hi l l Tribes of Thailand, 4d ed. Chiang Mai: Technical Service Club, Tribal Research Institute. Troeh, F.R., and Thompson, L . M . 1993. Soils and Soil Fertility, 5d ed. New York: Oxford University Press. Tsutsumi, T., Yoda, K. , Sahunalu, P., Dhanmanonda, P., and Prachaiyo, B. 1983. Forest: felling, 176 burning and regeneration. In Shifting Cultivation: A n Experiment at Nam Phrom, Northeast Thailand, and its Implications for Upland Farming in the Monsoon Tropics, ed. Kyuma, K. , and Pairintra, C , 13-62. Kyoto: Faculty of Agriculture, Kyoto University. Tulaphitak, T., Pairintra, C , and Kyuma, K . 1985a. Changes in soil fertility and tilth under shifting cultivation. II Changes in soil nutrient status. Soil Science and Plant Nutrition 31: 239-249. 1985b. Changes in soil fertility and tilth under shifting cultivation. I l l Soil respiration and soil tilth. Soil Science and Plant Nutrition 31:251-261. Uhl, C , Clark, K. , Clark, FL, and Murphy, P. 1981. Early plant succession after cutting and burning in the upper Rio Negro of the Amazon basin. Journal of Ecology 69:631-649. Unruh, J.D. 1990. Iterative increase of economic tree species in managed swidden-fallows of the Amazon. Agroforestry Systems 11:175-197. Ventura, W., and Watanabe, I. 1978. Growth inhibition due to continuous cropping of dryland rice and other crops. Soil Science and Plant Nutrition 24:375-389. Wada, S.I. and Wada, K . 1985. Charge characteristics and exchangeable cation status of Korean Ultisols and Alfisols and Thai Ultisols and Oxisols. Soil Science and Plant Nutrition 36:21-29. Wadsworth, G., Reisenauer, H . M . , Gordon, D.R., Singer, M.J . 1990. Effects of length of forest fallow on fertility dynamics in a Mexican ultisol. Plant and Soil 122:151-156. Walker, D.H., Sinclair, F.L., and Thapa, B. 1995. Incorporation of indigenous knowledge and perspectives in agroforestry development. Part I Review of methods and their application. Agroforestry Systems 30:235-248. Walkley, A . , and Black, L A . 1934. An examination of the Degtjarf method for determination of soil organic matter and a proposed modification of the chromic acid titration method. Soil Science 37:29-38. Warner, K . 1991. Shifting Cultivators: Local Technical Knowledge and Natural Resource Management in the Humid Tropics. Rome: FAO. Watters, R.F. 1960. The nature of shifting cultivation. Pacific Viewpoint 1:59-99. WCED. 1987. Our Common Future. World Commission on Environment and Development. Oxford: Oxford University Press. Ziegler, A . , and Giambelluca, T. 1997. Simulation of runoff and erosion on mountainous roads 177 in northern Thailand: A first look. In Human Impacts on Erosion and Sedimentation, ed. Walling, D.E., and Probst, J.-L., 21-29. Wallingford: International Association of Hydrological Sciences (IAHS) Press. Zinke, P., Sabhasri, S., and Kunstadter, P. 1978. Soil fertility aspects of the Lua' forest fallow system of shifting cultivation. In Farmers in the Forest: Economic Development and Marginal Agriculture in Northern Thailand, ed. Kunstadter, P., Chapman, E.C., and Sabhasri, S., 134-159. Honolulu: The University Press of Hawaii. Zou, X . M . , Valentine, D.W., Sanford, R.L., Jr., and Binkley, D. 1992. Resin-core and buried-bag estimates of nitrogen transformations in Costa Rican lowland rainforests. Plant and Soil 139:275-283. 178 Appendix 1 Weather Data Table A . 1.1 Temperature and rainfall data (1998 and 1999) at Mae Chaem Watershed Research Station*, Royal Forest Department, Chiang Mai, Thailand (RFD 1999) Month Minimum Maximum Mean Rainfall Temperature Temperature Temperature (mm) ( ° Q (°C) ( ° Q 1998 January 15 35 25 0 February 16 35 25 0 March 18 31 25 0 April 20 32 26 0 May 20 30 25 89 June 20 25 22 51 July 19 24 22 166 August 20 27 23 272 September 19 27 23 239 October 18 29 24 23 November 17 30 24 16 December 15 30 23 0 Mean 18 30 24 856 (Total) 1999 January 15 31 23 0 February 17 30 24 0 March 18 31 24 0 April 20 29 24 76 May 19 26 22 223 June 20 25 22 198 July 20 25 23 131 August 20 24 22 232 September 20 25 23 215 October 19 25 22 226 November 17 25 21 37 December 13 23 18 4 Mean 18 27 22 1,342 (Total) * The station is approximately 30 km east of Mae Hae Tai village, and was where the pot experiment was set up. 179 Table A.1.2 Mean annual temperature and annual rainfall data from 1985 to 1999 at Mae Chaem Watershed Research Station*, Royal Forestry Department, Chiang Mai, Thailand (RFD 1999) Year Mean Annual Temperature (° C) Annual Rainfall (mm) 1985 20 1,367 1986 22 1,063 1987 23 1,517 1988 23 1,236 1989 22 1,242 1990 22 1,175 1991 22 1,223 1992 22 1,187 1993 22 1,065 1994 15 1,133 1995 15 1,327 1996 22 1,363 1997 22 1,114 1998 24 856 1999 22 1,342 Mean 21 1,214 * The station is approximately 30 km east of Mae Hae Tai village, and was where the pot experiment was set up. 180 Appendix 2 Soil Moisture Soils from the shifting cultivation fields at 0-5 cm depth were collected in 1999 to determined monthly moisture contents. The collection was done from three points at approximately 2-m intervals apart at each slope position in each field. The soils were oven dried at 105° C for two days to determine the moisture content. The numbers presented are average percentage of moisture content calculated from all three slope positions (Table A.2.1). Table A.2.1 Surface-soil (0-5 cm) monthly moisture content (%) Month Rice 2 n Q Y r 2-Year 4-Year 4-Year 5-Year Secondary (1999) Field Rice Fallow Fallow Fallow Fallow Forest (1998) Field (1998) N W SW (1998) (1999) (1998) (1998) January 2.5 - 3.3 2.9 2.0 4.2 4.9 February 6.7 - 15 13 5.9 - 14 March* 11 - 21 21 17 - 22 April 2.2 - 5.5 7.2 5.1 - 4.8 May 9.3 - 13 15 14 10 19 June 4.1 7.5 14 4.3 7.5 3.5 5.5 July 5.3 10 17 13 14 15 18 August 48 57 39 35 42 49 53 September 28 24 24 29 27 26 29 October 24 19 27 31 26 19 33 November 29 27 29 31 29 26 37 1) The second-year rice field was a part of the 1998 rice field. 2) The 5-year fallow field became the 1999 rice field in April 1999. Italic months are in rainy season. * Rain at the beginning of the month. Appendix 3 Soil Temperature Two thermometers were buried under the soil surface in the rice field and the secondary forest to record monthly temperature from March to October 1999. Table A.3.1 Surface-soil (0-5 cm) monthly temperature Date Temperature (°C) Rice Field (1998) 4-Year Fallow Field (1998) Secondary Forest Mar 6, 99 Current 31 (4 PM) n.a. 20 (1 PM) Min 17 n.a. 18.5 Max 32 n.a. 22.5 Mar 28, 99 Current 22 (8 A M ) n.a. 23 (3 PM) M i n 17 n.a. 11 Max 41 n.a. 33 April 25, 99 Current 22 (10 A M ) n.a. 24 (3 PM) M i n 17 n.a. 19 Max 41 n.a. 24 June 13, 99 Current 22 (10 A M ) 23 (2 PM) 23 (2 PM) M i n 17 19 19 Max 40 37 37 July 17, 99 Current 23 (5 PM) 25 (4 PM) _* Min 17 21 _* Max 40 37 _* Aug 21, 99 Current 24.5 (1 PM) 25 (4 PM) 21 (11 A M ) Min 18 19 19 Max 40 37 30 Sept 19, 99 Current 23 (3 PM) 21.5 (10 A M ) 21 (12 PM) M i n 19 19 19 Max 29 37 31 Oct 9, 99 Current 23 (11 A M ) 24 (4 PM) 23 (3 PM) M i n 19 19 19 Max 29 37 30 Oct 31, 99 Current 22 (4 PM) 22 (2 PM) 20(11 A M ) M i n 17 17 18 Max 29 37 31 Mean Current 24 23 22 M i n 18 19 18 Max 36 37 30 n.a. = not available because another thermometer was added later in the 4-year fallow field. * The thermometer was broken. 182 Appendix 4 Sub-Soil Properties Table A.4.1 Soil texture at four depths in a chronosequence of shifting cultivation fields Field (in 1998) Texture 30-40 40-60 60-80 80-100 (%) cm cm cm cm Rice Field Clay 63 63 57 54 (after harvest) Silt 24 19 18 18 Sand 13 18 25 28 2-year Fallow Clay 40 31 49 65 Silt 28 29 25 20 Sand 32 40 26 15 4-year Fallow-NW Clay 62 65 69 69 Silt 20 22 19 20 Sand 18 13 12 11 4-year Fallow-SW Clay 54 42 58 46 Silt 34 27 24 26 Sand 12 31 18 28 5-year Fallow Clay 58 53 57 58 Silt 14 16 22 22 Sand 28 31 21 20 Secondary Forest Clay 65 48 44 49 Silt 19 14 16 18 Sand 16 38 40 33 Table A.4.2 Soil p H ( l : l in water) at four depth s in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 5.57 (0.33) 5.58 (0.43) 5.49 (0.34) 5.19(0.03) 2-year Fallow 4.73 (0.08) 4.91 (0.10) 5.09 (0.03) 5.14(0.04) 4-year Fallow-NW 4.86 (0.07) 5.08 (0.02) 5.26 (0.06) 5.12(0.03) 4-year Fallow-SW 5.08 (0.05) 5.09 (0.04) 5.14(0.04) 5.18(0.03) 5-year Fallow 4.82 (0.02) 4.95 (0.04) 4.97 (0.04) 5.03 (0.05) Secondary Forest 4.78 (0.03) 4.82 (0.11) 4.95 (0.01) 5.08 (0.02) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). 183 Table A.4.3 Soil pH (1:1 in KC1) at four depths in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 4.67 (0.41) 4.57 (0.48) 4.45 (0.40) 4.12(0.05) 2-year Fallow 4.06 (0.05) 4.10(0.06) 4.18(0.05) 4.21 (0.07) 4-year Fallow-NW 4.09 (0.05) 4.17(0.04) 4.23 (0.04) 4.20(0.02) 4-year Fallow-SW 4.18(0.07) 4.07 (0.02) 4.18(0.10) 4.16(0.05) 5-year Fallow 3.98 (0.03) 4.04 (0.05) 4.05 (0.05) 4.08 (0.03) Secondary Forest 3.93 (0.03) 3.94 (0.03) 4.00 (0.03) 4.01 (0.03) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). Table A.4.4 Soil organic matter concentrations (%) at four depths in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 1.47 (0.26) 0.71 (0.20) 0.64 (0.13) 0.46 (0.12) 2-year Fallow 3.64 (0.34) 2.13 (0.28) 1.32 (0.04) 0.85 (0.08) 4-year Fallow-NW 2.53 (0.22) 1.49 (0.17) 0.94 (0.03) 0.76 (0.08) 4-year Fallow-SW 2.14(0.56) 1.33 (0.14) 1.02 (0.13) 0.73 (0.07) 5-year Fallow 1.52 (0.17) 1.20 (0.15) 1.15(0.11) 0.86 (0.16) Secondary Forest 2.31 (0.38) 1.54 (0.32) 1.25 (0.24) 0.67 (0.14) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). Table A.4.5 Soil total N concentrations (%) at four depths in a cultivation fields chronosequence of shifting Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 0.10(0.01) 0.07 (0.01) 0.06 (0.02) 0.04 (0.02) 2-year Fallow 0.17(0.01) 0.11 (0.01) 0.10(0.004) 0.08 (0.01) 4-year Fallow-NW 0.11 (0.001) 0.10(0.002) 0.08 (0.01) 0.06 (0.01) 4-year Fallow-SW 0.11 (0.02) 0.09 (0.01) 0.07 (0.01) 0.28 (0.24) 5-year Fallow 0.12 (0.02) 0.10(0.01) 0.10(0.01) 0.07 (0.01) Secondary Forest 0.13 (0.02) 0.10(0.01) 0.29 (0.19) 0.07 (0.01) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). 