!"#$

#%&""

  '
   
 '
 
  
(
#)&

      
  
  
 
  
 *


  
   


  '
 
 




 '




  


* 
   &
+
       
  
 
* 
  

  

 
  

    
 '

*    





  
  

* 
  &
,

  

 


- + 
.)&%$/'
%0&"$/12'

  



3
 

  
,4 '
* 



 
 
   * 

 
 &

  
 
 
 
 

 
  

-
+ 
*
  
   5
 
.   '
 
   '

  
   2

  
 
.#)6

!#"2&
,
-
+   
*

  


   
   
 
1
 

 

 
 
 




 

  

 
  
  &
,
* 
*
5 

   
.#7$%8!"#02

5  

5 
  


- 
  
  &
  
  

    
* 
  


5 
*
 
 
9
 : &
,
 
 
 
*
  
    
   '
    
  '


 

 



  
  &


* ' 
-
+ 
 
  

;
  



 

 <
 * '


  
  

=
  

  
   &
,
 

 
 

1
 
* 
 
 

 *

  

+ '


  



5 



  


  

#)6
 
 &
>
  


5 
 
 
  


 4
*
  
    


 


  
 
 ;


 
#7 " &
,


* 
 

+ '

       

 
 

 
-
+
*
 &
 
 

*

  

    
  




 
4 
* 
 
  


 
 

- 
9


1  &
,
 





-
+ 


 
  

  

  
 
  



  

* 
  


 &

  


 

 

 

 
   
   
     
  

  
!"#$  ?@@ && @  A B ?? ? ? #)0)#0

9+C
7$D7#$ )7 D"" 9+C
7$D7#$ )7 D#$ 9C
# 0%$!##

!  
"
# 
$  !

  
  '
#"
7#
  

LINKING LAKE VARIABILITY, CLIMATE, AND HUMAN ACTIVITY IN BASOTU, TANZANIA

Lindsey Higgins

Linking lake variability, climate, and human activity in Basotu, Tanzania

Lindsey Higgins ©Lindsey Higgins, Stockholm University 2017

ISBN print 978-91-7649-680-0 ISBN PDF 978-91-7649-681-7 ISSN 1653-7211

Paper I ©Lindsey Higgins Paper II ©Open Access. First published in Geosciences. Paper III ©Lindsey Higgins

Cover Illustration by Lindsey Higgins Author Photo by Emanuel Ramsell Printed in Sweden by Universitetsservice US-AB, Stockholm 2017 Distributor: Department of Physical Geography, Stockholm University In memory of Lennon Bowie Zappa Erik Christoffer Wallgren (1990–2014)

Abstract

Paleoenvironmental investigations establish important baseline knowledge of the natural variability of lake systems, to better understand human impacts on the landscape, and the effects of climate change on water resources. By combin- ing long-term environmental history with investigations into modern land use patterns and climatological events, a wider perspective can be reached that has practical applications in water governance. This thesis presents a case study of Lake Basotu (4.37°S, 35.07°E), a crater lake in the Hanang district of north- central Tanzania, which acts as an important source of freshwater for local peo- ple. A three-meter long sediment core from an interior crater of Lake Basotu was investigated using proxy records (diatoms, magnetic parameters, and carbon content) and radiometric dating (14C and 210Pb). The Lake Basotu record was then compared to other sediment-based reconstructions from East Africa and records of historical famines to better place it into the timeline and understanding of regional climate dynamics. This work was extended into modern times (1973– 2015) by examining lake extent variations in the Landsat satellite archive. Shore- line boundaries for dry-season images were delineated and lake extent was cal- culated using GIS techniques. This remote sensing record was compared to cli- matological patterns, meteorological records, and the history of land-use changes in the surrounding district. As a whole, the Lake Basotu record indicates that major fluctuations in lake level are not abnormal; however, human influence has likely increased the lake’s sensitivity to climatic fluctuations. The timing of historical famines in East Africa were linked to periods of shallow lake condi- tions in Basotu, and the duration of the most extreme lake level changes correlate to a reversal in the 14C age-depth model. Recent variations in lake extent are likely connected to a mechanized wheat farming program implemented in the district as a foreign aid project in the early 1960s. To support the work done in Basotu, a preliminary investigation of sediment from the nearby Lake Babati was undertaken. Sediment from the two lakes indicates that their geographical location may be in a transition zone towards dryer conditions to the south during the Little Ice Age in East Africa. The results of this thesis support that Lake Basotu is an important location for understanding the potential impacts of cli- mate change and human activity on water resources in this region. Sammanfattning

Denna avhandling presenterar en studie av Lake Basotu (4.37° S, 35.07° Ö), en kratersjö i Hanang-distriktet i norra Tanzania som är en viktig sötvattenskälla för lokalbefolkningen. Sedan 1960-talet har konflikter uppstått över tillgången till vatten och användningen av marken kring sjön. För att bäst skilja effekterna av mänsklig aktivitet på Basotu-sjön från naturlig variation, tillämpas i avhandlingen en tvärvetenskaplig metod. Genom att kombinera långsiktig miljöhistoria med studier av nutida markanvändningsmönster och klimatologiska händelser nås en bred förståelse av sjön som vattenresurs, av praktisk betydelse för vattenförvaltning (governance). Strategier för bevarande av vattenresurser är särskilt viktiga, eftersom nederbörden i Tanzania förväntas bli alltmer opålitlig på grund av prognosticerade klimatförändringar. Basotu-sjöns naturliga variabilitet bedömdes genom analys av sediment från sjöns inre krater. En sedimentkärna från Basotu-sjön analyserades med avseende på diatoméer. Olika diatoméarter har olika preferenser gällande salthalt. Detta innebär att sedimentkärnan kunde delas in i åtta zoner, som var och en beskrevs med avseende på den generella miljön, baserat på diatoméerna som miljöindikatorer. Analys av mineralmagnetiska egenskaper och kolinnehåll användes tillsammans med diatomédata för att ge ett bredare perspektiv på sjönivåförändringar. Samtliga data användes för att utveckla en kronologi baserad på klimatförändringar som registrerats i andra studier av sediment från Östafrika och på historiska uppgifter om perioder av hungersnöd som har påverkat regionen. Analyserna tyder på att Basotu-sjön är känslig för torra förhållanden med potential att torka ut till en våtmark. Moderna förändringar i Basotu-sjöns utbredning registrerades genom manuell avgränsning av strandlinjer i Landsat-registreringar från 1973 till 2015. Förändringar mellan registreringsåren jämfördes med lokala nederbördsdata samt med ENSO- och IOD-index. Resultaten indikerar att Basotu-sjön har blivit mer känslig för extrem nederbörd sedan ca 1996. Översvämningar under de senaste åren föregicks av ett försök att implementera mekaniserad veteodling i området kring Basotu-sjön, och är en sannolik effekt av en ökad sedimenttransport till Basotu-sjön på grund av otillräckliga markvårdsåtgärder. För att fastställa hur markanvändningsaktiviteter kan ha påverkat sjöns ekologi jämfördes en markanvändningskronologi som går tillbaka till början av 1960- talet med de moderna sjönivåvariationerna. Telefonintervjuer genomfördes med två lokala informanter för att validera information i tidningsartiklar och rapporter från biståndsorganisationer. Den areella ökningen av veteodling i Hanang- distriktet uppvisade en negativ korrelation med sjöstorleken, sannolikt ett resultat av regnbevattning av veteodlingarna, vilket har minskat ytavrinningen och därmed tillskottet av vatten till sjön. Studien av Basotu-sjön indikerar att stora svängningar i sjöns vattennivå inte är onormala. Emellertid har människans miljöpåverkan sedan 1960-talet sannolikt ökat sjöns känslighet mot fluktuationer i nederbörd. Arbetet som presenteras i denna avhandling bidrar till att bättre förstå hur Basotu-sjön reagerar på såväl naturliga som antropogena processer, och ger riktlinjer för framtida forskning i liknande miljöer. Vidare presenteras belägg som betonar att Basotu-sjön är viktig för förståelsen av potentiella effekter av klimatförändringar och mänsklig aktivitet på vattenresurser i regionen.

Thesis Content

This thesis consists of three papers and a comprehensive summary. The three papers are referred to as papers I–III and include: one published, one in review, and one manuscript. The results of the three papers are summarized in chapter 4 of the comprehensive summary.

I Higgins, L., Westerberg, L. O., Risberg, J., Snowball, I. & Andersen, T.J. (2016). Sediment Based Reconstruction of Environmental Change in Basotu, Tanzania. Manuscript.

II Higgins, L., Koutsouris, A. J., Westerberg, L. O., & Risberg, J. (2016). Surface Area Variability of a North-Central Tanzanian Crater Lake. Geosciences, 6, 27.

III Higgins, L. & Caretta, M.A. (2016) Lake Extent Changes in Basotu, Tanzania: A Mixed Methods Approach to Understanding the Impacts of Anthropogenic Influence and Climate Variability. Accepted in current form by Landscape Research.

Author Contributions

Contributions from each author are divided as follows:

I L.H.: Participated in fieldwork, contributed to design of project, performed analysis, primary author of text, constructed figures and maps. L-O.W: Participated in fieldwork, contributed to design of project, consulted on analysis of data and editing of text. J.R.: Participated in fieldwork, contributed to design of project, consulted on analysis of data and editing of text. I.S.: Completed mineral magnetic measurements and was consulted in the analysis. Also contributed to methods section. T.J.A.: Completed lead-210 dating and was consulted in the analysis. Also contributed to methods section.

II L.H.: Conceived idea, designed project, performed analysis, primary author of text, constructed figures and maps. A.J.K.: Contributed data and text for methods section. Consulted on mapping of watershed. L-O.W.: Editing of text. J.R.: Editing of text.

III L.H.: Conceived idea, collected data, performed analysis, primary author of text, constructed figures and maps. M.A.C.: Performed telephone interviews and contributed to text

Contents

1 Introduction ...... 1 1.1 Objectives ...... 5 1.2 Research Strategy ...... 6

2 Study Area ...... 8 2.1 Climate ...... 9 2.2 Geology ...... 10 2.3 Soils ...... 10 2.4 Hydrology ...... 12 2.5 Lake Geochemistry ...... 13 2.6 Aquatic and Terrestrial Ecosystem ...... 14

3 Material and Methods ...... 15 3.1 Paper I ...... 15 3.1.1 Fieldwork and Sampling Procedure ...... 15 3.1.2 Diatoms ...... 16 3.1.3 Mineral Magnetic Measurements ...... 18 3.1.4 Carbon Content ...... 21 3.1.5 Radiometric Dating ...... 22 3.2 Paper II ...... 24 3.2.1 Remote Sensing and GIS Data ...... 24 3.2.2 Meteorological and Climatological Data ...... 25 3.3 Paper III ...... 26

4 Results ...... 27 4.1 Paper I ...... 27 4.2 Paper II ...... 34 4.3 Paper III ...... 38

5 Work in Progress ...... 40

6 Discussion ...... 47

7 Conclusions ...... 52

8 Acknowledgments ...... 54

9 References ...... 56

10 Appendices ...... 69 10.1 Appendix A: Lake Basotu Diatoms ...... 69 10.2 Appendix B: Reports obtained regarding wheat in the Hanang District ...... 71 10.3 Appendix C: Lake Babati Diatoms ...... 73

Abbreviations

AF alternating field AMS accelerator mass spectrometry ARM anhysteretic remanent magnetizations ASTER Advanced Spaceborne Thermal Emission and Reflection Radiometer BCR remanent coercivity bgl below ground level BGS British Geological Survey CIDA Canadian International Development Agency DC direct bias DEM digital elevation model DMI Dipole Mode Index EIA environmental impact assessment EM-DAT international disaster database ENSO El Niño-Southern Oscillation ESEM environmental scanning electron microscope ESRL Earth System Research Laboratory FAO the Food and Agriculture Organization GDPs global precipitation data sets GPCC Global Precipitation and Climatology Center v6 IOD Indian Ocean Dipole IPCC Intergovernmental Panel on Climate Change ITCZ intertropical convergence zone K volume magnetic susceptibility keV kiloelectron volt LIA Little Ice Age LOI loss on ignition MD multi-domain MEI Multivariate ENSO Index mT milliTesla NAFCO National Food and Agricultural Corporation NASA National Aeronautics and Space Administration NCCCS National Climate Change Communication Strategy NCCS National Climate Change Strategy NCEP-NCAR National Centers for Environmental Prediction-National Center for Atmospheric Research NERC Natural Environment Research Council NGA National Geospatial-Intelligence Agency NHMM non-homogeneous hidden Markov model NIR near-infrared NOAA National Oceanic and Atmospheric Administration PDB Pee Dee Belemnite P-E precipitation-evaporation PSD pseudo-single domain REAL Resilience in East African Landscapes SD single-domain SEI Stockholm Environmental Institute SHcal13 Southern Hemisphere calibration curve SIRM saturation isothermal remanence magnetization SP superparamagnetic SRTM Shuttle Radar Topography Mission SSA Sub-Saharan Africa SDGs Sustainable Development Goals SST sea surface temperatures SWIR shortwave-infrared TCWP Tanzania Canada Wheat Project TDS total dissolved solids TIC total inorganic carbon TMA Tanzania Meteorological Agency TOC total organic carbon UN United Nations UN-Water United Nations Water USDA United States Department of Agriculture USGS United States Geological Survey ı mass specific magnetization Ȥ mass specific magnetic susceptibility

1 Introduction

A long-time struggle in climate science has been the difference in opinion between what scientists and stakeholders deem useful information. As climate change impacts are far-reaching, it is more important than ever to bridge that gap and provide more usable information for adaptation and mitigation strategies. In 2015, 193 member states signed on to the United Nations (UN) Sustainable Development Goals (SDGs), which proposed 17 goals for meeting “the needs of the present without compromising the ability of future generations to meet their own needs” (UN, 2016). While these are not legally binding, they represent the acknowledgment that action is necessary. The concept of adaptation was nearly abandoned by political ecologists in the 1970s due to critiques of theoretical deficiencies by political economists claiming that the framework did not address the “social structural causes of vulnerability” (Bassett & Fogelman, 2013, p 45). However, since publication of the assessment reports of the Intergovernmental Panel on Climate Change (IPCC), adaptation has seen a resurgence in climate change literature (Bassett & Fogelman, 2013). Adaptation research is challenged by the need for designs and implementations to be context-specific as well as relevant to those in policy and decision making positions (Juhola & Kruse, 2015), thus bridging the gap between scientist and stakeholder. Adopting an approach, which utilizes multiple methods and disciplinary tools, speaks to the interconnectedness of the physical and social dimensions involved in climate change (Shaman et al., 2013). Nonetheless, adaptation strategies cannot be addressed before vulnerabilities are explored and risks are conceptualized (Füssel, 2007). A growing collaboration between traditional climate scientists working with paleoproxy records and archaeologists has emerged in recent years. The archaeological record is relevant to modern resilience, as it is an indicator of flexibility and adaptability in the face of a changing climate (Rockman, 2011). The Peloponnese region of Greece is one example of an area where there is a long comparison between multiple-proxy reconstructions and archaeological records, which focuses on how ancient societies responded to climate change (Weiberg et al. 2016). Similar work has been done in East Africa by the Resilience in East African Landscapes project (REAL); a Marie Curie funded partnership of seven research institutes bringing together multiple disciplines to understand the linkages between environmental change and social dynamics (www.real-project.eu). Separating natural variability from that which is anthropogenically induced is integral to planning efforts and assessment of the resources, as few ecosystems on Earth have been unaffected by human-driven processes (P.J. Lane, 2010). By understanding the dynamics of paleoenvironments we can better grasp how past societies were affected by and responded to fluctuations in climate, as well as make inferences as to the sustainability of their practices (P.J. Lane, 2015a). Taking this one-step further, the combination of paleoecological reconstructions

1 with modern observations allows assessment of the sustainability of practices associated with individual resources and fills gaps in data occuring due to the absence of physical monitoring. Dearing et al. (2010) state that:

A framework for studying land systems would ideally encompass the whole range of relevant temporal (and spatial) scales from the very long, multimillennial to the very short (annual, seasonal or even shorter), even though in practice the relevance of any timescale cannot normally be known a priori (p 21).

Whether past changes are appropriate analogues for present and future change is often debated, but it is not possible to fully understand socioenvironmental changes of today without looking to the past (Dearing et al., 2010; Stump, 2010). Throughout human history, settlements have developed alongside water resources such as seashores, rivers, and lakes. As lakes are sensitive to both human influence and climatic fluctuations, they are a suitable medium for paleoenvironmental research. In particular, small, closed-basin lakes are desirable due to their hydrologic simplicity (Verschuren, 1999; Wolin & Stone, 2010; Ryves et al., 2011). Methods of environmental reconstruction based on lake sediment allow us to ascertain the natural variability of a lake system and when it may have become stressed (Smol, 1992). Embedded in lake sediment is an archive of lake history in the form of physical and chemical signatures, which act as a proxy for a variety of information regarding the time of deposition (Smol, 1992; Cohen, 2003). These signatures include biological assemblages in the form of pollen, insects, macrofossils, and diatoms (microscopic silica-based algae). Physical parameters include variations with depth in grain size as well as the general appearance of the sediment in colour and texture. Chemical changes are noted in elemental and isotopic variations. Using these various proxies, paleolimnological reconstructions can determine the timing of events of anthropogenic origin, such as cultural eutrophication (Otu et al., 2011) and deforestation (Kiage & Liu, 2009). As a basic principle in the hydrology of lakes is that lake level rises during wet periods and falls during dry periods, and it is also possible to identify the timing of climatic events (Wolin & Stone, 2010). Employing a method that combines multiple proxies helps to balance out the strengths and weaknesses of the respective methods of analyses, as long as the interpreter avoids the tendency to make the data “fit” together (Birks & Birks, 2006). A wider understanding of the availability, use, and health of freshwater resources is important as water scarcity is predicted to affect two-thirds of the world’s population by 2025 (UNDESA, 2016). Of the continents, Africa has the second smallest percentage of the world’s freshwater resources at 9%, followed only by Australia (FAO, 2003). In particular, Sub-Saharan Africa (SSA) has more water-stressed countries than any other region (UNEP, 2010; UNDESA, 2016). In addition to being water-stressed, water access in SSA is deterred further as it is currently the poorest and least developed region in the world and lacks the economic resources to meet human demands (WWAP, 2012). Population growth in SSA is one of the highest in the world, with child mortality rates decreasing and fertility rates slow to follow (Canning et al., 2015). Within SSA, Tanzania is expected to experience the largest increase in population, potentially reaching 200 million people by 2060, a growth of ~340% (Canning et al., 2015).

