UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Sustainability of Leucaena leucocephala fallows in shifting cultivation on the Island of Mindoro, Philippines MacDicken, Kenneth G. 1994

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

Item Metadata

Download

Media
831-ubc_1995-983833.pdf [ 7.72MB ]
Metadata
JSON: 831-1.0088873.json
JSON-LD: 831-1.0088873-ld.json
RDF/XML (Pretty): 831-1.0088873-rdf.xml
RDF/JSON: 831-1.0088873-rdf.json
Turtle: 831-1.0088873-turtle.txt
N-Triples: 831-1.0088873-rdf-ntriples.txt
Original Record: 831-1.0088873-source.json
Full Text
831-1.0088873-fulltext.txt
Citation
831-1.0088873.ris

Full Text

SUSTAINABILITY OF LEUCAENA LEUCOCEPHALA FALLOWS IN SHIFTING CULTIVATION ON THE ISLAND OF MINDORO, PHILIPPINES by KENNETH G. MACDICKEN B.A., Washington State University, USA, 1980 M.S., The University of Hawaii, USA, 1983 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES Department of Forestry  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA October 1994 ©Kenneth Glenn MacDicken, 1994  In presenting this thesis in partial fulfillment 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 Forestry The University of British Columbia 1956 Main Mall Vancouver, Canada Date  2  dvttiu*- [<{<\L{  Abstract Increasing human populations, declining crop yields and reductions in per capita arable land area suggest that shifting cultivation as traditionally practiced in much of the humid tropics is no longer a sustainable farming practice. The use of nitrogen-fixing trees such as Leucaena leucocephala (Lam. de Wit) (leucaena) as a fallow species may be an important strategy to maintain or improve crop yields in shifting cultivation where natural resources are rapidly declining.  A series of experiments comparing leucaena and non-leucaena fallows was conducted on sites with fallows established between 1977 and 1992 on the Island of Mindoro, Philippines to determine the impacts of leucaena fallows on rice yields, soil nutrient status and the sustainability of production in shifting cultivation in the study area. Rice yields and yield components were measured on 10 shifting cultivation fields. One burning experiment and a series of chronosequence studies in fallows of 1-4 years of age were conducted. Informal interviews with farmers were also conducted to determine perceptions and social impacts of leucaena as a fallow crop.  Moisture content-adjusted grain yields (@0.14 g H20/g dry matter) were 3.8 t ha"1 following leucaena fallows and 2.71 ha"1 following non-leucaena fallows for an average yield response of 42%. This difference appears to relate most closely to soil N accumulated during the fallow period and N added through seedling biomass applied as a surface mulch during the cropping period. N-fixation of leucaena and possibly the inhibition of ii  nitrification of ammonium ions as demonstrated by higher NH4+ concentrations in soils under leucaena were associated with (and may explain) these differences. Higher N levels appear to delay crop maturity, as evidenced by higher grain moisture contents in the leucaena treatment.  Available phosphorus levels are low, but do not appear to differ between fallow types, although the size of phosphorus pools in the biomass and litter varied with fallow type. Ca and Mg were not limiting in either fallow type, due to the large soil pool of these nutrients. Organic phosphorus inputs to the rice crop are higher in the leucaena fallows when leucaena wood removals from the site are low. The greatest potential threat to sustainability of crop production following leucaena fallows is charcoal making and the potential losses of P from the system.  There appears to be little disadvantage to burning leucaena fallow fields prior to planting. The most significant advantages of burning appear to be increased P availability and reduced weeding costs, although no difference in grain yield due to burning was detected.  Leucaena can increase sustainability of shifting cultivation through higher N and P contributions that result in grain yield increases. Additional increases in yield are probably attainable with the use of a minimum fallow age of 3-4 years, timely weed control and use of high-yielding, traditional rice varieties.  iii  Table of Contents Abstract Table of Contents List of Tables List of Figures Acknowledgements  ii iv vii x xii  Chapter 1 Introduction The changing context of shifting cultivation Research objectives Organization of this thesis  1 4 5  Chapter 2 Shifting cultivation near Sto. Tomas and elsewhere in the humid tropics Shifting cultivation as practiced by the Iraya Mangyans Recent changes in the Sto. Tomas study area Soil changes under shifting cultivation Leucaena as a fallow improvement species The effects of burning  6 10 13 16 19  Chapter 3 Site description and research approach Climate Soils Vegetation General research approach Data management and analysis Limitations of the data  21 24 27 27 34 35  Chapter 4 Soil properties under leucaena and non-leucaena fallows Introduction Methods Results and discussion Soils at the time of rice harvest Geographic location as an explanatory variable Chronosequence soils Conclusions  38 38 40 40 43 45 51  Chapter 5 Grain and straw yields of upland rice under leucaena and non-leucaena fallows Introduction Methods Results and discussion Grain yield Straw yield and yield components  52 52 54 54 61  iv  Tiller number and 1000 grain weight Percent filled spikelets Straw nutrient content Conclusions  63 64 65 69  Chapter 6 Effect of burning on crop yields in leucaena fallows Introduction Methods Results and discussion Crop yields Soil properties Regeneration Conclusions  71 72 74 74 75 77 78  Chapter 7 Nutrient budgets Introduction Methods Results and discussion Nutrient contents of primary fallow species Nutrient budgets at the time of harvest Nutrient distribution changes over time Conclusions  82 83 87 87 89 92 99  Chapter 8 Regeneration Introduction Methods Results and discussion Vegetation changes during the fallow period Species diversity Conclusions  101 101 102 104 107 108  Chapter 9 Impacts of charcoal production Introduction Methods Results and discussion Nutrient removal impacts due to charcoal making Returns to labor Conclusions  110 Ill 112 114 117 118  Chapter 10 Farmers, fallows and sustainability Introduction Methods Results and discussion Changes in crop yields over time  120 121 123 123 v  Perceptions of leucaena Farmer-initiated adaptations Costs of leucaena fallows Conclusions  128 130 131 132  Chapter 11 Conclusions Response to key research questions Management implications Future research needs  134 137 140  References  143  Appendices Appendix 1 Pedon descriptions Appendix 2 Moisture content adjustment for grain yields Appendix 3 Straw nutrient contents Appendix 4 Nutrient budget worksheets for graphics in Chapter 7  VI  150 153 156 159  List of Tables Table 2.1  Upland rice varieties most commonly planted in the Sto. Tomas study area  2.2  Soil (0-10 cm) nitrogen status changes under sweet potatoes and leucaena in Sialum, Papua New Guinea  2.3  9  17  Exchangeable cations, CEC, total N and pH under bush and leucaena fallow regrowth  18  2.4  Soil properties under two bush and leucaena fallows at 0-10 cm soil depth  19  3.1  Climatic averages for the San Jose, Occidental Mindoro weather station  23  3.2 3.3  Summary of field characteristics for experiment sites used in rice-related experiments . 29 Summary of field characteristics for experiment sites in regeneration and nutrient balance studies 30  3.4  Summary of crop duration and typhoon effects on rice production  4.1  Soil properties and analysis of variance for soils sampled in September-October 1993 . 41  4.2  Differences in field means for soil calcium and magnesium at the time of harvest  43  4.3  Mean separation tests for pH and CEC (in NH4OAc at pH 7)  44  4.4  Soil properties and analysis of variance by geographic location  46  4.5  Soil properties and analysis of variance for chronosequence plots  49  4.6  Effects of fallow type and age on total soil nitrogen concentrations (eg g"1)  49  4.7  Effects of fallow type and age on soil nitrogen ion concentrations in March 1994  5.1  Differences in grain moisture content means by field and fallow type  57  5.2  Moisture content-adjusted means by field and fallow type at 14 eg g"1 m.c  58  5.3  Rice yield component means and analysis of variance  62  vii  36  . . . . 50  5.4  Straw dry mass means by  5.5  Differences between fields in grain to straw mass ratio means  64  5.6  Mean separation tests for tiller number and 1000-grain mass  65  5.7  Mean separation tests for percent filled spikelets  66  5.8  Treatment means and analysis of variance for straw nutrient content  67  5.9  Straw phosphorus concentration by field and fallow type  68  5.10  Straw nitrogemphosphorus ratios by field and fallow type  68  6.1  Rice grain and straw yield and analysis of variance in the burn vs. no bum experiment . 76  6.2  Total foliar nutrients (eg g"1 of dry mass) and analysis of variance  76  6.3  Soil properties and analysis of variance for post-burn treatments  77  6.4  Regeneration of leucaena stumps and seedlings in burned and unbumed treatments . . . 78  7.1  Nutrient concentration in primary woody fallow species (eg g"1)  88  7.2  Nutrient concentrations in Chromolaena odorata from five sites in the study area (in eg g*1 of dry mass)  89  Contribution of Chromolaena odorata to nutrient stocks in natural bush fallows in Southwestern Nigeria  89  Estimated soil organic inputs from litterfall, clearing and weeding in a fully-stocked, 4-year old leucaena fallow  98  7.3  7.4  field  63  8.1  Ratio of live stumps to seedlings one year after clearing in rice yield study fields . . . . 103  8.2  Survival of leucaena stumps following the 1993 rice harvest in the field adjacent to field 6  104  8.3  Regeneration summary worksheet  106  8.4  Species diversity in leucaena and non-leucaena fallow plots  108  9.1  Comparative charcoal yield, proximate analysis and heating value from different materials produced in a masonary block kiln (adapted from PCARRD, 1985:40) . . . . 114 viii  9.2  Biomass distribution in leucaena stands  115  9.3  Estimated nutrient removals through harvesting for charcoal  116  9.4  Average time required for charcoal production based on farmer estimates of time spent  118  Farmer-reported sowing rates and rice grain yields  128  10.1  Appendix Tables Appendix Table 2.1  Moisture content adjustment factors  155  Appendix Table 3.1  Critical contents of various elements for deficiency in the rice plant . . . 156  Appendix Table 3.2 Mean separation tests for straw nitrogen content  156  Appendix Table 3.3 Mean separation tests for straw potassium content  157  Appendix Table 3.4 Mean separation tests for straw calcium content  157  Appendix Table 3.5 Mean separation tests for straw magnesium content  158  Appendix Table 4.1  Nutrient budgets at the time of rice harvest  160  Appendix Table 4.2  Nutrient balance worksheet for leucaena fallow  161  Appendix Table 4.3  Nutrient balance worksheet for non-leucaena fallow  163  Appendix Table 4.4  Nutrient content differences between leucaena and non-leucaena fallows  165  IX  List of Figures Figure 2.1  Human population growth in the Sto. Tomas watershed from 1973-1993  11  3.1  Location of the study site  22  3.2  Soil changes through the profile for pedons 1 and 2  26  3.3.  Relative locations of the 10 experiment sites within the study area  28  4.1  Interaction plot for total soil nitrogen  47  4.2  Total soil nitrogen trends in chronosequence plots for leucaena and non-leucaena fallows (0-30 cm soil depth)  48  5.1  Unadjusted grain dry mass by fallow treatment and field  55  7.1  Net difference (leucaena - non-leucaena) in potentially available nutrients between fallow types in rice harvest sites  91  7.2  Nutrient distributions in leucaena fallow sites (above-ground + soil)  93  7.3  Nutrient distribution in non-leucaena fallow chronosequence plots (above-ground + soil) Net difference (leucaena - non-leucaena) in potentially available nutrients in chronosequence sites  94  8.1  Above-ground herbaceous and woody vegetation in chronosequence plots  104  8.2  Population density of woody species in chronosequence plots  107  10.1  Estimated rice yields in non-leucaena fallow field of Mr. Reynol Sanuton  124  10.2  Estimated rice yields in natural bush fallow field of Mr. Bino Sanuton  125  10.3  Estimated rice yields in fallow field of Tamayo Sanuton converted to leucaena fallow in 1990  126  7.4  x  96  Appendix Figures Appendix Figure 2.1 Reductions in rice grain yield with changes in grain moisture content (as calculated from Nangju and DeDatta, 1970)  153  Appendix Figure 2.2 Grain moisture content differences between leucaena and non-leucaena fallows  154  XI  Acknowledgements Many people contributed to the work in this dissertation. Foremost are the farmers who have for years planted leucaena as a fallow improvement crop and who were kind enough to offer their land, time and expertise to this study. To them I owe a debt of gratitude: Danny Carandang, Junior Constantino, Conrado Mansalansan, Paras Monoy, Bino Sanuton, Tamayo Sanuton, Gauding Zamonte and Ceilo Zoleta. The Philippine-American Educational Foundation, the J. William Fulbright Foreign Scholarship Board, the University of British Columbia, the Soil Management Support Services Program of the U.S. Soil Conservation Service, the University of the Philippines at Los Banos, the International Rice Research Institute and the Bureau of Soils and Water Management of the Philippines all provided essential support for various aspects of this work. I am thankful for their sponsorship. Dr. Tim Ballard provided helpful, friendly advice throughout. His thoughtful guidance and counsel was both intellectually and morally stimulating! Drs. Kimmins, Kozak and Schreier patiently provided useful inputs that are reflected throughout this manuscript. Dr. John McLean was most supportive of my entire program and demonstrated the ability to make the system work as it was intended. Friend and colleague Dr. Celso Lantican helped out with numerous laboratory arrangements in Los Banos. Arlene, Elijah and Aileen were generally patient in dealing with the hardship and stress associated with a complex set of problems spread over time and 10,000 km. The One Who is Faithful kept all of the loose ends from unravelling! For all of the friendship and support these people and institutions represent -1 am truly grateful.  xii  Chapter 1 Introduction The changing context of shifting cultivation The ancient practice of shifting cultivation1 remains a primary means of livelihood for some 300 million of people (Russell, 1988), most of whom live in the humid or sub-humid tropics of the third world. As recently as 1976 it was estimated that shifting cultivation is used in some form on approximately 30% of the world's arable land (Sanchez, 1976:346). Although shifting cultivation has long been criticized as a destructive practice that contributes to deforestation and land degradation, few alternatives exist for many shifting cultivators.  Human populations continue to increase rapidly, generally more than 2% per year in the Third World, resulting in increasing pressure on land for food and income generation. The "Green Revolution" technologies that have increased agricultural production in lowland farming systems have generally not reached upland farmers, particularly those who practice shifting cultivation. Many farmers who practice shifting cultivation have few alternative means of livelihood and face increasing poverty and decreasing land availability, soil fertility and crop yields. Land degradation leads to a need to plant larger areas to produce the same amount of food, thus accelerating the cycle of decline.  1  Shifting cultivation is a form of cyclical cultivation whereby cultivators cut some or all vegetation, burn it and raise agricultural crops before moving on to another site to repeat the process in subsequent years. 1  In many areas shifting cultivation is no longer able to sustain human life as it has in the past.  If shifting cultivators are to improve or maintain their standard of living, they must either improve production on swidden2 sites, find new land to farm or find employment off-farm. Although governments have tried for decades to ameliorate problems related to shifting cultivation through enactment of laws prohibiting shifting cultivation or the forced relocation of populations into other areas, there are few practical, just alternatives in many shifting cultivation areas. There are few realistic solutions that do not require massive government inputs or dramatic reductions in human populations.  One approach to reducing the negative impact of shifting cultivation in areas of high population density is to improve the nutrient availability for crop growth by manipulating vegetation during the fallow period, increase crop yields and thereby reduce the amount of land required to sustain farm families. The restoration of soil physical and chemical properties during the fallow period is an important factor in the long-term future of farmers who must continue to depend on shifting cultivation for livelihood and a way of life.  Given the amount of time and effort required to develop, encourage and extend improvements in agricultural practices it is essential that efforts to improve swidden agriculture be sustainable. This is particularly true in shifting cultivation areas that are often  2  A swidden is a plot of land cleared for farming by burning away vegetation. 2  untouched by government development efforts. If these "improvements" are not in the longterm sustainable, they provide only short-term respite from declines in soil quality.  While many definitions of sustainable agriculture exist, perhaps the most relevant to the consideration of shifting cultivation was offered by Greenland (1975), who described the conditions necessary for a stable agriculture as: 1. The nutrients removed by crops are replenished in the soil 2. The physical condition of the soil suited to the land use type is maintained, usually meaning that the soil humus level is constant or increasing. 3. There is no buildup of weeds, pests and diseases; 4. There is no increase in soil acidity or toxic elements; and 5. Soil erosion is controlled.  Preliminary studies of a Leucaena leucocephala (leucaena) based fallow system on the island of Mindoro, Philippines indicate substantial advantages of a leucaena fallow when compared to adjacent natural bush fallows (MacDicken, 1981; MacDicken, 1991). These preliminary studies were exploratory and were inadequate to draw solid conclusions on the sustainability of this practice. However, the widespread adoption of the practice, the potential for sustainable development of shifting cultivation and the lack of adequate study of several key aspects indicate the need and potential significance of additional research. The work described in this dissertation seeks to address several of these key issues.  3  Research objectives The overall goal of the research described in this dissertation is to answer key questions related to the sustainability impacts of a nitrogen-fixing fallow tree species in upland ricebased shifting cultivation.  The key research questions asked in this thesis are:  1. Are grain production and the soil properties that limit the yield of upland rice significantly different between leucaena and natural bush fallows? 2. Does leucaena contribute significantly greater amounts of nutrients to the upper horizons of the soil than natural bush fallows? 3. Do these nutrient additions contribute to significant increases in subsequent crop growth? 4. Is it likely that production of upland rice in shifting cultivation is more sustainable over time following leucaena fallows than non-leucaena fallows?  Specific research objectives are: 1. To determine soil chemical and physical properties under leucaena and nonleucaena fallows of known age and species composition. 2. To evaluate regeneration patterns in leucaena and non-leucaena fallows; 3. To determine the yield of upland rice crops planted in fields following both the leucaena and non-leucaena fallows. 4. To determine upland rice yields under burned and unburned leucaena fallows. 5. To produce nutrient budgets that can be used to assess the sustainability of agricultural production in leucaena and non-leucaena based shifting cultivation.  4  6. To assess the constraints to further adoption of this practice by other farmers on other sites.  Organization of this thesis This thesis is organized as a series of scientific papers in Chapters 4-9 introduced by a general description of shifting cultivation in Chapter 2 and an overview of the study site and research approach in Chapter 3. Chapter 4 describes changes in soil properties over time in non-leucaena and leucaena fallows. Chapter 5 discusses and interprets upland rice yield data from nine sites. The response of the rice crop to burning in a single leucaena fallow field is described in Chapter 6. Chapter 7 presents nutrient budgets for fallowfieldsthrough the time of the first rice crop. Regeneration in a chronosequence of fallowfieldsof 1 to 4 years of age is described in Chapter 8. The impacts of charcoal production are discussed in Chapter 9. Chapter 10 outlines the response farmers have to the use of leucaena as a fallow crop species and overall conclusions are found in Chapter 11.  5  Chapter 2 Shifting cultivation near Sto. Tomas and elsewhere in the humid tropics Shifting cultivation as practiced by the Iraya Mangyans The Iraya Mangyans of northern Occidental Mindoro are traditional shifting cultivators with centuries of cumulative experience in swidden agriculture. For most Iraya Mangyans, shifting cultivation is the activity around which they organize their lives and upon which they depend for most of their sustenance. It is estimated that there are some 60,000 to 80,000 Mangyans in six ethnic groups (Schult, 1991:145), most of whom practice some form of shifting cultivation.  Much of the land cultivated by the Iraya is characterized by steep topography and acid soils (pH 5.5-6.5) in a monsoon climate with a long dry season (>5 months). The rainy season generally begins in May or June and lasts until November or December. Many traditional Iraya Mangyan farming areas, particularly those near roads or the coast, are in degraded secondary forest or savannah vegetation. Most of these areas have been deforested within the last 50 years. Deforestation in this area has come primarily from three sources: 1) logging, both legal and illegal; 2) shifting cultivation, and; 3) cattle grazing and the annual burning of Imperata cylindrica grasslands for pasture improvement.  6  The form of shifting cultivation practiced by the Iraya is of relatively low-intensity. The land utilization intensity index described by Ruthenberg (1980:15-16) can be used to compare the relative land-use intensity of cropping systems. This index is calculated using the equation:  „  years of cultivation x 100 years of cultivation + years of fallow  Shifting cultivation as currently practiced by the Iraya Mangyans ranges from R values of 16 to 50. The most common practice is presently 1 year of cropping followed by 3-4 years of fallow for R values of 20-25. Using the classification system of FAO/SEDA (1974), shifting cultivation systems are those with R values of < 33 while fallow systems are those with values 33 < R < 66.  The shifting cultivation cycle in northern Occidental Mndoro begins with the clearing of a field site in January-March. Clearing is done manually with axes and machetes, with most vegetation left on the field to dry in preparation for burning. Prior to 1990, very little wood was removed from the field after clearing. Now, most stems > 10 cm dbh (diameter at breast height) are removed for charcoal production. Prior to 1990, only large trees with economic value for sawn timber, boat keels, or house construction or trees with recognized medicinal value were protected or extracted prior to burning. Some farmers distribute slash during clearing to provide a consistent burn intensity throughout the field; others burn vegetation in place.  7  After allowing the vegetation to dry for a period of between 1 week and 3 months3, firebreaks are constructed and fires set. Burning is generally monitored closely by the farm family since the accidental (or intentional) burning of adjacent lands is punishable by heavy fines imposed by the community. The time of burning appears to be set most commonly by two factors: phase of the moon and prediction of the onset of the rainy season. In most cases, farmers try to select times that are auspicious according to a lunar calendar and which are as close as possible to planting time in order to reduce weed regrowth.  Unburned or partially burned stems that are small enough to be moved by hand are removed from the field or may in some cases be placed in slash piles for a second burning. Small pieces of partially burned stems or branches are generally left in the field.  Minimum tillage is practiced, using either a dibble stick made of hardwood or a "balawang" - a bamboo pole with a steel blade. In either case, shallow holes are dug at approximately 20-25 cm intervals for upland rice and maize. These holes are prepared at the time of planting. Planting often takes place on a lunar calendar, with some days designated as more auspicious than others. In general, most Iraya farmers try to predict when the rains will begin and plant just before that time. The primary crop is upland rice, using a wide range of traditional varieties, the most common of which are described in Table 2.1. Maize is commonly intercropped to provide a staple crop in late July or August, although relatively  3  Primary forest vegetation takes the longest drying period, followed by secondary forest, degraded secondary forest, leucaena and Chromolaena odorata fallows. 8  few maize plants are intercropped, to reduce shading effects on the rice crop. After the maize harvest in July-August, the whole plant is often removed from the field to reduce shading of the rice crop. A large number of other food, fiber and medicinal plants are often planted in Iraya swidden fields, although the proportion of these crops in the cropping pattern is very small.  Table 2.1 Upland rice varieties most commonly planted in the Sto. Tomas study area Crop duration  Varieties  Early  Kinanda Binisaya Biniyabos Inunisan  Medium  Pulang balat Komoros (red) Komoros (white) Pinitogo Bolohan Inagdami  Late  Pinakutan Dinalaga  Most farmers try to do two weedings, although in communities where children attend school or where child mortality rates are high, labor may not be adequate to do two weedings. The first weeding is generally done in June or early July and the second in August or early September.  9  The rice harvest begins in mid-late September and lasts through early November. Many farmers plant several varieties of different maturity length and eating quality. Rice is generally harvested by removing the panicle by hand, although a curved, hand-held blade {karet) is used when available. The straw is left in the field, unburned. Some farmers plant cassava after the rice harvest is complete.  Recent changes in the Sto. Tomas study area Five factors have impacted the practice of shifting cultivation in the Sto. Tomas watershed over the last 20 years:  1. Population increases. The settlement of the village of Sto. Tomas is relatively recent most families have permanently resided in the village proper only since 1978. The presence of a school and a larger than average Iraya community have attracted additional settlers, both Iraya and lowland Filipinos. Thus population pressures on land and other resources (e.g. well water for drinking, fish, minor forest products) have intensified dramatically since the early 1980s. The population growth in the Sto. Tomas watershed is shown in Figure 2.14.  