232 American Academic Scientific Research Journal for Engineering, Technology, and Sciences ISSN (Print) 2313-4410, ISSN (Online) 2313-4402 https://asrjetsjournal.org/index.php/American_Scientific_Journal/index Total Nutrient Element Status of Subsistence Agriculture Soils in Angola Lídia P. de S. Teixeiraa*, Ian A. Simpsonb aSchool of Biological and Environmental Sciences, University of Stirling, Stirling FK9 4LA, Scotland, UK bArchaeology Department, Durham University, Durham DH1 3LE, UK aEmail: sousateixeira@hotmail.com bEmail: ian.simpson@durham.ac.uk Abstract Soil nutrient status in subsistence agricultural soils of Angola is poorly understood yet is vital in planning support and development of agricultural systems. This paper establishes the total nutrient status for two contrasting subsistence agricultural areas in Angola, within Huambo and Luanda provinces. Based on the World Reference Base for Soil Resources criteria (IUSS Working Group WRB, 2015, 2022) four soil catenas in each of Huambo and Luanda Provinces are classified as haplic Ferralsols and eutric Cambisols respectively. Mean and range total nutrient element values for twelve elements are determined (N, Ca, K, Mg, Mn, P, S, Zn, Cu, Mo, Fe, Al) with the results showing high variability and indicating that the haplic Ferralsols are below sub- Saharan averages for these elements while the eutric Cambisols are above these averages. Statistical analyses of relationships between soil nutrients and landscape factors by applying ANCOVA and pairwise comparisons using Tukey and Bonferroni tests indicate that underlying parent material has the biggest influence on element concentrations, further modified by slope processes; profile pedogenesis has had minimal contribution to element variances. Our findings highlight the need for detailed local analyses when planning supportive and effective nutrient management interventions. Keywords: Soil nutrients; nutrient variances; landscape relationships; Ferralsols; Cambisols. 1. Introduction Soils are an essential natural resource making major contribution to human well-being by providing important ecosystem services that includes vital nutrient elements for plant growth [1, 2]. Nutrients and their chemical and biological interactions in the soil form the basis for development and yield of agricultural crops, with knowledge of nutrient status in soils essential for assessment of soil quality leading to the practice of sustainable agriculture and influencing human, animal, and soil health [3]. ------------------------------------------------------------------------ Received: 3/9/2024 Accepted: 5/9/2024 Published: 5/19/2024 ------------------------------------------------------------------------ * Corresponding author. https://asrjetsjournal.org/index.php/American_Scientific_Journal/index mailto:sousateixeira@hotmail.com American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 233 Determining the amounts of nutrient elements in soil generates information about soil fertility and agricultural management potential. In the developing countries of sub-Saharan Africa knowledge of soil nutrient status is well recognised as an important tool in land assessment and endeavours to increase the success and sustainability of crop production [4]. Angola is no exception where satisfying basic food needs as environment changes requires development of new policies in the agricultural sector and, as a foundation for this, better understanding of the patterns and dynamics of nutrient variation in agricultural soils. This is particularly the case in systems of subsistence agriculture where nutrient status is poorly understood. In light of these imperatives, our research aim is to create new knowledge on the status of total soil nutrient elements in two contrasting but intensively cultivated small-farm subsistence regions in Angola - within Huambo and Luanda Provinces [5]. Focused on four representative catenas in each of these provinces, our first objective is to characterise and classify the soils of these localities based on the FAO World Reference Base for Soil Resources field and laboratory criteria [6, 7]. From this foundation, our second objective is to assess total levels of macro- and micro- nutrients, acidity (pH) with depth in profile and across the four catenary sequences in each of the two regions, indicating total nutrient elements reserves. To give context to these analyses the findings are compared with similar studies from sub-Saharan Africa, placing the Angola analyses on a continuum of nutrient element levels across sub-Saharan Africa [2, 4, 8, 9]. Our third objective is to assess the relationships between nutrient levels and landscape factors that includes site, position within the catena and profile depth. This offers a way of explaining, and potentially predicting, the distributions of soil elements within the Angolan landscape. Together, the analyses embedded within the aim and objectives of the research indicate nutrient reserves and potential of two important subsistence agricultural regions in Angola and give foundations and base lines for planning future land management that recognises regional and local variabilities. 