PAPEL DEL CONTADOR PÚBLICO

The study area is the watershed of the EAM in Tanzania, covering 33.9 million hectares (Figures 1.4-1.9). Historically, East Africa has experienced a hominid presence for over two million years, being home to some of the earliest known human fossils (Isaac and McCown, 1976). The Eastern Arc has been climatically stable for long periods, possibly preceding the end of the Miocene (Lovett, 1993a). Previous research and anecdotal evidence suggests large changes in land use over the past century, but no long-term spatial analysis has been conducted. The EAM are thought to have once been nearly covered by forest but it has been estimated that between 70% and 96% of the original forest cover has been lost (Newmark, 2002, Hall et al., 2009), mainly to agricultural encroachment (Burgess et al., 2001). It is thought that in 1900 there was about three times as much forest cover present compared with today (Madoffe et al., 2006) (Table 1.3; Table 1.4). Lowland forest in particular has been extensively exploited, with vast tracts of this forest type having been cleared for agriculture (Newmark, 2002, Lovett, 1993b). Unlike in many areas of the tropics, this time period is relatively well documented as several older land use maps are available (Gillman, 1949, Baumann, 1891, Engler, 1908-10, Shantz and Marbut, 1923). This is mainly due to Tanzania’s colonial past, firstly by Germany in the late 19th century, before being designated as a British Mandate from 1919 until 1961. During this period Tanzania’s natural resources were exploited, including timber resources and agricultural land for exports of timber, cloves and sisal. Despite much resource extraction, Tanzania has a long history of establishing protected areas, with the oldest protected area in my study area dating back to 1907 (IUCN and UNEP-WCMC, 2010, NPW, 2010).

The present day watershed is a heterogeneous mix of cropland, woodland and forest (the three major tropical biomes) and contains the administrative and commercial capitals of Dodoma and Dar es Salaam, respectively (see Swetnam et al. (2011) (2011) for further details) (Figure 1.5; Table 1.1). The watershed also shows a heterogeneous climate, under influence of the Indian Ocean (Mutai et al., 1998). Altitudinal ranges from sea level to over 2000m provide a wide temperature range (Lovett, 1993b) (Figure 1.6). Rainfall and dry season length are also extremely varied. In the northern

part of the study area, there are two peaks in rainfall (from October to December and from March to May). More southerly areas experience a single dry (June to September) and wet (November to May) season (Lovett, 1993b) (Figure 1.7). Further heterogeneity arises as large areas of woodland, savannah and croplands are burnt, often annually (Krawchuk et al., 2009) (Figure 1.9).

Figure 1.4 The Eastern Arc Mountains of Tanzania and Kenya (Platts et al.,

2011). The study area is the Eastern Arc watershed in Tanzania (Swetnam et al., 2011).

Broadly, the region can be subdivided into six distinct zones (the northern, western, central; eastern, southern and south eastern zones [Figure 1.10]) based on climatic, edaphic and anthropogenic criteria (Figures 1.5-1.10), giving rise to a natural, factorial experiment. The northern zone is typically characterised by a high anthropogenic disturbance and large monthly temperature ranges, but low levels of precipitation, soil fertility and fire

(Figure 1.10). The western zone, whilst similar to the northern zone, experiences low anthropogenic disturbance and a high annual probability of fire. The central zone, again similar to the northern zone in that it shows high anthropogenic disturbance and low levels of fire, is an area of high fertility, experiencing both low mean annual temperatures and small monthly temperature ranges but high levels of precipitation. The eastern zone is an area of high anthropogenic disturbance, similar to the northern zone but showing small monthly temperature ranges and high levels of precipitation. The southern zone shows many similarities with the central zone, but shows much lower levels of anthropogenic disturbance. The south eastern zone is relatively unique, being an infertile area with low levels of anthropogenic disturbance and small monthly temperature ranges, but high mean annual temperatures, levels of precipitation and occurrences of fire. Thus, in various combinations, the effects of climatic, edaphic and anthropogenic variables may be statistically isolated. For example, comparing tree inventory plots in the central and southern zones isolates the effect of varying anthropogenic disturbance, whilst keeping other variables relatively constant. Other examples include comparing the central and eastern zones (isolating the effects of mean annual temperature and soil fertility); the central and northern zones (isolating the effects of precipitation and soil fertility); the northern and western zones (isolating the effects of anthropogenic disturbance and fire); and the eastern and northern zones (isolating the effect of month temperature range and precipitation) (Figure 1.10).