2. Introduction of Leucaena leucocephala as a fallow improvement crop. In 1977, giant varieties of the nitrogen-fixing tree Leucaena leucocephala (leucaena) were introduced as a 4  Population was estimated through a reconstruction of the presence or absence of community members as recalled by female household heads. The assistance of Anna Ma. Avilla in this survey is gratefully acknowledged. 10  180  160  140  120 --  100  80 -In  1  1  1973 1975  1  1  1  1  1  1  1977 1979 1981  1  1  1  1  1  1  1  1  1  1  1983 1985 1987 1989 1991 Year  1  h-  1993  Figure 2.1 Human population growth in the Sto. Tomas watershed from 1973-1993  combination fallow improvement crop and commercial leaf meal source (see Chapter 10). By 1989 over 40 families had leucaena present in their fallow fields (MacDicken, 1991). Although the leaf meal production scheme failed due to marketing problems, leucaena has become a common fallow species in most fields.  3. Charcoal making. As noted above, using traditional shifting cultivation methods, the Iraya remove very little wood prior to burning a newly established field. Beginning in 1991, charcoal production changed this practice in the study area. Charcoal marketing agents, primarily lowland Filipinos from the barrio of Wawa, Abra de Hog or from Batangas  11  Province, began encouraging Iraya farmers to make charcoal they would purchase at a price of approximately US$.75 per 17 kg sack. Since there are few alternative sources of cash income in the study area, charcoal production has become a widespread activity, even though the returns to labor are low (Chapter 9).  4. Reduced labor availability. The presence of an elementary school in Sto. Tomas since 1979 has led to an increasing number of children who attend school full-time and are generally not available for as much farm labor as in years past. Children of working age, particularly those over 12 years of age, form an important labor pool for activities such as weeding. Since child mortality rates continue to be high (e.g. around 40-50% to age 12), the additional reduction in farm labor availability due to formal education is an important change in how fields are tended.  5. Rapid increase in the perennial asteraceous herb Chromolaena odorata. In the mid1970's, very little Chromolaena odorata (hagonoy) could be found in the study area. Scattered plants were seen, but were not dominant in fallow fields. By 1993, the predominant non-woody fallow species was hagonoy. Most farmers find it an easy species to manage and find it to be a useful fallow crop. In studies of shifting cultivation as practiced by another Mangyan tribe, the Hanunoo, in Southern Mindoro, Kasberg (undated) states: "According to my informant, most of the Hanunoo believe that hagunuy's (sic) benefits far outweigh its disadvantages. In comparing swidden sites from climax forests, secondary forest growth, and fields dominated by hagunuy, she draws several interesting conclusions. First, the problems with 12  weeds are most severe in swiddens of secondary forest growth, especially if the required ten tofifteen-yearfallow has not been observed. Weed competition in hagunuy dominatedfieldsis generally not as bad but is still a problem. Secondly, in comparing harvests from the respective swidden sites she has observed no major differences."  Soil changes under shifting cultivation The fallow period, during which physical and chemical properties of the soil are restored to a site, has been called the key to the long-term success of shifting cultivation (Ewel, 1976). Accumulation of nutrients and organic matter under various types of both mature tropical forest vegetation and herbaceous fallow crops has been described (Greenland and Kowal, 1960; Nye, 1961; Jaiyebo and Moore, 1964; Juo and Lai, 1977). Such studies show that both the type and age of a fallow crop may greatly influence the fertility status of a site by the end of the fallow period.  The importance of accumulation of organic matter and nutrients during the fallow period under native vegetation has been evaluated in a number of experiments and studies (Greenland and Kowal, 1960; Nye and Greenland, 1960; Zinke, Sabhasri and Kunstadter, 1978; Ewel, 1971). A number of other studies have examined the changes in the soils' chemical and physical properties under both arboreal and herbaceous fallow crops (Jaiyebo and Moore, 1964; Juo and Lai, 1977).  Jaiyebo and Moore (1964) found marked accumulation of exchangeable cations, nitrogen and organic matter under both planted herbaceous fallows and natural bush fallow. Little relationship between soil and plant Ca and Mg was found, but both the percentage and yield 13  of N in the plants related closely to total N levels in the soil. Maize crop yields were statistically the same following the bush fallow or tropical kudzu, but were markedly lower when preceded by the grass fallow. A high correlation (r=.87) was found between soil organic matter content and corn yields.  Juo and Lai (1977) also found an important relationship between soil organic matter content and productivity. They estimated that in order to prevent deterioration of chemical, physical and biological properties of the forest soil through maintenance of humified and partially decomposed organic matter, some 10-201 ha'1 yr"1 of dry matter would be required as a surface mulch.  Organic inputs from tropical forest trees come primarily through the following nutrient pathways (Jenny et al., 1949; Nye, 1961): 1. 2. 3. 4.  Litter fall Timber fall Root decomposition and nutrient excretion from roots and root nodules. Rain wash  Of these, litter fall has been found to be the single most important pathway of nutrient transfer (Jenny, 1949; Nye, 1961; Golley et al., 1975) and is perhaps the most commonly manipulated pathway.  A number of attempts have been made to improve the efficiency of the fallow period by speeding up the nutrient accumulation process through the use of fast-growing, and often  14  nitrogen-fixing tree species (Sanchez, 1976:384; Unruh, 1990). It is generally assumed that trees act as soil improvers, although most of this supposition is based on observations of soils under natural forest stands rather than planted stands. Sanchez et al. (1985) reviewed type I and type I I 5 experiments that examined the role of tree crops as soil improvers and found: 1) The fallow phase of tree crop production results in significant improvements in soil chemical properties - but that the extraction of nutrients through harvesting and leaching results in a depletion of key nutrients, primarily potassium. 2) These nutrient losses must be replaced by fertilization if yields are to be sustained.  Experience in producing agricultural crops has clearly shown that continued cropping and crop removal on the same piece of land without some sort of fertilization results in reduced yields. This situation can clearly be anticipated in the use of trees as soil improvers in agroforestry if the tree crops are to be harvested and removed from the site.  A major advantage in using nitrogen fixing trees (NFT) as soil improvers is their ability to fix atmospheric N into a form which can be utilized by plants and animals. Existing evidence indicates that when properly managed, plantations of NFT can significantly improve soil physical and chemical properties. Parfitt (1976), Juo and Lai (1977) and  5  Type I experiments are those in which soil dynamics are followed over time on the same site. Type II experiments are chronosequence studies. 15  MacDicken (1991) have reported several changes in soil chemical and physical properties under fallows of the NFT Leucaena leucocephala.  Leucaena as a fallow improvement species The leguminous tree Leucaena leucocephala (Lam.) DeWit (leucaena) has been widely used in a variety of indigenous and introduced agroforestry practices and has shown promise as an effective fallow improvement crop (Parfitt, 1976; IITA, 1980). However, there remains a lack of information on the effects of a leucaena fallow in shifting cultivation systems on soil erosion, soil nutrient contributions and sustainability.  Criteria for selection of species for use in community- based agroforestry systems have been suggested by Weaver (1979) and FAO (1977). Most important of these are: 1. 2. 3. 4. 5.  The capacity to produce foodstuffs and wood throughout the year. The ability to contribute to soil and water conservation. Low soil fertility requirements and fast-growth. Co-products which are easily stored. The ability to contribute to soil fertility improvement.  Leucaena meets these requirements. It produces wood used for light construction, fuel, pulp and several specialty uses. The foliage is used for animal feed, human consumption and green manure. It is a fast-growing species that has been successfully used in soil conservation practices in many countries. Perhaps most importantly, it appears to be an excellent soil improvement species.  16  Leucaena has been studied in a variety of production systems, including fallow improvement. Table 2.2 describes what appears to be a significant increase in total soil N after just 2 years of a leucaena fallow following cropping of sweet potatoes in Papua New Guinea. Table 2.3 compares leucaena as a fallow crop with natural bush regrowth on another site. The leucaena fallow was found to have resulted in significantly higher CEC, exchangeable Ca and K than did the bush fallow. Lack of improvement in total soil N suggests that much of the nitrogen in the leucaena leaf tissues may have been lost through volatilization, carried off in runoff and/or eroded sediments or leached out of the surface horizon.  Table 2.2 Soil (0-10 cm) nitrogen status changes under sweet potatoes and leucaena in Sialum, Papua New Guinea Crop Imperata cylindrica Ipomea batatas Leucaena leucocephala SOURCE: Parfitt, 1976  Cropping period 1 year 2 years  % Soil Nitrogen .35 .23 .75  Table 2.3 Exchangeable cations, CEC, total N and pH under bush and leucaena fallow regrowth pH Effective Ca Mg K Total N CEC Bush regrowth 6.5a 4.94a 3.34a 0.89a .42a .130a Leucaena 6.4a 6.22b 4.12b 1.14a .73b .146a Column values followed by the same letter are not significantly different at p=.05 SOURCE: Juo and Lai, 1977  17  MacDicken (1981) estimated that the improvements in soil chemical properties produced by a natural fallow could be produced by a leucaena fallow in a much shorter time under proper management. An evaluation of this leucaena fallow system in 1988 demonstrated the beneficial aspects of the leucaena fallow (MacDicken, 1991). Farmers using leucaena have successfully shortened fallow periodsfromover 6-8 years to 2-4 years with no perceived reduction in crop yields. Exchangeable calcium was significantly lower at 50 cm soil depth under the natural bush fallow. Soil pH in the 0-10 cm layer was significantly higher under the bush fallow, suggesting that pH was buffered by higher Ca deposition at the soil surface under the natural bush fallow, although there were no differences in exchangeable Ca in the surface layer (Table 2.4). It is also possible that acidification due to nitrification of larger quantities of organic N may explain the lower pH under the leucaena fallow. No significant differences between natural and leucaena fallows were found for other soil properties. The addition of nutrients during the fallow period takes place primarily through litterfall and from biomass felled at the time of clearing. Litterfall rates under two-year old leucaena stands in the Philippines are reported to be nearly 13 t ha'1" yr"1 compared with less than 91 ha"1 yr"1 for secondary forest stands (Saijse et al., 1979:320). These rates compare with litter fall rates of 7-15 t ha'1 yr_1 reported elsewhere for tropical secondary forests (Laudelot and Meyer, 1954; Nye, 1961; GoUey et al., 1975; Ewel, 1976). Van Den Beldt (1982) found that litterfall rates did not differ significantly under leucaena at population densities of 10,000 to 40,000 trees per ha and averaged 8.5 t ha"1 yr"1. The elemental composition of both fresh leaf tissue and senescent leaf tissue is generally higher for leucaena than for the mixed  18  tropical forest vegetation analyzed by Nye (1961), Greenland & Kowal (1960) and Ewel (1976). Table 2.4 Soil properties under two natural bush and leucaena fallows at 0-10 cm soil depth pH(H 2 0)  Organic matter  Ext. P (ppm)  Exch. Ca (ppm)  (%)  Fallow type  0-10 cm  50 cm  0-10 cm  50 cm  0-10 cm  Leucaena  5.8b  5.8a  3.8a  1.1a  27.5a  Nonleucaena  6.2a  6.1a  4.1a  1.0a  30.3a  50 cm  0-10 cm  50 cm  8.2a  1373a  1553b  23.4a  1380a  1173a  Means within a column followed by the same letter are not significantly different at P =0.05 SOURCE: MacDicken, 1991 Canopy leaf biomass of the Leucaena species hybrid K743 (L. leucocephala x L. diversifolid) has been reported to rangefrom3.2 to 10.01 ha"1 (dry matter) at the time of harvest of 3 year-old stands on four ustic moisture regime sites in Thailand (MacDicken, 1992:69-70). Similar quantities of canopy biomass might be expectedfromLeucaena leucocephala. The mean annual increment of woody biomass in 3 to 4 year-old stands on moderate-quality sites often ranges from 10-201 dry matter ha"1 yr"1.  The effects of burning In shifting cultivation, however, the greatest contribution in terms of available nutrients takes place just prior to planting (Nye and Greenland, 1964). Often thisflushin the release of nutrients is due to the burning of felled vegetation. Burning may be desirable as a 19  management tool in this system, as it is in a wide variety of other shifting systems, for a number of reasons such as improved seedbed preparation, more rapid release of nutrients, liming effects of the ash, and others (Rambo, 1981:5). Zinke et al. (1978) found that Ca, P and K are returned to the soil primarily as ash. Losses of nitrogen and sulphur due to volatilization during burning reduce the amounts of those nutrients available following fire (Sanchez, 1976:368).  20  Chapter 3 Site description and research approach  The experiments and studies described in this thesis were conducted in Sitio Sto. Tomas, Barangay Wawa, Abra de Hog, Occidental Mindoro, Philippines (13°29'N and 120°32'E) (Figures 3.1). The watershed in which the studies were conducted is approximately 800 ha with an estimated human population of 200. Leucaena fallows were established in this area beginning in 1976 on lands controlled by members of the Iraya Mangyan tribe.  A total of 18 experiment sites in the Sto. Tomas watershed were used during these studies. Research was conducted in the field in March 1993, September-November 1993 and March 1994. All of the experiment sites were managed by Iraya farmers on sloping lands currently in use for shifting cultivation.  Climate The study site is in a lowland, humid monsoonal climate. The area is classified as Philippine climate Type I - distinct wet and dry; wet from July to October and dry from November to June. Climatic averages covering the period 1981 to 1990 were obtained from the Philippine Weather Bureau (PAGASA) for the San Jose, Occidental Mindoro weather station, approximately 130 km South- Southeast of the research site in the same climate type.  21  Republic of the Philippines  Manila  Mindoro Island - * \  (  CS3  ^  Wawa, Abra de Hog  Figure 3.1 Location of the study site  VN5}  t  Table 3.1 Climatic averages for the San Jose, Occidental Mindoro weather station  Month  Rainfall (mm)  Number of rainy days  Max. monthly temperatures (°C)  Min. monthly temperatures (°C)  Mean monthly temperatures (°C)  January  2.3  3  31.8  22.2  27.0  February  6.7  1  32.1  21.8  26.9  March  11.5  2  33.6  23.2  28.4  April  15.0  3  34.3  23.8  29.0  May  112.5  8  33.7  24.1  28.9  June  415.4  16  31.5  23.4  27.4  July  369.5  19  30.5  23.2  26.9  August  443.2  20  30.5  23.1  26.8  September  378.4  17  30.6  23.0  26.8  October  270.3  15  31.0  23.1  27.1  November  107.6  7  32.0  23.0  27.5  December  19.2  3  31.8  22.2  27.0  ANNUAL  2152.0  114  31.9  23.0  27.5  Temperatures vary little throughout the year, meeting the criteria of an isohyperthermic temperature regime (Soil Survey Staff, 1975). Northeast trade winds prevail during January-April, while southwest winds prevail during June to October, bringing seasonally heavy rainfall. Rainfall in the study area meets the requirements of a ustic moisture regime.  23  Soils Soils in the Sto. Tomas watershed are generally moderately deep (i.e. 50-100 cm to lithic or paralithic contact), clay-loam to loam in texture and moderately acid (pH 5.5- 6.2). Topography is hilly to mountainous with most farm fields on slopes of 20-60%. Most fields are N, E or W facing.  In order to describe major soil types in the study area, two soil pits were prepared on two different landforms within the experimental area. Each pit was dug to lithic or paralithic contact (approximately 1 m) and described using Soil Taxonomy (Soil Survey Staff, 1990). A team of three soil scientists from the Philippine Bureau of Soils and Water Management assisted in collecting the samples, and providing the characterization for each pit. The complete characterizations for each pedon are found in Appendix 1.  The two sampled pedons were: Pedon No. 1 Parent material: Land use/vegetation: Soil temperature: Elevation: Taxonomic classification:  Residual, sedimentary (shale, siltstone) Leucaena, grasses, Saccharum spp., permanent crops. 26°C (0-10 cm) 45 m asl fine loamy, isohyperthermic, Typic Ustorthent  This pedon typifies crop lands in the upper-third slope position. Cultivation and erosion have truncated or removed the diagnostic horizons for other orders to be present. The fine loamy particle-size class, the presence of a paralithic contact at a depth of less than 1 m, no discernible diagnostic horizon nor evidence of ground water within 1.5 m of the surface 24  indicate this classification. Members of this soil subgroup are also found in extensive areas of the Great Plains of North America, where they are generally suited to grazing, forestry and non-irrigated grain crops (Soil Survey Staff, 1975). Figure 3.2 describes soil changes through the profile for both pedons.  Pedon No. 2 Parent material: Land use/vegetation:  Residual, sedimentary (shale, siltstone) grasses, upland rice, permanent crops grown in shifting cultivation Soil temperature: 26°C (0-10 cm) Elevation: 50 m asl Taxonomic classification: fine loamy, isohyperthermic, Typic Ustropepts Pedon 2 typifies mid-slope lands commonly used for upland rice production in Sto. Tomas. Deeper and of apparently higher base saturation than the upper slope soils described in Pedon No. 1, these soils are freely drained with an ochric epipedon present and high base saturation (>50%). As Inceptisols, they are generally more productive than associated soils of other orders (Buol et al., 1980). Common uses of this soil type on steep slopes such as Pedon No. 2 are woodland and non-irrigatedfieldcrops.  25  0  0 9 18 27 36 45  0 10 20 30 40 50  .0  0 10 20 30 40 50  0 9 18 27 36 45  Silt (cg/g)  Clay (cg/g)  0  9 18 27 36 45  20  20 E JO  40 O 73  •a  60  60  o  o CO  40  Q. U  CO  80  80 100  100  0 9 18 27 36 45 Sand (cg/g) 0  1 2  3  4  0™  0.1  0.2 0  20 6 o.  20 S  40  ja  I  •»-*  o. o •a  o  ^J J=  *->  o. u •a  60  o  CO  CO  100  1 2  3  4  80  Phosphorus (mg/kg)  O.O  40 60  o CO  10i 0  0 12 3 4 5 6 7 8  0.1  80 100  0.2  Potassium (cmol/kg)  012345678  Calcium (cmol/kg)  0 0 0.5 1.0 1.5 2.0  0  0 .0  20  20  /  , ;  40  40 o. u  a ID  •o  60  60 -  i  '5  o CO  /  i  E  •a  3.%i.fcv\.\5-'\9- 0  CO  80 10C (.0 0.5 1.0 1.5 2.0 Organic carbon (cg/g)  80  i.O 5.2 5.4 5.6 5.8 6.0 pH (H20)  i  100 0.0 3.% 1^\.\s."\9-° Cation exchange capacity (cmol/kg) Pedon #1 _ ^ Pedon #2  Figure 3.2 Soil changes through the profile for pedons 1 and 2 26  Vegetation The vegetation of Northern Occidental Mindoro is under substantial population pressure. Historically, Mindoro shares floristic origins with much of the island of Luzon and substantial portions of the humid lowland tropics of S.E. Asia. According to Dickerson (1928:286) it is likely that land links between Mindoro and Luzon during the Pliocene epoch of the Late Tertiary Period and between Mindoro and the island of Palawan during the Pleistocene epoch of the Glacial Period, resulted in substantial migration of flora. Much of northern Occidental Mindoro is of the monsoonal Molave forest type (characterized by the presence of the deciduous Vitex parviflora) and has in the last 50 years given way to coarse grasses (Imperata cylindrica and Saccharum spp.) and degraded secondary forest. The Molave forest type is typically found on low limestone hills in areas of shallow soils and long dry seasons. Very little Molave forest cover remains in the Philippines, most having been converted to other vegetation through logging, shifting cultivation and cattle grazing (de Guzman et al., 1986:77-78). Present day secondary forest cover in the Sto. Tomas watershed is characterized by species of Albizia, Intsia, Ficus, Antidesma, Alstonia, Trema, Diospyros, Mitragyna, Terminalia and Macaranga. Most areas are in open secondary forest cover characterized by <10 large trees per ha, scattered or dense shrubs, grasses, climbers and other herbs.  General research approach Five experiments were conducted on farmers'fieldsto meet the objectives stated above. Eighteen leucaena and bush fallowfieldsof known age, cropping history, slope and aspect 27  were used: 10 sites for rice-related experiments, and 8 for regeneration and nutrient budget studies. A series of informal interviews was used to characterize population and crop yield changes over time, farmer perceptions of leucaena fallows and current crop and charcoal production practices. Several ancillary studies were conducted to develop a weight table for leucaena seedlings, determine charcoal yields and conversion efficiencies and estimate seeding rates for upland rice. Table 3.2 summarizes the rice harvest sites and Table 3.3 the regeneration/nutrient budget sites. Location of the study sites is found in Figure 3.3.  N •  Verde Island Passage  Figure 3.3 Relative locations of the 10 experiment sites within the study area. Note: this topographic map was enlarged from a 1:24,000 scale map and is not true to scale. It is used here only to illustrate relative experiment site locations.  28  2  medium late early late medium late  Pulang balat Pinakutan Kinanda Dinalaga Pinitugo Bolohan  Junior Constantino  Tamayo Sanuton  Conrado Mansalansan  Conrado Mansalansan  Paras Monoy  Tamayo Sanuton  5  6  7  8  9  10  N  1  7  Slope position: U = upper, M = mid-slope, L = lower slope  Slope classes were defined as: Immoderate,<25%; 2=steep, 25-40%;3= very steep, >40%  N  E  3  1  S  2  W  L  U  M  M  U  M  M  M  U  M  Slope position7  M  L  M  M  M  M  L  L  L  M  Relative stoniness8  2  4  5  5  2  2  2  3  2  14  Fallow length  2  14  5  5  12  2  2  12  2  14  Total fallow age9  9  Total fallow age is years elapsed since the site was converted to leucaena  Stoniness was estimated based on the relative number of stones in core samples: F = few stones; M = moderate number of stones; S = stony  8  6  N  2  early  PSBURV1  Danny Carandang  4  3  N  2  late  Pinakutan  Bino Sanuton  3.00  N  S  1  medium  Inagdami  Gauding Zamonte  2  E  3  medium  Malagkit  Ceilo Zoleta  1  Aspect  Slope class6  Crop maturity class  Rice variety planted  Farmer  Field number  Table 3.2 Summary offieldcharacteristics for experiment sites used in rice-related experiments  Table 3.3 Summary offieldcharacteristics for experiment sites used in regeneration and nutrient balance studies Field number  Farmer  Slope class10  Aspect  Slope position11  Fallow length  1A  Tamayo Sanuton  2  W  M  1  2A  Nora Pagilagan  2  E  M  2  3A  Tamayo Sanuton  2  N  M  3  4A  Bino Sanuton  2  N  M  4  IB  Tamayo Sanuton  2  N  M  1  2B  Danny Carandang  2  E  M  2  3B  Nardo Pagilagan  2  W  M  3  4B  Tamayo Sanuton  2  N  M  4  Experiment sites were selected to allow: •  comparable stocking levels. Since the variability in leucaena population densities within and between sites is high, blocks with similar population densities were selected  •  homogeneity within blocks. Blocks were laid out along the contour with the long axis across the slope on soils of the highest degree of apparent homogeneity possible.  •  paired comparisons between fallow types. Where comparisons were made between fallow types, fields were selected where leucaena and natural bush fallows are adjacent to one another and of the same age.  Soil analyses were done by the Bureau of Soils and Water Management Central Research Laboratory in Diliman, Quezon City, Metro Manila. Soil pH was measured using glass  Slope classes: 1 = moderate, <25%; 2 = steep, 25-40%; 3 = very steep, >40% Slope position: U = upper; M = mid-slope; L = lower slope 30  electrodes in a 1:2 soil-water paste, organic carbon using wet digestion, available phosphorus using the Bray 1 method, Ca and Mg extracted with 1 N ammonium acetate at pH 7 and measured with atomic absorption spectrophotometry and total nitrogen using the micro-Kjeldahl method. Exchangeable K was also extracted with 1 N ammonium acetate at pH 7 and measured by atomic absorption spectrophotometry. Ammonium analysis of samples collected in March 1994 was conducted by the University of the Philippines at Los Baiios Dept. of Soils Analytical Laboratory using steam distillation with MgO. Nitrate analysis also used steam distillation with MgO and reduction to NH3.  Selection of the 0-30 cm soil depth for these studies is based on the rooting depth of the primary upland cereal crop grown in Occidental Mindoro - upland rice. Upland rice research suggests that there are substantial varietal differences in rooting depth and densities (Yoshida, 1981). However, it appears from root distribution studies of upland rice in heavy soils that more than 75% of the roots of most upland rice varieties are concentrated in the upper 30 cm (Yoshida, 1975:83).  The sites used for the regeneration and nutrient budget studies were uniform in slope class and position on the slope. Fallow ages were determined through interviews with each farmer.  31  Leucaena biomass estimation Diameter at breast height (dbh) and stump diameters of leucaena in experiment sites were measured for the purposes of stocking and biomass distribution. Total biomass and biomass components for trees > 2.0 cm dbh were determined using allometric relationships developed by Micosa-Tandug (1986) for the giant leucaena varieties used in the Sto. Tomas plantings (K8 and K28). Models were presented for six locations. Models for the Ilocos Sur location were selected due to similarities in climate, elevation, landform and soils. Equations of the following form were used to estimate biomass components for leucaena greater than 2.0 cm:  where:  b  W= oven-dry biomass of the component (kg) D= diameter at breast height (cm) b0, bj = regression constants  Component  b0  bx  R2  Stem wood Topwood and large branches Foliage Total biomass  0.1423 0.0142 0.0091 0.2399  2.39 2.55 2.16 2.31  .97 .90 .94 .98  A literature search for adequate models for leucaena seedlings and samplings failed to identify useful models. Mass tables were constructed for leucaena trees and tree seedling biomass by sampling of 30 seedlings in the range of 0.1 to 2.1 cm using the procedure described in MacDicken et al. (1991:83-84). Seedlings and saplings were selected from stands in four locations in the study area. Fresh mass was determined in the field. Subsamples were collected for oven-drying and moisture content calculation. Oven-dry mass 32  was fit to the model Y = a + b D2, where D = dbh (in cm) at 1.3 m and Y = oven-dry mass in kg.  The following equations were produced for use in modelling leucaena biomass of stems with a dbh of less than 2.0 cm: Whole tree: Wood: Foliage  Y = 0.001+0.164D2 Adjusted R2 = .96 Y = -0.002+ 0.153D2 Adjusted R2 = . 95 Y = 0.003 + 0.01 ID 2 Adjusted R2 = . 75  The use of chronosequences Chronosequence12 research permits the description of long-term temporal trends in a relatively short period of time. This approach was used to evaluate vegetation, soils and static nutrient budgets in leucaena and non-leucaena stands of varying ages. This approach assumes that "time-zero" conditions were the same and that the history of conditions and events were comparable between chronosequence plots (Kimmins, 1989:4). The advantages and disadvantages of this approach are well described (Cole and Van Miegroet, 1989), as are the risks of incorrect interpretation of results (Turvey and Smethurst, 1989).  