2. Materials and Methods 2.1 Study areas Huambo Province, Angola, is located in the highland central region of Angola, 450km southeast of the capital Luanda and is in the tropical climate zone with a rainy and a dry season (Köppen, Cwb) (Figure 1). The rainy season generally lasts eight months with the possibility of rain starting in September and ending in mid-May. The average annual precipitation ranges from ca. 1200 mm to 1500 mm with an average annual temperature between ca. 22°C and 24°C. The province is one of the richest agricultural in the country and in recent years (since the end of the civil war in 2002) agricultural land use extent is amongst the fastest increases in the country. More than 50% of the population works in agricultural production, with 85% of this group working in subsistence agriculture using rudimentary production tools [10]. The main crops are maize and millet/sorghum, following by beans, sweet potatoes, and coffee. Luanda Province, Angola, is located in lowland Angola, around Luanda, the country’s capital and largest city and with the Atlantic coast forming its western boundary (Figure 2). It has a hot semi-desert climate (Köppen, Bsh) with a mean temperature of ca. 25.4°C and a yearly precipitation of ca. 387mm but with high variability. It is the richest and most developed province in the nation, home to large industrial services, commercial centres American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 234 and one of the largest agricultural centres in country (the Quiminha area, government owned and managed) created after the end of the civil war. Despite these developments rudimentary subsistence agriculture continues to predominate and supports a significant population. Agricultural activity is based on the production of cassava, bananas, and vegetables. Figure 1: Study Provinces in their Angolan setting with location of study areas. a) Huambo Province - Bailundo, Lepi (Longonjo), Mungo and Ngongoinga; b) Luanda Province - Bom Jesus, Funda, Ramiro and Talelo (Calumbo) 2.2 Field Survey and sampling Four representative localities within Huambo Province and four representative localities within Luanda Province were selected for sampling and analyses. The locations in Huambo Province were at Bailundo (12°11′45″S 15°51′20″E), Mungo (11°40′S 16°10′E), Ngongoinga (12°54′24″S 15°11′11″E) and Lepi (12°52′S 15°24′E) (Figure 1) in undulating moderately sloping topography. Locations in Luanda Province were https://geohack.toolforge.org/geohack.php?pagename=Bailundo¶ms=12_11_45_S_15_51_20_E_region:AO_type:landmark https://geohack.toolforge.org/geohack.php?pagename=Mungo,_Angola¶ms=11_40_S_16_10_E_region:AO_type:city(113417) https://geohack.toolforge.org/geohack.php?pagename=Longonjo¶ms=12_54_24_S_15_11_11_E_region:AO_type:city(92103) https://geohack.toolforge.org/geohack.php?pagename=L%C3%A9pi¶ms=12_52_S_15_24_E_region:AO_type:city_source:GNS-enwiki American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 235 at Talelo (08O54’09’S 1322’20OE), Funda (08°46′37″S 13°22′18″E, Bom Jesus (9°10′07″S 13°34′00″E) and Ramiros (9.0513°S 13.0192°E) on undulating, gently sloping topography. (Figure 2). The catenary sequences at each locality were typically 500m in length and three soils profiles assessed corresponded to upper, middle and lower areas of the catena. The twenty-four 1xx0.5x0.8m profiles, approximately to crop rooting depth, were hand-dug. Soil profiles were described in the field to give horizon definitions (A1, A2, and B) using Munsell Colour, texture class (including stoniness) and structure class. Bulk soils sample were systematically collected at 10cm intervals down the profile giving a total of 152 samples for analyses and undisturbed samples were collected in 8x5x5cm Kubiena tins from the A2 horizon for micromorphological investigation. 2.3 Laboratory analyses All analyses were undertaken at the Instrumentation and Micromorphology Laboratories, University of Stirling. For particle size distributions analyses air dried <2mm fraction samples were dispersed with sodium hexametaphosphate with four-hour agitation. Particle size fractions were determined by laser diffraction with a calibrated LS Coulter Counter (model (LS 230) on three replicate samples. Representative samples of undisturbed soils were collected from the A2 horizon of each profile to enable assessment of weatherable minerals and clay pedofeatures for soil classification. Thin section manufacture following standard procedures of acetone replacement of water, resin impregnation and curing, mounting on a glass slide, slicing then lapping to 30μm thickness followed by petrological microscope assessment with description of features following international protocols [11, 12]. Further characterisation of mineral