Linking Land Use - Land Cover Change and Ecosystem Function in Tropical Lowland Watersheds of Belize, Central America

MISSING IMAGE

Material Information

Title:
Linking Land Use - Land Cover Change and Ecosystem Function in Tropical Lowland Watersheds of Belize, Central America
Physical Description:
1 online resource (148 p.)
Language:
english
Creator:
Buck, David G
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Interdisciplinary Ecology
Committee Chair:
Brenner, Mark
Committee Members:
Keys, Eric
Cohen, Matthew
Binford, Michael W

Subjects

Subjects / Keywords:
nutrients -- slash-and-burn -- tropical -- watershed
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre:
Interdisciplinary Ecology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Ecosystems in tropical regions are expected to experience substantial impacts related to human land use and land cover change in the coming decades. Freshwater ecosystems will likely experience a disproportionate impact relative to other tropical ecosystems. My dissertation research takes an interdisciplinary approach to examine interactions between humans and their environment in lowland tropical watersheds of Belize, Central America. I present results from a rapid assessment technique (Chapter 2) designed to estimate the intensity of impacts on freshwater ecosystems stemming from riparian zone land use practices. The technique provides an accessible, low-tech method for identifying river reaches experiencing high degrees of impact. Chapter 3 also examines riparian zone land use practices and focuses on the use of agrochemicals by small-scale Q’eqchi’ Maya farmers. A high percentage of farmers in the Temash River watershed of Belize are cultivating corn within riparian zones and are using herbicides without adequate training. The continued use of herbicides in these riparian zone farms represents an emerging threat to both human and ecosystem health in the Temash River. Chapters 4 and 5 look at slash-and-burn agriculture (locally referred to as milpa) in the Temash River watershed and its impact on soil nutrient dynamics (Chapter 4) and the loss of nutrients from terrestrial landscapes to streams (Chapter 5). In general, soil nutrients in the upper 10 cm are not significantly impacted by milpa. Plant-available nitrogen (as NO3--N) declines significantly in milpa fields and recovery to re-disturbance concentrations requires approximately 15 years, suggesting that 15 years is an ideal fallow period for milpa farmers in the Temash. In-stream nutrients in watersheds dominated by milpa are generally P-limited, as reflected in TN:TP concentrations. A seasonal flux of nutrients is exported from catchments at the onset of the rainy season but estimates for annual export of nutrients are similar to or less than other tropical watersheds.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by David G Buck.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Brenner, Mark.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-02-28

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0041467:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 LINKING LAND USE LAND COVER CHANGE AND ECOSYSTEM FUNCTION IN TROPICAL LOWLAND WATERSHEDS OF BELIZE, CENTRAL AMERICA By DAVID GRAY BUCK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PAR TIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 David Gray Buck

PAGE 3

3 This dissertation is dedicated to my wife, Ellie Harrison Buck for all of her support during this resea rch; and opportunity to live and learn in their watershed.

PAGE 4

4 ACKNOWLEDGMENTS I am grateful for the support of a number of individuals and organizations during my disserta tion research. I would like to thank my advisor and committee chair Dr. Mark Brenner as well as committee members Drs. Michael Binford, Matthew Cohen and provided access t o its geospatial laboratory and water quality laboratory. The NSF IGERT Working Forests in the Tropics program at the University of Florida awarded me a 3 year doctoral fellowship (NSF DGE 0221599) that helped create an ideal academic environment to explo re interdisciplinary questions related to human environment interactions and the conservation of tropical watersheds. In addition, an NSF Doctoral Dissertation Research Improvement Grant (NSF BSE 0825278) and a Sigma Xi Grants in Aid of Research (Grant ID # G2006351230333459) provided additional funding for field and analytical costs. Field work in the Monkey River watershed was done in collaboration with the Toledo Institute for Development and the Environment (TIDE) and Peter Esselman. Peter shared his experience and insights into the development of the ELSI and human impact mapping and was my partner in data collection for the 2007 impact mapping. Peter continues to be a leading advocate for freshwater conservation in Belize. Field work in the Temash R iver watershed was done in collaboration with the Sarstoon Temash Institute for Indigenous Management (SATIIM) Juan Pop (Crique Sarco village) was my primary field assistant during my dissertation field work and without him this work would not have been possible. Anasario Cal (Crique Sarco) served as my boat captain (while the boat was operational) and I am grateful for his time and knowledge of 10 horse power boat motors. In addition, Josiah Bo (Crique

PAGE 5

5 Sarco) and Mauricio Tush (Sunday Wood) were invalua ble assistants during the matambre interview process. Joel Wainwright (Ohio State University) also provided advice on early versions of the interview questionnaire used for the matambre study. Data collection on land use and land cover was conducted in c ollaboration with Sean Downey (University of Arizona). Sean developed a database used for storing literally thousands of data fields into a manageable and coherent format that made data analysis streamlined. Rain gauges installed in the Temash were purchased by Sean through a n NSF Doctoral Dissertation Improvement Grant (NSF Award # 0647832) I appreciate Sean s time, energy and commitment t o many of our shared research interests. Since completing the field work portion of my dissertation, I have worked at the Biodiversity Research Institute (BRI) in Gorham Maine. Dr. David Evers, the executive director of BRI, has provided endless support, including generous paid leave that helped me to see this project to completion. My parents, Steve and Bettie Buck, have been tremendously supportive during my graduate studies. I am forever grateful to my wife, Ellie Harrison Buck, and our daughter, El iza, who joined me during much of the 16 months of field work in southern Belize. Since then, our second daughter Natalie has joined the family. Natalie and Eliza have given me countless hours of enjoyment (and plenty of distractions) during the final wr iting phases of this dissertation. All three, particularly Ellie, have provided endless support and motivation for the completion of the dissertation. Without Ellie, this would not have been possible.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 11 ABSTRACT ................................ ................................ ................................ ................... 12 CHAPTER 1 BACKGROUND ................................ ................................ ................................ ...... 14 Human Environment Interactions in Tropical Watersheds ................................ ... 14 Aquatic Resources of Belize ................................ ................................ ................... 1 6 Statement of Objectives ................................ ................................ .......................... 18 2 MONITORING LAND USE CHANGES ALONG RIPARIAN CORRIDORS IN LOWLAND TROPICAL WATERSHEDS: APPLICATION OF HUMAN IMPACT MAPPING AND ES TIMATION OF LOCAL STRESS INTENSITY (ELSI) ............... 22 Introduction ................................ ................................ ................................ ............. 22 Study Site ................................ ................................ ................................ ............... 24 Methods ................................ ................................ ................................ .................. 25 Results ................................ ................................ ................................ .................... 27 Discussion ................................ ................................ ................................ .............. 27 Conclusions ................................ ................................ ................................ ............ 31 3 LAND USE PRACTICES, PESTICIDE APPLICATION S, AND FRESHWATER RESOURCE CONSERVAtiON IN A LOWLAND TROPICAL WATERSHED .......... 39 Introduction ................................ ................................ ................................ ............. 39 Study Area ................................ ................................ ................................ .............. 41 Study Villages and Watershed ................................ ................................ ......... 41 Riparian Zone Land Use within the Temash River Watershed ......................... 42 Methods ................................ ................................ ................................ .................. 43 Results ................................ ................................ ................................ .................... 44 Households and Matambre ................................ ................................ .............. 44 Water Use and Water Quality ................................ ................................ ........... 45 Matambre Field Management ................................ ................................ ........... 46 Discussion ................................ ................................ ................................ .............. 46 Matambre and Riparian Conservation ................................ .............................. 46 Matambre, Pesticides and Public Health ................................ .......................... 47 Environmental F ate of Agrochemicals in the Temash ................................ ...... 49 Conclusions ................................ ................................ ................................ ............ 51

PAGE 7

7 4 SOIL NUTRIENT DYNAMICS, ORGANIC MATTER TURNOVER AND LAND USE WITHIN AN ANTHROPOGENIC LANDSCAPE DOMINATED BY SHIFTING CULTIVATION ................................ ................................ ...................... 56 Introduction ................................ ................................ ................................ ............. 56 Study Area ................................ ................................ ................................ .............. 59 Methods ................................ ................................ ................................ .................. 60 Soil Sampling and Land Use History ................................ ................................ 60 Laboratory Analysis ................................ ................................ .......................... 61 Estimation of SOM Turnover ................................ ................................ ............ 62 Statistical Analyses ................................ ................................ .......................... 63 Results ................................ ................................ ................................ .................... 63 Soil Physical and Chemical Characteristics ................................ ...................... 63 Stable Carbon Isotopes in Vegetation and Soils and the MRT of SOM ............ 65 Discussion ................................ ................................ ................................ .............. 65 Soil Nutrient Dynamics in Milpa Versus Forest Soils ................................ ........ 65 Cacao Agroforestry and Pasture ................................ ................................ ...... 69 Stable Isotopes and MRT of SOM in Temash Soils ................................ ......... 71 Sustainability of Milpa in the Maya Forest ................................ ........................ 72 5 TEMPORAL AND SPATIAL VARIABILITY OF DISSOLVED NUTRIENTS IN A SMALL TROPICAL WATERSHED DOMINATED BY SHIFTING CULTIVATION ... 84 Introduction ................................ ................................ ................................ ............. 84 Methods ................................ ................................ ................................ .................. 87 Study Site ................................ ................................ ................................ ......... 87 Watershed Selection, Delineation, and Characterization ................................ 87 Annual Rainfall Pattern ................................ ................................ ..................... 88 Sample Collection and Laboratory Analysis ................................ ..................... 89 Statistical Analysis ................................ ................................ ............................ 90 Discharge, Nutrient Fluxes and Flow Weighted Mean Nutrient Concentrations ................................ ................................ .............................. 91 Results ................................ ................................ ................................ .................... 91 Precipitation, Watershed Characteristics, and Seasonal Discharges ............... 91 Seasonal and Longitudinal Variation of In stream Nutrient Concentrations ..... 94 Nutrient Fluxes and Flow Weighted Mean Nutrient Concentrations ................. 96 Discussion ................................ ................................ ................................ .............. 99 6 CONCLUDING REMARKS ................................ ................................ ................... 119 APPENDIX A ELSI RANK SCORES FOR OBSERVED STRESSES ................................ .......... 121 B MATAMBRE INTERVIEW QUESTIONAIRE ................................ ......................... 125 LIST OF REFERENCES ................................ ................................ ............................. 129

PAGE 8

8 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 148

PAGE 9

9 LIST OF TABLES Table page 1 1 Principal mechanisms by which land use influences stream ecosystems (modified from Allan 2004) ................................ ................................ ................. 20 1 2 Sixteen major watersheds of Belize, drainage areas and underlying geologies. Data from Lee et al. 1995. ................................ ................................ 20 2 1 Stress source relationships a nd the scientific literature used t o justify these relationships. ................................ ................................ ................................ ...... 32 2 2 Stress source rankings used for human impact mapping scoring criteria and the developing of the ELSI Index. ................................ ................................ ..... 33 3 1 Number of households surveyed and the mean household size (% of total households in village; = standard deviation) ................................ .................... 53 3 2 Summar y of matambre characteristics for each village ( = stdev) .................... 53 3 3 Agrochemical use among matambre farmers ................................ ..................... 53 3 4 Common agrochem icals of matambre farmers and their toxicity to aquatic organisms ................................ ................................ ................................ ........... 54 4 1 Dominant land cover classes in the Temash River watershed. Soils were collected from each land cover class (sample siz e in parentheses). .................. 75 4 2 Soil properties (mean standard error) of the seven primary land use/land cover classes in the Temash River Watershed. ................................ ................ 76 4 3 Macronutrients concentrations (mean standard error) of the seven primary land use/land cover classes in the Temash Watershed. 77 4 4 Micronutrient concentrations (mean standard error) of the seven primary land use/land cover classes in the Temash Watershed. 78 4 5 Pearson's product moment correlation matrix for soil characteristics from the Temash ................................ ................................ .......................... 79 4 6 13 C vs. PDB) isotopic concentrations of soil organic matter for the seven primary land cover cl asses in the Temash Watershed. ................ 80 5 1 Catchment characteristics of the four stud y watersheds and the Temash River watershed ................................ ................................ ................................ 102 5 2 Comparison of dry and wet season discharge rates (m 3 s 1 ) across all sampling sites ................................ ................................ ................................ ... 103

PAGE 10

10 5.3 discharge and antecedent rainfall ................................ ................................ ..... 104 5 4 Seasonal mean nutrient concentrations for DIN, TSN, and TN for all sampling sites within each study watershed. ................................ ................................ ... 105 5 5 Seasonal mean nutrient concentrations for SRP, TSP, and TP for all sampling sites within each study watershed. ................................ .................... 106 5 6 Seasonal mean ( stdev) TN:TP ratios for upstream and downstream sampling sites within each study watershed. ................................ .................... 107 5 7 Linear regression parameters fo r annual flow weighted DIN, TSN, and TN concentrations and watershed characteristics ................................ .................. 108 5 8 Linear regression parameters for a nnual flow weighted SRP, TSP, and TP concentrations a nd watershed characteristics ................................ .................. 109 5 9 A comparison of nutrient export coeff icients (kg ha 1 yr 1 ) from tropical watersheds ................................ ................................ ................................ ....... 110

PAGE 11

11 LIST OF FIGURES Figure page 1 1 The sixteen major watersheds of Belize ................................ ............................. 21 2 1 The Monkey River Watershed and the primary villages located within the catchment. ................................ ................................ ................................ .......... 34 2 2 Percent distribution of stres ses observed from human impact mapping in the MRW in 2000 and 2007. ................................ ................................ ..................... 35 2 3 The observed sources of stress in each of the three branches of the Monkey River ................................ ................................ ................................ .................. 36 2 4 Change in impacts and estimations of stress intensity for sedim entation (A) and flow alteration (B) for the MRW from 2000 2007. ................................ ...... 37 2 5 Change in the Overall ELSI Index for the MRW from 2000 2007. ...................... 38 3 1 Map of the Temash River watershed including the three study villages. The lower reaches include the Sarstoo n Temash National Park. .............................. 55 4 1 Map of study area. ................................ ................................ .............................. 81 4 2 Soil characteristics of land cover classes relative to forest soils. ...................... 82 4 3 Percentages of carbon derived from C3 forest and from C4 pasture. ................ 83 4 4 The Mean Residence Time (MRT) of carbon in SOM of pasture ........................ 83 5 1 Map of study area, including the Temash River watershed and the four study catchments were in st ream nutrients were monitored. ................................ ..... 111 5 2 Precipitation r ecord for three stations in the Temash River watershed from February 2, 20 07 through June 24, 2008. ................................ ....................... 112 5 3 Absolute and normalized (by watershed area) discharge rates for the study watersheds. ................................ ................................ ................................ ...... 113 5 4 Seasonal mean instantaneous discharge rates (m 3 sec 1 ) versus watershed area (ha). The two outliers (gray triangles) are YXL01 and SWD01. ............... 114 5 5 Daily nitrogen fluxes from the study watersheds ................................ .............. 115 5 7 Flow weighted mean concen trations for DIN, TSN, and TN. ............................ 117 5 8 Fl ow weighted mean concentrations for SRP, TSP, and TP. ........................... 118

PAGE 12

12 Abstract of Dissertation Presented to the Graduate Schoo l of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy LINKING LAND USE LAND COVER CHANGE AND ECOSYSTEM FUNCTION IN TROPICAL LOWLAND WATERSHEDS OF BELIZE, CENTRAL AMERICA By David G. Buck August 2012 Chair: Mark Brenner Major: Interdisciplinary Ecology Ecosystems in tropical regions are expected to experience substantial impacts related to human land use and land cover change in the coming decades. Freshwater ecosystems will li kely experience a disproportionate impact relative to other tropical ecosystems. My dissertation research takes an interdisciplinary approach to examine interactions between humans and their environment in lowland tropical watersheds of Belize, Central Am erica I present results from a rapid assessment technique (Chapter 2) designed to estimate the intensity of impacts on freshwater ecosystems stemming from riparian zone land use practices. The technique provides an accessible, low tech method for identif ying river reaches experiencing high degrees of impact. Chapter 3 also examines riparian zone land use practices and focuses on the use of agrochemicals by small watershed of B elize are cultivating corn within riparian zones and are using herbicides without adequate training. The continued use of herbicides in these riparian zone farms represents an emerging threat to both human and ecosystem health in the Temash River.

PAGE 13

13 Chapter s 4 and 5 look at slash and burn agriculture (locally referred to as milpa ) in the Temash River watershed and its impact on soil nutrient dynamics (Chapter 4) and the loss of nutrients from terrestrial landscapes to streams (Chapter 5). In general, soil n utrients in the upper 10 cm are not significantly impacted by milpa. Plant available nitrogen (as NO 3 N) declines significantly in milpa fields and recovery to re disturbance concentrations requires approximately 15 years suggesting that 15 years is an ideal fallow period for milpa farmers in the Temash. In stream nutrients in watersheds dominated by milpa are generally P limited, as reflected in TN:TP concentrations. A seasonal flux of nutrients is exported from catchments at the onset of the rainy se ason but estimates for annual export of nutrients are similar to or less than other tropical watersheds.

PAGE 14

14 CHAPTER 1 BACKGROUND Human Environment Interactions in Tropical Watersheds T he humid tropics are expected to experience significant declines in e cosystem services and functions due to land use and land cover change ( LULCC ) (Millennium Ecosystem Assessment 2005) and these impacts will likely not be distributed evenly across tropical biomes. Tropical rivers and their riparian zones serve as centers for rural livelihoods and high deforestation rates c oupled with population growth are expected to have a disproportionately large impact on tropical fr eshwater ecosystems (Sala et al. 2000) In tropical regions, b oth rural and urban people alike depend on freshwater resources for drinking and utility water, irrigation, protein via fishing, transportation and recreation. Durin g the next several decades, it is expected that (McClain 2002) As populations increase, greater dependence on freshwater resources will follow along with the potential for gre ater impact on these resources. E cologists have long understood the importance of connectivity between streams and rivers and their catchments (Likens and Bormann 1974, Hynes 1975, Weins 2002) and understanding the influence of land use practices on aquatic ecosystem structure and function often depends on the spa tial scale of the analysis. Regional conditions such as geomorphology, soils, topography, and land use land cover change can influence nutrient supply, sediment delivery, hydrology, and channel characteristics. Local riparian conditions such as vegetat ion cover can exert control over in stream habitat structure and organic matter input (Allan et al. 1997, Ankers et al. 2003) In some cases the most explanatory variable for variations in aquatic ecosystem structure

PAGE 15

15 may simply be proximity to human infrastructure (Scheuerell and Schindler 2004) Legacy effects from past land uses in the catchment may also be an important factor in defining the observed distribution of organisms within a catchment (Harding et al. 1998) In add ition, the annual hydrologic regime can have a significant effect and may overshadow the apparent control of physical/geologic characteristics in the catchment on dynamics such as nutrient supply and sediment delivery (Johnson et al. 1997) (Table 1 1). Although extensive literature exists linking human land use practices and aquatic ecosystem structure and function in temperate watersheds, studies incorporating both terrestrial and aquatic ecosystems in tropical regions are exceedingly rare (Allan et al. 1997, McClain 2002) Previous studies of nutrient dynamics in tropical watersheds suggest that nitrogen (N) load in rivers is slightly higher than in temperate systems (Downing et al. 1999, F iloso et al. 2003) Phosphorus (P) content in tropical streams is positively correlated with soil nutrient concentrations (Biggs et al. 2002) and ca n influence decomposition and N mineralization rates in tropical ecosystems (Hobbie and Vitousek 2000) Overall, l and use impacts on the biogeochemistry of tropical str eams are not well understood nor quantified (Biggs et al. 2004) In a large, tropical watershed (~12,400 km 2 ) with high human population densitie s, total N export was correlated with anthropogenic N inputs (Filoso et al. 2003) In wat ersheds converted to pasture, streams were N limited and had lower NO 3 concentrations than forested streams of similar size (Neill et al. 2001) In a small watershed (0.25 km 2 ) 80% deforested by slash and burn, high concentrations of TN and TP were observed in overland flow

PAGE 16

16 (Williams and Melack 1997) yet no significant increase s in NO 3 NH 4 + or PO 4 3 were observed in streams when compared with a forested catchme nt (Williams and Melack 1997) Biggs et al. (2004) observed a stream nutrient response to deforestation in watersheds (ranging in size from 2 to 300 km 2 ) that were 66 75% cleared, suggesting that stream nutrient concentrations are resistant to land cov er conversion below this threshold. Belize, Central America provides a unique opportunity to test hypotheses related to impacts of LULCC on aquatic systems. A substantial portion (~35%) of the country is held in some sort of natural or cultural protected area (Meerman and Wilson 2005) ) and the severity of ecosystem stress may be relatively low by regional standards The rivers of Belize provide a linkage between terrestrial ecosystems, wetlands and estuaries of t he coastal zone, and the longest barrier reef in the Western Hemisphere, the Mesoamerican Reef (MAR; or the Mesoamerican Barrier Reef System, MBRS ). Increased awareness of this connectivity has added to the value of a watershed scale perspective for conse rvation initiatives in Belize (Nunny et al. 2001, Bailey et al. 2007) Aquatic Resources of Belize A c omprehensive water quality survey of Belize identifi ed sixteen major watersheds and many small coastal tidal creek watersheds (Lee et al. 1995) These watersheds vary in size, traverse a variety of geologies and soil types and also drain diverse terrestrial ecosystems and human dominated landscapes (Lee et a l. 1995, Esselman and Boles 2001) The Maya Mountain Massif (MMM) forms the headwaters for 12 of the 16 major watersheds (Table 1 2 ), from the Belize River in central Belize to the Moho River in southern Belize. Most of these rivers originate a s high gra dient, relatively low pH streams within granite and metamorphic rock catchments, traverse

PAGE 17

17 limestone dominated landscapes, build up alluvial plains and wetlands, and ultimately discharge into the inner channel or shelf lagoon that separates the coast from t he barrier reef. The Rio Hondo (forming the northern border of Belize) and New River, originate in karst hills, drain the low relief limestone landscape of northern Belize and discharge into Chetumal Bay, which in turn discharges into the inner channel. The headwaters of the Temash River begin in Guatemala and flow eastward across the southern Toledo district. The Sarstoon River (demarcating the southern border of Belize) originates in Guatemala, within the mountain range and foothills of the Sierra de S anta Cruz (Esselman and Boles 2001, Esselman et al. 2006) In addition to lotic (flowing) waters, Belize has an abundance of lagoons and wetlands. Many of the freshwater wetlands occur in northern Belize and contain large complexes of lentic (st ill) water habitats including swamp forests, herbaceous marshes and open water areas (Esselman and Boles 2001) Of the many freshw ater wetlands in Belize, two have been internationally recognized as RAMSAR sites including Crooked Tree Lagoon and a Sphagnum bog within the boundaries of the Sarstoon Temash National Park. A few karst, sink hole lakes are found in the country, with Five Blues Lake in the Cayo Distric t being the most notable. Although recognized as a significant component of the freshwater resources of Belize, groundwater resources remain almost complet ely unstudied (ERMA 2007) Rese arch on freshwater biodiversity and the ecology of aquatic organisms is limited in Belize. Significant works on aquatic macroinvertebrates have been conducted in the Belize River ( Gonzalez 1980) and the Sibun River (Boles 1998) Both studies provide d descriptive information on the community of aquatic macroinvertebrates in

PAGE 18

18 these two watersheds and se rve as important reference document s for future monitoring efforts. The ecology and distribution of tropical disease carrying mosquitoes has also received significant research attention. Mosquitoes of the genus Anopheles (known malaria vectors) were obse rved to have a strong habitat preference within 1 km of rivers (Roberts et al. 1996) and larvae were significantly associated wit h certain riverine habitats (Manguin et al. 1996) Seasonal variation in the distribution of these mosquitoes also illustrates the role of increased rainfall in expanding the breeding habitat of these o rganisms (Roberts et al. 2002) The freshwater fishes o f Belize have been studied in greatest detail of any freshwater organism. Freshwater Fishes of the Continental Waters of Belize (Greenfield and Thomerson 1997) is the definitive work on freshwater fishes, and includes taxonomic keys importance of longitudinal connectivity and geology to native fish fauna has been documented in the Monkey River watershed (Esselman et al. 2006) whereas a country wide survey of freshwater fishes was recently completed that (1) document s national patterns in freshwater fish diversity, (2) predict s the eventual range of invasive tilapias as they continue to expand their range in Belize, and (3) iden tifies a network of freshwater fish conservation areas that, if protected, would conserve the most fish biodiversity with the lea st amount of investment (Esselman 2009) Statement of Objectives Following this introduction to Belize and its freshwater resources, I present four studies that examine impacts of LULCC on aquatic systems at var ious scales ranging from the riparian zone to the entire catchment. Chapter 2 addresses LULCC impacts at the scale of the riparian zone and presents results from a study utilizing a rapid impact

PAGE 19

19 assessment technique to estimate potential stresses on aquat ic ecosystems stemming from riparian land uses. Chapter 3 takes a closer look at riparian zone land use and Maya of southern Belize and their utilization of riparian zones for corn cultivation during the dry sea son. This chapter compares and contrasts traditional slash and burn agriculture with the slash and mulch agriculture that is practiced within riparian zones of the Temash River watershed and also discusses impacts related to an increased use of pesticides within these slash and mulch fields. Chapter 4 examines LULCC at the watershed scale in the Temash and focuses on soil nutrient dynamics across a chronosequence of active and abandoned agricultural fields, pasture, small holder cacao plantations, and mat ure forest. Chapter 5 examines the spatial and temporal variabi lity of in stream nutrients in four small watersheds within th e Temash River watershed. Land use within these catchments is primarily slash and burn and the discussion focuses on the relat ion ship between nutrients, land use p ractices, and abiotic variables such as soil characteristics and basement geology.