2 The Food and Agriculture Organization (FAO) of the United Nations (UN) has determined 1000 m3 of water per year per inhabitant to be the minimum to “sustain life and ensure agricultural production” in irrigation dependent countries (FAO, 2003 p. 21). Tanzania currently falls at 2591 m3 per year per inhabitant, but as this value is calculated per capita and based on the estimate of the amount of water available to the country after determining the quantity of reserved flows, it is in danger of falling (FAO, 2003). Understanding climatic variability in Tanzania therefore has widespread implications for its food and water security, especially as the UN Millennium Development Goals Indicators have reported that 32% of the population of c. 53 million are undernourished (UN, 2015). Food security in Tanzania is predicted to further decline with climate change as production capacity becomes increasingly limited by temperature rise and water scarcity, putting additional stress on people and resources (Arndt et al., 2012). Annual precipitation potentially represents up to 50% of evapotranspiration demands in semi-arid regions, meaning that a smaller proportion of fallen precipitation is actually available to plants (Stewart & Peterson, 2015). Due to increasingly unreliable precipitation, historically rain fed agricultural projects will need to explore water supplementation by irrigation, and water extraction from basins is expected to increase both for domestic and agricultural purposes (Lecoutere, 2011). The 2012 United Nations Water (UN-Water) country brief on Tanzania lists “inadequate hydrological data and information” as a contributing factor to mismanagement of water resources (p 5). Lyakurwa et al. (2014) have made the seriousness of such data deficiency clear in a study exploring the levels of water scarcity in the nine major water basins of Tanzania. Water scarcity levels in six of the nine basins were determined to be extreme, meaning low replacement related to extraction of freshwater. This was determined using the quantity of freshwater available, as well as its various uses, and a model accounting for the required amounts of maintenance of ecological systems. In semi-arid regions of the world, like Tanzania, the amount of rainfall is not the limiting factor in agricultural productivity, but rather the unreliability of that rainfall (Rockström et al., 2010). There is a pattern of variability within the sea surface temperature (SST) of the western and eastern Indian Ocean, known as the Indian Ocean Dipole (IOD), which influences precipitation in East Africa. In the positive phase of the IOD, warmer than average SST in the western Indian Ocean leads to enhanced convection along the coast of Africa. This brings severe rainfall and enhanced easterly winds (Saji et al., 1999; Ummenhofer et al., 2009). During the next century, the western Indian Ocean is expected to experience enhanced warming, with 2015 demonstrating a historical high in its sea surface temperature (IPCC, 2013; Johnson & Parsons, 2016). This would indicate more rainfall over East Africa. However, a non-homogeneous hidden Markov model (NHMM) for projecting 21st century precipitation in Tanzania has shown that annual precipitation is expected to decrease, while the intensity of rainfall will increase (Cioffi et al., 2016). This projection was based on 40 years of daily rainfall data and re-analysis data from the National Centres for Environmental Prediction-National Centre for Atmospheric Research (NCEP-NCAR). This trend towards more intense rainfall has already been noted. Between 1990 and 2014, the frequency of flood disasters, obtained from the international disaster database (EM-DAT), has increased across Africa as a whole, with losses

3 concentrated in the eastern region and Tanzania listed as a “flood-prone” country (Li et al., 2016). In April 2016, Tanzanian news reported flash flooding in northern Tanzania that was responsible for livestock deaths and destruction of crops and food storage (Xinhua, 2016). The unusually heavy rainfall over East Africa in 2015 and 2016 was most likely connected to the strong 2015–2016 El Niño (Becker, 2016). In the Hanang district, 900 acres of farmland were washed away, with deforestation and land-clearing for agricultural purposes suggested to be responsible (Xinhua, 2016). This shows how climate extremes and land use may interact in causing severe effects on the society. In a spatio-temporal analysis of the distribution of flood disasters across Africa, the frequency of flooding was significantly positively correlated with the rate of deforestation; meaning that as forest coverage is reduced, the rates of flooding increase (Li et al., 2016). Due to land degradation, severe precipitation will likely continue leading to flooding rather than effective recharge of groundwater, since degraded and compacted soils are unable to absorb the rainfall. Babati, a town in the Manyara Region of Tanzania, has experienced severe flooding in recent history due to its proximity to Lake Babati, which is highly responsive to precipitation (National Audit Office, 2012). A national disaster was declared in 1990 when lake levels rose so significantly that an outlet was created, running directly through the town (Gerdén et al., 1992). An environmental impact assessment (EIA) on the 1990 flood in Babati was conducted the following year, using climatological data and aerial photos from 1954, 1960, 1970, and 1990 (Strömquist, 1992). According to this EIA, flooding of Babati was most likely due to variability in rainfall rather than land-use changes, although the common opinion beforehand was that land clearing had increased run-off and sedimentation of the lake. This conclusion was made based on a 1964 flood, which occurred before substantial changes to the surrounding vegetation had taken place. A subsequent study (Sandström, 1995) determined that although the floods in 1964 and 1990 were triggered by heavy rainfall, the main driving mechanism was landscape change. The 1991 EIA did acknowledge that the current state of cultivation surrounding Lake Babati favours infilling of the lake due to erosion and that an analysis of past landscape changes would be necessary to better explain recent impacts (Strömquist, 1992). The flooding situation in Babati is an example of a situation where the modern understanding of lake dynamics could benefit from a paleolimnological investigation. Understanding the causes of flooding events in small lakes could mean the difference between substantial damage to environment and infrastructure, and the potential restoration of ecosystems. Holmgren and Scheffer (2001) suggested that precipitation related to the El Niño-Southern Oscillation (ENSO) might offer the potential to restore degraded ecosystems in arid and semi-arid regions of the world by inducing vegetation. In opposition, investigation of flooding in four Kenyan lakes (Obando et al., 2016) and the 2016 loss of crops in the Hanang district of Tanzania (Xinhua, 2016) have shown wide-spread negative impacts. By understanding the overall health and status of small lake ecosystems, better decision-making processes can be undertaken to ensure livelihood security for those dependent on their resources. Of equal importance to flood preparedness is the understanding and proper response to droughts. Agriculture in Tanzania accounted for 32% of its 2014 gross domestic product (down from 47% in 1995) and consequently, much of

4 the labour force is dependent on this sector (World Bank, 2016). As food security is declining with climate change, one of the primary unknowns is how to increase food security while not acting at the expense of other resources and exacerbating environmental degradation. Greater management efforts have been signalled in order to ensure that the agricultural sector does not deprive other users of water, and that safe freshwater levels are maintained in Tanzania (Lyakurwa et al., 2014). In 2012, the National Climate Change Strategy (NCCS) for Tanzania was launched. The NCCS was designed to help Tanzania address adaptation strategies in line with sustainable development and with a particular focus on the agricultural sector (Daly et al., 2015; FAO, 2015). Also in 2012, the National Climate Change Communication Strategy (NCCCS) was developed for Tanzania under the Vice President’s office in an effort to increase public awareness of climate change (Daly et al., 2015). The NCCCS program runs through 2017 and has aimed to target various audiences from the international and national levels down to wards and villages, where community and religious leaders are encouraged to assist in disseminating information and engaging the public (United Republic of Tanzania, 2012). The IPCC’s 5th assessment report highlighted the concern to focus on what could also be the negative impacts of adaptation strategies, or maladaptation, defined as “actions, or inaction that may lead to increased risk of adverse climate-related outcomes, increased vulnerability to climate change, or diminished welfare, now or in the future” (Noble et al., 2014, p 857). It is more important than ever to adopt holistic approaches to understanding complex landscape dynamics, as climate change is creating great uncertainty regarding the future availability of freshwater. In the past, once resources diminished beyond the point of usability, people would abandon areas of occupation and migrate (McLeman, 2011). Now in the face of unprecedented population growth and increasing stress on the environment, it is integral that efforts are made to protect our contemporary resources (Lambin, 2005).

1.1 Objectives A wider perspective of environmental variability in the context of both climate change and land use makes the work directly relevant to parties involved in planning and policy positions (Lambin, 2005). Paul Lane, Professor of Global Archaeology at Uppsala University wrote:

To my mind, a climate change archaeology devoid of such considerations, that examines changes in pollen concentrations and the nature of the sedimentary record without a consideration of whose lives these changes impacted; that evokes climatic stress without considering differential patterns of consumption or access to resources; that identifies resilience cycles, and phases of exploitation, collapse and re-structuring without a consideration of relations of social power and authority in these processes, however valuable towards enhancing understanding of the past, does little to advance our understanding of how archaeological studies might make the lives of ordinary people today any better, or help safeguard the future of the planet (P.J. Lane, 2015b, p 11).

This sentiment can and should be applied to paleoclimatological research itself, especially in areas of known long-term occupation. This doctoral thesis focuses

5 on Lake Basotu, a small freshwater lake in Tanzania as a case study for examining ties between lacustrine variability, climate, and anthropogenic influence. To accomplish this, three objectives were identified, and appropriate research methods were selected with those in mind. Specific methodologies are explained in more detail in chapter 3. The mixed method of this dissertation allows answers to be derived from multiple perspectives, which ensure fewer gaps in the information.

Objective #1: Show how water level has fluctuated in Lake Basotu over the past 1000 years, to establish a baseline for natural variability.

Objective #2: Reconstruct a history of modern lake extent (1973–2015) and identify potential drivers of observed changes.

Objective #3: Show if and how land-use practices in the area surrounding Lake Basotu may have influenced its function as a freshwater resource.

1.2 Research Strategy It was necessary for the Lake Basotu research strategy to be flexible and adapt methods to fit this type of environment, and Figure 1 depicts the flow of ideas producing this thesis. The original goal was to complete a paleoclimatic reconstruction for the last 1000 years using sediment from Lake Basotu and nearby sites. The idea was that this work would contribute to the regional understanding of climate change and human adaptation to it. Lake Basotu was selected due to its strategic location and appearance as a permanent water body. This area may lie in a region of transition for climate regimes in this region. Spatial patterns of precipitation in East Africa during the Little Ice Age likely occur from east to west (Russell et al., 2007) and north to south (Brown and Johnson, 2005). After examinations made during fieldwork and information gleaned from the literature review, it became clear that there was a much larger story in Basotu. This also presented the opportunity to look at Lake Basotu using a variety of techniques, while still contributing to the general understanding of how climate and human activity affect small freshwater lakes in semi-arid environments.

6

2012, all rights reserved), and used by MacDonald et al. (2012), the estimate is that 1–10 m of groundwater can be found below the Basotu watershed. According to MacDonald et al. (2012), in regions with annual rainfall between 250 and 500 mm/year, without the presence of a perennial river, the groundwater table can be as far as 100 m below ground level (b.g.l.) but at least 25 m b.g.l. depending on basement rocks. The surficial depth of groundwater near perennial rivers is always estimated to be below 7 m b.g.l. due to a set of empirical rules (BGS, 2011). OpenStreetMap (2016) data for this area shows a steam draining into the larger extent of Lake Basotu from the NNW, yet in Landsat satellite images this stream appears ephemeral. No other permanent rivers or streams have been observed in Basotu (Okedi et al., 1975; Chale 1989; C.R. Lane, 1990). There is a single reference to a Dongobesh River by Fenger et al. (1986), reporting Lake Basotu as the terminus, but this river has not appeared in other reports or maps available to this study. OpenStreetMap is a community-based- mapping project maintained by contributors with local knowledge (www.openstreetmap.org). Map data from OpenStreetMap contributors are available under the Open Database Licence and obtained through the Tanzania GIS User Group (http://www.tzgisug.org/). A water quality study listed Lake Basotu at risk of becoming saline due to the lack of a surficial outlet (Okedi, et al. 1975). Additionally, a subsequent report noted that high salinity in Lake Basotu was responsible for stunted fish growth (Chale, 1989). In opposition, recent news articles have stated Lake Basotu as an important source of freshwater for local people (Arusha Times, 2011a, 2011b, 2011c; PESA Times, 2013; The Citizen, 2013). Generally speaking, it is possible for lakes to switch between closed and open over time depending on climate conditions (Cohen, 2003). Based on water samples collected in 1969, Lake Basotu was found to follow a combination of gradual and rapid solute evolution paths that indicate solute evolution tightly coupled to evaporative loss (gradual path) with sudden mixing events (rapid path) causing H2S loss to the atmosphere (Kilham & Cloke, 1990). These rapid mixing events are typically associated with anoxic waters. However, the current presence of small-scale commercial fishing as a source of livelihood, as observed during fieldwork, indicates the presence of oxygen in Lake Basotu.

2.5 Lake Geochemistry The available water chemistry data on Lake Basotu are from samples collected in 1969 by a team investigating the series of crater lakes to which Basotu belongs (Kilham & Cloke, 1990). Compared to the other lakes in the series, Lake Basotu was found to be low in most of the ions examined: Na+, Mg2+, Cl-, F-, S+, S-, and 2- SO4 . All of the 11 lakes sampled were found to be supersaturated in CaCO3. Alkalinity was also measured, and reported as 4.52 meq/L of HCO3 + CO3 (Kilham & Cloke, 1990). Alkalinity acts a buffer against decreasing pH, but also increases the availability of inorganic carbon for use by plants (Boyd, 2015). Low alkalinity may also be a cause of low planktonic productivity within the lake water, as it is often associated with low phosphorous and nitrogen (Moyle, 1949). Measurements of alkalinity over 100 meq/L are considered high, but can reach up to 1500 meq/L in the high salinity lakes of Africa (Talling & Talling,

13 1965). Low alkalinity in lakes can fall below 0.4 meq/L (Charles & Smol, 1988), accordingly Lake Basotu’s alkalinity of 4.52 meq/L is not low enough to affect in-lake productivity.

2.6 Aquatic and Terrestrial Ecosystem Lake Basotu is primarily a freshwater-lentic system. A wetland system is also present in the area surrounding the shoreline, though the extent is dependent on precipitation and changing lake levels. The current trophic state of Lake Basotu is unknown, but likely mesotrophic to eutrophic as the lake supports a biological community. A widespread survey of diatom communities from East Africa found the composition of Lake Basotu to be planktonic, predominantly composed of varieties of Melosira granulata (Gasse et al., 1983). Melosira granulata is now taxonomically classified as Aulacoseira granulata (Simonsen, 1979). The primary species of fish in Lake Basotu are Tilapia spp., T. esculenta and T. melanopleura, and unspecified Clarias (Okedi et al., 1975). T. esculenta primarily feeds on phytoplankton, especially diatoms of the Melosira (=Aulacoseira) taxa, both in bottom sediments and suspended in the water body (Greenwood, 1953). T. melanopleura are limited to shallow environments as they primarily feed on aquatic plants (Lowe, 1959). Stomachs of T. melanopleura have also been found to contain aquatic insect larvae; juveniles will feed on diatoms, water fleas, and other algae (Munro, 1967). FishBase (http://www.fishbase.org) lists 32 species of Clarias, air breathing catfish, in Africa. All Clarias are freshwater fish (Nelson et al., 2016). Lake Basotu is a permanent and migratory home to at least 35 species of waterbirds, suggesting that the lake is an important food resource (Fyumagwa et al., 2014). It is also a site of refuge when nearby larger lakes dry out. A family of 15-20 hippopotami occupied Lake Basotu in the early 1960s (Downie and Wilkenson, 1962), the current numbers are unknown but their occupancy was confirmed by local fishermen during fieldwork in 2013. Woodland in Basotu is primarily Acacia, although there was significant clearing in the 1950s due to a tsetse fly outbreak, with Acacia fever trees in the area directly surrounding Lake Basotu. The primary grasses present are bamboo (Pennisetum mezianum), Rhodes (Chloris gayana), and thatching (Hyparrhenia anthistirioides) (Fenger et al., 1986).

14 3 Material and Methods

As each research paper uses different investigative and analytical methods, this section is subdivided to address them individually.

3.1 Paper I

3.1.1 Fieldwork and Sampling Procedure In March 2013, two parallel sediment cores were extracted from the bottom of the interior crater of Lake Basotu using a Russian corer. A Russian corer (length: 50 cm, inner diameter: 5 cm) is a stainless steel manual instrument for coring composed of a half tube and cover plate attached by a hinge. When inserted into the sediment, the turning handle is rotated 180° clockwise until the edge of the tube reaches the cover plate, cutting off the portion of material within the half tube, and the corer is then removed from the sediment (EPA, 1999). Two canoes were rented from local fishermen and a platform was constructed between them to facilitate coring and stabilize the boats (Figure 8). Sediment was relatively soft until 250 cm depth where a compacted layer was reached. Once punctured, gas bubbles accompanied by cooler temperature water were brought to the surface. No sediment could be retrieved beyond 300 cm depth. The sediment cores were partially subsampled in the field at 1 cm intervals using standard cubic plastic boxes for ease of transport (internal volume 7 cm3). The longer of the two cores (300 cm), core 1803, was chosen as the material for an environmental reconstruction based on proxy data contained within the sediment. Every sample was measured for mineral magnetic analysis (299 in total, due to one missing sample), every 4th sample was selected for diatom analysis (75 in total), and 148 samples were measured for carbon content at a resolution of at least every second cm for the top 190 cm of sediment and subsequently every fourth cm.

Figure 8. Coring platform for Lake Basotu (photo: Lars-Ove Westerberg, March 2013).

15 3.1.2 Diatoms Diatoms are single celled algae, which produce a silica shell, and are found in habitats containing some amount of water. The name “diatom” connotes the two valves that are held together by a girdle band to construct the frustule, or siliceous part of the cell wall (Dixit et al., 1992). The silica valves of diatoms are morphologically unique and have been used to develop a taxonomic system (Smol & Stoermer, 2010). Identification of diatoms is based on the following morphological variations:

1. Centric Diatoms a. Symmetrical along all axes i. The exception to this rule is Chaetoceros spp., which is rectangular in shape and therefore not symmetrical along all axes. Chaetoceros spp. is regarded as centric as it lacks a raphe. b. Size: diameter c. 6WULDHW\SHVKDSHRULHQWDWLRQGHQVLW\ QXPEHUȝP HWF 2. Pennate Diatoms a. Symmetry: Along two, one, or no axes (asymmetrical) b. Shape of valves: elliptical, ovate, semi-circle, etc. c. Raphe: presence, type, location d. 6WULDHW\SHVKDSHRULHQWDWLRQGHQVLW\ QXPEHUȝP HWF e. Size: length and width

Diatoms are a favoured biomonitor for environmental conditions (Dixit et al., 1992; Piirsoo et al., 2010), and in some cases considered more effective indicators of ecological change than water chemistry (Vilmi et al., 2015). Diatoms reproduce asexually and respond quickly to changes in the conditions of their environment (Patrick, 1948). Environmental factors affecting diatom distribution include pH, nutrient levels, salinity, temperature, and light availability. As diatom ecology has been studied extensively, the associated species are considered proxies for paleoenvironmental conditions (Battarbee, 1986). Taphonomy is an important consideration for paleoenvironmental reconstructions (Flower, 2017). After death, and before becoming trapped in sediment, the silica frustules of diatoms may be subject to dissolution, breakage, and grazing. Frustules can also be broken as a result of sample handling and slide preparation procedures (Flower, 1993). Taphonomic processes, occurring in both the water column and within the top sediment due to infiltration into open pore space (Iler, 1979), can lead to erroneous interpretations due to higher preservation of more robust species (Barker, 1992; Gasse & Van Campo, 2001). When concentrations of dissolved silica are low, some degree of dissolution will always occur (Flower and Ryves, 2009), but in pH environments above 9.0, silica dissolution can be substantial (Krausse et al., 1983). The degree of degradation to diatom frustules may also be a factor of sedimentation rates due to grazing by bottom feeders or dissolution from prolonged exposure at the sediment-water interface (Bradbury & Megard, 1972). Despite taphonomic drawbacks, fossil diatom assemblages have been found to capture environmental gradients (Hassan, 2015). Diatoms in sediment cores from Africa have been frequently used as a tool for reconstructing histories of environmental variability (e.g. Gasse et al., 1997;

16 Barker et al., 2002; Stager et al., 2005, 2009; Öberg et al., 2012). Modern diatom assemblages in Africa have shown significant relationships with hydrochemical variables (Gasse et al., 1995; Gasse, 2002). In the Eastern Highlands of Zimbabwe, diatoms from both remote streams and those near settled areas were collected and analysed. It was found that anthropogenic activity increased levels - of total dissolved solids (TDS) and NO3 , which was evident in the distribution of species of diatoms growing in the different sampling sites (Bere et al., 2013). The timing of recent eutrophication in Lake Victoria was determined through the analysis of nutrient preferences of diatoms in a sediment core from a Ugandan site (Stager et al., 2009). The discovery of that timing will be used in determining the primary causes behind algal blooms and anoxic conditions endangering fisheries in Lake Victoria. Additionally, in Lake Rukwa, Tanzania, diatoms were used to infer lake level changes over 21 700 years (Barker et al., 2002). For the diatom analysis in this dissertation, a standard slide-preparation method based on Battarbee (1986) was employed:

1. Drops of a 10% solution of HCl are added to each sample to test for carbonates. If the sample exhibits a reaction, carbonates are present and more drops are added to dissolve the material. 2. Samples are rinsed with distilled water and decanted in 2-hour intervals until a neutral pH is reached. 3. A 17% H2O2 solution is added and the samples are heated to 100°C to oxidize organic material. Samples are boiled until liquid in the beakers is no longer cloudy. If necessary, 30% H2O2 may be added slowly to samples with large amounts of organic matter; this is not done initially to avoid violent reactions once samples are placed on hot water bath to boil. 4. Samples are rinsed with distilled water and decanted after 2 hours, at least three times or until clear. 5. Ammonia (NH3) is added to disperse clay particles. 6. The clay fraction is removed by decanting in 2-hr intervals until the water is no longer cloudy. 7. The prepared material is allowed to dry on cover slips and mounted in Naphrax®.