184 Table A.4.6 Soil available P concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 1.2 (0.4) 1.0 (0.5) 0.5 (0) 1.5 (0.5) 2-year Fallow 2.2 (0.6) 1.2 (0.4) 1.0(0.3) 1.0 (0.3) 4-year Fallow-NW 2.2(1.2) 1.0 (0.3) 0.5 (0) 0.5 (0) 4-year Fallow-SW 1.8(0.7) 1.3 (0.6) 1.2 (0.4) 0.7 (0.2) 5-year Fallow 1.5 (0.6) 1.0(0) 0.5 (0) 0.5 (0) Secondary Forest 1.3 (0.2) 1.0 (0.3) 0.5 (0) 0.7 (0.2) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). Table A.4.7 Soil extractable K concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 212 (6) 173 (19) 151 (13) 116(3) 2-year Fallow 83 (11) 65(12) 56 (12) 51(13) 4-year Fallow-NW 65 (5) 57 (5) 55(4) 56(3) 4-year Fallow-SW 203 (52) 172 (33) 165 (41) 161 (35) 5-year Fallow 103 (27) 99 (23) 93 (23) 46 (27) Secondary Forest 97(18) 81 (16) 51(3) 43 (3) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). Table A.4.8 Soil extractable Ca concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 612 (239) 358 (206) 392 (311) 100 (25) 2-year Fallow 129 (54) 71 (29) 58(17) 42 (8) 4-year Fallow-NW 171 (76) 121 (65) 92 (40) 62 (26) 4-year Fallow-SW 87 (12) 150 (88) 171 (115) 133 (77) 5-year Fallow 54(11) 46 (4) 46(4) 50(7) Secondary Forest 117(85) 42 (4) 42 (8) 35 (14) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). 185 Table A.4.9 Soil extractable M g concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 145 (33) 121 (51) 112(65) 46 (14) 2-year Fallow 25 (11) 20 (8) 12(5) 9.2 (3) 4-year Fallow-NW 35 (11) 24(11) 20 (10) 18(6) 4-year Fallow-SW 49 (5) 60 (23) 64 (25) 68 (32) 5-year Fallow 25(4) 24(5) 25 (7) 24 (8) Secondary Forest 42 (28) 14(4) 15(9) 14(4) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). Table A.4.10 Soil CEC (cmol /kg) at four depths in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 29.8(9.1) 19.1 (3.4) 25.6 (0.8) 16.7 (0.8) 2-year Fallow 27.3 (0.4) 21.2(1.8) 22.2 (2.9) 31.3 (7.7) 4-year Fallow-NW 45.3 (11) 44.8 (2.2) 47.7(1.0) 39.5 (3.9) 4-year Fallow-SW 38.6 (4.6) 33.2 (3.1) 37.3 (1.1) 36.0 (0.9) 5-year Fallow 18.0(2.0) 18.9(2.9) 14.2 (3.3) 28.1 (3.0) Secondary Forest 41.6 (3.9) 34.4 (5.1) 38.9 (3.0) 31.8(4.3) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). Table A.4.11 Soil extractable Fe concentrations (ppm) at four depths in a chronosequence of shifting cultivation fields Field (in 1998) 30-40 cm 40-60 cm 60-80 cm 80-100 cm Rice Field (after harvest) 34 (23) 20(17) 11 (7.2) 6.5 (2.3) 2-year Fallow 50 (21) 15 (8.7) 3.3 (1.1) 2.3 (0.7) 4-year Fallow-NW 16 (3.9) 4.8(1.1) 1.7(0.4) 1.0 (0.3) 4-year Fallow-SW 33 (24) 15 (9.0) 6.2 (2.5) 4.2(1.5) 5-year Fallow 7.2 (1.9) 5.2 (0.4) 4.5 (1.1) 2.8 (0.6) Secondary Forest 27 (3.1) 14 (2.5) 6.5 (1.0) 2.8 (0.8) 1) The data were averaged from three slope positions. 2) Numbers in parentheses are standard errors (n=3). 186 Appendix 5 Soil Profile Description 1)1998 Rice Field I. Information on the Site : a. Profile number : 1998 Rice Field b. Soil name : Unknown c. Higher category classification : U S D A : Ultisols d. Date of examination : March 20 t h, 1999 e. Author : Wangpakapattanawong, P . , N . Anongrak and D. Saramathakul f. Location : Mae Hae Tai village, Mae Chaem District, Chiang Mai, Thailand (Grid Reference: 101388, Sheet 4645 IV) g. Elevation : 1,060 m asl. h. Land form : i . physiographic position : on convex slope i i . surrounding landform : mountainous i i i . microtopography : nil i . Slope on which profile is sited : steep slope (55%), north 60 west aspect j . Vegetation and Land-use : Under hill evergreen forest k. Climate : Data derived from Mae Chaem Watershed Research Station, Royal Forest Department, 30 km of the site at the elevation of approximately 1,000 m asl. II. General Information on the So i l : a- Parent material : Apparently derived "in situ" from shale in Paleozoic era (Carboniferous, Devonian and Silurian period) b. Drainage : Moderately well drained c. Moisture condition in profile : Moist below 50/55 cm d. Depth of groundwater table : Unknown e. Presence of surface stone and rock outcrops : Fairly stony and no rocks f. Evidence of erosion : Moderate sheet and rill erosion at site and in adjacent field g. Presence of salt or alkali: Soils free of excess salt or alkali h. Human influence : Shifting cultivation field 187 III. Profile Description Horizon Depth (cm) Description Ap 0-12/17 Dark yellowish brown (10 Y R 4/4) moist and yellowish brown (10 Y R 5/4) dry, clay loam; moderate medium and coarse granular and moderate medium subangular blocky; slightly sticky, slightly plastic, friable when moist, slightly hard when dry; few gravel of slightly to moderate weathered shale; abundant ubiquitous roots of rice; clear and smooth boundary; pH 5.8. B A 12/17-28/35 Brown to dark brown (7.5 Y R 4/4) moist and brown (7.5 Y R 5/4) dry, clay loam; moderate medium subangular blocky; slightly sticky, slightly plastic, friable when moist, slightly hard when dry; common gravel of moderate weathered shale; many ubiquitous roots of rice; clear and wavy boundary; pH 5.8. B t l 28/35-50/55 Brown to dark brown (7.5 Y R 4/4) moist and brown (7.5 Y R 5/4) dry, clay; moderate medium subangular blocky; sticky, slightly plastic to plastic, firm when moist, slightly hard when dry; common gravel of strongly weathered shale; few ubiquitous roots of rice; clear and smooth boundary; pH 5.6. Bt2 50/55-80/85 Yellowish red (5 Y R 5/6) moist, clay; moderate medium and coarse subangular blocky; very sticky, plastic to very plastic, firm to very firm when moist; common gravel of moderate weathered shale; few ubiquitous roots of rice; clear and broken boundary; pH 5.5. Bt3 80/85-95/114 Yellowish red (5 Y R 5/6) moist, clay; moderate medium and coarse subangular blocky; very sticky, very plastic, very firm when moist; common gravel of moderate weathered shale; no roots; clear and broken boundary; pH 5.2. Bt4 95/114-118+ Yellowish red (5 Y R 5/8) moist, gravelly clay; moderate fine and medium subangular blocky; sticky, slightly plastic to plastic, firm to very firm when moist; many gravel of slightly to moderate weathered shale; no roots; pH 5.2. 188 2) Two-Year Fallow Field fin 1998) I. Information on the Site : a. Profile number : Two-Year Fallow Field (in 1998) b. Soil name : Unknown c. Higher category classification : U S D A : Ultisols d. Date of examination : March 21 s t, 1999 e. Author : Wangpakapattanawong, P . , N . Anongrak and D. Saramathakul f. Location : Mae Hae Tai village, Mae Chaem District, Chiang Mai , Thailand (Grid Reference: 092367, Sheet 4645 IV) g. Elevation : 1,080 m asl. h. Land form : i . physiographic position : on convex slope, 100 m far from ridge i i . surrounding landform : mountainous i i i . microtopography : contour terracing i . Slope on which profile is sited : steep slope (38%), north 10 east aspect j . Vegetation and Land-use : Under hill evergreen forest k. Climate : Data derived from Mae Chaem Watershed Research Station, Royal Forest Department, 30 km of the site at the elevation of approximately 1,000 m asl. II. General Information on the So i l : a. Parent material : Apparently derived "in situ" from shale in Paleozoic era (Carboniferous, Devonian and Silurian period) b. Drainage : Moderately well drained c. Moisture condition in profile : Moist below 15 cm d. Depth of groundwater table : Unknown e. Presence of surface stone and rock outcrops : Fairly stony and no rocks f. Evidence of erosion : Slight sheet and rill erosion at site and in adjacent field g. Presence of salt or alkali: Soils free of excess salt or alkali h. Human influence : Two-year fallow field 189 III. Profile Description Horizon Depth (cm) Description A(p) 0-15 Dark reddish brown (5 Y R 3/4) moist and dark reddish brown (5 Y R 3/3) dry, clay loam; moderate fine and medium granular; slightly sticky, slightly plastic, friable when moist, slightly hard when dry; few gravel of strongly weathered shale; abundant ubiquitous roots of evergreen trees and grass; abrupt and smooth boundary; pH 4.9. A B 15-32 Reddish brown (5 Y R 4/3) moist, clay loam to clay; moderate fine and medium granular and moderate fine and medium subangular blocky; slightly sticky, plastic, friable to firm when moist; no gravel; abundant vertical and oblique roots of evergreen trees; abrupt and smooth boundary; pH 4.7. B t l 32-47/52 Reddish brown (2.5 Y R 4/4) moist, clay; moderate fine and medium subangular blocky; sticky, plastic, firm when moist; no gravel; many vertical and oblique roots of evergreen trees; clear and wavy boundary; pH 4.7. Bt2 47/52-83/92 Red (2.5 Y R 4/6) moist, clay; moderate fine and medium subangular blocky; very sticky, plastic to very plastic, very firm when moist; few gravel of strongly weathered shale; few vertical and oblique roots (1.0-3.0 cm.) of evergreen trees; clear and wavy boundary; pH 5.0. Bt3 83/92-112+ Red (2.5 Y R 4/8) moist, clay; moderate fine and medium subangular blocky; very sticky, very plastic, very firm when moist; few gravel of strongly weathered shale; few vertical roots (5.0-1.0 cm.) of evergreen trees; pH 5.1. 