PAGE 20

20 Table 1 1. Principal mechanisms by which land use influences stream ecosystems (modified from Allan 2004) Environmental Stressor Effec ts Sedimentation turbidity, scouring, abrasion primary production, depth heterogeneity of stream Nutrient enrichment autotrophic production, favorable conditions for filamentous algae, litter breakdown rates dissolved oxygen; species shift f rom sensitive to tolerant species Contaminant pollution heavy metals and toxic organic substances resulting in deformities and mortality growth and reproduction rates and survival among fishes Hydrologic alteration Alters runoff evapotranspiration ba lance vulnerability to erosion; potential for transport of nutrients, sediments and contaminants Riparian clearing light penetration and water temperature bank stability, litter and woody debris inputs and retention of nutrients and sediments Alte rs quantity and quality of organic matter Table 1 2. Sixteen major watersheds of Belize, drainage areas and underlying geologies. Data from Lee et al. 1995. Major Watershed Area (km 2 ) Geology Rio Hondo 15,075.5 Limestone rocks and variable soils; al kaline geochemistry Belize River 9,434.2* Mix of limestone, igneous and metamorphic rocks Sarstoon River 2,217.5 Limestone and tertiary coastal sediments New River 1864.0 Limestone rocks and variable soils; alkaline geochemistry Monkey River* 1275.4* Igneous rocks, metasediments, volcanic rocks, and limestone Moho River 1,188.5* Limestone Sibun River 967.8* Acidic igneous rocks, metasediments, and limestone Rio Grande 718.5* Limestone Manatee River 484.0* Metasediments and limestone Temash River* 474.6 Limestone and tertiary coastal sediments Sittee River 451.2* Metasediments and tertiary coastal sediments Deep River 347.9* Limestone North Stann Creek 281.4* Ancient igneous rocks and metasediments; some limestone South Stann Creek 258.0* Anci ent igneous rocks and metasediments Golden Stream 204.1* Limestone Mullins River 156.9* Metasediments and limestone

PAGE 21

21 Figure 1 1. The sixteen major watersheds of Belize

PAGE 22

22 CHAPTER 2 MONITORING LAND USE CHANGES ALONG RIPARI AN CORRIDORS IN LOWL AND TROP ICAL WATERSHEDS: APP LICATION OF HUMAN IMPACT MAPPING AND ESTIMATION OF LOCAL STRESS INTENSITY (EL SI) Introduction Riparian forests are central to river conservation activities because of their unique ecosystem structure and function within the landscape. Riparian forests maintain bank stability and serve as buffers against excess sediment and nutrient transport across the terrestrial aquatic interface. Vegetation in riparian zones regulates in stream temperatures and provides organic matter that is esse ntial for aquatic food webs. Riparian forests also exert control on the hydrologic balance in watersheds by influencing runoff, subsurface water storage, and evapotranspiration. In addition, riparian forests contain biological communities that are adapte d to disturbance regimes associated with hydrologic events. They are also key components for maintaining biological connectivity across the landscape. Together with aesthetic and cultural values related to water quality, reduced flood damage, and recreat ion, riparian forests are important target habitats in a conservation portfolio (Naiman and Decamps 1997, Naiman et al. 2005, Sabo et al. 2005) Conservation of riparian forests, their adjacent freshwater resources, and combined terrestrial and aquatic biodiversity, requires an integrated, whole catchment approach to management that incorporates a science based strategy for conservation and a plan for the resource needs of local stakeholders (Pringle et al. 1993, Saunders et al. 2002) This is particularly important in tropical regions where rivers and riparian zones have long served as sites for rural livelihoods. In the humid tropics, the deforestation rate is ~5.8 million ha/yr (Achard et al. 2002) and l and use and land cover

PAGE 23

23 change has a disproportionately large effect on tropical freshwater ecosystems and the goods and services they provide (Sala et al. 2000, Millennium Ecosystem Assessment 2005) Impacts on riparian forests occur across multiple spatial and temporal scales. Patterns of land use and land cover change at the catchment scale influence nutrient supply, sediment delivery, hydrology and geomorphic characteristics. Local riparian conditions such as vegetation cover can exert control over in stream habitat structure and organic matter input. In addition to spatial controls on freshwater ecosystems, the legacy of past land uses also in fluences freshwater ecosystems (Allan et al. 1997, Harding et al. 1998, Allan 2004) The scale of analysis for determining the impact of land use and land cover change on riparian zones and their associated freshwater resources is largely dependent on the available technology and the time allotted for analysis (Johnson and Gage 1997) In developing countries, technolog ical capability and the transfer of appropriate methodologies between developed and developing countries is key for the success of freshwater monitoring programs (Resh 2007) In Be lize, Central America, aquatic scientists, in collaboration with local natural resource managers and conservationists, have developed a monitoring technique for human impact mapping that estimates the severity and irreversibility of impacts on freshwater s ystems stemming from riparian zone land use (Esselman 2001) The Expected Local Stress Intensity (ELSI) Index has been utilized in numerous freshwater settings in the country (Esselman and Boles 2001, Lee 2002, TIDE 2003, Karper and Boles 2004, Esselman and Buck 2007, Requena 2008) Human impact mapping and ELSI represent a n

PAGE 24

24 communica te spatially explicit data on riparian zone land uses and estimate the impacts of these land uses on freshwater resources. Here I examine changes in riparian zone land use and the ELSI Index from the Monkey River watershed (MRW), Belize, between 2000 and 2 007. This paper addresses three primary concerns for the conservation of freshwater resources in the MRW: (1) What are the primary land uses within the riparian zone of the MRW responsible for stresses in riparian and aquatic ecosystems? (2) Where, within the watershed, do these land uses occur?, and (3) How severe are the stresses on riparian and aquatic ecosystems? A clear understanding of the drivers of ecological stress within riparian zones, together with knowledge of where land use change is occurring in the MRW, can enable effective management strategies that address riparian zone and freshwater conservation at the river reach, sub catchment, and catchment scale. Study Site The Monkey River is located on the southeastern flank of the Maya Mountains, i n southern Belize (Figure 2 1). The Monkey River watershed consists of three main branches (Bladen, Trio and Swasey) that join in the coastal plain and enter the Caribbean Sea as a 6 th order river. The Monkey River is the fourth largest watershed in Beli ze (1275 km 2 ). The headwaters of all three branches drain mountainous, primarily undisturbed tropical broadleaf forest. The middle reaches flow through human dominated landscapes that include commercial banana cultivation, pasture, gravel mining, and su bsistence agriculture. Human settlements are also concentrated within these middle reaches (Figure 2 1). The lower reaches, below the confluence of the Bladen and Swasey Branches, are largely undeveloped.

PAGE 25

25 The Monkey River is the largest of the six waters heds (Rio Grande, Middle River, Maya Mountain Marine Corridor (MMMC). The MMMC is a landscape scale conservation initiative that connects headwater regions of the Maya Moun tains to the coastal waters and coral reefs in the Gulf of Honduras. The MMMC is a matrix of national protected areas, extractive reserves, and human dominated landscapes that includes an expansive wetland and estuary network and a coastal embayment with ~130 cayes and a diverse fishery. The MMMC is a collaborative conservation initiative directed by the Toledo Trust, and The Nature Conservancy (TNC) Belize Program. A Conserva tion Action Plan was recently developed for the MMMC and freshwater systems were identified as a primary conservation target for the corridor (Salas and Meerman 2007) Within this conservation target, riparian zone connectivity was identified as a key ecological attribute. In the MMMC, riparian zone connectivity includes longitudinal connectivity the riparian zone and associated flood plain (Buck et al. 2008) Methods Human impact mapping was conducted in the MRW in 2000, 2002, and 2007 (Esselman 2001, TIDE 2003, Esselman and Buck 2007) Detailed descriptions of methods used in human impact mapping and the ELSI Index are presented elsewhere (Esselman 2001) Briefly, the methods include 7 primary steps: (1) Observations of human impacts within the riparian zone are made a long the main river channel, from a kayak or canoe. Each impact is marked using a global positioning system (GPS), and

PAGE 26

26 1). (2) The sources of stress are then linked to 6 primary stresse s (Table 2 1). (3) The contribution of each source of stress to the primary stresses are ranked using a combination of scientific literature review, best professional assessment, and pre determined ranking tables developed by TNC (TNC 2000, Esselman 2001, Esselman and Buck 2007) (4) Using a Geographic Information System (GIS), the river is segmented into 1 km reaches in an upstream direction beginning from the river mouth. Segment s are converted to polygons that include a 100 m buffer around the river channel. These segmented polygons become the basis of analysis for analyzing stresses across the riparian corridor. (5) Using the data collected with the GPS, the ranked stress scor es ( step 3 ) are summed for each river segment. (6) Spatially explicit maps, detailing expected stress intensity, are developed for each primary stress. For ease of communication, categorical break points were established by ranking each stress score and dividing the rankings into quartiles. Each quartile then represents one of four levels of stress (Low, Medium, High, Very High). (7) Rankings for each stress are then summed to provide a cumulative stress score for each river reach, thus providing an ove rall expected stress intensity. Human impact maps and ELSI scores for the MRW from 2000 and 2007 were analyzed to identify river reaches where changes in stress intensity had occurred. In the case of the Trio Branch of the Monkey River, impact data from 2 002 (TIDE 2003) were utilized because Trio Branch was not included in the original 2000 human impact mapping. The stress and source rankings from 2000/02 and 2007 are presented in Table 2 2. All rankings were equal across both study years except with respect to

PAGE 27

27 drainage ditches associated w ith industrial banana plantations within the MRW. As of 2002, the Belize Banana Growers Association has been working to improve drainage of Assistance to Traditional Sup pliers of Bananas (GOB 2002) Stress rank sc ores for 2000/02 were subtracted from stress rank scores from 2007. These data reveal the direction of land use change a positive value suggests increased impact between 2000 and 2007; a negative value suggests reduced impact between 2000 and 2007. Obs ervations of sources of stress were summarized for each branch of the MRW to further isolate areas of land use change and the drivers of change within each branch. Results Change in overall distribution of primary stress types within the MRW between 2000 a nd 2007 was minimal (Figure 2 2, Figure 2 3). The dominant stress in the MRW was sedimentation. Sources of stress contributing to flow alteration were minimal. The number of observed sources of stress within the riparian zone of the MRW increased from 13 8 observations in 2000 to 635 observations in 2007 (Figure 2 4). Riparian zones with no buffer or thin buffer (<10m) increased across all branches of the Monkey River. In addition, observations of cattle grazing within the riparian zone increased between 2000 and 2007 in each of the three main branches, with the greatest increase observed in the Trio Branch (0 observations in 2000; 44 observations in 2007) (Figure 2.4). Changes in the ELSI Index from 2000 to 2007 illustrate the spatial distribution of st ress within the MRW and how it has changed (Figure 2 5) Discussion Human impact mapping within the MRW identified river segments that require conservation action to alleviate further stresses on the aquatic resources of the

PAGE 28

28 watershed. The Trio Branch expe rienced the most rapid human induced changes. The expansion of cattle within Trio has severely impacted the middle and lower reaches of this tributary (Esselman a nd Buck 2007) Cattle contribute to multiple stresses including sedimentation, nutrient enrichment, and habitat alteration. Many of the new pasture areas along the Trio Branch use barbed wire fences that are stretched across the main channel. During fl ood events, riparian trees cleared during pasture development create large snags where barbed wire crosses the channel, impeding the natural flow of the river (Ess elman and Buck 2007) Small scale agricultural activity also impacts the riparian zone of the MRW. Slash and burn agriculture, primarily for corn cultivation, is a common practice in both the Bladen and Swasey Branches and contributes to the loss of ripa rian buffers (Esselman and Buck 2007) Human activities associated with slash and burn (e.g. irrigation, pesticide use) also contribute to the impact of this lan d use practice on aquatic ecosystems in the MRW. In stream gravel mining, largely in the Swasey Branch, is expected to change the natural flow regime and alter habitats. In addition, commercial banana farming in the Swasey and Bladen Branches represents a continued stress on riparian zones through reduced buffer width. Although the re engineering of drainage ditches has reduced the amount of run off from banana plantations reaching the main channel, irrigation for banana production may affect the hydrolog ic regime of the MRW. It is estimated that between 2.5 and 3.5 X 10 6 gallons of water per day are pumped from the Monkey River (primarily Swasey and Bladen Branches) for banana irrigation (GUARD 2007) In addition, Monkey River Village (Figure 2 1), located at the mouth of the Monkey River,

PAGE 29

29 vi llage (GUARD 2007) Although erosion is a natural process that occurs along the coastal margin, reduced transport and delivery of material from the middle reaches of Village. The combined impacts of channel alteration resulting from gravel mining and water abstraction contribute to reducing the amount o mouth (GUARD 2007) As with any rapid assessment technique, ELSI has several limitations. Fi rst, impact mapping provides an imperfect assessment of conditions and relies on assumptions regarding terrestrial aquatic interactions. These assumptions come from expert opinions and publications on temperate zone ecosystems, and require quantitative va lidation within tropical contexts. Second, ELSI provides an estimate of the intensity of stressors in each river reach, and does not consider propagated, cumulative downstream effects in the river network. This is evident in the ELSI percent distribution of stressors related to flow alteration and the resulting change detection map (Table 2 2, Figure 2 4). Although flow alterations are few relative to other observed stress types, the downstream effect of these impacts (e.g., in stream gravel mining, damm ing and pump houses for irrigation) may cause the most severe impact on the integrity of the Monkey River (GUARD 2007) In additio n, the calculation of ELSI scores assumes that stress ranks are additive and the interaction between stress types is linear. Finally, ELSI, because it draws on the location of visible activities in the riparian corridor, cannot detect unseen threats (e.g. agrochemical pulses), and has no way to identify legacy effects of past land use on freshwater systems. It is recommended that

PAGE 30

30 ELSI serve as a first step that should be incorporated into a multi scalar monitoring program that includes catchment scale mo nitoring, and detection of non visible threats such as toxics and heavy metals. ELSI provides an accurate assessment of the location and severity of stresses. Assessing these threats to riparian and freshwater ecosystem structure and function, however, is not enough. Conservation minded organizations must follow up on the information gathered to close the management gap between data collection and decision making. In Belize, a legal framework exists for conservation of riparian zones. The Crown Lands Ru le (Subsidiary Laws of Belize Vol. IV, 14) requires that 66 feet from the high water mark be left in un cleared vegetation along all water frontage in rural lands (Boles et al. 2008) In addition, the National Lands Act (No. 6 of 1992) requires that a 66 foot wide strip of land adjacent to streams, rivers or open water be left in its natural state unless the Ministry of Natural Resources and Environment gives approval for it to be used in a specific manner (Boles et al. 2008) Although these laws provide a means for riparian forest conservation, successful conservation of freshwater resources in Belize will require the c ombined efforts of the science, policy and education communities (Buck et al. 2008) Science including rapid assessment approaches like ELSI, backed up by more focused assessment of specific p roblems can be used to inform conservation strategies and define conservation priorities. Ultimately, however, political entities, educational institutions, NGOs and private industries need to develop integrated management structures that facilitate inter party coordination on water related issues.

PAGE 31

31 Conclusions Effective conservation of aquatic resources within the Maya Mountain Marine Corridor (MMMC) and the greater Port Honduras watershed requires an integrated management strategy that incorporates human a ctivities at the terrestrial aquatic interface. ELSI is a rapidly deployed, low cost approach to quickly gather information that can yield spatially and temporally detailed information about human activities that contribute to ecosystem stress. Resultin g maps communicate results in a simple, compelling way and are very useful in outreach activities and as primary information sources for decision making. When applied repeatedly through time, ELSI can detect trends in human uses of the riparian corridor. It is important that the impacts identified by ELSI be substantiated with follow up studies to assess quantitatively the greatest environmental threats. Complementary approaches to ELSI (e.g., biotic indices of environmental quality) should also be used t o identify invisible threats and legacy effects that are not detected by human impact mapping. Applications of ELSI in Belize demonstrate that the method can be applied across many contexts, suggesting it is an appropriate approach for use outside of Beli ze.

PAGE 32

32 Table 2 1. Stress source relationships and the scientific literature used to justify these relationships. When no scientific literature was available, personal observation was used to justify relationships (adapted from Esselman 2001) Stress Sourc es of Stress* References Sedimentation No riparian buffer (NB) Wood and Armitage 1997; Osborne and Kovacic 1993; Lowrance et al. 1997 Drainage ditches (DD) Usher and Pulver 1994 In stream gravel mining (GRV) Brown et al. 1998, Sandecki 1989 Channeli zation (CHN) Brookes 1986 Grazing (GRZ) Metzeling et al. 1995; Owens et al. 1996 Road access (RD) Cline et al. 1982; Metzeling et al. 1995 Thin riparian buffer (TB) Wood and Armitage 1997 Nutrient L oading Drainage ditches (DD) Usher and Pulve r 1994 No riparian buffer (NB) Osborne and Kovacic 1993; Peterjohn and Correll 1984; Lowrance et al. 1984 Community use (CU) Quddus 1980 Grazing (GRZ) Line et al. 2000 Thin riparian buffer (TB) Lowrance et al. 1997 Toxins/Contaminants Drain age ditches (DD) Usher and Pulver 1994 No riparian buffer (NB) Usher and Pulver 1994; Lowrance et al. 1997; Nearly et al. 1993 Thin riparian buffer (TB) Lowrance et al. 1997; Nearly et al. 1993 Altered Flow R egime Drainage ditches (DD) Poff et al. 1997 Water pumping (PH) Poff et al. 1997 In stream gravel mining (GRV) Mas Pla et al. 1999 In stream dam (DAM) Thermal A lteration No riparian buffer (NB) Osborne and Kovacic 1993; Gregory et al. 1991 Drainage ditches (DD) In stream da m (DAM) Direct H abitat A lteration No riparian buffer NB) Gregory et al. 1991; Harmon et al. 1986 In stream gravel mining (GRV) Brown et al. 1998; Sandecki 1989; Kondolf 1997 Channelization (CHN) Brookes 1986 Water pumping (PH) Codes for each source of stress are in parenthesis

PAGE 33

33 Table 2 2 Stress source rankings used for human impact mapping scoring criteria and the developing of the ELSI Index The contribution (contrib) and estimated irreversibility (Irrevers) of each source of stress is co mbined to create the qualitative source rank of low (L), medium (M), high (H), and very high (VH). These are converted to numberical values on a 1 10 scale to create the rank score. Stress Sources / (Source Code) Contrib. Irrevers. Source Rank Rank Score Sedimentation No riparian buffer (NB) VH H VH 10 Thin riparian buffer (TB) L H M 5 Drainage Ditch (DD) H H H / M 7.5 / 5 In stream gravel mining (GRV) H H H 7.5 Channelization (CHN) H H H 7.5 Cattle grazing (GRZ) VH H VH 10 Roa d access (RD) M H H 7.5 Nutrient Loading Drainage ditch (DD) VH M H 7.5 No riparian buffer (NB) H H H 7.5 Community Use (CU) M M M 5 Cattle grazing (GRZ) H M M 5 Thin riparian buffer (TB) L M L 2.5 Toxins / Contaminants D rainage ditch (DD) V H V 10 No riparian buffer (NB) H H H 7.5 In stream gravel mining (GRV) L H M 5 Thin riparian buffer (TB) M H M 5 Pump house (PH) L H M 5 Road access (RD) L H M 5 Altered Flow Regime Drainage ditch (DD) L M L 2.5 Pump house (PH) M M M 5 In stream gravel mining (GRV) L M L 2.5 Thermal Alteration No riparian buffer (NB) H H H 7.5 Drainage ditch (DD) L M M 5 Thin riparian buffer (TB) L M M 5 Direct Habitat Alteration No riparian buffe r (NB) M H M 5 In stream gravel mining (GRV) H M M 5 Channelization (CHN) H H H 7.5 Pump house (PH) L M L 2.5 Cattle grazing (GRZ) M M M 5 Sandbag dam (DAM) L L L 2.5 Thin riparian buffer (TB) L M L 2.5 betwe en 2000/02 and 2007 decrease from high to medium because of changes to drainage ditch design implemented by the Belize Banana Growers Assoc.

PAGE 34

34 Figure 2 1. The Monkey River Watershed and the primary villages located within the catchment.

PAGE 35

35 Figure 2 2 Percent distribution of stresses observed from human impact mapping in the MRW in 2000 and 2007. Overall Percent Distribution of Stresses (2000 2007) 0 0.05 0.1 0.15 0.2 0.25 0.3 Sed Nutr Toxic Flow Alt Therm Alt Habitat Stress Type Percent (%) 2000 2007

PAGE 36

36 Figure 2 3 T he observed sources of stress in each of the three branches of the Monkey River (Bladen Branch = A; Trio Branch = B; Swasey Branch = C) as well as the main stem of the Monkey River (D) bel ow the confluence of the Bladen and Swasey branches. Very few observations were made along the Monkey River (D). (Note the (NB=no bank; TB=thin bank; GRZ=grazing; CU=community use; RD=road; PH=pump house; CHN=channelization; GRV=gravel mining; DD=drainage ditch). C A B D

PAGE 37

37 Figure 2 4. Change in impacts and estimations of stress intensity for sedimentation (A) and flow alteration (B) for the MRW from 2000 2007. Similar maps were cre ated for all primary stresses. While ELSI does not accurately account for the upstream downstream communication of impacts within the riparian zone corridor, sedimentation and flow alteration are changing dynamics related to material transport in the MRW and severely impacting downstream habitats and human livelihoods (e.g., Monkey River Village).

PAGE 38

38 Figure 2 5 Change in the Overall ELSI Index for the MRW from 2000 2007.

PAGE 39

39 CHAPTER 3 LAND USE PRACTICES, PESTICIDE APPLICATIO NS, AND FRESHWATER RESOURCE CONS ERVATION IN A LOWLAND TROPICAL WATERSHED Introduction Pesticide use in Central America is driven by large scale agricultural production, particularly for export crops sugar cane, coffee, bananas, and other fruits. This agricultural sector accounts for ~85 % of pesticides used in the region (Galvao et al. 2004) Importation of pesticides in Central America nearly tripled between 1992 and 2001, t otaling more than 46,000 tons in 2001 (Galvao et al. 2004) In Belize, pesticide importation increased 36% between 1992 and 2000, from 952 to 1,296 tons. Lower pesticide importation in 2001 (831 tons: (Galvao et al. 2004) was followed by several years of high importation. Pesticid e imports in 2002, excluding chemicals associated with wood preservatives, totaled 1,311 tons. Import totals to Belize for 2003, 2004 and 2005 were 1,339, 1,235, and 1,290 tons, respectively (BPCB) dominated by l arge scale agricultural production. Agricultural exports in 2005 and 2006 totaled 85% and 74% of all exports, respectively, with sugar cane, papaya, citrus and banana accounting for approximately half of that production (BMAF 2006) In addition to this large scale agricultural production, small scale agriculturalists provide the domestic Belize market with a subst antial amount of food products. Belize has been self sufficient in rice production for approximately 10 years (B MAF 2006) Historically, scale farmers. For many years, and one half of the rice consumed in the country (Wilk 1997) Today they produce approximately 15% (BMAF 2006) cassava ( Manihot

PAGE 40

40 esculenta ) production occurs in southern Belize. In addition, cacao ( Theobroma cacao ) is supplying a growing international chocolate market (Levasseur and Olivier 20 00, Emch 2003) The trend towards market integration among rural farmers has been observed across Central America (Humphries 1993, Wilk 1997, Emch 2003, Keys 2004, Hamilton and Fischer 2005, Keys and Chowdhury 2006) Integration with larger markets and associated strategies for intensification, have been accompanied by increased introduction of pesticides into rural communities. Pesticide in troduction has been associated with an increase in public and environmental health concerns (Humphries 1993, Whit aker 1993, Popper et al. 1996, Azaroff and Neas 1999, Shriar 2001, Blanco et al. 2005) Education efforts in rural communities can increase awareness about safe handling of pesticides (Popper et al. 1996) but low literacy and education levels, inadequate protective clothing, and poorly functioning equipment preclude the safe use of pesticides, thereby impacting rural farmers and th eir households negatively (Popper et al. 1996, Azaroff and Neas 1999, Blanco et al. 2005) Although publ ic health concerns regarding pesticide use by rural, small scale farmers in Central America have received some attention in the literature (Azaroff and Neas 1999, Blanco et al. 2005) little att ention has been given to small scale pesticide use and its impacts on the natural resource base on which these rural communities rely. Rural communities throughout tropical regions tend to aggregate near freshwater resources that are used for drinking, wa shing, livestock, irrigation, and utilities, as well as sources of dietary protein (e.g., snails, fish, turtles) (Wilk 1985, Sala et al. 2000,

PAGE 41

41 McClain and Cossio 2003, McClain 2008) Pesticide use, particularly proximate to freshwater resources, and the cleaning of equipment such as backpack sprayers and pesticide containers in rivers and streams represents a thr eat to freshwater resources (Castillo et al. 1997) farmers of southern Belize. The survey focuses on riparian zone land use and the use of pesticides during corn cultivation in riparian zones. I identify the most commonly used pesticides and discuss potential threats to freshwater resource conservation in a lowland tropical watershed. Study Area Study Villages and Watershed Fieldw watershed (Figure 3 1). Crique Sarco, on the banks of the Temash River, was founded about 1912 (Grandia 2004) and is one of the oldest villages in the watershed. Lucky Strike lies near Tomagas Creek, a tributary of the Temash River. Lucky Strike, the youngest village in the watershed, was settled in 2003 by an extended family that came from nearby Crique Sarco. Sunday Wood is approximately 25 years old. It was founded by farmers from the nearby village of San Lucas (Grandia 2004) The village is near Sunday Wood Creek, where several small, 1 st order tri butary creeks drain the surrounding landscape. The headwaters of the bi national Temash River begin in Guatemala and flow eastward across the southern Toledo District of Belize towards the Caribbean Sea and Gulf of Honduras (Figure 3 1). Approximately 400 km 2 2 are in Belize. Soils are derived largely from Cretaceous age limestone with patches of

PAGE 42

42 dolomite and siliceous limestone (Wright et al. 1959) Mean annual rainfall is ~3800 mm with a dry season from December to April and a wet season from May to November. Mean monthly temperatures range from 18.7C (January) to 32.1C (May). Vegetation cover in the Temash River watershed is dominated by tropical evergreen lowland forests (Meerman et al. 2003b) The Sarstoon Temash National Park (STNP) is s ituated within the lower reaches of the Temash River (Figure 3 1) and the study villages are all located within the buffer zone of the national park. The STNP is a 40,000 acre community co managed national park. The Sarstoon Temash Institute for Indigenou s Management (SATIIM), a local non governmental organization, currently co manages the STNP in collaboration with the Belize Forest Department. SATIIM works with five communities adjacent to the park (including two of the three study villages, Crique Sarc management plan for the STNP (Herrera 2004) w as officially approved by the government of Belize in June 2005. Many of the ecosystems within the STNP depend on seasonal hydrology that leaves large areas of the park inundated for much of the year. The conservation and management of riparian zones wit hin the Sarstoon and Temash Rivers is a central component of the recently ratified management plan (Herrera 2004) Riparian Zone Land Use within the Temash River Watershed the cultivation of corn. Corn cultivation commonly occurs twice annually includ ing a slash and burn crop ( milpa ) that is grown during the rainy season (June November) and a slash and mulch crop ( matambre ) that is grown during the dry season (January May). Riparian zones and other low lying areas are preferred locations for plant ing

PAGE 43

43 matambre ( ecosystem services related to soil moisture retention and soil nutrient replenishment from occasional deposition of nutrient rich sediments during flooding. In addit ion, green mulch minimizes competition from weeds and protects the soil from compaction (Wilk 1997) repeatedly, and when fields are left fallow, mat ambre fallow periods are much shorter than wet season milpa fallow periods (Wilk 1985, 1997) Previous research suggests farmers (Wilk 1997) but is recognized as an integral component of the agricultural cycle. Contemporary matambre agriculture reflects long term cultural practices of the (Wilk 1985) Methods I conducted a series of semi structure Maya villages in the Temash River watershed. Semi structured interviews with seven village elders were conducted to better understand the history of matambre, general agricultural practices, and governance of n atural resources, with particular attention to riparian zones and associated freshwater ecosystems. Structured interviews were conducted with heads of households in the three villages. Ninety interviews were conducted out of a possible 103 households (87 %). Structured interviews were conducted between April and June 2008 and focused on practices associated with matambre, perceptions about water quality, use of pesticides in matambre fields, and the potential for conservation in the Temash River watershed Responses from structured interviews were summarized using SPSS software, release 10.0.0 (SPSS, Inc.).