The two-hour intervals used in steps 2 and 4 are based on Stokes’ law for determining the rate of sedimentation of an object falling through fluid. After two hours, diatom frustules will settle to the bottom while organic and clay particles remain in suspension. The Naphrax® used in slide mounting (step 7), is a microscope slide mountant in a toluene solvent, which has a refraction index of 1.65. The toluene solvent of Naphrax® is removed via heat during the slide mounting process and the refraction index increases to 1.73. (Brunel Microscopes Ltd). To increase identification accuracy, a guide was produced using a selection of samples examined under an environmental scanning electron microscope (ESEM). Prepared microscope slides were analysed and diatom valves were counted under a light microscope at X1000 magnification using oil immersion to further increase the refractive index. Identification was based on several references including a South African guide to common diatoms (Taylor et al., 2007), including reference material from Krammer and Lange-Bertalot (1986; 1988; 1991a, b) among others. A minimum of 300 valves were counted per

17 sample, but increased to 500 if one species accounted for more than one-third of the total valves counted. One of the strongest factors associated with diatom ecology is salinity, which is an indirect tracer of climatic variation (Battarbee, 1986; Gasse et al., 1997; Holmgren et al., 2012), and the focus of the analysis within this dissertation. When evaporation dominates the precipitation-evaporation (P-E) balance, ionic concentration within the body of water will result in flourishing of diatoms that prefer more saline environments. This concentration by evaporation also affects nutrient levels, but these are often altered by diffuse sources of anthropogenic pollution, primarily in the form of agricultural runoff and pastoral practices (de Jonge et al., 2002). Salinity preferences for the species identified in this dissertation were obtained from the preceding references as well as the online databases Diatoms of the United States (http://westerndiatoms.colorado.edu) and AlgaeBase (http://www.algaebase.org). Diatom species were classified into one of the following salinity preference groups:

- Brackish (demand relatively high salinity, but less than seawater) - Halophilous (able to tolerate low salinity, but generally freshwater) - Indifferent - Freshwater (cannot tolerate salinity) - Aerophilic (live exposed to air in moist areas; not submerged) - Unknown

Primary analysis of diatom distributions in this dissertation is qualitative and based on fluctuations in the ecological preference of identified taxa. Samples with a mixture of diatoms that have different salinity preferences may indicate fluctuations in shorter time scales than that captured by the bulk sampling process, which typically represent 10–100 years, depending on sampling interval (Gasse et al., 1997). Stratigraphic diagrams are expressed as relative abundances (percent of total) of each taxa and cluster analysis were performed using the TILIA program version 1.7.16 (Grimm, 1991).

3.1.3 Mineral Magnetic Measurements Iron oxides, and occasionally iron sulphides, have specific concentrations and properties, referred to as their mineral magnetic parameters (Dekkers, 2007). Magnetization occurs when electrons are moving in the same direction, which maintains, rather than cancels, their charges. These electrons are contained within small regions known as domains, each representing uniform magnetization directions (Dekkers, 2007). Domains are divided into the following types based on the size of the magnetic grain:

- superparamagnetic (SP): <20–25 nm, unstable - single-domain (SD): 25–80 nm, most stable - pseudo-single domain (PSD): 80 nm to 10–ȝPYHU\VWDEOH - multi-domain (MD): >10–ȝPOHVVVWDEOH

Minerals are divided into the magnetic classes of ferromagnetic, ferrimagnetic, and antiferromagnetic based on the spin of the electrons within the domains. This is referred to as magnetic moments (Dekkers, 2007). Figure 9 depicts the arrangement of these magnetic moments and the resulting net

18

Current practices also measure anhysteretic remanent magnetizations (ARM), a permanent magnetization produced in the laboratory by exposing the samples to an external magnetic field (Verosub & Roberts, 1995). ARM is typically analysed by its VXVFHSWLELOLW\ Ȥ$50 ZKLFKLVPDLQO\VHQVLWLYHWRVPDOOJUDLQV of SD or PSD size. However, ARM can decrease with high concentrations of magnetic particles due to interactions between them (Verosub & Roberts; 1995: Liu et al., 2012). Magnetic minerals in lake sediment can originate from both external and internal sources. External sources include both atmospherically deposited material and that which is washed into the water body (Thompson & Oldfield, 1986). Magnetite (Fe3O4) is the most important magnetic mineral of terrestrial RULJLQ IROORZHG E\ +HPDWLWH Į)H2O3) (Dekkers, 2007). Magnetite is ferrimagnetic (Banerjee & Moskowitz, 1985) and Hematite is antiferromagnetic but still able to carry a natural remanent magnetization (Dekkers, 2007). Internally, magnetic minerals can be produced by bacterial activity in anoxic bottom conditions; magnetosomes are bacteria that produce magnetite or greigite (Fe3S4) within their cells (Hesse & Stolz, 1999). Interpretation of magnetic parameters is not always straightforward, due to the possibility of different magnetic minerals having similar signatures. Magnetic mineral concentrations will also be diluted as organic matter accumulates and deposits in lake sediment (Thompson & Oldfield, 1986). This is typically observed as an inverse relationship between magnetic susceptibility and organic carbon content, although other factors such as CaCO3 and diatom silica deposition will also affect the measured values. Investigations of mineral magnetic parameters have previously been undertaken in African sediment cores. Magnetism, in particular magnetic susceptibility, was determined to be an effective tracer of rates of sedimentation in Lake Naivasha, Kenya (Boar & Harper, 2002). ARM measurements across a transect of the lake indicated that shoreline papyrus acted as a barrier to very fine-grained particles. A history of, primarily human-induced, soil erosion through sedimentation rates was established using mineral magnetic concentrations from sediment in Lake Haubi, Tanzania (Eriksson & Sandgren, 1999). In a study from Lake Tritrivakely, Madagascar, Williamson et al. (1998) suggested that varia-tions in mineral magnetics are due to fluctuations between oxic and anoxic depositional environments and that during dry periods, when oxic conditions prevailed, there was less dissolution of ferric oxides. Mineral magnetic parameters were also determined to be strong indicators of river- discharge and allowed identification of flooding events in a sediment core from the Limpopo River-plain in Mozambique (Sitoe et al., 2015). For the Basotu core, mineral magnetic analyses were undertaken at the Paleomagnetic Laboratory in the Department of Geology-Quaternary Sciences of Lund University, Sweden. The following procedure is further discussed in paper I of this thesis:

20 1. Standard plastic cubic boxes (internal volume 7 cm3) were packed with sediment, which was subsequently dried to provide distinct samples. 2. Initial volume magnetic susceptibility (K) was determined using a Geofyzica Brno KLY-2 Kappabridge, operating at a frequency of 920 Hz and a field intensity of 300 A/m. The Kappabridge automatically zeroed between each sample measurement and was calibrated with the manufacturer’s standard of 1171 x 10-6. 3. ARM was induced using 2G-Enterprises demagnetization coils at a peak alternating field (AF) of 100 milliTesla (mT) and a direct bias (DC) field of 0.1 mT (converts to magnetic field intensity [H] of 79.6 A/m). 4. SIRM was induced in a Redcliffe 700 BSM 700 pulse magnetizer acquired at 1 T and an IRM backfield at -100 mT was induced with a Molspin pulse magnetizer. The intensity of the fields produced by the pulse magnetizers were set using a Redcliffe Hall-effect probe and meter. 5. The isothermal remanences were measured with a Molspin Minispin that was calibrated with the manufacturers’ standard of 964 x 10-3 A/m.

The following magnetic calculations were supplied by Ian Snowball of Uppsala University, formerly of Lund Paleomagnetic Lab), and used in the analysis of magnetic measurements for core 1803:

ɖ =k×0.01÷md

ɐARM = ARM÷(md ÷ 1000) × 1000

ିଷ ି଺ ɖARM = ɐARM ×10 ÷79.6÷10

ɐSIRM =SIRM÷(md ÷ 1000) × 1000

ɖARM = ( ÷ ) ÷79.6÷10ିହ SIRM ɐARM ɐSIRM

Where md is the dry weight in g, 79.6 is the strength of the DC bias magnetic field to produce the ARM in A/m (H), ı represents the mass specific magnetization, and Ȥ is mass specific magnetic susceptibility.

3.1.4 Carbon Content The loss on ignition (LOI) method of determining carbon content can overestimate concentrations. Therefore, direct measurements of total organic carbon are generally preferred (Meyers & Teranes, 2001). Samples from Basotu were analysed using an ELTRA CS 500 in an oxygen atmosphere at 550°C to determine total organic carbon (TOC) and 900°C to determine carbonate contribution as total inorganic carbon (TIC) (Heiri et al., 2001). The concentration of CO2, formed when carbon components are oxidized during combustion, is displayed digitally as a percent of carbon present in the sample based on weight (ELTRA, 2004). Analysis was performed for at least every second sample for the top 190 cm of sediment and every fourth sample thereafter. To precisely determine which contributions to the carbon content of a sediment core that are land or aquatic based, isotopic measurements of carbon and nitrogen are necessary (Meyers & Teranes, 2001).

21 Deposition of organic carbon in lake sediments is of greater analytical importance than inorganic carbon (Cole et al., 2007). By analysing the amount of TOC in sediment, inferences can be made regarding the vegetation that was once living in and around the lake (Meyers & Teranes, 2001). Values for TOC are dependent on the initial biomass and its subsequent degradation through transport and preservation. For analytical purposes, carbon content in this thesis is interpreted as a factor of general lake productivity. TOC was found to positively correlate with lake depth in Lake Naivasha of Kenya during the historical record (Verschuren et al., 2000). High values of carbon (>20%) were also found to indicate deep water conditions associated with elevated production in Lake Victoria (Beuning et al., 1997).

3.1.5 Radiometric Dating

Radiocarbon 14C is the longest living radioisotope of carbon and is produced when 14N in the atmosphere collides with neutrons produced by cosmic radiation. When this occurs, the 14N atom loses a proton and turns into a carbon atom. The neutron is taken up by the carbon atom, thus forming 14C, while the released proton forms a hydrogen atom. The 14C atoms rapidly diffuse in the lower atmosphere, 14 eliminating any geographical variation, and oxidize into CO2, which then enters the biosphere where it is taken up by all living organisms (Bradley, 2014). Upon death, the 14C in an organism begins to decay with a half-life of approximately 5730 years. This process is measureable up until 10 half-lives or approximately 57 300 years after death with conventional methods (Bradley, 2014). As 14C is produced cosmogenically, variations in the rate of its productivity are not consistent through time (Krishnaswami & Lal, 1978). Solar activity varies due to sunspot activity and changes in the earth’s geomagnetic field (Bradley, 2014). During periods of low sunspot activity, there is an increase in the intensity of cosmic rays that reach Earth’s atmosphere, thus increasing 14C production. Conversely, 14C production is decreased when the magnetic field of the Earth increases as this prevents cosmic rays from reaching the atmosphere. To account for this, 14C calibration curves have been developed using statistical methods and annually resolved records such as tree-rings (Bronk Ramsey, 1995; 2009). Using 14C as a method to date sediment is not without potential difficulties. Contamination is a possible source of error in 14C dating, especially with bulk sediment samples. In areas with carbonate rocks, organisms may absorb “old” carbon from their surroundings, which leads them to appear older than they are in reality (e.g. Olsson, 1986). In addition, influx of dissolved carbonates or clasts of carbonate rock totally lacking 14C, will yield proportionally lower amounts of 14C in bulk samples, than what would be the result of dating organic matter alone (Walker, 2005). To avoid these forms of contamination, it is preferred to date terrestrial macrofossils. Macrofossils in volcanic lakes are rare (Marchetto et al., 2015), and in the Lake Basotu core no macrofossils were found. Consequently, 13 samples of bulk sediment were sent to Ångström Laboratory of Uppsala University in Sweden and one additional 14C date was performed by

22 Beta Analytics in Florida, USA. Both facilities used accelerator mass spectrometry (AMS) which counts the remaining number of carbon-14 ions present in each sample to determine a probable age range. AMS has the benefit of greater sensitivity, which requires smaller sample sizes than the traditional beta-counting method. Samples were pre-treated in 1% HCl, followed by 1% NaOH. OxCal (https://c14.arch.ox.ac.uk/oxcal.html), a freely available online platform using automatic calibration to build 14C chronological models, was used to construct a 14C age-depth model (Bronk Ramsey, 1995). The Southern Hemisphere calibration curve (SHcal13) was chosen as it accounts for the differences in the land to water ratio which produces older 14C ages due to larger expanses of ocean in the SH (Hogg et al., 2013).

Lead-210 210Pb is part of the series of uranium decay and has been successfully used in East Africa as a method for developing geochronologies (Barton & Torgersen, 1988; El-Daoushy & Eriksson, 1998; Stoof-Leichsenring et al., 2011; Öberg et al., 2012). Lake sediment contains two types of 210Pb, supported and unsupported. Supported 210Pb is constantly produced by the decay of 226Ra in the Earth’s crust, including its minerogenic sediment. The decay yields a constant ratio between 226Ra and supported 210Pb. Unsupported 210Pb, on the other hand, derives from atmospheric fallout. 226Ra that escapes into the atmosphere decays, via the intermediate isotope 222Rn, into 210Pb (Begy et al., 2009), which after fallout may enrich sediment with regards to 210Pb and disrupt the balance between 226Ra and 210Pb. Dating methods using 210Pb are thus based on the disequilibrium between 226Ra and 210Pb caused by the presence of the unsupported type in the sediment (Brenner et al., 1994). 210Pb has a relatively short radioactive half-life (22 years), and fallout of unsupported 210Pb varies with geography and precipitation patterns. However, its residence time in the atmosphere is short (5–10 days; Krishnaswami & Lal, 1978), and in water bodies it becomes associated with particulate matter and settles to the bottom quickly (Ivanovich & Harmon, 1992; Appleby, 1997). Accordingly, and in light of its short half-life, this rapid sedimentation of 210Pb in lakes makes it well representative of its time of deposition. To validate 210Pb dating, 137Cs is measured as a tracer of nuclear weapons testing which peaked in 1963; this peak can also be used to determine sedimentation rates since that time (Penningtonet al., 1973). Samples from Lake Basotu were analysed for the activity of 210Pb, 226Ra and 137Cs via gammaspectrometry at the Gamma Dating Centre, Department of Geosciences and Natural Resource Management, University of Copenhagen. The measurements were carried out on a Canberra ultralow-background Ge well-detector. 210Pb was measured via its gamma-peak at 46.5 kiloelectron volt (keV), 226Ra via the granddaughter 214Pb (peaks at 295 and 352 keV) and 137Cs via its peak at 661 keV.

23 3.2 Paper II

3.2.1 Remote Sensing and GIS Data In 1965, the director of the United States Geological Survey (USGS) proposed a satellite program for monitoring of natural resources, and in 1972 the first Landsat satellite was launched (NASA, 2017a). Since then, the Landsat program has seen the launch of eight missions, with plans for a 9th satellite announced in 2015, anticipated to launch in 2023 (NASA, 2015). The timeline for Landsat missions 1–8 is found in Table 1 (NASA, 2017b). Gaps in coverage over Tanzania occurred from 1974–1978 due to conservative image acquisition in Landsat-2, and from 1980–1983 as Landsat-3 was prone to technical difficulties (Goward et al., 2006). Primarily due to gaps in coverage, 20 years of dry-season images out of the 43 years the Landsat program has been active were available for analysis.

Table 1. Timeline of Landsat Satellites Launched Decommissioned Landsat-1 1972 1978 Landsat-2 1975 1982 Landsat-3 1978 1983 Landsat-4 1982 1993 Landsat-5 1984 2013 Landsat-6 1999 1999 Landsat-7 1999 --- Landsat-8 2013 ---

The Landsat data used in this thesis are from the Tanzanian long dry-season (June–October) to maximize the reflective differences between land and water as well as to avoid interference by clouds. No additional co-registration was performed as all scenes were processed by Standard Terrain Correction (Level 1T). Using GIS techniques, shorelines for Lake Basotu were delineated manually and surface area was calculated in ArcGIS 10.3.1® (ESRI, 2015). Sannel and Brown (2010) found, via an experiment using multiple users at different expertise levels, that errors from manual delineation may fall between ±1.5 m for distinct shorelines and ±3.0 m for diffuse shorelines. Using manual delineation, nevertheless, was found to be more accurate than semi-automatic classification in situations of poor or uncertain data quality. Due to its subjectivity and time-consuming nature, manual delineation should only be applied to small study areas (Sannel & Brown, 2010). In ArcGIS 10.3.1®, false-colour composites were created from the available bands for each of the 20 selected images of Lake Basotu. Depending on the satellite, to facilitate delineation, bands were alternated between the near-infrared (NIR), shortwave-infrared (SWIR), and blue-light wavelengths to emphasize shorelines. In the early satellites, Landsat 1–3, SWIR and blue bands were not available and composites were alternated between two NIR and green bands. A seamless digital elevation model (DEM) for the Basotu area was created from a mosaic of 12 Shuttle Radar Topography Mission (SRTM) raster tiles from 4–6°S and 34–36° E (USGS, 2000). Using flow paths from this DEM and the

24 hydrology toolbox of ArcGIS 10.3.1®, a historical outflow point was determined and the Lake Basotu watershed was delineated. The SRTM was a joint project of the National Aeronautics and Space Administration (NASA) and the National Geospatial-Intelligence Agency (NGA) in the United States (USGS, 2015). The product of this mission is a global dataset of land elevations every one arc-second (c. 30 m) which is published in rasters of one-degree tiles (USGS, 2015). This dataset is available via Earth Explorer, the bulk download service of the USGS (http://earthexplorer.usgs.gov).

