Effects of anthropogenic fires on soil properties and the implications of fire frequency for the Guinea savanna ecological zone, Ghana

Journal Pre-proof

Effects of anthropogenic fires on soil properties and the implications of fire frequency for the Guinea savanna ecological zone, Ghana Esther Ekua Amoako , James Gambiza PII: DOI: Reference:

S2468-2276(19)30762-8 https://doi.org/10.1016/j.sciaf.2019.e00201 SCIAF 201

To appear in:

Scientific African

Received date: Revised date: Accepted date:

6 January 2019 2 September 2019 10 October 2019

Please cite this article as: Esther Ekua Amoako , James Gambiza , Effects of anthropogenic fires on soil properties and the implications of fire frequency for the Guinea savanna ecological zone, Ghana, Scientific African (2019), doi: https://doi.org/10.1016/j.sciaf.2019.e00201

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier B.V. on behalf of African Institute of Mathematical Sciences / Next Einstein Initiative. This is an open access article under the CC BY-NC-ND license. (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Highlights  We investigated fire effects on soil properties in burned and unburned land use types 

We examined the implications of fire frequencies for the Guinea savanna, Ghana



Fire aided the mineralisation of some nutrients, as burned fields had higher mean values of Total N, OC, pH and Ca than unburned fields



Further studies are needed to fully understand the importance of fire regimes on soils in the Guinea savanna

Effects of anthropogenic fires on soil properties and the implications of fire frequency for the Guinea savanna ecological zone, Ghana Esther Ekua Amoako 1, 2*, James Gambiza1 1

Department of Environmental Science, Rhodes University, P.O. Box 94, Grahamstown, South Africa 2

Department of Ecotourism and Environmental Management, University for Development Studies, P.O. Box 1882 Tamale, Ghana *Corresponding author: Email: [email protected] (Esther Ekua Amoako) Declaration of interest None. ABSTRACT Fire is an important factor influencing the structure and function of tropical savannas. In spite of the extensive studies conducted on the effects of fire on soils in savannas, there are relatively few studies focusing on the Sudano-Guinean savanna of West Africa which experiences recurring fires in the dry season. The fires are anthropogenic and are mainly caused by hunters and farmers to flush out animals, remove debris from crop fields and to improve soil fertility. We investigated how the bush fires influence soil properties in four land use types in six districts in the Guinea savanna of Ghana. Data on fire counts were obtained from the CSIR Meraka Institute, South Africa and fires densities calculated for each district. Soils were sampled in burned and unburned woodlands and crop fields and analysed for pH, available P, Total N, OC, Ca, Mg, CEC, EC and texture. The fire densities varied amongst the selected districts. Of the six districts, the East Gonja district recorded the highest fire density (0.82 fires km-2). Tamale recorded the lowest density (0.32 fires km-2). Total N, OC, pH and Ca differed significantly across the different land use types. A principal component analysis showed a stronger association

2

and more positive gradient in woodlands than in crop fields. Total N and OC, showed a positive association, whereas silt showed a negative association to sand and clay. High fire frequencies were recorded in districts with high grass fuel loads and abundant wildlife. Fire aided the mineralisation of nutrients as burned fields had higher mean values of nutrients than unburned fields. Further studies are needed to fully understand the importance of fire regimes on soils in the Guinea savanna. Traditional ecological knowledge on the use of fire could be harnessed to reduce indiscriminate vegetation burning in the region. Keywords: bush burning; land use; soil properties; fire counts; Guinea savanna

3

1

Introduction

African savannas constitute roughly 50% of global terrestrial ecosystems (Lehmann et al. 2011; Osborne et al. 2018). Studies have revealed, however, that in recent years the African savanna has undergone a rapid transformation through anthropogenic activities including the indiscriminate anthropogenic use of fire (Tarimo et al. 2015; Bond et al. 2016; Dwomoh & Wimberly 2017). A number of researchers (Trollope & Trollope 2002; Andersson, Kjøller,&, Struwe 2003; Coetsee et al. 2010; Pricope & Binford 2012; Sluyter & Duvall 2016; Dwomoh & Wimberly 2017) found that there is continuous burning of vegetation in Africa. Satellite images and ground truthing confirm a pattern of burning from January to April in West Africa and from July to October in the East and parts of Southern Africa (N’Datchoh et al., 2015; Tarimo et al. 2015; Knowles et al. 2016; Archibald 2016; Dwomoh & Wimberly 2017). The authors highlight the extensive and frequent use of fire in the African savanna: this is why Africa is referred to as the ‘fire continent’ (Knowles et al., 2016). Land use practices – including the use of fire – have enormous effects on both vegetation and soils (Scholes and Walker, 1993; Van Langevelde et al., 2003; Veasey et al., 2013). It is noted that soil-fire relationships predominantly define the structure, function, and dynamics of savanna ecosystems (Bassett & Crummy 2003; Van Wilgen 2009; Certini 2014). Soil properties, therefore, is fundamental to the sustainability of savannas and requires further monitoring and greater attention than it has received (Basett et al. 2003; Vogt et al. 2015; Doerr & Santı 2016). Through the combustion of biomass, soil properties are altered and different soils are influenced differently. While some nutrients, such as nitrogen (N) and phosphorous (P), are volatilized and

lost, calcium (Ca) is made more available (St. John & Rundel 1977). Fire is thus a rapid mineralising agent, in contrast to natural processes, which may require years, or in some cases, decades, for decomposition to occur ( Nardoto & Bustamante 2003; Certini 2005; Pivello et al. 2010; Hanan et al. 2017). Additionally, the removal of vegetation cover during burning exposes the land to various forms of degradation, including erosion, leaching and reduction in soil porosity (Ferriera et al 2008). Slow infiltration and percolation in soils have also been attributed to the burning of vegetation, which leads to high runoff resulting in the topsoil being washed away (Bird et al. 1999; Kato & Haridasan 2002). The changes in soil composition and processes are, however, dependant on the fire regime (season, frequency, intensity, time, size and pattern, of fire). The study of fire effects on soils in fire-adapted Sudano-Guinean savannas (which cover approximately 1.3 million km-2 (35%) of the total West African region) is therefore important. There are a number of studies (Andersson et al 2003; Sackey & Hale, 2009; Dayamba et al. 2010; N’Datchoh et al. 2015; Cardoso et al. 2016; Amoako et al. 2018) on how biomass burning influences vegetation in these savannas, but very few studies have been done on soils, and how the multipurpose uses of fire influence fire occurrences. Kugbe et al. (2012, 2014) investigated seasonal burns and the impact of nutrient loss, whilst N’Datchoh et al. (2015) studied the interaction between climate-related and anthropogenic fire regimes. Bagamsah (2005) investigated the impact of a seasonal burn on both vegetation and soil in the Northern Region of Ghana. Pyne (2003) authored an essay on the use of fire in agriculture, and the names attributed to the the various occassions that fire is used within the cropping season in the Southern part of Ghana. Pyne (2003) also proposed ways of protecting forest reserves from wildfires in Ghana.

