Diameter growth of trees in miombo and acacia woodland in an eroded landscape in NE Tanzania

Diameter increment of trees typical miombo and acacia woodland was studied during a period 20 years in Kondoa district, Tanzania. The study performed permanent plots severely degraded area subjected to considerable restoration efforts. A total 15 species were selected from database collected within project for monitoring the landscape recovery. Growth performance African searched literature comparison, comprehensive list citations compiled. We found growth fall range reported earlier studies, although varied both between species. There are reports that radial increments unimodal over their lifespan, but we no clear support. In several species, annual increased with stem diameter. rainy ENSO year 1997/98 pairwise compared preceding two be significantly higher wet year, pointing soil water as limiting factor. conclude free development is an alternative tree planting on marginal land. L'augmentation du diamètre des arbres typiques forêts de et d'acacia été étudiée pendant une période ans dans le district Kondoa, en Tanzanie. L'étude réalisée sur parcelles permanentes zone sévèrement dégradée soumise à efforts restauration considérables. Un espèces ont sélectionnées partir d'une base données collectée cadre d'un projet suivi la paysage. Les performances croissance forestières africaines recherchées littérature fins comparaison, liste bibliographique complète compilée. Nous avons constaté que se situait fourchette indiquée études antérieures, bien varie fois entre au sein espèces. Nos résultats ne confirment pas l´augmentation est unimodale cours leur croissance. Chez plusieurs espèces, annuelle augmenté même temps tronc. La l'année pluvieuse 1997-98 comparée par paires avec les deux années précédentes, s'est avérée être significativement plus élevée humide, indiquant teneur eau sol un facteur limitant. Pour conclure, régénération naturelle aux plantations terres marginales. lack detailed information ecology trees. Knowledge rate crucial, also regeneration mortality, well other species-specific traits, such tolerance drought possession nitrogen fixation mycorrhiza. This coupled need find methods restore land semi-arid Africa. outside timber production systems has been pointed out by, i.a., Trouet al. (2006), Winters (2018) Chidumayo (2019). order increase this knowledge, extracted all data vegetation eroded landscape, here present overview 1991–2010. Stem depends traits varies habitat conditions. woodlands, generally correlated positively availability negatively high temperature, (Gourlay, 1995; see al., 2018) (Trouet 2006). Stems may even shrink temporarily (Baker 2002; Gerhardt & Todd, 2009) drought. arid regions, variation likely important due rainfall fluctuations. Most originally comes atmosphere accumulated by living organisms (Aerts Chapin, 2000). soils tropics have shortage (Aweto, 1981; Long 2012). could lead larger diameter nitrogen-fixing than others (Nasto 2019). Also, most tropical mycorrhiza, fungi rare little organic matter, slowing down establishment reducing survival (Frost 2001). It suggested its increasing small trees, reaching peak then declining (Bowman 2013; Hérault 2011). vary age diameter, giving rise largely positive or negative relationships increment. Methods studies areas (see Worbes, 1995). Measuring width rings, assumed annual, attempted researchers. more reliable way measure repeatedly circumference same individuals 2009). Tree slow, long-term (Chidumayo, probably first published work kind Mugasha (2017) who fifteen years. Our took place Irangi Hills (KIH), Dodoma Region, northeast Tanzania, (Mbegu Mlenge, 1984), where made halt erosion 1970s 1990s. late contribution MALISATA research programme (Man–Land Interrelations Semi-arid Tanzania) which included many disciplines active mainly 1990s (Kikula, 1999). report aims (1) possible basis work, (2) use test effect yearly precipitation (3) investigate differences without symbionts. Elevation KIH extends ca 2200 m above sea level (a.s.l.), our sites 1280 1830 a.s.l. (Supporting S1). town (4°54′14″S 35°46′56″E), at 1450 a.s.l., 652 mm (for 1931–1990; Ngana, 1996). Higher up, higher. season November April peaks December April. covers 125,600 ha surrounded plains about 1200 Bedrock consists highly weathered erodible Precambrian gneiss. Much