3.2.2 Meteorological and Climatological Data As previously mentioned, during a positive phase, the IOD can lead to increased convection along the coast of East Africa, which brings enhanced rainfall (Saji et al., 1999). The IOD is represented in climatological records as the Dipole Mode Index (DMI), a time series representing differences in the SST of the tropical western Indian Ocean and the tropical south-eastern Indian Ocean (Saji et al., 1999). Monthly DMI for 1960–2014 was calculated using the Hadley Centre Sea Ice and Sea Surface Temperature dataset. When a +IOD and an El Niño event co-occur, this has been associated with flooding in Northern Tanzania. Kijazi and Reason (2009) examined dry season floods in 1997 and 2006, when both anomalies occurred, and found similar flow patterns for both the entire season and the rainiest days. An El Niño event is associated with a basin-wide warming in the Pacific Ocean that disrupts global weather patterns (Ropelewski & Halpert, 1987; Trenberth, 1997), and is declared when the five-month running average of SST anomalies between 5°N–5°S and 120–170°W exceeds 0.4°C for at least six months (Trenberth, 1997). The atmospheric component of an El Niño event is the Southern Oscillation, and the two are often combined and referred to as the El Niño-Southern Oscillation (ENSO; Trenberth, 1997). Basotu falls within the region of East Africa that has shown the strongest relationship with ENSO in regards to enhanced precipitation, most likely due to proximity to the ocean (Ntale & Gan, 2004). Monitoring of ENSO is done via the Multivariate ENSO Index (MEI), which is maintained by the Earth System Research Laboratory (ESRL) of the National Oceanic and Atmospheric Administration (NOAA) in the United States (https://www.esrl.noaa.gov/psd/enso/mei). The MEI is calculated based on observed data over the tropical Pacific: sea-level pressure, zonal and meridional components of the surface wind, sea surface temperature, surface air temperature, and total cloudiness fraction of the sky (Wolter & Timlin, 1993). Values are updated monthly, and MEI data for this thesis were obtained in August 2015. Meteorological observations for the period of 1970–2015 were obtained from the TMA for two stations in north-central Tanzania: Singida Agricultural Station and Babati Station. The observations of these time series were discontinuous, with 91% and 81% available data, respectively. TMA data were also obtained for the shorter period of 2000–2015 from a third location, Singida Meteorological Station. A time series for precipitation in Basotu was created using a distance weighted average of the three stations and filled in with supplemental information from the Global Precipitation and Climatology Centre v6 (GPCC) provided by ESRL of NOAA (Udo et al., 2011). In a case-study

25 where seven global precipitation data sets (GDPs) were used to fill gaps in observed precipitation data from Tanzania, the GPCC was shown to have high spatial correlation (Koutsouris et al., 2015).The GPCC has the potential for worse performance during the dry season over the wet, but still clearly showed wet and dry periods and successfully followed the reference climatology (Koutsouris et al., 2015).

3.3 Paper III To ascertain the effects that land-use in the Basotu area has on the health of the lake, a chronology of events concurrent with the record of lake surface area, from paper II, was constructed. Newspaper articles from 2006–2016 were obtained from internet news searches of keywords related to Lake Basotu and are outlined in Table 2. A list of reports, which have been obtained and reviewed, are found in appendix B.

Table 2. Results of Internet News Search (25 articles retrieved, covering 2006–2016) Keywords Swahili terms Articles Irrigation umwagiliaji 4 Erosion/sediment mmomonyoko 3 NAFCO/Canada/Wheat ngano 20 Pesticides/fertilizers/chemicals/herbicides 4 Unrest/conflict//harassment 12 Search terms: Basotu, Bassotu (alternative spelling of Basotu), Hanang

Telephone interviews with two informants in Tanzania were conducted from Stockholm University by Martina Angela Caretta, currently Assistant Professor of Geography at West Virginia University. The informants were first contacted by email to inform them of the topic and to obtain permission to call, and each interview lasted approximately half an hour. Informant #1 was employed by CIDA in the 1970s, and is still residing in Basotu; #2 was working in Lake Basotu area in the late 2000’s. Following each interview, the findings were summarized and discussed with each informant to validate and ensure the reliability of the information (cf. Burke & Miller, 2001). The surface area record from Higgins et al. (2016) was compared to annual precipitation from the previous year, as well as the number of days in which the temperature exceeded 32°C for the same year. The rainy season in Basotu was also compared to the local growing season, to determine whether a relationship was present.

26 4 Results

4.1 Paper I Paper I presents a summary of the results of a multiproxy paleoenvironmental reconstruction from a Lake Basotu sediment core. Proxies examined include diatom assemblages (appendix A), mineral magnetic parameters, and carbon content. Samples were dated using 14C and 210Pb. Of initial difficulty to this environmental reconstruction was the radiometric chronology. The age-depth model derived from 14C dates exhibited a prominent reversal. In an attempt to constrain the dates, 210Pb was employed, but due to mixing in the top layer it was only possible to pinpoint the depth where the overlying sediment must be younger than 100 years. The 14C dates themselves did not seem random in regards to depth, but rather indicated that they may have been the result of an event in the lake’s history. Rather than abandoning the core as inappropriate material due to the dating difficulties, the decision was made to continue with the analysis in search of connections to that pattern within the proxy data. The most affirming discovery was the almost complete disappearance of diatoms at the same depth as the most extreme date and point of reversal in the age-depth model (~50 cm depth). Even more telling was the presence of aerophilic taxa at that depth, indicating that the lake had almost completely dried out. These results prove the importance of not disregarding troublesome data at first glance. As 14C dating was proven to be an interesting record of variability, it was also an ineffective method of determining the age of the sediment. In an attempt to infer a chronology for the sediment core a combination of proxy data, other sediment records, and historical climate events were consulted. The theory being that events of established dates could be placed along the core by examining the magnetic record, carbon measurements, and diatom observations as a whole. The subjective nature of this method is recognized, but without more reliable dating methods it is the only viable option to place this interesting record into the larger picture of climate variability in this region. Based on this inferred chronology, the Lake Basotu core was assumed to have been deposited primarily during the Little Ice Age (LIA), which occurred AD 1300–1850 in East Africa (Verschuren et al., 2000). ESEM images for prominent diatom taxa in the Basotu core are shown in Figure 10. Diatoms were grouped based on salinity preferences to draw inferences on water level changes. The mineral magnetic parameter ߯ARM/SIRM was selected as a proxy for erosion, with increasing values indicating decreasing ferrimagnetic grain size. A positive correlation between ߯ARM/SIRM and counts of freshwater diatom species in the core section, held to be most stable (249–120 cm; R2=0.33), is further justification for its use. These proxies, together with fluctuations in carbon content, present a picture of a highly variable lake with a substantial response to precipitation during the LIA.

27

The results of this recalculation of percent at depth after the removal of C. meneghiniana and N. amphibia are split into freshwater taxa (Figure 14) and non-freshwater taxa (Figure 15). Removal of the two species did not alter the overall interpretation of the Lake Basotu sediment core, but some obvious differences are noted below:

- The occurrence of Anomoeoneis sphaerophora, a brackish species preferring high pH (Schmid, 1979). In the revised diagrams, A. sphaerophora is present in low percentages in each zone of the core, but in fluctuating amounts. This lends to the original hypothesis of a rapidly changing lake level, as A. sphaerophora is most likely developing in the shallow shores as it has shown adaptation to changes in osmolarity (Schmid, 1979). A lowering of lake-levels would trap A. sphaerophora near the shore, which would then wash in as levels rise again; this also explains why the valves are primarily found as fragments (cf. Robertsson, 1973).

- Other species occurring in nearly all zones are Pinnularia acrosphaeria and Craticula ambigua. Both are grouped as freshwater species and considered cosmopolitan in the tropics (Taylor et al., 2007).

- Tryblionella hungarica, a halophilous species is quite prevalent in zone two of the reanalysis diagrams. This species is found concurrently with Thalassiosira faurii, which is also halophilous and is depicted in the original analysis.

- Rhopalodia gibba co-occurs with Epithemia sorex, both grouped as indifferent to salinity changes. R. gibba appears to replace another indifferent taxa, Epithemia adnata, around the beginning of zone III. Both species have similar nutrient requirements and occur in high pH environments. Thus, the transition is most likely due to increasing salinity, as R. gibba has a higher tolerance to salinity (Cumming & Smol, 1993).

31

4.2 Paper II Modern (1973–2015) surface area variability for Lake Basotu was reconstructed using dry-season Landsat satellite images and GIS techniques. The methods used are simple and effective for small areas, and easily repeatable. According to this record, in between flooding events, the size of Lake Basotu is in a state of decline. The surface area record was compared to indices for both ENSO and the IOD, as a co-occurrence of these have been linked to flooding events in northern Tanzania (Kijazi & Reason, 2009). This comparison indicated that, based on this co-occurrence theory, flooding should have occurred in years where it was not evident in the satellite record. The authors hypothesize that floods occurring in Lake Basotu after 1996 are due to sedimentation of the lake basin, which prevents infiltration of enhanced precipitation. Prior to 1996, with assistance from the Canadian government, mechanized wheat farming was implemented in the Hanang district where Basotu is located (Folster, 2001). After using a DEM to delineate the drainage area for Lake Basotu, it was determined that agricultural land within it increased by c. 20% between 1973 and 1997. When the landscape surrounding water bodies are developed in ways that affect the natural drainage patterns (i.e. agriculture and impervious surfaces such as roads) erosion is often the result. Rainfall erosion occurs when soil material is detached from the surface matrix due to the energy of falling raindrops and a subsequent transport process removes the detached particles (Kinnell, 2005). Soil surface can become sealed due to settling in open pore space of these dislodged particles (Greenland & Pereira, 1977; Carey & Paige, 2016).

Further Analysis

Lake Volume Multiple complications prohibited calculations of lake volume. No bathymetric or hydrologic data (i.e. water depth, inputs, or outputs) are available for Lake Basotu. Additionally, lake depth since the 1960s would be affected by the amount of sediment accumulating in the basin from soil erosion. An attempt was made to calculate the storage capacity of Lake Basotu using the 30 m resolution DEM from the SRTM. The DEM raster was cropped to the area of the maximum lake extent and converted to a triangulated irregular network (TIN) to allow for volume calculations. As each TIN classification falls within a range of ~11 m, and in field measurements of depth showed differences of around 4 m, the tools available are not sufficient to capture lake basin variations. Another possibility is that the DEM is representative of the water surface rather than the bottom topography, resulting in the “flat” appearance of the basin. The black outline in Figure 16 is representative of the lake extent observed in October 2002, the first measurement for Lake Basotu after the SRTM mission in 2000 (USGS, 2015). As the 2002 extent roughly follows the basin outline from the TIN, this reinforces the likelihood that the 2000 SRTM captured the lake surface.

34

(moderately suitable, but susceptible to gully erosion) at 14.25%. The percent of soils within agricultural land considered to be either marginally suitable or unsuitable at any time due to susceptibility to erosion, nutrient deficiency, and/or being prone to flooding is 43.75%. These results speak to the unsuitability of this area for mechanized farming and the need for protective measures for both soils and water resources.

4.3 Paper III Paper III is an interdisciplinary investigation into how activities in the surrounding area are both affecting and affected by the variability of Lake Basotu. In this paper, we argue that Lake Basotu is an example of how neglecting overall ecosystem health can lead to environmental degradation, which is long lasting and difficult to reverse. The Lake Basotu surface area record from Higgins et al. (2016) was compared to a chronology related to mechanized wheat farming in the surrounding region, which began in 1967. Events from the timeline were mainly extracted from newspaper articles and foreign aid reports. In 1975, after three years of “disastrous” farming (Macmillon & Ngoma, 1983, p 5), it was recommended that serious consideration be given to completely discontinuing wheat farming in Basotu, but new tillage techniques were implemented in 1976 and the program continued. By 1983, the Basotu Wheat Complex was extended to six farms (Table 5, Figure 19).

Table 5. Hanang Complex Farms (Macmillan & Ngoma, 1983) Farm Year Established Hectares Basotu 1968 4728 Setchet 1975 4828 Mulbadaw 1979 4600 Murjanda 1980 4800 Gawal 1982 3275 Gidagamowd* 1983 2690 Waret* 1983 ---- *located on eastern side of the Mulbadaw escarpment

38

Department of Geological Sciences at Stockholm University and reported as į13C (13C:12C) anGį15N (15N:14N) as well as the percentages of TOC, TIC, and nitrogen9DOXHVIRUį13C were plotted versus a Pee Dee Belemnite (PDB) FDOLEUDWHGVWDQGDUGZKLFKLVEDVHGRQDOLPHVWRQHIRVVLOZLWKDQDVVLJQHGį13C of 0 ‰, and the established standard for 13&DQDO\VLV %RXWWRQ į15N is plotted versus air which has an assumed value of zero. The ratio of 13C:12C is an indicator of whether organic material is terrestrial 12 or aquatic in origin and is based on how different classes of plants use CO2 13 13 over CO2 in photosynthesis (Bradley, 2014; Ehleringer et al., 1991). As CO2 12 is the heavier of the two variants, plants will preferentially use CO2. However, C3 plants are less efficient at using CO2 in general and will be more depleted 13 13 in C than the C4 cODVV %HQGHU :KHQį C is highly negative (< -25‰), the origin is most likely planktonic, while less negative values are produced by terrestrial contributions (Bradley, 2014). As nitrogen is a basic nutrient to plant life, the ratio of 15N:14N is indicative - of NO3 utilization within surface waters and therefore algal productivity. A study of the relationship between organic matter in the water column and within sediment for 11 lakes in Florida UHYHDOHGWKDWį15N increases as the lakes enter eutrophic states but will return to zero at its most hypereutrophic (Gu et al., 1996). According to a review of the isotopic values of various sources of nitrogen SROOXWLRQ +HDWRQ ORZYDOXHVRIį15N (-4–4‰) are indicative of fertilizer based inputs where higher values (10–20‰) would originate as human and animal waste. As nitrogen is used by aquatic organisms, the isotopic signatures of the source material are assimilated and subsequently incorporated into the sediment record. Measurements of nitrogen isotopes are not distorted by material that is nitrogen-poor within the bulk sediment (Talbot, 2001). The ratio of C/N in organic material is influenced by cellulose in plant source material and indicates whether organic matter is vascular in origin (Meyers & Ishiwatari, 1993). Non-vascular, i.e. aquatic plants have low C/N ratios due to a lack of cellulose (Duchesne & Larson, 1989). To explore the possibility of a similar dry phase to the Lake Basotu core, nine slides were prepared for basic diatom analysis. Availability of sediment for slide preparation was limited. Samples from the upper 100 cm had been partially prepared by previous researchers. Material available was in one of two states: suspended in an unidentified liquid or in dry state but having been homogenized into a fine powder. This is most likely due to material being used in the carbon and nitrogen analysis procedures. Slides were prepared from one of each type of processed material; 1 cm from the sample in liquid and 78 cm from the crushed material. Five transects were counted per slide, as in diatom analysis, a lower count is often sufficient in cases where stratigraphic correlation or simply delineating boundaries are the desired outcome (Battarbee et al., 2001).

Results and Interpretation As previously stated, the 14C dates on the Babati core are primarily from bulk sediment samples (Table 6). However, one piece of wood was extracted from 83 cm depth in the core, and was submitted for dating along with bulk sediment from that depth. The age of the wood was 250 ± 33 14C years BP, while the bulk sediment was placed at 1532 ± 30 14C years BP showing a potentially significant offset.

41

Table 6. Radiocarbon dates from Lake Babati core, primarily based on bulk sediment. Sample Depth in 14C age Calibrated age Distribution ranges ID core (cm) (years BP) 496 2 292 ± 30 AD 1507–1586 28.2%: AD 1507–1586 57.9%: AD 1619–1673 4.1% AD 1743–1768 5.1%: AD 1780–1797 518 24 100.5 ± 0.4 pM AD 1816–1922 67.4%: AD 1816–1830 28.0%: AD 1892–1922 535 41 106.0 ± 0.4 pM AD 1816–1922 49.8%: AD 1816–1830 45.6%: AD 1892–1922 566 72 676 ± 30 AD 1297–1394 95.4%: AD 1297–1394 577 83 1535 ± 30 AD 525–644 95.4%: AD 525–644 577* 83 250 ± 33 AD 1629–1807 36.9%: AD 1629–1696 58.5%: AD 1725–1807 582 88 1304 ± 30 AD 679–861 95.4%: AD 679–861 606 112 1387 ± 30 AD 644–766 64.6%: AD 644–692 5.6% AD 700–732 25.2%: AD 730–766 650 156 2195 ± 30 358–103 BC 30.9%: 358–274 BC 64.5%: 261–103 BC 726 232 538 ± 30 AD 1401–1450 95.4%: AD 1401–1450 730 236 1032 ± 30 AD 994–1148 46.5%: AD 994–1070 48.9%: AD 1077–1148 750 256 2401 ± 30 727–371 BC 0.8%: 727–718 BC 1.0%: 705–695 BC 93.5%: 542–371 BC pM: modern carbon contamination; *macrofossil based date

Figure 21 shows a comparison between the age-depth models from Lake Basotu and Lake Babati. Using two matching age-ranges, the Lake Babati age-depth model has been stretched to fit that of Lake Basotu and appropriately placed in the timeline in Figure 22. The age range for the top sample of the Babati core (AD 1507–1586) falls within that of the top sample from the Basotu core (AD 1457–1628). The second match is found at 150 cm in Lake Basotu (AD 1408–1459) and 232 cm in Lake Babati (AD 1401–1450). Both the top of each sediment core and the culmination of the age reversals lie within the LIA on the timeline, if the 14C dates are accepted as accurate. These results may indicate that the LIA in this region of Tanzania was more arid than what has been previously suggested. A similar reversal in dating has been observed in sediment from a wetland in the Ngorongoro Conservation Area of Tanzania (~150 km NNW of Basotu) but the culmination was placed earlier at AD 1280 (Svensson, 2010).

42

As the non-carbonate bedrocks surrounding Babati do not support the reservoir effect theory as suggested for Lake Basotu, older organic material introduced via erosion may be the source (Nambudiri et al., 1980). Due to extensive agriculture and land clearing, Lake Babati has suffered siltation which local fishermen have described as so severe that in some areas of the lake their nets can no longer be submerged (Kahurananga, 1992). If ploughing of land exposes material which is depleted in 14C, the incorporation of this into lake sediment may explain the older dates.

Preliminary Results The results of prior analyses performed on the Lake Babati core are presented in Figure 23, with the results of a brief exploration of identified diatom species in Appendix C. Overall vDOXHVRIį13C are typical of organic material terrestrial in RULJLQ+RZHYHULQDOJDHIURPIUHVKZDWHUODNHVWKHį13C signature cannot easily be differentiated from that of C3 land plants, as their primary source of CO2 is dissolved and at equilibrium with that of the atmosphere (Meyers, 1994). For this reason, the C/N ratio is likely more representative of the internal productivity of Lake Babati and thus indicates that organic material is primarily algal in origin. Sediment layers containing aggregates (113–88 cm and 258–235 cm) may be similar in origin to the desiccated material from Lake Basotu sediment, i.e. a result of prolonged drying.