PAGE 44

44 Results Households and Matambre In Crique Sarco village, 43 of 52 households (83%) were interviewed. Nine households either refused to participate or the head of the household was working outside the village at the time the survey was conducted (Table 3 1). In Lucky Strike village, eight of nine households (89%) were surveyed with one head of household working outside the village at the time the surve y was conducted (Table 3 1). In Sunday Wood village, 39 of 42 households (93%) were interviewed, with three households refusing to participate (Table 3 1). In Crique Sarco, 29 of the 43 interviewed households reported planting matambre in the dry season o f 2008. Sixty two percent of those households selected fields located within riparian zones, while 38% used areas classified as upland. Mean matambre field size in Crique Sarco was 1.03 hectares (ha). When asked to compare matambre and milpa field produ ctivity, approximately half of the farmers in Crique Sarco (48%) believed that matambre provided them with more corn (estimated as bags of corn per hectare) than the wet season milpa (Table 3 2). All households (100%) interviewed in Lucky Strike reported p lanting matambre fields in 2008, with 14% of those fields located within riparian zones. The reported mean field size of matambre in Lucky Strike was 1.45 ha. Matambre provided a higher yield of corn than milpa for all households surveyed in Lucky Strike (Table 3 2). In Sunday Wood, 32 of the 39 households surveyed (82%) reported planting matambre in 2008. Sixty two percent of these households reported planting their matambre in riparian zones, with an additional 19% planting in low lying areas away from a creek or river. Mean matambre field size in Sunday Wood was 0.96 ha. Seventy

PAGE 45

45 percent of the respondents in Sunday Wood stated that matambre provided more corn for their household than the wet season milpa (Table 3 2). Water Use and Water Quality In Cr ique Sarco, four households (<10%) reported collecting drinking water from either a creek or spring, whereas all respondents in Lucky Strike and Sunday Wood said they used either a creek or spring as their primary source for drinking water. All households in Crique Sarco reported utilizing the public water system that was installed in 2006, with 32% of the households augmenting with rainwater collection. None of the households in Lucky Strike reported utilizing rainwater and only 10% of households (n=4) r eported collecting rainwater in Sunday Wood. When asked about the water quality in the Temash River, respondents in Crique Sarco gave mixed responses, with 65% saying it was good. Twenty five percent of households in Lucky Strike and 19% in Sunday Wood th ought the Temash River water was clean enough to drink. When asked about the water quality in tributary creeks draining lands near their villages, responses differed. Fifty six percent in Crique Sarco and 63% in Lucky Strike felt that water in the tribut ary creeks was clean, whereas only 8% of the households in Sunday Wood felt so. In Crique Sarco, respondents considered pasture to have the largest impact on water quality, followed by domestic use (e.g., washing, bathing), and waste deposited directly in to the river. Households in Lucky Strike also considered pastures to be the primary stressor on water quality. Sunday Wood farmers reported that animals, particularly pigs, and domestic uses impacted water quality in Sunday Wood Creek and its tributaries.

PAGE 46

46 Matambre Field Management Matambre farmers who use riparian zones reported leaving a forested buffer strip between their fields and adjacent freshwater. In Crique Sarco, 69% of farmers reported leaving a buffer, while 75% of farmers in Lucky Strike and 8 9% of farmers in Sunday Wood reported leaving a buffer strip. The mean width of buffer strips reported by matambre farmers across all villages is 2.4 m. All matambre farmers reported weeding their fields at least once following planting. Eighty four per cent of these households used herbicides, typically applied four weeks after planting (Table 3 3). The most common herbicide used was Gramoxone followed by 2,4 D and Round Up (Table 3 4). Other agrochemicals used included Folidol and Tamaron. Of the farmers who reported using agrochemicals, 37% said they received training from the Belize Pesticide Control Board (BPCB). Discussion Matambre and Riparian Conservation Riparian forests and their associated floodplains provide an important buffer for freshw ater ecosystems against human impacts stemming from land use practices and land cover change (Naiman and Decamps 1997, Allan 2004) In Belize, a legal framework exists for the conservation of riparian zones. The Nati onal Lands Act of Belize (No. 6 of 1992) requires that a 66 foot wide strip of land adjacent to streams, rivers or open water be left in its natural state (Boles et al. 2008) Field obse rvations and interviews with village elders confirm that matambre fields are generally cleared to the edge of rivers or streams. Results from the structured household interviews contradict this claim. In part due to community outreach efforts by SATIIM, matambre

PAGE 47

47 matambre and its impact on riparian zones. The apparent disconnect between interview responses and matambre res pondent wishes to tell the interviewer what he/she perceives to be the desired response (Bernard 2006) SATTIM has engaged the local communities in a discussion about matambre through its outreach and education efforts, including a detai led mapping exercise of all matambre fields located within the STNP boundary. Likewise, the STNP management plan states that matambre within riparian zones is a direct threat to the management of the national park and its biodiversity (Herrera 2004) Despite these efforts, this conservation issue remains at odds with the land u se choices Matambre Pesticides and Public Health Of greater concern, from both an environmental and human health perspective, is the use of agrochemicals in matambre fields. Prior research on matambre was limite d River watershed (Wilk 1985, 1997) commonly used 2,4 D for weed control, Wilk (1997) made no mention of its use within matambre fields. Sim milpa farmers in the Temash to use machetes for weeding instead of agrochemicals because of both potential health risks and possible loss of wild foods that grow within agricultural fields. Resul ts from farmer interviews in this study point to increasing use of 3, Table 3 4). Public health concerns related to the improper use and disposal of agrochemicals in developing countries has been well documented (McConnell and Hruska 1993, Murray 1994, van Wendel de Joode et al. 1996, Escobichon 2001, Wesseling et al.

PAGE 48

48 2001a, Wesseling et al. 2001b, Wesseling et al. 2001c, Murray et al. 2002, Wesseling et al. 2005) Official surveys of regional hospitals, rural health posts and doctors from across Central America, conducted in coordination with the PanAmerican Health Organiz ation (PAHO) between 1992 and 2001, documented 43,368 acute intoxications, with an estimated 4,323 deaths resulting from exposure to pesticides (Galvao et al. 2004) No information on pesticide related illnesses or deaths are available for Belize and t hese values for Central America likely represent a minimum, as the majority of pesticide related illnesses go unreported in the region. Estimate s as high as 400,000 poisonings per year have been suggested for Central America (Murray et al. 2002) The agrochemical used most by matambre farmers in the Temash River bipyridinium dichloride) is one of the highest selling pesticides i n the world (Wesseling et al. 2001c) Paraquat, sold under many of its trade names, is listed as a res tricted use pesticide by (BPCB 2008) In Central America, farmers and plantation workers are most often exposed to paraquat via contact with the skin as a result of poorly functioning application equipment (e.g., backpack sprayers) or chemical spills. These occupational exposures cause chemical burns and lesions, most often on the back, wrists, and/or groin area (van Wendel de Joode et al. 1996, Penagos 2002) Paraquat is one of the primary pesticides associated with acute poisonings worldwide (Wesseling et al. 2005) and is the most common pesticide associated with fatal and suicidal poisonings in Central America (Wesseling et al. 2001a) The second most commonly used pesticide in the Temash River watershed is the herbicide 2,4 D (2,4 Dic hlorophenoxyacetic acid). Extensive research has been

PAGE 49

49 conducted in the developed world on 2,4 D and potential links to non Hodgkins lymphoma (Hardell 1979, Zahm et al. 1990, Cantor et al. 1992) Alt hough mechanisms of causality have been debated in the literature (Garabrant and Philbert 2002) studies have repeatedly identified farmers and agricultural workers as an at risk population for non Hodgkins lymphoma. Pesticide use, particularly use of 2,4 D, has been identified as a primary cause (Dreiher and Kordysh 2006) Exposu re to 2,4 D has also been linked with increased risk of gastric cancer (Ekstrm et al. 1999, Mills and Yang 2007) although this has been refuted by the pesticide industry (Burns et al. 2007) Results from the household survey show that only 37% of matambre farmers using pesticides have received training in sa fe handling of pesticides from the BPCB. In order to purchase pesticides, farmers are required to present proof of training by the BPCB. The Belize Pesticide Control Board struggles to cover rural areas of Belize, particularly the southern part of the co untry, where only two staff members are currently charged with pesticide training and monitoring. Farmers without official training likely obtain pesticides either from other farmers within their village that have the necessary proof of training or throug h the unregulated sale and trade of items with traveling sales people from neighboring Guatemala. Environmental Fate of Agrochemicals in the Temash Ecotoxicology studies and the assessment of environmental impacts of pesticides have historically been condu cted in temperate regions. There are insufficient data on almost all pesticides DDT being a notable exception (Kammerbauer and Moncada 1998) at higher temperature and higher humidity in tropical environments (Castillo et al. 1997, Kammerbauer and Moncada 1998, Marques et al. 2007) The research and regulatory capacity in most Central American countries to monitor pesticides in the

PAGE 50

50 environment is generally i nadequate, although Costa Rica and Nicaragua have extensive experience that could be applied to other areas in the region (Wesseling et al. 2001a) Studies in North America on the fate and transport of paraquat and 2,4 D suggest that they pose a threat to the environmental health of the watershed. Paraquat has a long half life in soils, ranging from 16 months to 13 years (Rao and Davidson 1980) Paraquat and its breakdown products can adsorb onto soil particles and be easily transpor ted to freshwater environments during runoff. Exposure to paraquat can directly alter the structure and function of aquatic and semi aquatic environments. Paraquat can cause fatalities in the zooplankton and benthic communities of lakes (Gagneten and Marchese 2003) and streams (Burnet 1972) resulting in prolonged alteration to community structure. It is considered slightly to moderately toxic to many species of fish and damages lung, liver and kidney tissue (Eisler 1990, Parma de Croux et al. 1999) Although 2,4 D decays rapidly in the environment (half life of 7 10 days), it is on water contaminants (USEPA 2008) and has been documented in groundwater in several states in the US and Canada (Howard 1989, Fitzgerald et al. 2001) Although its toxic effect on many trophic levels in aquatic environments is considered slight or negligible (Table 3 4), it has been documented to impact feeding behavior and can cause mortality in at least one fish species (NRCC 1978) Post mortality studies following a massive die off of corals in the Gulf of Chiriqui, Panama, documented high 2,4 D residues on coral heads (Glynn et al. 1984, Castillo et al. 1997) The source of pesticide was upstream rice farms (Castillo et al. 1997) This should be of particular

PAGE 51

51 concern for Belize, wh ere productive mangroves and sea grass beds along the coastal shelf give way to the longest barrier reef in the western hemisphere, the Mesoamerican Barrier Reef System (MBRS). The MBRS and the associated inner channel serve as rich nursery, breeding and feeding grounds, supporting a diverse and productive fishery. Current assessments of the coral reefs along the Belizean coast suggest the system is heavily impacted by tourism, coastal development and agricultural practices in the watersheds that drain in to the MBRS Organochlorines and heavy m etals have been documented in top predators across multiple aquatic habitats in Belize (Wu et al. 2000, Rainwater et al. 2002, Rainwater et al. 2007, Evers 2008) suggesting that agriculture practices may be responsible for contaminating freshwater, coastal and marine habitats. Bioaccumulation of these and other tox ins should be investigated in Belize and the greater MBRS region. Conclusions Riparian buffer strips can greatly reduce the amount of pesticides being transported across the terrestrial aquatic boundary (Asmussen et al. 1977, Lowrance et al. 1997a) Movement of herbicides in surface runoff typically occurs during a short period of time after application and the condition of the riparian zone at the time of application is an important factor in limiting pesticide runoff (Lowrance et al. 1997b) The infiltration of pesticides is strongly influenced by soil properties, particularly soil organic matter and pH (Reddy and Gambrell 1 987) and the presence of pesticides and their degradation products in shallow groundwater depends on these surface soil properties and the time lag required for infiltration (Lowrance et al. 1997a, Puckett and Hughes 2005) In riparian zones of southern Belize where matambre farmers are applying herbic ides annually, there is great potential for combined runoff and shallow

PAGE 52

52 groundwater infiltration. Research is needed on the movement and transport of pesticides in the lowland tropical watersheds that are home to these rural farming communities. Conservat ion of riparian zones within the Temash River watershed requires a focus on their combined social and ecological function. Riparian zones have a central role in the livelihood strategies and the cultural tradition of matambre farmers in the Te mash watershed and the rest of southern Belize (Wilk 1985, 1997) In addition to their ability to reduce the transport of contaminants across the terrestrial aquatic boundary, riparian zones provide numerous other ecosystem functions related to biodiversity conservation (Allan 2004, Naiman et al. 2005) Future directions for conservation of riparian zones should focus on: 1) education and outreach related to the importance of riparian zones in maintaining ecosystem function; 2) human and environment health concerns associated with agrochemical use; and 3) determination of optimum riparian buffer widths that will allow matambre farmers to cultivate within riparian zones while minimizing adverse effects from riparian forest removal and pesticide application. Land use activities within the riparian zone of the Temash River highlight the importance of reconciling the perceived differences between conservation goals and the land use practices and resource needs of Maya communities in southern Belize.

PAGE 53

53 Table 3 1. Number of households surveyed and the mean household size (% of total households in village; = standard deviation) Crique Sarco Lucky Strike Sunday Wood Overall households surveyed 43 (83%) 8 (89%) 39 (93%) 90 (87%) mean household size 5 ( 2.4) 4.6 ( 1.4) 5.4 ( 2.1) 5.1 (2.2) Table 3 2. Summary of matambre characteristics for each villa ge ( = stdev ) Crique Sarco Lucky Strike Sunday Wood Overall households practicing matambre 29 (67.4%) 7 (87.5%) 32 (82.1%) 68 (75.6%) % of households with higher yield from matambre 48% 100% 70% 64% % of matambre farms within riparian zone 62% 14% 62% 57% matambre in upland area 38% 86% 19% 34% matambre in lowland area 19% 8% mean travel t ime to field (minutes walking) 35 ( 22) 12 ( 8) 52 ( 17) 41 (22) mean size of field (hectares) 1.03 ( 0.74) 1.45 ( 0.32) 0.96 ( 0.39) 1.04 ( 0.57) Table 3 3. Agrochemical use among matambre farmers Crique Sar co Lucky Strike Sunday Wood Overall matambre farmers using herbicides 21 (n = 29) 7 (n = 7) 29 (n = 32) 57 (n = 68) frequency of herbicide application (% of farmers applying herbicide) once = 95% twice = 5% once = 72% twice = 28% once = 68% twice = 32% once = 79% twice = 21% matambre farmers receiving agrochemical training from Pesticide Control Board 6 1 18 25

PAGE 54

54 Table 3 4. Common agrochemicals of matambre farmers and their toxicity to aquatic organisms Acute Toxicity to Aquatic Organisms* # of farmer s Amphibian s Insect s Crustacea ns Mollusk s Fish Paraquat (sold as Gramoxone) 30 (53%) Slight Not Acute Moderate Slight Moderate to High 2,4 D 18 (32%) Slight Slight Not Acute Not Acute Moderate to High 9 (15%) n/a n/a n/a n/a n/a Toxicity summarized from (Kegley et al. 2008)

PAGE 55

55 Figure 3 1. Map of the Temash River watershed including the three study villages. Th e lower reaches include the Sarstoon Temash National Park.

PAGE 56

56 CHAPTER 4 SOIL NUTRIENT DYNAMI CS, ORGANIC MATTER T URNOVER AND LAND USE WITHIN AN ANTHROPOGE NIC LANDSCAPE DOMINATED BY SHIFTIN G CULTIVATION Introduction The Maya Forest region of northern Mesoame rica includes much of southern Mexico (the states of Chiapas, Campeche, Yucatn, and Quitana Roo), the northern Guatemalan department of the Petn, and Belize ( Figure 4 1 ). The region contains a wealth of natural and cultural resources and includes the mo st extensive tract of continuous tropical forest in the Americas, outside of Amaznia. The primary livelihood strategy in rural areas of the Maya Forest region is shifting cultivation, locally referred to as milpa The practice consists of clearing forest s by slash and burn, planting and cultivation of corn and other food crops, followed by fallowing before the area is re cleared and cultivated again. M ilpa has been practiced by indigenous people across Mesoamerica for millennia (Turner et al. 2001) and the contemporary expression of milpa use l ivelihood (Toledo et al. 2003) with corn ( Zea mays ) being the staple crop but farmers also actively engage in other land use activities including the cultivation of cash crops and pasture that serve to integrate rural farmers into local and regional markets (Humphries 1993, Keys 2004, Keys and Chowdhury 2006) Significan t research has been conducted on the impacts of milpa within the dry tropical forest of the Maya Forest region, particularly in the southern Yucatn peninsular region (SYPR). Milpa ejidos owned lands where ejido members practice milpa and often engage in the extraction of timber and non timber forest products as part of the land management (Turner et al.

PAGE 57

57 2001, Vance and Geoghegan 2002, Bray et al. 2004) Scholars consider milpa to be a dominant driver of land cover change in the SYPR resulting in large areas of secondary forest in varying stages of succession (Turner et al. 2 001, Chowdhury 2006) Above ground biomass within these secondary forests increases rapidly during succession while a return to values similar to mature forest stands requires between 55 95 years (Read and Lawrence 2003) A recent study also suggests that repeated cycles of clearing and burning can ultimately reduce the amount of carbon stored in the above ground biomass of secondary forests by 64% (Eaton and Lawrence 2009) However, in soils subjected to repeated cycles of clearing, burning, cultivation, and abandonment, the size of the carbon stock (i.e., g C/m 2 ) re mains relatively stable (Eaton and Lawrence 2009) Hughes et al. (1999) noted a similar pattern in the soils of secondary forests of the Los Tuxtlas region, northwest of the Yucatn, where milpa is also the dominant land use. Variability in the spatial distribution of soil nutrients in the SYPR is less influenced by land use than by tree related patterns of soil nutrients (Dckersmith et al. 1999) or variations in precipitation and topography across the peninsula (Lawrence and Foster 2002) Elsewhere in the Maya Forest region, the Lacandon Maya of southern Chiapas actively manage secondary forests to promote certain species with greater leaf litter production that aids in the re accumulation of soil organic matter following shifting cultivation (Levy Tacher and Golicher 2004) Several of these tree species have also been shown to limit harmful soil nematodes, effectively improving soil conditions during fallow (Diemont et al. 2006) The system of milpa as practiced by the Lacandon Maya

PAGE 58

58 maintains local biodiversity and soil ecology within a traditional subsistence livelihood strategy (Nations and Nigh 1980, Diemont and Martin 2009) Early studies of the impact of milpa on soils in the Petn region of the Maya Forest suggest that burning above ground biomass results in increases in some major exchangeable cations whereas organic matter, nitrogen, and phosphorus all decrease after burning (Cowgill 1962) During and dire ctly after cultivation, soil nutrients declined relative to neighboring forest soils and this decline was most pronounced in fields subject to successive years of cultivation (Cowgill 1962) Following milpa abandonment, early forest succession occurs rapidly in the Petn (Ferguson et al. 2003) However, episodes of human migration, dating back to the late 19 th century and continuing through the la te 20 th century, have resulted in significant changes in land use patterns in Petn. Approximately half of Petn forests have been cleared within the past 40 years as a result of rapid colonization, and milpa now competes with pasture development as the pr imary land use (Sader et al. 1997, Carr 2004, 2008) In addition, immigrants to Petn who practice milpa from other regions of Guatemala, are rapidly clearing forest, do not share the same degree of ecological knowledge about the forests of Petn as resident Itz Maya farmers, and do not follow some of the soil conservation practices of resident Itz Maya far mers (Atran 1993, Atran et al. 1999) This study was conducted in the southeastern lowlands of the Maya Forest region, in southern area in the late 19 th century from Guatemala, have no legal tenure to the land, but maintain customary land use practices that revolve around milpa agriculture and also

PAGE 59

59 include organ ic cacao ( Theobroma cacao ) agroforestry and pasture. I characterize soils within this anthropogenic landscape using a synchronic sampling design and present data on soils collected across a chronosequence of land covers including active milpa fields, fall owed fields of varying successional stages, mature forests, cacao orchards, and pasture. The objectives of the research were to: (i) describe soil nutrients within the milpa cacao agroforestry and pasture relative to mature forests by estimating the time needed for soils to recover to pre disturbance conditions; and (iii) consider the sustainability of these different land uses within the context of the Maya Forest. Stu dy Area The study was conducted in the Temash River watershed of southern Belize that Figure 4 1 ). Whereas prehistoric ived in southern Belize from the Alta Verapaz of Guatemala during the 1890s as part of a large scale labor migration to satisfy the demands of a growing cacao, coffee and banana plantation economy (Wilk 1997, Grandia 2004). Although the plantation economy diversified production strategy that included food staples and products for external milpa agriculture. Detai led descriptions of this land use practice for the Maya communities of Southern Belize can be found elsewhere (Wilk 1997, Grandia 2004) In addition to milpa agriculture for corn and other food cr Maya communities, have become suppliers to a rapidly expanding international market for organically grown cacao (Levasseur and Olivier 2000, Emch 2003)

PAGE 60

60 Maya also raise livestock and pasture is a rapidly expand ing land use in the Temash and other parts of southern Belize. Mean annual rainfall in southern Belize is ~3800 mm, but is highly seasonal with most precipitation occurring between May and October. Mean monthly temperatures range from 18.7C (January) to 32.1C (May). The natural vegetation cover in the watershed is dominated by tropical evergreen lowland forests (Meerman et al. 2003a) and is within the tropical moist forest life zone (Holdridge 1971) The soils in the Temash River watershed fall within the Temash sub suite of the Toledo Suite, as originally described by Wright et al. (1959) and generally correspond with gleyic and haplic acrisols of classification system (Baillie et al. 1993) The soils are derived from interbedded calcareous sandstones and mudstones (Wright et al. 1959) and are acidic, with thick clay horizons that result in imperfect drainage (Baillie et al. 1993, Holland et al. 2003) Methods Soil Sampling and Land Use H istory Soil samples were collected from plots (N = 74) representing seven different land use/land cover categor ies (Table 4 1). Five soil cores of the uppermost 10 cm (0 10 one core was collected at the approximate central point and four additional cores were collected at the e nd of 10 m long transects radiating outward from the central point in Whirl pak bags. Bulk density for each plot was estimated using the cylinder method (USDA 2004) Samples for nutrient and stable isotope ( 13 C) analysis were air dried and sieved to retain the < 2 mm fraction.

PAGE 61

61 The land use history of each soil sampling site was determined through interviews w ith farmers from each of the five villages. In active milpa fields, farmers were asked how many consecutive years the fields had been cultivated and to describe the land cover prior to clearing (i.e., type and age of land cover). In successional stages o f regrowth (SS1 SS3), the age of the fallowed field (locally referred to as huamil ), and the land cover prior to the original clearing was recorded. In addition, descriptions of the dominant vegetation within each successional stage plot were recorded, including species present, canopy height, canopy closure, and diameter at breast height of canopy and emergent trees. The age classes approximate previously described age (Wilk 1981) Simila r information was recorded for cacao orchards and pastures (Table 4 1). Plant samples of the dominant vegetation types within each land cover were also collected for isotope analysis. Laboratory Analysis Soil samples from each plot were homogenized and sp lit for individual analyses. Percent organic matter was determined by loss on ignition (LOI) following Nelson and Sommers (1996) and USDA (2004) Samples for macro and micronutrient analysis were sent to Waters Agricultural La bs, Inc., and analyzed using methods described in (Gavlak et al. 1994) Soil pH was determined in a 1:1, soil: water solution. Soil NO 3 was determined by colorimetry following cadmium reduction. All other macro and micronutrient analyses were conducted using a Mehlich I extraction and measured by inductively coupled plasma on a Thermo ICAP AES 6500 and reported as parts per million (ppm). Percent C and N were determined by flash combustion on a Carlo Erba NA 1500 CNS elemental analyzer. Values were converted to g/m 2 usi ng the bulk

PAGE 62

62 13 C) and isotopic analyses were measured simultaneously using a VG Prism Series II isotope ratio mass spectrometer with a triple trap preparation device linked to the elemental analyzer. Estimation of SOM Tur nover In landscapes with undisturbed C3 forest vegetation, active milpas pastures, and 13 C of SOM in surface soils can be used to determine the relative contribution of C3 vs. C4 vegetatio n to the t 13 C t ) and of all end members must be 13 C 3 veg ), C3 13 C 3 soil ), C4 13 C 4 veg ), and C4 13 C 4 soil ). In situations where a distinct SOM 13 13 C of the C4 13 C 4 veg ) can be used (Wiesenberg et al. 2004) From these values, it is po ssible to estimate the percent ( x ) of carbon coming fro E quation 4 1 (Balesdent et al. 1987) : 13 C t 13 C 4 veg 13 C 3 soil ) x 13 C 3 soil (4 1) Decomposition of SOM is assumed to follow first order kinetics (Balesdent and Mariotti 1996) and can be calculated using Equation 4 2 (Wiesenberg et al. 2004) : k = ln(C3 t /C3 t0 ) / (t t0) (4 2) Where k is the decay constant, C3 t equals the percentage of C3 carbon remaining in the soil after t years since clearing of original C3 vegetation, and C3 t0 is the percentage of C3 carbon in the soils at the time of clearing (assumed to equal 100). The mean residence time (MRT) equals the average time soil C resides in the SOM reservoir under steady state conditions (Derrien and Amelung 2011) MRT can be calculated using Equation 4 3 (Wiesenberg et al. 2004) :

PAGE 63

63 MRT = 1/ k (4 3) Statistical A nalyses The non parametric Mann Whitney U statistic (p significant differences between mature forest soils and soils from all other land cover classes. When ties in the rank scores for soil classes were present, the distribution estimate was corrected for ties, including a correction for continuity (Hintze 2001) The z score and corresponding probability value are provided for these cases. a post hoc bonferroni adjustment for multiple tests (adjusted P value = 0.00067). R esults Soil Physical and Chemical Characteristics Percent organic matter (%OM) varied little across land use/land cover classes. Milpa fields had the lowest %OM whereas cac ao orchards had the highest %OM, although no anthropogenic soil classes were significantly different from forest soils (Table 4 2). Similar patterns for carbon (g C/m 2 ) and nitrogen (g N/m 2 ) were also observed (Table 4 2 ). Median values for the C:N ratio for pasture soils were higher than the C:N ratio of forest soil s (U = 18, P = 0.027) (Table 4 2 Figure 4 2). Cation exchange capacity (CEC) for all soils ranged from 14.2 to 17.3 meq/100g and soils across all classes were acidic, with pH values between 5.0 and 5.7 (Table 4 2). Aluminum (Al) concentrations in SS1 soils were lower than forest soils (U = 45, z = 2.2181, P = 0.027). Soil nitrate (NO 3 N) concentrations from milpa (U = 17, P = 0.012), SS1 (U = 45, P = 0.027), SS2 (U = 23.5, z = 2.457, P = 0.0 14) and pasture (U = 8, P = 0.001) were

PAGE 64

64 significantly lower than forest soil NO 3 N (Table 4 3, Figure 4 2). Phosphorus (P) concentrations were low across all soil classes. Potassium (K) concentrations were elevated in milpa (U=8, z = 3.1161, P = 0.002), SS1 (U = 39, z = 2.489, P = 0.013), cacao (U = 0, P = 0.0014) and pasture (U = 9, P = 0.002) soils relative to forest soils (Table 4 3, Fig 4 2). Magnesium (Mg) concentrations in milpa soils were also elevated relative to forest soils (U = 19, P = 0.02) (Table 4 3, Figure 4 2). No significant differences were observed in calcium (Ca) concentrations across any soil classes. Sulfur (S) concentrations were lower than forest soils within milpa (U = 21, z = 2.1787, P = 0.029), cacao (U = 2.5, z = 2.3503, P = 0.019) and pasture (U = 19, z = 2.1411, P = 0.032) (Table 4 3, Figure 4 2). No significant differences were detected between the soils of anthropogenic land covers and forest soils for any of the mi cronutrients measured (Table 4 4 ). Strong positive cor relations (r > 0.85) were observed between %OM, g N/m 2 and g C/m 2 and these three soil characteristics were also correlated with CEC (r > 0.7) (Table 4 5 ). Ca was also correlated with %OM, g N/m 2 and CEC (r > 0.7), with a weaker correlation between Ca a nd g C/m 2 (r = 0.65). In addition, Boron (B) was positively correlated (r > 0.70) with g N/m 2 Ca, and pH, and had a weaker positive correlation with %OM (r = 0.63), g C/m 2 (r = 0.66), and CEC (r = 0.69). Boron also had a weak negative correlation with c opper (Cu) (r = 0.67). Cu was positively correlated (r > 0.70) with Iron (Fe) and both Cu and Fe were negatively correlated with pH (r = 0.72 and r = 0.68, respectively). Cu was also negatively correlated with Ca (r = 0.67). Aluminum (Al) had a weak negative correlation with CEC and a very weak positiv e correlation with Fe (Table 4 5 ).