5

The few studies conducted on savanna fires in Ghana have indicated that biomass burning, particularly in the dry season, has been a trend in the Northern Region of Ghana for decades, with records dating back to colonial times (Korem 1985; Pyne 1999; Andersson et al. 2003; Bagamsah 2005; Kugbe et al. 2012; Kugbe et al. 2014). The studies argue that bushfires are due to the uncontrolled use of fire for hunting, charcoal production, and crop production (precropping and post-harvest), as most farmers in the north of Ghana use fire for land preparation for cropping. Alhassan et al. (1999) also reported that fires frequently occur in areas with large populations of game and livestock and indicated that the effect of fire on natural pastures and parklands in the savanna is enormous. In this paper, we assessed the effects of fire on soils in four land use types – burned and unburned crop fields and woodlands. We also discuss some implications of fire frequencies for the study districts and communities. Thus data presented here answers the questions: 1. How does fire influence burned crop fields and woodlands in the Guinea savanna? 2. What are the implications of burning and non-burning for the study districts and communities?

6

3. Methods 1.1

Study area

The study was conducted in the Northern Region of Ghana (9.5439° N, 0.9057° W) with a population density of 35 persons / km-2 (Ghana Statistical Service 2012). The region has an average altitude of 150 m above sea level with an undulating topography. Geologically, the soils are Voltaian sandstone characterised by inherently low organic matter content and are prone to erosion (Vanlauwe et al. 2002; Braimoh & Vlek 2004). The soils are locally characterised as savanna Ochrosols (Obeng 1956) ranging from black, alluvial soils and brown, salty soils, to stony/gravelly soils and are classified as Luvisols and Fluvisols in the World Reference Base for Soil Resources (Mikkelsen & Langohr, 2004). The climate is tropical with long dry spells between November and April. The region has a unimodal rainfall distribution with an annual mean of 1 100 mm. The peak of the rainy season ranges from July to September. Thus, the region has only one cropping season, as opposed to two cropping seasons as found in the south of Ghana. Rainfall exceeds potential evaporation which occurs over a relatively short period (Kranjac-Berisavljevic et al. 1999). The mean annual temperature is 27oC. The annual potential open-water evaporation in the north is comparatively higher (about 2 000 mm) than the south, which records 1 350 mm. The region has a dry climate due to its proximity to the Sahel and experiences north-easterly winds (Harmattan) between November and April which facilitates vegetation burning( KranjacBerisavljevic et al. 1999). The region occupies about 62% of the Guinea savanna ecological zone (147 900 km-2) of Ghana (Raamsdonk et al. 2008). It is characterised by fire- and drought-resistant woody species such as

7

Vitellaria paradoxa, Combretum spp, Burkea africana, and Isoberlinia doka with common grasses such as Andropogon, Heteropogon and Hyparrhenia spp. A third of natural pastures (71% of 235 000 km-2) in the country fall within this region (Raamsdonk et al. 2008) (Fig.1).

Figure 1: Study area showing sampling sites. Farming is the predominant occupation in Ghana and this is not different in the Northern Region. Agriculture is practised on a small scale, with a farm-family cultivating one or more farm holdings of one to two hectares each (Gyasi et al. 1995; Braimoh and Vlek 2006). Farmers typically grow crops and keep animals: ruminants and poultry. Both land and crop rotation are commonly practiced in the region, and grass in the crop fields is burned, as part of land preparation before the soil is ploughed and hoed. The burning is done early (December to end

8

January) or late in the dry season (February to April) depending on the type of crop to be cultivated. Some fields may be burned twice to get rid of any debris from the previous burn. 1.1.1 Selection of study sites and fire count data Based on a report on daily fire counts in Ghana, received from the Advanced Fire Information System and the CSIR, Meraka Institute, 18 districts in the Northern Region, with fire frequency data available, were stratified into high (> 1 fires km-²) and low ( < 1 fires km-²) zones. Of the 18 districts, six districts and ten communities were purposively sampled with the assistance of the regional office of the Ghana Forestry Commission. Purposive sampling (Lavrakas, 2008) was necessary to include the few communities that practice non-burning and also for financial constraints. Community leaders were contacted to assist in the selection of sampling sites. The research team was comprised of teaching assistants from the University for Development Studies, Tamale, and guards from the Regional Forestry Commission. Unburned crop fields and woodlands were selected from two communities (Nwodua and Katabanawa) that practise no burning with a few found in the other communities. In the remaining communities we purposively selected: fire using (before/after harvest) and non-fire using households (herein after referred to as burned and unburned crop fields), and burned and unburned woodlands. We sampled at least two land use types in each of the selected communities. A summary of the characteristics of the selected districts and communities is presented in Table 1.

9

Table 1: Geographic, demographic and fire count characteristics of selected communities and districts. Community

District/municipality

Land area (km-²)

Population size

Average household size

Tugu

Tamale (9.4034°N,0.8424° W)

731

233,252

6 persons

Jagriguyilli & Nwodua

Tolon-Kumbungu (9.4310°N, 1.0649°W)

2,389

129,156

11 persons

96.6

0.36

4,892

121,117

8 persons

85.5

0.40

Wungu & Katabanawa Kpligini

West Mamprusi (10.352°N, 0.799°W) 4,160

199,600

9 persons

92.1

0.65

17,317

41,180

7 persons

82.5

0.60

9,351

135,450

4 persons

76.5

0.85

Damongo-Agric & Mognori

Mion (9.4450°N,0.0093° W) West Gonja 9.084°N, 1.818°W

Kushini &Kpalbe

East Gonja (8.5509°N 0.5183°W) Source: Ghana Statistical Service (2012)