190 3) Four-Year Fallow Field (NW) (in 1998) I. Information on the Site : a. Profile number : Four-Year Fallow Field (NW) (in 1998) b. Soil name : Unknown c. Higher category classification : U S D A : Ultisols d. Date of examination : March 20 t h, 1999 e. Author : Wangpakapattanawong, P . , N . Anongrak and D. Saramathakul f. Location : Mae Hae Tai village, Mae Chaem District, Chiang Mai, Thailand (Grid Reference: 100390, Sheet 4645 IV) g. Elevation : 1,020 m asl. h. Land form : i . physiographic position : on concave slope, 50 m downslope from ridge i i . surrounding landform : mountainous i i i . microtopography : nil i . Slope on which profile is sited : steep slope (50%), north 10 west aspect j . Vegetation and Land-use : Under hill evergreen forest k. Climate : Data derived from Mae Chaem Watershed Research Station, Royal Forest Department, 30 km of the site at the elevation of approximately 1,000 m asl. II. General Information on the So i l : a. Parent material : Apparently derived "in situ" from shale in Paleozoic era (Carboniferous, Devonian and Silurian period) b. Drainage : Moderately well drained c. Moisture condition in profile : Moist below 14 cm d. Depth of groundwater table : Unknown e. Presence of surface stone and rock outcrops : No stones and no rocks f. Evidence of erosion : Slight sheet and rill erosion at site and in adjacent field g. Presence of salt or alkali: Soils free of excess salt or alkali h. Human influence : Four-year fallow field 191 III. Profile Description Horizon Depth (cm) Description A(p) 0-14 Dark yellowish brown (10 Y R 4/4) moist and yellowish brown (10 Y R 5/4) dry, clay loam; moderate fine and medium granular and moderate medium subangular blocky; slightly sticky, slightly plastic, friable when moist, slightly hard when dry; few gravel of strongly weathered shale; many vertical and ubiquitous roots of evergreen trees and grass; abrupt and smooth boundary; pH 4.9. B(t)l 14-35 Brown to dark brown (7.5 Y R 4/4) moist, clay loam to clay; weak to moderate fine and medium subangular blocky; sticky, slightly plastic, friable to firm when moist; few gravel of strongly weathered shale; few vertical roots (6.0-10.0 cm.) of evergreen trees; clear and wavy boundary; pH 4.6. Bt2 35-60/65 Yellowish red (5 Y R 4/6) moist, clay; moderate medium and coarse subangular blocky; sticky, plastic, firm when moist; common gravel of strongly weathered shale; few vertical roots of evergreen trees; clear and wavy boundary; pH 5.0. Bt3 60/65-87/90 Yellowish red (5 Y R 5/6) moist, clay; moderate medium subangular blocky; sticky, plastic, firm when moist; common gravel of strongly weathered shale; no roots; clear and wavy boundary; pH 5.3. Bt4 87/90-115+ Yellowish red (5 Y R 5/8) moist, clay; moderate medium subangular blocky; sticky, slightly plastic, firm when moist; many gravel of moderate to strongly weathered shale; no roots; pH5.1. 192 4) Four-Year Fallow Field (SW) (in 1998) I. Information on the Site : a. Profile number : Four-Year Fallow Field (SW) (in 1998) b. Soil name : Unknown c. Higher category classification : U S D A : Ultisols d. Date of examination : March 20 t h, 1999 e. Author : Wangpakapattanawong, P . , N . Anongrak and D. Saramafhakul f. Location : Mae Hae Tai village, Mae Chaem District, Chiang Mai, Thailand (Grid Reference: 113393, Sheet 4645 IV) g. Elevation : 940 m h. Land form : i . physiographic position : on straight slope, 25 m downslope from ridge i i . surrounding landform : mountainous i i i . microtopography : nil i . Slope on which profile is sited : very steep slope (65%), south 30 west aspect j . Vegetation and Land-use : Under hill evergreen forest and dipterocarp forests k. Climate : Data derived from Mae Chaem Watershed Research Station, Royal Forest Department, 30 km of the site at the elevation of approximately 1,000 m asl. II. General Information on the So i l : a. Parent material : Apparently derived "in situ" from shale in Paleozoic era (Carboniferous, Devonian and Silurian period) b. Drainage : Well drained c. Moisture condition in profile : Moist below 28/30 cm d. Depth of groundwater table : Unknown e. Presence of surface stone and rock outcrops : No stones and no rocks f. Evidence of erosion : Moderate sheet and rill erosion at site and in adjacent field g. Presence of salt or alkali: Soils free of excess salt or alkali h. Human influence : Formerly four-year fallow field, now regenerating forest 193 III. Profile Description Horizon Depth (cm) Description A(p) 0-13 Brown to dark brown (10 Y R 4/3) moist and light yellowish brown (10 Y R 6/4) dry, clay loam; moderate fine and medium granular and moderate fine and medium subangular blocky; slightly sticky, slightly plastic, friable when moist, slightly hard when dry; few gravel of moderate to strongly weathered shale; many ubiquitous roots of deciduous trees and grass; abrupt and smooth boundary; pH 5.5. B t l 13-28/30 Brown to dark brown (7.5 Y R 4/4) moist and light brown (7.5 Y R 6/4) dry, clay; moderate fine and medium subangular blocky; sticky, plastic, firm when moist, slightly hard to hard when dry; common gravel of moderate to strongly weathered shale; many oblique roots of deciduous trees and grass; clear and wavy boundary; pH 4.9. Bt2 28/30-48 Strong brown (7.5 Y R 5/6) moist, clay; moderate medium subangular blocky; sticky, plastic, firm when moist; common gravel of moderate to strongly weathered shale; many oblique roots of deciduous trees and grass; clear and wavy boundary; pH5.1. Bt3 48-75/80 Strong brown (7.5 Y R 5/8) moist, clay; moderate medium subangular blocky; very sticky, very plastic, firm to very firm when moist; common gravel of moderate to strongly weathered shale; no roots; clear and wavy boundary; pH 5.1. Bt4 75/80-100/105+ Strong brown (7.5 Y R 5/6) moist, gravelly clay; moderate medium subangular blocky; sticky, plastic, firm when moist; many gravel of moderate to strongly weathered shale; no roots; clear and wavy boundary; pH 5.2. B C 100/105-115+ Strong brown (7.5 Y R 5/8) moist, gravelly clay; moderate medium and coarse subangular blocky; sticky, plastic, firm when moist; many gravel of moderate to strongly weathered shale; no roots; pH 5.2. 194 5) Five-Year Fallow Field fin 1998) I. Information on the Site : a. Profile number : Five-Year Fallow Field (in 1998) b. Soil name : Unknown c. Higher category classification : U S D A : Ultisols d. Date of examination : March 21 s t, 1999 e. Author : Wangpakapattanawong, P . , N . Anongrak and D. Saramathakul f. Location : Mae Hae Tai village, Mae Chaem District, Chiang Mai, Thailand (Grid Reference: 098380, Sheet 4645 IV) g. Elevation : 1,060 m asl. h. Land form : i . physiographic position : on convex slope i i . surrounding land form : mountainous i i i . micro topography : nil i . Slope on which profile is sited : steep slope (39%), north 85 east aspect j . Vegetation and Land-use : Under hill evergreen forest k. Climate : Data derived from Mae Chaem Watershed Research Station, Royal Forest Department, 30 km of the site at the elevation of approximately 1,000 m asl. II. General Information on the So i l : a. Parent material : Apparently derived "in situ" from shale in Paleozoic era Carboniferous, Devonian and Silurian period) b. Drainage : Moderately well drained c. Moisture condition in profile : Moist below 13/17 cm d. Depth of groundwater table : Unknown e. Presence of surface stone and rock outcrops : No stones and no rocks f. Evidence of erosion : Slight sheet and rill erosion at site and in adjacent field g. Presence of salt or alkali: Soils free of excess salt or alkali h. Human influence : Five-year fallow field 195 III. Profile Description Horizon Depth (cm) Description A(p) 0-13/17 Brown to dark brown (7.5 Y R 4/4) moist and light brown (7.5 Y R 6/4) dry, clay loam; moderate fine and medium subangular blocky and moderate very fine and fine granular; slightly sticky, slightly plastic, friable when moist, slightly hard when dry; few gravel of moderate weathered shale; many horizontal roots of evergreen trees; clear and smooth boundary; pH 4.8. B A 13/17-29/32 Strong brown (7.5 Y R 5/6) moist, clay loam; moderate fine and medium subangular blocky; slightly sticky, slightly plastic, friable when moist; few gravel of moderate weathered shale; many horizontal roots of evergreen trees; clear and smooth boundary; pH 4.6. B(t)l 29/32-49/51 Yellowish red (5 Y R 4/6) moist, clay loam to clay; moderate fine and medium subangular blocky; sticky, slightly plastic to plastic, firm when moist; common gravel of moderate to strongly weathered shale; few horizontal roots of evergreen trees; clear and smooth boundary; pH 4.8. Bt2 49/51-90/95 Yellowish red (5 Y R 5/6) moist, gravelly clay; moderate very fine and fine subangular blocky; very sticky, plastic, firm to very firm when moist; many to abundant gravel of moderate to strongly weathered shale; few horizontal and oblique roots of evergreen trees; clear and wavy boundary; pH 4.9. Bt3 87/90-115+ Yellowish red (5 Y R 5/8) moist, gravelly clay; weak very fine and fine subangular blocky; very sticky, plastic, firm when moist; abundant gravel of moderate to strongly weathered shale; few horizontal roots of evergreen trees; pH 5.0. 