PAGE 65

65 Stable Carbon Isotopes in Vegetation and Soils and the MRT of SOM Pasture grass ( B. humidicola ) and corn ( Z. mays 13 C values of C4 vegetation ( 13.13 and 12.84 respectively; Table 4 6 ). The pasture soils have 13 C values typical of SOM derived from C4 vegetation whereas milpa soils reflect the isotopic signature of C3 derived SOM (Table 4 5). All other l and cover classes have 13 C values typical of SOM derived from its dominan t C3 vegetation cover (Table 4 6 ). Within pasture soils, the relative contribution of C4 pasture grass to SOM ranged from 2.4% in a 1 year old pasture to 24% in a 28 year old pasture (Figure 4 2). Estimates for the MRT of SOM within pastures ranged from 14 yrs to 102 yrs (Figure 4 3 ). D iscussion Soil Nutrient Dynamics i n Milpa Versus Forest S oils The success of milpa agriculture is largely dependent on nutrient availabili ty in cultivated soils (Sanchez 1982, Kleinman et al. 1996) and the nutrient rich ash that remains following burning is a commonly cited mechanism for increasing soil fertility within milpa fields (Nye and Greenland 1960) The burning of above ground biomass release s a pulse of plant available nutrients and the resulting ash increases soil pH and cation exchange capacity (CEC) in tropical soils (Nye and Greenland 1960, Ewel et al. 1981, Tiessen et al. 1992, Giardina et al. 2000, McGrath et al. 2001, Arunachalam 2002) The pH and CEC of m ilpa soils in the Temash watershed did increase slightly relative to forest soils (Table 4 2) but this increase w as only significant at P=0.32 and P=0 24, respectively Wright et al. (1959) also noted little difference in pH between forested and cultivated soils in the Temash.

PAGE 66

66 Although ash can increase p H and CEC in tropical soils, nutrients particularly N, P and organic C can also be lost during and following the burn (Ewel et al. 1981, Tiessen et al. 1992, Giardina et al. 2000, Sommer et al. 2004) In Temash soils, total N and NO 3 were reduced in milpa fields relative to forests and NO 3 remaine d significantly lower than forest soils through the first two phases of succession (Table 4 3 Figure 4 2). The disruption of the N cycle by felling and burning tropical forests greatly reduces the amount of N uptake by vegetation, and increases in NO 3 i n soil solution have been observed following slash and burn (Uhl and Jordan 1984, Hlscher et al. 1997, Williams and Melack 1997) The accum ulation and immobilization of N in early successional weedy biomass helps to slow the loss of mineralized N via leaching (Lambert and Arnason 1986, Brubacher et al. 1989) but the fact that NO 3 in milpa soils does not return t o values similar to forest soils until > 15 years of fallow (SS3) is also a reflection of high N demand by early successional species (Ewel 1986) Available P content in Temash soils was relatively unchanged through cultivation and 2 nd succession and was never significantly differ ent from forest soils (Table 4 3 ). Soil P is derived from the physical and chemical weathering of geologic parent material, and in highly weathered tropical soils, any biologically available P is readily adsorbed onto cla ys and other inorganic constituents, becoming biologically unavailable (Vitousek and Sanford 1986, Tiessen et al. 1992, Lawrence and Schlesinger 2001) Although not measured in this study, approximately half of the P from bio mass burning reaches the soil surface (Giardina et al. 2000) and the organic P fraction in soils is often enriched following deforestation (Farella et al. 2007) In the acid soils of the Temash, Al and Fe likel y play an important role in reducing the amount of available P and lower

PAGE 67

67 concentrations of Al and Fe during early succession may aid in the maintenance of the limited amounts of P in these soils (Kleinman et al. 1996) The carbon (C) pool most heavily impacted by shifting cultivati on is above ground biomass whereas soil organic matter is the most stable (Ewel et al. 1981, Kotto Same et al. 1997) In a moist tropic al region of Costa Rica, Ewel (1981) estimated that approxi mately 31% of C stored in above ground biomass was lost due to burning. In the Amazon, estimates for the loss of C following slash and burn can be gre ater than 50% of the total above ground pool (Kauffman et al. 1995, Hughes et al. 1999) and as much as 75% of the initial C can be lost when shifting cultivation is carried out on the same plot for 3 4 years (Uhl 1987) Many other studies have also noted that land use int ensity including successive years of cultivation and repeated cycles of slash and burn is a critical factor when determining the impact of slash and burn on above ground C pools that can ultimately decrease the potential of second growth forests to seq uester C (Hughes et al. 1999, Zarin et al. 2005, Eaton and Lawrence 2009) How these land use practices impact soil C are not entirely clear In two of the above mentioned studies in which C pools in above ground biomass were impacted by the intensity of prior land use, C pools in soil remained stable (Hughes et al. 1999, Eaton and Lawrence 2009) Zarin et al. (2005) did not directly measure soil C. In other studies of slash and burn soils, the soil C content either changed l ittle or increased after felling and burning (Nye and Greenland 1964, Seubert et al. 1977, Kotto Same et al. 1997) In the Temash milpa soils, %OM and total C showed small but insignificant dec line s relative to forest soils (Table 4 2). However, some studies have documented declines in C following slash and burn. In her early study of milpa soils in the Petn,

PAGE 68

68 Cowgill (1962) documented decreases in %OM following the burn and Kotto Same et al. (1997) mention several other studies where OM C declined following burning. Salcedo et al. (1997) measured a 17% decrease in organic C following five years of cultivation on an Oxisol in northeastern Brazil. Organic matter is commonly volatilized at tem peratures between 200C and 315C (Lide 2004) During an expe rimental burn of slash produced from an 8 9 year old 2 nd growth forest in Costa Rica, Ewel (1981) recorded temperatures in excess of 400C within the burning slash 1 2 cm above the soil surface, but temperatur es at the soil surface averaged ~ 200C and dropped to ~100C at 1 cm soil depth and <38C at 3 cm depth. The C:N ratio decreased within the upper 3 cm during the slash and burn phase at this site but overall C storage was high in the soils (Ewel et al. 1981) In a mature (>100 yr old) dry forest site of central Mexico, surface temperatures exceeded 500C and declined to 100C at 3 cm soil depth (Giardina et al. 2000) Total C at this si te decreased by ~20% in the upper 2 cm following the slash burning but was unchanged from 2 5 cm soil depth (Giardina et al. 2000) The high soil surface temperatures at this site also alt ered soil N and P pools. Tota l N decreased in the upper 2 cm but was offset by an increase in plant available N within the 0 10 cm horizon with a similar decrease in organic P and increase in plant available P also occurring (Giardina et al. 2000) Variations in the maximum soil surface temperature during slash burning, and consequently the impact of burning on soil nutrient pools, are likely related to the age and stature of the cleared forest, the length of the dry season, and soil moisture (Giovannini et al. 1990, Giardina et al. 2000, Knicker 2007)

PAGE 69

69 Cacao Agroforestry and P asture In tropical regions, agricultura l practices that best mimic the structure and function of natural communities are considered more sustainable (Ewel 1986) Cacao agroforestry attempts to mimic the structure of a tropical forest by utilizing canopy trees for shade with cacao tre es occupying the under story C acao is cultivated across multipl e regions of the tropics both in large scale plantations and in smallholder orchards. Cacao cultivation by smallholder Maya farmers in southern Belize has expanded rapidly in recent years in an effort to meet an increasing international market demand for organic cacao beans (Emch 2003) The Toledo Cacao Growers Association (TCGA) was developed in the late 1980s to help facilitate the sale of cacao grown by farmers in southern Belize. Membership in th e TCGA has grown from 70 farmers in 1987 to more than 1000 members from 80 villages in 2008 (TCGA, unpublished data). Cacao orchards in the Temash were planted in 2003 2004 through assistance with TCGA and a local non governmental organization, the Sarstoo n Temash Institute for Indigenous Management (SATIIM). Cacao soils in the Temash have maintained forest soil characteristics (Tables 4 2, 4 3, 4 4) with K actually being significantly higher than forest soils (Table 4 3). Pasture expansion has long been c onsidered a driving force behind Central American deforestation, historically driven by external market demands for beef (Myers 1981). Government policies in Guatemala provided large tracts of land at low prices to land speculators in the Petn, resulting in rapid pasture expansion in that region (Schwartz 1990) Although policies have shifted slightly, pasture continues to expand, driven partly by new infrastructure development and by the need for land owners to

PAGE 70

70 establish clear land tenure (Kaimowitz 1995) Similar patterns are observed in Belize as well (Chomitz and Gray 1996) The Belize Ministry of Agricultur e conducts annual surveys of cattle pasture and, from 2005 2007, the number of cattle in the southern Toledo District increased by 47%, from 3515 to 5163, while pasture (both improved and natural) increased by 42%, from 6181 acres to 8789 acres (Ministry of Agriculture, unpublished data). During this same time period, the number of cattle in the Temash River watershed (within the Toledo District) increased by 110% from 194 head of cattle to 407 while pasture increased by 14% from 776 acres to 883 acres. Although this only accounts for three years of data, this may represent a general trend towards pasture expansion into the more remote areas of southern Belize like the Temash River watershed. Carbon content in pasture soils in the Temash is not signific antly different from the C content of forest soils. Several studies have noted the initial conservation of soil nutrient stocks following forest conversion to pasture (Buschbacher et al. 1988, Neill et al. 1997, Desjardins et al. 2004) Factors controlling the impact on soils of forest c onversion to pasture include land use history and pasture management (Buschbacher et al. 1988, Neill et al. 1997) Stabilization of organic C by Al organic matter complexes can also reduce the amount of carbon loss during conversion (Veldkamp 1994) Pastures do increase soil compaction and reduce root penetration, water infiltration and gas exchange (McGrath et al. 2001) and the low nitrate values in the Temash pasture soils (Table 4 2, Fig ure 4 2) may be related to compaction. Pastures in the Temash are frequently managed with fire and the increase in K in pasture soils (Table 4 2, Figure 4 2) is likely related to the repeated burning of above gr ound biomass.

PAGE 71

71 Stable Isotopes and MRT of SOM in Temash Soils When considering long term impacts of land use and land cover change on soils, the mean residence time ( MRT) of the carbon in soils is also important and may represent a more accurate measure of impacts from land cover change on soil structure and function than simply a measure of the amount of carbon per square meter in the soil (Trumbore et al. 1995, Balesdent and Mariotti 1996, Six and Jastrow 2002) The 13 C of milpa soils did not differ significantly from forest soils, thus precluding calculation of MRT using the above 13 C of corn is isotopically heavier than that of either forest plants or forest soils but it is no app arent in the 13 C of the SOM of Temash milpa soils. A primary reason for this is that litter fall from corn fields is only about 10% of forest litter production (Awiti et al. 2008) In addition, milpa sites in the Temash are rarely used for more than two consecutiv e years, effectively limiting the time for carbon from C4 plants to accumulate within the l eaf litter and topsoil. S tudies that have documented SOM changes under corn cultivation include sites where corn has been under cultivation for > 20 years (Cayet and Lightfouse 2001, Wiesenberg et a l. 2004) Awiti et al. (2008) estimate that maize derived ( C4 ) carbon would become the dominant source of carbon in soils after 38 years of continuous cultivation. Pasture soils in the Temash reflect th e shift in carbon input from C3 forests to C4 p asture grass. There is a strong positive correlation between the age of the pasture and the amount of C4 carbon in the soil (Table 4 6 Figures 4 3, 4 4). This pattern has been observed elsewhere in the tropics as well (Trumbore et al. 1995, Desjardins et al. 2004) and factors influencing the rate of incorporation of C4 carbon into soils include soil type (clayey versus sandy), soil particle size, climate, type of pasture grass, and pasture management practices (Trumbore et al. 1995, Desjardins et al. 2004) The

PAGE 72

72 MRT of Temash pasture soils is also related to pasture age the oldest pasture (28 ye ars) has an MRT of ~ 100 years. The estimates for MRT are conservative however, and may in fact be much longer. The model for determining the percent contribution of C3 vs C4 carbon assumes that inputs from C3 carbon sources stop after conversion to past ure. Tropical forest soils have active root zones down below 1 m in the soil and carbon cycling on the order of 100 1000 years can occur at depth (Nepstad et al. 1994, Trumbore et al. 1995) Replacing deep rooted trees with shallow rooted pasture grasses disrupts this soil carbon cycling at depth (Trumbore et al. 1995) Across the tropics, the potential for pasture to serve as a carbon sink varies widely and accurate estimates must also account for the loss of C stored in above ground biomass (Desjardins et al. 1994, Fearnside and Barbosa 1998) In the Temash soils, MRT of carbon is generally greater than estimates for MRT in tropical soils (~25 yrs), suggesting that in certain circumstance s pasture can serve as a carbon sink. However, no estimate for above ground biomass for Temash forests has been calculated that would allow for a more accurate assessment of the impact of forest conversion to pasture on soils. Sustainability of M ilpa in t he Maya Forest The persistence of milpa as a land use strategy in the Maya Forest region suggests that it is an effective means of adapting to the natural environment that capitalizes on ecosystem services without exhausting them (Alcorn and Toledo 1998, Berkes and Folke 2002, Toledo et al. 2003) In the Temash River watershed, data on soil NO 3 suggest that soil recovery following slash and bu rn may take upwards of 15 years to return to pre disturbance conditions. This same timeframe for recovery of soil has been observed in soils used by smallholder farmers in the Brazilian Amazon as well

PAGE 73

73 (Farella et al. 2007) Uhl and Jordan (1984) noted that s oil nutrients recovered within five years of fallow in the Venezuelan Amazon whereas numerous others have documented little i f any change in major soil nutrients following slash and burn (Uhl 1987, Hughes et al. 1999, Eaton and Lawrence 2009) An important factor in determining the ability of forests to recover f ollowing slash and burn is land use intensity, particularly the number of consecutive ye ars under cultivation (Uhl 1987, Zarin et al. 2005) Moran et al. (2000) however noted that relative to other land uses successiona l forests from slash and burn have higher overall above ground biomass and higher dominance of canopy biomass than forests succeeding pasture or abandoned mechanized agr iculture. An important distinction that should be made when assessing the sustainabil ity of slash and burn agriculture in the tropics is the relationship of the farmer to the area he is farming. Myers (1993) vators can be significantly different and is often tied up in larger national and regional development agendas that are not easily recognized at smaller scales (Lambin et al. 2001, Vance and Geoghegan 2002) In Petn, Guatemala, land development policies both inside and outside the region cre ated a group of farmers unfamiliar with the natural environment of the lowland forests (Atran et al. 1999, Carr 2002) cultiva tors have been a primary driver of deforestation in the Petn. The examples of the Lacandon Maya (Diemont et al. 2006) and the ejido system of the SYPR (Bray et al. 2004) suggest that Maya farmers can d evelop an agricultural syste m that is sustainable, and soil data from the Temash support this idea as well. A major challenge facing small scale agriculturalists in the Maya Forest region and throughout the tropics is

PAGE 74

74 related to market integration and the introduction of new crops that can negatively impact traditional agricultural practices (Humphries 1993, Ke ys 2004) Cash crops, such as cacao, grown within an agroforestry system can have minimal impacts on local biodiversity and ecosystem function (Steffan Dewenter et al. 2007) but smallholders would benefit from development assistance that provides trai ning in labor saving techn iques including intensive shade tree management, arboriculture, and green manure practices (McGrath et al. 2000, Rosenberg and Marcotte 2005)

PAGE 75

75 Table 4 1. Dominant land cover classes in the Tema sh River watershed. Soils were collected from each land c over class (sample size in parenthese s). Land Cover Class Q'eqchi' Name* Land Cover Age Dominant Vegetation milpa c'at c'al < 1yr Zea mays (n=11) Herbaceous species Heliconia sp., Solanum sp., SS1 coc' pim 1 3 yrs Neurolaena lobata Ipomoea violacea (n=21) Broadleaf species Cecropia sp., Vismia ferruginea Broadleaf species SS2 coc' che ru or 4 15 yrs Cecropia sp., Vismia ferruginea (n=14) coc' al c'al Melastomatacea, Acacia cornigera Atalea cohu ne Bursera simaruba Broadleaf species Atalea cohune Bursera simaruba SS3 ninki al c'al 16 30 yrs Vochysia hondurensis Terminalia (n=5) amazonia Understory species Sabal mauritiiformis Chamaedorea spp. Broadleaf s pecies mature Atalea cohune Vochysia hondurensis forest nink li q'uiche' > 30 yrs Terminalia amazonia Bursera simaruba (n=9) or q'uiche' Pithecellobium arboreum Understory species Sabal mauritiiformis Chamaedorea spp. Broadleaf canopy Atalea cohune Vismia ferruginea cacao kakaw ~ 5 7 yrs Inga sp. (n=4) Understory species Theobroma cacao pasture n.d. ~ 2 28 yrs Brachiaria humidicola (n=10) Q'eqchi' names from Wilk (1981) and/or Grandia (2004)

PAGE 76

76 Table 4 2. Soil properties (mean standard error) of the seven primary land use/land cover classes in the Temash River Watershed The med ian value is shown in parenthese s*. Land Use Categories Milpa SS1 SS2 SS3 Forest Cacao Orchard Pasture (n = 11) (n = 21) (n = 14) (n = 5) (n = 9) (n = 4) (n = 10) OM 10.2 0.6 12.4 0.7 11.9 0.9 12.0 1.5 12.4 1.7 13.8 0.4 11.4 0.9 (%LOI) (9.5) (12.3) (12.9) (12.5) (10.4) (14.0) (11.9) C 3805 143 4450.3 311 4498 417 3219 294 4526 743 5012 320 4440 314 (g C/m2) (3673) (4259) (4240) (3332) (3521) (4937) (4555) N 293 16 365 31 367 38 261 38 392 75 411 37 332 25 (g N/m2) (290 ) (331 ) (365) (250) (302) (404) (352) C:N 13.1 0.3 12.5 0.3 12.5 0.3 12.8 0.9 12.1 0.5 12.3 0.3 13.5 0.3 (12.9) (12.8) (12.3) (12.3) (12.6) (12.3) (13.5)* CEC 14.7 0.6 17.3 0.8 16.5 1.6 15.4 2.1 14.2 2.0 16.1 0.4 14.2 0.9 (meq/100g) (14.0) (16.9) (17.1) (13.6) (12.6) (16.0) (1 4.4) pH 5.3 0.1 5.7 0.1 5.5 0.2 5.3 0.2 5.0 0.3 5.1 0.1 5.6 0.1 (in H 2 O) (5.2) (5.7) (5.3) (5.2) (4.9) (5.0) (5.6) Al 144.0 12.0 137.1 19.3 171.7 35.1 155.8 28.9 243.5 44.7 179.6 33.6 169.9 15.2 (ppm) (15 6.5) (110.8)* (112.3) (149.5) (205.0) (156.8) (179.5) median values with asterisk (*) indicate significant difference from the median forest soil value (Mann Whitney U test statistic, P<0.05).

PAGE 77

77 Table 4 3. Macro nutrients concentrations (mean standar d error) of the se ven primary land use/land cover classes in the Temash Watershed. The med ian value is shown in parenthese s*. Land Use Categories Milpa SS1 SS2 SS3 Forest Cacao Pasture (n = 11) (n = 21) (n = 14) (n = 5) (n = 9) (n = 4) (n = 10) Mac ronutrients NO3 N (ppm) 46 9 65 11 57 7 73 14 105 18 129 21 27 10 (60)* (43)* (49)* (72) (91) (118) (15)* P (ppm) 3 0.3 2.0 0.1 2.1 0.2 2.4 0.5 2.2 0.3 1.9 0.0 1.8 0.0 (2.7) (1.8) (1.9) (1.9) (1.8) (1.9) (1.8) K (ppm) 91 15 59 5 45 4 47 4 39 3 79 8 86 10 (84)* (57)* (41) (49) (40) (75)* (89)* Mg (ppm) 283 19 193 20 170 22 194 26 226 49 253 22 192 20 (281)* (147) (177) (187) (183) (266) (207) Ca (ppm) 1359 156 2223 143 2094 290 1845 414 1488 352 1707 146 1516 112 (1116) (2143) (2110) (1713) (1613) (1804) (1525) S (ppm) 3.5 0.4 7.3 1.3 9.7 2.2 3.9 1.3 7.9 2.0 2.8 0.5 3.6 0.3 (4.0)* (4.5) (7.0) (2.5) (5.0) (2.8)* (4.0)* median concentrations with asterisk ( *) indicate significant difference relative to the median forest soil value (Mann Whitney U test statistic, P< 0.05).

PAGE 78

78 Table 4 4. Mi cronutrient concentrations (mean standard error) of the se ven primary land use/land cover classes in the Temash Waters hed. The med ian value is shown in parenthese s*. Land Use Categories Milpa SS1 SS2 SS3 Forest Cacao Pasture (n = 11) (n = 21) (n = 14) (n = 5) (n = 9) (n = 4) (n = 10) Micronutrients B (ppm) 0.3 0.0 0.5 0.0 0.4 0.0 0.5 0.2 0.5 0.1 0.3 0 .0 0.3 0.0 (0.3) (0.5) (0.3) (0.4) (0.5) (0.3) (0.3) Zn (ppm) 1.0 0.0 1.1 0.1 1.0 0.1 1.2 0.3 0.9 0.1 0.9 0.1 0.9 0.1 (1.0) (0.9) (0.8) (0.9) (0.8) (0.9) (0.7) Mn (ppm) 40.4 11.2 26.6 3.1 32.6 5.1 29.6 3.0 31.9 7.4 38.0 7. 4 36.9 3.7 (34.0) (26.0) (28.5) (28.0) (30.5) (38.3) (34.0) Fe (ppm) 12.5 1.5 6.1 0.8 12.2 2.8 11.0 4.5 16.3 6.3 9.5 3.1 12.8 1.5 (11.9) (5.7) (9.4) (6.4) (8.4) (7.7) (13.6) Cu (ppm) 0.4 0.0 0.3 0.0 0.3 0.0 0.3 0.0 0.4 0.1 0 .3 0.0 0.4 0.0 (0.4) (0.2) (0.4) (0.3) (0.2) (0.3) (0.3) median concentrations with asterisk (*) indicate significant difference relative to the median forest soil value (Mann Whit ney U test statistic, P< 0.05).