Magnetic Zone I (258–235 cm): Aggregates in the sediment correspond to LQFUHDVHGȤDQGWKXVHLWKHULQFUHDVHGHURVLRQRUDORZHUODNHOHYHO3UHYDOHQFHRI Cyclotella meneghiniana suggests slightly saline water, which lends to the interpretation of a low lake level (cf. Holmgren et al., 2012). According to field notes from 2009, the amount of aggregates in this core section was large, and as further penetration into the sediment by the coring equipment was not possible, there may have been complete desiccation below.

Magnetic Zone II (235–114 cm): Low and relatiYHO\VWDEOHȤPHDVXUHPHQWV indicate low erosion intensity, or a coring location farther from the shoreline. Aulacoseira ambigua and Aulacoseira granulata var. angustissima are both freshwater species, but flourish in nutrient rich environments. The co-occurrence with Cyclotella meneghiniana suggests variable lake level between the wet and dry seasons. As water levels in Lake Babati increase and salinity decreases, Aulacoseira resting cells would be rejuvenated (Sicko-Goad et al., 1989).

Magnetic Zone III (114–73 cm  $ SURORQJHG SHDN LQ Ȥ FRUUHVSRQGV WR aggregates in the sediment. The extremely low counts and presence of Pinnularia borealis at 103 cm suggests of drought conditions that would substantially have lowered the lake level.

Magnetic Zone IV (75–53 cm $GURSLQį15N at approximately 70 cm may represent the introduction of N into the Lake Babati catchment, probably from an increase in pastoral activities. This is accompanied by an increase in carbon content, which may represent increased productivity within the lake due to the higher nutrient availability.

44

Magnetic Zone V (53–15 cm): carbon content decreases DQGȤLQFUHDVHVWRZDUGV the sediment surface. This indicates a trend towards a larger Lake Babati and thus a more humid period in the area.

Magnetic Zone VI (15–0 cm): Peaks LQȤDWDQGFPFRXOGEHUHSUHVHQWDWLYH of the 1964 and 1990 floods (cf. Sitoe et al., 2015). Sediment material from this level contained no diatoms as it has been previously prepared for pollen analysis.

Without further analyses, it is not possible to make ecological inferences to specific periods. 210Pb dating may help to identify recently deposited material to verify the hypothesis that siltation from agricultural activities has led to the introduction of old carbon into lake sediment. In addition, for sediment older than 100 years, it may be possible to apply the inferred chronology method from paper 1. Magnetic zone II in the Babati core closely resembles the section of Lake Basotu sediment that has been matched to the beginning of the LIA (AD 1300), and determined to be a time of a relatively stable lake. Consequently, increased erosion and soil formation processes in zone III could then be linked to either the Nyarubanga famine of AD 1560–1620 or the Lapanarat-Mahlatule famine of AD 1760–1840 (cf. Verschuren et al., 2000). A more thorough investigation of sediment from Lake Babati is suggested. This would offer additional insights into the paleoclimate and paleohydrology of north-central Tanzania and in particular the flooding history of Lake Babati. Smaller sampling intervals will capture more of the seasonal variation and lend to investigations of changing precipitation patterns.

Further Implications for Lake Basotu The reversal in the age depth model of the Lake Babati core strengthens the chronological interpretation of Lake Basotu. As both have similar shapes and dates, it is unlikely that user error played a part in the dating process. Additionally, the difference between TOC and TIC is less in the Lake Basotu sediment, where carbonate rocks are present, than in Lake Babati sediment. It follows that allochthonous contributions to sediment were likely low in Lake Basotu, strengthening the hypothesis that, prior to intensive agricultural practices, erosional contributions into Lake Basotu were generally low.

46 6 Discussion

Proper governance of water resources requires an understanding of the multiple scales and processes involved in those landscapes (Savenije et al., 2014). By combining studies involving natural and human systems, patterns may arise that were previously missed by separating the social and environmental aspects of ecosystem studies (Liu et al., 2007). A comprehensive understanding of environmental variability in the context of both climate change and land use makes the work directly relevant to parties involved in planning and policy positions (Lambin, 2005). This work begins with a discussion of the ecological history of Lake Basotu and how the lake has experienced changes over the last c. 1000 years (paper I). In paper II, that record is extended into modern times by examining lake extent variations using satellite images. Finally, in paper III the modern variability of Lake Basotu is compared to land use in the surrounding area. Figure 24 highlights how the three manuscripts contained within this dissertation are interconnected. Papers I and II directly contribute to the understanding of one another. Changes observed in the satellite record (paper II) confirm that the water-level changes determined by the paleolimnological reconstruction (paper I) are reasonable. Based on the results from paper II, paper III further explores what factors are influencing Lake Basotu today and highlights how neglecting interconnections between environmental parameters can have lasting implications for freshwater resources. The baseline of variability established by paper I further demonstrates the importance of the work done in paper III, as water stress is likely to increase in these types of environments. This mixed method approach, which has been applied to the material and area of Lake Basotu, combines an understanding of past variability with present day changes to create a useful policy making case study. Overall, this work demonstrates that although the natural state of Lake Basotu is variable, land use in the surrounding area has likely increased the sensitivity of the basin to climatic events. As precipitation becomes increasingly unreliable, and the demands of an increasing population place stress on the environment, it is more important than ever to develop strategies for conserving freshwater resources. Naiman et al. (1998) made a call for scientific work regarding freshwater management, which addresses two challenges: “(1) to use sound ecological principles to inform the decision-making process and (2) to provide an effective interdisciplinary perspective to resolve freshwater issues.” (p 570). This thesis addresses both of these challenges by presenting a combination of investigations into the past and present of Lake Basotu. In the subsections below, other aspects of relevance for the interpretation of the results presented in the thesis are discussed.

47 may have proved useful in analysing where there is a groundwater contribution and the past sensitivity of Lake Basotu to precipitation.

Water Volume and Residence Time As the volume and water inputs/outputs of Lake Basotu are unknown, it is not possible to calculate water residence time, which is a function of both parameters (Dodds, 2002). Additionally, residence time is likely complicated by increased sediment inputs from soil erosion in recent years (Higgins et al., 2016). General estimates of lake volume going back in time were made using changes in diatom species composition in paper I. To strengthen this interpretation, it may have been possible to use isotopic variations in oxygen as a proxy. As the lighter isotope (O16) preferentially evaporates, enrichment in the heavier isotope (O18) in lake water will occur during periods of high evaporation and low precipitation (Wolfe et al., 2001). Generally speaking, this isotopic signature is preserved in sediments through incorporation into aquatic organisms as well as carbonate precipitation (Leng & Marshall, 2004). The source material of sediment is needed to determine which, if any, of these potential proxies would be appropriate.

Chronological Difficulties The amount of 14C isotope in sediment is determined by a combination of organic material in the lake and from its surroundings (Björck & Wohlfarth, 2001). When using bulk sediment for 14C dating, uncertainty will arise from contamination by other sources of carbon (Grimm et al., 2009). Aquatic organisms will assimilate carbon from clastic, carbon-rich material dissolved in the water (Figure 25). This clastic material is free of 14C, and will therefore “age” the assimilating organisms, by reducing the proportion of 14C in the organisms (Geyh et al., 1998). Moreover, organisms living in the lake may assimilate carbon from decomposing limnic and terrestrial organisms. Such organic material has a certain amount of 14C that depends on the time of death of the decomposing organisms. Combined, these ageing effects are known as the “reservoir effect” (Deevey et al., 1954). Further affecting the isotopic composition of aquatic organisms is the lag time regarding incorporation of 14 atmospheric CO2 into waterbodies, which is directly related to lake size (Olsson, 1991). Direct contamination of the bulk sediment also arises from increased erosion bringing in detrital material, which completely lacks the 14C isotope (Nambudiri et al., 1980; Davies et al., 2004) and older organic material which is depleted in 14C (Björck et al., 1998). The 14C age-depth models for Lake Basotu and Lake Babati have been compared in this dissertation. The general shape of the model in Lake Basotu indicates relatively stable accumulation at the bottom of the core with one period of fluctuating lake levels, which led to increasingly older dates and a distinct reversal. The model for Lake Babati exhibits a more back and forth pattern in the 14C ages, which may be due to the absence of carbonate bedrock to influence the isotopic composition of the water. The timing of the age reversals in Lakes Basotu and Babati indicates a sub-regional drought during the LIA, the main phase of which is considered cool and wet in East Africa (e.g. Verschuren et al., 2000; Stager et al. 2005). However, Russel et al. (2007) suggested an east-west spatial pattern to the LIA with arid conditions prevailing in westernmost East

49 as it can be certain it has some affinity for salinity. Thalassiosira faurii is generally considered an indicator of salinity (Barker et al., 2002), but recent work has questioned that affiliation (Roubeix et al., 2014). For Lake Basotu, the classification as a halophilous species was maintained, as further research is necessary to determine if further taxonomic distinction is necessary.

Remote Sensing The Landsat archive is freely available, easily accessible and enables catchment scale monitoring of land cover (Giardino et al., 2010). This archive offers a means to monitor remotely located lakes in Africa where on site monitoring may not be possible (Ballatore et al., 2014). Although somewhat subjective, manual delineation of shorelines was selected for this task, as Lake Basotu is relatively small with distinct shorelines (cf. Sannel & Brown, 2010). It was hoped that if successful, this method could be expeditiously employed with little training for other important freshwater resources under conflict, as Tanzania is a country lacking in hydrological data (UN Water, 2013). For larger areas, automatic detection is generally preferred (Ouma & Tateishi, 2006). Supervised classification of Landsat has been used to track lake extent changes in Lake Urema of Mozambique (Böhme et al., 2006). This was done in concert with Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) data. The use of ASTER allows the user to designate samples that represent the desired classification, which is then performed automatically by the software. Landsat images were initially consulted to account for discrepancies between field observations of Lake Basotu in 2013 and work done by the British Geological Survey (BGS) in the early 1960s (Survey Division of the Ministry of Lands, 1966). The surface area for both the 2013 Landsat image (10.90 km2) and the lake as depicted by the BGS (0.75 km2) were calculated through manual delineation of shorelines. As the difference was substantial, a more complete record of modern variability using the full range of archived Landsat images (as available via USGS Earth Explorer in August 2015) was constructed. The discovery of these considerable variations in size was also enlightening for the paleoenvironmental reconstruction in paper I and lends to the hypothesis that the lake is highly dynamic.

51 7 Conclusions

The interdisciplinary approach adopted in this dissertation has contributed to a better understanding of how Lake Basotu responds to both natural and anthropogenic forcing and provides a guideline for future research in similar settings.

The objectives stated in chapter 1.1 were addressed as follows:

Objective #1: Show how water level has fluctuated in Lake Basotu over the past 1000 years, to establish a baseline for natural variability.

To assess the natural variability of Lake Basotu, an environmental reconstruction was completed using proxy data; specifically diatoms, carbon content, and mineral magnetic analysis. Sediment was dated using the radiometric methods of radiocarbon and lead-210. Due to a reversal in radiocarbon dates, an inferred chronology was developed using the available proxy data and a history of regional famines. Accordingly, Lake Basotu has shown sensitivity to drought conditions with the potential to dry out to a marsh-like state, at least since the end of the Medieval Climate Anomaly (c. AD 1250). It is likely that the MCA lead to Lake Basotu completely drying, as is evident by the desiccation of sediment below 250 cm, and for at least 50 cm.

A similar pattern in the radiocarbon dates of material from Lake Babati reinforces the hypothesis that a regional event is responsible for the dating reversals, rather than user mishandling of material.

Objective #2: Reconstruct a history of modern lake extent (1973–2015) and identify potential drivers of observed changes.

Landsat satellite images show three periods of declining lake extent, separated by flooding events in 1997 and 2006. It was determined that since 1996 Lake Basotu has been more sensitive to extreme precipitation, which often accompanies the co-occurrence of a positive Indian Ocean Dipole and an El Niño event. Flooding observed in recent years follows an attempt to install mechanized wheat farming in the area surrounding Lake Basotu and is a likely effect of siltation of the Basotu basin due to inadequate soil conservation measures.

Objective #3: Show if and how land-use practices in the area surrounding Lake Basotu may have influenced its function as a freshwater resource.

The history of human agricultural and pastoral activity around Lake Basotu was compared to the chronology of lake extent changes to look for commonalities.

52 An internet search of newspaper articles (2006–2016) and foreign aid reports provided the basis for this comparison. Telephone interviews with informants living and working in the area around Lake Basotu provided a more complete picture of farming practices. Data indicate that mechanized wheat farming in the district surrounding Lake Basotu has had negative effects. Diverting water during the growing season for watering of wheat crops likely led to the diminishing lake extent observed in paper II. Furthermore, as cropland was left barren after harvesting, and during part of the rainy season, loose top soil could then be washed into Lake Basotu. The case of Lake Basotu highlights the necessity of taking a broad perspective in environmental planning. Primarily through inadequate soil conservation measures, the health of a vital freshwater resource for many people was put at risk.

53 8 Acknowledgments

I would not be here without a few rock star women in science who inspired me to pursue a PhD during my “younger” academic years; thank you Kelly M. Frothingham, Camille A. Holmgren, Lisa Marie Anselmi, and Ellen Mosley-Thompson. Thank you to my supervisors, Lars-Ove Westerberg and Jan Risberg, for giving me the opportunity to conduct this research, and for your time and effort in the production of this dissertation. I would also like to thank Karin Holmgren and Malin Kylander for their support and encouragement during my time at Stockholm University. I have many others to thank for their assistance with this project. The text of this kappa was improved by comments and suggestions from Ninis Rosqvist, Regina Lindborg, and Peter Jansson. I thank Meighan Boyd, Natasha Webster, and Benedict Reinardy for their time spent proofreading. For their help in procuring data for my analyses, I thank Alexander Koutsouris (for precipitation records and watershed mapping) and Qiong Zhang (for calculating DMI values). Ian Snowball, your insights and assistance with the interpretation of my mineral magnetic data have been incredibly appreciated, and I look forward to future collaborations. Martina A. Caretta, your drive and work ethic are an inspiration to me; thank you for being my cross-disciplinary guru. Further thanks go to the constantly amazing library staff of Stockholm University, and their detective like ability to find the most obscure of texts. Funding for this project has been provided by the Bolin Centre for Climate Research. Field work in Tanzania would not have been possible without the generous funding by the Swedish Society for Anthropology and Geography, on- site support by Alfred Muzuka (The Nelson Mandela African Institution of Science and Technology, Arusha), and the British Institute in East Africa. Analytical costs have been kindly supported by the Swedish Society for Anthropology and Geography, Ahlmanns Fund, and Research Area 5 of the Bolin Centre for Climate Research. Support for attending conferences has been provided by the Climate Research School and Research Area 5 of the Bolin Centre for Climate Research. Although sediment material collected from Kenya was not used in the writing of this dissertation, I owe many thanks to David M. Mburu (Jomo Kenyatta University of Agriculture and Technology) for his assistance in the field and action when I fell ill. I also thank the Albert & Maria Bergströms Fund for financing Kenyan fieldwork and the Carl Mannerfelts Fund for analytical support.

And now the mushy stuff…

My time as a PhD student at Stockholm University has been both challenging and rewarding. With the support of colleagues, friends, and family, I was able to come back from tragedy to produce something I am very proud of. So it goes. First, I would like to thank and acknowledge my family and friends in the United States. Without their support, none of this would have been possible. I

54 love and miss you all. Thank you for the care packages, video chats, letters, and visits. To my sambo and rock, Patrik Mattsson. You came into my life at a time when I thought I had no more love left to give, and showed me it was possible. Your love and patience with me during the push to finish were endless. I know this academic universe is a strange place to you, and I deeply appreciate your support. I love you. Muito obrigado to Eva Rocha who has been the best office-mate I could ask for during these last years. I appreciate the support, laughs, and sushi dates. I hope that banana stays lost.

To everyone: Thank you very much Asante sana Tack så mycket Muito obrigado 㠀ᖖឤ寊ġ ǼȣȤĮȡȚıIJȫʌȠȜȪ Grazie mille

“I do have hope. Nature is enormously resilient, humans are vastly intelligent, the energy and enthusiasm that can be kindled among young people seems without limit, and the human spirit is indomitable. But if we want life, we will have to stop depending on someone else to save the world. It is up to us-you and me, all of us. Myself, I have placed my faith in the children.” -Jane Goodall (Alexandra Morton, Listening to Whales: What the Orcas Have Taught Us, 2002)

55 9 References

Al-Handal, A. Y., & Abdullah, D. S. (2010). Diatoms from the restored Mesopotamian marshes, South Iraq. Algological studies, 133, 65–103. Appleby, P. G. (1997). Dating recent sediments by 210Pb: Problems and solutions. In Proceedings of the 2nd NKS/EKO-1 Seminar, Helsinki (Vol. 2, No. 4). Arndt, C., Farmer, W., Strzepek, K., & Thurlow, J. (2012). Climate change, agriculture and food security in Tanzania. Review of Development Economics, 16, 378–393. Arusha Times (30 July–5 August, 2011a). Generators ‘disturbing Lake Bassuto gods and spirits’. Retrieved from www.arushatimes.co.tz/2011/28/Local%20News_5.htm Arusha Times (13–19 August, 2011b). Local conflicts fueled by politicians - claim. Retrieved from www.arushatimes.co.tz/2011/30/Local%20News_4.htm Arusha Times (10–16 December, 2011c). RC Mbwilo wants lasting solution for Lake Bassotu crisis. Retrieved from http://www.arushatimes.co.tz/2011/45/Local%20News_7.html Ballatore, T. J., Bradt, S. R., Olaka, L., Cózar, A., & Loiselle, S. A. (2014). Remote Sensing of African Lakes: A Review. In V. Barake & M. Gade (Eds.) Remote Sensing of the African Seas (pp. 403–422). The Netherlands: Springer. Banerjee, S. K. & Moskowitz, B. M. (1985). Ferrimagnetic Properties of Magnetite. In J. L. Kirschvink, D. S. Jones, & B. J. MacFadden (Eds.) Magnetite Biomineralization and Magnetoreception in Organisms: A New Biomagnetism (pp. 17–41). United States: Springer. Barker, P. (1992). Differential diatom dissolution in Late Quaternary sediments from Lake Manyara, Tanzania: an experimental approach. Journal of Paleolimnology, 7, 235–251. Barker, P., Telford, R., Gasse, F., & Thevenon, F. (2002). Late Pleistocene and Holocene palaeohydrology of Lake Rukwa, Tanzania, inferred from diatom analysis. Palaeogeography, Palaeoclimatology, Palaeoecology, 187, 295–305. Barton, C. E., & Torgersen, T. (1988). Palaeomagnetic and 210Pb estimates of sedimentation in Lake Turkana, East Africa. Palaeogeography, Palaeoclimatology, Palaeoecology, 68, 53–59. Bassett, T. J., & Fogelman, C. (2013). Déjà vu or something new? The adaptation concept in the climate change literature. Geoforum, 48, 42–53. Battarbee, R. W. (1986). Diatom Analysis. In B. E. Berglund (Ed.) Handbook of Holocene Palaeoecology and Palaeohydrology (pp. 527–570). Chichester: John Wiley & Sons Ltd. Battarbee, R. W., Jones, V. J., Flower, R. J., Cameron, N. G., Bennion, H., Carvalho, L., & Juggins, S. (2001). Diatoms. In J. P. Smol, H. J. B. Birks & W. M. Last (Eds.). Tracking Environmental Change Using Lake Sediments. Volume 3: Terrestrial, Algal, and Siliceous Indicators (pp. 155–202). The Netherlands: Kluwer Academic Publishers. Becker, E. (2016). January El Niño update: It’s got a lot going on. ENSO Blog, National Oceanic and Atmospheric Adminstration (NOOA). Retrieved from: https://www.climate.gov/news-features/blogs/enso/january-el-ni%C3%B1o- update-it%E2%80%99s-got-lot-going