Household in agriculture (%) 26.1

Fire density (km-²) 0.32

1.1.2 Sampling design and laboratory analyses Three plots of 100 m by 150 m were randomly selected in each of the four land use types. Plot sizes were determined based on the average household farm size which range from one to two hectares. Soils were collected at the depth of 0 - 5 cm (Rashid 1987; Pivello et al. 2010) in five (20 m by 20 m) subplots, which were randomly demarcated in the larger plots (100 m by 150 m). Samples were collected from the four corners and the centre of each subplot and bulked together to form a composite sample of the five, for each large plot. Soils were sampled from 3 March to 27 April 2017, during the late dry-season. A total of 151 composite samples from the four land use types were analysed for pH, organic carbon (OC), Total nitrogen (N), Available phosphorous (P), potassium (K), magnesium (Mg), calcium (Ca), cation exchange capacity (CEC), electrical conductivity (EC) and soil texture (Sand, silt and clay) at the Ecological Laboratory (Ecolab), University of Ghana. Samples were 10

air-dried, crushed and sieved through a 2 mm mesh. Soil pH was determined in a soil suspension of 1:2 and 0.01M CaCL2 using a pH meter (Kalra & Maynard, 1991). OC content was determined by wet oxidation – a modified Walkley-Black method (Walkley & Black, 1934). Total N determination was based on the macro Kjeldahl method (Bremer 1960). Exchangeable bases (Ca, Mg and K) were determined in 1N NH4 OAc extract of soil. Ca and Mg were determined by atomic absorption spectrometry (AAS) and K by flame photometry (Moss, 1961). Available P was determined by the Bray 1 method (Bray & Kurtz, 1945) using 1 g air dry soil in an extraction solution of 15 ml of 1M NH4F and 25 l of 0.5M HCl and 460 ml distilled water. Electrical conductivity was determined by a conductivity meter and cell method (Piper, 1942; Rayment & Lyons, 2012). Soil texture was determined by the sieve and hydrometer method (Wen, Aydin, & Duzgoren-Aydin, 2002). 1.1.3 Data analysis Data on soil nutrients from the four land use types and fire count data were checked for normality using the Shapiro Wilk test. Mean fire counts and fire densities were calculated for each district. One way analysis of variance (ANOVA) was used to test for the significance differences in the fire counts for a five year period and data on soil properties were tested with Welch’s ANOVA (McDonald, 2014) for statistical significances among the four land use types. Welch’s ANOVA allows for heterogeneity of variance. Tukey’s Honestly Significant Difference (HSD) was used to compare the differences in means among the land use types. Results were reported as statistically significant at α = 0.05. A principal component analysis (PCA) and cluster were used

to analyse the soil data. The data analyses were done using R version 3.4.2 and

RcmdrPlugin FactoMineR (Fox & Bouchet-Valat 2018).

11

2

Results

2.1

Fire frequencies of study districts

There were significant differences (F

5, 80.49

= 7.4, p <0.001) in monthly fire counts across the

study districts (November to March) over a period of five years (2013 to 2017). The highest mean monthly fire count for East Gonja (256.4±608.19) was 33 times higher than the counts in Tamale (7.83±19.22) which recorded the lowest fire count (Table 2). Monthly fire counts in Tamale differed significantly from counts in East Gonja (p < 0.001), West Gonja (p < 0.01), and Mion (p = 0.01) (Table 2).

Table 2: Mean monthly counts (±SD) in the study districts in the dry seasons (2013 – 2017). District Mean ±SD

Fire count Maximum Minimum

Tamale

7.8± 19.22a

122.0

38.0

Tolon-Kumbungu

52.9 ±12.69ab

585.0

343.0

West Gonja

56.7±686.81b

3958.0

2889.0

West Mamprusi

81.5±75.61b

1160.0

713.0

Mion

91.4 ±201.49b

786.0

619.0

East Gonja

256.4± 608.19c

4068.0

1958.0

Source: AFIS/CSIR Meraka Institute 2.2

Effect of fire on soil properties

Soil nutrient levels varied across the four land use types (Table 3). Total N, OC, Ca and pH showed significant differences between burned or unburned land use types. Soil pH levels varied among the land use types

(F3, 70 = 4.36, p = 0.01). pH in burned woodlands was significantly

12

higher than that found in unburned woodland (p = 0.05) and unburned crop field (p = 0.01). pH in woodlands were 1.03 times higher than crop fields. There were significant differences in OC levels across the land use types (F3, 63 = 13.84, p < 0.001). Percentage OC in burned woodlands differed significantly from unburned woodland (p = 0.01), burned crop fields (p = 0.02) and unburned crop fields (p < 0.001). Percentage OC was 1.2 times higher in burned land use types than in unburned land use types and 1.2 times in woodland than that found in crop fields (Table 3). Total N was also significant across all the land use types (F 3, 65 = 20.04, p < 0.001) with levels in unburned crop fields significantly different from burned crop fields (p = 02). Burned woodlands was also significantly different from burned crop fields (p = 03), unburned crop fields (p < 0.001) and unburned woodland (p < 0.001). The levels of N were 1.14 times higher in burned than the unburned land use types. Ca levels were significantly different (F

3, 65.

= 3.32, p = 0.05). Ca in burned woodland was

significantly different from unburned crop fields (p = 04). The mean Ca was 1.04 times higher in burned than unburned land use types and 1.22 times higher in woodlands than in crop fields. There were no significant differences in available P (F 3, 58 = 1.69, p = 0.17), Mg (F 3, 59.83 = 0.76, p = 0.52), K (F 3, 58.16 = 0.57, p = 0.64), EC (F 3, 60.21 = 1.03, p = 0.38), CEC (F 3, 63.55 = 0.86, p = 0.46) across the land use types. For the textural properties, there were significant differences in the clay content of soils across the four land use types (F

3, 64.03

= 4.86, p = 0.004). Unburned crop fields was significantly

different from burned crop field (p = 0.02), and unburned woodland also differed significantly from burned crop fields (0.03). Silt also differed significantly (F 3, 58 = 1.34, p = 0.04) across the 13

land use types, with burned crop fields showing high significant differences from unburned woodlands (p = 0.03), and burned woodland (p = 0.05). There were no significant differences in the sand content across the different land use types (F 3, 65.09 = 0.49, p = 0.68) (Table 3). Table 3: Mean (±SD) of soil properties in burned and unburned crop fields and woodlands. For a given variable, values labelled with the same letter in a row were not significantly different at p > .05 (Tukey’s HSD). Soil variable

Crop fields Unburned Burned

Woodlands Unburned Burned

pH

6.5± 0.66a

6.8±0.26ab

6.7±0.58ab

6.9±0.59b

OC (%)

1.3±0.34a

1.6±0.51a

1.6±0.55a

1.9±0.50b

Avail. P (mg kg-1)

14.6±8.54 a

16.1±6.82a

17.3±6.66a

17.1±8.14 a

Total N (%)

0.07±0.01a

0.08±0.01b

0.07±0.01ab

0.08±0.01c

K (cmol kg-1)

0.7±0.54a

0.6±0.68a

0.6±0.53a

0.7±0.63a

Ca (cmol kg-1)

3.7±1.49a

3.7±0.86a

4.1±1.37b

4.5±1.36b

Mg (cmol kg-1) CEC (cmol kg-1)