PAGE 79

79 Table 4 5. Pearson's product m %OM g N/m2 g C/m2 C:N CEC Ca S B Fe Cu Al pH %OM 1.00 g N/m2 0.85* 1.00 g C/m2 0.86* 0.96* 1.00 C:N 0.44* 0.62* 0.41 1.00 CEC 0.82* 0.75* 0.71* 0.52* 1.00 Ca 0.73* 0.70* 0.65* 0.54* 0.93* 1.00 S 0.2 0.31 0.28 0.23 0.30 0.44* 1.00 B 0.63* 0.73* 0.66* 0.49* 0.69* 0.73* 0.35 1.00 Fe 0.43* 0.44* 0.43* 0.29 0.54* 0.65* 0.18 0.56 1.00 Cu 0.46* 0.48* 0.49* 0.26 0.57* 0.69* 0.33 0.67* 0.74* 1.00 Al 0.40 0.33 0.32 0.27 0.65* 0.64 0.15 0.35* 0.47* 0.35 1.00 pH 0.39 0.49* 0.47* 0.25 0.63* 0.75* 0.42 0.71* 0.68* 0.72* 0.42 1.00 P < 0.01 ( with post hoc bonferroni adjustment for multiple comparisons) only characteristics that had a significant correlation are shown. All others were insignificant at P < 0.01 and P < 0.05

PAGE 80

80 Table 4 13 C vs. PDB) isotopic concentrations (mean standard error) of soil organic matter for the seven primary land cover classes in the Temash Watershed. The median value is shown in parentheses*. Soil Dominant Vegetation Land Cover Category 13 C 13 C Milpa 27.71 0.11 12.84 ( 27.6 3) ( Z. mays ) SS1 27.79 0.10 28.27 ( 27.77) ( Heliconia sp.) SS2 27.67 0.12 32.16 ( 27.81) ( V. ferruginea ) SS3 28.03 0.12 30.29 ( 28.07) ( A. cohune ) Forest 27.99 0.11 30.29 ( 27.90) ( A. cohune ) Cacao Or chard 28.14 0.10 n.d. ( 28.09) ( T. cacao ) Pasture 25.28 0.36 13.13 ( 25.22) (B. humidicola ) median values with asterisk (*) are significantly different from forest soil (Mann Whitney U test statistic, P< 0.05)

PAGE 81

81 Figure 4 1. Map of study area.

PAGE 82

82 Figure 4 2. Plot of select soil characteristics that include at least one land cover class that is significantly different from forest soils (see Tables 2 and 3 for details; asterisks (*) denote soils that are significantly differe nt from forest soils, Mann Whitney U statistic, P < 0.05). (SS1 = 1 st succesional stage; SS2 = 2 nd successional stage; SS3 = 3 rd successional stage)

PAGE 83

83 Figure 4 3. Percentages of carbon derived from C3 forest and from C4 pasture. Figure 4 4. The Mea n Residence Time (MRT) of carbon in SOM of pasture

PAGE 84

84 CHAPTER 5 TEMPORAL AND SPATIAL VARIABILITY OF DISSOLVED NUTRIENTS IN A SMALL TROPICAL WATERSHED D OMINATED BY SHIFTING CULTIVATION Introduction Nutrient concentrations transported in rivers and streams ca n provide insight into interactions with, and characteristics of the adjacent terrestrial landscape. The transfer of nutrients across the terrestrial aquatic boundary and the flux of nutrients through stream and river networks depends on a combination of biotic and abiotic factors including catchment and riparian zone vegetation, hydrology, geology (slope and soils), and local/regional climatic patterns. Human land use and land cover conversion in particular conversion for agricultural purposes, is consi dered one of the primary drivers of rapid change in nutrient concentrations resulting in the impairment of freshwater and near shore coastal environments (Carpenter et al. 1998, Caraco and Cole 1999, Downing et al. 1999, Arbuckle and Downing 2001) I n humid tropical watersheds, h uman induced change to nutrient export occurs in a context that, in some cases already experiences nutrient export rates higher than those of temperate watersheds with similar runoff rates (Downing et al. 1999, Lewi s et al. 1999) Humid tropical forest ecosystems have high rates of nitrogen (N) fixation and are generally not N limited (Vitousek and Sanford 1986, Lewis et al. 1995) leading to weak N retention in tropical catchments. However, many aquatic ecosystems in the tropics are considered N limited because organic N, rather than p lant available N forms, constitutes a large proportion of the total nitrogen export (Downing et al. 1999, Neill et al. 2001) Phosphorus (P) concentrations in tropical stream networks are dependent on terrestrial sources and reflect the weathering of local catchment geology (Meybeck 1982) and are also positively correlated with soil nutrient concentrations (Biggs et al.

PAGE 85

85 2002) In stream P can also be strongly influenced by interactions with groundwater and geothermal springs (Pringle and Triska 1991) Understanding large tropical river responses to catchment level changes is difficult because of heterogeneity of disturbance at larger spatial scales (McDowell and Asbury 1994) It is generally considered that small catchments are more sensitive to changes in land use, in part because of their small surface area to channel length ratio (Thomas et al. 2004) and because they display higher rates of nutrient transformation relative to larger rivers (Peterson et al. 2001) L and cover conversion from forest to agricultural land in small tropical watersheds result s in increasing stream discharge because of reduced evapotranspirat ion (ET) and increased overland flow (Bruijnzeel 2004, Moraes et al. 2006) However, identifying changes in dissolved in stream nutrient concentrations as a consequ ence of land cover conversion in small tropical watersheds is understudied and consequently not well understood (Biggs et al. 2004) Future proj ections of land use and land cover changes associated with population growth suggest that disproportionately large impact s will be observed in tropical freshwater ecosystems (Sala et al. 2000, Dudgeon et al. 2006, McClain 2008) Although it has often been argued that small scale farmers are the most dominant agents in tropical forest clearing (Myers 1991, Houghton 1994) large scale global market processes are considered the ultimate drivers of land use and land cover change in the tropics (Lambin et al. 2001, Geist and Lambin 2002, Filoso et al. 2006) However, m ore than 250 million people in the tropics p ractice slash and burn agriculture (Attiwill 1994) a land use strategy that includes clearing forest patches, burning the fallen biomass, p lanting and cultivating corn and other food crops, followed by fallowing

PAGE 86

86 before the area is re cleared and cultivated again I n the Maya Forest of northern Central America, slash and burn, locally referred to as milpa most pervasive land use strategy for millennia (Turner et al. 2001) Although an extensive literature exists on int eractions between milpa and wildlife (Gomez Montes and Bayly 2010) s oil nutrients (Diekmann et al. 2007) and forest structure (Dckersmith et al. 1999, Lawrence and Foster 2002) there h ave been no studies relating milpa agriculture and contemporary aquatic ecosystems in northern Central America 1 and there is only one study investiga ting slash and burn agriculture and its impact on stream nutrients from Amazonia (Williams and Melack 1997 ; Williams et al. 1997). In this study, I investigate spatial and temporal differences of in stream nutrient concentrations in four watersheds of varying sizes where the dominant land use is milpa agriculture. The objectives for this study are to (i) char acterize discharge patterns in relation to precipitation and watershed characteristics for each study catchment, (ii) examine differences in seasonal and longitudinal in stream nutrient concentrations and nutrient fluxes within each study catchment, (iii) determine relationships between seasonal and annualized flow weighted mean nutrient concentrations and watershed characteristics, and (iv) compare nutrient export coefficients of the study watersheds of the Tema sh River to other tropical and temperate wate rsheds. 1 An extensive literature exists on the use of sediment records from lakes and fluvial geomorphology in the Maya Forest region to better understand ancient human environment interactions. See (Brenner et al. 2002) for a review.

PAGE 87

87 Methods Study Site The headwaters of the binational Temash River begin in Guatemala and flow eastward across the southern Toledo District of Belize towards the Caribbean Sea and Gulf of Honduras (Fig ure 4.1 ). Approximately 400 km 2 of the watershed km 2 are located in Belize. The Temash River is one of 16 major watersheds of Belize and is one of only two of these major watersheds to have never been instrumented with a stream gauge station or other hydrologic monitoring equipment (Lee et al. 1995) Soils are derived largely from Cretaceous limestone with patches of dolomite and siliceous limestone (Wright et al. 1959). Vegeta tion cover in the Temash River w atershed is dominated by tropi cal evergreen lowland forests (Meerman et al. 2003b) In addition, the lower reaches of the watershed have large stands of red mangrove ( Rhizophora mangle ) as well as the only documented Sphagnum bogs in lowland Central America (Meerman et al. 2003b) The lower reache s of the Temash River are within the Sarstoon Temash National Park, a 41,000 acre community co managed park. The Temash watershed Maya villages. The dominant land use milpa agriculture, though pasture and cacao ( Theobroma cacao ) agroforestry are expanding in the region (Levasseur and Olivier 2000, Emch 2003) Watershed Selection, Delineation, and Characterization The four study watersheds are 2 nd order streams. Yax Cal creek (YXL) and Crique Sarco creek (CRS) drain into the middle reaches of the Temash River. Sunday Wood creek (SWD) and Conejo creek (CON) drain into the lower reaches of the Temash River and a large R. mangle estuary. Upstream sampling sites (e.g., YXL01, CRS01, SWD01, CON01) were selected during the dry season and as such represent the

PAGE 88

88 farthest upstream reach that retained water during the dry season. Downstream sites in YXL and CRS (YXL05 and CRS0 5) are close to the confluence with the Temash River. Downstream sites in SWD and CON (e.g., SWD04 and CON04) were selected to includ e the nearest village within the catchment area s A geographic information system (ArcView and ArcInfo, ESRI, Redlands, CA) was used to delineate watershed boundaries and the catchment areas for each sampling point. Land cover percentages within each watershed were derived from the 1:250,000 Belize Ecosystems Map (Meerman and Sabido 2005). The distribution s of soil types within each watershed were derived from a digitized version of the 1:250,000 map of the soils of Belize (Wright et al. 1959) Annual Rainfall Pattern R ain gauges (Tru Chek graduated 1 150 mm) were installed i n three communities within the watershed (Crique Sarco, Lucky Strike, and Conejo) beginning in February 2007 (Figure 5.1) Rain gauges were positioned in the open, at least 5 meters from any structure. Total rainfall was mon itored daily and recorded. The transition between the dry and rainy season s in the Temash Watershed was determined using methods described in Marengo et al. (2001). A five day running mean was calculated for the total rainfall record from each station. The onset (and end ) of the rainy season must satisfy three criteria : (1) mean daily precipitation is more (less) than 4.0 mm; (2) six out of eight of the preceding (subsequent) days must have mean precipitation less (more ) than 3.5 mm; (3) six out of eight subsequent (preceding) days must have mean precipitation more (less) than 4.5 mm (Marengo et al. 2001)

PAGE 89

89 Sample Collection and Laboratory Analysis Water samples were collected using a modified synoptic sampling design (Grayson et al. 1997) to achieve both spatial coverage and longitudinal sampling within each of the study water sheds. Samples were collected twice per month, except during October, between April 2007 and June 2008. Longitudinal samples within each catchment were collected on the same day and individual catchments were sampled within 1 3 days of one another. Wate r samples for dissolved nutrients were filtered in the field through 0.7 mL HDPE bottles. Samples for dissolved inorganic N (DIN) analyses were 2 SO 4 to bring sample pH to <2. Samples for total soluble N (TSN) and P (TSP) and soluble reactive phosphorus (SRP) were not treated with acid. Whole (unfiltered) water samples for total nitrogen (TN) and total phosphorus (TP) were collected during three dry season months and four rainy season months. Samples were placed on ice in the field and stored frozen (<3 month s) until analysis. Concentrations of dissolved and total N and P were analyzed on a Technicon Autoanalyzer II with a single channel colorimeter using standard colorimetric techniques. Dissolved inorganic nitrogen (NO 3 + N O 2 ) was measured by cadmium re duction and SRP concentrations were analyzed using the ascorbic acid ammonium molybdate method. Total dissolved (and whole water) P was measured on filtered (and unfiltered) samples by ascorbic acid ammonium molybdate colorimetry following acidic persulfa te digestion. Total dissolved and whole water N was measured by cadmium reduction, a fter basic persulfate digestion on filtered and unfiltered samples respectively (Davis and Simmons 1979, APHA 1989)

PAGE 90

90 Statistical Analysis Seasonal differences (dry vs. rainy) in discharge were determined using the non parametric Mann Whitney U statistic (p and antecedent rainfall (1 day, 7 day, 14 day, and 28 day prior) were determined using ea ch study watershed were determined using linear regression. Seasonal differences (dry vs. rainy) in nutrient concentrations at the upstream and downstream sampling sites were assessed using t tests (p when the nutrient concentra tions were not normally distributed, the non parametric Mann Whitney U statistic (p rences between seasons. Within stream differences between the upstream and downstream sampling sites was also assessed for the dry season and rainy season using the same test parameters. The calculation of flow weighted mean concentrations (FWMC) does not provide an error term. However, the variance of a weighted average can be est imated using Equation 5 1 : (5 1) The standard deviation of x ( ) is derived from the in stream nutrient data and w is the weighting factor, in this case the flow (Q). This est imate of variance for the weighted mean nutrient concentrations is used to calculate a standard deviation of the FWMC ( for statistical comparisons of FWMC from the four study watersheds.

PAGE 91

91 Discharge, Nutrient Fluxes and Flow Weighted Me an Nutrient Concentrations Instantaneous discharge ( m 3 s 1 ) at each sampling site was estimated during each sampling event using the float method (Gore 2006) Mean surface flow velocity (m s 1 ) was measured using a surface float and adjusted (multi plied by 0.85) to estimate mean velocity Between two and four floats were used to estimate mean velocity. The mea n velocity was then multiplied by the cross sectional area of the stream channel (m 2 ) at the time of sampling to calculate instantaneous discharge rate (m 3 s 1 ) (Gore 2006) Instantaneous nutrient fluxes were calculated for each sampling event by multiplying the instantaneous discharge ( m 3 s 1 ) by the nutrient concentration (mg L 1 = g m 3 ) to obtain g s 1 and ultimately expressed as load ( kg day 1 ). Instantaneous discharge rates and instantaneous nutrient fluxes were assumed to represent the average discharge and flux for each sampling interval (~15 days or 30 days in months when only a single sample was colle cted). Annual nutrient fluxes (kg ha 1 yr 1 ) from each study watershed were then calculated by multip lying instantaneous fluxes (kg day 1 ) by the number of days in each sampling interval, summing the resulting nutrient load (kg) over the year, and dividin g by watershed area (ha). Seasonal (dry and wet) and annual flow weighted mean nutrient concentrations (mg/L), calculated by dividing the sum of the seasonal (or annual) nutrient loads by the total seasonal (or annual) discharge rate, were used to examine water quality patterns independent of discharge. Results Precipitation, Watershed Characteristics, and Seasonal Discharges The rainy season began on June 4 and the dry season start ed on December 19 in 2007 (Figure 5 2) The 2007 2008 dry season lasted un til May 20, 2008. Total annual precipitation (rainy season 2007 to onset of rainy season 2008 ) at each of the three

PAGE 92

92 stations was >3600 mm, and > 80 % of the total annual rainfall fell during the rainy season. Although several large rainfall events occurred during Dec 2007 Feb 2008, these events were preceded and followed by extended periods of rain free days (Figure 5 2 ) rainfall criteria. At the Conejo rain gauge station, t here was a brief period from November 4 18, 2007 that satisfied the criteria for onset of the dry season. T his period however, was preceded and followed by a period of 18 days when the 5 day moving average precipitation did not fall below 4 mm, and con sequently was not used to mark the onset of the dry season. The four study watersheds differ in size, geology, soils, and to a degree, vegetation cover (Table 5 1). Crique Sarco creek (CRS) has a mean stream channel slope of 6.6% and the basement geology is predominantly clastic material. Soils are young (cambisols) and shallow (leptosols). Yax Cal creek (YXL) is the smallest of the 4 study catchments and has a mean stream channel slope of 2.9%. The bedrock of YXL is entirely clastic sedimentary rock and soils are almost entirely young cambisols. Sunday Wood creek (SWD) is the largest of the four study catchments and has a mean channel slope of 5.4% and its bedrock includes alluvial, clastic, and limestone deposits. The soils are predominantly fluvial i n origin. Conejo creek (CON) has a mean channel slope of 2.9%. Conejo creek bedrock is predominantly clastic sedimentary rock with some alluvial deposits. Soils are young and of fluvial origin. Land cover in the four watersheds is dominated by agricult ural lands. Yax Cal and Crique Sarco creeks have the highest overall percent land cover as agriculture (77% and 75%, respectively). Agricultural lands in Sunday Wood and Conejo creeks compromise approximately 60%

PAGE 93

93 of the total land cover (Table 5 1). The only other dominant land cover in the study catchments includes combined semi deciduous and evergreen tropical forest. Discharge varied across watersheds and between sites within watersheds. Mean dry season instantaneous discharge from each watershed ran ged from 0.015 0.04 m 3 s 1 (YXL) to 0.31 0.24 m 3 s 1 (CRS). Wet season instantaneous discharge was greatest in CRS (0.66 0.83 m 3 s 1 ). Dry season and wet season discharge rates were significantly different across all sampling sites with the excepti on of CRS05, YXL01 and YXL05 (Table 5 2). Absolute discharge rates generally increased from upstream sites to downstream sites within each watershed (Table 5 2, Figure 5 3). Mean instantaneous discharge rates were correlated with antecedent rainfall, al though the strength of the correlation with different periods of antecedent rainfall (e.g., 1 day, 7 days, 14 days, and 28 days prior) varied across each watershed. Discharge rates in CRS had a strong correlation (r=0.73) with 1 day prior rainfall. The c orrelation between discharge rate and 1 day prior rainfall was strongest relative to other correlation coefficients for SWD, but the correlation was weak (r=0.52). YXL discharge was strongly correlated with 14 day prior rainfall (r=0.68) while CON dischar ge was correlated with 28 day prior rainfall (r=0.62) (Table 5 3). Mean seasonal discharge rates were also significantly related to watershed area (Figure 5 4). To remove the effect of watershed area, discharge rates were normalized by watershed area (Fi gure 5 3). When normalized for watershed area, the general trend towards increasing discharge rates downstream was maintained with the exception of the most upstream sites in YXL and SWD (Figure 5 3).

PAGE 94

94 Seasonal and Longitudinal Variation of In stream N utrient Concentrations Mean nutrient concentrations for each sampling site within each study watershed are shown in Table 5 4 (N species) and Table 5 5 (P species). In the upper reaches of CRS (CRS01), TN concentrations during the dry season were higher than in the wet season (U=0.0, P=0.024). Seasonal d ifferences in DIN, TSN, SRP, TSP, and TP were not statistically significant (P>0.05) (Table 5 4, 5 5). In the lower reaches of CRS (CRS05), TSN concentrations were significantly higher during the dry sea son (U=18.5, P=0.020). During the dry season, n utrient concentrations at upstream and downst ream sites were not statistically different in CRS. During the rainy season, SRP and TSP concentrations at CRS05 were greater than at CRS01 (t = 2.824, P=0.010 ( SRP); U=16.5, P=0.005(TSP)). No other differences between upstream and downstream wet season nutrient concentrations in CRS were observed. In YXL, no seasonal differences in N species were observed at the upstream sampling site (YXL01). However, dry s eason concentrations of SRP (t = 4.969, P<0.001), TSP (t = 3.669, P=0.001) and TP (t = 4.687, P<0.001) were all higher than wet season concentrations at YXL01. These seasonal differences were not observed at the downstream site with only dry season TSN c oncentrations being higher than wet season TSN concentrations at YXL05 (U=28, P=0.038). During the dry season, DIN and TSN concentrations were greater at YXL05 than at YXL01 (U=37, P=0.046; U=29, P=0.014, respectively) while the opposite occurred with P n utrients. Dry season concentrations at YXL01 of SRP (U=12, P<0.001), TSP (U=12, P<0.001) and TP (U=7, P=0.026) were all greater than at YXL05. No differences in nutrient concentrations between the upstream and downstream site were observed during the rai ny season.

PAGE 95

95 Nutrient concentrations in SWD varied little between the seasons. At SWD01, dry season TN was greater than wet season TN (U=1.0, P=0.048). No other seasonal differences in nutrient concentrations were observed at the upstream or downstream sit es. In addition, no longitudinal differences in nutrient concentrations were observed during the wet or dry season (Table 5 4, 5 5). In CON, dry season TN concentrations were greater than wet season concentrations at the downstream site (U=0.0, P=0.017). Differences between upstream and downstream nutrient concentrations were only observed during the wet season with TP being greater at the downstream site (t = 3.50, P=0.025). No other longitudinal differences in nutrient concentrations were observed (T able 5 4, 5 5). The contribution of DIN to the TSN pool was highly variable across seasons and across study watersheds. DIN comprised approximately 58% of both the dry season and rainy season TSN pool at both upstream and downstream sites. TSN comprised approximately 60% of the TN pool on average during the dry season and contributed more than 85% during the wet season. In YXL the average contribution of DIN to the TSN pool was between 64% and 88%. The TN pool in YXL averaged more than 60% TSN during th e dry season and during the rainy season included more than 90% TSN. In SWD, DIN comprised approximate 50% of the TSN pool during the dry season and more than 75% during the rainy season, with a similar pattern of DIN:TSN observed in CON. During the rain y season in both SWD and CON, DIN contributed more than 70% to the TSN pool on average. TSN comprised approximately 90% of the TN pool during the rainy season in SWD and averaged between 73% and 100% in CON.

PAGE 96

96 In CRS, SRP comprised between 17% and 46% of TS P, on average, during the dry season. During the wet season, SRP averaged between 30% and 44% of TSP in CRS. TSP contributed a larger portion to the TP pool in CRS, with SRP:TSP averaging between 50% and 90%. During the dry season in YXL, SRP contribute d more than 85% to the TSP pool at the upstream site (YXL01). Lower ratios of SRP:TSP were observed downstream. Dry season average contributions of TSP to the TN pool were similar to SRP:TSP while rainy season TSP comprised more than 90% of the TP pool. In SWD, SRP averaged between 30% and 53% of the TSP pool, regardless of season. The TSP contribution to the TP pool averaged between 38% and 63% during the dry season and between 75% and 97% during the rainy season. Dry season and wet season SRP:TSP ra tios were similar in CON, averaging between 32% and 55%. During the dry season, TSP contributions to the TP pool averaged between 30% and 53% while the contribution averaged between 58% and 76% during the rainy season. Ratios of TN:TP at upstream and dow nstream sites within each watershed are shown in Table 5 6. Dry season TN:TP values are generally higher than the wet season value at the same site although only the ratios at the downstream site in CON (CON04) and the downstream site in SWD (SWD04) were significantly different (t=2.946, P=0.022 and t=2.484, P=0.048, respectively). Overall, TN:TP ratios suggest waters are P limited across all sites though large standard deviations during the rainy season in YXL, SWD, and CON reflect reduced N concentrati ons during periods of high flow (see also Table 5 4) Nutrient Fluxes and Flow Weighted Mean Nutrient Concentrations Nutrient fluxes (kg/day) varied over time and across catchments within the Temash River (Figures 5 5 and 5 6). In all catchments, daily l oads were low at the end

PAGE 97

97 of the 2007 dry season. An increase, or pulse in daily N flux was observed at the onset of the 2007 rainy season, although its timing and duration varied across catchments. In CRS, the first increase in N flux, from <1 kg/day to almost 10 kg/day, was observed between the June and July sampling period (Figure 5 5). A second, much larger pulse of N occurred in CRS later in the rainy season, at the end of August. Subsequent N fluxes remained low throughout the rainy season in CRS ( Figure 5 5). In YXL, a pulse of N was observed during June and persisted through the July sampling period. In SWD, N fluxes were low throughout the 2007 sampling period (Figure 5 5). In Conejo creek, the initial pulse of N was observed in June 2007, dec lined in July 2007, and then remained low throughout the year. The final sampling event (June 2008) occurred within 3 days of the onset of the 2008 rainy season and all study catchments showed a large change in nutrient flux at that time. Daily phosphorus loads were low across all study watersheds, rarely exceeding 0.5 kg/day (Figure 5 6). A small increase was observed following the onset of the rainy season (late June 2007) in CRS, YXL, and CON. A large pulse of TP, TSP, and SRP was observed in CRS in l ate August, with a TP flux of ~2.0 kg/day. A series of sampling events in CRS after the onset of the dry season (January 2008) captured additional fluxes of P. P fluxes remained low through the dry season with the onset of the rainy season in June 2008 s howing a small increase in P flux occurring at the end of the sampling (Figure 5 6) Flow weighted mean (FWM) concentrations for dissolved N (inorganic and total) varied little seasonally within catchments or between catchments (Figure 5 7). In CRS, no s tatistically significant difference between wet and dry FWM concentrations of DIN, TSN, or TN were detected. In YXL, dry season FWM TSN

PAGE 98

98 concentrations were higher than wet season FWM concentrations (t = 2.154, P=0.049). Dry season FWM TN was also higher than wet season FWM TN in YXL (t = 3.889, P=0.012). In SWD, FWM TN was also higher in the dry season than the wet season (t = 5.516, P=0.005). There were no seasonal differences in FWM concentrations of N species in CON (Figure 5 7). Flow weighted me an concentrations of P across all sampling sites were low and there were no significant differences between season P concentrations (Figure 5 8). Concentrations were generally below 0.001 mg L 1 in all catchments and samples included high variability, evi dent in the large standard deviations for each estimated FWMC (Figure 5 8). Between catchment differences in the annual FWMC were observed for some nutrients. The FWM DIN in CON was greater than the FWM DIN in YXL (t = 2.070, P=0.046) and SWD (t = 2.690 P=0.011). The FWM TSN in CON was also greater than the FWM TSN in SWD (t = 3.034, P=0.005). In YXL, the FWM TSN was greater than the FWM TSN in SWD as well (t=2.298, P=0.028). No other differences in total annual FWM nutrient concentrations were obse rved between catchments (Figures 5 7 and 5 8). Relationships between annual flow weighted mean concentrations and land cover and soil classes were determined by linear regression (Table 5 7, 5 8). Flow weighted means from each of the 19 sample sites were regressed against percent land cover and percent soil class cover. Outliers were determined using the studentized residual and concentrations were positively related to % silty clay alluvium (soil class 28) although the coefficient of determination was weak (r 2 = 0.28). FWM TSN was also positively

PAGE 99

99 related to soil class 28, although again the relation was weak (r 2 = 0.28). When outliers were removed, TSN was negatively related to soil class 49 (sandy alluvium). Flow weighted mean concentrations of TN were positively related to % forest cover and negatively related to % agriculture. The strongest predictors of TN were % coastal loamy sand (soil class 24; r 2 = 0.56) and % sandy loam (soil class 49; r 2 = 0.47). Land cover and catchment soils were generally weak predictors of FWM phosphorus concentrations (Table 5 8). SRP TSP, and TP were positively related to % forest cover and negatively related to % agriculture in catc hments, although the coefficients of determination (r 2 ) were weak. TSP was also positively related to the % clay soils in the study watersheds. (Table 5 8). Discussion After forests are cleared for milpa, decaying organic matter decomposes and remineraliz ed nitrogen is able to enter the soil (Williams et al. 1997) The capacity of the soil to absorb this nitrogen is reduced because of the felling of trees and the dying of root systems. Consequently, leached nitrate accumulates and moves downslope At the onset of the rainy season, this soil nitrate can be released as a pulse into adjacent waterbodies (Downing et al. 1999) The capacity of the riparian zone to minimize the amount of nitrate entering adjacent waterbodies via denitrification depends on the extent of land clear ing and th e presence of organic rich soils (McClain et al. 1994) The burning of slash also le ads to a pulse of dissolved N able to reach adjacent waterbodies via subsurface flow. Although much of the available N in slash is released into the atmosphere during the burning, a portion of it will settle on the landscape. As the cycle of slash and b urn continues, areas repeatedly used for slash and burn will

PAGE 100

100 have reduced export of dissolved nitrogen because of a gradual decrease in N stocks (Downing et al. 1999) In the four study watersheds, this initial pulse of N was observed at the onset of the rainy season in three of the four study watersheds and was dominated by dissolved N. Following this initial pulse, daily loads of N remained low throughout the rainy season (Figure 5 5). Concentrations also remain low well into the dry season. The dry season is a period of peak leaf litter fall in much of the tropics and leaf litter decomposition can contribute to N immob ilization in both terrestrial and aquatic ecosystems (McDowell and Asbury 1994) These low N concentrations during the dry season differ from watersheds dominated by large scale agriculture in that base flow conditions can often be associat ed with increases in dissolved N because of the increased contribution of groundwater to stream hydrology (Kemp and Dodds 2002) The nitrogen/phosphorus ratios indicate the study watersheds are all P limited. Phosphorus concentrations at all sites (except YXL01) were measured at or near the detection limit of the laboratory instrumentation. There is a weak correlation be tween TSP and clay soils in the study area. It is possible that any available P is derived from clay soils where P is adsorbed to metal oxides (Pacini et al. 2008) In YXL01 and CON01 high er SRP concentrations were observed during the dry season (Table 5 5). It is possible that P is released from the mineralization of accumulated organic matter during the dry season (Saunders et al. 2006) The four study watersheds exhibited different patterns of N and P export (Table 5 9). CRS and CON exported the largest volume of N, relative to the other study watersheds. In CRS, DIN comprised approximately 46% of the TSN exported while in