grains was undertaken by scanning electron microscopy (SEM/EDX), Zeiss EVO/MA15 operating under variable pressure (60 Pa). Slides were viewed using a backscatter detector with an accelerating voltage of 20 kV, a filament current of 2.542 A and a beam current of 100 μA with working distance of 8.5 mm to optimise EDX detector geometry Point counts were performed with a count time of 20 seconds with identified automatically using Zeiss Aztec software; element concentrations (wt %) for each analysis were normalised to 100%. Air dried <2mm fraction samples for macro- and micro- element analyses were digested with HNO3 and sealed heating with Sartorius brand Mars / CEM model microwave synthesis giving total extraction. Element determination with replicates was then undertaken by ICP-MS (iCAP 6000 series) to give ppm values. Nitrogen and carbon contents were analysed using a FlashSmart elemental analyser. Samples were samples combusted at 9500C in oxygen with alumina and copper oxide catalyst and helium gas carrier. N and C were determined using a Multi-separation column (SS, 2m, 6x5mm. Part # 260 07920) at 500C with standards used to calculate percentage. pH measurement was undertaken in both H2O and in CaCl2 (0.125m) solution. pH values of suspensions were given by pH meter model 292 Pye Unicam. Exchangeable cations (CEC) of the top 20 cm within soil profiles were displaced by leaching soil within a solution of potassium chloride. In the leachate, exchangeable calcium and magnesium were determined by atomic absorption spectrometry and exchangeable acidity by nitration against standard sodium hydroxide solution. Determination of CEC and the related base saturation (BS) BS was undertaken on the 10 and 20 cm depth samples (A1 horizon) based on the formulae CEC = Ca + Mg + H and BS = 100 (Ca + Mg)/CEC. https://geohack.toolforge.org/geohack.php?pagename=Cacuaco¶ms=08_46_37_S_13_22_18_E_ https://geohack.toolforge.org/geohack.php?pagename=Bom_Jesus,_%C3%8Dcolo_e_Bengo¶ms=9_10_07_S_13_34_00_E_region:AO-BGO_type:adm1st_source:kolossus-dewiki https://geohack.toolforge.org/geohack.php?pagename=Ramiros¶ms=9.0513_S_13.0192_E_type:city(28708)_region:AO-LUA American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 236 2.4 Statistical analyses Through application of software jamovi version 2.3 debug, the macro- and micro- nutrient total levels data were subjected to descriptive statistical analysis to obtain the mean, maximum, minimum, median, and standard deviation, together with ANCOVA analyses and pairwise comparison using Tukey and Bonferroni tests. Due to limited degrees of freedom in the sampling design, interaction terms could not be fully evaluated; nonetheless, trends in the data by site, hillslope position and soil depth are identified. Statistical test results were assessed at 0.05 significance level. 3. Results and Discussion 3.1 Field survey and soil classification The Soil Atlas of Africa [13] classifies the soils of Huambo Province as haplic and xanthic Ferralsols and soils within Luanda Province as predominantly chromic Luvisols with areas of eutric Cambisols and eutric Fluvisols [6, 7, 14]. Our observation of field and laboratory properties tests and refines these classifications giving a secure soil classification foundation for the assessment of total nutrient element status. At Huambo, soil profiles (Table 1) typically comprise A1, A2 and B horizons with diffuse boundaries separated by Munsell hues that range from 7.5R through 7.5YR to 10YR, values ranging between 2 and 6, and chromas ranging from 1 to 8. Soils typically have reduced hue and chroma values in lower profiles of the catena. Field textures are silt loams and silty clay loams throughout the profiles and across the catenas with well-developed granular micro-aggregations. Clay content is relatively high, and down profile increases together with micromorphological evidence of clay accumulation as coatings and fills indicates clay mobilisation. Organic carbon content in the A1 horizon is low, although can be as high as 5% at 10cm depth. Cation exchange capacities of A1 horizons are low and range from 3.95 to 13.96 cmolc kg-1 clay1, while % base saturation is in general also low although with some high peaks. This reflects the high frequency of weathering resistant quartz and low frequencies (<10%) of more weatherable Ca- and Na- plagioclases and K- feldspars. These observations are consistent with Ferralic horizon diagnostic criteria and a haplic Ferralsol soil classification across the Huambo study areas [6, 7]. There are however localised profile variations reflecting movement of clay through the profile and across the catena resulting in argic horizon attributes [15]. Munsell colour contrasts indicating redox reactions suggest that profiles lower in the catena have a greater soil wetness periodicity than those higher in the catena. American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 237 Table 1: Field and laboratory data sets for soil