The EAM themselves (5.2 million ha, as delimited in Platts et al. (2011)) are nested within the broader study area (Figure 1.11), and are considered a global priority for biodiversity conservation due to the high levels of plant and animal endemism (Lovett, 1990, Myers et al., 2000, Burgess et al., 2007). At the time of the last national census, the population of Tanzania was 34.4 million people (NBS, 2006), of which 2.2 million lived in the EAMs and 12.9 million lived within the wider watershed catchment. Over the last 14 years, the national population growth rate has been 2.9% yr-1, tending to increase pressure on land and resources (NBS, 2006). Pressure on local resources derives from global, as well as local demand. Each year, a large amount of, mostly illegally felled, timber is exported from the study area (Milledge et al., 2007). The true extent of timber removal is highly uncertain, but is estimated to have a significant economic impact, with 58 million US$ in timber royalties lost annually (Milledge et al., 2007). Through a combination of external and internal demand, waves of forest degradation radiate from Dar es Salaam

(Ahrends et al., 2010). This is of concern as the importance of this region to global biodiversity is well recognised (Myers et al., 2000).

Despite broad climate stability since the Miocene (Lovett, 1993a), the region is predicted to experiences alterations in the future. The current population increase is expected to continue, reaching 43.9 million by 2015 (World Bank, 2004). East Africa is one of few tropical regions where future climate projections are in broad agreement (Hulme et al., 2001, Ruosteenoja et al., 2003, Christensen JH et al., 2007, Sitch et al., 2008). Over the next century, most simulations show a robust future warming and general annual-mean rainfall increases, divided into more precipitation during rainy seasons but less or no change during dry seasons (Doherty et al., 2009). However, the simulations provide highly uncertain projections future extreme precipitation anomalies and do not account for any changes in land use or vegetation structure (Doherty et al., 2009). In addition, fires are predicted to become less frequent, although this prediction could also be affected by anthropogenic activities (Krawchuk et al., 2009). Furthermore, plausible storylines estimating the land cover distribution within my study area in the year 2025 have been developed (Swetnam et al., 2011). In combination, modelling these climatic and anthropogenic changes provides the potential to make predictions of the effect of valuable ecosystem services, such as carbon storage. In this thesis, I use the heterogeneous landscape to model and map carbon storage, in both the past and present. In addition, I assess the variables potentially causing the observed spatial distribution. I hope that, as Tanzania is a United Nations REDD+ pilot country, a better understanding of LCC and the current carbon stock in Tanzania will inform policy makers (Burgess et al., 2010), both nationally and internationally.

Table 1.3 Eastern Arc forest type categories and characteristics (Lovett, 1993b).

Forest Type Altitude (m above sea level) Rainfall (mm) Canopy height (m) Emergent’s height (m) Basal area (m2/ha) DBH (cm) Stem density (>20cm dbh) Number of species (Lovett, 1999) Montane >1500 1000-1200 10-20 Up to 30 20-40 Few > 100 Most < 40 240 42 Upper montane forest >1800 >1200 10-20 Up to 25 30-70 Few > 100 Most < 40 330 57 Montane 1200-1800 >1200 25-40 Up to 50 30-120 Many > 50 High proportion > 100 250 120 Submontane 800-1400 >1500 25-40 Up to 50 30-70 Many > 50 High proportion > 100 170 114 Lowland <800 >1500 25-40 Up to 50 - Many > 50 High proportion > 100 140 91 Dry lowland <800 1000-1500 15-20 Up to 35 - - - 52