PAGE 101

101 CON, DIN accounted for more than 80% of the dissolved N exported from the watershed (Table 5 9). YXL and SWD both had l ow TSN export coefficients (0.53 and 0.18, respectively). Dissolved inorganic N comprised approximately 50% of the TSN exported from YXL while DIN accounted for ~65% of the TSN exported from SWD (Table 5 9). Phosphorus export was low across all study wat ersheds (Table 5 10). Total soluble P export was lowest from SWD and highest from CRS. Soluble reactive P contributed between 33% and 75% of the TSP exported. When compared to other tropical watersheds, the export coefficients presented here are lower. This is most notable in YXL and SWD where export coefficients for DIN and TSN are at least one order of magnitude less than other tropical systems (Table 5 9, 5 10). Overall, nutrient export from the study watersheds is likely underestimated because of t he method used (Table 5 9). Daily and monthly loads of nutrients were derived from only one or two samples and assumed to be representative of the sampling interval. Discharge in these watersheds is correlated with rainfall (Table 5 3) and sampling was n ot conducted to develop a stronger estimate of export during storm events. It is likely that a large percentage of the annual export occurs during a few single events at the onset of the rainy season and finer sampling intervals would capture some of this export (McDowell and Asbury 1994)

PAGE 102

102 Table 5 1. Catchment characteristics of the four study watersheds and the Temash River watershed (all data are for the catchment area of the farthest downstream sampling site). Crique Sarco (CRS 05 ) Yax Cal (YXL 05) Sunday Wood (SWD 04) Conejo (CON 04) area (ha) 1084 469 3204 1384 mean elevation (m) 53 26 41 28 mean slope (%) 6.6 2.9 5.4 2.9 Geology (%) Alluvial 0 0 60 27 Limestone 23 0 15 0 Clastic Se dimentary 77 100 25 73 Soils (%) Cambisol 79 95 30 52 Fluvisol 0 5 61 48 Leptosol 11 0 9 0 Leptosol vertisol 10 0 0 0 Land Cover (%) Semi decid./evergreen Forest 23 23 40 36 Lowland Swamp Forest 0 0 0 5 A gricultural Lands 75 77 60 59 Waterbody 2 0 0 0

PAGE 103

103 Table 5 2. Comparison of dry and wet season discharge rates (m 3 s 1 ) across all sampling sites (Mann Whitney U statistic, P<0.05) Dry Season Discharge (m 3 sec 1 ) Wet Season Discha rge (m 3 sec 1 ) Sampling Site median 25% 75% median 25% 75% U statistic P value CRS01 0.047 0.017 0.248 0.353 0.088 0.642 15.0 0.037 CRS02 0.060 0.000 0.263 0.446 0.167 0.970 10.0 0.017 CRS03 0.123 0.000 0.234 0.680 0.257 1.541 10.0 0.01 7 CRS04 0.181 0.047 0.257 0.383 0.171 0.859 14.0 0.045 CRS05 0.288 0.116 0.539 0.365 0.142 0.926 23.5 0.613 YXL01 0.095 0.036 0.199 0.210 0.155 0.291 15.0 0.079 YXL02 0.037 0.019 0.088 0.148 0.056 0.246 16.0 0.046 YXL03 0.053 0.033 0.113 0.236 0.089 0 .403 8.0 0.007 YXL04 0.088 0.035 0.133 0.382 0.137 0.716 8.0 0.007 YXL05 0.000 0.000 0.000 4.2 x 10 5 0.000 0.222 17.0 0.232 SWD01 0.000 0.000 0.000 0.031 0.000 0.080 13.5 0.029 SWD02 0.000 0.000 0.000 0.135 0.010 0.264 6.0 0.013 SWD03 0.019 0.011 0.08 3 0.250 0.100 0.299 7.0 0.006 SWD04 0.031 0.006 0.043 0.293 0.169 0.316 2.0 0.002 CON01 0.000 0.000 0.006 0.059 0.017 0.155 9.5 0.005 CON02 0.009 0.000 0.025 0.124 0.057 0.280 7.0 0.006 CON03 0.000 0.000 0.000 0.260 0.129 0.406 5.0 0.002 CON04 0.000 0 .000 0.000 0.380 0.076 0.844 4.0 0.001

PAGE 104

104 discharge and antecedent rainfall Antecedent Rainfall Catchment 1 day 7 day 14 day 28 day CRS* 0.73 0.53 0.41 0.39 YXL* 0.48 0. 63 0.68 0.33 SWD^ 0.52 0.34 0.30 0.26 CON^ 0.47 0.61 0.59 0.62 all correlations are significant at P < 0.01 *CRS and YXL correlated with the CRS rain gauge station ^SWD and CON correlated with the CON rain gauge station

PAGE 105

105 Table 5 4. Seasonal mean nutrient concentrations for DIN, TSN, and TN for all sampling sites within each study watershed. Values are mean 1 standard deviation (sample size in parenthesis). DIN (mg/L) TSN (mg/L) TN (mg/L) dry wet dry wet dry wet Crique Sarco CR S01 0.276 0.442 (10) 0.142 0.269 (13) 0.391 0.401 (10) 0.207 0.267 (13) 0.586 0.298 (6) 0.097 0.063 (3) CRS02 0.291 0.443 (11) 0.159 0.342 (11) 0.351 0.416 (11) 0.219 0.330 (11) 0.496 0.377 (6) 0.106 0.065 (2) CRS03 0.292 0.45 2 (11) 0.178 0.360 (11) 0.355 0.421 (11) 0.218 0.350 (11) 0.513 0.364 (6) 0.108 0.079 (3) CRS04 0.270 00.407 (12) 0.169 0.355 (11) 0.406 0.386 (11) 0.204 0.346 (11) 0.517 0.323 (7) 0.093 0.051 (3) CRS05 0.329 0.434 (11) 0.218 0.457 (9) 0.507 0.374 (11) 0.275 0.437 (9) 0.602 0.275 (6) 0.124 0.038 (2) Yax Cal YXL01 0.154 0.259 (12) 0.214 0.511 (12) 0.179 0.252 (12) 0.225 0.508 (12) 0.378 0.401 (7) 0.058 0.026 (3) YXL02 0.227 0.374 (12) 0.204 0.3 09 (12) 0.255 0.360 (12) 0.227 0.305 (12) 0.405 0.408 (7) 0.040 0.027 (3) YXL03 0.216 0.428 (12) 0.205 0.293 (12) 0.314 0.408 (12) 0.220 0.296 (12) 0.446 0.371 (7) 0.047 0.040 (3) YXL04 0.278 0.483 (12) 0.191 0.291 (12) 0.334 0.459 (12) 0.222 0.288 (12) 0.432 0.401 (7) 0.047 0.023 (3) YXL05 0.272 0.399 (12) 0.170 0.315 (10) 0.357 0.366 (12) 0.201 0.306 (10) 0.491 0.342 (7) 0.082 0.055 (3) Sunday Wood SWD01 0.120 0.204 (11) 0.161 0.401 (10) 0.143 0.187 (11) 0.183 0.394 (10) 0.424 0.405 (6) 0.027 0.009 (3) SWD02 0.118 0.190 (10) 0.163 0.369 (11) 0.240 .0230 (10) 0.177 0.364 (11) 0.431 0.395 (6) 0.064 0.052 (3) SWD03 0.121 0.198 (11) 0.219 0.480 (11) 0.308 0.163 (10) 0. 246 0.472 (11) 0.522 0.300 (7) 0.057 0.037 (3) SWD04 0.143 0.219 (11) 0.286 0.630 (9) 0.213 0.192 (11) 0.306 0.623 (9) 0.494 0.341 (6) 0.063 0.039 (2) Conejo CON01 0.178 0.302 (11) 0.381 0.723 (11) 0.353 .0477 (11) 0.402 0.714 (11) 0.556 0.369 (6) 0.071 0.033 (3) CON02 0.123 0.191 (11) 0.932 0.764 (10) 0.216 0.207 (10) 0.432 0.760 (10) 0.435 0.348 (5) 0.087 0.050 (3) CON03 0.0117 0.0171 (11) 0.313 0.465 (10) 0.226 0.181 (11) 0.348 0.482 (10) 0 .556 0.328 (6) 0.080 0.020 (3) CON04 0.128 0.176 (11) 0.303 0.450 (10) 0.280 0.167 (11) 0.338 0453 (10) 0.636 0.258 (6) 0.086 0.028 (3)

PAGE 106

10 6 Table 5 5. Seasonal mean nutrient concentrations for SRP, TSP, and TP for all sampling sites with in each study watershed. Values are mean 1 standard deviation (sample size in parenthesis). SRP (mg/L) TSP (mg/L) TP (mg/L) dry wet dry wet dry wet Crique Sarco CRS01 0.001 0.002 (9) 0.001 0.001 (13) 0.004 0.005 (10) 0.003 0.0 03 (13) 0.009 0.004 (6) 0.006 0.003 (3) CRS02 0.002 0.002 (10) 0.002 0.003 (11) 0.004 0.003 (11) 0.005 0.003 (11) 0.009 0.003 (6) 0.010 0.003 (2) CRS03 0.001 0.001 (10) 0.001 0.001 (11) 0.003 0.002 (11) 0.005 0.005 (11) 0.007 0.003 (3) 0.007 0.003 (3) CRS04 0.004 0.004 (10) 0.005 0.008 (11) 0.006 0.004 (11) 0.009 0.007 (11) 0.009 0.004 (7) 0.015 0.012 (3) CRS05 0.005 0.006 (10) 0.002 0.001 (9) 0.010 0.010 (11) 0.006 0.003 (9) 0.023 0.017 (6) 0.008 0.001 (2) Yax Cal YXL01 0.019 0.008 (12) 0.006 0.006 (12) 0.022 0.008 (12) 0.010 0.007 (12) 0.029 0.010 (7) 0.011 0.010 (3) YXL02 0.004 0.002 (11) 0.004 0.005 (12) 0.006 0.003 (12) 0.006 0.004 (12) 0.011 0.004 (7) 0.010 0.008 (3) YXL03 0.002 0.002 (11) 0.001 0.001 (12) 0.006 0.005 (12) 0.004 0.002 (12) 0.011 0.005 (7) 0.004 0.002 (3) YXL04 0.001 0.001 (11) 0.001 0.001 (12) 0.004 0.003 (12) 0.004 0.001 (12) 0.008 0.003 (7) 0.005 0.002 (3) YXL0 5 0.005 0.014 (11) 0.005 0.014 (10) 0.007 0.013 (12) 0.011 0.015 (10) 0.011 0.017 (7) 0.018 0.023 (3) Sunday Wood SWD01 0.002 0.003 (10) 0.002 0.002 (10) 0.004 0.005 (10) 0.005 0.003 (10) 0.009 0.006 (6) 0.007 0.003 (3) SWD02 0.011 0.022 (9) 0.002 0.001 (10) 0.012 0.025 (10) 0.003 0.002 (11) 0.009 0.012 (6) 0.006 0.002 (3) SWD03 0.007 0.008 (9) 0.002 0.002 (11) 0.014 0.018 (10) 0.007 0.004 (11) 0.020 0.020 (7) 0.008 0.003 (3) SWD04 0.001 0. 001 (10) 0.002 0.002 (9) 0.005 0.002 (11) 0.005 0.003 (9) 0.010 0.005 (6) 0.009 0.005 (2) Conejo CON01 0.010 0.019 (10) 0.004 0.009 (11) 0.012 0.020 (11) 0.005 0.008 (11) 0.017 0.021 (6) 0.005 0.001 (3) CON02 0.001 0.002 (9) 0.002 0.001 (10) 0.004 0.003 (10) 0.014 0.033 (10) 0.006 0.001 (5) 0.039 0.059 (3) CON03 0.010 0.020 (10) 0.003 0.002 (10) 0.013 0.018 (11) 0.006 (0.003) (10) 0.028 0.010 (6) 0.007 0.002 (3) CON04 0.002 0.003 (10) 0.002 0.0 02 (10) 0.005 0.004 (11) 0.004 0.001 (10) 0.017 0.009 (6) 0.007 0.001 (3)

PAGE 107

107 Table 5 6. Seasonal mean ( stdev) TN:TP ratios for upstream and downstream sampling sites within each study watershed. TN:TP Dry Season Rainy Season Crique Sar co upstream 167 115 31 8 downstream 92 85 33 4 Yax Cal upstream 33 26 18 17 downstream 248 174 19 14 Sunday Wood upstream 106 99 9 1 downstream 109 45 23 23 Conejo upstream 105 61 32 12 downstream 9 0 36 26 6

PAGE 108

108 Table 5 7. Linear regression parameters for annual flow weighted DIN, TSN, and TN concentrations with watershed characteristics of land cover and soil classification as independent variables. Dependent Coefficient of Independent Vari able Regressio n variable Land Cover Soil Class # mg L 1 r 2 p value Intercep t For. Agr. 15 24 28 49 Outliers 1 DIN 0.0068 0.75 0.1246 0.0004 CON03; CON04 2 DIN 0.051 0.39 0.1534 0.0007 CON03; CON 04 3 DIN 0.0008 0.91 0.1094 0.0001 CON03 4 DIN 0.0043 0.8 0.1 0.0001 CON03 5 DIN 0.28 0.03 0.0948 0.0018 CON03 6 DIN 0.21 0.075 0.115 0.0009 CON03; CON04 7 TSN 0.0073 0.74 0.174 0.0004 CON03; CON04 8 TSN 0.039 0.47 0.1979 0.0007 CON03; CON04 9 TSN 0 0 YXL01 10 TSN 0.12 0.19 0.1142 0.0006 CON03; CON04 11 TSN 0.26 0.03 0.1481 0.002 12 TSN 0.39 0.01 0.1705 0.0012 CON03; CON04 13 TN 0.37 0.01 0.0628 0.0036 YXL05 14 TN 0.35 0.016 0.3645 0.003 CRS05; YXL05 15 TN 0.14 0.156 0.123 0.0016 CRS05; YXL05 16 TN 0.56 <0.001 0.0184 0 .0021 YXL01; YXL05 17 TN 0.092 0.24 0.1675 0.0016 YXL05 18 TN 0.47 0.003 0.1958 0.0023 YXL05 soil class 15 = limestone/dolomite clay; class 24 = coastal loamy sand; class 28 = silty clay alluvium; class 49 = sandy clay alluvium (*) indicates significant linear relationship between nutrient and cland cover or soil class

PAGE 109

109 Table 5 8. Linear regression parameters for annual flow weighted SRP, TSP, and TP concentrations with watershed characteristics of land cover and soil classification as independent variables. Dependent Coefficient of Independent Variable Regressio n variable Land Cover Soil Class # mg L 1 r 2 p value Intercep t For. Agr. 15 24 28 49 Outliers 1* SRP 0.32 0.017 0.0008 0.0001 YXL05 2* SRP 0.29 0.025 0.0069 0.0001 YXL05 3 SRP 0.19 0.08 0.002 0 YXL05 4 SRP 0.005 0.78 0.0022 0 YXL05 5 SRP 0.003 0.85 0.0025 0 YXL05 6 SRP 0.12 0.18 0.0029 0 YXL05 7* TSP 0.3 0.027 0.0026 0.0001 YXL05; CON02 8 TSP 0.24 0.054 0.0115 0.001 YXL05; CON02 9* TSP 0.26 0.043 0.0042 0.0001 YXL05; CON02 10 TSP 0.005 0.79 0.0047 0 YXL05; CON02 11 TSP 0.06 0.36 0.0054 0 YXL05; CON02 12 TSP 0.1 0.23 0.0057 0 YXL05; CON02 13* TP 0.28 0.028 0.0043 0.0002 CON02 14* TP 0.238 0.047 0.0226 0.0002 CON02 15 TP 0.003 0.83 0.0086 0 YXL01; CON02 16 TP 0.038 0.46 0.0071 0 CON02 17 TP 0 0 CON01; CON02 18 TP 0.21 0.067 0.011 0.0001 CON02 soil class 15 = limestone/d olomite clay; class 24 = coastal loamy sand; class 28 = silty clay alluvium; class 49 = sandy clay alluvium (*) indicates significant linear relationship between nutrient and cland cover or soil class

PAGE 110

110 Table 5 9. A comparison of nutrient export coef ficients (kg ha 1 yr 1 ) from tropical watersheds Yield Stream/River size (ha) land cover DIN TSN TN SRP TSP TP Location/Reference kg ha 1 yr 1 Crique Sarco creek 1084 75% agr 1.04 2.25 0.03 0.09 southern Belize / this study Yax Cal creek 469 77% agr 0.28 0.53 0.03 0.04 southern Belize / this study Sundaywood creek 3204 60% agr 0.12 0.18 0.00 0.01 southern Belize / this study Conejo creek 1384 59% agr 2.04 2.48 0.02 0.03 southern Belize / this study Braco do Mota 23.4 80% agr 3.64 6.44 9.14 0.08 0.33 0.48 central Amazon (A) Igarape de Mota 18 >95% forest 2.67 3.61 4.31 0.02 0.05 0.08 central Amazon (A) Tempisquito 319 >95% forest 6.10 0.57 Costa Rica (B) Tempisquito Sur 311 >95% forest 4.90 0.33 Costa Rica (B) Kathia 264 >95 % forest 5.60 0.34 Costa Rica (B) Marilin 36 >95% forest 4 0.46 Costa Rica (B) El Jobo 55 >95% forest 4.3 0.34 Costa Rica (B) Zompopa 37 >95% forest 6 0.43 Costa Rica (B) Icacos 326 >95% forest 3.2 8.01 9.8 0.07 Puerto Rico (C) Sonad ora 262 >95% forest 1.69 5.43 5.9 0.05 Puerto Rico (C) Toronja 16.2 >95% forest 1.16 3.96 4.4 0.03 Puerto Rico (C) (A) Williams and Melack 1997; (B) Newbold et al. 1995; (C) McDowell and Asbury 1994

PAGE 111

111 Figure 5 1. Map of study area, including (A) Belize and the location of the Temash River watershed in southern Belize. (B) Four catchments were selected for in stream nutrient monitoring Crique Sarco (CRS), Yax Cal (YXL), Sunday Wood (SWD), and Conejo (CON).

PAGE 112

112 Figure 5 2. Precipitation reco rd for three stations in the Temash River watershed from February 2, 2007 through June 24, 2008. Shaded vertical bars mark the seasonal boundaries between dry and rainy seasons.

PAGE 113

113 Figure 5 3. Absolute and normalize d (by watershed area) discharge rates for the study watersheds.

PAGE 114

114 Figure 5 4. Seasonal mean instantaneous discharge rates (m 3 sec 1 ) versus watershed area (ha). The two outliers (gray triangles) are YXL01 and SWD01.

PAGE 115

115 Figure 5 5. Daily nitrogen flux es from the study watersheds

PAGE 116

116 Figure 5 6 Daily phosphorus fluxes from the study watersheds

PAGE 117

117 Figure 5 7. Flow weighted mean concentrations for DIN, TSN, and TN. Concentrations are averaged across seasonal (wet vs. dry) and annual (wet and dry comb ined) time periods.

PAGE 118

118 Figure 5 8. Flow weighted mean concentrations for SRP, TSP, and TP. Concentrations are averaged across seasonal (wet vs. dry) and annual (wet and dry combined) time periods.

PAGE 119

119 CHAPTER 6 CONCLUDING REMARKS Land use and land cover change occurs across varying spatial and temporal scales and can alter ecosystem structure and function in the lowland tropical watersheds of Belize. Local conditions within the riparian zone can alter thermal conditions and nutrient biogeochemistry in t ropical rivers. Aquatic ecosystems are impacted most severely when land use change within riparian zones occurs rapidly across multiple river reaches. Human impact mapping within riparian zones is an important conservation tool that can detect trends in land use change and estimate the severity of stress stemming from these changes. The resulting maps are easily interpreted and provide spatially explicit data on aquatic ecosystem stresses useful for natural resource managers, policy decision makers, and the general public. Human impact mapping represents a rapidly deployable and low cost method for collecting spatially explicit information about potential threats to aquatic ecosystems. However, one challenge with such rapid assessment techniques is that they do not accurately quantify threats not immediately observed during the mapping effort. In a more detailed assessment of riparian zone land use, interviews with small scale farmers revealed that agrochemicals represent an emerging threat to a quatic ec osystems. Whereas agrochemical use within plantation scale farming systems (i.e., banana and citrus plantations) is well documented, my research in the Temash River watershed documented the common practice of agrochemical use within riparian zone farming plots. This research points to the need for more detailed studies of the impact of agrochemical use on aquatic ecosystems in Belize.

PAGE 120

120 LULCC occurring at larger spatial scales can also influence aquatic ecosystem function. The dominant land use within the Temash River watershed is slash and burn agriculture and the survey of soils along the chronosequence revealed that, when practiced with a 15 year fallow period, this land use practice does not significantly impact soil nutrients. The slash and burn agric the Temash River watershed is not detrimental to the conservation of tropical forests. However, pasture expansion does threaten to alter land use practices within the watershed. Pasture requires large tracts of lan d and it permanently removes land from the traditional fallow rotation used for slash and burn. Carbon isotopic analysis of soil organic matter revealed that the introduction of C4 pasture grasses has altered the carbon isotopic signature of pasture soils relative to forest soils and the organic matter turnover rate in pastures varies with regard to the age of the pasture. Although C4 corn is the dominant crop that is cultivated within the Temash, soils under shifting cultivation still reflect the is otopic signature of C3 forest derived organic matter. R esults from the in stream nutrient analysis suggest that nutrient concentrations within the tributaries and main channel of the Temash River watershed are largely controlled by the seasonal fluctuations in hy drology. During periods of low flow (i.e., dry season), nutrient concentrations remain low and vary little across time and space. At the onset of the rainy season, nutrient concentrations peak, suggesting a pulse of nutrients being transported from terrest rial areas into the aquatic ecosystem. Following this initial peak, nutrient concentrations vary little throughout the remainder of the rainy season.

PAGE 121

121 APPENDIX A ELSI RANK SCORES FOR OBSERVED STRESSES

PAGE 122

122 Stress Sources Source Code Contrib. Irrevers. Rank Rank Score Sedimentation No riparian buffer commercial agriculture NB (BAN) H H H 10 No riparian buffer milpa NB (MLP) M H M 5 No riparian buffer old milpa NB (OLDMLP) L H M 5 No riparian buffer cattle grazing NB (GRZ) H H H 7.5 No riparian buffer buildings/residential NB (house) L H M 5 Thin Buffer commercial agriculture TB (BAN) M H M 5 Thin Buffer milpa TB (MLP) L H M 5 Thin Buffer old milpa TB (OLDMLP) L H M 5 Thin Buffer cattle grazing TB (GRZ) H H H 7.5 Thin Buffer buildings/residen tial TB (house) L H M 5 In stream gravel mining GRV V H V 10 Channelization CHN H H H 7.5 In stream Grazing GRZ V H V 10 Road access RD H H H 7.5 Drainage ditch DD M H M 5 Nutrient loading No riparian buffer commercial agriculture NB (BA N) H M M 5 No riparian buffer milpa NB (MLP) L M L 2.5 No riparian buffer old milpa NB (OLDMLP) L M L 2.5 No riparian buffer cattle grazing NB (GRZ) H M M 5 No riparian buffer buildings/residential NB (house) L M L 2.5 Thin Buffer commercial agri culture TB (BAN) M M M 5 Thin Buffer milpa TB (MLP) L M L 2.5 Thin Buffer old milpa TB (OLDMLP) L M L 2.5 Thin Buffer cattle grazing TB (GRZ) M M M 5 Thin Buffer buildings/residential TB (house) L M L 2.5 Laundry CU L M L 2.5

PAGE 123

123 In stream Grazing GRZ H M M 5 Drainage ditches DD V M H 7.5 Toxins/Contaminants No riparian buffer commercial agriculture NB (BAN) H H H 7.5 No riparian buffer milpa NB (MLP) M H M 5 In stream gravel mining GR L H M 5 Thin Buffer commercial agriculture TB ( BAN) M H M 5 Thin Buffer milpa TB (MLP) L H M 5 Pump house PH L H M 5 Road access RD L H M 5 Drainage ditches DD V H V 10 Altered flow regime Water pumping PH M M M 5 In stream gravel mining GRA L M L 2.5 Drainage ditches DD L M L 2.5 Habitat alteration/ fragmentation No riparian buffer commercial agriculture NB (BAN) M M M 5 No riparian buffer milpa NB (MLP) M M M 5 No riparian buffer old milpa NB (OLDMLP) M M M 5 No riparian buffer cattle grazing NB (GRZ) M M M 5 No r iparian buffer buildings/residential NB (house) M M M 5 Thin Buffer commercial agriculture TB (BAN) L M L 2.5 Thin Buffer milpa TB (MLP) L M L 2.5 Thin Buffer old milpa TB (OLDMLP) L M L 2.5 Thin Buffer cattle grazing TB (GRZ) L M L 2.5 Thin Buff er buildings/residential TB (house) L M L 2.5 In stream gravel mining GR H M M 5 In stream Grazing GRZ M M M 5 Channelization CHN L M L 2.5 Pump house PH L M L 2.5 Sandbag dam DAM L L L 2.5 Gravel mining GRV L L L 2.5

PAGE 124

124 Thermal alteratio n No riparian buffer commercial agriculture NB (BAN) H M M 5 No riparian buffer milpa NB (MLP) H M M 5 No riparian buffer old milpa NB (OLDMLP) H M M 5 No riparian buffer cattle grazing NB (GRZ) H M M 5 No riparian buffer buildings/residential NB ( house) H M M 5 Drainage ditches DD L M M 5 Thin Buffer commercial agriculture TB (BAN) L M M 5 Thin Buffer milpa TB (MLP) L M M 5 Thin Buffer old milpa TB (OLDMLP) L M M 5 Thin Buffer cattle grazing TB (GRZ) L M M 5 Thin Buffer buildings/reside ntial TB (house) L M M 5