The chronosequence work described in chapters 4 and 7-9 appear to meet the assumptions noted above. Potential sources of variation in the history of each plot are in the cropping  12  an ecological time series where the differences observed between the units (i.e. stands or ecosystems) that comprise the sequence are the result of differences in age or time. 33  history prior to the last rice crop and the potential differences in wood removal due to charcoal production. These limitations may limit the confidence with which the results can be used. However, this approach provides an adequate basis for several working hypotheses.  Data management and analysis Site and farmer data were entered into a series of database tables using Microsoft Access. Measurement data were entered using the spreadsheet Excel and the statistics package Systat. Data were proofread upon entry for mostfilesusing the on-entry proofreader provided by the Microsoft Sound System. Files were printed and re-checked against the original data sheets. Data analysis was done with Systat for Windows (5.03), Statgraphics Plus (6.0) and Statgraphics for Windows. Adjusted R2 values were calculated as per Systat procedures (Systat, 1992:158). The following steps were generally taken for each data set: 1. Scatterplot or SPLOM plots were printed to identify potential errors in entry or outliers that require re-checking, and to help recognize potentially important relationships between variables. 2. Box and whisker plots were printed to allow preliminary estimates of paired comparisons. 3. An analysis of variance was conducted for each of the planned experiments. This included the output of least-squares means and post-hoc comparisons using Tukey's HSD test. For the analysis of multi-site data the replications within site mean square term was used as the denominator in the F test. Explanatory relationships  34  with environmental variables such as geographic location, aspect, slope class and relative stoniness were also tested in analyses of variance. 4. Pearson product-moment correlation coefficients were calculated and compared using the Bartlett chi-square test as a general guide to probability levels for twovariable correlations. For multiple variable correlations, Bonferroni coefficients were calculated. 5. Where appropriate, line plots of planned relationships were produced using treatment means and a distance-weighted least squares smoothing technique. Linear and non-linear regression models were fit to relationships that appeared logical.  Limitations of the data All of the data collected for this series of studies were for a 1 year period, and included rice yields for only one harvest. Clearly, there is year-to-year variation in crop development and yield that was not sampled for this thesis. In addition, Typhoon Kadiang affected the study area on 5-7 October 1993. Although the typhoon passed through central Luzon, more than 300 km north of the study site, strong winds and rain lodged a substantial portion of the unharvested rice crop. Lodging resulted in spikelet removal, rat damage, restricted photosynthesis and translocation. The estimated maturity at the time of the typhoon ranged from the milk to hard dough stage. However, given the high yields (mean grain yields were 3 t ha"1 in the non-leucaena fallow fields and 41 ha"1 following leucaena fallows), it is unlikely that lodging due to the typhoon caused substantial reductions in grain yield.  35  An attempt was made to standardize rice variety use in each site to eliminate varietal response as a confounding variable. This proved to be impossible due to a lack of adequate seed quantities. However, farmer-cooperators took care to use both the same variety and same management practices in both fallow treatments in each field.  Table 3.4 Summary of crop duration and typhoon effects on rice production Field number  Rice variety planted  Crop maturity class  Days to harvest  Days from typhoon to harvest  Estimated maturity stage at time of typhoon  1  Malagkit  medium  138  9  medium dough  2  Inagdami  medium  138  8  medium dough  3  Pinakutan  late  149  25  milk or soft dough  4  PSBURV1  early  127  3  maturity  5  Pulang balat  medium  131  10  hard dough  7  Kinanda  early  115  na  na  8  Dinalaga  late  148  27  milk or soft dough  9  Pinitugo  medium  134  10  medium dough  10  Bolohan  late  147  24  milk or soft dough  Since all of thefieldsused for experimentation were farmer-managed, there was unavoidable variation in the intensity of management applied in each site. Differences in planting density and weeding intensity were observed between several fields. There were, however, no differences observed in management between fallow treatments within a field.  36  Cropping histories and fallow lengths were established through reconstruction of events for a particular farm field. Generally, this was checked against verifiable events - the eruption of Mount Pinatubo, the assignment of a local school teacher, or the year in school of a farm family member. Yet, even though care was taken to accurately establish these points in time, there is still the possibility that there are errors in some of these reconstructions.  Pre-planting measurements of available soil N (extractable nitrate and ammonium) and periodic measurements up to panicle initiation would likely provided the best estimates of the influence of leucaena on soil nitrogen related to crop growth and yield. However, these measurements are sensitive to the time span between sampling and analysis. The researcher was not present at the study site during this time of the year and the site was too distant from the laboratory to take these measurements during the harvest season when many other measurements were also being taken. For this reason, total soil N was used to provide an estimate of soil N changes under the leucaena fallow.  37  Chapter 4 Soil properties under leucaena and non-leucaena fallows Introduction Soil fertility changes are perhaps the most studied biophysical aspects of shifting cultivation. Yet, most of this research has focused on changes in natural fallows and the response to management practices such as burning, clearing, fertilization or the use of nitrogen-fixing, non-woody legumes. Relatively little is known of the soil changes due to the substitution of fast-growing legume trees for natural vegetation even though nitrogen-fixing trees have long been thought to have substantial soil improvement potentials.  This chapter examines the hypothesis that there are significant differences in soil properties between natural bush and leucaena fallows on steeply-sloping shifting cultivation sites in Occidental Mindoro. The specific experimental objective was to determine soil chemical and physical properties under leucaena and natural bush fallows of known age and species composition.  Methods Two studies were conducted to evaluate soil nutrient status under leucaena and nonleucaena fallows. Thefirstwas a series of nine single-factor experiments comparing soil nutrients in two levels of fallow types (leucaena and non-leucaena). These experiments used the same randomized complete block design and plots as the rice yield studies 38  described in Chapter 3. The second study utilized 16 chronosequence plots (2 fallow types x 4 ages x 2 replications) as described in Chapter 3.  Soil samples were collected during three different periods: March 1993 - pre-burn samples for the burn vs. no-burn experiment described in Chapter 6; September to November 1993 rice field and chronosequence sampling for total N, extractable P, exchangeable K, Ca, Mg, CEC, pH and organic carbon; March 1994 for ammonium and nitrate. Simple random sampling was used with a 5 x 30 cm split-core sampler with hammer attachment. Four core samples were taken from each treatment plot and bulked by plot in the burning and rice field studies and eight core samples were collected per plot in the chronosequence studies. Surface litter was removed from each sampling site prior to sampling and coarse fragments removed from the sample at the time of core removal. Samples were screened in the field using a 5mm screen and air-dried immediately after collection in a shaded area for approximately 3 days at maximum temperatures ranging from 25-35 °C. Samples for ammonium and nitrate analyses were air-dried and placed in air-tight plastic sample bags. The average time from sampling to analysis for these samples was 1 week.  Samples were analyzed by the Bureau of Soils and Water Management Central Research Laboratory in Diliman, Quezon City, Metro Manila. Soil analyses included soil pH with glass electrodes in a 1:2 soil-water paste, organic carbon using wet digestion, available phosphorus using the Bray 1 method, Ca and Mg extracted with 1 N ammonium acetate at pH 7 and measured with a flame photometer and total nitrogen using the micro-Kjeldahl  39  method. CEC was determined using the method of Gillman (1979) and unbuffered solutions. Exchangeable K was requested but inadvertently omitted from all but 8 samples. Ammonium analysis of samples collected in March 1994 was conducted by the University of the Philippines at Los Bafios Dept. of Soils Analytical Laboratory using steam distillation with MgO. Nitrate analysis also used steam distillation with MgO and the reduction to NH3. Carbon-nitrogen and nitrogen-phosphorus ratios were calculated using Systat 5.03.  Analyses of variance were conducted on soil properties for both studies. Tukey's HSD test was used to test mean differences for factors with significant F values. Significant interactions were plotted, then tested with Tukey's HSD. Soil properties were correlated with grain and straw yields and yield components using Bonferroni's adjustment for multiple comparisons. Soil nutrient levels over time were plotted by fallow type for every age using the distance-weighted least squares method of Systat. Average values from the 1993 rice harvest sampling forfieldsites on the eastern ridge were used for fallow age 0 since all but two of the chronosequence plots were established on the eastern ridge.  Results and discussion Soils at the time of rice harvest Initial analyses comparing sites and fallows detected no significant differences between fallow types at the time of harvest, although differences were detected betweenfieldsfor Ca, Mg, pH and CEC (Table 4.1). There were no significant differences between fallow types in total nitrogen, phosphorus or organic carbon at the time of harvest, nor were there 40  differences in the carbon:nitrogen or nitrogemphosphorus ratios. None of the soil chemical properties measured at the time of harvest were significantly correlated with any of the measured rice yields or yield components.  Table 4.1 Soil properties and analysis of variance for soils sampled in September-October 1993 Total N(cg g')  Ext. P (mg g-1)  Exch Ca (cmol kg 1 )  Exch. Mg (cmol kg 1 )  pH (H 2 0)  Organic C(cg g"1)  CEC (cmolc kg"')  C:N ratio  N:P ratio  Nonleucaena  0.17  6.6  8.7  3.6  6.0  1.41  16.5  8.49  0.03  Leucaena  0.16  6.4  8.8  3.6  6.1  1.45  16.9  8.94  0.03  Treatment  Source  df  Total N  P  Ca  Mg  pH  Organic carbon  CEC  C:N ratio  N:P ratio  Site  8  NS  NS  **  **  **  NS  **  NS  NS  Reps in site  11  NS  NS  NS  NS  *  NS  *  NS  NS  Fallow  1  NS  NS  NS  NS  NS  NS  NS  NS  NS  Sitex fallow  8  NS  NS  NS  NS  NS  NS  NS  NS  NS  Pooled residual  10  * = F is significant at p=.05, ** = F is significant at p= 01, NS = non-significant at p=.05 Total soil nitrogen and magnesium levels were moderately high for tropical soils; calcium, organic carbon and cation exchange capacity were at moderate levels. Soil pH was in the near optimum range for upland rice. The only difference in soil calcium was between field 41  5 and field 9. Field 9 produced the highest grain and stover yields of all fields, yet had the lowest soil calcium levels.  Field number 1 had significantly higher exchangeable magnesium than fields 7, 8 and 9. Deficiencies of Ca and Mg are common in upland rice in the tropics (Fageria et al., 1991:190), although the high yields found in the field with the lowest Ca and Mg levels (field 9) suggest that none of the sites in this experiment suffered from either Ca or Mg deficiency.  pH and CEC Field 9 had the lowest pH (Table 4.3) and produced the highest grain and stover yields (Chapter 5). The four lowest pH values were all found in soils on the western side of the study area (i.e. west of the Sto. Tomas River). Rice is adapted to a wide range of soil conditions and is acid-tolerant. According to Fageria et al. (1991:165), the optimum pH for upland rice culture is around 6.0.  CEC was lower in fields 7 and 9 than in fields 1,3,5 and 10. Field 7 had the lowest grain and stover yields and field 9 had the highest yields, suggesting that CEC is not a useful explanatory variable for rice yield in this set of experiments. CEC is often related to organic carbon content of soils under shifting cultivation (Sanchez, 1976:372), although CEC was not significantly correlated with organic carbon in these studies.  42  Table 4.2 Differences in field means for soil calcium and magnesium at the time of harvest13 Field number  Exch. Ca (cmol kg"1 soil)  Field number  Exch. Mg (cmol kg"1 soil)  5  11.0a  1  4.7a  3  9.8ab  2  4.2ab  4  9.8ab  3  4. lab  1  9.4ab  10  4.0ab  10  8.8ab  5  3.5ab  8  8.2ab  4  3.4ab  2  8.0ab  8  3.2b  7  7.3ab  9  2.8b  9  6.5b  7  2.7b  Geographic location as an explanatory variable Grain yields and a number of soil properties appeared to differ between geographic locations. When grain and stover yields were sorted by soil nutrient levels and reviewed irrespective of the significance of the differences in treatment means, thefieldswest of the Sto. Tomas River generally had lower nutrient levels and grain yields than those east of the  13  Values followed by the same letter are not significantly different at p=.05 using Tukey's HSD test. 43  Table 4.3 Mean separation tests for pH and CEC (in NF^OA, at pH 7) Field number  pH(H 2 0)  Field number  CEC (cmolc kg 1 )  10  6.4a  3  18.4a  5  6.4a  5  18.3a  3  6.3a  10  18.3a  4  6.3a  1  18.0a  2  6.0ab  2  17.2ab  7  5.9ab  4  17.0ab  8  5.9ab  8  15.2ab  1  5.8b  9  14.1b  9  5.6b  7  13.9b  river. This is not true forfieldno. 9, which had the highest grain and stover yields (Chapter 5). When soil properties were analyzed using ridge as a source in the analysis of variance, significant differences between these geographic clusters were detected in soil calcium, pH and CEC (Table 4.4). Calcium, pH and CEC were higher infieldseast of the river than in fields west of the river.  According to 2 of the 3 persons who can recall pre-World War II vegetation and land-use patterns, the western ridge (see Figure 3.2 for a topographic map) has over the past 50-80 years been subjected to different land uses than has the eastern ridge. The western ridge was  44  inhabited and farmed earlier than the eastern ridge. According to Mr. Ignacio Zoleta14 the western ridge was regularly farmed before the war, while the eastern ridge was undisturbed during this period. Cattle grazing was conducted on the pasture lease of a Mr. Zoleta15 from approximately the 1950's to the early 1970's on most of the study area on the western ridge. Pasture management in this part of northern Occidental Mindoro largely consists of the annual burning oflmperata cylindrica grasslands. As a consequence, much of the western ridge was subjected to morefrequentand widespread fire than was the eastern ridge. By the late 1970's, many of thefieldson the western ridge had been abandoned, largely due to low crop yields and severe weed competition from Imperata cylindrica. Ocular surveys of both ridges suggest that unused grasslands cover a larger proportion of the study area on the western ridge. The longer period of cultivation and the broader use of indiscriminantfireon the western ridge may explain the generally lower fertility and crop yields on these soils.  Chronosequence soils The only significant differences between fallow age or type on soils in the chronosequence plot soils were related to soil nitrogen (Table 4.5). The interaction between fallow age and type is described in Figure 4.1 and Table 4.6. No differences were detected in total soil Nbetween fallow treatments in fallow ages 1 and 2. This is consistent with the findings reported above for soil N levels at the time of rice harvest. Nitrogen levels were higher in  14  Mangyan Mayor of western Abra de Hog  15  A Tagalog landowner - no relation to Ignacio Zoleta 45  the leucaena fallow than in the non-leucaena fallow in years 3 and 4. This relationship is described graphically in Figure 4.2. Table 4.4 Soil properties and analysis of variance by geographic location Geographic location  Total N (eg g')  ExtP (mg kg')  Exch. Ca (cmol kg')  Exch. Mg (cmol kg')  pH (H20)  Organic carbon (egg 1 )  CEC (cmol kg 1 )  C:N ratio (kg kg')  N:P ratio (kg kg')  Eastern ridge  0.169  7  9.9  3.7  6.3  1.46  18.0  8.68  0.03  Western ridge  0.162  6  7.9  3.5  5.9  1.41  15.7  8.74  0.03  Source  df  Total N  P  Ca  Mg  pH  Organic carbon  CEC  C:N ratio  N:P ratio  Location  1  NS  NS  **  NS  **  NS  **  NS  NS  Fallow  1  NS  NS  NS  NS  NS  NS  NS  *  NS  Location x fallow  1  NS  NS  NS  NS  NS  NS  NS  NS  NS  Fields in locations  7  NS  NS  NS  *  NS  NS  **  NS  NS  Reps in fields  11  NS  NS  NS  NS  NS  NS  *  NS  NS  Fallow x fields in locations  7  NS  NS  NS  NS  NS  NS  NS  NS  NS  Pooled residual  11  * = F is significant at p=.05, ** = F is significant at p=01, NS = non-significant at p=.05  46  0.20 0.19 0.18 0.18  * d 0.17 u 60 0.16 O w a 0.15  * J  Year 4  •>H •TH  o 09 ^ •M  o H  014  0.13 0.13 0.12 0.11 0.10 Non-leucaena  Leucaena  Fallow type Figure 4.1 Interaction plot for total soil nitrogen  Declines in total soil nitrogen to the lowest point 3-4 years after the beginning of the fallow period have been reported elsewhere (Szott et al., 1991; Nakano, 1978; Zinke et al., 1978; Sabhasri, 1978). For example, Zinke et al. (1978:145-147) found a decrease in total soil N levels between years 1 and 4 of the fallow period in traditional shifting cultivation among the Lua' in Northern Thailand.  The declines in total soil nitrogen in chronosequence plots for years 3 and 4 relate to lower biomass and soil organic additions (Chapter 7). Year 4 plots had likely been thinned for charcoal production. The difference in rate of decline in total soil N between fallow types appears to be due to the greater accumulation of NH4+ under the leucaena fallow (Table 4.7)  47  0.20  0.18  G  0.15  60  o  \  t-l  a  \  0.13  \  o  3 o.io  e  E2 4 o  0.08  Leucaena Non-leucaena  0.05 0.0  1.0  2.0  3.0  4.0  Fallow age (years) Figure 4.2 Total soil nitrogen trends in chronosequence plots for leucaena and non-leucaena fallows (0-30 cm soil depth). Note: each data point represents one chronosequence plot and the greater resistance of soil ammonium to leaching. The higher total soil N:NH4+ ratio in the leucaena fallow treatment suggests that not only is total soil N higher under the leucaena fallow, but a greater proportion of this soil N is in the ammonium ion form. Although these data are based on sampling during a single period in the dry season, they are consistent with the findings of Sanginga et al. (1989) in Nigeria, who also found significantly greater accumulation of NH4+ following the addition of leucaena prunings.  48  Table 4.5 Soil properties and analysis of variance for chronosequence plots  Total N(cg g"1)  Ext.P (mg kg')  Exch. Ca (cmol kg"')  Exch. Mg (cmol kg')  Nonleucaena  0.15  4.9  7.6  Leucaena  0.17  8.2  8.5  Treatment  pH (H 2 0)  Organic carbon (egg"')  CEC (cmol kg')  C:N ratio (kg kg')  N:P ratio (kg kg')  3.9  5.9  1.25  16.7  8.6  0.04  3.8  6.1  1.39  17.6  8.2  0.03  Source  df  Total N  P  Ca  Mg  pH  Organic carbon  CEC  C:N ratio  N:P ratio  Year  3  **  NS  NS  NS  NS  NS  NS  *  NS  Fallow  1  **  NS  NS  NS  NS  NS  NS  NS  NS  Yearx fallow  3  *  NS  NS  NS  NS  NS  NS  NS  NS  Pooled residual  8  Table 4.6 Effects of fallow type and age on total soil nitrogen concentrations (eg g"1) Fallow type 16 Year Leucaena  Non-leucaena  1992  .160a  .160ab  1991  .195a  .190a  1990  .180a  .150b  1989  .145b  .105c  16  Average of two replications. Mean separations in a column by Tukey's HSD at p=.05. Column values followed by the same letter are not significantly different at p=05. LSD05 value for comparing fallow type means in a row is .022 49  The literature suggests that the dominance of soil ammonium is not common in upland tropical soils. Mueller-Harvey et al. (1989:321) demonstrated the dominance of soil nitrate nitrogen in a tropical secondary forest soil in southern Nigeria during the dry season. Soil ammonium was low throughout the season and seasonal changes were small, but nitrate concentrations were much higher in the high organic matter plots and showed marked seasonal trends. Nitrate was highest in April at the onset of the rains (>60 ug g'1) and declined rapidly in May and June (-10 ug g"1) due to uptake by the crop and losses to leaching and denitrification. Yoshida (1975) also noted that nitrates commonly predominate in upland soils.This suggests that the dominance of NH4+ in the leucaena fallow soils may be unique in upland rice soils and may explain why total soil N reductions in fallow years 3 and 4 are less in the leucaena fallow than in the non-leucaena fallow.  Table 4.7 Effects of fallow type on soil nitrogen ion concentrations in March 1994 Treatment  Nitrogen ion form NOngkg 1 ) asNH 4 +  NOngkg- 1 ) as N0 3 -  Total soil NH4+:N ratio  Total soil N03":N ratio  Leucaena fallow  27.4a  17.2a  .20a  .12a  Non-leucaena fallow  17.8b  13.2b  .13b  .10a  Note: Column values followed by the same letter are not significantly different at p=.05. The LSD (p=.05) for comparison of N ion concentration means within rows is 5.4  50  Conclusions Soil nutrients generally did not differ significantly betweenfieldsat the time of rice harvest, even though there were significant differences in grain and stover yields between fields. Total soil nitrogen under the year 3 and 4 leucaena fallows was higher than in the nonleucaena fallows and suggests the potential contributions of nitrogenfromleucaena in the field under the conditions found in Sto. Tomas take at least 3 years to be potentially useful to crops. The crop yield data presented in Chapter 5 do not conclusively demonstrate a direct fallow age to yield relationship, but yields in the longer fallows were generally higher than those in the shorter fallows. This suggests that the relationship between fallow age and total soil nitrogen detected in this experiment is an important indicator of the amount of time required to accumulate enough soil nitrogen to achieve a positive yield response.  The evidence of greater ammonium ion concentrations in the leucaena fallow treatment is preliminary, but does indicate an important area of further study. If nitrification inhibitors are present in or associated with leucaena foliage, this would perhaps explain why ammonium levels are substantially higher in upland soils that are generally higher in nitrates. The non-protein amino acid mimosine found in concentrations of around 4 eg g"1 (dry mass) in leucaena foliage is a known toxin for non-ruminant livestock; it or some other component may well be a nitrification inhibitor, although no mention of this was found in the literature. Lower rates of nitrification and higher soil ammonium could increase soil nitrogen accumulation in a fallow system due to reduced leaching of nitrates.  51  Chapter 5 Grain and straw yields of upland rice under leucaena and non-leucaena fallows Introduction Rice is the most sought-after staple food in northern Occidental Mindoro. Although other starch crops such as cassava, banana and sweet potato may be as important as rice in terms of the number of calories produced by farmers in Sto. Tomas, rice is a crop of high status and traditional value. It is also the crop around which much of the yearly schedule is organized for the Iraya Mangyans (see Chapter 2). The need to sustain rice yields over time is strongly felt and widely recognized among Mangyan farmers in the Sto. Tomas area (Chapter 10). Thus, for improvements in farming practice to be widely adopted in this and similar areas, they must be capable of sustaining rice production over time.  An experiment was conducted in the 1993 rainy season to evaluate rice yields on nine fields on two ridges in the Sto. Tomas watershed (Chapter 3). The objective of this experiment was to determine the yield and yield components of upland rice crops planted infieldsunder both leucaena and non-leucaena fallows. The hypothesis tested was that grain and straw yields of upland rice are different following fallows of different species composition.  Methods Nine farmfieldswere used for this study (see Table 3.2). A randomized complete block design was used with 2 or 3 replications, the number of replicates depending onfieldsize, 52  slope position, homogeneity of slope and aspect. Plot size was 2 x 4 m with the long axis oriented up and down the slope. This single factor experiment included two levels: leucaena and non-leucaena fallows.  Rice crop measurements were rice grain and straw fresh weight, percent filled spikelets, 1000 grain weight and number of tillers per m2. Eight 50 x 50 cm sub-plots were harvested per treatment plot using the methods described by Gomez (1972). Both treatments were harvested together in eachfield,with the date of harvest set by the farmer-cooperator. Grain was harvested by removing the panicles in the field. Spikelets were removed by hand before weighing. All tillers were counted - effective and ineffective. All count measurements were made using a multiple tally meter. Sub-samples of grain and straw were oven-dried to constant weight for use in adjusting fresh weight to oven-dry weight.  The analysis of variance and Tukey's HSD were conducted on grain and straw yield, 1000 grain weight, tiller number, square-root transformed values for percentfilledspikelets, and arc-sine transformed values for grain to straw ratio and grain moisture content. Since field sites represent both different environments and different rice varieties, data were plotted and analyzed to explain the influence of environmental factors, primarily nutrient-related, and varietal factors using the approach described by Gomez and Gomez (1984:567-571).  Grain moisture content differences between fallow types within eachfieldwere adjusted at the plot level using the relationship between grain yield and moisture content as described  53  by Nangju and De Datta (1970). A two-way table was constructed relating increases in yield to decreasing grain moisture content. Adjustments were made by using the tabular adjustment factor on the plot mean (within the same replication) with higher grain moisture content (Appendix 2). All values were adjusted to 14 g H20 g dry matter"1 moisture content using the formula: Adjusted grain weight = A x W, where W = fresh mass of harvested grain, and: , 100 - M A= 86 where M = moisture content (percent, wet mass basis, i.e. 100(mass of water / mass of water + solids)).  Results and discussion Grain yield Initial combined analysis of variance using all 9 fields detected a significant interaction in grain yields between fields and fallows (Figure 5.1). This interaction was difficult to interpret due to the use of different varieties in each field, and inconclusive differences in soil and foliar nutrient contents. However, consistent and significantly higher (p<01) grain moisture content following the leucaena fallow suggested that the development and maturation of grain following leucaena fallows was delayed.  