56 Begy, R., Cosma, C., & Timar, A. (2009). Recent changes in Red Lake (Romania) sedimentation rate determined from depth profiles of 210 Pb and 137 Cs radioisotopes. Journal of Environmental Radioactivity, 100, 644–648. Bender, M. M. (1968). Mass spectrometric studies of carbon 13 variations in corn and other grasses. Radiocarbon, 10, 468–472. Bere, T., Phiri, C., Kadye, W. T., & Utete, B. (2013). Benthic diatom assemblages in mountain streams: community structure in relation to environmental and human pressures. African Journal of Ecology, 51, 625–634. Beuning, K. R., Talbot, M. R., & Kelts, K. (1997). A revised 30,000-year paleoclimatic and paleohydrologic history of Lake Albert, East Africa. Palaeogeography, Palaeoclimatology, Palaeoecology, 136, 259–279. BGS (2011). An initial estimate of depth to groundwater across Africa. Groundwater Science Programme Open Report. British Geological Survey. Retrieved from: http://nora.nerc.ac.uk/17907/1/OR11067.pdf Birks, H. H., & Birks, H. J. B. (2006). Multi-proxy studies in palaeolimnology. Vegetation History and Archaeobotany, 15, 235–251. Björck, S., Bennike, O., Possnert, G., Wohlfarth, B., & Digerfeldt, G. (1998). A high- resolution14C dated sediment sequence from southwest Sweden: age comparisons between different components of the sediment. Journal of Quaternary Science, 13, 85–89. Björck, S., & Wohlfarth, B. (2001). 14C chronostratigraphic techniques in paleolimnology. In W. M. Last & J. P. Smol (Eds.) Tracking environmental change using lake sediments. Volume 1: Basin Analysis, Coring, and Chronological Techniques. (pp. 205–245). The Netherlands: Springer. Boar, R. R., & Harper, D. M. (2002). Magnetic susceptibilities of lake sediment and soils on the shoreline of Lake Naivasha, Kenya. In D. M. Harper, R. R. Boar, M. Everard, & E. P. Hickley (Eds.) Lake Naivasha, Kenya (pp. 81–88). The Netherlands: Springer. Böhme, B., Steinbruch, F., Gloaguen, R., Heilmeier, H., & Merkel, B. (2006). Geomorphology, hydrology, and ecology of Lake Urema, central Mozambique, with focus on lake extent changes. Physics and Chemistry of the Earth, Parts A/B/C, 31, 745–752. Boutton, T. W. (1991). Stable carbon isotope ratios of natural materials: I. Sample preparation and mass spectrometric analysis. Carbon isotope techniques, 1, 155. Boyd, C. E. (2015). Water Quality: An Introduction (2nd ed., 357 pp.). Switzerland: Springer International Publishing. Bradbury, J. P., & Megard, R. O. (1972). Stratigraphic record of pollution in Shagawa Lake, northeastern Minnesota. Geological Society of America Bulletin, 83, 2639– 2648. Bradley, R. S. (2014). Paleoclimatology: reconstructing climates of the quaternary. (3rd ed., 675 pp.) Oxford: Academic Press. Brenner, M., Peplow, A. J., & Schelske, C. L. (1994). Disequilibrium between 226Ra and supported 210Pb in a sediment core from a shallow Florida lake. Limnology and Oceanography, 39, 1222–1227. Bronk Ramsey, C. (1995). Radiocarbon calibration and analysis of stratigraphy: The OxCal program. Radiocarbon, 37, 425–430. Bronk Ramsey, C. (2009). Bayesian analysis of radiocarbon dates. Radiocarbon, 51, 337–360. Brown, E. T., & Johnson, T. C. (2005). Coherence between tropical East African and South American records of the Little Ice Age. Geochemistry, Geophysics, Geosystems, 6, 11 pp.. Brunel Microscopes Ltd. Naphrax Diatom Mountant. Retrieved from http://www.brunelmicroscopes.co.uk/naphrax.html

57 Burke, L. A., & Miller, M. K. (2001, May). Phone Interviewing as a Means of Data Collection: Lessons Learned and Practical Recommendations. Forum Qualitative Sozialforschung/Forum: Qualitative Social Research, 2. Camberlin, P., Moron, V., Okoola, R., Philippon, N., & Gitau, W. (2009). Components of rainy seasons’ variability in Equatorial East Africa: onset, cessation, rainfall frequency and intensity. Theoretical and Applied Climatology, 98, 237–249. Canning, D., Raja, S., & Yazbeck, A. S. (Eds.). (2015). Africa's Demographic Transition: Dividend Or Disaster?. Washington, DC: World Bank Publications. Carey, A. M., & Paige, G. B. (2016). Ecological Site-Scale Hydrologic Response in a Semiarid Rangeland Watershed. Rangeland Ecology & Management, 69, 481–490. Chale, F. M. M. (1989). Some ecological studies on Oreochromis esculentus Graham (Pisces: Cichlidae) in two small tropical lakes. Indian Journal of Fisheries, 36, 41– 46. Charles, D. F. & Smol, J. P. (1988). New methods for using diatoms and chrysophytes to infer past pH of low-alkalinity lakes. Limnology and Oceanography, 33, 1451– 1462. Cioffi, F., Conticello, F., & Lall, U. (2016). Projecting changes in Tanzania rainfall for the 21st century. International Journal of Climatology, 36, 4297–4314. The Citizen (9 February, 2013). District authorities fail to protect Lake Bassotu: councillor. Retrieved from http://www.thecitizen.co.tz/News/District-authorities- fail-to-protect-Lake-Bassotu/-/1840392/1976394/-/2h8so2/-/index.html Cohen, A. S. (2003). Sedimentological Archives in Lake Deposits. In Paleolimnology: the history and evolution of lake systems (pp. 162–207). Oxford University Press. Cole, J. J., Prairie, Y. T., Caraco, N. F., McDowell, W. H., Tranvik, L. J., Striegl, R. G., Duarte, C.M., Kortelainen, P., Downing, J.A., Middelburg, J.J., & Melack, J. (2007). Plumbing the global carbon cycle: integrating inland waters into the terrestrial carbon budget. Ecosystems, 10, 172–185. Cumming, B. F., & Smol, J. P. (1993). Development of diatom-based salinity models for paleoclimatic research from lakes in British Columbia (Canada). In Twelfth international diatom symposium (pp. 179–196). The Netherlands: Springer. Daly, M. E., Yanda, P. Z., & West, J. J. (2015). Climate change policy inventory and analysis for Tanzania. CICERO Report. Davies, S. J., Metcalfe, S. E., MacKenzie, A. B., Newton, A. J., Endfield, G. H., & Farmer, J. G. (2004). Environmental changes in the Zirahuén Basin, Michoacán, Mexico, during the last 1000 years. Journal of Paleolimnology, 31, 77–98. Dearing, J. A., Braimoh, A. K., Reenberg, A., Turner, B. L., & van der Leeuw, S. (2010). Complex land systems: the need for long time perspectives to assess their future. Ecology and Society, 15, 21. Deevey, E. S., Gross, M. S., Hutchinson, G. E., & Kraybill, H. L. (1954). The natural C14 contents of materials from hard-water lakes. Proceedings of the National Academy of Sciences, 40, 285-288. Dekkers, M. J. (2007). Magnetic proxy parameters. In D. Gubbins & E. Herrero-Bervera (Eds.) Encyclopedia of Geomagnetism and Paleomagnetism (pp. 525–534). The Netherlands: Springer. Dixit, S. S., Smol, J. P., Kingston, J. C., & Charles, D. F. (1992). Diatoms: powerful indicators of environmental change. Environmental Science & Technology, 26, 22– 33. Dodds, W. K. (2002). Physiography of Lakes and Reservoirs. In Freshwater ecology: concepts and environmental applications (pp. 91–112). Academic press. Downie, C., & Wilkinson, P. (1962). The explosion craters of Basotu, Tanganyika Territory. Bulletin of Volcanology, 24, 389–420. Duchesne, L. C., & Larson, D. W. (1989). Cellulose and the evolution of plant life. Bioscience, 39, 238–241.

58 Ehleringer, J. R., Sage, R. F., Flanagan, L. B., & Pearcy, R. W. (1991). Climate change and the evolution of C4 photosynthesis. Trends in Ecology & Evolution, 6, 95–99. El-Daoushy, F., & Eriksson, M. G. (1998). Radiometric dating of recent lake sediments from a highly eroded area in semiarid Tanzania. Journal of Paleolimnology, 19, 377–383. ELTRA (2004). CS-500 Operation Manual. Retrieved from https://tp.dresden- concept.de/userdata/7049ca454ff977f8235a8e1ce3f465fc.pdf EPA (1999). Innovative Technology Verification Report Sediment Sampling Technology Aquatic Research Instruments Russian Peat Borer. United States Environmental Protection Agency. Retrieved from https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P10011T1.txt Eriksson, M. G., & Sandgren, P. (1999). Mineral magnetic analyses of sediment cores recording recent soil erosion history in central Tanzania. Palaeogeography, Palaeoclimatology, Palaeoecology, 152, 365–383. ESRI (2015). ArcMap 10.3.1; ESRI: Redlands, CA, USA, 2015. Environmental Systems Resource Institute. FAO (2003). Review of World Water Resources by Country [report]. Food and Agriculture Organization of the United Nations. Retrieved from ftp://ftp.fao.org/agl/aglw/docs/wr23e.pdf FAO (2015). Review of Climate Change Adaptation and Mitigation in Agriculture in the United Republic of Tanzania [report]. Food and Agriculture Organization of the United Nations. Retrieved from http://www.fao.org/3/a-i5042e.pdf Fenger, M., Hignett, V., & Green, A. (1986). Soils of the Basotu and Balangida Lelu areas of northern Tanzania and their suitability for mechanized dryland farming. Agriculture Canada. Flower, R. J. (1993). Diatom preservation: experiments and observations on dissolution and breakage in modern and fossil material. In Twelfth International Diatom Symposium (pp. 473–484). The Netherlands: Springer. Flower, R. J. (2017). A personal perspective on four decades of paleolimnology and environmental change research. Journal of Paleolimnology, 57, 1-17. Flower, R. J. & Ryves, D. B. (2009). Diatom preservation: differential preservation of sedimentary diatoms in two saline lakes. Acta Botanica Croatica, 68, 381–399. Folster, N. (2001). Systemic constraints on aid policy and aid outcomes: the history of Canadian official development assistance to Tanzania (Doctoral dissertation, London School of Economics and Political Science (United Kingdom)). Füssel, H. M. (2007). Adaptation planning for climate change: concepts, assessment approaches, and key lessons. Sustainability science, 2, 265-275. Fyumagwa, R. D., Mwakalebe, G., Mwita, M., & Mmassy, E. C. (2014). Lake Basotu and Lake Tlawi are essential wetlands for waterbirds. South African Journal of Wildlife Research, 44, 84–89. Gasse, F. (2002). Diatom-inferred salinity and carbonate oxygen isotopes in Holocene waterbodies of the western Sahara and Sahel (Africa). Quaternary Science Reviews, 21, 737–767. Gasse, F., Barker, P., Gell, P. A., Fritz, S. C., & Chalie, F. (1997). Diatom-inferred salinity in palaeolakes: an indirect tracer of climate change. Quaternary Science Reviews, 16, 547–563. Gasse, F., Juggins, S., & Khelifa, L. B. (1995). Diatom-based transfer functions for inferring past hydrochemical characteristics of African lakes.Palaeogeography, Palaeoclimatology, Palaeoecology, 117, 31–54. Gasse, F., Talling, J. F., Kilham, P. (1983). Diatom assemblages in East Africa: classi- fication, distribution and ecology. Revue d'Hydrobiologie Tropicale, 16, 3–34. Gasse, F. & Van Campo, E. (2001). Late Quaternary environmental changes from a pollen and diatom record in the southern tropics (Lake Tritrivakely, Madagascar). Palaeogeography, Palaeoclimatology, Palaeoecology, 167, 287–308.

59 Gell, P. A., Sluiter, I. R., & Fluin, J. (2002). Seasonal and interannual variations in diatom assemblages in Murray River connected wetlands in north-west Victoria, Australia. Marine and Freshwater Research, 53, 981–992. Gerdén, C. A., Khawange, G. M. O., Mallya, J. M., Mbuya, J. P., & Sanga, R. C. (1992). The Wild Lake: The 1990 floods in Babati, Tanzania–rehabilitation and prevention (No. 2). RCU Report. Geyh, M. A., Schotterer, U., & Grosjean, M. (1998). Temporal changes of the 14 C reservoir effect in lakes. Radiocarbon, 40, 921–931. Giardino, C., Bresciani, M., Villa, P., & Martinelli, A. (2010). Application of remote sensing in water resource management: the case study of Lake Trasimeno, Italy. Water Resources Management, 24, 3885–3899. Goward, S., Arvidson, T., Williams, D., Faundeen, J., Irons, J., & Franks, S. (2006). Historical Record of Landsat Global Coverage. Photogrammetric Engineering & Remote Sensing, 72, 1155–1169. Greenland, D. J., & Pereira, H. C. (1977). Soil damage by intensive arable cultivation: Temporary or permanent? [and discussion]. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 281, 193–208. Greenway, P. J. (1955). Ecological observations on an extinct East African volcanic mountain. The Journal of Ecology, 544–563. Greenwood, P. H. (1953). Feeding mechanism of the cichlid fish, Tilapia esculenta Graham. Nature, 172, 207–208. Grimm, E. C. (1991). Tilia and Tiliagraph. Illinois State Museum, Springfield. Grimm, E. C., Maher, L. J., & Nelson, D. M. (2009). The magnitude of error in conventional bulk-sediment radiocarbon dates from central North America. Quaternary Research, 72, 301–308 Gu, B., Schelske, C. L., & Brenner, M. (1996). Relationship between sediment and SODQNWRQ LVRWRSH UDWLRV į& DQG į1  DQG SULPDU\ SURGuctivity in Florida lakes. Canadian Journal of Fisheries and Aquatic Sciences, 53, 875–883. Hassan, G. S. (2015). On the benefits of being redundant: low compositional fidelity of diatom death assemblages does not hamper the preservation of environmental gradients in shallow lakes. Paleobiology, 41, 154–173. Hastenrath, S. (2007). Circulation mechanisms of climate anomalies in East Africa and the equatorial Indian Ocean. Dynamics of Atmospheres and Oceans, 43, 25–35. Heaton, T. H. (1986). Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere: a review. Chemical Geology: Isotope Geoscience Section, 59, 87–102. Hecky, R. E. & Kilham, P. (1973). Diatoms in Alkaline, Saline Lakes: Ecology and Geochemical Implications. Limnology and Oceanography, 18, 53–71. Heiri, O., Lotter, A. F., & Lemcke, G. (2001). Loss on ignition as a method for estimating organic and carbonate content in sediments: reproducibility and comparability of results. Journal of Paleolimnology, 25, 101–110. Hesse, P., & Stolz, J. F. (1999). Bacterial magnetite and the Quaternary climate record. Quaternary Climates, Environments and Magnetism, 390, 161–198. Higgins, L., Koutsouris, A. J., Westerberg, L. O., & Risberg, J. (2016). Surface Area Variability of a North-Central Tanzanian Crater Lake. Geosciences, 6, 11pp.. Hogg, A. G., Hua, Q., Blackwell, P. G., Niu, M., Buck, C. E., Guilderson, T. P., Heaton, T. J., Palmer, J. G., Reimer, P. J., Reimer, R. W., Turney, C. S. M., & Zimmerman, S. R. H. (2013). SHCal13 Southern Hemisphere Calibration, 0-50,000 Years cal BP. Radiocarbon, 55, 1–15. Holmgren, M. & Scheffer, M. (2001). El Niño as a window of opportunity for the restoration of degraded arid ecosystems. Ecosystems, 4, 151–159. Holmgren, K., Risberg, J., Freudendahl, J., Achimo, M., Ekblom, A., Mugabe, J., Nor- ström, E., & Sitoe, S. (2012). Water-level variations in Lake Nhauhache, Mozam- bique, during the last 2,300 years. Journal of Paleolimnology, 48, 311–322.

60 Iler, R.K. (1979). The Chemistry of Silica: Solubility, Polymerization, Colloid and Surface Properties and Biochemistry of Silica (896 pp.). New York: Wiley. IPCC (2013). Climate Change 2013: The Physical Science Basis. In T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P. M. Midgley (Eds.) Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (1535 pp). Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press. Ivanovich, M., & Harmon, R. S. (1992). Uranium-series Disequilibrium: Applications to Earth, Marine, and Environmental Sciences (2nd ed, 910 pp.). Oxford: Clarendon Press. Johnson, G. C. & Parsons, A. R. (Eds.) (2016). Global Oceans. In J. Blunden & D. S. Arndt (Eds.) State of the Climate in 2015. Bulletin of the American Meteorological Society, 97, 63–91. de Jonge, V. N., Elliott, M., & Orive, E. (2002). Causes, historical development, effects and future challenges of a common environmental problem: eutrophication. In M. Elliot & E. Orive (Eds.) Nutrients and Eutrophication in Estuaries and Coastal Waters (pp. 1–19). The Netherlands: Springer. Juhola, S., & Kruse, S. (2015). A framework for analysing regional adaptive capacity assessments: challenges for methodology and policy making. Mitigation and Adaptation Strategies for Global Change, 20, 99–120. Kahurananga, J. (1992). Lake Babati, Tanzania and Its Immediate Surroundings. Part I: Baseline Information. Regional Soil Conservation Unit (SIDA). Nairobi, Kenya. Kebede, S., Lamb, H., Telford, R., Leng, M., & Umer, M. (2002). Lake-groundwater relationships, oxygen isotope balance and climate sensitivity of the Bishoftu crater lakes, Ethiopia. In E. O. Odada & D. O. Olago (Eds.) The East African Great Lakes: Limnology, Palaeolimnology and Biodiversity (pp. 261–275). The Netherlands: Kluwer Academic Publishers. Kiage, L. M. & Liu, K. B. (2009). Palynological evidence of climate change and land degradation in the Lake Baringo area, Kenya, East Africa, since AD 1650. Palaeogeography, Palaeoclimatology, Palaeoecology, 279, 60–72. Kijazi, A. L. & Reason, C. J. C. (2009). Analysis of the 2006 floods over northern Tanzania. International Journal of climatology, 29, 955–970. Kilham, P. & Cloke, P. L. (1990). The evolution of saline lake waters: gradual and rapid biogeochemical pathways in the Basotu Lake District, Tanzania. Saline Lakes (pp. 35–50). The Netherlands: Springer. Kinnell, P. I. A. (2005). Raindrop-impact-induced erosion processes and prediction: a review. Hydrological processes, 19, 2815–2844. Kókai, Z., Bácsi, I., Török, P., Buczkó, K., Balogh, C., & Tóthmérész, B. (2015). Halophilic diatom taxa are sensitive indicators of even short term changes in lowland lotic systems. Acta Botanica Croatica, 74, 287–302. Koutsouris, A. J., Chen, D., & Lyon, S. W. (2015). Comparing global precipitation data sets in eastern Africa: a case study of Kilombero Valley, Tanzania. International Journal of Climatology, no vol., 15pp.. Krabbenhoft, D. P., Bowser, C. J., Anderson, M. P., & Valley, J. W. (1990). Estimating groundwater exchange with lakes: 1. The stable isotope mass balance method. Water Resources Research, 26, 2445–2453. Krammer, K. & Lange-Bertalot, H. (1986). Süßwasserflora von Mitteleuropa. Band 2. Bacillariophyceae. Teil 1. Naviculaceae. Germany: Gustav Fischer Verlag, Stuttgart. Krammer, K. & Lange-Bertalot, H. (1988). Süßwasserflora von Mitteleuropa. Band 2. Bacillariophyceae. Teil 2. Bacillariaceae, Epithemiaceae, Surirellaceae. Germany: Gustav Fischer Verlag, Stuttgart.