PAGE 125

125 APPENDIX B MATAMBRE INTERVIEW QUESTIONAIRE Matambre Practices in the Temash River Watershed, Belize David G. Buck School of Natural Resources & Environment University of Florida Informed Consent Please read this consent docu ment carefully before you decide to participate in this study. My name is David Buck. I am a graduate student supervised by Dr. Mark Brenner, a professor at the University of Florida, in the USA. His address is P.O. Box 112120, University of Florida Gainesville, FL 32611, USA. I would like to ask a few questions and talk with you about land use history and agricultural practices on your land and the overall landscape in this region. It will take about one hour to ask these questions. Answer ing these questions will not affect you for better or worse. You do not have to answer any question you do not wish to answer. Your participation is completely voluntary and I can offer you no compensation for your participation in this survey. I have h ad other participants refuse before. You do not have to stop working to speak with us. If you would prefer, I can come back at another time. I will not write down your name and all answers or information you share with us will be private. If you hav e any questions or concerns about your rights, they can be directed to the UFIRB office, P.O. Box 112250, University of Florida Gainesville, FL 32611, USA. If you have additional questions regarding the research, data analysis and/or use of the results I can be contacted locally by phone (669 4336). Do you have any questions? May I begin asking my questions? Remember, you can stop us at any time or we can schedule to meet some other time. I have read the procedure described above. I voluntarily agree to participate in the procedure and I have received a copy of this description. A. Interview Info A1. Participant A2. Village A3. Interviewer A4. P.I. Notes: B. Household Info data provided elsewhere C. Matambre Qu estions C1. Did you plant matambre this year ? [Yes] [No] C2. Where did you plant? a) along Temash River c) on a hillside b) along creek [name:_______________ ] d) other (where: _________________) C3. How do you travel to your matambre field and how lo ng does it take? a) Walk __________(time) d) Horse__________(time) b) Bike__________(time) e) Other _______ ; __________(time) c) Dory__________(time) C4. If you were to walk to your field, how long would it take you? ________________ (time)

PAGE 126

126 C5. What is the s ize of your field? _______ (circle one:acres / manzanas / hectares) C6. How long did it take to clear? __________________ (days) C7. Did anyone help you clear the field? [Yes] [No] If NO, proceed to C9 C8. If yes to C5, how many men helped you clear ? ____________ (# of men) C9. Did anyone help you plant your field? [Yes] [No] If NO, proceed to C11 C10. How many people helped you plant your field? ______________ (# of men) C11. Did you plant anything else other than corn in your field? [Yes] __ ________, __________, __________, __________, __________ (specify) [No] C12. On average, do you reap more corn from your matambre or milpa? (circle one) Average number of bags of unshelled corn/acre for matambre ______________ Average number of bags o f unshelled corn/acre for milpa ______________ D. Field Maintenance /Use D1. Will you weed your matambre after planting? [Yes] [No] If NO, proceed to D3 D2. How frequently will you weed your field? a) once c) three times b) twice d) other __________ (# of times) D3. Do you use herbicides on your field? [Yes] [No] If NO, proceed to D6 D4. What type of herbicide do you use? a) 2, 4 D c) Round Up b) Gramoxone d) other __________ D5. How frequently do you apply herbicides to your field? a) once c) three t imes b) twice d) other __________ (# of times) D6. Have you received training in how to use agrochemicals? [Yes] [No] If NO, proceed to D8 D7. Did you receive this training from the Pesticide Control Board? If so, where? [Yes] ________________ [No] D8. If you have excess corn from your matambre, will you sell it? [Yes] [No] If NO, proceed to Section E. Riparian Zone D9. How much will you sell it for? ____________ (amount $BZ/pound) D10. Will you sell it in the village or in town? [Punta Gorda] [village ___________ ]

PAGE 127

127 E. Riparian Zone / Water Quality / Conservation a) _____________ b) _____________ c) _____________ d) _____________ e) _____________ f) _______ ______ g) _____________ h) _____________ i) _____________ j) _____________ k) _____________ l) _____________ E2. Is water in the Temash River clean? [Yes] [No] E3. Are the creeks near your village clean? [Yes] [No] E4. What makes the creeks and river dirty and not sa fe to drink? a) cattle d) chemicals b) clearing land e) other _________; _________; _________; _________ c) matambre E5. Do other communities in the watershed effect the water near your village? [Yes] How? _________________________________________ __________ _____________________________________________________________ [No] E6. Have you ever gotten sick from drinking or swimming in the river/creek? [Yes] _______________ (what kind of sickness) [No] E7. What caused your sickness? _____________ _________________________________________ E8. Do you collect drinking water from the river, creek or spring? [Yes]____________(specify which river/creek/spring) [No] E9. Do you wash clothes/dishes in the river or creek? [Yes]_________(specify river/cr eek) [No] E10. Does anyone control access to the river or creeks? Who? [Yes] _________________ [No] E11. Has there ever been any conflict in the village about water or water resources? [Yes] [No] briefly describe: ____________________________ _________________________________ __________________________________________________________________________________ E12. Who resolves those conflicts? a) Alcalde c) government b) village council d) other: ___ ____________________

PAGE 128

128 E13. When you farm, harvest timber, or clear for pasture, do you leave a buffer of trees along the edge of the stream or river? [Yes] ___________ (how wide meters) [No] E14. Would you participate in a conservation program that encouraged maintaining a buffer of forest along the edges of all creeks and the river in the Temash Watershed? [Yes] [No] E15 Would you participate in a tree planting program that planted trees along the edges of the creeks and river? [Yes] [No]

PAGE 129

129 LIST OF REFERENCES Achard, F., H. D. Eva, H. J. Stibig, P. Mayaux, J. Gallego, T. Richards, and J. P. Malingreau. 2002. Determination of deforestation rates of the World's humid tropical forests. Science 297 :999 1002. Alcorn, J., and V M. Toledo. 1998. Resilient resource management in Mexico's forest ecosystems: the contribution of property rights. Pages 216 249 in F. Berkes and C. Folke, editors. Linking social and ecological systems. Cambridge University Press, Cambridge, UK. Allan, J. D. 2004. Landscapes and Riverscapes: The Influence of Land Use on Stream Ecosystems. Annual Review of Ecology, Evolution and Systematics 35 :257 284. Allan, J. D., D. L. Erickson, and J. Fay. 1997. The influence of catchment land use on stream integrity across multiple spatial scales. Freshwater Biology 37 :149 161. Ankers, C., D. E. Walling, and R. P. Smith. 2003. The influence of catchment characteristics on suspended sediment properties. Hydrobiologia 494 :159 167. APHA. 1989. Standard methods for the ex amination of water and wasterwater, 17 edition. American Public Heath Association. Arbuckle, K., and J. A. Downing. 2001. The influence of watershed land use on lake N:P in a predominantly agricultural landscape. Limnology and Oceanography 46 :970 975. Arun achalam, A. 2002. Dynamics of soil nutrients and microbial biomass during first year cropping in an 8 year jhum cycle. Nutrient cycling in agroecosystems 64 :283 291. Asmussen, L. E., A. W. White, E. W. Hauser, and J. M. Sheridan. 1977. Reduction of 2,4 D l oad in surface runoff down a grassed waterway. Journal of Environmental Quality 6 :159 162. Assessment, M. E. 2005. Ecosystems and Human Well being: Synthesis. Island Press, Washington, D.C. Atran, S. 1993. Itza Maya tropical agro forestry. Current Anthropo logy 34 :633 700. Atran, S., D. Medin, N. Ross, E. Lynch, J. Coley, E. U. Ek', and V. Varnarsky. 1999. Folkecology and commons management in the Maya Lowlands. Proceedings of the National Academy of Sciences 96 :7598 7603. Attiwill, P. M. 1994. The disturban ce of forest ecosystems: the ecological basis for conservative management. Forest Ecology and Management 63 :247 300.

PAGE 130

130 Awiti, A. O., M. G. Walsh, and J. Kinyamario. 2008. Dynamics of topsoil carbon and nitrogen along a tropical forest cropland chronosequence : Evidence from stable isotope analysis and spectroscopy. Agriculture, Ecosystems and Environment 127 :265 272. Azaroff, L. S., and L. M. Neas. 1999. Acute health effects associated with nonoccupational pesticide exposure in rural El Salvador. Environmental Research Section A 80 :158 164. Bailey, M., P. H. Kirshen, E. Spang, S. Morrison, C. Pizarro, G. Kayser, G. Cid, and A. Parish. 2007. Rapid assessment of anthropogenic impacts on select transboundary watersheds of the Mesoamerican Barrier Reef System (MBRS ) region. MBRS Project, National Oceanic and Atmospheric Administration and Tufts University. Baillie, I. C., A. C. S. Wright, M. A. Holder, and E. A. Fitzpatrick. 1993. Revised Classification of the Soils of Belize. NRI Bulletin 59, Natural Resources Inst itute, Chatham, UK. Balesdent, J., and A. Mariotti. 1996. Measurement of soil organic matter turnover using 13 C natural abundance. Mass Spectrometry of Soils:83 111. Balesdent, J., A. Mariotti, and B. Guillet. 1987. Natural 13C abundance as a tracer for s tudies of soil organic matter dynamics. Soil Biology & Biochemistry 19 :25 30. Berkes, F., and C. Folke. 2002. Back to the future: ecosystem dynamics and local knowledge. Pages 121 146 in L. H. Gunderson and C. S. Holling, editors. Panarchy: understanding t ransformations in human and natural systems. Island Press, Washington. Bernard, H. R. 2006. Research methods in anthropology: qualitative and quantitative approaches, 4th edition. Altamira Press, New York. Biggs, T. W., T. Dunne, T. F. Domingues, and L. A. Martinelli. 2002. Relative influence of natural watershed properties and human disturbance on stream solute concentrations in the southwestern Brazilian Amazon basin. Water Resources Research 38 :(10.1029/2001WR000271). Biggs, T. W., T. Dunne, and L. A. Ma rtinelli. 2004. Natural controls and human impacts on stream nutrient concentrations in a deforested region of the Brazilian Amazon basin. Biogeochemistry 68 :227 257. Blanco, L. E., A. Aragn, I. Lundberg, C. Lidn, C. Wesseling, and G. Nise. 2005. Determi nants of dermal exposure among Nicaraguan subsistence farmers during pesticide applications with backpack sprayers. The Annals of Occupational Hygiene 49 :17 24. BMAF. 2006. Belize Ministry of Agriculture and Fisheries Annual Report 2006.

PAGE 131

131 Boles, E. 1998. Developing a Community Based Rapid Watershed Assessment Protocol Focusing on the Sibun River Watershed in Belize as a Case Study. Jackson State University, Jackson, MS. Boles, R., D. G. Buck, and P. C. Esselman. 2008. Synthesis of Water Resource Conservati on, Management and Research Activities in the Watersheds of Belize. The Nature Conservancy Belize program. BPCB. 2008. Belize Restricted Use Pesticides. in http://www.pcbbelize.com/ BPCB. unpublished data. Be lize Pesticide Imports from 2002 to 2007. in Belize Pesticide Control Board. Bray, D. B., E. A. Ellis, N. Armijo Canto, and C. T. Beck. 2004. The institutional drivers of sustainable landscapes: a case study of the 'Mayan Zone' in Quintana Roo, Mexico. La nd Use Policy 21 :333 346. Brenner, M., M. F. Rosenmeier, D. A. Hodell, and J. H. Curtis. 2002. Paleolimnology of the Maya Lowlands: long term perspectives on interactions among climate, environment, and humans. Ancient Mesoamerica 13 :141 157. Brubacher, D. J. T. Arnason, and J. Lambert. 1989. Woody species and nutrient accumulation during the fallow period of milpa farming in Belize, CA. Plant and Soil 114 :165 172. Bruijnzeel, L. A. 2004. Hydrological functions of tropical forests: not seeing the soil for the trees? Agriculture, Ecosystems and Environment 104 :185 228. Buck, D. G., P. C. Esselman, and R. Boles. 2008. Status, Needs and Future Directions for Effective Freshwater Conservation in Belize, Central America. The Nature Conservancy Belize program. Burnet, A. M. R. 1972. Effects of paraquat on invertebrates in a Canterbury stream, New Zealand. New Zealand Journal of Marine and Freshwater Research 6 :448 455. Burns, C. J., W. M. Mahlburg, and J. P. Dutra. 2007. Pesticide exposure among farm workers. En vironmental Research 105 :285 286. Buschbacher, R., C. Uhl, and E. A. S. Serro. 1988. Abandoned pastures in Eastern Amazonia. II. Nutrient stocks in the soil and vegetation. Journal of Ecology 76 :682 699. Cantor, K. P., A. Blair, G. Everett, R. Gibson, L. F. Burmeister, L. M. Brown, L. Schoman, and F. R. Dick. 1992. Pesticides and other agricultural risk factors for non Hodgkin's lymphoma among men in Iowa and Minnesota. Cancer Research 52 :2447 2455.

PAGE 132

132 Caraco, N. F., and J. J. Cole. 1999. Human impact on nit rate export: an analysis using major world rivers. Ambio 28 :167 170. Carpenter, S. R., N. F. Caraco, D. L. Correll, R. W. Howarth, A. N. Sharpley, and V. H. Smith. 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecological Applicat ions 8 Carr, D. L. 2002. Rural Frontier Migration and Deforestation in the Sierra de Lacandon National Park, Guatemala. University of North Carolina. Carr, D. L. 2004. Ladino and Q'eqch Maya land use and land clearing in the Sierra de Lacandn National P ark, Petn, Guatemala. Agriculture and Human Values 21 :171 179. Carr, D. L. 2008. Farm households and land use in a core conservation zone of the Maya Biosphere Reserve, Guatemala. Human Ecology 36 :231 248. Castillo, L. E., E. d. l. Cruz, and C. Ruepert. 1 997. Ecotoxicology and pesticides in tropical aquatic ecosystems of Central America. Environmental Toxicology and Chemistry 16 :41 51. derived n alkanes in soil particle size fractions. Organic Geochemistry 32 :253 258. Chomitz, K. M., and D. A. Gray. 1996. Roads, Land Use, and Deforestation: A Spatial Model Applied to Belize. The World Ban k Economic Review 10 :487 512. Chowdhury, R. R. 2006. Driving forces of tropical deforestation: the role of remote sensing and spatial models. Singapore Journal of Tropical Geography 27 :82 101. Cowgill, U. M. 1962. An agricultural study of the Southern Maya Lowlands. American Anthropologist 64 :273 286. Davis, C. O., and M. S. Simmons. 1979. Water chemistry and phytoplankton field and laboratory procedures. Derrien, D., and W. Amelung. 2011. Computing the mean residence time of soil carbon fractions using sta ble isotopes: impacts of the model framework. European Journal of Soil Science 62 :237 252. Desjardins, T., F. Andreux, B. Volkoff, and C. C. Cerri. 1994. Organic carbon and 13C contents in soils and soil size fractions, and their changes due to deforestati on and pasture installation in Eastern Amazonia. Geoderma 61 :103 118. Desjardins, T., E. Barros, M. Sarrazin, C. Giradin, and A. Mariotti. 2004. Effects of forest conversion to pasture on soil carbon content and dynamics in Brazilian Amazon. Agriculture Ec osystems & Environment 103 :365 373.

PAGE 133

133 Diekmann, L. O., D. Lawrence, and G. S. Okin. 2007. Changes in the spatial variation of soil properties following shifting cultivation in a Mexican tropical dry forest. Biogeochemistry 84 :99 113. Diemont, S. A. W., and J F. Martin. 2009. Lacandon Maya ecosystem management: sustainable design for subsistence and environmental restoration. Ecological Applications 19 :254 266. Diemont, S. A. W., J. F. Martin, S. I. Levy Tacher, R. B. Nigh, P. R. Lopez, and J. D. Golicher. 20 06. Lacandon Maya forest management: Restoration of soil fertility using native tree species. Ecological Engineering 28 :205 212. Dckersmith, I. C., C. P. Giardina, and R. L. Sanford. 1999. Persistence of tree related patterns in soil nutrients following s lash and burn distrubance in the tropics. Plant and Soil 209 :137 156. Downing, J. A., M. McClain, R. Twilley, J. M. Melack, J. Elser, N. N. Rabalais, W. M. Lewis, R. E. Turner, J. Corredor, and D. Soto. 1999. The impact of accelerating land use change on t he N Cycle of tropical aquatic ecosystems: Current conditions and projected changes. Biogeochemistry 46 :109 148. Dreiher, J., and E. Kordysh. 2006. Non Hodgkin Lymphoma and pesticide exposure: 25 years of research. Acta Haematologica 116 :153 164. Dudgeon, D., A. H. Arthington, M. O. Gessner, Z. I. Kawabata, D. J. Knowler, E. Lvque, R. J. Naiman, A. H. Prieur Richard, D. Soto, M. L. J. Staissny, and C. A. Sullivan. 2006. Freshwater biodiversity: importance, threats, status, and conservation challenges. Bio logical Reviews 81 :163 182. Eaton, J. M., and D. Lawrence. 2009. Loss of carbon sequestration potential after several decades of shifting cultivatoin in the Southern Yucatan. Forest Ecology and Management 258 :949 958. Eisler, R. 1990. Paraquat Hazards to F ish, Wildlife, and Invertebrates: A Synoptic Review. U.S. Fish and Wildlife Service, Laurel, Maryland. Ekstrm, A. M., M. Eriksson, L. E. Hansson, A. Lindgren, L. B. Signorello, O. Nyrn, and L. Hardell. 1999. Occupational exposures and risk of gastric can cer in a population based case control study. Cancer Research 59 :5932 5937. Emch, M. 2003. The Human Ecology of Mayan Cacao Farming in Belize. Human Ecology 31 :111 131. ERMA. 2007. Draft National Integrated Water Resources Management Policy (Including Clim ate Change) for Belize. Consultants report to the Ministry of Natural Resources and the Environment, Belmopan, Belize. Escobichon, D. J. 2001. Pesticide use in developing countries. Toxicology 160 :27 33.

PAGE 134

134 Esselman, P. C. 2001. The Monkey River Baseline Stud y: Basic and Applied Research for Monitoring and Assessment in Southern Belize. University of Georgia, Athens, GA. Esselman, P. C. 2009. Fish Communities and Conservation of Aquatic Landscapes in Northeastern Mesoamerica. University of Michigan, Ann Arbor, MI. Esselman, P. C., and E. Boles. 2001. Status and future needs of limnological research in Belize. Pages 35 68 in R. G. Wetzel and B. Gopal, editors. Limnology in Developing Countries. Esselman, P. C., and D. G. Buck. 2007. Hydrologic Assessment of the Monkey River Watershed, Belize. Phase 1: Human Impact Mapping Along the Monkey River and its Main Tributaries. Toledo Institute for Development and the Environment. Esselman, P. C., M. C. Freeman, and C. M. Pringle. 2006. Fish assemblage variation between geologically defined regions and across a longitudinal gradient in the Monkey River Basin, Belize. Journal of the North American Benthological Society 25 :142 156. Evers, D. C. 2008. Assessing the global threat of mercury: A case study of the Mesoamerican Reef with preliminary findings for sharks and groupers. Report BRI 2007. BioDiversity Research Institute, Gorham, ME. Ewel, J., C. Berish, B. Brown, N. Price, and J. Raich. 1981. Slash and burn impacts on a Costa Rican wet forest site. Ecology 62 :816 82 9. Ewel, J. J. 1986. Designing agricultural ecosystems for the humid tropics. Annual review of ecology and systematics 17 :245 271. Farella, N., R. Davidson, M. Lucotte, and S. Daigle. 2007. Nutrient and mercury variations in soils from family frams of the Tapajs region (Brazilian Amazon): Recoomendations for better farming. Agriculture, Ecosystems and Environment 120 :449 462. Fearnside, P. M., and R. I. Barbosa. 1998. Soil carbon changes from conversion of forest to pasture in Brazilian Amazonia. Forest Ec ology and Management 108 :147 166. Ferguson, B. G., J. Vandermeer, H. Morales, and D. M. Griffith. 2003. Post agricultural succession in El Petn, Guatemala. Conservation Biology 17 :818 828. Filoso, S., L. A. Martinelli, R. W. Howarth, E. W. Boyer, and F. D entener. 2006. Human activities changing the nitrogen cycle in Brazil. Biogeochemistry 79 :61 89. Filoso, S., L. A. Martinelli, M. R. Williams, L. B. Lara, A. Krusche, M. V. Ballester, R. Victoria, and P. B. deCamargo. 2003. Land use and nitrogen export in the Piracicaba River basin, Southeast Brazil. Biogeochemistry 65 :275 294.

PAGE 135

135 Fitzgerald, D., D. S. Chanasyk, D. R. Neilson, D. Kiely, and R. Audette. 2001. Farm well water quality in Alberta. Water Quality Research Journal of Canada 36 :565 588. Gagneten, A. M ., and M. Marchese. 2003. Effect of paraquat on freshwater zooplankton and zoobenthos assemblages in enclosure experiments. International Journal of Ecohydrology and Hydrobiology 3 :389 398. Galvao, L. A., J. A. Escamilla, S. Henao, E. Loyola, C. S. Castill o, and P. Arbelaez. 2004. Pesticides and Health in the Central American Isthmus. PAHO, Washington, D.C. Garabrant, D. H., and M. A. Philbert. 2002. Review of 2,4 Dichlorophenoxyacetic acid (2,4 D) epidemiology and toxicology. Critical Reviews in Toxicology 32 :233 257. Gavlak, R. G., D. A. Horneck, and R. O. Miller. 1994. Plant, Soil and Water Reference Methods for the Western Region. WREP 125. Geist, H. J., and E. F. Lambin. 2002. Proximate Causes and Underlying Driving Forces of Tropical Deforestation. Bio Science 52 :143 150. Giardina, C. P., J. Robert L. Sandford, I. C. Dckersmith, and V. L. Jaramillo. 2000. The effects of slash burning on ecosystem nutrients during the land preparation phase of shifting cultivation. Plant and Soil 220 :247 260. Giovannini, G., S. Lucchesi, and M. Giachetti. 1990. Effects of heating on some chemical parameters related to soil fertility and plant growth. Soil Science 149 :344 350. Glynn, P. W., L. S. Howard, E. Corcoran, and A. D. Freay. 1984. The occurrence and toxicity of he rbicides in reef building corals. Marine Pollution Bulletin 15 :370 374. GOB. 2002. The Europoean Union Grants 6.00 Million Belize Dollars to the Banana Industry, Government of Belize, accessed 11/27/2008 ( http://www.governmentofbelize.gov.bz/press_release_details.php?pr_id=2016) in Press Office, Government of Belize. Gomez Montes, C., and N. J. Bayly. 2010. Habitat use, abundance, and persistence of Neotropical migra nt birds in a habitat matrix in northeast Belize. Journal of Field Ornithology 81 :237 251. Gonzalez, V. J. B. 1980. A Limnological Investigation of a Tropical Fresh Water Ecosystem: The Belize River, Belize, Central America. Western Michigan University, Ka lamazoo, MI. Gore, J. A. 2006. Discharge measurements and streamflow analysis. Pages 51 78 in F. R. Hauer and G. A. Lamberti, editors. Methods in Stream Ecology. Elsevier Inc., London.

PAGE 136

136 Grandia, L. 2004. The Wealth Report: Q'eqchi' traditional knowledge and natural resource management in the Sarstoon Temash National Park. Krishna Copy Center, Berkeley, California. Grayson, R. B., C. J. Gippel, B. L. Finlayson, and B. T. Hart. 1997. Catchment wide impacts on water quality: the use of 'snapshot' sampling durin g stable flow. Journal of Hydrology 199 :121 134. Greenfield, D. W., and J. E. Thomerson. 1997. Fishes of the Continental Waters of Belize. University of Florida Press, Gainesville, FL. GUARD. 2007. Monkey River Village Erosion Study: An Assessment and Prop osed Mitigation. Monkey River Village Council and the Protected Areas Conservation Trust. Hamilton, S., and E. F. Fischer. 2005. Maya farmers and export agriculture in highland Guatemala: implications for development and labor relations. Latin American Per spectives 32 :33 58. Hardell, L. L. 1979. Malignant lymphoma of histiocytic type and exposure to phenoxyacetic acids or chlorophenols. Lancet 1 :55 56. Harding, J. S., E. F. Benfield, P. V. Bolstad, G. S. Helfman, and E. B. D. Jones. 1998. Stream biodiversit y: the ghost of land use past. Proceedings of the National Academy of Sciences 95 :14843 14847. Herrera, A. 2004. Sartoon Temash National Park Management Plan. in Punta Gorda, Belize. Hintze, J. 2001. NCSS and PASS. Number Cruncher Statistical Systems. in Kaysville, Utah. Hobbie, S. E., and P. M. Vitousek. 2000. Nutrient Limitation of Decomposition in Hawaiian Forests. Ecology 81 :1867 1877. Holdridge, L. R. 1971. Forest Environments in Tropical Life Zones: A Pilot Study. Pergamon Press, New York. Holland, B., N. Brorsen, and B. Johnson. 2003. Assessment of Geology and Soils in the Sarstoon Temash Region, Toledo District, Belize, Central America. Sarstoon Temash Institute for Indigenous Management, Punta Gorda. Hlscher, D., B. Ludwig, R. F. Mller, and H. Fster. 1997. Dynamic of soil chemical parameters in shifting cultivation in the Eastern Amazon. Agriculture, Ecosystems and Environment 66 :153 163. Houghton, R. A. 1994. The Worldwide Extent of Land Use Change. BioScience 44 :305 313.

PAGE 137

137 Howard, P. H. 1989. H andbook of Environmental Fate and Exposure Data for Organic Chemicals. Lewis Publishers, Chelsea, MI. Hughes, R. F., J. B. Kauffman, and V. J. Jaramillo. 1999. Biomass, carbon and nutrient dynamics of secondary forests in a humid tropical region of Mexico. Ecology 80 :1892 1907. Humphries, S. 1993. The intensification of traditional agriculture among Yucatec Maya farmers: facing up to the dilemma of livelihood sustainability. Human Ecology 21 :87 102. Hynes, H. B. N. 1975. The stream and its valley. Verh. Int ernat. Verein. Limnol 19 :1 15. Johnson, L. B., and S. H. Gage. 1997. Landscape approaches to the analysis of aquatic ecosystems. Freshwater Biology 37 :113 132. Johnson, L. B., C. Richards, G. E. Host, and J. W. Arthur. 1997. Landscape influences on water c hemistry in Midwestern stream ecosystems. Freshwater Biology 37 :193 208. Kaimowitz, D. 1995. The end of the hamburger connection? Livestock and deforestation in Central America in the 1980s and 1990s. in H. Cortes Salas, R. d. Camino, and A. Conteras, edit ors. Readings of the workshop on government policy reform for forestry conservation and development in Latin America. Inter American Institute for Cooperation on Agriculture, San Jose, CR. Kammerbauer, J., and J. Moncada. 1998. Pesticide residue assessment in three selected agricultural production systems in the Choluteca River basin of Honduras. Environmental Pollution 103 :171 181. Karper, J., and E. Boles. 2004. Human Impact Mapping of the Temash River, Belize. Sarstoon Temash Institute for Indigenous Man agement. Kauffman, J. B., D. L. Cummings, D. E. Ward, and R. Babbitt. 1995. Fire in the Brazilian Amazon: 1. Biomass, nutrient pools, and losses in slashed primary forests. Oecologia 104 :397 408. Kegley, S. E., B. R. Hill, S. Orme, and A. H. Choi. 2008. PA N Pesticide Database. in Pesticide Action Network, North America (San Francisco, CA, 2008), http://www.pesticideinfo.org Kemp, M. J., and W. K. Dodds. 2002. Spatial and temporal patterns of nitrogen concentrat ions in pristine and agriculturally influenced prairie streams. Biogeochemistry 53 :125 141. Keys, E. 2004. Commercial agriculture as creative destruction or destructive creation: a case study of chili cultivation and plant pest disease in the southern Yuca tan region. Land Degradation and Development 15 :397 409.