classification, Huambo Province Soil Profile Horiz on Depth (cm) Munsell colour Field Texture Field Structure Clay % Silt % Sand % Organic carbon % % Weatherable coarse minerals Clay accumulation features Base Saturation (%) CEC cmolc kg-1 clay1 Bailundo upper A1 10 10YR 6/6 silty loam Granular, well developed 17.38 81.13 1.0 1.4 43.88 7.13 20 10YR 6/6 16.30 69.67 14.01 0.8 45.88 4.43 30 10YR6/6 21.82 71.90 6.28 0.6 - - A2 40 10YR 6/8 silty loam Granular, well developed 31.42 68.59 0.01 0.4 (<10%), Quartz dominant Coatings - - 50 10YR 6/8 29.42 70.09 0.49 0.2 (<10%), Quartz dominant Coatings - - B 60 10YR 5/8 silty loam Granular, well developed - - - 0.2 - - 70 10YR 5/8 - - - - - Bailundo middle A1 10 10YR 6/8 silty loam Granular, well developed 20.12 78.74 1.11 2.4 62.27 6.36 20 10YR 6/8 11.11 75.3 13.58 1.7 71.23 5.56 30 10YR 6/8 23.61 75.70 0.65 1.3 - - A2 40 10YR 5/8 silty loam Granular, well developed 24.09 75.11 0.80 1.6 (<10%), Quartz dominant None observed - - 50 10YR 5/8 21.20 76.83 1.98 0.8 (<10%), Quartz dominant None Observed - - B 60 10YR 5/6 silty loam Granular, well developed 29.82 70.17 0.00 1.1 - - 70 10YR 5/6 - - - - - Bailundo lower A1 10 10YR 4/3 silty loam Granular, well developed 12.93 76.59 10.49 0.9 56.44 7.35 20 10YR 4/3 15.89 81.08 3.02 0.8 54.74 7.07 30 10YR 5/4 16.29 80.13 3.56 0.3 - - A2 40 10YR 5/4 silty loam Granular, well developed 18.72 78.58 2.71 0.5 (<10%), Quartz dominant Coatings, Fills - - 50 10YR 4/4 18.19 69.82 11.98 0.2 (<10%), Quartz dominant Coatings, Fills - - B 60 10YR 4/4 silty loam Granular, well developed 18.55 72.92 8.52 0.0 - - 70 10YR 4/4 20.91 73.74 5.35 - - Mungo upper A1 10 7.5R 3/6 silty loam, silty clay loam Granular, well developed 22.56 70.45 7.21 1.5 41.29 4.09 20 7.5R 3/6 25.12 69.89 5.0 1.0 39.29 3.95 30 7.5R 3/6 31.54 68.45 0.00 0.6 - - A2 40 7.5R 3/8 silty loam, silty clay loam Granular, well developed 30.77 69.16 0.4 0.4 (<10%), Quartz dominant None observed - - 50 7.5R 3/8 30.41 68.73 0.86 0.4 (<10%), Quartz dominant None observed - - B 60 7.5R 3/8 silty loam, silty clay loam Granular, well developed - - - - - 70 7.5R 3/8 - - - - - Mungo middle A1 10 7.5R 3/2 silty loam, silty clay loam Granular, well developed 14.42 70.60 14.98 2.8 41.47 9.57 20 7.5R 3/2 18.43 61.88 19.70 1.6 23.09 9.36 30 7.5R 3/2 25.25 64.84 9.91 1.2 - - A2 40 7.5R 3/6 silty loam Granular well developed 31.78 67.71 0.51 0.9 (<10%), Quartz dominant Coatings - - 50 7.5R 3/6 30.69 69.31 0.00 1.1 (<10%), Quartz dominant Coatings - - B 60 7.5R 3/6 silty loam Granular, well developed 21.05 64.46 15.48 0.3 - - 70 7.5R 3/6 - - - - - - Mungo lower A1 10 7.5YR 6/2 silty loam, silty clay loam Granular, well developed - - - 3.2 10.14 8.9 20 7.5YR 6/2 - - - 4.6 42.68 13.96 30 7.5YR 6/2 - - - 7.4 - - A2 40 7.5YR 3/1 silty loam, silty clay loam Granular, well developed - - - 15.0 (<10%), Quartz dominant Coatings - - 50 7.5YR 3/1 - - - 3.7 (<10%), Quartz dominant Coatings - - B 60 7.5YR 2/1 silty loam Granular, well - - - 5.5 - - American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 238 70 7.5YR 2/1 developed - - - - - - 80 7.5YR 2/1 - - - - - - Lepi upper A1 10 7.5 YR 4/3 silt loam Granular, well developed 5.05 71.06 6.01 2.1 18.56 2.95 20 7.5 YR 4/3 19.23 79.35 1.51 1.5 (<10%), Quartz dominant Coatings, Fills 73.18 5.97 30 7.5 YR 4/3 22.90 51.48 43,47 0.6 (<10%), Quartz dominant Coatings, Fills - - Lepi middle A1 10 7.5YR 3/6 silt loam, slightly silt clay loam Granular, well developed 29.74 66.60 0.02 1.1 23.51 5.23 20 7.5YR 3/6 29.74 69.43 0.82 0.8 (<10%), Quartz dominant Coatings 20.96 5.06 30 7.5YR 3/6 33.38 70.34 29.74 0.4 (<10%), Quartz dominant Coatings - - Lepi lower A1 10 7.5R 4/4 silt loam, silt clay loam Granular, well developed 18.91 67.01 0.00 0.4 75.53 3.02 20 7.5R 4/4 22.09 67.72 10.19 1.0 (<10%), Quartz dominant Fills 27.94 7.77 30 7.5R 4/4 32.99 58.41 22.71 - (<10%), Quartz dominant Fills - - Ngongoinga upper A1 10 10YR 5/3 silty loam, silty clay loam Granular, well developed, subangular blocky 17.19 72.48 10.53 5.0 33.07 4.78 20 10YR 5/3 26.92 72.02 1.05 1.1 32.1 7.07 30 10YR 5/3 25.83 70.74 3.45 0.4 - - A2 40 10YR 6/6 silty loam, silty clay loam Granular, well developed, subangular blocky 29.46 70.43 0.09 0.3 (<10%), Quartz dominant Coatings - - 50 10YR 6/6 - - - 0.3 (<10%), Quartz dominant Coatings - - B 60 10YR 6/6 silty loam, silty clay loam Granular, well developed, subangular blocky - - - - - - 70 10YR 6/16 - - - - - - Ngongoinga Middle A1 10 10YR 5/3 silty loam Granular, well developed, subangular blocky 16.12 68.34 15.55 0.9 18.03 5.86 20 10YR 5/3 27.53 71.63 0.82 0.6 23.85 5.25 30 10YR 5/3 26.65 73.09 0.24 0.5 - - A2 40 10YR 6/4 silty loam Granular, well developed, subangular blocky 32.75 67.25 0.00 0.4 (<10%), Quartz dominant Coatings - - 50 10YR 6/4 37.55 62.43 0.00 0.3 (<10%), Quartz dominant Coatings - - B 60 10YR 6/6 silty loam Granular, well developed, subangular blocky - - - 0.2 - - 70 10YR 6/6 - - - - Ngongoinga Lower A1 10 10YR 4/1 silty loam, silty clay loam