hillslopes believed formerly carried (Brachystegia–Julbernardia) (Backéus 1994), still existing less disturbed sites. Lower parts catena had (the genera Senegalia Vachellia) kinds woodland. However, degraded. gullies formed naturally 10,000 ago (Eriksson During last millennium, cut iron making (Lane Further, there intensive cultivation cash crops 19th-century caravans (Mungo'ng'o, appearance Rangi people 200 meant grazing (Christiansson, Finally, colonial administration cleared eradicate tsetse flies 1930s 1940s Sandgren, By then, much topsoil lost large areas. Residual deficient matter and, particularly, boron. phosphorus, sulphur, manganese, zinc copper low supply (Munkert, Eight types distinguished Backéus (1994). All various degrees affected previous human activity. Apart secondary mixed forest, 2000 distinguished, particularly Brachystegia microphylla elevations, B. spiciformis lower levels, Markhamia obtusifolia strong disturbance Vachellia tortilis toe slopes. few grasslands, treeless scattered acacias (Vachellia nilotica, V. seyal). correlation sequence weak overshadowed degree disturbance. 1973, Tanzanian government launched District, known HADO (Hifadhi Ardhi Dodoma, translated Region Soil Conservation; Christiansson 1991). At time, virtually (Backéus, pers. obs. 1975). Ongoing overgrazing livestock time main reason behind precarious situation. evicted 1979 used emanate succession absence fire. With purpose, 27 permanently marked × plots, representing different S1), established 1991. fenced, since trespassing herds cattle occurred KIH, despite bylaws intended prevent this. Plots no, little, field layer not fenced. Plot corners marked. As fires frequent, firebreak around each plot. Fences firebreaks maintained until 1998. Woody plants monitored modified line intercept method (Canfield, 1941). Along five parallel transects plot, long 4 apart, diameters woody >50 cm high, having vertical crown projections intercepting one lines, measured 30 ground. lengths lines recorded. sum ‘intercept’ per plot expresses percentage cover sampling done April, some extending days into May. 1991, Subsamples recorded 1992–1995, 1997 Therefore, died new ones appeared. final 2010, become widespread. fences broken. Full attempted, 16 re-measured identified help notes remaining markings. aim now restricted collection individual presented here. For study, only ≥50 ≥3 measuring height (30 cm) least three times 1991 2010. Only limited number set throughout whole Some often referred shrubs, leading stems called ‘trees’. Trees tagged, excluded when it fully ascertained consecutive measurements individual. reduced set, minimum six kept further study. all, represented 239 1069 included. yr−1 calculated slope regression against levels irregularities time. Mean standard error calculated. Possible influence analysed population level. whether best described simple linear quadratic regressions, comparisons p- R2-values (R2 adjust regressions) residual distribution. useful, because did seasons, except 1997/1998, El Niño Southern Oscillation (ENSO; Lyaruu, 1998). paired two-sample t-tests means actual contrasting seasons 1996/97 1997/98. Species 6 end 1996/1997 represent growth, 1995/1996 season. values therefore divided two. 749 585 mm. 1131 rain (Lyaruu, highest mean (0.95 yr−1) (Table 1). Average 0.15 0.56 yr−1. lowest Combretum zeyheri. Variation among considerable. biggest average 3.55 2.00 Euphorbia candelabrum, these largest amplitude growth. certain periods Differences taken together (ANOVA, p < 0.001; <5 included). fairly small, rather big microphylla, spiciformis, Julbernardia globiflora, senegal nilotica (p 0.05; Figure 1a–e). zeyheri relation (Figure 1f), significant function. seemed independent 10 13 14 1; 1a). t = 2.73, 0.013). No found. (cm potentially fix those do ability. former group, Dalbergia nitidula (0.21), Dichrostachys cinerea (0.25), (0.55), (0.54), seyal (0.33) (0.95), whereas (0.36), (0.56), (0.15) globiflora (0.36) nitrogen. difference groups. 