3.0±1.97a 9.0±3.63a

3.0±0.91a 8.8±1.66a

3.12±1.07a 9.1±2.23a

3.26±1.69a 4.9±1.59a

172.6±137.80a 42.0±7.62a

146.0±74.30a 40.6±8.54a

190.6±119.55a 40.6±12.52a

174.0±107.31a 40.4±9.57a

43.8±16.39b

36.9±13.88a

44.6±12.62b

44.3±12.34b

Clay (%) 14.2±15.02a Source: Field data, 2017

22.6±14.26b

14.9±14.34a

15.2±13.04a

EC µs/CM Sand (%) Silt (%)

Burned crop field n=20; unburned crop field n=27; burned woodland n=53, unburned woodland n=51

The output of the Principal Component Analysis (PCA) of nutrient variables showed more than 51 percent of the variance in the dataset (Fig 2). Dim 1 and Dim 2 accounted for 32.88 percent and 18.63 percent respectively, which showed a positive association and an increasing gradient between Total N and OC (0.7), and Available P and Ca (0.5) in woodland. Soil pH and the 14

exchangeable ions also showed a somewhat good association in unburned crop fields and unburned woodland, whereas silt showed a negative association (-0.5) to sand, clay, TN and OC. The individual samples in woodlands showed a much wider variability with burned woodland showing a positive association in clay, OC and TN. More individuals in crop fields showed a negative association than individual in woodlands. There were a few outliers from the woodlands, burned and unburned crop fields (Fig.2)

15

Fig. 2: Biplots of soil variables in the four land use types (cb, cub, wb and wub denotes burned crop fields, unburned crop field, burned wood land and unburned woodland, respectively). (Fig. 2 COLOUR IN PRINT) The dendrogram showed four major clusters from the 151 samples from the four land use type. The clustering shows similarities between the soil samples in the various land use type with a high similarity between samples in woodland burned and unburned in the first cluster. The last cluster shows more similarities across burned and unburned woodlands and unburned crop than burned crop fields with a few outliers in the woodlands (Fig 3).

16

Figure 3 A dendrogram showing four clusters in samples from the different land use types (Fig. 3 COLOUR IN PRINT)

3 3.1

Discussion Fire effects on soil properties

Generally, soils in the north of Ghana are low in nutrients, however, the levels found in this study are slightly higher than what has been reported in other studies (Bagamsah 2005; Tahiru et al. 2015). The varying nutrient levels found in this study may be attributed to the depth at which soils were sampled, the time of burning, the intensity of fire, the moisture content in soil, and the fuel load at the time of burning, which differ between woodlands and crop fields (Sharrow & Wright, 1977; Bird et al., 1999; Kugbe, Fosu & Vlek, 2014).

17

The significant levels of pH, OC, Total N and Ca content in burned woodland and crop fields showed that fire chemically aids in the mineralisation of decomposable material on the surface or in topsoil, (Owensby & Wyrill 1971; Sharrow & Wright, 1977; Mitros et al. 2002) which was the case of this study as the soils were sampled at the depth of 0-5cm. The mean pH range was neutral (6.5 - 6.9) found in this study is rather common to mineral soils in humid regions (FAO 1984). The mean values of pH across the land use types were somewhat higher than that reported in a similar study (Bagamsah 2005) in the Guinea savanna of Ghana. Again, this is probably due to the depth (0 - 5) at which soils were sampled in this study whereas it was 0 -10 cm in Bagamsah 2005 as found by Bessah et al. (2016) that soil nutrient levels varied with depth. The intensity of fire and the time at which samples were taken could also be a reason for the high pH values recorded in this study, as the peak burning time between December and January which is early dry season burn is associated with low intensity fires (Bagamsah, 2005). Owensby and Wyrill (1971) found that the charring of fuel during burning influenced the higher level of pH in soils in Kansa rangelands during early spring burns than late spring burns. Wardle et al. (1998) found that high pH levels in burned debris aids in binding and releases phenolic compounds that enhance nutrient availability in soils. The significantly high levels of OC found in burned crop fields (1.6±0.51) burned woodland (1.9±0.50) unburned crop field (1.3±0.34) and unburned woodlands (1.6±0.55) fall within the optimum levels (0.6% - 2%) of the Guinea savanna woodland (FAO 2005). The significant differences in burned land use types from unburned land use types could also be attributed to charring due to incomplete combustion of organic material (Kauffman, Cummings, & Ward, 2018). This confirms a study that found an increase in OC content in the top soil (0 - 10 cm) of the Zambian Miombo woodland after fires (Stromgaard 1992). Furthermore, Rau et al. 2008 18

reported an increase in OC at the top surface soils of both burned and burned woodlands but which decreased with increased depth; this could be attributed to the significantly high level of OC found in this study, as soils were sampled at the topsoil. Contrarily, Certini 2005 argued that fire induced increase in OC could be due to the decline in mineralisation rate of organic matter as a result of decrease in microbial activity during and immediately after fires. The relatively low mean OC content found in both burned and unburned crop fields could be attributed to plant uptake (Stromgaard 1992) as well as tillage practices which includes the use of ploughs for land preparation (Diao, Cossar, Houssou, & Kolavalli, 2014), which has been shown to decrease OC content in crop fields (Riezebos & Loerts, 1998; Chan & Heenan; Flávia, Pinheiro, Vilas, & Campos, 2015). Chan & Heenan (2005) found that unburned crop fields had higher OC than burned crop fields. The highly significant levels of Total N found in burned plots indicate the rapid oxidation of N in burned fields, which normally occurs four days after the fire as shown in similar studies (Sharrow & Wright 1977; Nardoto & Bustamante 2003; Pivello 2011). However, the studies indicated a decline in N levels a month after fire, when compared to pre-burned levels (Sharrow & Wright 1977; Nardoto & Bustamante 2003; Duguy et al. 2007; Pivello 2011). Nardoto & Bustamante (2003) also reported higher mineralisation of N in unburned than burned fields, a year after fire. Contrarily, Mitros et al. 2002, found no significance difference in Total N in both burned and unburned plots. The significant levels of Total N found in unburned crop fields could be attributed to the addition of inorganic fertilisers, as most farmers in the study area use compound fertilisers with nitrogen phosphorous and potassium (NPK to supplement crop nutrition (Kanton et al., 2016; Liu, Sung, Chen, & Lai, 2014; Martey et al., 2014). A study in