Granular, well developed, subangular blocky - - - 2.3 28.92 7.88 20 10YR 4/1 - - - 1.4 14.64 7.5 30 10YR 4/1 - - - 1.2 - - A2 40 10YR 6/3 silty loam, silty clay loam Granular, well developed, subangular blocky - - - 0.7 (<10%), Quartz dominant Coatings - - 50 10YR 6/3 - - - 0.7 (<10%), Quartz dominant Coatings - - B 60 10YR 6/3 silty loam, silty clay loam Granular, well developed, subangular blocky - - - 0.5 70 10YR 6/3 - - - - The Luanda Province soil profiles (Table 2) typically comprise A1, A2 and B horizons with diverse Munsell hues that range from 5R to 10YR, but which retain the same hue in each profile, values ranging between 1 and 8 and varying between profile, and chromas ranging from 1 to 8 also varying between profile horizons. Field textures are silt loams and silty clay loams throughout the profiles and across the catenas, with the exception of the coastal Ramiro site where field textures are dominantly sandy loams. Field structures are well-developed granular and sub-angular blocky aggregations. Percentage clay content is relatively high with low variability through the profile although at some locations clay content declines down the profile. Ramiro is again an exception, having low percentage clay contents. Micromorphological evidence of clay accumulation as coatings American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 239 and fills indicates mobilisation of the clay fraction in all profiles. Organic carbon content in A1 horizons are low, ranging from <0.1 to 4.3% and declining through the top 30cm of the profile. Cation exchange capacities of A1 horizons are also low and range from 0 to 16.84 cmolc kg-1 clay1. In contrast, % base saturation is typically high and above 50% but with a range from 0 - 90.97%. The weatherable mineral fraction is >10% and comprises Ca- and Na- plagioclases and K- feldspars; quartz is the dominant weathering resistant mineral. These observations contrast with the Luvisol classification criteria set out in the World Reference Base (WRB) for Soil Resources [6,7] with clay enhanced argillic horizon expected to be lower in the profile. Rather, and although Munsell colour hues are consistent within each profile, the criteria presented in Table 2 are more diagnostic of Cambic horizons and a Cambisol classification across the Luanda study area. Base saturation data (%) are only available for the A1 horizon at 10 and 20cm depths and so it is not possible to be definitive on the eutric nature of the Luanda soils, but the high % base saturation values strongly indicate that these soils may be classified as eutric Cambisols. Table 2: Field and laboratory data sets for soil classification, Luanda Province Soil Profile Horizon Depth (cm) Munsell colour Field Texture Field structure Clay % Silt % Sand % Organic carbon % % Weatherable coarse minerals Clay accumulation features Base Saturation (%) CEC cmolc kg-1 clay1 Bom Jesus upper A1 10 7.5R 3/1 silt loam Granular, well developed 23.59 75.72 0.67 0.8 60.71 4.07 20 7.5R 3/1 23.91 74.24 1.85 0.8 83.08 9.45 30 7.5R 3/1 23.6 75.18 1.22 0.7 - - A2 40 7.5R 3/2 silt loam Granular, well developed 22.25 76.58 1.18 0.6 (>10%) Coatings, Fills - - 50 7.5R 3/2 22.02 75.43 2.53 0.7 (>10%) Coatings, Fills - - B 60 7.5R 3/3 silt loam Granular, well developed 19.11 79.04 0.70 0.7 - - 70 7.5R 3/3 - - - - - - Bom Jesus middle A1 10 7.5R ½ silt clay loam Granular, well developed 30.34 68.36 1.30 1.4 76.51 10.22 20 7.5R ½ 27.08 72.59 0.35 1.0 84.14 10.09 30 7.5R ½ 32.07 67.95 <0.01 0.8 - - A2 40 7.5R 1/3 silt clay loam Granular, well developed 29.47 70.39 0.15 0.8 (>10%) Coatings, Fills - - 50 7.5R 1/3 28.26 71.61 0.28 0.8 (>10%) Coatings, Fills - - B 60 7.5R 3/3 silt clay loam Granular, well developed 25.78 71.61 0.14 0.7 - - 70 7.5R 3/3 - - - 0.6 - - Bom Jesus lower A1 10 7.5R ½ silt clay loam, silt loam Granular, well developed 28.64 70.52 0.83 1.8 85.34 10.91 20 7.5R ½ 29.84 69.75 0.39 1.2 - - 30 7.5R ½ 28.53 70.63 0.85 1.2 - - A2 40 7.5R 1/3 silt clay, slightly silt loam Granular, well developed 23.33 75.36 1.30 1.0 (>10%) Fills - - 50 7.5R 1/3 28.37 70.58 1.06 1.0 (>10%) Fills - - B 60 7.5R 3/3 silt clay, silt loam Granular, well developed 24.44 73.83 1.74 1.0 - - 70 7.5R 3/3 - - - - - - American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 240 Funda upper A1 10 2.5YR 7/3 silt clay loam, silt loam Apedal, crack 22.38 69.37 8.29 2.5 84.43 10.28 20 2.5YR 7/3 22.18 73.69 4.13 1.6 79.77 11.86 30 2.5YR 6/3 21.89 78.02 2.76 1.6 - - A2 40 2.5YR 6/3 silt clay loam, silt loam Apedal, crack 21.58 73.25 5.19 1.9 (>10%) Coatings, Fills - - 50 2.5YR 5/2 19.8 76.39 3.79 5.6 (>10%) Coatings, Fills - - B 60 2.5YR 5/2 silt clay loam, silt loam Apedal, crack 10.91 86.08 2.99 10.1 - - Funda middle A1 10 2.5YR 5/2 silt loam Subangular blocky, well developed 21.39 69.70 8.88 4.3 87.89 13.21 20 2.5YR 5/2 25.65 72.10 2.24 1.6 0 0 30 2.5YR 5/2 23.83 74.08 2.09 1.5 - - A2 40 2.5YR 5/4 silt loam Subangular blocky, well developed 24.98 73.73 1.28 1.5 (>10%) Coatings, Fills - - 50 2.5YR 5/4 24.28 73.25 2.49 1.2 (>10%) Coatings, Fills - - B 60 2.5YR 4/4 silt loam Subangular blocky, well developed 22.31 75.44 2.24 1.2 - - 70 2.5YR 4/4 22.78 75.39 1.82 1.5 - - 80 2.5YR 4/4 24.1 73.13 2.77 - - - Funda lower A1 10 2.5YR 5/2 silt loam Subangular blocky, well developed 17.86 70.99 11.13 1.7 0 0 20 2.5YR 5/2 17.59 73.87 8.51 1.2 0 0 30 2.5YR 5/2 19.02 75.27 5.67 1.2 - - A2 40 2.5YR 5/4 silt loam Subangular blocky, well developed 16.19 73.41 10.40 1.1 (>10%) Fills - - 50 2.5YR 5/4 13.12 73.67 13.31 1.0 (>10%) Fills - - B 60 2.5YR 4/4 silt loam Subangular blocky, well developed 14.84 76.70 8.46 1.1 - - 70 2.5YR 4/4 13.78 77.47 8.75 1.1 - - 80 2.5YR 4/4 13.61 74.47 11.92 - - - Talelo upper A1 10 5R 5/6 silt loam Granular, well developed 16.24 78.75 4.99 0.8 87.13 12.43 20 5R 5/6 20.03 77.85 2.15 0.4 89.23 14.86 30 5R 4/6 24.60 74.33 1.07 0.4 - - A2 40 5R 54/6 silt loam Granular, well developed 20.92 75.15 3.92 0.4 (>10%) Coatings, Fills - - 50 5R 4/4 21.28 75.39 3.22 0.4 (>10%) Coatings, Fills - - B 60 5R 4/4 silt loam Granular, well developed 17.17 76.66 6.18 0.7 - - 70 5R 4/4 - - - - - - Talelo middle A1 10 5R 5/6 silt loam Granular, well developed 16.29 69.71 13.96 0.6 87.54 16.84 20 5R 5/6 17.82 70.64 11.53 0.9 62.00 16.84 30 5R 6/6 17.63 69.67 12.71 0.7 - - A2 40 5R 6/6 silt loam Granular, well developed 21.72 74.90 3.39 0.6 (>10%) Coatings - - 50 5R 5/4 18.46 74.85 6.68 0.4 (>10%) Coating - - B 60 5R 5/4 silt loam Granular, well developed 16.53 67.30 16.14 0.3 - - 70 5R 5/4 16.81 76.65 6.54 0.3 - - 80 5R 5/4 17.49 79.51 2.98 - - - Talelo lower A1 10 7.5YR 7/6 silt loam Granular, well developed 16.73 66.03 16.98 1.0 90.97 8.86 20 7.5YR 7/6 18.00 68.47 13.53 0.8 81.45 8.63 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 241 3.2 Mean and range nutrient element status Summary statistics of twelve elements (N, Ca, K, Mg, Mn, P, S, Zn, Cu, Mo, Fe, Al together with pH) including mean and range values of macro- and micro- elements are given for the Huambo Province haplic Ferralsols and the Luanda Province eutric Cambisols (Table 3); these are also expressed as box plots in Figure 3. The statistics indicate variability in nutrient reserves across the study areas and give a basis for broader sub-Saharan regional comparisons (Table 4). Analytical methods are different in these regional studies and caution is needed when considered in relation to the Angola data sets, but Total X-ray Fluorescence (TXRF) does give comparable results corrected to ICP-MS while low total nutrient reserves are indicated when less that plant available nutrient 30 7.5YR 7/8 22.84 69.73 7.45 0.6 - - A2 40 7.5YR 7/8 silt loam Granular, well developed 21.01 65.03 13.94 0.6 (>10%) Coatings - - 50 7.5YR 4/6 22.50 63.90 13.61 0.3 (>10%) Coatings - - B 60 7.5YR 4/6 silt loam Granular, well developed 24.43 61.74 13.80 0.3 - - 70 7.5YR 4/6 17.11 42.20 40.69 0.4 - - 80 7.5YR 4/6 - - - - - - Ramiro upper A1 10 10YR 8/4 sandy loam, silt loam Granular, well developed 26.59 72.45 0.86 0.2 50.53 3.23 20 10YR 8/4 11.33 37.56 51.11 0.1 33.15 3.59 30 10YR 8/6 20.31 65.56 14.11 0.09 - - A2 40 10YR 8/6 sandy loam, silt loam Granular, well developed 4.35 18.41 77.25 0.07 ND ND - - 50 10YR 8/6 2.42 9.92 87.65 0.04 ND ND - - B 60 10YR 8/6 sandy loam, silt loam Granular, well developed 9.86 26.60 63.53 0.07 - - 70 10YR 8/6 - - - - - - Ramiro middle A1 10 10YR 8/4 sandy loam, silt loam Granular, well developed 3.52 17.36 79.30 0.1 87.89 13.21 20 10YR 8/4 11.87 58.49 29.62 0.2 53.99 3.48 30 10YR 8/6 4.29 18.89 76.83 <0.1 - - A2 40 10YR 8/6 sandy loam, silt loam Granular, well developed 3.57 14.85 81.59 0.06 ND ND - - 50 10YR 8/6 9.49 42.55 47.96 0.05 ND ND - - B 60 10YR 8/6 sandy loam, silt loam Granular, well developed 7.49 35.89 56.59 0.06 - - 70 10YR 8/6 2.39 11.47 86.15 0.04 - - 80 10YR 8/6 3.89 17.55 78.56 0.03 - - Ramiro lower A1 10 10YR 8/4 sandy loam, silt loam Granular, well developed 7.69 45.60 46.71 0.1 46.46 4.48 20 10YR 8/4 7.75 44.00 48.96 0.1 43.38 0 30 10YR 8/6 6.87 33.35 59.77 0.1 - - A2 40 10YR 8/6 sandy loam, silt loam Granular, well developed 7.52 37.60 54.88 0.06 ND ND - - 50 10YR 8/6 3.21 13.31 83.58 0.07 ND ND - - B 60 10YR 8/6 sandy loam, silt loam Granular, well developed 7.69 36.83 55.50 0.06 - - 70 10YR 8/6 8.19 37.42 54.40 0.06 - - 80 10YR 8/6 6.26 26.55 67.19 0.06 - - American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 242 comparators. The values for nitrogen (N; ppm) range between 21-7,842 for Huambo Province and 0.03-16,975 for Luanda Province; mean values range across the four catenas from 449-1,245 at Huambo and 37-4,386 at Luanda. Boyer [16] reports that tropical soils generally have low N levels, but apart from the high mean value evident at Bom Jesus the results reported here are either close to or below those predicted for this region of Africa [9]. Explanation for the low levels may be related to lack of grazing and associated microbial activity [17] and in lower profile horizons that