61 Krammer, K. & Lange-Bertalot, H. (1991a). Süßwasserflora von Mitteleuropa. Band 2. Bacillariophyceae. Teil 3. Centrales, Fragilariaceae, Eunotiaceae. Germany: Gustav Fischer Verlag, Stuttgart. Krammer, K. & Lange-Bertalot, H. (1991b). Süßwasserflora von Mitteleuropa. Band 2. Bacillariophyceae. Teil 4. Achnanthaceae, Kritische Ergänzungen zu Navicula (Lineolatae) und Gomphonema. Germany: Gustav Fischer Verlag, Stuttgart. Krausse, G. L., Schelske, C. L., & Davis, C. O. (1983). Comparison of three wet- alkaline methods of digestion of biogenic silica in water. Freshwater Biology, 13, 73–81. Krishnaswami, S. & Lal, D. (1978). Radionuclide limnochronology. In A. Lerman (Ed.) Lakes (pp. 153–177). New York: Springer. Lambin, E. F. (2005). Conditions for sustainability of human–environment systems: Information, motivation, and capacity [editorial]. Global Environmental Change, 15, 177–180. Land Resource Research Institute (1986). Soils of the Basotu Area, Map Sheet 84/3 Tanzania [map]. (1:50,000). Agriculture Canada. Retrieved from http://eusoils.jrc.ec.europa.eu/content/soils-basotu-area-map-sheet-843 Lane, C. R. (1990). Barabaig Natural Resource Management: Sustainable Land Use under Threat of Destruction. United Nations Research Institute for Social Development. Discussion Paper 12. Retrieved from http://www.unrisd.org/unrisd/website/document.nsf/(httpPublications)/CAB206729 493A99580256B67005B6057?OpenDocument. Lane, C. R. (1991) Wheat at What Cost? CIDA and the Tanzania-Canada Wheat Program. In J. Swift & B. Tomlinson (Eds.) Conflicts of Interest: Canada and the Third World (pp 111–160). Toronto, Ontario: Between the Lines. Lane, C. R. (1994). Pastures lost: alienation of Barabaig land in the context of land policy and legislation in Tanzania. Nomadic Peoples, 81–94. Lane, P. J. (2010). Developing landscape historical ecologies in eastern Africa: an outline of current research and potential future directions. African Studies, 69, 299– 322. Lane, P. J. (2015a). Just how long does ‘long-term’have to be?: Matters of temporal scale as impediments to interdisciplinary understanding in historical ecology. In C. Isendahl & D. Stump (Eds.) The Oxford Handbook of Historical Ecology and Applied Archaeology. Oxford Handbooks Online. Lane, P. J. (2015b). Archaeology in the age of the Anthropocene: A critical assessment of its scope and societal contributions. Journal of Field Archaeology, 40, 485–498. Lane, P. J. (2015c). Sustainability: Primordial conservationists, environmental sustainability and the rhetoric of pastoralist cultural heritage in East Africa. In K. L. Samuels & T. Rico (Eds.) Heritage Keywords: Rhetoric and Redescription in Cultural Heritage (pp 259–283). Boulder, Colorado: University Press of Colorado. Lecoutere, E. (2011). Institutions under construction: resolving resource conflicts in Tanzanian irrigation schemes. Journal of Eastern African Studies, 5, 252–273. Leng, M. J. & Marshall, J. D. (2004). Palaeoclimate interpretation of stable isotope data from lake sediment archives. Quaternary Science Reviews, 23, 811–831. Li, C. J., Chai, Y. Q., Yang, L. S., & Li, H. R. (2016). Spatio-temporal distribution of flood disasters and analysis of influencing factors in Africa. Natural Hazards, 82, 721–731. Liu, J., Dietz, T., Carpenter, S. R., Alberti, M., Folke, C., Moran, E., Pell, A. N., Deadman, P., Kratz, T., Lubchenco, J., Ostrom, E., Ouyang, Z., Provencher, W., Redman, C. L., Schneider, S. H., & Taylor, W. W. (2007). Complexity of coupled human and natural systems. Science, 317, 1513–1516. Liu, Q., Roberts, A. P., Larrasoana, J. C., Banerjee, S. K., Guyodo, Y., Tauxe, L., & Oldfield, F. (2012). Environmental magnetism: principles and applications. Reviews of Geophysics, 50.

62 Lowe, R. H. (1959). Breeding behaviour patterns and ecological differences between Tilapia species and their significance for evolution within the genus Tilapia (Pisces: Cichlidae). Proceedings of the Zoological Society of London, 132, 1–30. Lyakurwa, F. S., Song, G., & Chen, J. (2014). Quantitative modeling of freshwater stress in the nine water basins of Tanzania. Chinese Journal of Population Resources and Environment, 12, 309–315. Lydig, A. (2009). Factors conditioning the distribution of fresh water pulmonates, Biomphalaria spp., Bulinus spp., and Lymnea spp., in Babati District, Tanzania. Bachelor’s Thesis. Södertörns Högskola. MacDonald, A. M., Bonsor, H. C., Dochartaigh, B. É. Ó., & Taylor, R. G. (2012). Quantitative maps of groundwater resources in Africa. Environmental Research Letters, 7, 8pp.. MacMillan, R. A., & Ngoma, A. (1983). Vertisolic Soils of the Hanang Wheat Complex, Tanzania: Characteristics, Distribution, Use and Management. East Africa Soil Correlation Committee. Retrieved from http://www.wossac.com/search/wossac_detail.cfm?ID=42023 Marchetto, A., Ariztegui, D., Brauer, A., Lami, A., Mercuri, A.M., Sadori, L., Vigliotti, L., Wulf, S. & Guilizzoni, P. (2015). Volcanic lake sediments as sensitive archives of climate and environmental change. In D. Rouwet, B. Christenson, F. Tassi, & J. Vandemeulebrouck (Eds.) Volcanic Lakes (pp. 379–399). Berlin Heidelberg: Springer. Markham, B. L., Storey, J. C., Williams, D. L., & Irons, J. R. (2004). Landsat sensor performance: history and current status. IEEE Transactions on Geoscience and Remote Sensing, 42, 2691–2694. Martin, U., Németh, K., Lorenz, V., & White, J. D. (2007). Introduction: Maar-diatreme volcanism. Journal of Volcanology and Geothermal Research, 159, 1–3. Mbanguka, R. P., Lyon, S. W., Holmgren, K., Girons Lopez, M., & Jarsjö, J. (2016). Water Balance and Level Change of Lake Babati, Tanzania: Sensitivity to Hydroclimatic Forcings. Water, 8, 21 pp.. McLeman, R. A. (2011). Settlement abandonment in the context of global environmental change. Global Environmental Change, 21, S108–S120. Mdadisi, U. (2009). The State of the then NAFCO, NARCO and Absentee Landlords’ Farms/Ranches in Tanzania. Land Rights Research and Resources Institute. Retrieved from http://www.academia.edu/28534494/The_State_of_NAFCO_Farms_and_NARCO _Ranches_in_Tanzania Meyers, P. A. (1994). Preservation of elemental and isotopic source identification of sedimentary organic matter. Chemical Geology, 114, 289–302, Meyers, P. A. & Ishiwatari, R. (1993). Lacustrine organic geochemistry–an overview of indicators of organic matter sources and diagenesis in lake sediments. Organic geochemistry, 20, 867–900. Meyers, P. A. & Teranes, J. L. (2001). Sediment organic matter. In Tracking environmental change using lake sediments. Volume 2: Physical and Geochemical Methods (pp. 239–269). The Netherlands: Springer. Milne, G. (1947). A soil reconnaissance journey through parts of Tanganyika Territory December 1935 to February 1936. Journal of Ecology, 35, 192-265. Moyle, J. B. (1949). Some Indices of Lake Productivity, Transactions of the American Fisheries Society, 76, 322–334. Munro, J. L. (1967). The food of a community of East African freshwater fishes. Journal of Zoology, 151, 389–415 Naiman, R. J., Magnuson, J. J., & Firth, P. L. (1998). Integrating cultural, economic, and environmental requirements for fresh water. Ecological Applications, 8, 569– 570.

63 Nambudiri, E. M. V., Teller, J. T., & Last, W. M. (1980). Pre-Quaternary microfossils– a guide to errors in radiocarbon dating. Geology, 8, 123–126. NASA (2015). Landsat 9 is Coming. National Aeronautics and Space Administration. Retrieved from http://landsat.gsfc.nasa.gov/nasa-usgs-begin-work-on-landsat-9-to- continue-land-imaging-legacy/ NASA (2017a). Landsat Science: History. National Aeronautics and Space Administration. Retrieved from http://landsat.gsfc.nasa.gov/about/history/ NASA (2017b). A Landsat Timeline. National Aeronautics and Space Administration. Retrieved from http://landsat.gsfc.nasa.gov/a-landsat-timeline/ National Audit Office (2012). Floods in Babati: A Performance Audit of the Management of Prevention and Mitigation of Floods at Central, Regional and Local Levels of the Government of Tanzania. A report from the controller and auditor general of Tanzania. The United Republic of Tanzania. NBS (2013). 2012 Population and Housing Census, Population Distribution by Administrative Areas. Dar es Salaam: National Bureau of Statistics, Ministry of Finance. National Bureau of Statistics. Retrieved from http://ihi.eprints.org/1344/ NERC (2012). Download digital groundwater maps of Africa. Retrieved from http://www.bgs.ac.uk/research/groundwater/international/africangroundwater/maps Download.html. Natural Environment Research Council. Nelson, J. S., Grande, T. C., & Wilson, M. V. (2016). Fishes of the World (5th ed, p 230). Hoboken, New Jersey, USA: John Wiley & Sons. Nicholson, S. E. & Kim, J. (1997). The relationship of the El Niño-Southern oscillation to African rainfall. International Journal of Climatology, 17, 117–135. Noble, I.R., S. Huq, Y.A. Anokhin, J. Carmin, D. Goudou, F.P. Lansigan, B. Osman- Elasha, & A. Villamizar (2014). Adaptation needs and options. In C.B. Fields, V.R. Barros, D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Mastrandrea, and L.L. White (Eds.) Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (pp. 833–868). Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press. Ntale, H. K. & Gan, T. Y. (2004). East African rainfall anomaly patterns in association with El Niño/Southern Oscillation. Journal of Hydrologic Engineering, 9, 257–268. Obando, J. A., Onywere, S., Shisanya, C., Ndubi, A., Masiga, D., Irura, Z., Mariita, N., & Maragia, H. (2016). Impact of Short-Term Flooding on Livelihoods in the Kenya Rift Valley Lakes. In M. E. Meadows & J. C. Lin (Eds.) Geomorphology and Society (pp. 193–215). Japan: Springer. Okedi, J., Basaibwaki, P., & Chale, F.M. (1975). A Survey of the Fisheries of Lakes Babati and Basotu. Annual Report. Olsson, I. U. (1986). Radiometric dating. In B. E. Berglund (Ed.) Handbook of Holocene Palaeoecology and Palaeohydrology (pp. 273–312). Chichester and New York: John Wiley. Olsson, I. U. (1991). Accuracy and precision in sediment chronology. Hydrobiologia, 214, 25–34. OpenStreetMap. 2016 February. Retrieved from http://www.tzgisug.org/. Otu, M. K., Ramlal, P., Wilkinson, P., Hall, R. I., & Hecky, R. E. (2011). Paleolimnological evidence of the effects of recent cultural eutrophication during the last 200years in Lake Malawi, East Africa. Journal of Great Lakes Research, 37, 61–74. Ouma, Y. O. & Tateishi, R. (2006). A water index for rapid mapping of shoreline changes of five East African Rift Valley lakes: an empirical analysis using Landsat TM and ETM+ data. International Journal of Remote Sensing, 27, 3153–3181.

64 Panizzo, V. N., Mackay, A. W., Ssemmanda, I., Taylor, R., Rose, N., & Leng, M. J. (2008). A 140-year record of recent changes in aquatic productivity in a remote, tropical alpine lake in the Rwenzori Mountain National Park, Uganda. Journal of Paleolimnology, 40, 325–338. Patrick, R. (1948). Factors Effecting the Distribution of Diatoms. Botanical Review, 14, 473–524. Pennington, W., Tutin, T. G., Cambray, R. S., & Fisher, E. M. (1973). Observations on lake sediments using fallout 137Cs as a tracer. Nature, 242, 324–326. Perrolf, K. & Sandström, K. (1995). Correlating Landscape Characteristics and Infiltration. A Study of Surface Sealing and Subsoil Conditions in Semi-Arid Botswana and Tanzania. Geografiska Annaler, 77, 119–133. PESA Times (30 Jul, 2013). Human Activities Threaten Lake Basotu. Retrieved from http://pesatimes.co.tz/news/environment/human-activities-threaten-lake-bassotu Piirsoo, K., Pall, P., Tuvikene, A., Viik, M., & Vilbaste, S. (2010). Assessment of water quality in a large lowland river (Narva, Estonia/Russia) using a new Hungarian po- tamoplanktic method. Estonian Journal of Ecology, 59, 243–258. Present, E.W. (1973). Report on the Soils of the Basotu Wheat Farm, Tanzania for the Agronomic Research Project (Wheat). Canadian International Development Agency. Retrieved from http://www.wossac.com/search/wossac_detail.cfm?ID=36329 Risberg, J., Sandgren, P., Teller, J. T., & Last, W. M. (1999). Siliceous microfossils and mineral magnetic characteristics in a sediment core from Lake Manitoba, Canada: a remnant of glacial Lake Agassiz. Canadian Journal of Earth Sciences, 36, 1299– 1314. Robertsson, A. M. (1973). Late-glacial and Pre-boreal pollen and diatom diagrams from Skurup, southern Scania (75 pp.). Stockholm: Geological Survey of Sweden. Rockman, M. (2011). The necessary roles of archaeology in climate change mitigation and adaptation. In M. Rockman & J. Flatman (Eds.) Archaeology in Society (pp. 193–215). New York: Springer. Rockström, J., Karlberg, L., Wani, S. P., Barron, J., Hatibu, N., Oweis, T., Bruggeman, A., Farahani, J., & Qiang, Z. (2010). Managing water in rainfed agriculture–The need for a paradigm shift. Agricultural Water Management, 97, 543–550. Ropelewski, C. F. & Halpert, M. S. (1987). Global and regional scale precipitation patterns associated with the El Niño/Southern Oscillation. Monthly Weather Review, 115, 1606–1626. Roubeix, V., Chalié, F., & Gasse, F. (2014). The diatom Thalassiosira faurii (Gasse) Hasle in the Ziway–Shala lakes (Ethiopia) and implications for paleoclimatic reconstructions: Case study of the Glacial–Holocene transition in East Africa. Palaeogeography, Palaeoclimatology, Palaeoecology, 402, 104–112. Russell, J. M., Verschuren, D., & Eggermont, H. (2007). Spatial complexity of ‘Little Ice Age’climate in East Africa: sedimentary records from two crater lake basins in western Uganda. The Holocene, 17, 183–193. Ryves, D. B., Mills, K., Bennike, O., Brodersen, K. P., Lamb, A. L., Leng, M. J., Russell, J. M., & Ssemmanda, I. (2011). Environmental change over the last millennium recorded in two contrasting crater lakes in western Uganda, eastern Africa (Lakes Kasenda and Wandakara). Quaternary Science Reviews, 30, 555–569. Saji, N. H., Goswami, B. N., Vinayachandran, P. N., & Yamagata, T. (1999). A dipole mode in the tropical Indian Ocean. Nature, 401, 360–363. Sandström, K. (1995). The recent lake Babati floods in semi-arid Tanzania: a response to changes in land cover?. Geografiska Annaler. Series A. Physical Geography, 35– 44. Sannel, A. B. K. & Brown, I. A. (2010). High-resolution remote sensing identification of thermokarst lake dynamics in a subarctic peat plateau complex. Canadian Journal of Remote Sensing, 36, S26–S40.

65 Savenije, H. H. G., Hoekstra, A. Y., & Van der Zaag, P. (2014). Evolving water science in the Anthropocene. Hydrology and Earth System Sciences, 18, 319–332. Schön, S. J. (2011). Magnetic Properties. In J. H. Schön (Ed.) Physical Properties of Rocks: a workbook (pp 373–391). Oxford: Elsevier. Retrieved from http://www.sciencedirect.com.ezp.sub.su.se/science/handbooks/15678032/8 Schmid, A. M. M. (1979). Influence of environmental factors on the development of the valve in diatoms. Protoplasma, 99, 99–115. Schulz, R., Buness, H., Gabriel, G., Pucher, R., Rolf, C., Wiederhold, H., & Wonik, T. (2005). Detailed investigation of preserved maar structures by combined geophysical surveys. Bulletin of Volcanology, 68, 95–106. Shaman, J., Solomon, S., Colwell, R. R., & Field, C. B. (2013). Fostering advances in interdisciplinary climate science. Proceedings of the National Academy of Sciences, 110, 3653–3656. Sicko-Goad, L., Stoermer, E. F., & Kociolek, J. P. (1989). Diatom resting cell rejuvenation and formation: time course, species records and distribution. Journal of Plankton Research, 11, 375–389. Simonsen, R. (1979). The diatom system: ideas on phylogeny. Bacillaria, 2, 9–71. Sitoe, S. R., Risberg, J., Norström, E., Snowball, I., Holmgren, K., Achimo, M., & Mugabe, J. (2015). Paleo-environment and flooding of the Limpopo River-plain, Mozambique, between c. AD 1200–2000. Catena, 126, 105–116. Smol, J. P. (1992). Paleolimnology: an important tool for effective ecosystem management. Journal of Aquatic Ecosystem Health, 1, 49–58. Smol, J. P. & Stoermer, E. F. (Eds.). (2010). The diatoms: applications for the environmental and earth sciences (469 pp.). Cambridge: Cambridge University Press. Stager, J. C., Hecky, R. E., Grzesik, D., Cumming, B. F., & Kling, H. (2009). Diatom evidence for the timing and causes of eutrophication in Lake Victoria, East Africa. Hydrobiologia, 636, 463–478. Stager, J. C., Ryves, D., Cumming, B. F., Meeker, L. D., & Beer, J. (2005). Solar variability and the levels of Lake Victoria, East Africa, during the last millenium. Journal of Paleolimnology, 33, 243–251. Stewart, B. A. & Peterson, G. A. (2015). Managing green water in dryland agriculture. Agronomy Journal, 107, 1544–1553. Stoof-Leichsenring, K. R., Junginger, A., Olaka, L. A., Tiedemann, R., & Trauth, M. H. (2011). Environmental variability in Lake Naivasha, Kenya, over the last two centuries. Journal of Paleolimnology, 45, 353–367. Strömquist, L. (1992). Environmental impact assessment of natural disasters, a case study of the recent Lake Babati floods in northern Tanzania. Geografiska Annaler. Series A. Physical Geography, 81–91. Stump, D. (2010). “Ancient and backward or long-lived and sustainable?” The role of the past in debates concerning rural livelihoods and resource conservation in Eastern Africa. World Development, 38, 1251–1262. Survey Division of the Ministry of Lands (1966). Tanzania Quarter Degree Sheet 84 [map]. (1:125,000). Government of Tanzania: Dodoma, Tanzania. Svensson, E. (2010). Climate and environmental dynamics during the last 2500 years as inferred from dambo sediment, Irkeepusi, northern Tanzania (Master’s thesis). Talbot, M. R. (2001). Nitrogen isotopes in palaeolimnology. In W. M. Last & J. P. Smol (Eds.) Tracking environmental change using lake sediments. Volume 2: Physical and Geochemical Methods (pp. 401–439). The Netherlands: Springer. Talling, J. F. & Talling, I. B. (1965). The Chemical Composition of African Lake Wa- ters. International Review of Hydrobiology, 50, 421–463. Taylor, J.C., Harding, W.R., & Archibald, C.G.M. (2007). An Illustrated Guide to Some Common Diatom Species from South Africa. Report to the Water Research Commission.