19

Ghana (Rhodes, 1995) also showed evidence of N deposition in soils from rainfall, which may contribute to the levels found – across the land use types. The association between Total N and OC in the PCA implies that an increase in Total N will correspond positively to OC (Schrumpf, Kaiser & Schulze, 2014). The strong positive association between individual samples in burned woodland implies that fire, to a large extent, had a positive influence on soils in woodlands – probably through the release of nitrogen that has been tied up in growing and dead plants and through rapid decomposition. This could also be due to the peak period of burning being in the early dry season, so that burning may not have a devastating effect on soil OC and N but rather enhance their availability which confirms studies (Sharrow & Wright, 1977; Raison 1979; Stromgaard 1992; Pivello et al. 2010) that have reported an increase in soil nutrient contents after low intensity fires. The levels, however, declined with time, particularly for N (Sharrow & Wright, 1977). Studies (Raison 1979; Stromgaard 1992; Pivello et al. 2010) have reported an increase in the soil nutrients after exposure to low intensity fire, which is most probably the case in this study, as the peak period of burning occurs during the early dry season. Other studies, however, reported a decline in N and P levels with time (Munk 1987; Bird et al. 1999; Jobbagy & Jackson; 2001; Bessah et al. 2016). The results could imply that burning enhances some soil nutrients only for a short period. However, burning for nutrients as practised by most subsistence farmers may not be a sustainable way of increasing soil productivity (Chan & Heenan, 2005). Both burned crop fields and burned woodlands showed relatively high levels clay. The highest level of clay in burned crop field can be attributed to negative effect of fire on soils, as was observed by Varela, Benito and Keizer, 2010. The same authors observed that effect of fire could cause water repellent on the topsoil leading to erosion. The levels of clay and silt found across 20

the land use types could enhance colloidal levels in the soils (FAO1985; Braimoh & Vlek 2004; Bessah et al. 2016). Some studies on the effects of fire on soils have found more sand in burned fields than unburned fields (Ulery & Graham 1993; Bird et al. 1999). 3.2

The implications of fire frequency in the study districts This study confirmed seasonal fire recurrence in the Guinea savanna (November to April)

(Korem 1985; Andersson et al. 2003; Bagamsah 2005; Sackey & Hale 2008; Kugbe et al. 2014). Previous studies on fire in the study area (Korem 1985; Bagamsah 2005; Sackey & Hale 2008) and observations made during the fieldwork are evidence of burning in most of the study communities. The data on fire in the savannas of Ghana is, however, limited to frequency, time or season, and does not include intensity/severity, characteristics or type of fire. The discussion on fire frequency in this study is based on fire counts and the occurrence of fire in the dry season. The districts that recorded high mean fire counts (Table 2) are large areas of open savanna woodland with high game populations and tall grass growth (Andropogon and Hyparrhenia species) (Andersson et al. 2003; Bagamsah 2005; Bassett et al 2003; Kugbe 2014). A good rain season promotes high growth of grass and coupled with the Harmattan conditions, could cause extensive fires. The high fire usage in this area may be attributed to hunting: to flush out rodents and small animals and attract animals for hunting, as well as for farming and charcoal burning (Korem 1985; Alhassan 1999; Day et al. 2014, Odadi et al. 2017). The West Gonja – for instance, is characterised by high game population and contains the Mole National Park (4 840.4 km-2). Most communities in this area engage in hunting and charcoal burning as an alternative livelihood activity during the dry season which involves the use of fire (Adongo, R., Nkansah, D.O. & Salifu & Adongo, 2012; Blench, Dendo, Road, & Kingdom, 2007). The provision of

21

alternative livelihood income sources to these communities and the responsible use of fire could curb the indiscriminate use of fire (CBD, 2011; Bonye, 2013). The fire low count districts are relatively dry with low rainfall for grass growth (Andersson, Kjøller & Struwe 2003). Burning in these areas may be of low intensity, however the fires may spread faster for a relatively short period than the areas with high grass growth and fuel load (Andersson, Kjøller & Struwe 2003; Hantson et al. 2017). Moreover, some of the communities, such as Nwodua, in the low counts districts are known to have practised non-burning for more than 20 years (Andersson, Kjøller & Struwe 2003). An observation made during sampling was that the communities, such as Nwodua and Katabanawa, have relatively small areas of arable land, hence the intensification of crop farming in these communities (Rhodes 1999; Kotu et al. 2017), thus found the community found non-burning as a way of improving soil productivity as observed by Chan & Heenan (2005). Interactions with farmers revealed that they do not burn the crop residue but rather plough the residue into the soil. In addition, some of them supplement soil nutrients by applying inorganic fertiliser, in order to maximise crop yield per unit area and to continuously use the same piece of land. Jagriguyilli community, on the other hand, is known for the preservation of one the largest sacred groves in the north of Ghana, the Jaagbo sacred grove (1 km-2). The grove is believed to house a twin god which protects the surrounding communities from diseases, thus the area is protected against fire through the creation of fire belts around the grove to ensure a peaceful abode for the gods (Oteng-yeboah 1996; Telly 2006; Omsby 2012). This was also observed in the Tugu community where there was evidence of fire very close to the sacred grove but not in the grove because the gods must be protected against fire. Again, large unburned areas were identified in Jagriguyilli and Nwodua, in the Tolon-Kumbungu district. These two communities 22

have been internationally recognised, with Jagriguyilli having been awarded a UNESCO heritage status since 1993 (Oteng-yeboah 1996; Telly 2006; Poreku 2014). These traditional practices could be a model for other communities within the study region. This could be achieved through community mobilization and sensitization. The lowest mean fire count recorded in the Tamale area, however, is attributed to the large human population with a relatively small land area, the expansion of the built environment and a decrease in the natural environment.

4 Conclusion This study confirms that vegetation burning is recurrent in the dry season in Guinea savanna ecological zone of Ghana – with East Gonja district recording the highest density amongst the study districts. The results showed varied nutrient levels across burned and unburned woodlands and crop fields. There was, however, a seemingly positive influence of fire on pH, Total N, OC and Ca in the soils. Total N, OC and pH were at optimum levels, according to the savanna recommended levels for crop production, and the results revealed a positive association between Total clay, N and OC. The proportions of sand and silt found in each land use were nearly homogeneous in content with very low clay content. The relatively high N and OC contents in burned crop fields and woodland confirm why farmers would burn: to increase soil nutrients for crop production. Nevertheless, rural communities that use fire for daily livelihood activities can be sensitized on the long-term implications of indiscriminate burning on soils. Cultural and traditional belief systems on burning and non-burning could also be harnessed to reduce the indiscriminate bush burning.

23

The conclusions drawn from this study are based on the frequency and the seasonality of fire in the Northern Region. The severity, intensity, size and pattern, which determine the behaviour and types of fire, are still understudied in the Guinea savanna. These factors are however, crucial in determining wholistic effects of fire on soils in the savannas of Ghana, and require further investigation. Acknowledgements We want to express our profound gratitude to the community leaders – the Regional Forestry Commission – for their immense support. We are very grateful to the research assistants of the Faculty of Natural Resources and Environment, University for Development Studies, Tamale, Ghana, who assisted in the collection of field data. Our sincere appreciation is extended to Dr Amos Karbo-Bah of the University of Energy and Natural Resources, Sunyani, Ghana, as well as to the CSIR Meraka Institute, South Africa, for providing a five-year data set on daily fire counts for Ghana. The financial support from the Organisation for Women in Science in the Developing World (OWSD) in Collaboration with Rhodes University is very much appreciated.