54  600 r550 500  %  450  M  „  400  03 CO  S 350 >> •° 300 a |  250 200 150  100  Leucaena Non-leucaena Fallow type  Figure 5.1 Unadjusted grain dry mass by fallow treatment and field  Grain moisture content was higher in the leucaena fallow treatments and differed significantly amongfields(Table 5.1). Rice grain moisture content generally decreases from around 60 eg g"1 during the initial stages of grainfillto around 20 eg g"1 or less at maturity (Yoshida, 1981:59). The higher grain moisture content in the leucaena fallow treatment suggests the crop develops more slowly following a leucaena fallow. An analysis of variance using grain moisture content categories as sources demonstrated two significant categories: 1.fieldswith grain moisture content of <25 eg g"1 and, 2.fieldswith grain moisture content of >25 eg g'1. Fields in which moisture content was <25 eg g'1 m.c. had higher grain yields under leucaena fallows, and those with higher m.c. had lower grain yields or there was no difference between fallow types.  55  Nanju and DeDatta (1970) studied changes in grain moisture and the optimum time of harvest in paddy rice for two seasons. The optimum grain moisture content at harvest was between 18-22 eg g"1. Plants harvested earlier than this time (i.e. with higher grain moisture contents) produced substantially lower yields. For example, a difference of 3 eg g'1 in grain moisture indicated a maturity difference of 6-8 days and a grain yield reduction of around 10-20 eg g'1 (Appendix 2). This is a significant delay considering the ripening stage of rice in the tropics (from flowering to maturity) is between 25-35 days regardless of variety (DeDatta, 1981:160). Moisture content differences of the magnitude found in these experiments (as high as 14 eg g"1 in one plot) indicate substantial differences in maturation.  Adjusted grain yield differences were highly significant (p < 01) between fallow treatments and between fields (Table 5.2). Adjusted yields were higher following the leucaena fallow and were higher in field 9, planted to a tall, medium-maturity variety (Pinitugo), than in the three lowest yielding fields. Rice in the leucaena treatment in this field was taller by an average of 15 cm with foliage that was visibly a darker green and more vigorous than the foliage in the non-leucaena fallow treatment.  The most likely reason for these differences is the greater amount of nitrogen accumulated during the fallow period in the leucaena fallow (see Figure 4.1). The influence of nitrogen on crop development is dependent on species, soils and climate. The maturation of cerealcrops may be delayed by higher nitrogen levels in the temperate zone and may or may  56  Table 5. 1 Differences in grain moisture content means byfieldand fallow type Field  Fallow type  Field number  Mean (eg g"1)  Fallow type  Mean (eg g"1)  8  30.0a  Leucaena  27.2a  1  30.0a  Non-leucaena  23.9b  10  28.8ab  7  26.3ab  9  25.8ab  3  23.3ab  4  22.3ab  2  21.5b  5  21.5b  not be affected in the tropics (Russell, 1980:32-34). Sanchez (1976:451) states that neither subspecies of rice shows significant changes in growth duration with varying nitrogen rates and that nitrogen uptake at harvest is essentially identical, although this may vary with season. Nangju and DeDatta (1970) compared grain yields and time of harvest for 0, 60 and 120 kg ha"1 N at the IRRI campus in Los Banos in a dry season study of four rice varieties, including one tall indica type. The "window" for maximum grain yields was approximately 4 days later in the 120 kg ha"1 N treatment than in the 0 N treatment, and about 1 day later in the 60 kg ha"1 N treatment. The same experiment conducted during the rainy season failed to produce these kinds of treatment differences in relative grain maturity.17  17  The Sto. Tomas watershed is approximately 100 km S.E. of Los Banos and has a wet season climate that is drier than the Los Banos dry season. 57  Table 5. 2 Moisture content adjusted grain yields byfieldand fallow type at 14 eg g"1 m.c. Field  Fallow type  Field number  R value18  Mean (t ha"1)  Fallow type  Mean (t ha'1)  9  0.14  5.14a  Leucaena  3.8a  3  0.17  3.61ab  Non-leucaena  2.7b  1  0.07  3.51ab  2  0.33  3.34ab  5  0.33  3.21ab  10  0.33  3.20ab  4  0.50  2.70b  7  0.20  2.66b  8  0.20  1.66b  Nitrogen additions in the leucaena fallow comefromtwo primary sources - litterfall and leucaena seedlings weeded during thefirst8 weeks of the rice crop. Approximately 100200 kg ha"1 of nitrogen are added annually during the fallow period through leaf litter, with another 32-280 kg N ha"1 added during thefirst8 weeks of crop growthfromweeded leucaena seedlings left as a surface mulch (Chapter 7). Some of this N comesfromfixation and somefromthe soil N pool.  18  The R value as used here is the land-use intensity index described by Ruthenburg (see Chapter 2). 58  Grain yields were generally high in allfieldscompared to production reported elsewhere in the Philippines. IRRI (1975:87) suggested the yield of traditional varieties is limited to about 41 ha"1, although this was based on upland rice research at a very early stage of development. The same source reports that C22, an upland high-yielding variety released in the mid-1970s, frequently yielded 3-41 ha"1 under upland conditions. This variety was tested by farmers in Sto. Tomas in 1979-80 and was rejected due to low yield. Kawano et al. (1972) reported yields of up to 61 ha"1 under high fertilization in the Peruvian jungle.  The high yields in these experiments are in part explained by two factors: 1. The farmer-cooperators were all experienced, traditional shifting cultivators with high levels of skill in a no-tillage system using crop varieties they had in most cases used for decades. Weed control was generally good in allfieldsexceptfields7 and 8 where thefirstweeding was seriously delayed. 2. The soils of the study area are generally fertile and the weather during the 1993 cropping season was, with the exception of one typhoon of moderate strength in early October, well-suited to rice production.  There is also some evidence that crop yields in experimental plots were substantially higher than the whole-field grain yields reported by farmers (Chapter 10). Higher yields following the leucaena fallows are not explained by soil nutrients at the time of harvest (Chapter 4) nor by the foliar P difference in fallow treatments. This is consistent with thefindingsof other investigators working with leucaena prunings in alley cropping systems (Kang et al., 1985;  59  Sanginga et al., 1989). However, the form of inorganic nitrogen provided to rice may be important to crop performance. In a study of maize response to leucaena prunings on an Ultisol in Fashola, Nigeria maize yields were higher with the addition of leucaena prunings (Sanginga et al., 1989). Thye found no differences between soil nitrate levels, but soil ammonium levels were higher in the soils with prunings added than in soils with no prunings added.  Analysis of soil ammonium and nitrate conducted in March 1994 indicated that soil ammonium levels were higher in leucaena fallows (Chapter 4). It remains unknown whether or not this difference is constant through the year. However, higher levels of ammonium during the month of March (a period of virtually no rainfall) are likely to remain high through the dry season due to a lack of nitrification. Russell (1980:337) cites a study by Semb and Robinson (1969) in which they found flushes of 13 to 180 kg N ha"1 shortly after the onset of rains in fallow fields in East Africa. Accelerated nutrient inputs during the "assart" period of accelerated mineralization of the forest floor following clearcutting are commonly found in a variety of forest types (Vitousek, 1981). Mueller-Harvey et al. (1989:321) demonstrated the flush of nitrate in a tropical forest soil following clearing and the onset of the rainy season.  Farmers in Sto. Tomas generally plant rice so that seed germinates with the first reliable rains of the season, usually during the month of May (Chapter 2). Soils under leucaena fallows of over 2 years in length are higher in total soil N, nitrate and ammonium at the time  60  of clearing than soils under non-leucaena fallows (Chapter 4). This plus the generally low expected rate of nitrification in dry soils through the dry season likely provide a larger flush of nitrate following the onset of rains. Since planting is done to coincide with these rains, the young rice seedlings are able to use this nitrate to speed early development.  Several alley cropping studies have found advantages in the incorporation of leucaena green manure, suggesting substantial N losses due to higher volatilization of N in the surface mulch (Evensen, 1984). However, Read et al. (1985) found no difference in maize yields or ear leaf N values between incorporation and surface application of leucaena leaves. Decomposition of surface applied leucaena foliage (55 eg g"1 loss in weight after 20 days) was rapid. They also found residual benefits of a single foliage application to a subsequent maize crop.  Straw yield and yield components Grainfilland grain moisture content differed signficantly between treatments, while differences were noted betweenfieldsfor all measured yield components (Table 5.3). Straw weight was higher in the tallest varieties than in the short-statured HYV infield4 and field 7, which had unusually low straw weight (Table 5.4) and did not differ between fallow types. Straw dry weight and grain weight were linearly correlated (R = .64, p<001).  61  Table 5.3 Rice yield component means and analysis of variance Straw (t ha 1 )  Tiller number perm 2  1000 grain wt. (g)  Percent filled spikelets (eg g"1)  Grain: Straw ratio  Grain m.c. (eg g 1 )  Natural fallow  5.41  194.9  25.43  90.9  .71  23.9  Leucaena fallow  6.16  194.5  24.62  86.5  .74  27.2  Treatment  Source  df  Straw  Tiller number  1000 grain wt.  Grain fill  Grain: Straw ratio  Grain m.c.  Field  8  **  **  **  **  **  **  Reps in Field  11  NS  NS  NS  NS  NS  NS  Fallow  1  NS  NS  NS  *  NS  **  Field x fallow  8  NS  NS  NS  NS  NS  NS  Pooled residual  11  * = F is significant at p=.05, ** = F is significant at p=01, NS = non-significant at p=.05  In general, the grain-straw ratio was lowest forfieldsin which grain moisture content was highest andfilledspikelet percentage was lowest (fields 1, 10, 3 and 8). Grain to straw ratio is a relative measure of a crop's economic value. For paddy rice this value ranges from about 0.5 for tall, traditional varieties to about 1.0 for improved, short-stemmed varieties (Yoshida, 1981).  62  Table 5.4 Straw dry mass means by field Field number  Maturity  Stature  Mean straw yield (t ha"1)  3  medium-long  tall  8.75a  9  medium  tall  8.15a  10  medium  medium  7.65ab  1  medium  medium  6.69ab  2  medium  medium  5.70ab  8  long  short  5.25ab  5  short  short  4.97abc  4  medium  short (HYV)  3.64bc  7  short  medium  1.25c  Grain to straw ratio was remarkably high infield7 (Table 5.6). Rice in both treatments in thisfieldwas chlorotic, lacked vigor and was infected with rust disease. The next highest ratio was in the high-yielding variety PSBURV1, a short-stemmed variety bred for highnitrogen inputs. Tiller number and 1000 grain weight Field 8 produced more tillers than all butfields10 and 3 (Table 5.6). In general,fieldson the western ridge tended to produce more tillers than those on the eastern ridge, although this difference was not significant. The development of tiller primordia into tillers depends on plant nutritional status, the supply of carbohydrate, light and temperature. Tillering is also impaired by N or P deficiency (Yoshida, 1981:131). The values for tiller number are 63  Table 5.5 Differences betweenfieldsin grain to straw mass ratio means Field number  Mean grain to straw ratio  7  2.1a  4  0.79b  5  0.67bc  2  0.62bc  9  0.60bc  1  0.53bcd  10  0.48cd  3  0.43cd  8  0.30d  for all culms, including tillers with panicles and those without. Grain mass is generally a very stable varietal character (Yoshida, 1981:238). The differences in 1000 grain mass detected betweenfieldsare very likely varietal or grain maturity differences.  Percent filled spikelets Percentfilledspikelets was significantly higher in the non-leucaena fallow treatment and was lowest infields8 and 3 (Table 5.7). Three factors that affectfilled-spikeletpercentage under the conditions found in these experiments are nitrogen levels, lodging or bending and strong winds (Yoshida and Parao, 1976). Total soil nitrogen levels are moderately high (Chapter 4) and do not appear to be limiting. Lodging was observed in all of the experiments due to Typhoon Kadiang which generated strong winds in the experimental  64  Table 5.6 Mean separation tests for tiller number and 1000 grain mass Field number  Number of tillers perm 2  Field number  1000 grain mass (g per 1000 grains)  8  263a  1  30.8a  10  245ab  7  28.7ab  3  203abc  4  26.3abc  7  192bc  3  24.8a-d  9  192bc  9  24.5bcd  4  189bcd  10  24.2bcd  1  187cd  2  23.5bcd  5  155cd  8  21.7cd  2  126d  5  20.8cd  area on October 6, 1993. Given windspeeds for this typhoon in Southern Luzon, the windspeeds in Sto. Tomas likely reached peaks of 60-100 kph.  Straw nutrient content There were significant differences betweenfieldsfor all of the nutrients measured in this experiment (Table 5.8). The only differences between fallow types were in foliar phosphorus, where foliar P concentration was higher in the non-leucaena fallow treatment (Table 5.9) and in higher foliar N:P mass ratio in the leucaena fallow treatment (Tables 5.8 and 5.10). Mean separations for foliar nutrients other than P are found in Appendix 3.  65  Table 5.7 Mean separation tests for percentfilledspikelets Field  Fallow type  Field number  Mean (eg g"1)  Fallow type  Mean (eg g 1 )  2  96.6a  Leucaena  86.5b  1  96.0ab  Non-leucaena  90.1a  5  94.4abc  9  93.9abc  10  93.4abc  7  85.7a-d  4  84.7a-d  8  77.9d  3  75.5d  Chromolaena odorata, which dominates the non-leucaena fallows has foliar P concentrations that are 3 times that of leucaena foliage (Chapter 7). This is reflected in higher P concentrations in the litter layer and in straw P concentrations. Straw P and adjusted grain yields were inversely proportional (R2=52, p=.001), although concentrations were well below the 1 eg g"1 reported as the critical toxicity level (Yoshida, 1981:183). This inverse relationship between straw biomass and P concentration may be due to dilution as straw biomass increases. Such a relationship was detected (straw P = 1/(4.473 + 0.0103 *straw dry weight, R^.24), although weak.  66  Table 5. 8 Treatment means and analysis of variance for straw nutrient content Total N (eg g'1)  P (eg g 1 )  K (eg g"1)  Ca (eg g"1)  Mg (eg g 1 )  N:P ratio (kg kg 1 )  Nonleucaena fallow  0.653  0.154  1.7  0.037  0.074  5.97  Leucaena fallow  0.69  0.107  1.76  0.039  0.068  7.58  Treatment  Source  df  Total N  P  K  Ca  Mg  N:P  Field  8  **  **  **  **  **  **  Replications within Field  11  NS  NS  NS  **  NS  **  Fallow  1  NS  **  NS  NS  NS  *  Field x fallow  8  NS  NS  NS  NS  NS  NS  Pooled residual  11  The highest grain yields were found infieldswith the highest straw N:P ratio (Table 5.10). The N:P ratio was also higher in the leucaena fallow treatment which produced the highest moisture content-adjusted grain yields, suggesting that P alone is not a limiting factor to grain yield even though P levels in all sites appear to be low. N:P ratio and moisture content-adjusted grain yield were linearly correlated (R=78, p=.02)19. Higher N:P ratios in the rice straw indicate higher available N for rice after leucaena, further suggesting greater available soil N after the leucaena fallow.  19  Field 10 was dropped from this analysis due to unusually high P values thought to be due to contamination 67  Table 5.9 Straw phosphorus concentration byfieldand fallow type Field  Fallow type  Field number  Mean (eg g"1) Fallow type  Mean (eg g"1)  10  .225a  Leucaena  .107a  4  .183a  Non-leucaena  .154b  7  .170ab  8  .147ab  5  .145ab  2  .HObc  1  .083bc  3  .058c  9  .057c  Table 5.10 Straw nitrogen:phosphorus ratios byfieldand fallow type Field  Fallow type  Field number  Mean  Fallow type  Mean  9  11.05a  Leucaena  7.58a  1  10.45a  Non-leucaena  5.97b  3  9.36ab  2  7.66abc  8  6.01abc  7  5.60a-d  5  •  4.84bcd  4  3.77cd  10  2.26d 68  Conclusions Differences in grain yield were related to both fallow type and site. Moisture contentadjusted grain yield was higher by an average of 42 eg g"1 in the leucaena fallow treatment, with a range of differences of 19 to 106 eg g"1. The higher grain moisture content of rice from the leucaena fallow treatments appears to relate to greater total soil nitrogen accumulation as demonstrated in the chronosequence sites.  Nitrogen additions from leucaena occur during three periods: litterfall during the fallow period, additions to the soil after clearing and at the time of burning, and from seedlings and coppice shoots removed and applied to the soil surface as a mulch. Of these sources, the two most important in terms of grain yield appear to be litterfall and surface mulch from weeding.  The greater accumulation of soil nitrogen in leucaena fallows of > 2 years of age and additions from seedling biomass are the most likely explanations of higher yields following leucaena fallows. The difference in soil N accumulation during the fallow period may be due to lower nitrification in soils with high leucaena foliage inputs over time. The higher levels of soil ammonium and the work of Sanginga et al. (1989) suggest a nitrification inhibitor may slow nitrification of ammonium mineralized from leucaena foliage. This would reduce leaching of nitrates and result in higher total soil N accumulations.  69  Chromolaena biomass contains a very high concentration of P, which is related to higher foliar P contents and lower N.P in the non-leucaena treatment. This primary value of Chromolaena in swidden as practiced by the Iraya is in the accumulation of P in the biomass which upon burning provides an important source of available P for crop growth. No explanation for the inverse relationship between foliar P and grain yield was found, although Chromolaena may produce chemicals that can inhibit other plant species (Holm et al., 1977:212).  Farmers in Sto. Tomas tend to harvest the rice crop while grain moisture content is higher than the optimum for maximum yield, presumably due to a desire to reduce the risk of grain loss due to weather or herbivory. Higher grain and straw moisture contents, plus lower grainfillpercentages in rice following the leucaena fallow suggest that upland rice grown under leucaena fallows was physiologically less mature than rice grown in non-leucaena fallows.  70  Chapter 6 Effect of burning on crop yields in leucaena fallows Introduction Burning is often cited as an integral and critical component of shifting cultivation systems (Nye and Greenland, 1960:66-67; Ruthenburg, 1980:36-38; Rambo, 1981:5-9). The functions of burning are generally thought to be reduced weed and insect pests, improved nutrient availability through ash additions, soil pH changes and changes in microbiological populations, and improved soil tilth. All of the nutrient elements in the above-ground fallow biomass are added to the soil in ash through burning with the exception of nitrogen and sulphur. Most of the nitrogen in the biomass is lost to the atmosphere as gaseous nitrogen, nitrogen oxides and ammonia. Sulphur is lost as sulphur dioxide. Nitrogen losses through burning may negate some of the value of enhanced fallows using trees with high nitrogen foliage.  Nitrogen fixing trees often have foliage of higher N concentration than associated nonnitrogen fixing trees (MacDicken, 1994:4-5), and are often advocated for agroforestry systems. This nitrogen is primarily contributed to the soil through leaf litterfall during the fallow, additions to the soil after clearing (but before burning) and, in cases such as leucaena fallows in Sto. Tomas, through surface mulched seedlings that are weeded early in the cropping season. Annual leucaena litterfall rates of over 121 ha"1 have been reported in a humid climate in the Philippines (Sajise et al., 1979:320) and 101 ha"1 in a semi-arid monsoonal climate on saline soils in India (Puri et al.,1992). Litterfall rates may be 71  independent of spacing in high density plantations of 10,000-40,000 trees per ha (Van Den Beldt, 1982). While these rates are within the range of values often reported for secondary or mature tropical forests, the N, P and K concentrations of leucaena foliage (the largest component of litter) are generally higher (Chapter 7). Leucaena foliage has a low C:N ratio and is composed of very fine pinnae that decompose rapidly, so most of the additions from litterfall are rapidly converted into soil organic matter. Fire in shifting cultivation does not appear to destroy humified organic matter from the soil, although the non-humified litter layer is lost (Nye and Greenland, 1960:70). Nitrogen from leucaena leaf litter is not substantially affected by burning.  The source of organic inputs most affected by burning is the biomass on the soil surface at the time of burning. In the case of leucaena fallows this is most likely in the range of 50150 dry t ha'1 depending on age, population density, vigor and intensity of wood removals. While burning does not affect mineral nutrients other than N and S, the losses of N alone due to burning are likely in the range of 75-200 kg ha"1. If burning could be eliminated from site preparation, it is possible that long-term buildup of organic matter and soil nitrogen under a nitrogen fixing tree fallow would be enhanced and subsequent crop yields increased. An experiment was conducted in a leucaena fallow field to test the null hypothesis that there is no difference in crop yields following burned and unburned leucaena fallows.  Methods A single experiment was conducted on the field of Mr. Tamayo Sanuton in a randomized complete block design with four pseudo-replications. Burning was conducted in a single  72  strip aligned up and down the slope with replications placed across the slope. This singlefactor experiment included two levels - burning and no burning. Leucaena population density prior to burning was 3,625 trees per ha and was nearly equal in each set of plots.  Rice yields were determined by simple random sampling within treatment plots using eight 50 x 50 cm sample plots per treatment plot. Pre-burn and post-burn soil samples for the 030 cm horizon were collected in March 1993 and 182 days later in September 1993. Burning was done in April 1993 and the site planted to upland rice in May 1993. Soils were analyzed for N, P, Ca, Mg, pH, CEC and organic carbon. Straw was analyzed for N, P, K, Ca and Mg. Rice grain and straw fresh weight, %filledspikelets, 1000 grain weight and number of tillers per m2 were measured. Sub-samples were oven-dried to constant weight for use in adjusting fresh weight to oven-dry weight.  The number of leucaena seedlings and the diameter classes of living and dead leucaena stumps were recorded in both the burned and unburned treatments in November 1993 following the rice harvest. Stumps were identified by species, examined for coppice shoots, measured using a vernier caliper and placed into 5 cm diameter classes. Leucaena stumps were evenly distributed throughout the experiment, with 29 stumps in the unburned treatment plot and 28 stumps in the burned treatment plot. A t-test of stump diameters showed no significant difference in diameter between the two treatments. The lack of difference between treatments in stem number and stump diameter indicates homogeneity in  73  pre-clearing leucaena vegetation. The analysis of variance and Tukey's HSD test were conducted on crop yield, soil and foliar nutrients.  Results and discussion Crop yields There were no significant differences in grain or straw yields between burning treatments (Table 6.1), nor did any of the measured yield components differ between treatments. Foliar nutrient contents were not significantly different between treatments at the p=.05 level. (Table 6.2). However, phosphorus content of the straw was significantly higher in the burned treatment at p=.06, suggesting greater soil P availability after burning.  Leucaena wood is a major pool of phosphorus in leucaena-dominated fallows (Chapter 7). Burning concentrates readily available P in the ash. In the system of charcoal production presently used in Sto. Tomas, substantial P removals take place due to removal of wood for charcoal making. In the unbumed treatment, the net available P losses were higher due to a lack of P contributed to the soil through ash. Nutrient budgets from the chronosequence studies demonstrate the importance of potentially available P in leucaena wood to available P useful for crop growth.  Mr. Sanuton, the owner and farmer who managed thisfieldduring the experiment, reported greater weed regrowth in the unbumed treatment, requiring two weedings as opposed to one weeding in the burned treatment. He also predicted there would be no difference in rice 74  yields between the burn and no-burn treatment before harvest. These observations were based on both his knowledge of the rice crop as it developed and the lack of apparent crop yield reductions in at least two poorly burned fields in previous years. On two reported occasions in leucaena fallowfields,very low intensity burns were observed by farmers in neighboring fields. Crop yields were said to be "similar" to yields in burned fields and were not different in weed regrowth. He did report purpling of leaf margins at early stages of crop growth in the unburned treatment. This had disappeared by the heading stage.  Soil properties There were no differences in measured soil chemical properties between treatments in either the pre-burn or post-burn soil analyses conducted on soils collected in March 1993 (Table 6.3). Any changes in soil nutrients that may have taken place following burning were no longer present 180 days after burning. The lack of change in nutrient levels six months after burning is consistent withfindingselsewhere. Nye and Greenland (1960) indicated several studies in which organic matter either remained constant or increased following burning. Brinkmann and Nascimento (1973) reported increases in Ca, Mg, P, K and pH immediately after burning but these declined to very near the pre-burn levels within 148 days after burning. It is unlikely that soil pH changes would be detected in this experiment due to the thickness of the samples (30 cm) and the likelihood that any pH changes only occurred in the top 15-20 cm.  75  Table 6.1 Rice grain and straw yield and analysis of variance in the burn vs. no burn experiment  Grain (t ha 1 )  Straw (t ha 1 )  Tiller number per m2  1000 grain wt. (g)  Grain fill (eg g"1)  Burned  2.26  8.18  237.5  21.7  78.6  .276  Unburned  2.32  7.51  243.8  20.6  80.6  .310  Source  df  Grain  Straw  Tiller number  1000 grain wt.  Grain fill  Grain: Straw ratio  Replication  3  NS  NS  NS  NS  NS  NS  Burn type  1  NS  NS  NS  NS  NS  NS  Residual  3  Treatment  Grain: Straw ratio (kg kg 1 )  Table 6.2 Total foliar nutrients (in eg g'1 of dry mass) and analysis of variance in the burning experiment N  P  K  Ca  Mg  Burned  .538  .085  1.31  .04  .10  Unburned  .498  .070  1.33  .03  .12  Treatment  19  Source  df  N  P  K  Ca  Mg  Replication  3  NS  NS  NS  NS  NS  Burn type  1  NS  NS19  NS  NS  NS  Residual  3  The p value for phosphorus was .058 76  Table 6.3 Soil properties and analysis of variance for post-burn treatments Treatment  Organic carbon (eg g"1)  pH (H 2 0)  Total N(cg g1)  Ext. P (mg kg 1 )  Exch. Ca (cmol kg 1 )  Exch.Mg (cmol kg 1 )  CEC (cmolc kg 1 )  N/P (kg kg 1 )  Burned  1.39  6.35  .163  5.88  9.28  6.35  16.1  .028  Unburned  1.38  6.32  .165  5.68  9.0  6.32  17.0  .028  Source  df  Organic carbon  pH  Total N  Ext. P  Ca  Mg  CEC  N/P  Replication  3  NS  NS  NS  NS  NS  NS  NS  NS  Burn type  1  NS  NS  NS  NS  NS  NS  NS  NS  Residual  3  Regeneration There were substantial differences in regeneration between treatment types (Table 6.4). Burning and weeding of coppice shoots killed all but the largest stump in the burned treatment. In the unburned treatment, nearly 60% of the stumps were killed by the removal of coppice shoots. Results of a stump mortality survey (Chapter 8) show the higher survival rate of larger stumps following burning in other fields.  