1995/96 n ≥ senegal, Thespesia (Azanza) garckeana, tortilis, 2). results substantially add Africa put general context. summarised Supporting S2. Although rates scarce 2019), expected, cases explicitly dealing topic. Several below 1.3 (e.g. Mbow Shackleton, Steenkamp 2008), cases, researchers mixed, comments, breast ground (Sass-Klaassen 2008; Therrell 2007). According Van Holsbeeck (2016), ‘sampling … stump heights similar’. environments vary, reports. applies S2 S3). Thus, nothing indicate environment caused should emphasised nutrient Gondwanaland plateau adapted environment. exhibited tortilis. sampled valley bottom close seasonal stream, conditions comparatively good. Canopy (0.2%–199%; investigated predominantly grew relatively cover. appeared (Brachystegia 9 each; Searsia (Rhus) natalensis 8 each). failed detect microphylla. globiflora) 0.41 ± 0.046 (all together), regrowth (0.34–0.50 given (2019) Zambia (B. boehmii, Isoberlinia angolensis, J. Uapaca kirkiana). He (1000–1500 dry (700–1000 similar his state reduction Chidumayo's Particularly rapid based ring Gourlay (1995) Kanowski (1991) S2), 1 Apparently, good rings S2). suggest reliably Stahle 1999; Others disagree. (2010) warn increases old ages. (2008) recognise 89% rings. Double (Gammadid, 1989; Gourlay, Martin Moss, 1997) occur seasons. can erratic rainfall. February (2006) concluded ‘without cross-dating, chronology precise determination area’. must treated care. Still, give support hypothesis too reached supposed peak, zeyheri, contrary, early Holdo (2006). shows trend (significant regression; 1f). Other authors depending (Gourlay 1996; Holdo, 2006; 2017). term, cannot firmly refute any hypothesis. showing fast traditionally protected rocky hill, obviously cut. They bigger stature recent after protection. hypothesised ability would without. importance itself, surroundings, Högberg, 1986; Bakhoum 2018; Nasto Acacias (albeit variable species), not. found, however, Högberg (1986) points limit non-nitrogen-fixing phosphorus fixers. mentioned above, denuded up 1970s. nutrients fixers phosphorus. Except cinerea, 1997/1998 expected result conforms numerous Gammadid, Chidumayo, 2005; Gebrekirstos 2008). deviating D. shrubby form, causing reaction measured. Namibia (Shikangalah 2020), albeit drier climate. Years extreme KIH. mm, eight 900 1932–1990 (Ngana, Such appear plants. (2005, interacted temperature. (2006, 2010), Sahel Nicolini (2010). noted temperature precipitation. if factors overshadow Zimbabwe depend (Grundy, showed cent accounted Pterocarpus angolensis Shackleton (2002). related amounts ‘during heart season’ (Stahle 2007; 2010). amount distribution appears important. agriculture, forest use, consequences erosion, complex, nutrient-deficient. Restoration activities included, example fast-growing Eucalyptus spp. Grevillea robusta. Decisions local involvement. Ca. aside fuelwood, amidst people's fields. variety seedlings produced rural people, fruit wood lots. Measures improve infiltration slow 1984). Burning cutting banned. These measures, motivated technical viewpoint, never accepted conflicted traditional conservation (Östberg, Planting modify assemblies natural vegetation, thereby affecting subsistence (Fleischman 2020). Many confirm changes following heavy restrictions Africa, Kebrom Tekle (2001), O'Connor (2014) Demel Teketay (2018). 1970s, decided abandon leave considered costly management. 1979, approach effects improved grass enforced unpopular owners. ban adhered to. When started already changed partly surface mostly well-vegetated landscape. Harpachne schimperi Melinis repens stage succession. successional Themeda triandra rare. 1990s, once again entered scale, illegal (Kangalawe widespread practice uncontrolled. Individual indigenous, self-sown re-sprouting normal area. Generally, regenerate quite sites, far motivate measures. seed bank lacked (Lyaruu Backéus, 1999), suggesting dispersal Shortage will eventually through processes crusts (Aranibar 2003; Dalton 2004; 1986). Regeneration outcome great interacting anthropogenic practices environmental heterogeneity. undertaken suitable areas, seems viable heavily landscapes remnants left original vegetation. grateful A.C. Mbegu, founding director project, facilitated ways, successor Z.A. Matari, part fieldwork. Dr Z.K. Rulangaranga team Among helped us fieldwork, want mention A.A. Mbonde, C.B. Kweyunga H.O. Suleimani. Dr. L. Skoglund fieldwork sorting. Prof. H. Rydin considerably statistical analyses. None. findings openly available https://doi.org/10.5878/5s6y-wv36. Please note: publisher responsible content functionality supporting supplied authors. Any queries (other missing content) directed corresponding author article.