PAGE 138

138 Keys, E., and R. R. Chowdhury. 2006. Cash crops, smallholder decision making and institutional interactions in a closing frontier: Calakmul, Campeche, Mexico. American Geographer 5 :75 90. Kleinman, P J. A., R. B. Bryant, and D. Pimentel. 1996. Assessing ecological systainability of slash and burn agriculture through soil fertility indicators. Agronomy Journal 88 :122 127. Knicker, H. 2007. How does fire affect the nature and stability of soil organic nitrogen and carbon? A review. Biogeochemistry 85 :91 118. Kotto Same, J., P. L. Woomer, M. Appolinaire, and Z. Louis. 1997. Carbon dynamics in slash and burn agriculture and land use alternatives of the humid forest zone in Cameroon. Agriculture, Ecosyste ms and Environment 65 :245 256. Lambert, J. D. H., and J. T. Arnason. 1986. Nutrient dynamics in milpa agriculture and the role of weeds in initial stages of secondary succession in Belize, CA. Plant and Soil 93 :303 322. Lambin, E. F., B. L. Turner, H. J. G eist, S. B. Agbola, A. Angelsen, J. W. Bruce, O. T. Coomes, R. Dirzo, G. Fischer, and C. Folke. 2001. The causes of land use and land cover change: moving beyond the myths. Global Environmental Change, Part A: Human and Policy Dimensions 11 :261 269. Lawren ce, D., and D. Foster. 2002. Changes in forest biomass, litter dynamics and soils following shifting cultivation in Southern Mexico: an overview. Interciencia 27 :400 408. Lawrence, D., and W. H. Schlesinger. 2001. Changes in soil phosphorus during 200 year s of shifting cultivation in Indonesia. Ecology 82 :2769 2780. Lee, M. 2002. Human Impact Mapping in Crooked Tree Lagoon. School for International Training. Lee, M. D., J. D. Stednick, and D. M. Gilbert. 1995. Belize Environmental Water Quality Monitoring Final Report. Submitted as part of USAID NARMAP project to the Government of Belize. Levasseur, V., and A. Olivier. 2000. The farming system and traditional agroforestry systems in the Maya community of San Jose, Belize. Agroforestry Systems 49 :275 288. Levy Tacher, S. I., and J. D. Golicher. 2004. How predictive is traditional ecological knowledge? The case of the Lacandon Maya fallow enrichment system. Interciencia 29 :496 503.

PAGE 139

139 Lewis, W. M., S. K. Hamilton, and J. F. Saunders. 1995. Rivers of northern S outh America. in C. E. Cushing, K. W. Cummins, and G. W. Minshall, editors. Rivers and stream ecosystems. Elsevier, Amsterdam. Lewis, W. M., J. M. Melack, W. H. McDowell, M. McClain, and J. E. Richey. 1999. Nitrogen yields from undisturbed watersheds in th e Americas. Biogeochemistry 46 :149 162. Lide, D. R. 2004. CRC handbook of chemistry and physics. CRC Press, Boca Raton, FL. Likens, G. E., and F. H. Bormann. 1974. Linkages between Terrestrial and Aquatic Ecosystems. BioScience 24 :447 456. Lowrance, R., L. S. Altier, J. D. Newbold, R. R. Schnabel, P. M. Groffman, J. M. Denver, D. L. Correll, J. W. Gilliam, and J. L. Robinson. 1997a. Water quality functions of riparian forest buffers in Chesapeake Bay watersheds. Environmental Management 21 :687 712. Lowrance R., G. Vellidis, R. D. Wauchope, P. Gay, and D. D. Bosch. 1997b. Herbicide transport in a managed riparian forest buffer system. Transactions of the Ameican Society of Agricultural Engineers 40 :1047 1057. Manguin, S., D. R. Roberts, R. G. Andre, E. Rejm nkov, and S. Hakre. 1996. Characterization of Anopheles darlingi (Diptera: Culicidae) larval habitats in Belize, Central America. Journal of Medical Entomology 33 :205 211. Marengo, J. A., D. Liebmann, V. E. Kousky, N. P. Filizola, and I. C. Wainer. 2001. Onset and end of the rainy season in the Brazilian Amazon Basin. Journal of Climate 14 :833 852. Marques, M. N., A. I. Badiru, O. Beltrame, and M. A. F. Pires. 2007. Pesticide leaching and run off hazard in the Ribeira de Iguape river basin in So Paulo sta te, Brazil. Journal of the Brazilian Society of Ecotoxicology 2 :179 185. McClain, M. E. 2002. The application of ecohydrological principles for better water resources management in South America. Pages 193 209 in M. E. McClain, editor. The Ecohydrology of South American Rivers and Wetlands. IAHS Press, Wallingford. McClain, M. E. 2008. Ecohydrology as a tool in the sustainable development of large tropical rivers. in D. Harper, M. Zalewski, and N. Pacini, editors. Ecohydrology: Processes, Models and Case St udies. CABI, Oxfordshire. McClain, M. E., and R. E. Cossio. 2003. The use of riparian environments in the rural Peruvian Amazon. Environmental Conservation 3 :242 248.

PAGE 140

140 McClain, M. E., J. E. Richey, and T. P. Pimentel. 1994. Groundwater nitrogen dynamics at the terrestrial lotic interface of a small catchment in the Central Amazon Basin. Biogeochemistry 27 :113 127. McConnell, R., and A. Hruska. 1993. An epidemic of pesticide poisoning in Nicaragua: implications for prevention in developing countries. American Journal of Public Health 83 :1559 1562. McDowell, W. H., and C. E. Asbury. 1994. Export of carbon, nitrogen, and major ions from three tropical montane watersheds. Limnology and Oceanography 39 :111 125. McGrath, D. A., M. L. Duryea, N. B. Comerford, and W. P. Cropper. 2000. Nitrogen and phosphorus cycling in an Amazonian agroforest eight years following forest conversion. Ecological Applications 10 :1633 1647. McGrath, D. A., C. K. Smith, H. L. Gholz, and F. d. A. Oliveira. 2001. Effects of land use change o n soil nutrient dynamics in Amaznia. Ecosystems 4 :625 645. Meerman, J., P. Herrera, and A. Howe. 2003a. Rapid Ecological Assessment: Sarstoon Temash National Park, Toledo District, Belize. Sarstoon Temash Institute for Indigenous Management. Meerman, J., and J. R. Wilson. 2005. The Belize National Protected Areas System Plan. Taskforce on Belize's Protected Areas Policy and Systems Plan, Belmopan, Belize. Meerman, J. C., P. Herrera, and A. Howe. 2003b. Rapid Ecologial Assessment: Sarstoon Temash National P ark, Toledo District, Belize. Sarstoon Temash Institute for Indigenous Management. Meybeck, M. 1982. Carbon, nitrogen, and phosphorus transport by world rivers. American Journal of Science 287 :401 450. Mills, P. K., and R. C. Yang. 2007. Agricultural expos ures and gastric cancer risk in Hispanic farm workers in California. Environmental Research 104 :282 289. Moraes, J. M. d., A. E. Schuler, T. Dunne, R. d. O. Figueiredo, and R. L. Victoria. 2006. Water storage and runoff processes in plinthic soils under fo rest and pasture in eastern Amazonia. Hydrologic Processes 20 :2509 2526. Moran, E. F., E. S. Brondizio, J. M. Tucker, M. C. d. Silva Forsberg, S. McCracken, and I. Falesi. 2000. Effects of soil fertility and land use on forest succession in Amaznia. Fores t Ecology and Management 139 :93 108. Murray, D. L. 1994. Cultivating crisis: the human cost of pesticides in Latin America. University of Texas Press, Austin, TX.

PAGE 141

141 Murray, D. L., C. Wesseling, M. Keifer, M. Corriols, and S. Henao. 2002. Surveillance of pest icide related illness in the developing world: putting the data to work. International Journal of Occupational and Environmental Health 8 :243 248. Myers, N. 1991. Tropical forests: Present status and future outlook. Climatic Change 19 :3 32. Naiman, R. J., and H. Decamps. 1997. The ecology of interfaces: riparian zones. Annual Review of Ecology and Systematics 28 :621 658. Naiman, R. J., H. Decamps, and M. E. McClain. 2005. Riparia: Ecology, Conservation, and Management of Streamside Communities. Elsevier Aca demic Press, Burlington, MA. Nations, J. D., and R. B. Nigh. 1980. The evolutionary potential of Lacandon Maya sustained yield tropical forest agriculture. Journal of Anthropological Research 36 :1 30. Neill, C., L. A. Deegan, S. M. Thomas, and C. C. Cerri. 2001. Deforestation for pasture alters nitrogen and phosphorus in soil solution and stream water of small Amazonian watersheds. Ecological Applications 11 :1817 1828. Neill, C., J. M. Melillo, P. A. Steudler, C. C. Cerri, J. F. L. d. Moraes, M. C. Piccolo, and M. Brito. 1997. Soil carbon and nitrogen stocks following forest clearing and pasture in the southwestern Brazilian Amazon. Ecological Applications 7 :1216 1225. Nelson, D. W., and L. E. Sommers. 1996. Total carbon, organic carbon and organic matter. P ages 961 1010 in J. M. Bigham, editor. Methods of soil analysis. Part 3: Chemical and microbiological properties. ASA, CSSA, SAAJ, Madison, WI. Nepstad, D. C., C. R. d. Carvalho, E. A. Davidson, P. H. Jipp, P. A. Lefebvre, G. H. Negreiros, E. D. d. Silva, T. Stone, S. Trumbore, and S. Viera. 1994. The role of deep roots in the hydrological and carbon cycles of Amazonian forests and pastures. Nature 372 :666 669. NRCC. 1978. Phenoxy Herbicides Their Effects on Environmental Quality with Accompanying Scient ific Criteria for 2,3,7,8 Tetrachlorodibenzo p Dioxin (TCDD). Ottawa, Canada. Nunny, R., M. Santana, P. Stone, D. Tillet, and D. Walling. 2001. An investigation of the impact on reef environments of changing land use in the Stann Creek district of Belize. The Watershed Reef Interconnectivity Scientific Study (WRIScS) Module 3. EU Project No. B7 6200/97 02/VII/ENV. Nye, P. H., and D. J. Greenland. 1960. Soil Under Shifting Cultivation. Commonwealth Agricultural Bureaux.

PAGE 142

142 Nye, P. H., and D. J. Greenland. 1964 Changes in the soil after clearing tropical forest. Plant and Soil 21 :101 112. Pacini, N., D. M. Harper, V. Ittekkot, C. Humborg, and L. Rahm. 2008. Nutrient Processes and Consequences. Pages 30 45 in D. M. Harper, M. Zalewski, and N. Pacini, editors. Ec ohydrology: processes, models and case studies. CAB International, Cambridge. Parma de Croux, M. J., P. Arquiel, H. Ortego, and J. A. Lorente. 1999. Acute toxicity of paraquat to a common neotropical fish species (Apareiodon affinis) (Pisces, Hemiodidae). Polish Archive of Hydrobiology 46 :57 62. Penagos, H. 2002. Contact dermatitis caused by pesticides among banana plantation workers in Panama. International Journal of Occupational and Environmental Health 8 :14 18. Peterson, B. J., W. M. Wollheim, P. J. Mul holland, J. R. Webster, J. L. Meyer, J. L. Tank, E. Marti, W. B. Bowden, H. M. Valett, A. E. Hershey, W. H. McDowell, W. K. Dodds, S. K. Hamilton, S. Gregory, and D. D. Morrall. 2001. Control of nitrogen export from watersheds by headwater streams. Science 292 :86 90. Popper, R., K. Andino, M. Bustamante, B. Hernandez, and L. Rodas. 1996. Knowledge and beliefs regarding agricultural pesticides in rural Guatemala. Environmental Management 20 :241 248. Pringle, C. M., C. F. Rabeni, A. C. Benke, and N. G. Aumen. 1993. The role of aquatic science in freshwater conservation: cooperation between the North American Benthological Society and organizations for conservation and resource management. Journal of the North American Benthological Society 12 :177 184. Pringle, C. M., and F. J. Triska. 1991. Effects of geothermal groundwater on nutrient dynamics of a lowland Costa Rican stream. Ecology 72 :951 965. Puckett, L. J., and W. B. Hughes. 2005. Transport and fate of nitrate and pesticides: hydrogeology and riparian zone processes. Journal of Environmental Quality 34 :2278 2292. Rainwater, T. R., B. M. Adair, S. G. Platt, T. A. Anderson, G. P. Cobb, and S. T. McMurry. 2002. Mercury in Morelet's Crocodile eggs in northern Belize. Archive of Environmental Contamination and T oxicology 42 :319 324. Rainwater, T. R., T. H. Wu, A. G. Finger, J. E. Caas, L. Yu, K. D. Reynolds, G. Coimbatore, B. Barr, S. G. Platt, G. P. Cobb, T. A. Anderson, and S. T. McMurry. 2007. Metals and organochlorine pesticides in caudal scutes of crocodile s from Belize and Costa Rica. Science of the Total Environment 373 :146 156.

PAGE 143

143 Rao, P. S. C., and J. M. Davidson. 1980. Estimation of pesticide retention and transformation parameters required in nonpoint source pollution models. in M. R. Overcash and J. M. D avidson, editors. Environmental Impact of Nonpoint Source Pollution. Ann Arbor Science, Ann Arbor, MI. Read, L., and D. Lawrence. 2003. Recovery of biomass following shifting cultivation in dry tropical forests of the Yucatn. Ecological Applications 13 :85 97. Reddy, K. S., and R. P. Gambrell. 1987. Factors affecting the adsoprtion of 2,4 D and methyl parathion in soils and sediments. Agriculture, Ecosystems and Environment 18 :231 241. Requena, E. L. 2008. Human Impact Mapping of the Belize River Watershed: Branch Mouth to the Coast Line. University of Belize, Belmopan. Resh, V. H. 2007. Multinational, freshwater biomonitoring programs in the developing world: lessons learned from African and southeast Asian river surveys. Environmental Management 39 :737 748 Roberts, D. R., S. Manguin, E. Rejmnkov, R. Andre, R. E. Harbach, E. Vanzie, S. Hakre, and J. Polanco. 2002. Spatial distribution of adult Anopheles darlingi and A. albimanus in relation to riparian habitats in Belize, Central America. Journal of Vecto r Ecology 27 :21 30. Roberts, D. R., J. Paris, S. Manguin, R. Harbach, R. Woddruff, E. Rejmnkov, J. Polanco, B. Wullschleger, and L. Legters. 1996. Predictions of malaria vector distributions in Belize based on multispectral satellite data. American Journ al of Tropical Medicine and Hygeine 54 :302 308. Rosenberg, D. E., and T. P. Marcotte. 2005. Land use system modeling and analysis of shaded cacao production in Belize. Agroforestry Systems 64 :117 129. Sabo, J. L., R. Sponseller, M. Dixon, K. Gade, T. Harms J. Heffernan, A. Jani, G. Katz, C. Soykan, J. Watts, and J. Welter. 2005. Riparian zones increase regional species richness by harboring different, not more, species. Ecology 86 :56 62. Sader, S. A., C. Reining, T. Sever, and C. Soza. 1997. Human Migratio n and Agricultural Expansion: A Threat to the Maya Tropical Forests. Journal of Forestry 95 :27 32. Sala, O. E., F. S. Chapin, J. J. Armesto, E. Berlow, J. Bloomfield, R. Dirzo, E. Huber Sanwald, L. F. Heuenneke, R. B. Jackson, A. Kinzig, R. Leemans, D. M. Lodge, H. A. Mooney, M. Oesterheld, N. L. Poff, M. T. Sykes, B. W. Walker, M. Walker, and D. H. Wall. 2000. Global biodiversity scenarios for the year 2010. Science 287 :1770 1774. Salas, O., and J. Meerman. 2007. Maya Mountain Marine Corridor Conservation Action Strategy. The Nature Conservancy Belize program.

PAGE 144

144 Salcedo, I. H., H. Tiessen, and E. V. S. B. Sampaio. 1997. Nutrient availability in soil samples from shifting cultivation sites in the semi arid Caatinga of NE Brazil. Agriculture, Ecosystems and E nvironment 65 :177 186. Sanchez, P. A. 1982. Nitrogen in shifting cultivation systems of Latin America. Plant and Soil 67 :91 103. Saunders, D. L., J. J. Meeuwig, and C. J. Vincent. 2002. Freshwater protected areas: strategies for conservation. Conservation Biology 16 :30 41. Saunders, T. J., M. E. McClain, and C. A. Llerena. 2006. The biogeochemistry of dissolved nitrogen, phosphorus, and organic carbon along terrestrial aquatic flowpaths of a montane headwater catchment in the Peruvian Amazon. Hydrological P rocesses 20 :2549 2562. Scheuerell, M. D., and D. E. Schindler. 2004. Changes in the spatial distribution of fishes in lakes along a residential development gradient. Ecosystems 7 :98 106. Schwartz, N. 1990. Forest society: A social histroy of Peten, Guatema la. University of Pennsylvania Press, Philadelphia. Seubert, C. E., P. A. Sanchez, and C. Valverde. 1977. Effects of land clearing methods on soil properties of an ultisol and crop performance in the Amazon jungle of Peru. Tropical Agriculture (Trinidad) 5 4 :307 321. Shriar, A. J. 2001. The dynamics of agricultural intensification and resource conservation in the buffer zone of the Maya Biosphere Reserve, Peten, Guatemala. Human Ecology 29 :27 48. Six, J., and J. D. Jastrow. 2002. Organic matter turnover. Pag es 936 942 in R. Lal, editor. Encyclopedia of Soil Science. Taylor and Francis, London. Sommer, R., P. L. G. Vlek, T. D. d. A. S, K. Vielhauer, R. d. F. R. Coelho, and H. Flster. 2004. Nutrient balance of shifting cultivation by burning or mulching in th e Eastern Amazon evidence of subsoil nutrient accumulation. Nutrient cycling in agroecosystems 68 :257 271. Steffan Dewenter, I., M. Kessler, J. Barkmann, M. M. Bos, D. Buchori, S. Erasmi, H. Faust, G. Gerold, K. Glenk, S. R. Gradstein, E. Guhardja, M. Ha rteveld, D. Hertel, P. Hohn, M. Kappas, S. Kohler, C. Leuschner, M. Maertens, R. Marggraf, S. Mige Kleian, J. Mogea, R. Pitopang, M. Schaefer, S. Schwarze, S. G. Sporn, A. Steingrebe, S. S. Tjitrosoedirdjo, S. Tjitrosoemito, A. Twele, R. Weber, L. Woltmann M. Zeller, and T. Tscharntke. 2007. Tradeoffs between income, biodiversity, and ecosystem functioning during tropical rainforest conversion and agroforestry intensification. Proceedings of the National Academy of Sciences 104 :4973 4978.

PAGE 145

145 Thomas, S. M., C. Neill, L. A. Deegan, A. V. Krusche, V. M. Ballester, and R. L. Victoria. 2004. Influences of land use and stream size on particulate and dissolved materials in a small Amazonian stream network. Biogeochemistry 68 :135 151. TIDE. 2003. The Freshwater Initia tive: River Assessment Project. Toledo Institute for Development and Environment. Tiessen, H., I. H. Salcedo, and E. V. S. B. Sampaio. 1992. Nutrient and soil organic matter dynamics under shifting cultivation in semi arid northeastern Brazil. Agriculture, Ecosystems and Environment 39 :139 151. TNC, editor. 2000. The Five S Framework for Site Conservation, 2nd edition. The Nature Conservancy, Arlington. Toledo, V. M., B. Ortiz Espejel, L. Corts, P. Moguel, and M. d. J. Ordoez. 2003. The multiple use of tr opical forests by indigenous peoples in Mexico: a case of adaptive management. Conservation Ecology 7 :9. Trumbore, S. E., E. A. Davidson, P. B. d. Camargo, D. C. Nepstad, and L. A. Martinelli. 1995. Belowground cycling of carbon in forests and pastures of Eastern Amazonia. Global Biogeochemical Cycles 9 :515 528. Turner, B. L., S. C. Villar, D. Foster, J. Geoghegan, E. Keys, P. Klepeis, D. Lawrence, P. M. Mendoza, S. Manson, Y. Ogneva Himmelberger, A. B. Plotkin, D. P. Salicrup, R. R. Chowdhury, B. Savitsky, L. Schneider, B. Schmook, and C. Vance. 2001. Deforestation in the southern Yucatn peninsular region: an integrative approach. Forest Ecology and Management 154 :353 370. Uhl, C. 1987. Factors controlling succession following slash and burn agriculture in Amazonia. Journal of Ecology 75 :377 407. Uhl, C., and C. F. Jordan. 1984. Succession and nutrient dynamics following forest cutting and burning in Amazonia. Ecology 65 :1476 1490. USDA. 2004. Soil Survey Laboratory Methods Manual. USDA Natural Resources Co nservation Service, Lincoln, NE. USEPA. 2008. List of Drinking Water Contaminants and Their MCLs ( http://www.epa.gov/ogwdw/contaminants/index.html) in United States Environmental Protectio n Agency. van Wendel de Joode, B., I. A. M. de Graaf, C. Wesseling, and H. Kromhout. 1996. Paraquat exposure of knapsack spray operators on banana plantations in Costa Rica. International Journal of Occupational and Environmental Health 2 :294 304. Vance, C ., and J. Geoghegan. 2002. Temporal and spatial modeling of tropical deforestation: a survival analysis linking satellite and household survey data. Agricultural Economics 27 :317 332.

PAGE 146

146 Veldkamp, E. 1994. Organic carbon turnover in three tropical soils under pasture after deforestation. Soil Science Society of America Journal 58 :175 180. Vitousek, P. M., and R. L. Sanford. 1986. Nutrient cycling in moist tropical forest. Annual review of ecology and systematics 17 :137 167. Weins, J. A. 2002. Riverine landscap es: taking landscape ecology into the water. Freshwater Biology 47 :501 515. Wesseling, C., A. Aragn, L. Castillo, M. Corriols, F. Chaverri, E. d. l. Cruz, M. Keifer, P. Monge, T. J. Partanen, C. Ruepert, and B. v. W. d. Joode. 2001a. Hazardous Pesticides in Central America. International Journal of Occupational and Environmental Health 7 :287 294. Wesseling, C., M. Corriols, and V. Bravo. 2005. Acute pesticide poisoning and pesticide registration in Central America. Toxicology and Applied Pharmacology 207 :S 697 S705. Wesseling, C., B. v. W. d. Joode, and P. Monge. 2001b. Pesticide related illness and injuries among banana workers in Costa Rica: a comparison between 1993 and 1996. International Journal of Occupational and Environmental Health 7 :90 97. Wesselin g, C., B. van Wendel de Joode, C. Ruepert, C. Len, P. Monge, H. Hermosillo, and T. J. Partanen. 2001c. Paraquat in developing countries. International Journal of Occupational and Environmental Health 7 :275 286. Whitaker, M. J. 1993. The challenge of pesti cide education and training for tropical smallholders. International Journal of Pest Management 39 :117 125. Wiesenberg, G. L. B., J. Schwarzbauer, M. W. I. Schmidt, and L. Schwark. 2004. Source and turnover or organic matter in agricultural soils derived f rom n alkane/n carboxylic acid compositions and C isotope signatures. Organic Geochemistry 35 :1371 1393. Wilk, R. R. 1981. Agriculture, Ecology and Domestic Organization Among the Kekchi Maya. The University of Arizona. Wilk, R. R. 1985. Dry season agricul ture among the Kekchi Maya and its implications for Prehistory. Pages 47 57 in M. Pohl, editor. Prehistoric Lowland Maya Environment and Subsistence Ecology. Papers of the Peabody Museum, Cambridge. Wilk, R. R. 1997. Household ecology: economic change and domestic life among the Kekchi Maya in Belize. Northern Illinois University Press, Dekalb, IL. Williams, M. R., T. R. Fisher, and J. M. Melack. 1997. Solute dynamics insoil water and groundwater in a central Amazon catchment undergoing deforestation. Bioge ochemistry 38 :303 335.

PAGE 147

147 Williams, M. R., and J. M. Melack. 1997. Solute export from forested and partially deforested chatchments in the central Amazon. Biogeochemistry 38 :67 102. Wright, A. C. S., D. H. Romney, and R. H. Arbuckle. 1959. Land in British Hon duras, Report of the Land Use Survey Team. Colonial Research Publication (24) HMSO. 327pp & maps. London. Wu, T. H., T. R. Rainwater, S. G. Platt, S. T. McMurry, and T. A. Anderson. 2000. Organochlorine contaminants in Morelet's crocodile ( Crocodylus mor eletii ) eggs from Belize. Chemosphere 40 :671 678. Zahm, S. H., D. D. Weisenburger, P. A. Babbit, R. C. Sall, J. B. Vaught, K. P. Cantor, and A. Blair. 1990. A case control study of non Hodgkin's lymphoma and the herbicide 2,4 dichlorophenoxyacetic acid (2, 4 D) in eastern Nebrask. Epidemiology 1 :349 356. Zarin, D. J., E. A. Davidson, E. Brondizio, I. C. G. Viera, T. S, T. Fedpausch, E. A. G. Schuur, R. Mesquita, E. Moran, P. Delamonica, M. J. Ducey, G. C. Hurtt, C. Salimon, and M. Denich. 2005. Legacy of fi re slows carbon accumulation in Amazonian forest regrowth. Frontiers in Ecology and the Environment 3 :365 369.

PAGE 148

148 BIOGRAPHICAL SKETCH David Buck is from Charlotte, North Carolina and attended the University of North Carolina at Chapel Hill from 1992 to 199 American Studies and a minor in geology. Between undergraduate and graduate school, David worked as the field station manager for Monkey Bay Wildlife Sanctuary. At Monkey Bay, he was responsible for facilit ating international study abroad programs that focused on the natural and cultural history of the Maya Forest region. of Natural Resources and Environment (SNRE) and gradua ted in 2004 with a degree in paleolimnology of hypersaline Lago Enriquillo, Dominican Republic. In 2004 David started in the PhD program in SNRE as an NSF IGERT fellow in the W orking Forests in the Tropics program. He conducted his PhD research in the Temash River watershed of southern Belize, Central America. He received his PhD from the University of Florida in the summer of 2012. David and his wife Ellie Harrison Buck have two children and currently live in Durham, NH, where his wife is an assistant professor of archaeology at the University of New Hampshire. David works for the Biodiversity Research Institute (BRI) in Gorham, s tropical research program.