196 6) Secondary Forest I. Information on the Site : a. Profile number : Secondary Forest b. Soil name : Unknown c. Higher category classification : USDA :? d. Date of examination : March 21 s t, 1999 e. Author : Wangpakapattanawong, P . , N . Anongrak and D. Saramathakul f. Location : Mae Hae Tai village, Mae Chaem District, Chiang Mai , Thailand (Grid Reference: 093373, Sheet 4645 TV) g. Elevation : 1,100 m asl. h. Land form : i . physiographic position : on ridge i i . surrounding land form : mountainous i i i . microtopography : nil i . Slope on which profile is sited : steep slope (45%), north 5 west aspect j . Vegetation and Land-use : Under hill evergreen forest k. Climate : Data derived from Mae Chaem Watershed Research Station, Royal Forest Department, 30 km of the site at the elevation of approximately 1,000 m asl. II. General Information on the So i l : a. Parent material : Apparently derived "in situ" from shale in Paleozoic era Carboniferous, Devonian and Silurian period) b. Drainage : Imperfectly drained to moderately well drained c. Moisture condition in profile : Moist below 40 cm d. Depth of groundwater table : Unknown e. Presence of surface stone and rock outcrops : Fairly stony and no rocks f. Evidence of erosion : Slight sheet and rill erosion at site and in adjacent field g. Presence of salt or alkali: Soils free of excess salt or alkali h. Human influence : Partially disturbed 197 III. Profile Description Horizon Depth (cm) Description A 0-15 Brown to dark brown (7.5 Y R 4/4) moist and strong brown (7.5 Y R 5/6) dry, clay loam; moderate medium and coarse granular; slightly sticky, slightly plastic, friable when moist, slightly hard when dry; few gravel of moderate weathered shale; many horizontal roots (1.0-2.0 cm.) of evergreen trees; abrupt and smooth boundary; pH 4.5. B A 15-40 Dark reddish brown (5 Y R 3/4) moist and reddish brown (5 Y R 4/4) dry, clay loam to clay; moderate fine and medium subangular blocky and moderate fine and medium granular; sticky, slightly plastic to plastic, friable to firm when moist, slightly hard when dry; few gravel of moderate weathered shale; many horizontal roots of evergreen trees; clear and smooth boundary; pH 4.4-4.8. B t l 40-60/62 Yellowish red (5 Y R 4/8) moist, clay; moderate fine and medium subangular blocky; sticky, plastic, firm when moist; common gravel of moderate weathered shale; few horizontal roots of evergreen trees; clear and wavy boundary; pH 4.8. Bt2 60/62-78/83 Red (2.5 Y R 4/6) moist, gravelly clay; moderate very fine and fine subangular blocky; sticky, plastic, firm when moist; many gravel (0.5-3 cm.) of slightly to moderate weathered shale; few oblique roots of evergreen trees; gradual and wavy boundary; pH 5.0. B C 78/83-112+ Red (2.5 Y R 4/6) moist, gravelly clay; moderate very fine and fine subangular blocky; sticky, slightly plastic to plastic, friable to firm when moist; abundant gravel (1-5 cm.) of slightly to moderate weathered shale; few oblique roots of evergreen trees; pH 5.0. Appendix 6 Comparison of Soil Properties Amongst Various Studies Table A.6.1 Comparison of soil pH (1:1 in water) amongst various studies in northern Thailand Field Depth This Sabhasri Nakano Andriesse Funakawa (cm) Study (1978) (1978) and Schelhaas etal. (1997a) (1998) (1987a) Crop Field 0-5 6.0 6.4 5.7-6.5 n.a. n.a. 5-15 5.7 5.5 -15-30 5.8 5.5 -1-year Fallow 0-5 n.a. n.a. 6.4 n.a. n.a. 5-15 -15-30 -2-year Fallow 0-5 5.0 n.a. n.a. n.a. n.a. 5-15 4.8 15-30 4.7 4-year Fallow 0-5 5.1 5.7 5.5 n.a. n.a. 5-15 4.7 5.7 -15-20 - - 5.2 15-30 4.6 5.8 -5-year Fallow 0-5 4.9 n.a. n.a. n.a. n.a. 5-15 4.7 15-30 4.6 7-year Fallow 0-5 n.a. 5.8 n.a. n.a. -5-15 5.5 -15-30 5.5 -0-10 - 5.2 10-20 - 5.1 20-30 - 5.2 Secondary 0-5 4.6 n.a. 5.2 5.2-5.7 n.a. Forest 5-10 - - 5.2-5.6 5-15 4.4 - -10-25 - - 5.2-5.5 15-30 4.4 _ n.a. = not available. 199 Table A.6.2 Comparison of S O M concentrations (%) amongst various studies in northern Thailand Field Depth This Sabhasri Nakano Andriesse Funakawa (cm) Study (1978) (1978) and Schelhaas et al. (1998) (1987a) (1997a and b) Crop Field 0-5 6.8 6.1 7.7-9.7 n.a. -0-10 - - - 5.1-5.7 5-15 4.4 2.4 - -15-30 2.7 1.6 - -1-year Fallow 0-5 n.a. n.a. 8.0 n.a. -0-10 - 2.0-2.7 5-15 - -15-30 - -2-year Fallow 0-5 5.8 n.a. n.a. n.a. n.a. 5-15 5.3 15-30 4.5 4-year Fallow 0-5 6.7 5.0 7.5 n.a. -0-10 4.4 - - 5.7-6.2 5-15 3.6 1.5 3.2 -15-30 1.0 - -5-year Fallow 0-5 6.5 n.a. n.a. n.a. n.a. 5-15 4.1 15-30 3.1 7-year Fallow 0-5 n.a. 5.4 n.a. n.a. -0-10 - 3.3-3.7 5-15 2.4 -15-30 1.7 -Secondary 0-5 7.9 n.a. 7.3 4.9-5.7 -Forest 0-10 - - - 5.0-6.6 5-10 - - 3.3-3.7 -5-15 4.8 - - -10-25 - - 2.3-2.4 -15-30 3.6 -n.a. = not available. 200 Table A.6.3 Comparison of soil total N concentrations (%)amongst various studies in northern Thailand Field Depth This Nakano Andriesse Funakawa (cm) Study (1978) and Schelhaas etal. (1998) (1987a) (1997a and b) Crop Field 0-5 0.31 0.34-0.38 n.a. -0-10 - - 0.23-0.26 5-15 0.44 - -15-30 0.13 - -1-year Fallow 0-5 n.a. 0.32 n.a. -0-10 - 0.07-0.11 2-year Fallow 0-5 0.26 n.a. n.a. n.a. 5-15 0.25 15-30 0.21 4-year Fallow 0-5 0.31 0.28 n.a. -0-10 - - 0.21-0.30 5-15 0.21 - -15-30 0.17 - -5-year Fallow 0-5 0.31 n.a. n.a. n.a. 5-15 0.20 15-30 0.14 7-year Fallow 0-10 n.a. n.a. n.a. 0.11-0.12 Secondary 0-5 0.37 0.22 0.15-0.17 -Forest 0-10 - - - 0.28-0.33 5-10 - - 0.10-0.12 -5-15 0.23 - - -10-25 - - 0.08-0.09 -15-30 0.17 - _ n.a. = not available. 201 Table A.6.4 Comparison of soil available P concentrations (ppm) amongst various studies in northern Thailand Field Depth This Study Sabhasri Nakano Andriesse (cm) (1998) (1978) (1978) and Schelhaas (Bray II) (Bray II) (Bray I) (1987a) (Bray II) Crop Field 0-5 21.5 18.5 7.6-9.4 n.a. 5-15 4.8 4.7 -15-30 3.3 3.4 -1-year Fallow 0-5 5-15 15-30 n.a. n.a. 8.0 n.a. 2-year Fallow 0-5 5-15 15-30 2.9 1.7 1.3 n.a. n.a. n.a. 4-year Fallow 0-5 2.5 3.7 7.4 n.a. 5-15 1.7 3.7 -15-20 - - -15-30 1.6 3.1 -5-year Fallow 0-5 5-15 15-30 2.9 1.3 1.1 n.a. n.a. n.a. 7-year Fallow 0-5 5-15 15-30 n.a. 4.6 5.8 4.2 n.a. n.a. Secondary 0-5 2.2 n.a. 14.6 11.3-29.7 Forest 5-10 - - 5.6-17.8 5-15 1.6 - -10-25 - - 2.9-5.9 15-30 1.6 n.a. = not available. 202 Table A.6.5 Comparison of soil extractable K concentrations (ppm) amongst various studies in northern Thailand Field Depth This Study Sabhasri Andriesse and Schelhaas (1987a) (cm) (1998) (1978) Crop Field 0-5 360 177 n.a. 5-15 292 114 15-30 313 107 2-year Fallow 0-5 141 - n.a. 5-15 100 -15-30 93 -4-year Fallow 0-5 195 127 n.a. 5-15 143 105 15-30 127 102 5-year Fallow 0-5 209 - n.a. 5-15 173 -15-30 150 -7-year Fallow 0-5 n.a. 152 n.a. 5-15 132 15-30 127 Secondary 0-5 193 n.a. 169-188 Forest 5-10 - 140-158 5-15 117 -10-25 - 126-152 15-30 96 n.a. = not available. 203 Table A.6.6 Comparison of soil extractable Ca concentrations (ppm) of the secondary forest site with another study in northern Thailand Depth (cm) This Study (1998) Andriesse and Schelhaas (1987a) 0-5 182 282-843 5-10 - 199-302 5-15 51 -10-25 - 180-250 15-30 42 -Table A.6.7 Comparison of soil extractable Mg concentrations (ppm) of the secondary forest site with another study in northern Thailand Depth (cm) This Study (1998) Andriesse and Schelhaas (1987a) 0-5 64 152-227 5-10 - 118-138 5-15 19 -10-25 - 96-111 15-30 15 -Table A.6.8 Comparison of soil CEC (cmofTkg) of the secondary forest site with another study in northern Thailand Depth (cm) This Study (1998) Andriesse and Schelhaas (1987a) 0-5 32.4 19.8-22.9 5-10 - 13.5-15.7 5-15 27.9 -10-25 - 12.0-13.0 15-30 29.1 -204 Appendix 7 Pot Experiment A.7.1 Introduction Shifting cultivators have been forced to reduce their fallow periods, e.g. from 10 to five or six years due to population pressure and land scarcity. This trend is likely to continue. As a result, it is crucial to know whether shorter-age of fallow fields will be able to support crop growth with and without external nutritional inputs. A.7.2 Methods The experiment design was factorial composing of soils from the 2- and 4-year fallow fields (in 1998) at two depths, which were 0-15 and 15-30 cm. Initial soil properties were reported in Chapter 3, and were shown here (Table A.7.1). The soils were then transferred into black plastic pots of 12 cm in diameter and 12 cm in height. Each pot was planted with 10 seeds of upland rice. Each set of the experiment was divided into four treatments of control, and fertilized with 21-0-0, 16-20-0, and 15-15-15 fertilizers, totaling of 30 pots. Initially, rice biomass and nutrient analyses were to be done three times, i.e. three pots for each sampling time, throughout the growing season from middle April to late October 1999. This was designed to examine biomass-and nutrient-accumulation patterns of the rice. However, the total number of crops for each treatment were reduce to five to seven due to crop damaged by cattle and over-saturated of soil moisture because the rainfall was rather high in 1999 and poor drainage in the pots. The fertilizers were applied two months after planting. The rice crop was harvested at the end of October 1999. As a whole, the grain production was very low, so the biomass was only divided into above- and below-ground components. The samples were air- and oven-dried, weighed, and 205 analyzed for N , P, K , Ca, and Mg. Due to limited oven capacity, a set of samples (4% of the total) was selected for oven-drying at 60 °C, and weighed to correct moisture content. Table A.7.1 Comparison of biophysical characteristics of the 2-year and 4-year fallow fields Parameter Depth 2-year Fallow 4-year Fallow (cm). (in 1998) (in 1998) Soil 0 - 5 pH (1:1 in water) 5.02 (0.07) 5.07 (0.05) p H ( l : l i n K C l ) 4.3 (0.1) 4.3 (0.1) Total O.M. (%) 5.8 (0.3) 6.7 (0.3)* Organic Carbon (%) 3.4 3.9* Total N (%) 0.26 (0.02) 0.31 (0.01)* Available P (ppm) 2.9 (0.4) 2.5 (0.2) Extractable K (ppm) 141 (10) 195(23)* CEC (cmol7kg) 26.7 (0.8) 24.1 (3.0) 5 - 1 5 pH (1:1 in water) 4.78 (0.06) 4.65 (0.04)* p H ( l : l i n K C l ) 4.0 (0.1) 3.9 (0.1) Total O.M. (%) 5.3 (0.5) 4.4 (0.2) Organic Carbon (%) 3.1 2.6 Total N (%) 0.25 (0.02) 0.21 (0.01) Available P (ppm) 1.7(0.1) 1.7(0.2) Extractable K (ppm) 100 (12) 143(19)* CEC (cmof/kg) 25.5 (2.1) 25.0 (0.9) 15-30 pH (1:1 in water) 4.67 (0.06) 4.59 (0.04) p H ( l : l i n K C l ) 3.9(0.1) 4.1 (0.1) Total O.M. (%) 4.5 (0.2) 3.6 (0.2)* Organic Carbon (%) 2.6 2.1* Total N (%) 0.21 (0.01) 0.17(0.01)* Available P (ppm) 1.3 (0.2) 1.6 (0.2) Extractable K (ppm) 93 (5.7) 127 (20) CEC (cmof/kg) 26.4 (0.8) 17.3 (1.5)* Vegetation Estimated Tree Biomass (t/ha) 1.4 6.7 Litterfall (t/ha) 4.0 (total) 5.9 (total) 3.8 (leaf) 4.9 (leaf) 1) Numbers in parentheses are standard errors (n=9). 2) Within each soil depth, asterisks indicate significant (P<0.10) differences between the two fallow fields. 