Table 1.4 A summary of the Eastern Arc Mountains

Mountain Block Coordinates (degrees and minutes) (Burgess et al., 2007) Block Area (km2) (Burgess et al., 2007) Forest Cover (km2) (Mbilinyi and Kashaigili, 2005) Altitudinal range of forest (m

above sea level) (Burgess et al., 2007) Number of forest patches (Newmark, 1998) Number of single block endemics (MNRT, 2006) Forest cover loss 1995- 2000 (km2) (Hall et al., 2009) North Pare 0335-0346 S, 3733-3740 E 454 27 1300-2113 2 0 27.8 South Pare 0404-0434 S, 3745-3801 E 1578 138 820-2463 5 2 28.7 West Usambara 0420-0507 S, 3806-3841 E 2507 319 1200-2200 17 5 39.9 East Usambara 0445-0520 S, 3826-3848 E 1082 263 130-1506 8 4 38.1 Nguu 0527-0538 S, 3736-3732 E 1591 188 1000-1500 Included within Nguru 0 9.2 Nguru 0527-0613 S, 3726- 3737 E 1673 297 400-2000 8 0 6.3 Ukaguru 0619-0635 S, 3653-3703 E 1259 172 1500-2250 1 1 16.5 Uluguru 0651-0712 S, 3736-3745 E 1478 278 300-2400 5 14 17.5 Rubeho 0648-0772 S, 3634-3658 E 4637 464 520-2050 6 2 26.8 Malundwe 0724 S, 2718 E 1662 13 1200-1275 - 0 0.0 Udzungwa 0722-0843 S, 3507-3658 E 16131 1353 300-2580 26 17 22.4 Mahenge 0837-0838 S, 3642-3644 E 2802 19 460-1040 3 0 31.4

Figure 1.5 Examples of the anthropogenic heterogeneity of the study area, illustrated using (a) the natural logarithm of the population

Figure 1.6 Differences in temperature across the study area, illustrated using (a) the mean annual temperature (MAT) and (b) the

mean annual monthly temperature range (Hijmans et al., 2005, Jarvis et al., 2008).

(oC) (oC)

Figure 1.7 Difference in precipitation across the study area, illustrated using (a) the mean maximum cumulative water deficit and (b)

Figure 1.8 Examples of the edaphic heterogeneity of the study area, illustrated using (a) soil fertility and (b) the percentage sand

Figure 1.9 Further examples of the heterogeneity of the study area, illustrated using (a) the annual mean burned area probability

(Roy et al., 2005) and (b) the mean annual global horizontal solar radiation (Perez et al., 2002, NREL, 2010).

Daily Solar Radiation (Watt-hours m-2) Annual Fire Probability

Figure 1.10 The six zones (red) broadly describing the heterogeneity of the Eastern Arc Mountains of Tanzania and Kenya (Platts et

al., 2011). The study area is the Eastern Arc watershed in Tanzania (Swetnam et al., 2011). Northern Zone

Anthropogenic disturbance: High Mean annual temperature: Medium Monthly temperature range: Large Precipitation: Low

Soil fertility: Low

Burned area probability: Low

Eastern Zone

Anthropogenic disturbance: High Mean annual temperature: High Monthly temperature range: Small Precipitation: High

Soil fertility: Low

Burned area probability: Low Central Zone

Anthropogenic disturbance: High Mean annual temperature: Low Monthly temperature range: Small Precipitation: High

Soil fertility: High

Burned area probability: Low

Western Zone

Anthropogenic disturbance: Low Mean annual temperature: Medium Monthly temperature range: Large Precipitation: Low

Soil fertility: Low

Burned area probability: High

Southern Zone

Anthropogenic disturbance: Low Mean annual temperature: Low Monthly temperature range: Small Precipitation: High

Soil fertility: High

Burned area probability: Low

South Eastern Zone

Anthropogenic disturbance: Low Mean annual temperature: High Monthly temperature range: Small Precipitation: High

Soil fertility: Low