Leucaena seed with their waxy seed coat are known to respond well to scarification treatments, as shown by the profuse germination of seed observed in many leucaena fallow fields following burning. According to Mr. Sanuton, thefieldowner, there were more leucaena seedlings in the burned treatment during early stages of crop development. The 77  number of leucaena seedlings present later in development was higher in the unburned treatment. This is likely due to the effects of seed scarification by fire on early germination of leucaena seedlings in the burned treatment. These seedlings were weeded in June. Seeds that germinate later were not as likely to be weeded as they posed little competition to the taller rice crop. Since germination was not stimulated byfirein the unburned treatment, germination was delayed and these seedlings were not as heavily weeded as those in the burning treatment. Table 6.4 Regeneration of leucaena stumps and seedlings in burned and unburned treatments Leucaena stumps  Treatment  Mean stump diameter (cm) Living stumps  Basal area (m2)20  Dead stumps  Percent living stumps  Number of seedlings following rice harvest  Burned  2521  8.9  0.049  3.6  30  Unburned  10.4  8.2  0.135  41.4  48  Conclusions The lack of grain yield response to burning suggests that nutrient additions through burning are not the only reason farmers burn slash prior to planting upland rice. There are at least six major benefits attributed to burning in shifting cultivation (Rambo, 1981:5): clearing of  20  Basal area of living stumps only  21  There was only one living stump in the burned treatment 78  unwanted vegetation, alteration of soil structure making planting easier, enhancement of soil fertility by ashes, decrease in soil acidity, increased nutrient availability and reductions in microbial, insect and weed populations. None of these benefits were observed in this experiment through the pre-burn and post-burn samplings or indirectly through grain yields, although the lower foliar P in the unburned treatment suggests that perhaps repeated rotations with no burning could lead to P deficiencies.  The generally assumed benefits of burning are often associated with forest fallows in which large trees are a predominant part of the fallow vegetation. These benefits may not necessarily apply to leucaena fallows in the study area for the following reasons:  1. Unwanted vegetation: Mr. Sanuton reported only 1 weeding was required in the burned treatment compared to two weedings in the unburned treatment. Given the decreasing availability of child labor for weeding (Chapter 2), reduced weed requirements are an important advantage of burning.  2. Improvement of soil structure: High litterfall rates during the fallow period result in visibly improved soil tilth under leucaena stands that results in seedbeds that are easily planted without burning.  3. Enhanced soil fertility: Given the lack of significant differences between treatments in grain yield, it appears that any changes in soil reaction or nutrient availability due to burning  79  are not enough to be important in the short-term. There were no mineral nutrient differences between treatments after 6 months.  4. Reduced risk from pests: This is difficult to assess and was not evaluated in this experiment.  The advantages of burning are the potential for increased P availability and reduced weeding, particularly in fields with shorter fallow periods or in fields in which openings in the canopy have been created by removal of trees for charcoal. The potential loss of P from the site for charcoal makes burning even more important as a source of mineral P additions. When site preparation does not include burning, it is very likely that greater amounts of wood will be removed for charcoal or placed in slash piles that concentrate nutrients in a smaller portion of the field.  Burning directly influences the form of regeneration in leucaena fallows by killing most of the stumps. Stump survival following clearing and two weedings was nearly 40% higher in the unburned treatment. It is clear that leucaena stumps can survive fire and that the stump mortality observed in this experiment is due to the combination of fire and weeding of coppice shoots. The only apparent advantage of the no-burning treatment is in the survival of stumps and initial seedling regeneration. It is unknown whether this is of long-term consequence.  80  These observations need to be repeated in future years to determine if this lack of response to burning is a constant under leucaena fallows and if repeated cropping without burning continues to provide stable yields. Given the present study, burning should remain as a management tool in leucaena fallows.  81  Chapter 7 Nutrient budgets in leucaena and non-leucaena fallows  Introduction Changes in the relative size of nutrient pools over time can have important land management implications for the sustainability of production. In an ideal crop production system, soil nutrients would be available during critical times for crop growth and would be immobilized during the rest of the year to prevent losses due to leaching. To be sustainable in real terms this cropping system would need not only to maintain nutrient stocks but to increase the long-term supply of available nutrients to meet expanding demands of increasing populations and higher material expectations. By understanding the relative sizes of nutrient pools in fallows of varying composition, as well as the changes that take place over time, management practices can be developed or refined to optimize nutrient use and minimize nutrient losses from the system.  The evaluation of nutrients lost through harvest is a critical factor in determining the sustainability of cropping systems, particularly those that use high-yielding or fastgrowing species or where whole tree harvesting is practiced (Kimmins, 1977). This may be particularly true on the steep slopes and shallow soils on which shifting cultivation is often practiced in the humid tropics, or where all types or diameter classes of wood are utilized. 82  The objective of this study was to produce nutrient budgets of leucaena and non-leucaena fallows and to use these to evaluate the potential sustainability of each fallow type. The study used a chronosequence approach (see Chapter 3) and was designed only to produce a static nutrient budget to help explain differences in nutrients between the two fallow types. The study does not measure inputs from precipitation or below-ground litterfall nor does it attempt to account for leaching or erosion losses.  Methods Soils and litter Soil nutrients were determined through sampling of the 0-30 cm horizon in each of the 9 rice harvest fields and each of the 16 chronosequence plots. Soil sampling methods are described in detail in Chapters 3 and 4. Constant bulk density (1.4 Mg m"3) and coarse fragment (25 eg g"1) values from the soil pit characterizations were used to calculate soil mass per ha.  The standing litter crop was sampled in October 1993 by randomly collecting litter in 50 x 50 cm sampling frame at 8 locations in each of the 16 chronosequence plots. All samples were bulked by chronosequence plot, weighed and sub-sampled for nutrient analysis and moisture content. Nutrient analysis samples were air-dried for three days, oven dried at 65° C to a constant weight, ground with a Wiley mill through a 0.85 mm (20 mesh) sieve and submitted to the Bureau of Soils and Water Management Central Research Laboratory for analysis. Sub-samples for moisture content were dried at 80°  83  C to constant weight and weighed to the nearest 0.1 g. Annual litterfall was estimated from literature sources.  Vegetation Rice straw was sampled and weighed as described in Chapter 5. Herbaceous vegetation in 10 x 10 m chronosequence plots was sampled in October 1993 by removing all aboveground herbaceous vegetation from 50 x 50 cm sampling frames at 8 locations in each of the 16 chronosequence plots. Sampling plots were distributed randomly in strata defined by understory vegetation. The number of samples in each stratum was set in proportion to the size of the stratum. All samples were bulked by chronosequence plot, weighed and sub-sampled for nutrient analysis and moisture content. Nutrient analysis samples for herbaceous vegetation were air-dried for three days, oven dried at 65° C to a constant weight, ground with a Wiley mill through a 0.85 mm (20 mesh) sieve and submitted to the Bureau of Soils and Water Management Central Research Laboratory for analysis. Sub-samples for moisture content were dried at 80° C to constant weight and weighed to the nearest 0.1 g.  Tree seedlings of < 1.5 m in height were included with herbaceous vegetation. All trees, including seedlings of > 1.5 m in height, were enumerated by the species local name. Dbh was measured for every tree using a vernier caliper or diameter tape. Wood and tree leaf samples were collected from groups of five trees and bulked into composite samples. Three bulked composite samples were taken for each treatment, representing a  84  total of 15 trees per treatment. Leaf samples were taken from four locations in the canopy - two from young leaves and two from older leaves. Wood samples were taken from lower, mid and upper stem or branch portions and included bark, heartwood and sapwood. Below-ground biomass was not measured or estimated. These samples were air-dried, ground in a Wiley mill and analyzed by the Dept. of Soils laboratory at the University of the Philippines at Los Banos. Grain nutrient contents were taken from Yoshida (1981:129) for the sub-species indica, variety Peta - a widely studied, mediumtall variety.  Several assumptions were used in producing the nutrient budgets described in this chapter, all relating to the source of organic inputs. It is assumed that the following four sources of organic inputs are of greatest importance for shifting cultivation as practiced by the Mangy ans of Sto. Tomas:  1. Litterfall during the fallow period, primarily consisting of leaf litter, contributes approximately 6-12 t ha'1 (dw) per year on a lowland tropical forest site (Sandhu et al., 1990; Sajise et al., 1979; Vitousek, 1984; Anderson and Swift, 1983). Leaf litterfall is greatest during the dry season.  2. Slash additions resulting from clearing done prior to the planting of annual crops provide important quantities of P, K, Ca and Mg following burning, although N and S are generally lost during combustion (Nye and Greenland,  85  1960:68). Increases in pH following burning also result in increased nutrient availability which is particularly important for P on acid soils.  3. Weeds and thinned coppice shoots are left as a surface mulch 1-2 times during the cropping period. This can be a particularly important source in weedy sites (i.e. short fallows) or in fields with fallow species that are heavy seeders. Fallow rotations in Sto. Tomas are generally short (e.g. < 4 years) and are dominated by heavy seeders (leucaena and Chromolaena odorata).  4. Crop residues left at the end of the rainy season provide substantial organic inputs used by the fallow vegetation. Rice straw alone contributed between 1 and 9 t dry matter ha"1 in study area field following the rice harvest in October November.  5. Soil movement due to water erosion is similar between fallow types. Both the Chromolaena and leucaena fallows quickly revegetate the site following cropping. By the time the rainy season begins (May-June) the site is generally fully occupied in both fallow types. Under the present system of management, vegetation, including the leucaena, rarely exceeds 15 m in height and the understory is seldom completely suppressed. Below-ground biomass was not included due to a lack of data in the literature and the fact that this component is not easily manipulated through management.  86  Root biomass  may play an important role in nitrogen additions from both the leucaena and Chromolaena fallows, although the limited literature available on this subject suggests that these contributions are relatively small. Palm et al. (1988) reported that after 84 days, a Sesbania sesban green manure crop accumulated 83 kg N ha"1 in above ground biomass and 9 kg N ha"1 in below ground biomass. S. sesban is a tropical tree legume with many characteristics similar to those of leucaena. They also found that the low nitrogen content and slow mineralization of Sesbania roots provided little N to a subsequent rice crop. It is likely however, that below-ground litter additions play an important long-term role in maintaining soil organic matter levels.  Results and discussion Nutrient contents of primary fallow species The two primary fallow species in Sto. Tomas are leucaena and Chromolaena odorata, based upon observations of frequency in the chronosequence plots and in ocular observations in fallow fields in the study area. Leucaena is the most common of 15 woody species found in study area fallow fields. Foliar N, P and K were all higher in the leucaena foliage than in foliage of associated tree and shrub species (Table 7.1). There was no difference in foliar Ca, Mg or in the concentrations of N, P, K, Ca or Mg in stem or branch wood.  87  Table 7.1 Nutrient concentration in primary woody fallow species (eg g"1) Material  N  P  K  Ca  Mg  Leucaena foliage  3.99  .15  2.14  .65  .25  Mixed species foliage  2.18  .11  1.48  .78  .29  Significance  **  *  *  NS  NS  Leucaena wood22  .52  .05  .54  .15  .09  Mixed species wood  .57  .05  .67  .16  .08  Significance  NS  NS  NS  NS  NS  ** = significant at p=.01, *= significant at p = .05  Fresh Chromolaena odorata plants have high nitrogen, phosphorus and potassium contents, both in the leaves and stem (Table 7.2). Foliar P content is more than three times greater than that found in leucaena. Herren-Gemmill (1991) found the contribution of C. odorata to nutrient stocks in fallow vegetation is also substantial, and exceeds its contribution of biomass (Table 7.3), particularly for N, P and K.  22  Stem and branch wood were collected and analyzed separately. An analysis of variance for all measured nutrients detected no significant differences. These values and all subsequent means for wood nutrients represent both branch and stem wood samples combined. 88  Table 7.2 Nutrient concentrations in Chromolaena odorata from five sites in the study area (in eg g"1 of dry mass) Plant part  N  P  K  Ca  Mg  Leaves  3.72  0.51  2.64  0.09  0.24  Stems  2.74  0.23  2.93  0.1  0.09  Whole plant average  2.42  0.33  2.43  0.14  0.18  Table 7.3 Contribution of Chromolaena odorata to nutrient stocks in natural bush fallows in Southwestern Nigeria Element C N P Ca Mg K  Total nutrient stocks Contribution of C. odorata (g m"2) (eg g1) 477.9 50.5 13.3 64.5 1.2 73.9 3.1 24.8 1.3 52.8 14.8 74.2  Adapted from Herren-Gemmill, 1991  Nutrient budgets at the time of harvest Nutrient budgets for each of the nine fields in the grain harvest studies showed that, at the time of harvest, nitrogen was predominantly found in the soil, with relatively little removed in the grain (Appendix Table 4.1). In contrast, nearly 50 eg g"1 of (extractable P + total plant tissue P) was in the biomass at the time of harvest. In general, roughly 25 eg g"1 of the potentially available phosphorus was removed from the system through  89  grain removal. Insignificant amounts of calcium and magnesium were present in the above-ground biomass relative to the pool of these exchangeable bases in the soil.  Differences between fallow types in total system nitrogen were small between six of the nine fields (Figure 7.1), but were greater in fields 2, 4 and 7. Nitrogen was substantially higher in the leucaena treatment in field 4, but lower in the leucaena fallow in fields 2 and 7. Differences in phosphorus were more varied. Phosphorus was unusually high in the leucaena fallow in field number 4 due to high laboratory values for soil P concentration. This is most likely due to sample contamination or analytical error rather than exceptionally high soil P. Phosphorus was also higher for the non-leucaena fallow treatment in fields 2,5, 8 and 10. The differences between fallow types were small for the remaining fields.  Differences between fallow types were generally small for Ca and Mg. Calcium was much higher in the soil under the leucaena fallow in field 1 and was lower in fields 3 and 8. Differences between treatments in the other fields were relatively small. Magnesium was also higher in the leucaena fallow field in field 1. The leucaena fallow in field 4 had substantially greater N and P than the non-leucaena fallow, but was the only field with substantially lower Mg.  90  Note: Error bars represent the standard error of the mean  Net difference (leucaena-non-leucaena) in potentially available nutrients between fallow types  in rice harvest sites  Figure 7.1  I • I  Exchangeable + plant magnesium  Exchangeable + plant calcium  t—  Extractable + plant phosphorus  Total nitrogen  Nutrient distribution changes over time Nutrient budget tables for leucaena and non-leucaena fallows, and the differences between fallow types, are found in Appendix 4. Data from those tables are summarized in Figures 7.1 to 7.4. The only significant pool for Ca and Mg is the soil pool in both the leucaena and non-leucaena treatments at all ages. In the leucaena fallow chronosequence, soil N makes up the majority of total N in the system, although N in leucaena wood is also substantial (52 to 164 kg ha"1). However, phosphorus is the nutrient at greatest risk in terms of loss from the site through harvest removals (Figure 7.2).  The total quantities of elemental P that are potentially available for crop use are low, less than 24 kg ha"1. The largest pool of P in the leucaena fallows of every age is leucaena wood and since the introduction of charcoal making, leucaena wood has become the largest source of nutrient removals - a potentially much more important source than harvested grain. The implications of these removals are discussed in Chapter 9.  Nutrient distributions are similar in the non-leucaena fallow chronosequence for Ca and Mg. Nitrogen in the non-leucaena fallow treatments is largely found in the soil nutrient pool. Again P is the element at greatest risk in terms of potential losses, although to a lesser degree than in the leucaena fallows because of the high proportion of total P in the litter and herbaceous vegetation (Figure 7.3).  92  :  tf&ftratW  r  1400 T 1200 1000 800 600 400 200 f 0 "  m  «MW$m Fallow age  •"•••'••*• w*% n**mgmm  Exchangeable + plant magnesium  Fallow age  Extractable + plant phosphorus  Nutrient distributions in leucaena fallow sites (above-ground + soil)  Fallow age  Hi  I  0  V*"*-*!: •  ^500- w H  ID  t  Figure 7.2  ra O  Exchangeable + plant calcium  Fallow age  Total nitrogen  3500 3000 2500 + 2000 1500  800  • Other species • Leucaena foliage • Leucaena wood D Herbs • Litter PSoil DGrain • Straw  3000 2500 • 2000 • 1500 1000 500 0 2 Fallow age  •• ••  m^^ri f  Exchangeable + plant calcium  Fallow age  i-:i-;:- '••'•  Mi:  TO  I  1400 1200 1000 800 600 400 + 200 0 1  Fallow age  2  W**W«- t:M  •mi  Exchangeable + plant magnesium  Fallow age  Extractable + plant phosphorus  • Straw  D Grain  DSoil  • Litter  DLeucaena wood D Herbs  BLeucaena foliage  mother species  Figure 7.3 Nutrient distribution in non-leucaena fallow chronosequence plots (above-ground + soil)  ! S  TO  Total nitrogen  This potential risk to future crop production excludes P that is not available (i.e. not extractable). Although the fixed P that will with weathering become available will contribute to future crop nutrition, it will do so at a rate that is most likely the same between fallow types. The increased risk comes from increased removals of potentially available P in wood with the removal of the leucaena stems.  At the time of harvest, nearly all of the potentially available P is accumulated in the straw and grain. Most of this is left on the site in the form of straw, but approximately 40% is removed in the grain harvest. In the year 1-4 plots, biomass P was much higher than extractable P in the soil. The largest pool was herbaceous vegetation - reflecting the high P concentrations in Chromolaena odorata biomass. Woody biomass, consisting of both leucaena and other species, is the second largest pool followed by litter.  The net difference in nutrient accumulations between fallow types shows that N and P were substantially higher in the leucaena fallows (Figure 7.4). In the case of nitrogen, these differences were due to soil and wood storage (see Appendix Table 4.4). Phosphorus differences were due to the P in the leucaena wood. Extractable soil P was slightly greater in the leucaena plots, but P levels in the litter and herbaceous vegetation were greater in the non-leucaena fallow. Calcium and Mg were generally higher in the leucaena fallows, but varied greatly.  95  a-  o o  3  3  Fallow age (years)  1  2  Fallow age (years)  2  Calcium  1  4  o c  01  ll  -10 -L  20 x  15 -, 10 5 0 0  Fallow age (years)  2  Fallow age (years)  Magnesium  1  Phosphorus  Net difference (leucaena - non-leucaena) in potentially available nutrients in  1500 1000 500 0  o o o o o  chronosequence sites  Figure 7.4  Z  0)  •a J *  H-  *0)  (0 £  Net difference (kg/ha)  a> o c  0  Nitrogen Net differe (kg/ha)  3  4  Although the sources of soil organic inputs were not studied intensively for this thesis, an estimate was made of the approximate contribution of each source based on nutrient concentrations, regeneration and biomass measurements. From these estimates it appears that the greatest nitrogen additions come from litterfall, followed by seedlings weeded and left as a surface mulch (Table 7.4). All N stored in the above-ground biomass at the time of burning is lost through combustion.  Leaf litter appears to be the most important component of leucaena litter. In a study of leucaena in a dry environment in Varanasi, India, Sandhu et al. (1990) found annual litterfall of 10 t/ha, with approximately 50 eg g"1 occurring during the dry summer months. Leaves accounted for 88 eg g"1 of total litterfall and provided the largest N contribution from the litter (132 kg/ha) compared to fruits (105 kg/ha), twigs (44 kg/ha) and roots (34 kg/ha).  Phosphorus additions from the biomass are distributed in litterfall, ash from the burning of biomass prior to planting and from seedlings applied as a surface mulch. Ash is probably the largest source followed by additions through litterfall. The surface-mulched seedlings are probably the least important source of P additions.  Seedlings present at the time of weeding in leucaena fallows may be as high as 180 seedlings per m2, although population densities of <50 seedlings per m2 appear to be more common. Biomass estimations of small seedlings using the relationship described in  97  Table 7.4 Estimated soil organic and nutrient inputs from litterfall, clearing and weeding in a fully-stocked, 4 year old leucaena fallow23 Input source  Biomass (t ha 1 )  Annual litterfall  Estimated nutrient inputs (kg ha"1) N  P  K  Ca  Mg  5-10  100-200  7-15  100-200  32-64  12-24  Clearing  15-40  0  10-22  na  25-62  15-35  Weeding  0.8-7  32-280  1-10  17-150  5-45  2-18  132-480  18-47  117-350  62-171  29-77  TOTAL  Chapter 3 suggest that the N additionsfromseedling foliage alone is likely in the range of 32 to 280 kg N ha-1 during thefirst2-3 months of the rice crop. Some of the N accumulated in these seedlings is nitrogen derived from the atmosphere through symbiotic fixation, although nodulation does not appear to begin until 4-6 weeks (Ezenwa and Atta-Krah, 1990; Fegbemi and Nwoboshi, 1991). Since most of the seedlings weeded during the rice crop are more than 6 weeks old, it is probable that fixation is responsible for at least part of this N.  These quantities of N, P and K are significant for rice production. For example, rice grain yield responses to N fertilization are frequently 20 kg or more per kg of N applied (FAO, 1984). Phosphorus requirements are approximately 3.5 kg elemental P per ton of grain produced. If the inputs from biomass to the soil noted in Table 7.4 can be  23  Estimate is based on literature values for litterfall (Sajise et al, 1979) and canopy biomass (MacDicken, 1992). Foliar biomass of seedlings was based on unpublished 1990 datafromtwo lm2 quadrats from high and low density regeneration in fully stocked leucaena fallow. 98  maintained, and it appears that they can given current land-use practices, it is likely that nutrient stocks are sufficient to maintain crop yields at the currently levels.  Conclusions Nutrient storage patterns for nitrogen are similar between fallow types, with relatively little N storage in the above-ground biomass compared to the total soil nitrogen pool. The leucaena fallow accumulates more total N, presumably as a result of biological nitrogen fixation by the leucaena. Since leaf litter decomposes rapidly in the tropical environment of the study site, no differences in soil organic matter accumulations were detected.  The leucaena fallow effectively stores a large proportion of potentially available P in leucaena wood, where it is at risk of removal due to harvest for charcoal production. However, the leucaena fallow is able to accumulate nearly two times as much P than the non-leucaena fallow. The non-leucaena fallow stores most of the potentially available P in herbaceous vegetation and in the litter layer, which is likely a very stable storage pool in the form of shifting cultivation practiced in the study area. The implications of P removal due to charcoal production are discussed in Chapter 9.  Leucaena foliage is higher in N,P and K than other predominant tree species in the study area. However, foliage of the predominant non-leucaena fallow species, Chromolaena odorata, is nearly as high as leucaena in N and is substantially higher in P and K. The  99  differences in foliar P are clearly reflected in higher levels of phosphorus in the litter layer under the non-leucaena fallow.  The static inventory conducted in this chronosequence study does not address the complexities of nutrient cycling required to draw confident conclusions. It does however, provide adequate data for the working hypothesis that given the size of nutrient pools and the present levels of soil organic inputs, it appears that nutrients are not likely to be a limiting factor for rice yields in the leucaena fallow unless charcoal production increases. This may not be true for the non-leucaena fallows due to the smaller size of the total nutrient pools for N and P. The differences in grain yields reported in Chapter 5 suggest this difference has important consequences for the well-being of the residents of Sto. Tomas as well as the future sustainability of crop production.  100  Chapter 8 Regeneration Introduction The type and pattern of regeneration are important to the evaluation and selection of fallow management alternatives. The rate at which regeneration covers a site after clearing influences soil erosion rates as well as the pattern of nutrient cycling, both of which have direct sustainability implications. The type of regeneration also limits the range of possible management alternatives.  A series of surveys was conducted in the study area to determine the relative importance of different regeneration types in fallow fields of ages 1 to 4 years.  Methods Three types of surveys were conducted: l)an enumeration of seedlings and living stumps in rice harvest study fields one year after clearing; 2) a stump mortality survey of coppice regeneration in a 300 m2 newly-harvested rice field, and; 3) a complete vegetation survey of a chronosequence of 1, 2, 3 and 4 year-old fallow fields.  In the enumeration survey, the number of live stumps and seedlings were counted in all of the rice harvest study fields with the exception of field 3 which was under cassava during the March 1994 survey. In the stump mortality survey, leucaena stumps were  101  identified, examined for living coppice shoots, measured with a vernier caliper or diameter tape and placed into 5-cm diameter classes.  