206 A.7.3 Results Results from the pot experiment were inconclusive. The results of this experiment should be inferred with caution because the poor drainage condition might confound the actual affects of experimental treatments. The pot experiment results are presented in Tables A.7.2-A.7.7. Table A.7.2 Upland rice biomass (g/pot) of the pot experiment Soil Depth Treatment Number Shoot Response Root Response (cm) of Pots (g/pot) (%) (g/pot) (%) 2-yr Fallow 0-15 Control 6 2.1 (0.8) 0 0.6 (0.2) 0 21-0-0 4 2.1 (1.5) 4.1 1.0 (0.7) 60 16-20-0 6 6.0(1.5) 192 1.6 (0.5) 150 15-15-15 6 6.4 (2.3) 211 2.5 (1.4) 301 2-yr Fallow 15-30 Control 5 1.4 (0.3) 0 0.5 (0.1) 0 21-0-0 6 5.4 (0.9) 282 1.4 (0.3) 197 16-20-0 6 18(4.6) 1,150 5.1 (1.6) 949 15-15-15 5 8.5 (3.2) 499 2.7(1.2) 469 4-yr Fallow 0-15 Control 7 3.1 (0.7) 0 0.6 (0.2) 0 21-0-0 7 5.7 (1.3) 82 1.1 (0.4) 76 16-20-0 7 19 (3.7) 488 4.0 (0.6) 528 15-15-15 5 9.1 (1.6) 187 2.0 (0.5) 209 4-yr Fallow 15-30 Control 6 6.7 (2.8) 0 2.1 (1.2) 0 21-0-0 6 5.2 (2.0) -22 1.2 (0.4) -45 16-20-0 6 28 (6.8) 313 8.3 (2.0) 298 15-15-15 6 14(3.5) 103 5.5 (2.2) 166 1) Fertilizers were applied at 20 kg/ha when the rice was two months old. 2) Numbers in parentheses are standard errors. 207 Table A.7.3 Upland rice biomass (g/pot) of the control treatment of the pot experiment Field (in 1998) Soil Depth (cm) Number of Pots Shoot (PO.05) Root (PO.10) 2-yr Fallow 0-15 6 2.1 (0.8)b 0.6 (0.2)ab 2-yr Fallow 15-30 5 1.4 (0.3)b 0.5 (0.1)b 4-yr Fallow 0-15 7 3.2 (0.7)ab 0.6 (0.2)ab 4-yr Fallow 15-30 6 6.7 (2.8)a 2.1 (1.2)a 1) Numbers in parentheses are standard errors. 2) Within each biomass component, different letters indicate significant differences among experimental units. Table A.7.4 Upland rice biomass (g/pot) of the 21-0-0 fertilizer treatment of the pot experiment Field (in 1998) Soil Depth (cm) Number of Pots Shoot Root 2-yr Fallow 0-15 4 2.1 (1.5)a 1.0 (0.7)a 2-yr Fallow 15-30 6 5.4 (0.9)a 1.4 (0.3)a 4-yr Fallow 0-15 7 5.7(1.3)a 1.1 (0.4)a 4-yr Fallow 15-30 6 5.2 (2.0)a 1.2 (0.4)a 1) Numbers in parentheses are standard errors. 2) Within each biomass component, different letters indicate significant differences (P<0.1) among experimental units. 208 Table A.7.5 Upland rice biomass (g/pot) of the 16-20-0 fertilizer treatment of the pot experiment Field (in 1998) Soil Depth (cm) Number of Pots Shoot Root 2-yr Fallow 0-15 6 6.0(1.5)b 1.6 (0.5)b 2-yr Fallow 15-30 6 18 (4.6)ab 5.1 (1.6)ab 4-yr Fallow 0-15 7 19 (3.7)ab 4.0 (0.6)b 4-yr Fallow 15-30 6 28 (6.8)a 8.3 (2.0)a 1) Numbers in parentheses are standard errors 2) Within each biomass component, different letters indicate significant differences (P<0.05) among experimental units. Table A.7.6 Upland rice biomass (g/pot) of the 15-15-15 fertilizer treatment of the pot experiment Field (in 1998) Soil Depth (cm) Number of Pots Shoot Root 2-yr Fallow 0-15 6 6.4 (2.3)b 2.5(1.4)a 2-yr Fallow 15-30 5 8.5 (3.2)ab 2.7(1.2)a 4-yr Fallow 0-15 5 9.1 (1.6)ab 2.0 (0.5)a 4-yr Fallow 15-30 6 14 (3.5)a 5.5 (2.2)a 1) Numbers in parentheses are standard errors 2) Within each biomass component, different letters indicate significant differences (P<0.1) among experimental units. Table A.7.7 Nutrient concentrations (%) of the upland rice of the pot experiment 209 Field Soil Depth Treatment N P K Ca M g (in 1998) (cm) Shoot 2-yr Fallow 0-15 Control 1.6 0.05 0.8 0.005 0.06 21-0-0 1.7 0.05 1.0 0.003 0.03 16-20-0 2.3 0.17 1.1 0.003 0.08 15-15-15 1.8 0.15 1.4 0.003 0.07 2-yr Fallow 15-30 Control 1.2 0.06 0.9 0.005 0.06 21-0-0 2.0 0.06 1.4 0.013 0.06 16-20-0 1.5 0.11 1.5 0.005 0.07 15-15-15 1.5 0.11 1.5 0.005 0.05 4-yr Fallow 0-15 Control 1.4 0.06 1.0 0.013 0.09 21-0-0 1.8 0.05 1.2 0.013 0.06 16-20-0 1.6 0.14 1.2 0.025 0.11 15-15-15 1.8 0.18 1.4 0.025 0.07 4-yr Fallow 15-30 Control 1.1 0.07 1.4 0.013 0.09 21-0-0 1.7 0.05 1.2 0.005 0.06 16-20-0 1.4 0.12 1.6 0.003 0.11 15-15-15 1.2 0.13 1.3 0.005 0.09 Root 2-yr Fallow 0-15 Control 0.9 0.07 0.8 Trace 0.04 21-0-0 1.3 0.06 0.7 Trace 0.04 16-20-0 1.3 0.12 1.0 Trace 0.05 15-15-15 1.0 0.08 0.8 Trace 0.04 2-yr Fallow 15-30 Control 0.7 0.06 0.9 0.003 0.05 21-0-0 1.2 0.05 0.9 0.003 0.04 16-20-0 1.1 0.09 0.8 0.003 0.04 15-15-15 1.0 0.13 1.1 0.005 0.05 4-yr Fallow 0-15 Control 1.0 0.07 1.0 0.003 0.05 21-0-0 1.1 0.05 0.8 0.005 0.04 16-20-0 1.2 0.11 0.7 0.003 0.05 15-15-15 1.3 0.17 1.2 0.006 0.06 4-yr Fallow 15-30 Control 0.7 0.07 0.7 Trace 0.05 21-0-0 1.2 0.12 0.7 Trace 0.04 16-20-0 0.9 0.11 1.0 Trace 0.04 15-15-15 0.8 0.11 1.0 Trace 0.05 A l l replications for each treatment were pulled before the chemical analyses, therefore, statistical analyses could not be done. Appendix 8 Wood Density Wood density (kg/m ) was used to convert slash volume (kg/m ) estimated by the line intersect method into biomass (kg/ha) in Chapter 5. Wood density from published articles of species found elsewhere in the tropics similar to those found in the fallow fields and the forests in my study was used (Table A.8.1). Table A.8.1 Wood density of various tropical tree species Source Tree Species Wood Density (kg/W) Kayama(1978) Calophyllum sp. 479 Cratoxylum sp. 352 Quercus spp. 797 Shorea spp. 721 660 828 824 579 Terminalia sp. 388 340 Tristania sp. 893 Martawijaya and Sumarni (1978) Ilex sp. 570 Lagerstroemia sp. 860 Shima wallichii 700 Mean 642 Standard Error 52 211 Appendix 9 DBH Distribution and Aboveground Biomass of Fallow Trees Table A.9.1 Numbers of trees in each D B H class in shifting cultivation fields (0.04 ha) Field 3-5 5-9 10-14 15-19 20-24 25-29 30-34 35-39 (in 1998) cm cm cm cm cm cm cm cm 2-year Fallow 45 7 0 0 0 0 0 0 4-year Fallow- 80 22 1 0 0 0 0 0 N W 5-year Fallow 95 41 1 1 2 1 1 1 Secondary 46 34 11 3 1 6 0 0 Forest Burial Forest 55 34 12 4 4 4 1 1 Table A.9.2 Fresh aboveground tree biomass (kg/0.04 ha) in each D B H class in shifting cultivation fields Field 5-9 10-14 15-19 20-24 25-29 30-34 35-39 Total (in 1998) cm cm cm cm cm cm cm 2-year 57 0 0 0 0 0 0 57 Fallow (100) (100) 4-year Fallow- 237 30 0 0 0 0 0 267 N W (89) (11) (100) 5-year 430 26 72 351 236 482 554 2,151 Fallow (20) (1) (3) (16) (11) (22) (26) (100) Secondary 375 399 304 208 1,876 0 0 3,162 Forest (12) (13) (10) (7) (59) (100) Burial 422 494 405 718 1185 441 682 4,347 Forest (10) (11) (9) (17) (27) (10) (16) (100) Numbers in parentheses are percentage of the total aboveground biomass in each D B H class. 212 Appendix 10 Farmer's Informal Interviews about Forest-Fallow Shifting Cultivation Management A. 10.1 Introduction Agricultural researchers in Thailand have been concerned about the relatively low yields of upland rice from shifting cultivation system, and have been doing research on yield improvement. However, the average yields of upland rice (Chapter 4) reported for northern Thailand (955 and 1,032 kg/ha in 1967 for the Lua and the Karen, respectively; Kunstadter 1978), for northern Laos (1,100 kg/ha in 1991; Roder et al. 1997b), and the values I found in this study (1,126 and 902 kg/ha from two Karen villages in 1998 and 1999, respectively) suggest little change over the past two decades. In a purely ecological sense, the better the crop growth and yield, the greater the nutrients the crops require from the soils and external inputs, and the greater the export of nutrients in harvested biomass. The low yields in the study area thus suggest modest nutrient demands on the soils. In addition, even though the overall rice yield is low, the farmers benefit from other crops (e.g. sweet potato, cucumber, gourd, and onion) that are grown in sparse mixture with the upland rice. The fast-paced expansion of the road network into the region occupied by the Karen has led to rapid transfers of new farming technology into the area, and increased erosion problems. Use of chemical fertilizers has become a common practice where commercial fixed-field farming is 213 practiced, but is still unaffordable in most shifting-cultivation villages. The rapid transfers of farming technology and increasing market accessibility have affected the way the younger farmers (20 to 30 years old) farm. Modern farming technology (the Green Revolution) has helped to increase crop production around the world by improving the genetic characteristics of crops and by increasing external chemical inputs. This arguably has led to an improvement in food supplies in various parts of the world. However, recently there have been concerns about excessive usage of chemicals in agriculture, and the possible ecological effects of genetically modified crops on the environment. A.10.2 Objectives and Rationale The objectives of this part of the thesis were to gather and analyze indigenous knowledge of the Karen about shifting cultivation management, and to seek experience-based knowledge concerning the sustainability of their agricultural system. For the Karen of Mae Hae Tai and other villages, places are recognized by names that are associated, one way or another, with nature. Each of their shifting-cultivation fields has a name that has a descriptive meaning; for example, "the red creek field", and "the field with a big-tree on the mountain ridge". It is evident that the villagers connect to their natural surroundings. In 214 addition, their farming (i.e. shifting cultivation) in remote areas is practiced without any chemical inputs, and weeds are eradicated, and rice is harvested and threshed manually. Without modern technical aids to agricultural production, the Karen produce subsistence-level food crops in the environment where they live. Therefore, it is logical to document indigenous knowledge that the Karen have about shifting cultivation management, as this system has been providing the villagers with stable yields over many centuries. Walker et al. (1995) stated that even though scientific synthesis of the indigenous ecological knowledge has a potentially important role to play, it has received inadequate attention. A . 