Funding The financial support from the Organisation of Women in Science in the Developing World in Collaboration with Rhodes University for the PhD Fellowship is limited to tuition and living expenses for the fellowship period. Ethics approval and consent to participate Not applicable Authors’ Contribution This is part of a PhD thesis by Esther Ekua Amoako (main author) under the supervision of Professor James Gambiza. References Adongo, R., Nkansah, D.O. & Salifu, S. M. A., & Adongo, R. (2012). Social and psychological aspects of communal hunting (pieli) among residents of Tamale Metropolis in the Northern Region of Ghana. African Journal of Hospitality, Tourism and Leisure, 2(2), 1–15. Andersson, Mi., Kjøller, A., Struwe, S. (2003). Soil emissions of nitrous oxide in fire-prone 24

African savannas. Journal of https://doi.org/10.1029/2002JD003345

Geophysical

Research,108(D20),4630.

Anthony D. Griffiths, Stephen T. Garnett, B. W. B. (2015). Fire frequency matters more than fire size: Testing the pyrodiversity–biodiversity paradigm for at-risk small mammals in an Australian tropical savanna. Biological Conservation, 186, 337–346. https://doi.org/https://doi.org/10.1016/j.biocon.2015.03.021 Bagamsah, T. T. (2005). The Impact of Bushfire on Carbon and Nutrient Stocks as Well as Albedo in the savanna of northern Ghana. (P. lL. G. Vlek, Ed.), Ecology and Development Series 25. Göttingen: Cuvillier Verlag. Basett, T. J., Koli Bi, Z., & Ouattara, T. (2003). Land Users and Landscapes: Fire in Savanna, Environmental Change and land Reform in Northern Cote d’Ivoire. In T. J. Basett & D. Crummy (Eds.), African Savannas: Global Narratives & Local Knowledge of Environmental Change (pp. 53–71). Irthlingborough: Woolnough. Bessah, E., Bala, A., Agodzo, S. K., & Okhimamhe, A. A. (2016). Dynamics of soil organic carbon stocks in the Guinea savanna and transition agro-ecology under different land-use systems in Ghana. Cogent Geoscience, 2(1), 1–11. https://doi.org/10.1080/23312041.2016.1140319. Bird, M. ., Veenendaal, E. ., Moyo, C., Lloyd, J., & Frost, P. (2000). Effect of fire and soil texture on soil carbon in a sub-humid savanna (Matopos, Zimbabwe). Geoderma, 94(1), 71– 90. https://doi.org/10.1016/S0016-7061(99)00084-1 Blench, R., Dendo, M., Road, G., & Kingdom, U. (2007). Working paper : agricultural production and the potential for commodity chain expansion in the three northern regions of Ghana in 2006. Bond, W. J., Laestadius, L., Veldman, J. W., Bond, W. J., Parr, C. L., Parr, C. L., Scheiter, S. (2016). Ecology. Ancient grasslands at risk. Science (New York, N.Y.), 351(6269), 120– 122. https://doi.org/10.1126/science.aad5132. Bonye, S. Z. (2013). Rural Livelihood diversity: Coping Strategies in Wa West Districts. European Scientific Journal, 9(35), 139–156. Bray, R. H., & Kurtz, L. T. (1945). Determination of total, organic, and available forms of phosphorus in soils. Soil Science, 59(1), 39–46. https://doi.org/10.1097/00010694194501000-00006 Cardoso, A. W., Medina-Vega, J. A., Malhi, Y., Adu-Bredu, S., Ametsitsi, G. K. D., Djagbletey, G., Oliveras, I. (2016). Winners and losers: tropical forest tree seedling survival across a West African forest-savanna transition. Ecology and Evolution, 6(10), 3417–3429. https://doi.org/10.1002/ece3.2133 Certini, G. (2005). Effects of fire on properties of forest soils: a review. Oecologia, 143(1), 1–10. https://doi.org/10.1007/s00442-004-1788-8 Certini, G. (2014). Fire as a Soil-Forming https://doi.org/10.1007/s13280-013-0418-2 25

Factor.

Ambio,

43(2),

191–195.

Chan, K. Y. and Heenan, D. P. (2005) ‘The effects of stubble burning and tillage on soil carbon sequestration and crop productivity in southeastern Australia’, Soil Use and Management, 21(4), pp. 427–431. doi: 10.1079/SUM2005357. Coetsee, C., Bond, W. J., & February, E. C. (2010). Frequent fire affects soil nitrogen and carbon in an African savanna by changing woody cover. Oecologia, 162(4), 1027–1034. https://doi.org/10.1007/s00442-009-1490-y Day, M., Gumbo, D., Moombe, K. B., Wijaya, A., & Sunderland, T. (2014). Zambia country profile Monitoring , reporting and verification for REDD +. Bogo: CIFOR. Diao, X., Cossar, F., Houssou, N., & Kolavalli, S. (2014). Mechanization in Ghana: Emerging demand, and the search for alternative supply models. Food Policy, 48, 168–181. https://doi.org/10.1016/j.foodpol.2014.05.013 Duguy, B., Rovira, P., & Vallejo, R. (2007). Land-use history and fire effects on soil fertility in eastern Spain. European Journal of Soil Science, 58(1), 83–91. https://doi.org/10.1111/j.1365-2389.2006.00802.x Dwomoh, F., & Wimberly, M. (2017). Fire Regimes and Their Drivers in the Upper Guinean Region of West Africa. Remote Sensing, 9(11), 1117. https://doi.org/10.3390/rs9111117 Fernandes, P. M. P. M. P. M., Davies, G. M. M., Ascoli, D., Fernández, C., Moreira, F., Rigolot, E., … Tekkaya, C. (2013). Prescribed burning in southern Europe: Developing fire management in a dynamic landscape. Journal of Environmental Management, 11(SUPPL. 1), 1606–1616. https://doi.org/10.1890/120298. Flávia, É., Pinheiro, M., Vilas, D., & Campos, B. De. (2015). Tillage systems effects on soil carbon stock and physical fractions of soil organic matter. Agricultural Systems, 132, 35– 39. https://doi.org/10.1016/j.agsy.2014.08.008 Gyasi, E., Agyepong, G., Ardayfio-Schandorf, E., Enu-Kwesi, L., Nabila, J., & Owusu-Bennoah, E. (1995). Production pressure and environmental change in the forest-savanna zone of Southern Ghana. Global Environmental Change, 5(4), 355–366. https://doi.org/10.1016/0959-3780(95)00070-5 Hanan, E. J., Tague, C. N., & Schimel, J. P. (2017). Nitrogen cycling and export in California chaparral: The role of climate in shaping ecosystem responses to fire. Ecological Monographs, 87(1), 76–90. https://doi.org/10.1002/ecm.1234 Hantson, S., Scheffer, M., Pueyo, S., Xu, C., Lasslop, G., Van Nes, E. H., … Mendelsohn, J. (2017). Rare, Intense, Big fires dominate the global tropics under drier conditions. Scientific Reports, 7(1). https://doi.org/10.1038/s41598-017-14654-9 Hassan, S. N., & Rija, A. A. (2011). Fire history and management as determinant of patch selection by foraging herbivores in western Serengeti, Tanzania. International Journal of Biodiversity Science, Ecosystem Services and Management, 7(2), 122–133. https://doi.org/10.1080/21513732.2011.617710 Jobbagy, E. G., & Jackson, R. B. (2001). The distribution of soil nutriments with depth : Global patterns of the imprint of plants. Biogeochemistry, 53, 51–77. 26