The vegetation survey used 10 x 10 m plots in leucaena and non-leucaena fallow fields of 1,2,3 and 4 years of age. A total of 16 plots were surveyed (2 fallow types x 4 ages x 2 plots per fallow type and age). Herbaceous vegetation and litter were collected using eight 50 x 50 cm quadrats per plot. These samples were bulked, weighed and subsampled for moisture content and nutrient analyses (see Chapter 7). Herbaceous vegetation of less than 1.5 m in height was included in these samples. All woody vegetation of greater than 1.5 m in height was measured at dbh and the local species name recorded. Basal area and biomass calculation methods are described in Chapter 3.  Results and discussion The March 1994 enumeration survey showed a difference in the primary means of regeneration between leucaena and other woody species (Table 8.1). For leucaena, the ratio of living stumps to seedlings 12 months after clearing was less than 0.4, with the exception of field 9 which had a near equal number of living stumps and seedlings (0.9).  In contrast, for other woody species the ratio of living stumps to seedlings averaged 1.7, indicating coppice regeneration was much more important than seedlings for the nonleucaena species. Most of these species produce coppice shoots more slowly than  102  leucaena and are not thinned as frequently during the cropping season as leucaena. They also produce little seed in most fallow fields.  The stump mortality survey indicated the importance of stump diameter to survival of leucaena stumps (Table 8.2). The ratio of living stumps to dead stumps was 1:2.2. The mean diameter of live stumps was approximately twice that of dead stumps. As noted in Chapter 6, the mean diameter of living stumps in the unburned treatment was less than 40% that of living stumps in the burned treatment. Smaller diameter leucaena stumps were clearly more susceptible to fire than larger stamps.  Table 8.1 Ratio of live stamps to seedlings one year after clearing in rice yield study fields  Field number  Leucaena  Other species  1  .27  .97  2  .10  .35  4  .38  .85  5  .20  .69  6  .14  4.42  7  .14  1.67  8  .04  2.86  9  .89  2.87  10  .03  .33  103  Table 8.2 Survival of leucaena stumps following the 1993 rice harvest in the field adjacent to field 6 Living stumps  Dead stumps  Number per ha  613  1,355  % of total stumps  31.2  68.8  Stump diameter (cm)  16.6  8.9  Vegetation changes during the fallow period Vegetation changes as estimated through the chronosequence plots are described in Table 8.3. In the leucaena fallows, understory vegetation was constant over year 1 and 2 and increased in year 4 (Figure 8.1). This appears to be due to openings in the canopy as indicated by lower population densities and total basal area of both leucaena and other tree species in the leucaena-dominated fallows.  Figure 8.1 Above-ground herbaceous and woody vegetation in chronosequence plots  104  Herbaceous vegetation in the non-leucaena fallows was substantially greater than in the leucaena fallows. Biomass declined in year 2 plots and rapidly increased in year 3 plots. This is consistent with Chromolaena odorata stand development described by HerrenGemmill (1991). This pattern is attributed to rapid vegetative growth in year 1, followed by self-thinning due to competition and reduced vegetative growth due to prolific seed production. Toky and Ramakrishnan (1983) found that Chromolaena dominated shifting cultivations sites in N.E. India for the first 4-5 years. Fallows that extended beyond 5 years were rapidly colonized by bamboos and shade-intolerant tree species.  The total woody biomass in the leucaena fallows was much higher than in the nonleucaena fallows, but fluctuated substantially between chronosequence years. It is unclear whether this difference is due to age, site quality, total fallow age, population of seed trees or other factors. The apparent decline in leucaena population density between years 3 and 4 (Figure 8.2) is likely due to natural mortality and to selection thinning for charcoal production. However, the year 4 plot population density of just over 2,000 trees ha"1 is close to that found in a 6-year old leucaena stand (Chapter 9). Population density of woody species in the non-leucaena fallows remained fairly constant between years, although there was also a substantial decline between year 3 and 4. This suggests that the selection thinning for charcoal and/or natural mortality was similar between fallow types and chronosequence plots, thus satisfying this key assumption for the use of the chronosequence approach (see Chapter 3).  105  Age 1 1 2 2 3 3 4 4 1 1 2 2 3 3 4 4  type NL NL NL NL NL NL NL NL L L L L L L L L  Fallow Species L O L O L 0 L O L 0 L 0 L O L 0  1  (cm) 1.6 1.9 1.2 1.2 2.3 2.3 3.6 2.7 1.7 1.2 1.9 2.0 1.9 1.1 3.6 2.8  DBH per ha 1900 950 2200 1150 700 2900 100 350 15700 1800 7800 1300 11750 4200 2050 350  Stems Foliage 0.1 0.06 0.06 0.02 0.81 0.47 0.97 0.41  Wood 1.93 0.92 1.17 0.34 15.55 9.18 20.9 9.35  Leucaena  ' L = leucaena, 0 = other woody species  (m2) 0.66 0.48 0.38 0.17 0.39 1.77 0.11 0.23 5.42 0.31 3.2 0.5 6.3 1.39 2.72 0.45  BA  Table 8.3 Regeneration summary worksheet  Total 2.03 0.00 0.98 0.00 1.23 0.00 0.36 0.00 16.36 0.00 9.65 0.00 21.87 0.00 9.76 0.00  Height 0.00 4.33 0.00 3.69 0.00 4.71 0.00 5.12 0.00 3.27 0.00 4.28 0.00 3.13 0.00 5.25  Wood 0.00 1.85 0.00 0.34 0.00 5.73 0.00 0.59 0.00 0.70 0.00 1.29 0.00 4.84 0.00 3.01  BIOMASS (t/hal Other:spp.  12.36  25.74  75.65  81.20  87.68  95.69  16.25 10.47  36.56  16.96  73.02  Leucaena 51.06  Percent  0.93  6.90  1.26  3.78  Total wood  -O-NL  Figure 8.2 Population density of woody species in chronosequence plots  Species diversity Common varieties of Leucaena leucocephala were introduced into the Philippines from Mexico in the 16th century and have been naturalized in many locations throughout the country. However, leucaena was not present in northern Occidental Mindoro until the introduction of the giant varieties K8 and K28 in 1976. As a vigorous, prolific seed producer it has spread by natural regeneration outside of the areas initially planted. Most of the conversion of new fallow sites to leucaena now takes place through natural regeneration rather than planting (MacDicken, 1991).  Given leucaena's ability to aggressively dominate the site once established, it is likely that the widespread use of leucaena fallows will result in some impact on species diversity in the Sto. Tomas watershed. However, indigenous species persist in comparable numbers in both the leucaena and non-leucaena dominated fallows (Table  107  8.4). While leucaena biomass clearly dominates in the leucaena fallows (Appendix Table 4.1), it is also clear that indigenous species are able to persist under a predominantly leucaena canopy. It is likely that any species diversity impacts are minimal at present.  Table 8.4 Species diversity in leucaena and non-leucaena fallow plots Fallow age  Mean number of tree species per 100 m2 Leucaena fallow  Non-leucaena fallow  1  11.0  6.0  2  8.0  9.0  3  11.0  13.0  4  9.0  6.0  Mean  9.8  8.5  Conclusions Earlier evaluations of leucaena as a fallow improvement crop indicated the value of coppice shoots in regeneration of the fallow after clearing and cropping. However, the vegetation surveys demonstrate the greater importance of seedling regeneration.  This  has significance where seedling biomass is a potentially important source of available nutrients.  Indigenous tree species in the study area appear to revegetate cultivated sites largely through coppice regrowth. Their ability to survive under heavy competition as indicated by the lack of difference between fallow types in the number of woody species present 108  suggests this coppicing ability provides them with the capacity to compete with species that are much more aggressive.  Herbaceous vegetation remained fairly constant under the leucaena fallow. There was a slight increase in herbaceous biomass in the year 4 plots. The herbaceous vegetation changes in the non-leucaena plots followed patterns cited elsewhere for Chromolaena.  In the leucaena fallow, the protection of stumps may not be important given the contribution of seedling regeneration to revegetation of the site and to nutrient availability to the rice crop. In this context, fire becomes a useful management tool because of the potential benefit of enhanced leucaena seed germination due to seed scarification, without seriously affecting regeneration of the fallow at the end of the cropping season.  109  Chapter 9 Impacts of charcoal production Introduction The reduction or loss of traditional income sources such as rattan gathering and small boat keel production have forced residents of the Sto. Tomas watershed to seek alternative sources of livelihood. Three current alternative sources of cash income are illegal logging, collection of beach rocks for sale as construction materials and charcoal production. Of these three, the most pervasive among households in the village of Sto. Tomas is charcoal production.  Charcoal production is, perhaps more so than any other economic activity in the study area, a potentially critical threat to nutrient stocks in shifting cultivation. Traditionally, very little wood is removed from the site in the form of shifting cultivation practiced by the Iraya (see Chapter 2). The introduction of charcoal making has meant the extraction of large quantities of wood and the nutrients contained in that wood.  An understanding of the impacts of charcoal production is critical to the understanding of sustainability trends in shifting cultivation. Particularly important are nutrient removals due to charcoal making and the implications of expanded regeneration and planting of leucaena as a combination woodfuel source and fallow improvement crop. The extent of charcoal production and the impacts of harvest on nutrient stocks was evaluated through an assessment of production rates, efficiency and wood use rates. 110  Methods Charcoal efficiency was calculated by dividing wood dry-mass of the feedstock entered into the pit by the number of kg of charcoal produced. Wood mass was determined by measuring dbh and length of the wood harvested for one charcoal pit and estimating stem wood biomass using the models in Chapter 3. Data on wood productionfromthe chronosequence studies were used to calculate wood yields for fallow ages 1-4.  Dbh was measured on all leucaena trees on two other leucaena fallow sites. Thefirstsite was owned by Mrs. Loreta Tobias and was estimated by the owner to be 6 years of age. This site was used because it was the only site identified from which no trees had been removed for charcoal production. The second site, owned by Mr. Bino Sanuton, represented a leucaena stand planted in 1987 andfromwhich charcoal had been madefroma heavy thinning in 1992. Biomass was calculated using the models cited in Chapter 3.  Estimates of the nutrient impacts of charcoal removal were made by modeling the proportion of total biomass removedfromundisturbed stands assuming that stemwood of trees with dbh > 5 cm is removed for charcoal production. Nutrient concentrations reported in Chapter 7 were used for these estimates.  Returns to labor were estimated by interviewing four charcoal producers. In these interviews farmers were asked about the amount of time (usually given in days or half-days)  111  required to produce their most recent batch of charcoal. The number of sacks actually taken from the site was used to calculate income and return to labor (US$ per day).  Results and discussion Charcoal production began in the Sto. Tomas watershed in about 1991 (see Chapter 2). It has become a major source of non-crop income and is practiced by 82% of the households living in the village of Sto. Tomas. A similar survey in 1990 indicated that while interest in charcoal making was high, no one was producing charcoal at that time. Total annual production in the study area in 1993 was approximately 5,000 sacks (85 tons), approximately 60% of which comesfromleucaena fallow fields.  The number of batches of charcoal produced and the quantity of charcoal produced per batch varies widely among Sto. Tomas farmers. Of the 14 charcoal producers interviewed, 2 claimed to produce more than 500 sacks per year, 8 estimated they produced between 100 to 400 per year and 4 said they produce less than 100 per year. The batch size generally ranges from 5-60 sacks (85-1020 kg) per pit. The average mass per sack is 17 kg.  The earliest reported prices offered for charcoal in Sto. Tomas were PI5 (US$0.58) to P20 (US$0.77) per sack in 1991. Pricesfluctuatewith time, charcoal quality and negotiating skills of both buyer and producer. Commonly cited prices in September 1993 were P25 (US$0.96) per sack delivered to the seashore for transport. They include the sack, which may be provided by the buyer at a cost of P2-3 per sack. At least two Iraya Mangyan 112  farmers have marketed charcoal directly in Batangas Province and report net prices of P3035 per sack.  Species composition of charcoal produced varies depending on the area harvested. Pits are prepared adjacent to the wood source, used once and abandoned, although in some cases the pits are later re-used. All of the charcoal producers use the earthen pit method, which is the least capital intensive and the least efficient. All of the 14 charcoal producers interviewed stated that leucaena charcoal was of better quality than that from most species available to them within carrying distance to the shore. The primary quality criterion is the friability of the charcoal. Charcoal that breaks apart easily is either not purchased by middlemen or is bought at a lower price. Leucaena charcoal does not break apart as easily as other common species, particularly Albizia procera, also commonly used for charcoal in Sto. Tomas. Leucaena charcoal in the Philippines generally compares favorably to other types of charcoal materials (Table 9.1) and can produce charcoal that meets metallurgical Grade A2 and A3 standards (PCARRD, 1985: 3). The conversion efficiency of leucaena wood to charcoal on a weight basis was 14 eg g"1.24  24  Note: charcoal efficiency is generally calculated on the basis of wood wet weight rather than oven-dry weight. Efficiency based on wet wood weight would have been substantially less than 10 eg g"1. The net efficiency calculations here also include the loss of fines not sold. 113  Table 9.1 Comparative charcoal yield, proximate analysis and heating value from different materials produced in a masonary block kiln (adapted from PCARRD, 1985:40) Heating value (J/E)  Material  Charcoal yield (eg g 1 )  Fixed carbon (eg g"1)  Ash content (eg g"1)  Coconut trunk  25.0  77.8  6.4  26,900  Mixed sawmill waste  31.3  80.5  4.3  28,475  Leucaena  27.4  83.1  2.4  29,230  Nutrient removal impacts due to charcoal making Removals of wood from fallow fields in Sto. Tomas are nearly always stem wood only. Foliage is rarely removed and no evidence of root removal was seen. Charcoal production has potentially significant impacts on nutrient cycling given the distribution of nutrients in leucaena and it's relative importance as a charcoal species in the study area.  The nutrient removal impacts of leucaena wood harvest for charcoal depend primarily on standing biomass in trees with dbh >5-10 cm. For older stands, such as the 6-year-old stand cited in Table 9.2, all of the stems have dbh > 5 cm. These stands would be clearcut for charcoal production and virtually all of the stem wood would be removedfromthe site. This would result in very high nutrient removals (Table 9.3). However, given the current practice of selection thinning in leucaena stands prior to clearing for crop production, it is likely that nutrient removals are not as significant, perhaps with the exception of phosphorus.  114  Table 9.2 Biomass distribution in leucaena stands Stand age  Average dbh (cm)  Stems per ha  6-year-old, undisturbed  12.7  1,339  131.0  173.8  131.0  6-year-old, 1 charcoal removal  6.2  1,173  57.7  82.0  51.0  4-year-old, prior removals  3.6  2,050  120.5  120.0  7.0  3-year-old, undisturbed  1.9  11,750  31.6  32.5  11.7  2-year-old, undisturbed  1.9  7,800  15.9  16.4  1.8  1-year-old, undisturbed  1.7  15,700  31.0  31.8  1.6  Stemwood (t ha 1 )  Total biomass (t ha 1 )  Stem wood > 5 cm (t ha 1 )  Charcoal removalsfromthe 6-year-old fallow with one charcoal removal were approximately 440 kg in 1992, or the equivalent of 3,330 kg ha-1. Using the charcoal conversion efficiency recorded on an adjacentfield,this would require the removal of approximately 23.81 ha"1 of oven-dry leucaena wood. Per ha nutrient removals in this quantity of wood would be 124 kg N, 12 kg P and 128 kg K.  Phosphorus removals in 3 and 4 year old leucaena stands were estimated to be between 16 and 30 (eg g"1) of the potentially available P. This is roughly equivalent to the amount of elemental P removed in the rice grain harvest. In a system in which P is an important  115  1  6 year old stand undisturbed 6 year old stand, 1 removal 4 year old, prior removals 3 year old, undisturbed 2 year old, undisturbed 1 year old, undisturbed  N  % of total' na na 5 6 1 1  kg/ha 87 25 3 6 1 1  P % of total na na 16 30 4 6  K kg/ha 939 275 38 63 10 9  kg/ha 261 76 10 18 3 2  na na 0.3 0.4 0.1 0.1  %of total  Ca  Total biomass nutrients = nutrients in biomass + soil extractable nutrients to a depth of 30 cm  kg/ha 904 265 36 61 10 8  Table 9.3 Estimated nutrient removals through harvesting for charcoal  kg/ha 156 46 6 11 2 1  Mg % of total na na 0.4 0.6 0.1 0.1  limiting factor, these removals may have significant long-term consequences.  However, the risks of critical nutrient removals are in the clearcutting and removal of older stands. Assuming similar soil nutrient status between the 4 and 6 year old sites, the harvest of the six-year-old undisturbed stand of Mrs. Tobias would result in N and P removals that are probably greater than the soil pool of total N and available P. The data are inadequate to draw firm conclusions, but it is certain that clearcutting of older leucaena stands would result in substantial N and P removals. These stands are the type preferred by Sto. Tomas charcoal producers since they have large quantities of wood that make pit construction and wood transport to the pit more efficient than in younger stands where less wood of the minimum diameter is available.  The amount of N removed in the wood is probably not critical to future rice crops - even though the quantities are high. This is due to the fact that if this wood were left and cleared prior to planting of food crops, it would be burned and most of the above-ground biomass N lost. The potential P losses are the most critical in these older stands. For example, in the six-year-old undisturbed stand, the P loss amounts to that removed in approximately 201 of rice grain - or the equivalent of about 5-10 years rice grain production.  Returns to labor Returns to labor were very low due to low charcoal prices, use of simple hand tools, inefficient production methods and long transport distances from pit to seashore. The 117  average amount of time per sack of charcoal is given in Table 9.4. The return to labor at P22.50 per sack (P25 - P2.50 for empty sack) is P20.45 (US$0.82) per day, which is approximately 25% of the official minimum wage and about 50% of the predominant daily wage in northwestern Mindoro. In terms of purchasing power, daily earnings of P20 will purchase perhaps enough rice to provide 2 light meals a day for a family of 5 and nothing more.  Table 9.4 Average time requirements for charcoal production based on farmer estimates of time spent Activity  Man-days per sack  % of total labor  Pit preparation  .17  15  Wood preparation  .25  23  Pit loading/covering  .30  27  Charcoal extraction  .17  15  Bagging and transport  .2  18  1.1  100  TOTAL n=4  Conclusions Charcoal production provides such an important source of cash income in the study area that it is unlikely to disappear in the absence of other economically viable alternatives. At present rates of production, the most important nutrient impacts will be the removal of phosphorus, particularly in clearcut stands. Selection thinning in fallow fields < 4 years old does not likely result in large enough nutrient removals to pose a threat to sustainability.  118  However, the greatest threat to sustainability is clearcutting. The losses of P from the clearcutting of stands >4 years of age are equivalent to several years removal due to rice cultivation. Given the generally low availability of soil P, the removal of P in wood for charcoal making is a real threat in coming years - particularly as human populations increase and the pressures for cash income increase.  It is clear that an improvement in conversion methods (e.g. the use of metal or brick kilns, better combustion rate controls) would increase efficiency and reduce waste. However, as the returns to labor increase with increased conversion efficiency, so do the incentives to increase production rates and therefore nutrient removals.  119  Chapter 10 Farmers, fallows and sustainability  Introduction Indigenous technical knowledge of the management of forest lands is a topic long neglected by governments and development agencies dealing with shifting cultivators and swidden agriculture. Yet in recent years research has demonstrated that in many cases traditional swidden agriculturists possess substantial local technical knowledge that has allowed them to use a complex agricultural system in a manner well-adapted, under certain conditions, to upland environments in the tropics (Warner, 1991:2).  The sustainable management and use of land relates directly to the extent and "quality" of this indigenous knowledge. In most cases, shifting cultivators are not effectively assisted by extension workers nor do they have access to research-generated information in the way lowland farmers might. They depend on knowledge from experience. This knowledge guides how they manage their land, which in turn influences the capacity of the land to sustain agricultural production.  The farmers of Sto. Tomas generally have extensive knowledge of the environment around them. For example, Iraya men can describe the wood or medicinal properties of scores, and, in many cases, hundreds of plant species. They describe soils as "warm" or "cold" depending on texture, relative organic matter content, color and apparent fertility. They can  120  describe in some detail the site requirements, growth patterns and eating qualities of a large number of rice varieties and other crops. Yet, plant knowledge and experience alone are not likely to result in sustainable land-use without innovations in management practices. The ability to generate and evaluate alternatives is an essential step to these improvements. Innovation requires an understanding of the constraints to current practices and to the potential alternatives.  If the ultimate objective of fallow improvement innovations such as the introduction of leucaena as a fallow improvement crop is the enhanced sustainability of shifting cultivation, there must be some understanding of the human and social constraints to traditional and introduced practices. This section describes a series of informal interviews that were conducted to assess some of these constraints. This survey was not definitive in that it produced little quantitative data with enough farmers to provide solid conclusions. For this purpose a more elaborate survey would be needed to obtain a more thorough perception of fanners views.  Methods Interviews were conducted with approximately 50% of the household heads in Sto. Tomas to determine what modifications have been made to the leucaena fallow practice since it's inception in 1977. Interviews were unstructured, informal and generally conducted over several meetings from September 1993 to March 1994. The interview approach followed the guidelines established for the diagnosis and design of agroforestry (Raintree, 1987). 121  Interviews were held with farmers who actively maintain leucaena fallows, and with several farmers from the study area who do not have leucaena fallows. Approximately 70% of the interviews were made during visits to fallow sites. This allowed the farmer-respondents to point out areas they wished to talk about. Interviews were conducted with some 35 individual farmers - 29 men and 6 women, although each interview did not cover all of the topics reported in this chapter.  The most reliable respondents were asked about crop yields over time. These respondents were asked to recall crop yields on the samefieldsfor different crop years. Those who seemed to remember with some confidence were asked the same questions again in a subsequent interview. Only those responses that were consistent in both interviews are reported. Crop yields in 1993 for the ricefieldsincluded in the grain yield study were also given by farmers in local measurement units, as was the amount of seed sown. The rankings of crop yields were compared to yields obtained through the yield experiments using Spearman's rank correlation test.  Questions were asked about charcoal production rates, costs and prices, preferred species for charcoal and thefrequencyof harvest for charcoal. The resultsfromthese interviews are reported in Chapter 9. Often the interviews took several other directions relating to vegetation changes, alternative crops and the specific methods used for crop cultivation.  122  A primary limitation to the data collected through these interviews is the potential reluctance of some farmers to speak openly of problems in the leucaena-based fallows. Although no outward evidence of this was seen, the researcher's long association with leucaena in Sto. Tomas has left the impression among some that negative statements about leucaena might "embarrass" the interviewee. An attempt was made throughout the interviews to be as open as possible to any response without advocating any particular practice.  Results and discussion Changes in crop yields over time Three respondents provided consistent, confident responses to questions about rice yields over time. In each of these selected cases, the samefieldwas reported using the same rice variety. Rice variety use in Sto. Tomas tends to be stable over time. Once a variety has been adopted by a farmer, it is generally retained for many years.  Mr. Reynol Sanuton provided estimates that fit an expected pattern of yield decline when a field is cropped using short fallow periods (Figure 9.1). Upland rice yields declined over three successive crops, as did the yield per unit seeded (kg yield/kg seeded). He reported that thisfieldwas colonized by Imperata cylindrica in 1975 and was not cropped thereafter.  123  A second respondent, Mr. Bino Sanuton, spoke of how their forefathers had taught them to farm in ways that were no longer suitable. He attributed this to encroaching weeds. His  Legend kg harvested yield per unit seeded  i s CL  n>  a.  Figure 10.1 Estimated rice yields in non-leucaena fallowfieldof Mr. Reynol Sanuton recollection of crop yields on the samefieldwere limited to thefirstand last crops planted on thisfield(Figure 9.2)25. The ratio of rice yield to seed planted declined over this 22 year period declined by 50% from around 40:1 in 1970 to 20:1 in 1992.  25  Eight farmers said they could confidently recall details of their first crop as a farmer with his own field. 75% of these respondents (6 of 8) could remember the variety planted, the amount of seed sown and the number of sacks harvested. 124  800-  40  Legend Grain harvested (kg/ha)  700  35  Yield per unit sown  I  600-  30  500-  25  < Q. T3 to —i  •o  B <n  20  JO to to D. CD  CO  c '5  300  I—  o  C  400  200  - ro  roo  -5  a. ca  1992  1970  Year Figure 10.2 Estimated rice yields in non-leucaena fallow field of Mr. Bino Sanuton  Mr. Tamayo Sanuton recalled rice yields in a field that began in natural bush fallow with a high percentage of Chromolaena odorata that was converted to a leucaena fallow in 1990 (Figure 9.3). Although several farmers reported yield increases in leucaena fallow fields,  125  Mr. Sanuton was the only farmer able to recall specific yield values before and after the vegetation conversion. Rice yields were slightly higher after the conversion to leucaena.  