10.3 Methods To achieve the objective i f this sub-project, I gathered and analyzed traditional knowledge of the Karen with respect to the characteristics of fertile soils and the role of vegetation in maintenance and restoration of soil fertility. Originally, a structured interview questionnaire was prepared that was to be used with sufficient numbers of farmers to get statistically analyzable data. However, due to time limitations, the questionnaire was only presented to 15 individuals, including leaders in the village (the current and former village headmen, the deputy village headman), some elders, and a schoolteacher. It was assumed that the individuals that were interviewed would have the greatest accumulated experience, but would also adequately represent younger farmers. A l l the questions were asked in an informal conservation format. 215 Interview Questions 1. What was the upland-rice planting area in 1999, and what was the yield? 2. What was the paddy-rice planting area in 1999, and what was the yield? 3. Why is rice planted for only one year, with no consecutive-year cropping? 4. What are characteristics of good-quality soils? 5. What are characteristics of poor-quality soils? 6. Is C. odorata good or bad for the soils? Why or why not? 7 . Which is the field with the best soil quality? What are the indicators? 8. Which is the field with the poorest soil quality? What are the indicators? 9. Why are some trees left standing in the field? 10. Are fertilizers used? Why or why not? 11. Are herbicides used? Why or why not? 12. How many times weeding is done in one growing season? 13. What is the shortest acceptable fallow period? The farmers use the volume of a bucket as their yield measurement method, and they approximately equal one bucket to 13-14 kg of rice. Therefore, a value of 13.5 kg/bucket was used to convert their reported yields to kg/ha. 216 A. 10.4 Results and Discussion A.10.4.1 Farmers' Responses a) Mr. Decha, the former village headman, 52 years old He stated that growing upland rice without fertilizer can yield about 1,600 kg/ha, and that fertilization can improve the yield by as much as two fold. The NP 16-20-0 fertilizer is the most commonly used fertilizer because it is the cheapest. A few farmers plant cabbage in the first year follow by upland rice in a second year of cropping. They apply the 16-20-0 fertilizer to the cabbage crops, and claim that the residual effect of the first year fertilizer improves the rice yield without re- fertilization. The field with the best soil is the 3-year fallow field (in 1998), which wil l be cleared and cropped again in 2001, while the rest of the fields have lower soil fertility, but little variation between the fields. In 1999, he grew upland rice on a total area of 1.28 ha, which yielded about 2,430 kg (1,900 kg/ha) without fertilizer. Good soils are those with dark color and are rather clayey and shiny. Bad soils are dry and gravelly. He has been seeing C. odorata (formerly Euphatorium odoratum) in the village since he was a child. Euphatorium adenoforum is another species of this genus that is common in the village vicinity. He claimed that about 30 years ago, seeds of E. adenoforum were aerially dispersed in the area by some kinds of airplane, and it became common since then. E. adenoforum is also called the Communist Grass or the Black C. odorata. Like many of the 217 farmers, he believes that C. odorata is related to good rice yield. Many farmers believe that the bigger the C. odorata in the fallow before burning, the better the rice yield. Cycas, i f present, is not good for the soils. He does not grow rice for a second year because weeds become more prominent after the first year of harvest, and there will be no fertilizer effect from the ash for the rice. This wil l result in slow growth and low yield. Herbicide is not used. Weeds are manually eradicated three to four times within one growing season. Some trees are left standing and trimmed, and some are kept about 1-m in height to keep soils in good structure. Earthworms are present in good soils. Second to upland rice, corn is the most common crop species grown in this system. Other species planted along with rice are sweet potato, cucumber, gourd, and onion. However, the amounts planted are rather small. For example, a fistful of seeds of these species is mixed with about 40 kg of rice seeds before planting. In his opinion, the best fallow period is seven to eight years, and the shortest viable fallow period is two years (i.e. a 3-year cycle). Decaying leaf litterfall in fallow fields transforms into soils, but the fresh leaf litterfall after slashing for crop production needs to be burnt before it can be utilized by plants. If the fresh leaf litter is incompletely burned, it wil l turn into a termite mound, which is not good for the rice. The burning can provide good rice yield, suppress weed growth, and ki l l insects. 218 b) Mr. Lele, the village headman, 36 years old Most of his responses were similar to those of the former headman. In 1999, his family grew about 0.5 ha of upland rice, which provided 810 kg of grain (1,620 kg/ha). He also has 1.28 of paddy rice-fields, which gave 5,400 kg of grain (4,218 kg/ha) in 1999. The main reason that he does not grow upland rice for a second year is that there will be more weeds. C. odorata is one of the indicators of good soils, and Cycas is the opposite. He thinks that the minimum fallow period viable and sustainable for the village's rotational farming is three years. c) Mr. Kawa, the deputy headman, 36 years old Mr. Kawa is one of the most active individuals in the village. He has been involved in many activities, including the Village Volunteer program, and is the coordinator for various natural resources management projects in cooperation with CARE-Thailand. He was also among the first farmers to initiate rotational cropping of cabbage followed by upland rice within fixed-fields. One of the interesting features of this type of farming is that the farmers take advantage of the residual effect of fertilizer applied in year-one cropping to improve rice production in year-two cropping. The cropped field is then left to fallow. This change in farming practice was a result of a recent change in land ownership. In general, villagers have no legal ownership of the land they farm; it belongs to the government. In addition, the area of Mae Hae Tai was recently included within the boundaries of the newly proposed Mae Tho National Park. This has 219 complicated matters regarding rotational farming or shifting cultivation practices by the Karen. The government, via the Royal Forest Department (RFD), has required that the villagers must demonstrate that they can exist within the national park boundaries without the practice of shifting cultivation, which the RFD views as an environmentally degrading activity. The RFD has been investigating how to sustain the villages within national parks for a number of years. However, some NGOs are arguing that shifting cultivation is a part of the Karen culture. This debate is ongoing. In 1999, Mr. Kawa got about 810 kg of upland rice grain from about 0.5 ha of land (1,620 kg/ha). He described how one of the farmers tried to grow rice in a second and third consecutive year. It resulted in lower yield in the second and no grain yield at all in the third year. Mr. Kawa believes that good soils are dark in color. Bad soils are red and sandy with gravels. Having C. odorata results in fertile soils and good rice yields. When asked about terracing and non-terraced fields, he believes that the non-terraced field is better in terms of soil fertility because topsoils are often buried by sub-soil materials in process of the terrace construction. d) Mr. Somsak, 23 years old In 1999, he grew about 0.5 ha of upland rice, but he did not know the exact yield. He also has 0.50 ha of paddy rice-field. He believes that it is not worth growing rice for the second year 220 because the soils wil l be too infertile for rice growing. Good soils are those with dark color. C. odorata is beneficial for both rice growth and the soils. The field with the best quality soils is the 3-year fallow (in 1998), which wil l be used in the 2001 growing season. Fertilizers and herbicides are not used. If fertilizer is to be used, 16-20-0 is the best one. In general, this person appeared to know about traditional knowledge only superficially. This might be related to his relatively young age. e) Mr. Preecha, 30 years old In 1999, he got about 1,620 kg of upland rice grain from 0.65 ha of rice planting (2,492 kg/ha). His reasons for not planting rice for the second year were poor rice yield and increasing weeds. Good soils are dark in color. Bad soils normally have high amounts of gravel and consist of red dusty particles. He has no knowledge about C. odorata. The field related to the best rice yields is the 3-year fallow (in 1998), in which all the villagers can obtain the highest yields regardless of the locations within the field. For the other fields, only some villagers can obtain good rice yields because soil fertility is rather patchy in those fields. Neither fertilizer nor herbicide is used. f) Mr. Pasi, > 60 years old He said that the fallow period used to be seven to eight years instead of six to seven years as at present, which suggests that the decrease of fallow period is not recent. In 1999, he had 0.65 ha 221 of upland rice, which gave about 1,350 kg of grain (2,077 kg/ha). There is no second-year planting because of increasing weeds. Good soils are those with no gravelly and stony components. Bad soils are the opposite. In addition, he has seen Imperata cylindrica (which renders soils infertile) for a long time, and he added that it is very difficult to eradicate. However, I. cylindrica is not commonly found in the area (my personal observation). In general, big trees render soils fertile, as does the Communist Grass (E. adenoforum). Neither fertilizer nor herbicide is used because it is too expensive to purchase. He would buy fertilizer i f he had enough money, and his answers regarding the possible use of fertilizer were the same as for the other farmers. He weeds three to four times a year. The most fertile soils are at the 3-year fallow field (in 1998). g) Mr. Salu, > 60 years old The main reason for practicing rotational farming is that they get natural fertilizer from burned slash, and he believes that five to seven years of fallow is adequate for good grain yields. The continuous planting of rice is not practiced because of poor yield in the following years. In 1999, a grain yield of 1,350 kg was gained from 0.8 ha of upland rice (1,688 kg/ha). He is not certain about properties that would define good and bad soils. He uses neither fertilizer nor herbicide. He weeds three to four times a year. 222 h) Mr. Surachai, a local school teacher, > 40 years old He claimed to be the first one who initiated the practice of crop (cabbage and rice) rotation in fixed fields. He raised the issue at the urging of the government to replace shifting cultivation with permanent agriculture. He divided his plot of land of 3.2 ha into five pieces of approximately 0.6 ha each. Each of these five subplots was further divided it into two halves. The first half is planted with cabbage, and 16-20-0 fertilizer is applied once within a 3-month-long cropping period. He claimed that the other cabbage-growing hill tribes such as Hmong apply fertilizer at least three times in one crop. He also applies insecticide once in the growing season, but no herbicide. Upland rice is planted in the following year in the post-cabbage half, and it can utilize fertilizer residue supposedly unused by the cabbage. Mr. Kawa, the deputy headman, is another person who is practicing this new type of farming, along with four other farmers out of the 57 households. The teacher thinks that some other farmers want to try this new way of farming but it is impractical for them. This is because the farming areas are normally family owned, which makes it difficult for families with many members to decide what to do, so the most common way is to continue traditional shifting cultivation practices. It is interesting that this new type of farming does not involve burning. I think that this new farming practice is feasible for him because he is a teacher earning a stable monthly salary, so he is less worried about making a living than the common villagers who depend for their livelihood mainly on rice production from shifting cultivation. 223 i) Mr. Subur, 40 years old The 3-year fallow field (in 1998), which will be cropped in 2001, has the best soils. Poor soils are red and hard. If they continue growing rice for the second year, there would be too much weed growth (grasses). He believes that, unlike the 3- and 4-year fallow fields, 1- and 2-year fallows would not provide sufficient plant biomass to give adequate ash from burning, which is very important for good rice growing. j) Mr. Pakaew, 50 years old In 1999, he planted 0.8 ha of upland rice in a field that he believes to have poor soil-quality. The field with the best soil quality is the 3-year fallow (in 1998) because it has C. odorata, and Schima wallichii. He does not use any fertilizer. The weeding is done manually. He plants rice for only one year because the second-year planting results in excessive weeds. The secondary forest has the poorest soil quality for farming. k) Mr. Su Kae, 36 years old He planted 0.5 ha of upland rice in 1999, and the yield was 1,620 kg (3,240 kg/ha). He and other villagers claimed that the amount of rainfall was unusually high in 1999, and the rain extended into the harvesting time in early November, which was supposed to be dry. This resulted in excessive moisture causing the harvested grain to germinate in the field, and subsequently poor 224 yields. In fact, the annual rainfall was 1,342 mm in 1999, and it was 856 mm in 1998, while the 15-year average annual rainfall was 1,214 mm (RFD, 1999), so 1999 was only slightly above average. He believes that the 3-year fallow (in 1998) is the field with the best soils; they are dark in color, with no gravels, also with the presence of earthworms. C. odorata is good for the soils. Chemical inputs are not used. He said that a few farmers started to use herbicide about two years ago. However, he believes that using herbicide can have harmful side effects on the rice (e.g. dehydrating the rice). 1) Mr. Kawae, 40 years old He only has 0.32 ha of paddy rice, and is therefore not involved in upland rice farming. However, he had some opinions about this system. The 3-year fallow (in 1998) has the best soils. Most of the grasses are not good for the soils. He mentioned that some old farmers separate good and bad soils by tasting. It is considered good soils i f they are sweet and sour in taste. Another indicator of good soils is its surface. If the surface is coated by green mosses, it is considered to be good soil. He never grew rice for the second year because there is inadequate amount of ash. Most trees in the fallow fields can sprout. m) Mr. Tenu, age unavailable He had 0.32 ha of upland rice in 1999, and got about 270 kg of grain (844 kg/ha). Good soils, 225 which have dark and dark-reddish colors, result in good rice yields. Earthworms and their excretion, and bryophytes are indicators of good soils. High-density tree cover can protect surface soils from erosion, but the farmers utilize runoff to carry sediment from uplands down to valley-bottom paddy rice fields. Such erosion is greater in young fallow fields with few standing trees in comparison to older fallow fields and secondary forest. The highest erosion occurs after the first rain following the burning, and it washes away some of the ash. He notes that having terraced rice fields would require a high water input. The 3-year fallow field (in 1998) has good soils resulting in good rice yields. n) Mr. Dechi, 26 years old He believes that continuous cropping results in excessive weeds. Good soils are dark in color and have earthworms. Ash from slash burning can be a good fertilizer. He believes that the ash is beneficial for the soils because it is salty in taste. One of the benefits of burning is killing weeds. He does the weeding five times for one growing season in May, June, July, August, and September, which is more frequent than the other farmers who weed about three times. The farmers do the weeding by using a knife because it is faster than pulling. Besides, pulling normally kills the weeds leaving no rootstocks to re-emerge back in the fallow periods, while clearing by the knife only temporarily suppress the weeds during the growing season. 226 o) Mr. Keka, 27 years old His family has eight members, so his rice growing area is rather large (2.4 ha). In 1999, the family obtained about 2,700 kg of upland rice (1,125 kg/ha). They also possess 0.32 ha of paddy rice-field, which yielded about 810 kg (2,531 kg/ha) in 1999. Normally, the headman chooses and announces which field is to be cleared and cropped. However, it is a common knowledge among the villagers of where they will be cropping each year. He also believes that C. odorata is good for the soils. In his opinion, the 3-year fallow field (in 1998) has the best soils. Paddy rice receives natural fertilizer from sediment in runoff water. Therefore, having terraces in upper slope position also results negatively in less sediment, which is supposed to be beneficial to paddy rice. A.10.4.2 Summary of the Farmers' Common Responses 1. The shifting cultivation field with the best soil quality is the 3-year fallow field (in 1998), which wil l be cleared and cropped in 2001. 2. Good-quality soils are rather clayey, with dark color, and small amounts or no gravel. Poor-quality soils are the opposite, i.e. sandy, with red color, and gravely. 3. Earthworms and mosses are also indicators of good-quality soils. 4. Soil fertility is highly variable spatially within a field and this is reflected in the high variability in rice reported by the different farmers. In 1999, the grain yield (all from the 227 same field) ranged from 844 to 3,240 kg/ha with an average of 1,845 kg/ha (S.E. = 238, n=9). 5. Fertilizer is not commonly used for the rice crop because it is unaffordable. 6. The main reasons for growing rice for only one crop are reported to the increasing weeds and decreasing productivity because of declining soil fertility. 7. C. odorata is good for soil fertility and rice productivity. 