Kalra, Y. P., & Maynard, D. G. (1991). Methods Manual for Forest Soil and Plant Analysis. Edmonton, Alberta: Natural Resources Canada. Kanton, R. A. L., Prasad, P. V. V, Mohammed, A. M., Bidzakin, J. K., Ansoba, E. Y., Asungre, P. A., … Bidzakin, J. K. (2016). Organic and Inorganic Fertilizer Effects on the Growth and Yield of Maize in a Dry Agro-Ecology in Northern Ghana Organic and Inorganic Fertilizer Effects on the Growth and. Journal of Crop Improvement, 30(1), 1–16. https://doi.org/10.1080/15427528.2015.1085939 Kato, E., & M. Haridasan. (2002). The effects of fire on infiltration rates and surface sealing in a latosol under cerrado vegetation in central Brazil. In Forest Fire Research & Wildland Fire Safety (pp. 1–10). Millpress. Kauffman, B. J., Cummings, D. L., & Ward, D. E. (2018). Relationships of fire, biomass and nutrient dynamics along a vegetation gradient in the Brazilian Cerrado. Ecology, 82(3), 519–531. Knowles, T., Assede, E., Daitz, D., Geldenhuys, C., Scholes, R. J., & Trollope, Winston S.W., Zietsman, J. (2016). Africa- the Fire Continent: Fire in Contemporary African Landscapes. Korem, A. (1985). Bush fire and agricultural development in Ghana. (A. Korem, Ed.). Tema: Ghana Publishing Corporation. Kotu, B. H., Alene, A., Manyong, V., Hoeschle-Zeledon, I., & Larbi, A. (2017). Adoption and impacts of sustainable intensification practices in Ghana. International Journal of AgriculturalSustainability, 15(5), 539–554. https://doi.org/10.1080/14735903.2017.1369619 Kranjac-Berisavljevic, G., Bayorbor, T. B., Abdulai, A. S., Obeng, F., Blench, R. M., Turton, C. N., … Drake, E. (1999). Rethinking Natural Resource Degradation in Semi-Arid SubSaharan Africa : Tamale: Overseas Development Institute (ODI). Kugbe, J., Fosu, M., & G.Vlek, P. L. (2014). Impact of season, feul load and vegetation cover on fire mediated nutrient losses accross savanna agro-ecosystem the case of northern Ghana. Nutr Cycl Agroecosys, 102, :113–136. https://doi.org/DOI 10.1007/s10705-014-9635-8 Kugbe, J. X., Mathias, F., Desta, T. L., Denich, M., & Vlek, P. L. G. (2012). Annual vegetation burns across the northern savanna region of Ghana: period of occurrence, area burns, nutrient losses and emissions. Nutrient Cycling in Agroecosystems, 93(3), 265–284. https://doi.org/10.1007/s10705-012-9514-0 Lehmann, C. E. R., Archibald, S. A., Hoffmann, W. A., & Bond, W. J. (2011). Deciphering the distribution of the savanna biome. New Phytologist, 191(1), 197–209. https://doi.org/10.1111/j.1469-8137.2011.03689.x Liu, C., Sung, Y., Chen, B., & Lai, H. (2014). Effects of Nitrogen Fertilizers on the Growth and Nitrate Content of Lettuce ( Lactuca sativa L .), 4427–4440. https://doi.org/10.3390/ijerph110404427 Martey, E., Wiredu, A. N., Etwire, P. M., Fosu, M., Buah, S. S. J., Bidzakin, J., … Kusi, F. (2014). Fertilizer Adoption and Use Intensity Among Smallholder Farmers in Northern Ghana : A Case Study of the AGRA Soil Health Project, 3(1). https://doi.org/10.5539/sar.v3n1p24 27

McDonald, J. H. (2014). Handbook of Biological Statistics. Sparky House Publishing (3rd ed.). Baltimore: Sparky House Publishing,. https://doi.org/10.1017/CBO9781107415324.004 Mikkelsen, J. H., & Langohr, R. (2004). Indigenous knowledge about soils and a sustainable crop production, a case study from the Guinea Woodland Savannah (Northern Region, Ghana).GeografiskTidsskrift,104(2),13–26. https://doi.org/10.1080/00167223.2004.10649515 Mitros, C., Mcintyre, S., & Moscato-goodpaster, B. (2002). Annual burning affects soil pH and total nitrogen content in the CERA oak woodlands. Tillers, 3, 29–32. Munk, I. (1987). The Influence of Texture, Soil Depth and Geology on pH in Farmland Soils. ActaAgriculturae Scandinavica, 37(4), 407–418. https://doi.org/10.1080/00015128709436572 N’Datchoh, E. T., Konaré, A., Diedhiou, A., Diawara, A., Quansah, E., & Assamoi, P. (2015). Effects of climate variability on savannah fire regimes in West Africa. Earth System Dynamics, 6(1), 161–174. https://doi.org/10.5194/esd-6-161-2015 Nardoto, G. B., Maria, M., & Bustamante, M. M. D. C. (2003). Effects of fire on soil nitrogen dynamics and microbial biomass in savannas of Central Brazil. Pesquisa Agropecuaria Brasileira, 38(8), 955–962. https://doi.org/10.1590/S0100-204X2003000800008 Oteng-yeboah, A. . A. (1996). Biodiversity in three tradition groves in Ghana. In van M. de R. J. M. Hedberg, O., van der Maesen L. J. G., van der Burgt, X. M. (Ed.), The Biodiversity of African Plants : Proceedings XIVth AETFAT Congress (p. 863). Wageningen. Owensby, C. E., & Wyrill, J. B. (1971). Effects of Range Flint Hills Soil on Kansas. Range Management, 26(3), 24–27. Piper, C. S. (1942). Soil and plant analysis : a laboratory manual of methods for the examination of soils and the determination of the inorganic constituents of plantsNo Title. University of Adelaide, Adelaide. Pivello, V. R. (2011). The use of fire in the cerrado and Amazonian rainforests of Brazil: Past and present, 7(1), 24–39. https://doi.org/10.4996/fireecology.0701024 Pivello, V. R., Oliveras, I., Miranda, H. S., Haridasan, M., Sato, M. N., & Meirelles, S. T. (2010). Effect of fires on soil nutrient availability in an open savanna in Central Brazil. Plant and Soil, 337(1–2), 111–123. https://doi.org/10.1007/s11104-010-0508-x Poreku, G. (2014). Sacred groves and biodiversity conservation in the Tolon District, Northern Region, Ghana. Iceland: UNU-LRT. Pricope, N. G., & Binford, M. W. (2012). A spatio-temporal analysis of fire recurrence and extent for semi-arid savanna ecosystems in southern Africa using moderate-resolution satellite imagery. Journal of Environmental Management, 100, 72–85. https://doi.org/10.1016/j.jenvman.2012.01.024 Pyne, S. (2003). Ghana, a Land between: In Stephen J. Pyne, Smokechasing University of Arizona Press.