Of the farmers who could not recall specific yields with any consistency, the opinion of 82% of those interviewed was that yields were declining. Several farmers attributed this to conversion of secondary forest to grasslands and to shorter fallow periods. One man reported his parents said there used to be fewer weeds than there are now - a view both he and his wife shared.  so  200  Legend —-H"—-  Grain harvested (kg/ha)  150 Q. •o  CD J * 30  100  (A 0) CO  .  _c 'co  ^  ^  •  11—  i_  O  in CD CD Q. (D Q.  50  10-  1988  1991  1993  Year  Figure 10.3 Estimated rice yields in fallowfieldof Mr. Tamayo Sanuton converted to leucaena fallow in 1990 126  One important aspect of sustainability in an environment such as Sto. Tomas is human health. Lower crop yields are only one part of a cycle that reduces the overall well-being of shifting cultivators. In Sto. Tomas, where malaria, typhoid, amoebic dysentery, measles and a host of other illnesses are commonplace, food production and the sustainability of production are closely intertwined. The implications of reduced swidden crop yields on sustainability are both direct (i.e. reduced quality of life through increased lack of food/income), and indirect through increased susceptibility to illness and the lack of ability to seek medical treatment. Illness then contributes to decreased crop yields through reduced ability to tend crops and livestock.  Rice crop yields for the 1993 season were reported by farmers on a whole-field basis by variety. Table 10.1 describes area planted andfreshweight crop yield for thosefieldsand rice varieties in the rice harvest experiments. Wholefieldyields were substantially lower than those reported in the grain yield trial, with yields ranging from .5 to 4.3 t ha"'. These reports are subject to a substantial amount of error, primarily since farmers presently rely to a large extent on labor of non-family members to assist in the harvest. These workers receive a portion of the grain they harvest (usually 20 to 25%) as wages26. This grain was generally not included in the reported grain yields. The ranking of sites by grain yield according to farmer reports and the moisture content-adjusted grain yield was similar, with a Spearman's r of 0.67.  26  Although this is a traditional practice among the Iraya, the degree to which the work is done by people outside the nuclear family has increased since children began attending school in the late 1970's. 127  Table 10.1 Farmer reported sowing rates and grain yields  Field number  1 2 4 5 6 7 9 10  Seed sown Grain harvested Seed sown (kg) (kg) to grain harvested  1.3 3.3 2.6 5.2 5.2 6.5 5.2 3.9  109.2 54.6 54.6 163.8 54.6 122.9 245.7 109.2  84.0 16.8 21.0 31.5 10.5 18.9 47.3 28.0  Area planted (m2)  Yield (t ha 1 )  256 640 512 1024 1024 1280 1024  768  4.3 0.9 1.1 1.6 0.5 1.0 2.4 1.4  Perceptions of leucaena Perceptions of leucaena varied substantially among farmers. Most felt there were soil fertility benefits, although it was difficult to know how much of this reflected suggestions the researcher had provided in years past. Everyone interviewed felt it was a useful tree for charcoal production. Those farmers who had the fewest leucaena stands were the most likely to see cash income through charcoal as the greatest benefit.  Some claimed leucaena fallows increase the size of rice spikelets. Others noted color or rice height differences in adjacent fields of leucaena and non-leucaena fallows. Some farmers noted varietal differences in the response to leucaena fallows. Pulang balat (a medium maturity variety) and dinalaga (a high eating quality, late maturing variety) were reported to have excellent response while others such as red or white bolohan (medium-late maturing varieties) lodge frequently.  At least two farmers saw a direct link between fallow length  and degree of lodging. One farmer suggested that fallow lengths should be 2-3 years on "flat" lands and 3-4 years on sloping lands. These statements were made in early September  128  prior to any of the rice harvest studies and proved to be an accurate management prescription based on the data.  The reasons given for soil improvement by leucaena were varied. Several people thought the fact that leucaena leaflets close at night was important because it increases the infiltration of dew to the rice plants. Others felt the sap was the main source of improvement. Perhaps the most perceptive comment was that leucaena does not "fight" the rice crop.  Two farmers felt leucaena had negative effects associated with insects. The leucaena psyllid (Heteropsylla cubana) was a serious pest in the mid-1980's, disappeared by the late 1980's and in 1993 was found in small numbers in roughly half of the stands visited. One farmer, a Visayan immigrant, kills his leucaena by girdling because he believes the psyllid attacks his rice crop. The other, who is also the Barangay Kapitan27, was convinced that psyllids had indirectly damaged his rice crop by exuding a sticky substance that attracted ants and eventually resulted in a rust infection and reduced yields (fields 7 and 8).  When leucaena seedlings overtop the young rice plants, psyllids were reported to jump from the leucaena foliage onto the rice. The Barangay Kapitan was the only participant in the rice harvest study to be late in weeding - due to illness and other reasons. Leucaena seedlings grew to at least 50 cm before weeding when psyllids rested on the rice plants. All  27  Senior elected official at the village level 129  of the farmers interviewed, including the Barangay Kapitan, noted that when weeding was done before the leucaena reached 25 to 40 cm, there were no shading or psyllid problems.  Some farmers also noted rat damage to leucaena, particularly in the dry season when rats remove the bark apparently to get at the moist cambium and vascular tissues. In the most severely affected stands it appeared that about 5% of the trees were affected.  Farmer-initiated adaptations The original design of this system called for the use of stump cuttings which were pruned to 5-10 cm above and 20 cm below the root collar (MacDicken, 1981). As previously described (MacDicken, 1991), farmers preferred taller stump cuttings and successfully used stumps with tops of 100-130 cm above the root collar. The taller cuttings had the advantage of sprouting well above the rice crop, allowing the newly emergent foliage to grow in full sunlight without shading the rice or maize.  Most farmers who have established leucaena in natural bush fallowfieldssince the mid1980s have done so by broadcasting seed at about the same time upland rice is sown. Quantities of <1 kg are generally broadcast in upper slope positions with the expectation that natural regeneration will move the leucaena further down the slope with time. Only 7% of the respondents had attempted to convert entire fields at one time. This use of natural regeneration has proven to be the least labor-intensive method of establishment, and is presently in use by all of the respondents intentionally expanding their leucaena fallows. 130  Only one respondent reported recent use of bareroot wildlings as a means of establishment. His use was on a limited scale (<20 wildlings) and was done as an informal experiment. He reported successful establishment, but would not likely use this method in the future since broadcasting of seed was easier and almost as effective.  The vigorous seeding of leucaena has often been cited as a potential disadvantage as an intercrop species because of the large number of seedlings produced under leucaena stands. However, in this fallow system, seedlings are used by farmers as an unincorporated green manure. Seed from mature leucaena fallows are scarified during the pre-plantfireand germinate shortly after the initial rains. The rice crop, which is planted at about the time of the first reliable rainfall, germinates at about the same time but grows much more quickly than the leucaena. The leucaena seedlings are pulled or cut at the time of thefirstweeding and left as a surface mulch.  Costs of leucaena fallows Only two additional costs of the leucaena fallow were noted in any of the interviews, both related to weeding. Thefirstcost is the additional weeding required in leucaena when weeding is delayed. All of the respondents said that weeding leucaena seedlings was less time-consuming than weeding of other species when leucaena seedlings were less than about 25-40 cm in height. Above that height, the seedlings become nearly impossible to pull and need to be cut - a less preferred weeding method.  131  The second cost was that of removing coppice shoots during the rice crop. Since some leucaena stumps resprout during the rice crop, the thinning or removal of coppice shoots is necessary to prevent shading. Most farmers cut back the coppice shoots periodically - at weeding times and whenever they come across shoots long enough to cut. Several farmers reported removing the shoots 3-4 times per season. Other vigorous coppicing species such as Albiziaprocera are also present in fields in the Sto. Tomas watershed, but are fewer in number than leucaena.  Leucaena as a species in the fallow costs less in time spent for weeding and management when weeding is done early. Several farmers said the Chromolaena fallows have more weeds and take longer to weed than the leucaena fallows. Most respondents try to weed early in the rice crop for the purpose of increasing rice yield and to make the job as easy as possible.  Most farmers (92%) said they are either expanding their leucaena plantings by broadcasting seed or are allowing natural regeneration to expand the area. Only one farmer - the Visayan immigrant who kills his trees by girdling - reported he would not expand his plantings because of the perceived danger to the rice crop.  Conclusions Farmers in Sto. Tomas report declining crop yields and relate these declines directly to reductions in forest cover. They also face declining labor availability and increasing 132  population pressures. Traditional income sources from the harvest of forest products have declined for most farmers. The only current replacement source of traditional wood products is charcoal production which is now widely spread. This decline and the increasing pressures on land are recognized by many farmers - most of whom have no alternative but to continue the present form of shifting cultivation.  Leucaena fallows are recognized by approximately 92 % of the farm families as providing a net positive influence on rice crop yields. This, plus the increased utilization of leucaena wood for charcoal has led to increases in the amount of land intentionally managed for leucaena production - although the extent of this increase is difficult to quantify. Leucaena has increased sustainability through increased crop yields, thus reducing the pressure to further expand the swidden area.  However, the improvement of soil fertility is insufficient to guarantee the sustainability of production in areas such as the Sto. Tomas watershed. Labor inputs must be available to provide a level of management that can effectively use increased nutrient supplies. In Sto. Tomas, the education of children in primary school (grades 1-6) has substantially reduced the pool of labor for weeding. A second critical factor is illness, which also has significant negative effects on the amount of time farmers can spend infieldmanagment and on the vigor with which they work in the field. Increased efforts to remove the sources of typhoid, malaria and dysentary plus the availability of medical treatment would have direct impacts on the sustainability of crop production in Sto. Tomas.  133  Chapter 11 Conclusions Increasing human populations, declining crop yields and reductions in per capita arable land suggest that shifting cultivation as traditionally practiced in the Sto. Tomas watershed is no longer a sustainable practice. The use of Leucaena leucocephala as a fallow improvement crop has helped to increase crop yields and may be an important part of a land-use strategy to maintain or improve crop yields in an area where natural resources are rapidly declining.  The overall goal of the research described in this dissertation was to answer questions related to the sustainability impacts of a nitrogen-fixing tree species fallow in upland ricebased shifting cultivation. The conclusions drawnfromthe research described in this thesis are described here in terms of answers to the key research questions posed in Chapter 1.  Response to key research questions 1. Are grain yields and the soil properties that limit production of upland rice significantly different between leucaena and non-Ieucaena fallows?  Yes. Grain yields were higher by an average of 42% following leucaena fallows. This difference appears to relate most closely to nitrogen as evidenced by total soil N accumulated during the fallow period, N added through seedling biomass applied as a surface mulch during the cropping period and higher plant N in the rice crop following the leucaena fallow. The lack of significant soil N differences between fallow treatments at the 134  time of harvest and the higher grain yields following leucaena fallows suggest the rice crop effectively used the increased N from leucaena.  Available phosphorus levels were low, but did not appear to differ between fallow types, although the size of phosphorus pools in the biomass and litter varied with fallow type. Ca and Mg were not limiting in either fallow type due to the large soil pool of these nutrients.  2. Does leucaena contribute significantly greater amounts of nutrients to the upper horizons of the soil than non-leucaena fallows?  Yes. The chronosequence studies suggest that total soil N accumulates at higher rates in the leucaena fallows than in the non-leucaena fallows. Primary reasons for this are. greater N inputs resulting from N-fixation of leucaena and possibly the inhibition of nitrification of ammonium ions as demonstrated by higher NH4" concentrations in soils under leucaena. Total soil N differences disappeared by the time of rice harvest. The primary sources of soil organic inputs are litterfall and weeding of leucaena seedlings. The contributions of weeded seedlings may be of greatest importance to the rice crop since these additions are generally made during the first 4-8 weeks after planting. Optimum N fertilizer applications in rice are generally split to increase uptake and efficiency. The accumulated soil N from litterfall provides a basal N application and high C:N biomass from seedlings provides side-dressed nitrogen during the first two months after planting.  135  Phosphorus contributions are higher in the leucaena fallows when leucaena wood removals from the site are low. Phosphorus additions take place through litter, ash left on the soil surface at the time of burning and through surface-mulched seedlings.  3. Do these nutrient additions contribute to significant increases in subsequent crop growth? Yes. The higher grain yields following the leucaena fallows are evidence of increased nutrient contributions compared to the non-leucaena fallow. This appears to be due to foliage that is higher in N, P and K than other woody fallow species in the same area. Leucaena also has foliage that is slightly higher in N and and Ca than Chromolaena odorata, although leucaena foliage is lower in P and K than chromolaena.  Increases appear to be stable across a wide range of rice varieties. The yield experiment included short, medium and long-duration varieties of short, medium and tall stature. It also included one glutinous variety. Although the interaction between fallow type and sites or variety was not statistically significant, the differences between fallow type ranged from 19106%.  4. Is it likely that production of upland rice in shifting cultivation is more sustainable over time following leucaena fallows than non-leucaena fallows? Yes. Leucaena fallows accumulate greater quantites of N, primarily derived from the atmosphere. Examination of leucaena root nodules at the peak of the dry season (when  136  moisture stress is at its maximum level) found leghaemoglobin present and nodules that appeared to be fixing nitrogen. Given leucaena N-fixation research elsewhere, it is likely that the proportion of nitrogen derived from the atmosphere is high (e.g. 60-80% of total N uptake). This N is obtained with very little capital cost since most of the plantings are established through broadcasting of seed or through natural regeneration with no use of inoculum. There is also no increase in management costs incurred by farmers. N-fixation is likely maintained at fairly high levels over time since periodic cropping with rice effectively reduces soil N levels so that N-fixation is not inhibited.  Increased crop yields reduce the pressure to expand the size of swidden fields, therefore allowing greater concentration of weeding/tending efforts on a smaller area.  The greatest potential threat to sustainability of crop production following leucaena fallows is charcoal production and the potential losses of P from the system. Low returns to labor do not discourage farmers from producing charcoal, primarily due to a lack of alternative sources of cash income.  Management implications Proper management will be required to convert larger nutrient pools in the leucaena fallows into economic yield. The following management considerations appear to be important to make use of the potential benefits of the leucaena fallow.  137  Optimum leucaena fallow age The highest grain yields were generally in the fields with the lowest land utilization intensity index (R) values (largest fallow:cropping period ratios). It appears that greater N accumulation begins in fallow ages of about 3 years, which most likely explains generally higher grain yields in older fallows. A minimum fallow age of 3 to 4 years would likely provide greater N for crop growth than younger fallows. If substantial quantities of wood are not removed, it is likely that longer fallows (e.g. 6-7 years) would result in higher grain yields. These estimates are based only on soil N accumulations. In practice, optimum fallow length calculations by farmers often include a wide range of other considerations such as land, labor and capital availability.  Limited removal of large stems The potentially important P losses through removal of larger diameter stems for charcoal making suggests that these removals should be limited if long-term sustainability is to be achieved.  Burning There appears to be little or no disadvantage to burning leucaena fallow fields prior to planting. The most significant advantages of burning appear to be increased P availability and reduced weeding costs, although there appears to be no difference in grain yield due to burning.  138  Weeding The importance of early weeding of leucaena seedlings and the thinning of coppice shoots was consistently noted by farmers. The impact of delayed weeding was demonstrated in low grain yields in fields 7 and 8. Weeding needs to take place within the first 4-6 weeks or before the leucaena seedlings reach a height of 25-40 cm. Weeding not only reduces competition with the rice crop but also appears to reduce the risk of psyllid-related damage to the rice crop and provides a high quality surface mulch.  Grain moisture content The differences in rice maturation indicate a need to delay the harvest of rice in leucaena fallowed fields to maximize grain yields. Grain moisture content diferences between crops grown in leucaena and non-leucaena fallows may not be readily apparent to farmers who monitor crop maturity based on experience with crops under lower N content soils. Care needs to be taken to harvest at a time when grain moisture content is 18-22%, recognizing the periodic need to harvest earlier than this due to the arrival of typhoon-strength winds and rain during the harvest season.  Varietal susbstitution The giant leucaena varieties used in Sto. Tomas (K8 and K28) produce much higher wood yields and less seed than the common (Hawaiian-type) leucaena varieties. This has led to their pan-tropical use for wood and foliage production. However, given the value of seedling regeneration in leucaena fallows, it is highly likely that the common varieties  139  would be better suited to sustainable production than the giants. Lower wood yields on stems of smaller diameter would result in fewer wood removals, and hence lower phosphorus removals. The higher seed production rates of the common varieties would result in more seedlings in the early stages of crop growth, providing more N-rich foliage as a green manure.  Crop varieties An unknown portion of the site-to-site differences in grain yield and the response to the leucaena treatment can be attributed to varietal differences. Increased lodging is a risk in the higher N environment following a leucaena fallow. A logical management practice would be to use short-statured varieties that respond well to N fertilization. Given the generally higher eating quality of traditional upland rice varieties, emphasis should be placed on traditional varieties.  Future research needs 1. Permanent plot studies. The use of chronosequence plots in these studies has been a useful means of reducing the time and resource requirements for the research described in Chapters 4, 7-9. However, the variability in some measurements (e.g. woody biomass, population density) demonstrate the limitations of this approach. To fully understand the dynamics of leucaena vs. non-leucaena fallows, permanent plots must be established and monitored. This should include regular, periodic monitoring of changes in vegetation, litter  140  and soil nutrients. Measures of available soil N (e.g. extractable nitrate and ammonium) should be monitored throughout the fallow period and cropping season.  2. Soil organic inputs. While it is apparent that the primary sources of organic inputs are litterfall, biomass left on the site at the time of clearing and the weeding of seedlings, the relative importance of these three sources in leucaena fallows is still unclear. An important next step will be to monitor and quantify each of these for the purpose of determining their relative importance to nutrient availability. This includes root turnover and carbon allocation studies.  3. Role of Chromolaena odorata in P cycling. Chromolaena appears to be an effective accumulator of P. Phosphorus is the element most likely to be limiting in the study area. The use of Chromolaena on a short-rotation fallow might be a useful means of increasing available P. The combination of this species and leucaena might be a low-cost means of improving plant nutrition without external inputs. However, the inverse relationship between straw P concentrations and grain yield suggest additional research is necessary.  4. Other species. Clearly Leucaena leucocephala is not the only nitrogen fixing tree useful for fallow improvement, although the list of species for this purpose is still limited. Future research on this approach to fallow improvement should include additional species such as Sesbania sesban, Calliandra calothyrsus and Gliricidia sepium.  141  5. Effects of leucaena on crop development and maturation. The grain moisture content differences observed in these studies are significant in that they are a potential source of error. Future grain yield research will need to carefully monitor grain moisture content and the timing of harvest to ensure treatments are harvested at the same moisture content stages.  6. Rice varietal trials. It is highly likely that the rice varieties used in Sto. Tomas respond differently to leucaena fallows and that they have different yield potentials. These varieties should be placed in a multi-location varietal trial.  142  References  Anderson, J.M. and M.J. Swift. 1983. Decomposition in tropical forests, in S. Sutton, T. Whitmore and A. Chadwick (eds.) Tropical rain forest: Ecology and management. Blackwell Scientific Publications, Oxofrd. pp. 287-309. Brinkmann, W.L.F. and J.C. De Nascimento. 1973. The effect of slash and burn agriculture on plant nutrients in the Tertiary region of Central Amazonia. Turrialba 23(3): 284-290. Buol, S.W., F.D. Hole andRJ. McCracken. 1980. Soil genesis and classification. Iowa State Univ. Press, Ames. Iowa. Second Edition. 404pp. Cole, D.W. and H. Van Miegroet. 1989. Chronosequences: a technique to assess ecosystem dynamics. In W.J. Dyck and C.A. Mees (Ed.). Research Strategies for Long-term Site Productivity. Proceedings, JJEA/BE A3 Workshop, Seattle, WA, August 1988. IEA/A3 Report No. 8. Forest Research Institute, New Zealand, Bulletin 152. DeDatta, S.K. 1981. Principles andpractices of rice production. John Wiley and Sons, New York. 618pp. de Guzman, E.D., RM. umali and E.D. Sotalbo. 1986. Guide to Philippine flora andfauna. Vol. 3, Dipterocarps and non-Dipterocarps. Min. of Natural Resources, Quezon City. 414pp. Dickerson, R.E. 1928. Distribution of life in the Philippines. Monograph 21, Bureau of Science, Manila. 322pp. Ewel, J. 1971. Biomass changes in early tropical succession. Turrialba, 21, 110-2. Ewel, J. 1976. Litter fall and leaf decomposition in a tropical forest succession in Eastern Guatemala, Jour, of Ecology, 64:1, 293-308. Ezenwa, I., A.N. Atta-Krah. 1990. Initial growth and nodule development of leucaena and gliricidia. Leucaena Res. Reports 11:99-101. Fageria, N.K., V.C. Baligar and C.A. Jones. 1991. Growth and mineral nutrition of field crops. Marcel Dekker, Inc. New York. Fagbemi, T. and L. Nwoboshi. 1991. Behavior of leucaena nodules in the Southern Guinea savanna zone of Nigeria. Leucaena Res. Reports 12:78-79.  143  FAO and SID A. 1974. Report on regional seminar on shifting cultivation and soil conservation in Africa. Food and Agriculture Organization, Rome. FAO. 1977. Forestry for local community development. FO:MISC/77/22. Rome, Italy, 113 pp. FAO. 1984. Fertilizer and plant nutrition guide. FAO Fertilizer and Plant Nutrition Bulletin 9. FAO, Rome. 176pp. Gillman, G.P. 1979. A proposed method for the measurement of exchangeable properties of highly weathered soils. Aust. J. Soil Res. 17:129-139. Golley, F.B., J.T. McGinnis, RG. Clements, G.I. Child and M. Duever. 1975. Mineral Cycling in a Tropical Moist Forest Ecosystem. Univ. of Georgia Press, Athens, Georgia. Gomez, K.A. 1972. Techniques for field experiments with rice. IRRI, Los Banos. 48pp. Gomez, K.A. and A. A. Gomez. 1984. Statistical procedures for agricultural research. John Wiley and Sons, New York. 680pp. Greenland, D.J. and J.M.L. Kowal. 1960. Nutrient content of the moist tropical forest of Ghana. Plant and Soil, 12, 154-74. Greenland, D.J. 1975. Bringing the Green Revolution to the shifting cultivator. Science 180: 841-844. Herren-Gemmill, B. 1991. The ecological role of the exotic asteraceous Chromolaena odorata in the bush fallow farming system of West Africa. Biotrop Special Publication No. 44 (11-22). Holm, L.G., D.L. Plucknett, J.V. Pancho and J.P Herberger. 1977. The world's worst weeds. East-West Center. Honolulu. 609pp. International Institute of Tropical Agriculture (JITA). 1980, Annual Report for 1979. Ibadan, Nigeria. IRRI. 1975. Major research in upland rice. Los Banos, Philippines. 255pp Jaiyebo, F.O. and A.W. Moore. 1964. Soil Fertility and Nutrient Storage in Different SoilVegetation Systems in a Tropical Rain-Forest Environment, Trop. Agriculture (Trin.), Vol. 41:129-139. Jenny, H , S.P. Gessel and F.T. Bingham. 1949. Comparative study of decomposition rates of organic matter in temperate and tropical regions. Soil Sci. 68, 419-32.  144  Juo, A.S.R. and R. Lai. 1977. The effect of fallow and continuous cultivation on the chemical and physical properties of an Alfisol in Western Nigeria. Plant and Soil, 47, 567-584. Kang, B.T., H. Grimme and T.L. Lawson. 