8. The major benefit from burning is the fertilizer effect of wood ash. A.10.5 Analysis of the Karen Knowledge on Shifting Cultivation Management In general, most farmers independently gave the same or similar answers to the questions. The following discusses the responses to each of the major topics covered in the questions A.10.5.1 Roles of the Fallow The farmers only grow rice for one year because; 1) the increasing weeds in the following years result in a higher labor demand for weeding, and 2) poor rice yields. One of the reasons given to explain why rice yields are reduced in the second and subsequent years is the lack of ash fertilizer in the subsequent years. A question was asked as to which of 1) increasing the cropping period, or 2) decreasing the fallow period, was more preferable i f the area of land under cultivation needed to be reduced because of population growth. Most farmers responded that 228 increasing the cropping period would result in lower yields in subsequent years, while decreasing the fallow period would have a less dramatic negative result. When asked what is the shortest fallow period acceptable, the answers varied slightly among the farmers. Most responded that the most important role of the fallow is to allow the fallow species to grow until they provide sufficiently large biomass to produce enough ash fertilizer for rice growing. The former headman also mentioned that the fallow vegetation could help to keep good soil structure. The shortest fallow period that was thought to acceptable by the farmers was two to three years. At present, the fallow period is five years, of which about 10 t/ha of tree biomass is slashed and burned at the end of the fallow (Table 5.2, Chapter 5). The farmers perceived that the shortest viable fallow period was two or three years. A.10.5.2 Soil Fertility According to all the farmers interviewed, the field with the best soils is the 3-year fallow field (in 1998), which wil l be cleared and cropped in 2001. The other fields have lower soil quality with little variation between fields. One farmer stated that one good rice yield could be obtained from this field (the 3-year fallow field in 1998) no matter where within the field he farms, which is apparently not true for the other fields in which productivity is very variable. However, this best field was not included in the soil sampling (chronosequence) of this thesis work because it was 229 not a legitimate part of the chronosequence (because of its higher fertility), so no data are available on which to explain the farmers' experience. Good soils are dark in color and clayey, with no gravels. The presence of earthworms is another indicator of good soils. Bad soils are red in color and sandy. These indicators of soil quality were cited by all the farmers. A.10.5.3 Roles of Trees and Litterfall The village headman raised one issue concerning the role of standing trees in fallow fields, which is to sustain the soil structure. Other farmers indirectly cited a contribution of the fallow trees in the form of litter, which decomposes to be natural fertilizers. Therefore, the farmers assert that the two main functions of the fallow trees with regard to site fertility and productivity: maintain physical characteristics of soil, and maintain soil chemical characteristics. The lack of vegetation cover increases the impacts of sunlight and rain on the soil, which makes the soils become more compact, i.e. higher bulk density. Once the fallow trees are cut, their belowground portions are still intact, and the continuing belowground activity should be beneficial for the maintenance of low bulk density and of soil structure. 230 The nutritional contribution of litter by the fallow trees is similar to mulching in other agroforestry systems. In farming with mulching, litter decomposition rates and nutrient releasing patterns play a very important role in providing plants with available nutrients. The concept of synchronization of nutrients released in mulching with nutrient uptake demands of crop plants emphasizes the importance of the timing of mulching for crop productivity. However, at the end of the fallow, the slash and litter are burned to provide ash fertilizer. Therefore, it is assumed that the total nutrient accumulation in litter rather than its decomposition rate that is more important in the shifting cultivation system. Pampasit (1998) studied the nutrient composition of different components of living trees of a hill-evergreen forest in northern Thailand that was similar to the secondary forest of Mae Hae Tai village. There were 10 tree genera that were found in both my 5-year fallow field (in 1998) and the study by Pampasit (1998). Of these 10 genera, the average nutrient concentrations in their fresh green leaves was 45.4, 1.9, 0.6, 1.0, 0.6, and 0.2% for C, N , P, K , Ca, and Mg, respectively (Table 5.8, Chapter 5; Pampasit 1998). As a result, the litter contributes substantial amounts of C and N , which are incorporated into soil organic matter during the fallow period. The last important contribution of the fallow trees is to provide food and medicinal products. The fallow consists of various tree species. For example, there were 40 different tree species in the 231 40x40 m area of the 5-year fallow field (in 1998). This contribution has been used as one of the arguments of the pro-shifting-cultivation groups, which claim that changing from the current practice to fixed-field agriculture may cause losses in useful plants that are found exclusively in the fallow. A number of ICRAF studies are being undertaken to investigate this issue. Schmidt-Vogt (1999) reported that the Karen gave no specific reasons for keeping the fallow trees in the shifting cultivation fields. However, Mischung (1990 cited in Schmidt-Vogt 1999) stated that trees are left in the fields to promote forest regeneration from seeds. However, this should only apply to a small number of species because most of the common species in the fallow fields of Mae Hae Tai village are able to regenerate vegetatively by sprouting. A.10.5.4 Roles of C. odorata A l l the farmers interviewed believe that C. odorata is good for soils, and that its presence results in good rice yields. The bigger the biomass of C. odorata, the better the rice yields. The biomass data in my study indicated that C. odorata produced about 7 t/ha of aboveground biomass after one growing season. The farmers do the last of three weedings about one month before rice harvesting because they believe that it is useless to weed again even though some weed species that re-establish after the 232 third weeding still continue to grow. They allow the weeds to continue growing continuously into the fallow period because there are no farming activities in the cropped field after rice threshing. Most of the "weed" species are desired fallow species that only constitute "weeds" during the rice growing season; they turn into fallow species during the later part of shifting cultivation cycle. Some "weed" species found in both crop and fallow fields are used as food or medicines. Schmidt-Vogt (1999) studied different types of shifting cultivation practices in northern Thailand and pointed out that what appears to outsiders to be degraded forests in fact contain a higher number of species than previously thought. This can be viewed in terms of plant succession. In general, the number of all plant species found in early successional stages (e.g. the young fallow fields) is higher than the later (e.g. the older fallow fields and secondary forests). A.10.5.5 Farming Technology Generally, no external chemical inputs such as fertilizer and herbicide are commonly used in upland rice farming. However, a number of farmers said that i f they had enough money, they would buy fertilizer; they would choose the 16-20-0 fertilizer because it is the most inexpensive one. This means that the farmers realize the benefits of using chemical fertilizer in yield improvement, and the experiments in this thesis show that the rice responded positively to added 233 fertilizers. Therefore, the lack of use of fertilizer is more of a socio-economic consideration than an agricultural one per se. However, some old farmers mentioned that litter in the fallow is the best source of fertilizer. Some farmers have begun to use herbicides in their upland rice-fields, but again it is considered too expensive by most farmers. One farmer noticed that broadcast application of herbicide could also negatively affect the rice by dehydrating and then suppressing the rice growth. Some farmers grow cabbage cash crops with the use of fertilizer (16-20-0), and the residual nutritional effects improve rice production, i f it is grown after the cabbage as is practiced by a number of the farmers. Terracing is not thought to be as beneficial as some development projects believe it to be. First of all, terrace construction is intensively labor demanding. For Mae Hae Tai village, terraces of the upland rice fields were constructed by the US AID funded project called Mae Chaem Watershed Development project. The villagers were benefited by receiving wages for terrace construction. However, they haven't maintained the terraces for several reasons; 1) top soils, which are very important for crops, were buried during the construction under infertile sub-soils, 2) having paddy rice on steep slopes requires a large amount of water, but their farming system is rainfed and rainfall can be irregular from year to year, and 3) the terraces act as a trap for nutrients, 234 which previously flowed down slope to nourish paddy rice at the valley bottom. A.10.6 Summary A number of the farmers' responses to the interview questions concerning their farming practices were explainable in an ecological context; for example, the variability in crop yields which reflects variability of soil fertility, which in turn depends on localized topography and physical characteristics of the soils. Some of the responses were intriguing and deserved further attention, such as the development of termite mounds as a consequence of incomplete burning of slash. The interviews raised a number of interesting differences between the knowledge of young and old farmers regarding shifting cultivation management. It is thought that this reflects changes in farming practices by younger farmers. Because changed systems do not have years of experience to draw on, studies are urgently needed to evaluate the nutritional, yield and sustainability aspects of these changed systems. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0090492/manifest

Comment

Related Items