are less weathered there may be changes in nitrogen forms as a result of water movement within the profile [18]. Considering calcium (Ca; ppm), potassium (K; ppm) and magnesium (Mg; ppm) macro- nutrients, the calcium range at Huambo is 25-34,758 with mean ranges from 653-7,674 and at Luanda there is a 0.03-224,403 range with mean ranges from 209-79,236. Potassium values range from 0.02-1,372 with mean ranges of 327-570 at Huambo and ranges of 0.02-23,968 and mean ranges of 970-14,678 at Luanda. Magnesium values range from 0.03-844 with mean ranges from 121-340 at Huambo, and at Luanda there is a 0.02-26,063 range and mean range of 214-9,388. Total calcium, potassium and magnesium levels from the Huambo Ferralsols are at or below the means for these elements across sub-Saharan Africa and which may the result of long-term weathering, pH mediated mobilisation and high iron and aluminium concentrations (Table 3; [19, 4]. Ferralsols are typically nutrient poor with potassium becoming scarcer before magnesium and calcium in managed soils [20, 21]. At Luanda, mean values for these elements are generally high with with two catenas having higher means for calcium, three catenas having higher or close to means for potassium compared to the sub-Saharan Africa means, and with magnesium much higher than modelled spatial predictions of available concentrations in sub- Saharan Africa. These high values are associated with parent material weathering and limited down profile mobilisation because of the hot semi-desert climate [22]. Results for the phosphorus macro-nutrient range from 0.02-537 ppm at Huambo with a mean range across the four catenas of 49-329 ppm. These values are generally low in comparison to the regional data sets, again emphasising the low nutrient status in Huambo Ferralsols. Phosphorus values at Luanda are higher with a range from 0.02-2257 ppm and a mean range from 116-1559 ppm although only two catena locations are consistently higher than the regional data set indicators; such variability may be the result of the type of soil treatment [23]. Regardless of soil classification phosphorus is a relatively little leached element and bound by abundant aluminium and iron oxides in highly weathered soils and as evidenced by high iron and aluminium values in the profiles examined [24]. Furthermore, the very weak leaching of phosphorus for hundreds of thousands of years is one of the major causes of the low level of phosphorus in ferralitic soils [16]. Soil phosphorus scarcity in family farming regions of sub-Saharan Africa is common [25]. In considering micro- element concentrations attributed to weathering processes [26], at Huambo the manganese range is 3.75-540 ppm with mean range across the four catenas of 106-301 ppm; the zinc range is 0.02-36.9 ppm with a mean range from 6.9-22.3; the copper range is 0.02-326 ppm with a mean range from 6.16-43.9 ppm and the molybdenum range is 0.02-0.04 ppm. Observed micro-nutrient values from the Huambo catenas are generally lower than regional total nutrient values, with the exception of copper at the Lepi catena. At Luanda, American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 243 the manganese range is 0.04-15,632 ppm with a mean range across the four catenas of 632-9573 ppm; the zinc range is 0.04-49,260 ppm with a mean range from 25.9-12,484; the copper range is 0.04-48,744 ppm with a mean range of 5.66-16,577 and the molybdenum range is 0.03-25,921 with a mean range of 213-9336 ppm. Luanda soil micro-nutrient concentrations are greater than regional and global averages (Table 4) and confirm high to low micro-nutrient status contrasts between the Luanda Cambisols and Huambo Ferralsols respectively. Table 3: Summary statistics of total soil nutrient elements (ppm/kgmg-1), haplic Ferralsols, Huambo Province* and eutric Cambisols, Luanda Province** Angola (N, Ca, K, Mg, Fe, Mn, P, S, Zn, Cu, Mo, Al and pH. N Ca Mean Median St. Dev Max Min Mean Median St. Dev Max Min Bailundo* 463 365 407 1493 37 653 615 332 1215 104 Lepi* 1245 388 1943 7842 21 1112 591 1337 5303 155 Mungo* 1008 601 1022 3451 113 7674 571 11529 34758 25.5 Ngongoinga* 449 281 388 1327 78 925 831 537 1822 305 Talelo** 501 451 340 1314 0.03 2651 2412 1302 5368 0.02 Ramiro** 37 35.1 23.9 92.9 0.03 209 141 231 835 0.03 Funda** 640 519 367 1777 204 79236 75555 48482 224403 4.44 Bom Jesus** 4386 764 5723 16975 495 19865 20810 11023 36045 6888 K Mg Mean Median St. Dev Max Min Mean Median St. Dev Max Min Bailundo* 327 317 161 756 104 127 129 38.1 205 61.4 Lepi* 570 525 310 1372 125 340 300 217 844 102 Mungo* 450 407 230 909 151 121 96.7 90.4 413 0.03 Ngongoinga* 386 447 191 562 0.02 170 173 58.9 242 46.6 Talelo** 14678 14482 6257 23968 0.02 9388 11182 4449 14967 0.02 Ramiro** 970 977 277 1930 556 214 200 68.2 424 94.7 Funda** 9223 9401 3159 14931 0.82 7231 7937 4998 26063 0.03 Bom Jesus** 13275 11546 3514 19914 9858 7062 6421 1566 9983 5525 Fe Mn Mean Median St. Dev Max Min Mean Median St. Dev Max Min Bailundo* 