28

Raison, R. J. (1979). Modification of the soil environment by vegetation fires, with particular reference to nitrogen transformations: a review. Plant and Soil, 108, 73–108. Rashid, G. H. (1987). Effects of fire on soil carbon and nitrogen in a Mediterranean oak forest of Algeria. Plant and Soil, 103, 89–93. Rayment, G. E., & Lyons, D. J. (2012). New, comprehensive soil chemical methods book for australasia. Communications in Soil Science and Plant Analysis, 43(1–2), 412–418. https://doi.org/10.1080/00103624.2012.641802 Rhodes, E. R. (1995). Nutrient depletion by food crops in Ghana and soil organic nitrogen management. Agricultural Systems, 48(1), 101–118. https://doi.org/10.1016/0308521X(95)93648-W Riezebos, H. T., & Loerts, A. C. (1998). In influence of land use change and tillage practice on soil organic matter in southern Brazil and eastern Paraguay, 49, 271–275. Sackey, I., & Hale, W. H. G. (2008). Effects of perennial fires on the woody vegetation of Mole National Park, Ghana. Journal of Science and Technology, 28(2), 36–47. https://doi.org/10.4314/jgsa.v10i2.18038 Santín, C., Doerr, S. H., & Santı, C. (2016). Fire effects on soils : the human dimension. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1696), 20150171. https://doi.org/10.1098/rstb.2015.0171 Scholes, R. J., & Walker, B. H. (1993). An African Savanna. Managing. Cambridge: Cambridge University Press. https://doi.org/10.1017/cbo9780511565472 Schrumpf, M., Kaiser, K., & Schulze, E. (2014). Soil organic carbon and total nitrogen gains in an old growth deciduous forest in Germany. PloS One, 9(2), e89364. https://doi.org/10.1371/journal.pone.0089364 Sharrow, S. H., & Wright, H. A. (1977). Effects of litter on soil nitrate , temperature, moisture and Tobosagrass production in the rolling plains. Journal of Range Management, 30(July), 266–270. Sluyter, A., & Duvall, C. (2016). African Fire Cultures, Cattle Ranching, and Colonial Landscape Transformations in the Neotropics, 106(2), 294–311. https://doi.org/10.1111/j.1931-0846.2015.12138.x Stromgaard, P. (1992). Immediate and long-term effects of fire and ash- fertilization on a Zambian miombo woodland soil. Agriculture, Ecosystems & Environment, 41, 19–37. Tahiru, F., Fosu, M., Gaiser, T., Becker, M., Inusah, B. I. Y., Mutari, A., … Mohammed, A. M. (2015). Fertilizer and Genotype Effects on Maize Production on Two Soils in the Northern Region of Ghana. Sustainable Agriculture Research, 4(4), 76. https://doi.org/10.5539/sar.v4n4p76 Tarimo, B., Dick, Ø. B., Gobakken, T., & Totland, Ø. (2015). Spatial distribution of temporal dynamics in anthropogenic fires in miombo savanna woodlands of Tanzania. Carbon Balance and Management, 10(1), 18. https://doi.org/10.1186/s13021-015-0029-2

29

Trollope, W. S. W., & Trollope, L. A. (2002). Fire behaviour a key factor in the fire ecology of African grassland and savannas. Forest Fire Research and Wildland Fire Safety, 1–15. https://doi.org/ISBN 90-77017-72van Langevelde, F., Van De Vijver, C. A. D. M., Kumar, L., van de Koppel, J., de Ridder, N., van Andel, J., … Rietkerk, M. (2003). Effects of Fire and Herbivory on the Stability of Savanna Ecosystems. Ecology, 84(2), 337–350. Van Wilgen, B. W. (2009). The evolution of fire management practices in savanna protected areas in South Africa. South African Journal of Science, 105(9–10), 343–349. Vanlauwe, B., Diels, J., Lyasse, O., Aihou, K., Iwuafor, E. N. O., Sanginga, N., … Deckers, J. (2002). Fertility status of soils of the derived savanna and Northern Guinea savanna and response to major plant nutrients, as influenced by soil type and land use management. Nutrient Cycling in Agroecosystems, 62(2), 139–150. https://doi.org/10.1023/A:1015531123854 Varela, M. E., Benito, E., & Keizer, J. J. (2010). Effects of wildfire and laboratory heating on soil aggregate stability of pine forests in Galicia: The role of lithology, soil organic matter contentandwaterrepellency.CATENA,83(2–3),127–134. https://doi.org/10.1016/j.catena.2010.08.001 Vogt, D. J., Tilley, J. P., & Edmonds, R. L. (2015). Ecosystem Science and Applications. (K. A. Vogt, J. Chen, & H. Asbjornsen, Eds.). Berlin: Higher Education Press. Walkley, A., & Black, I. A. (1934). An examination of the Degtjareff method for determining soil organic matter, and a proposed modification of the chromic soil titration method. Soil Science, 37(1), 29–38. https://doi.org/10.1097/00010694-193401000-00003 Wen, B. P., Aydin, A., & Duzgoren-Aydin, N. S. (2002). A comparative study of particle size analyses by sieve-hydrometer and laser diffraction methods. Geotechnical Testing Journal, 25(4),434–442.

30