66 Thompson, R. (1973). Paleolimnology and Paleomagnetism. Nature, 242, 182–184. Thompson, R. & Oldfield, F. (1986). Environmental Magnetism (227 pp.). The Netherlands: Springer Science & Business Media. Thompson, R., Stober, J. C., Turner, G. M., Oldfield, F., Bloemendal, J., Dearing, J. A., & Rummery, T. A. (1980). Environmental applications of magnetic measurements. Science, 207, 481–486. TNRF (2006). The Place of Pastoralism in Modern Tanzania [seminar proceedings]. Tanzania Natural Resource Forum. Retrieved from http://www.ereto- npp.org/download_documents/RLWG_2006_the_place_of_pastoralism_in_modern _Tanzania.pdf Trenberth, K. E. (1997). The definition of El Niño. Bulletin of the American Meteorological Society, 78, 2771–2777. Udo, S., Becker, A., Finger, P., Meyer-Christoffer, A., Rudolf, B. & Ziese, M. (2011). GPCC Full Data Reanalysis Version 6.0 (at 0.5°, 1.0°, 2.5°): Monthly Land-Surface Precipitation from Rain-Gauges built on GTS-based and Historic Data. Retrieved from ftp://ftp.dwd.de/pub/data/gpcc/html/fulldata_v6_doi_download.html Ummenhofer, C. C., Sen Gupta, A., England, M. H., & Reason, C. J. (2009). Contributions of Indian Ocean sea surface temperatures to enhanced East African rainfall. Journal of Climate, 22, 993–1013. UN (2015). Millennium Development Goals Indicators: United Republic of Tanzania. Retrieved from http://mdgs.un.org/unsd/mdg/Data.aspx UN (2016). Sustainable Development Goals: The Sustainable Development Agenda. United Nations. Retrieved from http://www.un.org/sustainabledevelopment/development-agenda/ UNDESA (2016). International Decade for Action ‘WATER FOR LIFE’ 2005-2015. United Nations Department of Economics and Social Affairs. Retrieved from http://www.un.org/waterforlifedecade/scarcity.shtml UNEP (2010). “Africa Water Atlas”. Division of Early Warning and Assessment (DEWA). United Nations Environment Programme (UNEP). Nairobi, Kenya. Retrieved from http://www.unep.org/pdf/africa_water_atlas.pdf United Republic of Tanzania (2012). National Climate Change Communication Strategy (2012–2017). Retrieved from https://www.undp-aap.org/sites/undp- aap.org/files/National_CC_Communication_Strategy_2012_2017.pdf UN WATER (2013). United Republic of Tanzania UN-Water Country Brief. Retrieved from http://www.unwater.org/fileadmin/user_upload/unwater_new/docs/Publications/TZ A_pagebypage.pdf USGS (2000). Shuttle Radar Topography Mission, 1 Arc-Second Global scenes: SRTM1S04E033V3, SRTM1S04E034V3, SRTM1S04E035V3, SRTM1S04E036V3, SRTM1S05E033V3, SRTM1S05E034V3, SRTM1S05E035V3, SRTM1S05E036V3, SRTM1S06E033V3, SRTM1S06E034V3, SRTM1S06E035V3, & SRTM1S06E036V3. United States Geological Survey. Global Land Cover Facility; University of Maryland: College Park, MD, USA. USGS (2015). Shuttle Radar Topography Mission (SRTM) 1 Arc-Second Global. United States Geological Survey. Retrieved from https://lta.cr.usgs.gov/SRTM1Arc Van Dam, H., Mertens, A., & Sinkeldam, J. (1994). A coded checklist and ecological indicator values of freshwater diatoms from the Netherlands. Aquatic Ecology, 28, 117–133. Verosub, K. L. & Roberts, A. P. (1995). Environmental magnetism: past, present, and future. Journal of Geophysical Research: Solid Earth, 100, 2175–2192. Verschuren, D. (1999). Sedimentation controls on the preservation and time resolution of climate-proxy records from shallow fluctuating lakes. Quaternary Science Reviews, 18, 821–837.

67 Verschuren, D., Laird, K. R., & Cumming, B. F. (2000). Rainfall and drought in equatorial east Africa during the past 1,100 years. Nature, 403, 410–414. Vilmi, A., Karjalainen, S. M., Landeiro, V. L., & Heino, J. (2015). Freshwater diatoms as environmental indicators: evaluating the effects of eutrophication using species morphology and biological indices. Environmental monitoring and assessment, 187, 1–10. Walker, M. (2005). Quaternary Dating Methods. Chichester: John Wiley. Weiberg, E., Unkel, I., Kouli, K., Holmgren, K., Avramidis, P., Bonnier, A., Dibble, F., Finné, M., Izdebski, A., Katrantsiotis, Stocker, S. R., Andwinge, M., Baika, K., Boyd, M., & Heymann, C. (2016). The socio-environmental history of the Peloponnese during the Holocene: Towards an integrated understanding of the past. Quaternary Science Reviews, 136, 40–65. Whalen, J. K., Chang, C., Clayton, G. W., & Carefoot, J. P. (2000). Cattle manure amendments can increase the pH of acid soils. Soil Science Society of America Journal, 64, 962–966. Williamson, D., Jelinowska, A., Kissel, C., Tucholka, P., Gibert, E., Gasse, F., Massault, M., Taieb, M., Van Campo, E., & Wieckowski, K. (1998). Mineral-magnetic proxies of erosion/oxidation cycles in tropical maar-lake sediments (Lake Tritrivakely, Madagascar): paleoenvironmental implications. Earth and Planetary Science Letters, 155, 205–219. Wolfe, B. B., Edwards, T. W. D., Elgood, R. J., & Beuning, K. R. M. (2001). In W. M. Last & J. P. Smol (Eds.) Tracking environmental change using lake sediments. Volume 2: Physical and Geochemical Methods (pp. 373–400). The Netherlands: Springer. Wolin, J. A. & Stone, J. R. (2010). Diatoms as indicators of water-level change in freshwater lakes. In J. P. Smol & E. F. Stoermer (Eds.), The Diatoms: Applications for the Environmental and Earth Sciences (pp. 174–185). Cambridge University Press. Wolter, K. & Timlin, M. S. (1993). Monitoring ENSO in COADS with a seasonally adjusted principal component index. In Proc. of the 17th Climate Diagnostics Workshop (pp. 52–57). Oklahoma, USA: NOAA/NMC/CAC, NSSL, Oklahoma Climatological Survey, CIMMS and the School of Meteorology University of Oklahoma. World Bank (2016). Agriculture, value added (% of GDP). Retrieved from http://data.worldbank.org/indicator/NV.AGR.TOTL.ZS WWAP (2012). The United Nations World Water Development Report 4: Managing Water under Uncertainty and Risk. World Water Assessment Programme. Paris: UNESCO. Xinhua (17 April 2016). Feature: Hunger looms in Northern Tanzania floodhit district. Retrieved from http://news.xinhuanet.com/english/2016-04/17/c_135287353.htm. Öberg, H., Andersen, T. J., Westerberg, L. O., Risberg, J., & Holmgren, K. (2012). A diatom record of recent environmental change in Lake Duluti, northern Tanzania. Journal of Paleolimnology, 48, 401–416.

68 10 Appendices

10.1 Appendix A: Lake Basotu Diatoms A list of diatom species identified in the Lake Basotu sediment core, 1803.

Amphora copulata (Kützing) Schoeman & Archibald Amphora veneta Kützing Anomoeoneis sphaerophora (Ehrenberg) Pfitzer Aulacoseira granulata (Ehrenberg) Simonsen Aulacoseira sp. Thwaites Caloneis bacillum (Grunow) Cleve Cocconeis placentula Ehrenberg Craticula ambigua (Ehrenberg) D.G. Mann Craticula buderi (Hustedt) Craticula molestiformis (Hustedt) Lange-Bertalot Craticula sp. Grunow Cyclotella meneghiniana Kützing Cyclotella ocellata Pantocsek Cyclotella sp. (Kützing) Brébisson Cymatopleura solea (Brébisson) W. Smith Cymbella kappii (Cholnoky) Cholnoky Cymbella spp. Agardh Cymbopleura subaequalis (Grunow) Krammer Diadesmis confervacea (Kützing) D.G. Mann Diploneis smithii (Brébisson) Cleve Encyonema silesiacum (Bleisch) D.G. Mann Encyonema sp A Epithemia adnata (Kützing) Brébisson Epithemia sorex Kützing Epithemia argus var. alpestris (W.Smith) Grunow Epithemia spp. Kützing Eunotia minor (Kützing) Grunow Eunotia pectinalis var. undulata (Ralfs) Rabenhorst Eunotia spp. Ehrenberg Fallacia pygmaea (Kützing) Stickle & D.G. Mann Fragilaria biceps (Kützing) Lange-Bertalot Fragilaria capucina Desmazières Fragilaria ulna (Nitzsch) Lange-Bertalot Gomphonema affine Kützing Gomphonema gracile Ehrenberg sensu stricto Gomphonema minutum (Agardh) Agardh Gomphonema parvulum (Kützing) Gomphonema truncatum Ehrenberg

69 Gomphonema spp. Ehrenberg Gyrosigma parkerii (Harrison) Elmore Hantzschia amphioxys (Ehrenberg) Grunow Luticola kotschyi (Grunow) Luticola mutica (Kützing) D.G. Mann Luticola ventricosa (Kützing) D.G.Mann Luticola sp. Mann in Round, Crawford and Mann Navicula erifuga Lange-Bertalot Navicula placentula (Ehrenberg) Kützing Navicula radiosa Kützing Navicula rostellata Kützing Navicula viridula (Kützing) Ehrenberg Navicula spp. Bory de Saint-Vincent Nitzschia amphibia Grunow Nitzschia compressa (Bailey) C.S. Boyer Nitzschia frustulum (Kützing) Grunow Nitzschia linearis W.Smith Nitzschia linearis var. subtilis Hustedt Nitzschia obtusa W.Smith Nitzschia palea (Kützing) W. Smith Nitzschia spp. Hassall Pinnularia acrosphaeria W. Smith Pinnularia borealis Ehrenberg Pinnularia gibba Ehrenberg Pinnularia spp. Ehrenberg Rhopalodia gibba (Ehrenberg) O. Müller Sellaphora pupula (Kützing) Mereschkovsky Stauroneis phoenicenteron (Nitzsch) Ehrenberg Stauroneis spp. Ehrenberg Staurosira elliptica (Schumann) D.M. Williams & Round Stephanodiscus sp. Ehrenberg Surirella ovalis Brébisson Surirella spp. Turpin Thalassiosira faurii (Gasse) Hasle Tryblionella apiculata W. Gregory Tryblionella calida (Grunow) D.G. Mann Tryblionella hungarica (Grunow) Frenguelli Tryblionella sp. W. Smith

Single occurrence, not included in diagrams: Caloneis silicula (Ehrenberg) Cleve Caloneis sp. Cleve Cymbella muelleri Hustedt Eunotia formica Ehrenberg Gomphonema pumilum var. rigidum Reichardt & Lange-Bertalot Nitzschia sigma (Kützing) W. Smith Pinnularia viridis (Nitzsch) Placoneis clementis (Grunow) E.J. Cox Staurosira construens Ehrenberg

70 10.2 Appendix B: Reports obtained regarding wheat in the Hanang District

International Institute for Environment and Development (IIED) - Swedish International Development Agency

- Displaced Pastoralists and Transferred Wheat Technology in Tanzania (1990, 21 pp)

Canadian International Development Agency

- Report on the Soils of the Basotu Wheat Farm, Tanzania for the Agronomic Research Project (Wheat) (1973, 73 p) - Soils of the Basotu and Balangida Lelu Areas of Northern Tanzania (1986, 194 pp) - A Regional Travelling Workshop on Long-Term Wheat-Based Sustainability Trials in East Africa (1994, 104 pp)

Legal and Human Rights Center – Tanzania

- Human Rights and Business Report 2013 (212 pp)

Regional Wheat Workshop for Eastern, Central, and Southern Africa

- 4th (1985): Njoro, Kenya - 5th (1987): Antsirabe, Madagascar - 8th (1993): Kampala, Uganda - 9th (1995): Addis Ababa, Ethiopia - 10th (1998): University of Stellenbosch, South Africa - 11th (2000): Addis Ababa, Ethiopia

United States Agency for International Development (USAID)

- Improving Capacity of Cususwash Universities for Water Management for Agriculture (1973, 141 pp) - Potential Groundwater and Land Resource Analysis for Planning and Development, Arusha Region, United Republic of Tanzania (1975, 238 pp) - Soil Resource Inventories; Proceedings of a Workshop (1977, 365 pp) - Appraisal Report of the Aid Program in Tanzania (1978, 49 pp) - The Context for Village-Level Development in Hanang District, Tanzania (1978, 71 pp) - Trends and Interrelationships in Food, Population, and Energy in Eastern Africa: A Preliminary Analysis; Volume 3: Literature Summaries and Reviews (1980, 201 pp) - Agricultural Livelihoods in Eastern Africa: Classification and Distribution, A First Approximation (1980, 28 pp)

71 - Agricultural Research Resource Assessment in the SADCC Countries; Volume II: Country Report – Tanzania (1985, 147 pp) - Tanzania Private Sector Assessment: Appendices (1992, 476 pp) - Assessment of CBNRM Best Practices in Tanzania, Final Report (2002, 118 pp) - USAID Country Profile: Tanzania (unknown, 40 pp)

United Nations:

- Tanzania Weed Management Training Course Summary Report (1984, 15pp)

Miscellaneous:

- Environmental Impact of Agricultural Development A.I.D. in Tanzania (1975, 63 pp) - Arusha Region Integrated Development Plan; Volume Five: Information Strategy and Documentation (1981, 337 pp) - East Africa Soil Correlation Committee: Vertisolic Soils of the Hanang Wheat Complex, Tanzania (1983, 38 p) - Court Summary: Mulbadaw Village Council and 67 Others v National Agricultural and Food Corporation (1984) - The World Conservation Union; Proceedings of a Seminar on Wetlands of Tanzania (1991, 181 pp) - Study on Options for Pastoralists to Secure their Livelihoods; Experiences in the Defense of Pastoralists Resource Rights in Tanzania: Lessons and Prospects (2007, 34pp) - The State of the then NAFCO, NARCO and Absentee Landlords’ Farm/Ranches in Tanzania (2009, 114pp)

72 10.3 Appendix C: Lake Babati Diatoms A list of diatom species identified in the Lake Babati sediment core, and subsequent counting data.

Amphora sp. Ehrenberg Anomoeoneis sphaerophora (Ehrenberg) Pfitzer Aulacoseira ambigua (Grunow) Simonsen Aulacoseira granulata var. angustissima (O Müller) Simonsen Cocconeis placentula Ehrenberg Cocconeis placentula var. euglypta (Ehrenberg) Grunow Craticula sp. Grunow Cyclotella meneghiniana Kützing Cymbella sp. Agardh Diadesmis confervacea (Kützing) DG Mann Encyonema caespitosum Kützing Encyonema silesiacum (Bleisch) DG Mann Epithemia adnata (Kützing) Brébisson Epithemia sorex Kützing Epithemia sp. Kützing Fragilaria sp. Lyngbye Gomphonema affine Kützing Gomphonema clevei Fricke Gomphonema minutum (Agardh) Agardh Gomphonema parvulum (Kützing) Luticola mutica (Kützing) DG Mann Nitzschia amphibia Grunow Nitzschia spp. Hassall Pinnularia borealis Ehrenberg Pinnularia sp. Ehrenberg Rhopalodia gibba (Ehrenberg) O Müller Rhopalodia operculata (Agardh) Håkansson Rhopalodia sp. Muller Thalassiosira faurii (Gasse) Hasle Tryblionella calida (Grunow) DG Mann Surirella ovalis Brébisson

73 Results of Lake Babati Diatom Analysis

Sample Depth (cm) 253– 217– 193– 163– 151– 133– 103– 79– Species 252 216 192 162 150 132 102 78 Amphora sp. x Anomoeoneis sphaerophora x x Aulacoseira x xx xx xx x ambigua Aulacoseira granulata var. x xx x xx xx xx x angustissima Cocconeis x placentula Cocconeis placentula var. x euglypta Craticula sp. x x Cyclotella meneghiniana xx xx xx xx xx xx xx xx Cymbella sp. x Diadesmis x x confervacea Encyonema x caespitosum Encyonema x x x x silesiacum Epithemia adnata x x x x Epithemia sorex x xx x x x Epithemia sp. x Fragilaria sp. x Gomphonema x x affine Gomphonema clevei x Gomphonema minutum x x Gomphonema x parvulum Luticola mutica x Nitzschia amphibia x x Nitzschia spp. x x x x Pinnularia borealis x Pinnularia sp. x x x Rhopalodia gibba x Rhopalodia operculata x Rhopalodia sp. x x x Thalassiosira faurii xx x x Tryblionella calida x Surirella ovalis x x Surirella spp. x x x Fragmented x x Fragilaria (x: present, xx: prominent)

74 Comments on Lake Babati Diatom Analysis

253–252 cm Sample contains primarily fragmented and/or dissolved diatom frustules.

217–216 cm Two transects were counted due to high abundances. Fragmented Fragilaria were unidentifiable and hence not counted.

193–192 cm Relatively large variety of taxa with fair preservation. Two transects were counted due to high abundances. Fragmented Fragilaria were unidentifiable and not counted.

163–162 cm Degraded with unidentifiable fragments.

151–150 cm Degraded with unidentifiable fragments.

133–132 cm Sparse material with unidentifiable fragments.

103–102 cm Very sparse and fragmented diatoms.

79–78 cm Prepared from crushed material, contains unidentifiable fragments.

2–1 cm Prepared from suspended material, no diatoms present, Lycopodium pollen suggests sediment preparation was for pollen analysis.

75