1985. Alley cropping sequentially cropped maize and cowpea with leucaena on sandy soil in Southern Nigeria. Plant and Soil 85:267-277. Kasberg, R.H. undated. The impact of Chromolaena odorata (L.) on swidden cultivation in Southern Mindoro. Mimeo. 12pp. Kawano, K., P.A. Sanchez, M.A. Nurena and J.R. Velez. 1972. Upland rice in the Peruvian jungle, in IRRI, Rice Breeding. Los Banos, Philippines. Kimmins, J.P. 1977. Evaluation of the consequences for future tree productivity of the loss of nutrients in whole-tree harvesting. For. Ecol. Manage. 1:169-183. Kimmins, J.P. 1989. Summary of Session 1. In W.J. Dyck and C.A. Mees (Ed.). Research Strategies for Long-term Site Productivity. Proceedings, EEA/BE A3 Workshop, Seattle, WA, August 1988. JJEA/A3 Report No. 8. Forest Research Institute, New Zealand, Bulletin 152. Laudelot, H. and Meyer, J. Les cycles l'elements mineraux et de matiere organique en foret equatoriale congolaise. Trans. Fifth Int. Cong. Soil Sci. (Comm. II), 267-72. 1954. MacDicken,K.G. 1981. Leucaena as a fallow improvement crop: A first approximation. Paper presented at EWC Workshop on Environmentally Sustainable Agroforestry and Fuelwood Production with Fast-Growing, Nitrogen-Fixing, Multi-Purpose Legumes. East-West Center, Honolulu, Nov. 1981. MacDicken, K.G. 1991. Impacts of Leucaena leucocephala as a fallow improvement crop in shifting cultivation on the Island of Mindoro, Philippines. Forest Ecology and Management 45:185-192. MacDicken, K.G. 1992. 1987 humid zone network trials: Analysis, interpretation and evaluation. Winrock International. Arlington, Virginia. 230pp. MacDicken, K.G. 1994. Selection and Management of Nitrogen Fixing Trees. FAO and Winrock International. 290 pp. MacDicken, K.G., G.V. Wolf and C.B. Briscoe (Eds.). 1991. Standard research methods for multipurpose trees and shrubs. Winrock International and the International Center for Research in Agroforestry. Arlington, Virginia, USA.  145  Micosa-Tandug, L. 1986. Biomass prediction equations for giant ipil-ipil (Leucaena leucocephala (Lam.) de Wit). Sylvatrop Philippine Forest Research Journal 11(1-2): 122. Mueller-Harvey, I., A. Juo and A. Wild. 1989. Mineralization of nutrients after forest clearance and their uptake during cropping, in J. Proctor (Ed.) Mineral Nutrients in Tropical Forest and Savanna Ecosystems. Blackwell Scientific Publications, Oxford, pp.315-324. Nakano, K. 1978. An ecological study of swidden agriculture at a village in Northern Thailand. South East Asian Studies 16:411-446. Nangju, D. and S.K. DeDatta. 1970. Effect of time of harvest and nitrogen level on yield and grain breakage in transplanted rice. Agron. J. 62:468-474. Nye, P.H. 1961. Organic matter and nutrient cycles under moist tropical forest. Plant and Soil 13:333-46. Nye, P.H. and D.J. Greenland. 1960. The Soil under Shifting Cultivation. Tech. Comm. No. 51, Comm. Bureau of Soils, Harpenden, England. Nye, P.H. and D.J. Greenland. 1964. Changes in the Soil After Clearing Tropical Forest, Plant and Soil 21(1): 101-112. Parfitt, R.L. 1976. Shifting cultivation - How it affects the soil environment, Harvest, Vol. 3, No. 2. Palm, O., W.L. Weerakoon, M.A. DeSilva and T. Rosswall. 1988. Nitrogen mineralization of Sesbania sesban used as green manure for lowland rice in Sri Lanka. Plant Soil 108:201-209. Philippine Council for Agriculture and Resources Research and Development (PCARRD). 1985. The Philippine recommends for fuehvood and charcoal utilization. Technical Bulletin Series No. 56. Los Bafios. 95pp. Puri, S., S.R. Gupta and B.B. Bhardwaj. 1992. Litterfall quantity and decomposition rate in a Leucaena leucocephala plantation on a saline soil. Leucaena Research Reports 13:4042. Raintree, J.B. (Ed.). 1987. D&D User's Manual. An Introduction to Agroforestry Diagnosis and Design. International Centre for Research in Agroforestry, Nairobi. 110pp. Rambo, A T . 1981. Fire and the Energy Efficiency of Swidden Agriculture. EAPI Reprint, EastWest Center, Honolulu.  146  Read, M.D., B.T. Kang and G.F. Wilson. 1985. Use of Leucaena leucocephala (Lam. de Wit) leaves as a nitrogen source for crop production. Fertilizer Research 8: 107-116. Russell, E.W. 1980. Soil conditions and plant growth (10th edition). Longman, London. 849pp. Russell, W.M.S. 1988. Population, swidden farming and the tropical environment. Population and Environment 10(2):77-94 Ruthenburg, H. 1980. Farming Systems in the Tropics. Clarendon Press, Oxford, England. Sajise, P.E., V. Reyes, V. Cudera and R. Guce. 1979. Litterfall studies of five upland ecosystems. J n P.E. Sajise and R. Raros (Eds.). Biophysical characterization of upland ecosystems at Mount Makiling, Calamba, Laguna, Philippines. Upland Hydroecology Program, University of the Philippines at Los Bafios. Sanchez, P. A. 1976. Properties and Management of Soils in the Tropics. J. Wiley and Sons, New York. Sanchez, P.A., C.A. Palm, C.B. Davey, L.T. Szott, and C.E. Russell. 1985. Tree crops as soil improvers in the humid tropics? in M.G.R. Cannel and J.E. Jackson (eds), Attributes of trees as crop plants. Institute of Terrestrial Ecology, Huntingdon, England. Sandhu, J., M. Sinha and R.S. Ambasht. 1990. Nitrogen release from decomposing litter of Leucaena leucocephala in the dry tropics. Soil Biol. Biochem. 22(6): 859-863. Sanginga, N., K. Mulongoy and M.J. Swift. 1989. Contribution of nitrogen by Leucaena leucocephala and Eucalyptus grandis to soils and a subsequent maize crop, j n Proceedings of a regional seminar on trees for development in Sub-Saharan Africa, February 20-25, 1989, Nairobi, Kenya. International Foundation for Science, Stockholm, pp. 253-258. Schult, V. 1991. Mindoro: A social history of a Philippine Island in the 20th century. Divine Word Publications, Manila. 213 pp. Semb, G. and J.B.D. Robinson. 1969. East Africa Agric. For. Jour. 34:350. Soil Survey Staff. 1975. Soil Taxonomy. Agriculture Handbook 436. Soil Conservation Service, Washington, D.C. 754pp. Soil Survey Staff. 1990. Keys to Soil Taxonomy. SMSS Technical Monograph No. 6, Fourth Edition, Blacksburg, VA.  147  SYSTAT, Inc. 1992. Systatfor Windows: Statistics, Version 5 Edition. Evanston, IL. USA. 750 pp. Szott, L.T., C.A. Palm and PA. Sanchez. 1991. Agroforestry in acid soils of the humid tropics. Advances in Agronomy 45:275-301. Toky, O.P. and P.S. Ramakrishnan. 1983. Secondary succession following slash and burn agriculture in North-Eastern India. Jour, of Ecology 71:735-745. Turvey, N.D. and P.J. Smethurst. 1989. Apparent accumulation of nitrogen in soil under Radiata pine: Misleading results from a chronosequences. In W.J. Dyck and C.A. Mees (Ed.). Research Strategies for Long-term Site Productivity. Proceedings, JEA/BE A3 Workshop, Seattle, WA August 1988. IEA/A3 Report No. 8. Forest Research Institute, New Zealand, Bulletin 152. Unruh, J.D. 1990. Iterative increase of economic tree species in managed swidden-fallows of the Amazon. Agroforestry Systems 11:175-197. Van Den Beldt, R. 1982. Litterfall as a function of population in a 1 year old leucaena (K8) planting. Leucaena Research Reports 3:95. Vitousek, P. 1981. Clearcutting and the nitrogen cycle, in F.E. Clark and T. Rosswall (eds.). Terrestrial nitrogen cycles. Ecol. Bull. 33:631-642. Vitousek, P.M. 1984. Litterfall, nutrient cycling and nutrient limitation in tropical forests. Ecology 65(1): 285-298. Warner, K. 1991. Shifting cultivators: Local technical knowledge and natural resource management in the humid tropics. Community Forestry Note 8, FAO. Rome. 80pp Weaver, Peter. 1979. Agri-silviculture in tropical America, Unasylva 31(126) 1979:2-12. Yoshida, S. 1975. Characteristics of upland rice, in IRRI. Major research in upland rice. International Rice Res. Inst., Los Banos, Philippines, pp. 46-90. Yoshida, S. 1981. Fundamentals of rice crop science. International Rice Res. Inst., Los Banos, Philippines. 269pp. Yoshida, S. and F.T. Parao. 1976. Climatic influence on yield and yield components of lowland rice in the tropics, in International Rice Research Institute. Climate and Rice. Los Banos, Philippines, pp. 471-494.  148  Zinke, P. J., S. Sabhasri, P. Kunstadter. 1978. Soil Fertility Aspects of the Lua' Forest Fallow System of Shifting Cultivation in: P. Kunstadter, E.D. Chapman and S. Sabhasri (Editors) Farmers in the Forest, Univ. Press of Hawaii, Honolulu.  149  Appendix 1 Pedon descriptions Pedon No. 1 Location:  Physiographic position: Slope: Parent material: Land use/vegetation: Soil temperature: Elevation: Climate: Taxonomic classification: Date described/sampled: Described by: No. of samples:  Sitio Sto. Tomas, Wawa, Abra de Hog, Occidental Mindoro, 13°29'14", E 120°37'34"). about 12.8 km. West of Wawa, Abra de Hog and .33 km. Barrio is north of pedon on bearing of343° Lower ridge top of sedimentary hill 2-3% Residual, sedimentary (shale, siltstone) Ipil-ipil, grasses, (aquingay, Saccharum spp. /permanent crops. 26°C up to 10 cm. of topsoil 45 masl Distinct wet and dry (Type I); wet from July to October and dry from November to June fine loamy, isohyperthermic, Typic Ustorthent Sept. 23, 1993 Q. A. Navero, J. Gerpacio and K. MacDicken 3 disturbed; 4 undisturbed core samples  HORIZON DEPTH (cm) DESCRIPTION AC  0-9  Dark grayish brown (10 YR 4/2) moist, silt loam to silty clay loam; no mottles; weak fine to medium granular to sub angular blocky structure; slightly sticky, blocky structure; slightly plastic; friable when moist; presence of few (5%) partly to highly weathered soft platy structure probably shale and talc with olive gray (5Y 5/2) color; many fine vertical and horizontal tubular open pores; many small to medium roots; diffuse smooth boundary.  C,  9-33  Strong brown (7.5 YR 4/6) moist, silt loam to loam; no mottles; weak fine to medium granular to subangular blocky structure; non-sticky, non-plastic, friable moist; few fine tubular and interstitial open pores; many (50-60%) olive (5Y 5/4) weathering soft rock fragments probably shale or talc common fine and few medium roots; diffuse irregular boundary  33-76  Yellowish red (5YR 5/6) moist, silt loam; no mottles; weak fine to medium granular to subgranular blocky structure; non-sticky, non150  plastic, friable moist; many (about 60-70%) weathering soft rock fragments to siltstone or shale; diffuse irregular boundary; few fine and medium roots CR  76-103  Brown to dark brown (7.5YR 4/4) moist, silt loam; no mottles; weak fine to medium granular to subangular, blocky structure; non-sticky, non-plastic, friable moist; many olive (5Y 5/3) soft partly weathered rock fragments probably shale, talc or siltstone. below 103 cm. can be dug with shovel with some difficulty.  R  Pedon No. 2 Location:  Taxonomic classification: Date described/sampled: Described by: No. of samples:  Sitio Sto. Tomas, Wawa, Abra de Hog, Occidental Mindoro, 13°29'08", E 120°37'35"). about 12.8 km. West of Wawa, Abra de Hog and .47 km. Barrio is north of pedon on bearing of 342° Upper side slope of sedimentary hill 40% Residual, sedimentary (shale, siltstone) Leucaena leucocephala, Anacardium occidentale, Artocarpus altilis, Chromolaena odorata, upland rice grown in shifting cultivation, permanent crops. 26°C up to 10 cm. of topsoil 50 masl Distinct wet and dry (Type I); wet from July to October and dry from November to June fine loamy, isohyperthermic, Typic Ustropepts Sept. 23, 1993 Q. A. Navero, J. Gerpacio and K. MacDicken 3 disturbed; 6 undisturbed core samples  HORIZON DEPTH (cm)  DESCRIPTION  A  0-16  Very dark grayish brown (10 YR 3/2) moist, silty clay loam; no mottles; moderate medium subangular blocky breaking to fine subangular blocky structures with application of moderate pressure; slightly sticky, slightly plastic, slightly firm; few small and medium ant burrows with horizontal and diagonal orientation; few weathered and unweathered small rock fragments, probably shale, siltstone and talc; diffuse small boundary; common small and medium roots.  B  16-37  Brown to dark brown (7. SYR 4/2) moist, silty, clay loam; no mottles; moderate medium subangular blocky breaking to fine subangular  Physiographic position: Slope: Parent material: Land use/vegetation:  Soil temperature: Elevation: Climate:  151  blocky structures with application of moderate pressure; slightly sticky, slightly plastic, slightly firm; few small and medium ant burrows with horizontal and diagonal orientation, few small and medium (about 3%) partly weathered and unweathered rock fragments (light olive brown, 2.5Y 5/4); common fine and medium vertical and horizontal tubular open pores; diffuse smooth boundary; few medium and common fine roots. 37-63  Reddish-brown (5YR 4/3) moist, silty clay loam, no mottles; moderate medium subangular blocky structure breaking to fine subangular block structure when moderate pressure is applied; slightly sticky, slightly plastic, slightly firm; common small and medium (20-30%) partly weathered and unweathered olive (5Y 5/3) rock fragments; few small and medium ant burrows; few small vertical and horizontal tubular pores; diffuse smooth boundary; few fine roots  63-102  Yellowish red (5YR 4/6) moist, silty clay loam; no mottles; moderate medium subangular blocky structure; slightly sticky, slightly plastic, slightly firm; many partly to unweathered pale olive (5Y 6/4) and olive (5Y 4/3) rock fragments; few small tubular open pores; few fine roots.  152  Appendix 2 Moisture content adjustment for grain yields The basis for the moisture content adjustments reported in Chapter 5 comes from Nangju and DeDatta (1970). The relationship they described between grain moisture content and reduction in grain yield is shown in Appendix Figure 2.1. The need for moisture content adjustments between treatments is demonstrated in Appendix Figure 2.2. Grain moisture content was consistently and significantly higher in the leucaena fallow treatment than in the non-leucaena treatment.  19.0 20.0 21.0 22.0 23.0 24.0 25.0 26.0 27.0 28.0 Grain moisture content (eg g"1) Appendix Figure 2.1 Reductions in rice grain yield with changes in grain moisture content (as calculated from Nangju and DeDatta, 1970) 153  40.0  'a 3 300 d E c CO  O)  c CD  £  20O Nlon-leucaena Leucaena  10.0 20.0  25.0  30.0  35.0  Overall m.c. mean (eg g"1)  Appendix Figure 2.2 Grain moisture content differences between leucaena and nonleucaena fallows The moisture content to maturity relationship was used to produce Appendix Table 2.1 which shows the adjustments made in treatment plot means. The reduction factor was determined by calculating the differences in grain yield with changes in grain moisture content from the Nangju and DeDatta paper and applying the same percent yield reductions to the Sto. Tomas grain moisture content differences of the same magnitude in the same range. The adjustment factor is equal to 1/yield reduction factor. Where the difference in moisture contents exceeded the range described the maximum value was used (0.5).  This adjustment provided treatment plot means of the same moisture content in each replication. This moisture content was used to adjust grain yields to a standard 14% as described in Chapter 5. For example, for thefirstpair of datafromfieldnumber 1, the calculation changed the grain dry mass estimate to what it might have been at 26% moisture  154  J  (ONj  J  0)010sJslCnt003«WWvl01M4  ! k  a3^CO-»tDCJ)0)-' : r  5>03©-i(DO(Qa  O N > 0 - » 0 - > - » 0 - > 0  O ;-> O p  I ;-> p  O -» - • O - • O O ;-" O ;-> p  o o o o o -  -> O N> O  3  OSS  0 1 ( » 0 1 ( X l U O M O - ' 0 1 » * I O - ' a ! f f i U O ^ M 0 1 0 0 ) O H > M « l - ' ~ J M U -i ' * 0u) Uu I 1u C *u (o ro £^ C D 0 1 ( O W - ' J N l N l ( D ^ O O O * « 0 > 0 - C O O N l > J 60 ai -f* OO>cnC0ON3-*O^IWOCOOOO-»  3 -J°  ^G0^Cd60CJt«-"C0C0roN>-»M - ' W I O M U N ^ - ' ^ U M M C O O U Q a i o a i t o t i ^ t . -» 00 2 <Q C O - J ^ O I _ A W W W ( D W - » W 0 1 ^ N l ( H S ( [ £ I O U ( 0 0 ) ^ O i p ^ C O - 1 0 ) _ - J ^ 0 ) 0 ) 0 ^ 0 ) ( O a i to -* (a' 3 - » -F» O M -» ^J O) ! | ^ ^ ^ f o b i M ® ^ ^ ^ © b i ^ b o ^ c o b f o k ) C o c o i ! k C o c * 3 ^ c j i b o c o b o ^  b b b b b b o b u b b b i o b ^ v i b b b b ^ b b b b b b ^ b b b b b b j ' oooooooocoooovjoc»oiotooo-'Oorv300tDooo*.oooioco  -* O O O O O O :  I  ^ W N N 3 C O W ^ W N ) M W ^ IJ « M W M N ) M r O t O M M - » I O I O M M N l M N ) W I O - » ( O M M f O W M W 00^(»OOOOtONJMNJsl- \ltOOCO(DUO)sJ-»-(i(OMfO-'M-|i01-*^W(0(OOCJ1C£l-*0)^.  O^^CJI^WW  3 S  (O0000C0C0CD00siMs|slCnu1CtI(J1^^^^C0C0a]C«}MMrolO->->^-'3  Zr-zr-zr-zr-zr-zr-zr-zr-i-zr-zzr-zi-i-zzr-r-z  o o o o  —i—i —i -» (o (£) (£) (o  Appendix 3 Straw nutrient contents Appendix Table 3.1 Critical contents of various elements for deficiency in the rice plant Element  Critical content  Plant part analyzed  Growth stage  N P K Ca Mg  2.5% 0.1% 1.0% 0.15% 0.10%  Leaf blade Leaf blade Straw Straw Straw  Tillering Tillering Maturity Maturity Maturity  Source: Yoshida, 1981 Appendix Table 3.2 Mean separation tests for straw nitrogen content Field Field number  Mean (eg g"1)  1  .86a  7  .84a  8  .80a  2  .75ab  4  .66ab  9  .59b  3  .53b  5  .52b  10  .48b  156  Appendix Table 3.3 Mean separation tests for straw potassium content Field Field number  Mean (eg g"1)  2  2.0a  5  2.0a  10  2.0ab  3  1.8ab  1  1.7ab  9  1.6ab  7  1.6ab  4  1.5ab  8  1.4b  Appendix Table 3.4 Mean separation tests for straw calcium content Field Field number  Mean (eg g'1)  7  .06a  4  .06a  1  .04ab  2  .04bc  5  .04bc  9  .04bc  3  .03bc  10  .03bc  8  .02c  157  Appendix Table 3.5 Mean separation tests for straw magnesium content Field  Fallow type  Field number  Mean (eg g'1)  Fallow type  Mean (eg g"1)  7  .11a  Leucaena  .074a  8  .09ab  Non-leucaena  .068b  3  .09ab  4  .07bc  1  .06cd  2  .06cd  9  .06cd  5  .05cd  10  .04d  158  Appendix 4 Nutrient budget worksheets for graphics in Chapter 7 The following tables are the basis for the nutrient budget graphics found in Figures 7.1-7.4. The methods used in these calculations are described in Chapter 7.  159  Appendix Table 4.1 Nutrient budgets at time of rice harvest Soil depth (cm): Soil bulk density: Kg soil/ha:  Field number 1  30 1.40 315000 Biomass (kg/ha) Leucaena Non-leucaena Straw Grain Straw Grain  2  5890 4794  3 4  9529 4212  5 7 8  5250  9  10278  10  2963 3647  3696 3055  3994  7486 6597 7974  3432  3074  2312  6422  4442  3526  2257  1160  2368 1717  1342  2763  5253 6021  1483 3791  7418  3772  Soil 457  Total  Straw  Grain  533 552 636  73 43  40  Soil 441  Total 553  33 37  551  627  567  641  695  20 18  25  536  580  24  536  578  7876  5273 3432  Straw  Grain 32  3489  Nutrient content (kg/ha) Nitrogen Leucaena Field number 1  Non-leucaena  2  45 41 58  39 43  473  3 4  28  37  5  33  48  630 504  536  585  36  7  10  25  488  524  11  30  567  607  8  43  18  504  565  41  63  56  546  34  10  38  37  488  666 563  515 567  571  9  16 41  36  40  473  549  642  Phosphorus Leucaena Field number  Non-leucaena  Straw  Grain  Soil  Total  Straw  Grain  Soil  6  11  22  6  7  12  Total 26  2  5 4  7  15  26  20  35  6  8  18  32  9 4  6  3  7  15  26  4  6  7  47  61  7  5  11  22  5  5  8  5  34  2  23 16  36  7  9 5  3  8  6  3  22  22 31  9  6 3  46 24  40  9  6  11  18  3  8  15  7  13  35 34  20  8  Straw  Grain  Soil  1 2  3 2  1 2  3528  3 4  3 2  2 2  2473 2851 3134  2477 2855  5  2  2  3339  7  1  1  2363  8  1 4  1  2384  2  3  2  1  10  16 23 17  52 34 44  Calcium Leucaena Field number  9 10  Non-leucaena Total 3532  Straw 3 2  Grain 2  Soil 2378  1  2599 3355 3024  2 2  2 1  3343 2364  1  1  1 1  2048  2385 2054  1 1  2  2  2882  2887  1  2  3138  Total 2383 2602 3358 3027  3591 2237  3593  2793 2037  2795 2041  2662  2665  2239  Magnesium Leucaenai  Non-leucaena  Straw  Grain  Soil  Total  1  4  4  1748  1756  Straw 4  2  3  4  1292  1299  3  5 4  1197  1211  4  9 3  756  5  3  5  1150  7  1  3  8  5  9  7 4  Field number  10  Grain  Soil  Total  4  1229  1237  4  4  1339  1346  7  4  1370  1381  763  2  3 3  866  2 1  1402 1071  1407  1158 870  824  2  1008  1015  5  2  819 987  6  893  906  3  5  861  869  4  1292  1299  3  5  1197  1204  160  3  1075 994  Straw 6160 0 0 0 0  30 1.4 315000  Year 0 1 2 3 4  Calcium  Year 0 1 2 3 4  Phosphorus  Year 0 1 2 3 4  Nitrogen  Straw 0.04  Straw 0.15  Straw 0.69  Nutrient concentrations (% of DW)  Year 0 1 2 3 4  Soil depth (cm): Soil bulk density: Kg soil/ha:  Grain 0.05  Grain 0.20  Grain 1.09  Grain 3470 0 0 0 0  Soil 0.88 0.76 1.02 0.80 0.83  Soil 0.00 0.00 0.00 0.00 0.00  Soil 0.16 0.16 0.20 0.18 0.15  Herbs 0 210 340 290 580  Herbs 0.00 0.06 0.41 0.06 0.03  Herbs 0.00 0.23 0.30 0.14 0.19  Herbs 0.00 2.49 1.81 2.09 2.50  Litter 0 1950 1030 1400 1340  Litter 0.00 0.05 0.06 0.02 0.06  Litter 0.00 0.09 0.14 0.09 0.10  Litter 0.00 1.30 1.57 1.16 1.41  Leucaena wood 0 15500 9180 20900 9350  Leucaena wood 0.00 0.15 0.15 0.15 0.15  Leucaena wood 0.00 0.05 0.05 0.05 0.05  Leucaena wood 0.00 0.52 0.52 0.52 0.52  Leucaena foliage 0 810 470 970 410  Leucaena foliage 0.00 0.65 0.65 0.65 0.65  Leucaena foliage 0.00 0.15 0.15 0.15 0.15  Leucaena foliage 0.00 3.99 3.99 3.99 3.99  Other wood 0 700 1290 4840 3010  Biomass (kg/ha, dry-weight)  Appendix Table 4.2 Nutrient balance worksheet for leucaena fallow  Other wood 0.00 0.16 0.16 0.16 0.16  Other wood 0.00 0.05 0.05 0.05 0.05  Other wood 0.00 0.57 0.57 0.57 0.57  Other foliage 0 17 32 143 89  Other foliage 0.00 0.78 0.78 0.78 0.78  Other foliage 0.00 0.11 0.11 0.11 0.11  Other foliage 0.00 2.18 2.18 2.18 2.18  Total 9630 19187 12342 28543 14779  Magnesium Year 0 1 2 3 4  1 2 3 4  Calcium Year 0  1 2 3 4  Phosphorus Year 0  Nitrogen Year 0 1 2 3 4  Nutrient content (kg/ha)  Year 0 1 2 3 4  Magnesium Grain 0.11  Straw 4.56 0.00 0.00 0.00 0.00  Straw 2.40  Straw 9.49 0.00 0.00 0.00 0.00  Soil 2.02 1.67 5.67 1.01 1.80  Soil 504.00 504.00 614.25 567.00 456.75  Soil 0.36 0.34 0.39 0.42 0.37  Grain 3.82 0.00 0.00 0.00 0.00  Soil 1134.00 1055.25 1212.75 1323.00 1149.75  Grain Soil 1.74 2 7 7 2 . 0 0 2394.00 3197.25 2520.00 2598.75  Grain 6.94 0.00 0.00 0.00 0.00  Straw Grain 4 2 . 5 0 37.82 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00  Straw 0.07  Litter 0.00 9.95 1.03 1.54 2.41  Litter 0.00 0.98 0.62 0.28 0.74  Litter 0.00 1.76 1.39 1.26 1.27  Litter 0.00 25.25 16.12 16.17 18.83  Herbs 0.00 0.14 0.13 0.13 0.12  Herbs 0.00 0.29 0.43 0.38 0.70  Herbs 0.00 0.13 1.40 0.17 0.17  Herbs 0.00 0.48 1.03 0.39 1.10  Herbs 0.00 5.22 6.16 6.06 14.47  Litter 0.00 0.51 0.10 0.11 0.18  Leucaena wood 0.00 13.49 7.99 18.18 8.13  Leucaena wood 0.00 23.25 13.77 31.35 14.03  Leucaena wood 0.00 8.22 4.87 11.08 4.96  Leucaena wood 0.00 80.60 47.74 108.68 48.62  Leucaena wood 0.00 0.09 0.09 0.09 0.09  Leucaena foliage 0.00 2.03 1.18 2.43 1.03  Leucaena foliage 0.00 5.27 3.06 6.31 2.67  Leucaena foliage 0.00 1.22 0.71 1.46 0.62  Leucaena foliage 0.00 32.32 18.75 38.70 16.36  Leucaena foliage 0.00 0.25 0.25 0.25 0.25  Other species 0.00 0.59 1.09 4.14 2.58  Other species 0.00 1.27 2.35 9.00 5.60  Other species 0.00 0.35 0.64 2.44 1.52  Other species 0.00 4.36 8.05 30.71 19.10  Other wood 0.00 0.08 0.08 0.08 0.08  Total 1142.38 1081.59 1224.46 1349.67 1164.59  Total 2776.14 2424.89 3218.45 2567.11 2621.95  Total 18.44 13.69 14.30 17.63 11.26  Total 584.33 651.75 711.07 767.32 574.12  Other foliage 0.00 0.29 0.29 0.29 0.29  Straw 5410 0 0 0 0  30 1.4 315000  Grain 1.09  Grain 0.20  Grain 0.05  Straw 0.69  Straw 0.15  Straw 0.04  Year 0 1 2 3 4 Calcium  Year 0 1 2 3 4  Grain 2957 0 0 0 0  Year 0 1 2 3 4 Phosphorus  Nitrogen  Nutrient concentrations (% of DW)  Year 0 1 2 3 4  Soil depth (cm): Soil bulk density: Kg soil/ha:  Herbs 0.00 1.50 1.90 0.85 2.40  Herbs 0.00 0.19 0.32 0.10 0.16  Herbs 0.00 0.05 0.08 0.04 0.05  Soil 0.00 0.00 0.00 0.00 0.00  Soil 0.88 0.76 0.70 0.67 0.78  Litter 0 2050 1530 1500 2020  Soil 0.14 0.16 0.19 0.15 0.10  Herbs 0 1680 1180 3260 1380  Other wood 0.00 0.05 0.05 0.05 0.05 Other wood 0.00 0.16 0.16 0.16 0.16  Leucaena foliage 0.00 0.15 0.15 0.15 0.15 Leucaena foliage 0.00 0.65 0.65 0.65 0.65  Leucaena wood 0.00 0.05 0.05 0.05 0.05 Leucaena wood 0.00 0.15 0.15 0.15 0.15  Litter 0.00 0.14 0.10 0.14 0.14  Litter 0.00 0.06 0.04 0.06 0.10  Other wood 0.00 0.57 0.57 0.57 0.57  Other foliage 0 10 12 20 30  Leucaena foliage 0.00 3.99 3.99 3.99 3.99  Other wood 0 1850 340 5730 590  Leucaena wood 0.00 0.52 0.52 0.52 0.52  Leucaena foliage 0 100 60 60 20  Biomass (kg/ha, dry-weight)  Litter 0.00 1.38 1.57 1.29 1.22  Leucaena wood 0 1930 920 1170 340  Appendix Table 4.3 Nutrient balance worksheet for non-leucaena fallow  Other foliage 0.00 0.78 0.78 0.78 0.78  Other foliage 0.00 0.11 0.11 0.11 0.11  Other foliage 0.00 2.18 2.18 2.18 2.18  Total 8367 5570 2512 10240 2360  Magnesium Year 0 1 2 3 4  Calcium Year 0 1 2 3 4  1 2 3 4  Phosphorus Year 0  1 2 3 4  Nitrogen Year 0  Nutrient content (kg/ha)  1 2 3 4  Year 0  Magnesium Grain 0.11  Straw 4.00 0.00 0.00 0.00 0.00  Straw 2.11  Straw 8.33 0.00 0.00 0.00 0.00  Soil 1.26 1.26 2.90 0.95 1.07  Soil 441.00 504.00 598.50 472.50 315.00  Soil 0.36 0.34 0.39 0.42 0.37  Grain 3.25 0.00 0.00 0.00 0.00  Soil 1134.00 1071.00 1228.50 1323.00 1165.50  Soil Grain 1.48 2772.00 2394.00 2205.00 2110.50 2457.00  Grain 5.91 0.00 0.00 0.00 0.00  Straw Grain 37.33 32.23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00  Straw 0.07  Litter 0.00 2.05 1.68 7.80 2.83  Litter 0.00 1.23 0.61 0.90 2.02  Litter 0.00 2.87 1.53 2.10 2.83  Litter 0.00 28.29 23.94 19.35 24.64  Herbs 0.00 0.13 0.16 0.10 0.12  Herbs 0.00 2.18 1.89 3.26 1.66  Herbs 0.00 0.84 0.94 1.30 0.69  Herbs 0.00 3.19 3.78 3.26 2.21  Herbs 0.00 25.20 22.42 27.71 33.12  Litter 0.00 0.10 0.11 0.52 0.14  Leucaena wood 0.00 1.68 0.80 1.02 0.30  Leucaena wood 0.00 2.90 1.38 1.76 0.51  Leucaena wood 0.00 1.02 0.49 0.62 0.18  Leucaena wood 0.00 10.04 4.78 6.08 1.77  Leucaena wood 0.00 0.09 0.09 0.09 0.09  Leucaena foliage 0.00 0.25 0.15 0.15 0.05  Leucaena foliage 0.00 0.65 0.39 0.39 0.13  Leucaena foliage 0.00 0.15 0.09 0.09 0.03  Leucaena foliage 0.00 3.99 2.39 2.39 0.80  Leucaena foliage 0.00 0.25 0.25 0.25 0.25  Other species 0.00 1.45 0.30 4.47 0.54  Other species 0.00 3.09 0.65 9.50 1.20  Other species 0.00 0.88 0.17 2.72 0.31  Other species 0.00 10.76 2.20 33.10 4.02  Other wood 0.00 0.08 0.08 0.08 0.08  Total 1141.26 1078.62 1233.32 1339.70 1170.87  Total 2775.59 2402.71 2208.97 2124.34 2461.55  Total 15.51 9.38 8.95 9.73 6.63  Total 510.56 582.28 654.24 561.14 379.35  Other foliage 0.00 0.29 0.29 0.29 0.29  « ^ t "t 9 ^  n cq i n cq (N co O CO O  00 CD 00  I— rv r-» in o CD CN  o CN in r^ r^ od  75 in cvj in oo in o O Pooi ^r-i pin' ^CO CN « - tfr o •*  « O m in ro CN o O O O O rco a w  5 ° *7 "~d •*'  r-  « O ^ 03 *fr r•5 O cbi ini CN in Q. CO  i - o co o) ^t i^  w O oo r>- in ^1-  (0 CO 'o  CO Q. CO  q co co co q d d i d d CN  CD  a. co  u. CD -C +J  l_  CD  CD -C  ^-' CO CN CN  O CD CN i -  » o co * n q ra d od co co in •— CM t - oo « o  co ^ :  CD O  co c  co i n  • -^  i-  CD  * ^;  O  <- O  3  CD  o CD 0 5 o O CN r> * 3  TJ  T3 O CN ^  CO 00  §d 3  CN CD O  <*  rv !>>- cn cq <t CN If) CN  q  T3 O n  OO  r^ ^ d <t  o CO * o o CN CN  3  m  O O  O ) CO CN  *~  CO  CD  c  c  c  CD CD U 3 CD - J  CD CO o 3 CD _ l  o o o o  „ O N h ffl i •9 d CN oi CN ^  CO CO CD  CO i — 0 0 CN T — CN T —  X  15 I  j ; q q co CN oq S d f i N ri ui  S o -' d d ^  =5 9  o o  00  in • *  *— OT  '  O  00  I  5  00  '  r- i -  <t  00  o o  CN  '  CO -Q  '  o r>> o o  CD  00 CO  «- ro o  o o o  1o  1  s—  CO  OJ  > o  T—  CN CO  <*  a. >- o  CO -Q  O  c  1 ,  m oo  co  o <*  o o o o o o o o o  .E  CO  s 5  o o o o o o o o o  CO  1  r^  r>  o o  CO i n  o o  CO  r* o  CO  • *  rv  CO  o  o  00  00  m  in  o oo o m  1o  CN CO r—  CN CN CO  *•*  '  I— CD  00  o  i  *•"  Oi  o  CN 1  i  •  CO o o o o o o o o o  o o o o o o o o o  CO  1 E 3  E  a CO i O ro J= CD  i-  i: ™ S CD  •  o o o o o o o o  CO 3 w O -C  cCD  CO  05  CN  1  jj o co o co co  6 5  I—  Hd d d d ^  £ o o o o o o o o o  3 «N o o o o 2 w o o o o  ,_ m  O  '  «— c co o o o o 2 in o o o o  in  O  3  CO  CI)  Z  "~  CD  CO  CD CD O 3 CD  q d  co c CD CD U 3 CO  cD C m O 3 CD _l  cn JO  O  CD O OO q CN O) • CD O t - i-" CN  i  CD  c CD CD O 3 CD  CD CD O  cu q C? 6  CO CO  c  3  O) CO  ° is CD  t-  CN CO >tf  165  O  CD  >  O  « - CN CO <fr  i>  CO CD  > o  i-  CM CO  ^  

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.831.1-0088873/manifest

Comment

Related Items