1704 922 2005 8882 635 106 99.1 61.1 205 10 Lepi* 45788 59274 24126 69029 12287 195 198 122 408 22.8 Mungo* 4909 978 6810 21520 407 115 92.2 126 502 6.53 Ngongoing* 8569 3298 15906 50583 478 301 231 178 540 2.75 Talelo** 48709 54500 19084 81212 0.04 9573 9445 4081 15632 0.04 Ramiro** 2320 2258 664 4405 1302 632 637 180 1259 363 Funda** 32337 35190 9110 39682 3.63 6016 6131 2060 9738 0.53 Bom Jesus** 24052 32925 19432 47186 35.7 9533 9933 2306 12988 6429 P S Mean Median St. Dev Max Min Mean Median St. Dev Max Min Bailundo* 49 22.2 56.6 189 0.07 33.6 25.3 25.7 96.6 6.46 Lepi* 329 313 98.5 537 159 156 80.6 185 736 24.9 Mungo* 64 31.4 65.5 155 0.03 42.7 37.8 32.9 101 3.52 Ngongoinga* 84 6.01 121 270 0.03 42 37.3 24.6 82.7 11 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 244 Talelo** 217 179 214 1112 0.02 167 190 74.5 310 0.02 Ramiro** 116 92.4 109 502 0.03 62 58.4 23.9 132 25.2 Funda** 521 489 251 1366 0.05 1061 1100 299 1375 0.11 Bom Jesus** 1559 1741 544 2257 827 365 281 130 698 268 Zn Cu Mean Median St. Dev Max Min Mean Median St. Dev Max Min Bailundo* 8.2 5.63 7.31 23.2 0.1 3.23 1.66 5.28 21.9 0.03 Lepi* 22.3 21.4 8.21 36.9 10 43.9 27.4 73 326 9.51 Mungo* 3.74 3.03 2.9 8.72 0.02 4.52 4.21 3.83 12.2 0.02 Ngongoinga* 6.9 4.87 6.13 15.8 0.03 6.17 6.37 4.56 16.2 0.52 Talelo** 179 153 82.3 353 0.04 37.6 40 17.5 64.6 0.04 Ramiro** 25.9 16.3 39.1 191 1.89 5.66 3.37 7.24 31.1 0.41 Funda** 12484 2264 17228 49260 0.1 37.1 38.6 11.2 50.2 0.04 Bom Jesus** 7023 1663 8925 27134 687 16577 41.2 22449 48744 25.6 Mo Al Mean Median St. Dev Max Min Mean Median St. Dev Max Min Bailundo* 0.0317 0.03 0.00786 0.04 0.02 8047 7378 4048 20366 4188 Lepi* 0.0265 0.03 0.00493 0.03 0.02 0.0282 0.02 0.0101 0.04 0.02 Mungo* 0.0218 0.02 0.00393 0.03 0.02 21384 17646 13322 51993 6691 Ngongoinga* 0.03 0.03 0 0.03 0.03 13885 9546 11293 43108 7706 Talelo** 9336 11121 4425 14885 0.04 76407 81495 22893 106278 0.03 Ramiro** 213 199 67.9 422 94.2 7268 8356 3086 12182 0.02 Funda** 7191 7894 4971 25921 0.03 94743 102425 30777 134294 8.64 Bom Jesus** 7871 7945 1664 9928 5495 108860 110106 26982 153770 59901 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 245 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 246 Figure 3: Box plots showing variations in soil nutrient element concentrations in four catenas in Huambo province (H1 - H4; haplic Ferralsols) and Luanda Province (L1 - L4; eutric Cambisols). y axis is nutrient concentration in ppm and pH value Table 4: Mean values of total soil nutrient elements (ppm / mgkg-1) in haplic Ferralsols, Huambo* Province Angola and eutric Cambisols, Luanda** Province Angola with published element contents in African and global soils Element Bailundo* Ngongoinga* Mungo* Lepi* Bom Jesus** Funda** Talelo** Ramiro** A B C D Ppm N 463 449 1008 1245 4386 640 501 37 - - - 981 Ca 653 925 7674 1112 19865 79236 2651 209 13300 - 9780 330-1820 K 327 386 450 570 13275 9223 14678 970 26600 - 10893 62-180 Mg 127 170 121 340 7062 7231 9388 214 - - - 60-170 P 49 84 64 329 1559 521 217 116 2000 - 143 480 S 33.6 42 42.7 156 365 1061 167 62 - - - - Fe 1704 8569 4909 45788 24052 32337 48709 2320 67300 - 27954 -200 Mn 106 301 115 195 9533 6016 9573 632 1600 437 466 117 Zn 8.2 6.9 3.74 22.3 7023 12484 179 26 120 64 29 5 Cu 3.23 6.17 4.52 43.9 16577 37.1 37.6 6 60 20-30 17 3 Mo 0.0317 0.03 0.0218 0.0265 7871 7191 9336 213 - 0.1-> 7 - - Al 8047 13885 21384 0.0282 108860 94743 76407 7268 91500 - 33927 900-998 A: [2] total element concentration by TXRF analyses corrected to ICP-MS Na2O2, sub-Saharan Africa; B: [8] soil trace elements, worldwide mean contents; C: [4] further total element concentration by TXRF analyses, sub- Saharan Africa (0-20cm and 20-50cm profile depths; D: [9] Mehlich 3 plant available nutrients, except total P - 0-20cm, 20-50cm and up to 125cm profile depths - and machine learning modelling to give soil nutrient maps of Sub-Saharan Africa. 3.3 Nutrients and landscape relationships. Statistical relationships between soil nutrients and landscape factors including site, profile depth and position in the catena (profile) give explanation of soil nutrient concentrations and highlight variabilities in nutrient concentrations across the study areas. Overall - for both study areas and all catenas - ANCOVA analyses indicates that (i) ‘Site’ was highly statistically significant (P < 0.001); (ii) ‘Hillslope Position’ was significant (P < 0.05); and (iii) ‘Depth’ was not statistically significant (P > 0.05). Pairwise comparison using Tukey and American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 247 Bonferroni tests, which show similar test probabilities, disaggregates the ‘site’ statistics for the nutrients considered (Table 5). Of the sixteen different Huambo and Luanda sites pairwise combinations, there are significant differences in twelve pairs for both Mn and Al, eleven for both K and Mo, ten for Mg, nine for Fe, eight for S, six for P, four for N, Ca and Cu and three for Zn. These observations highlight soil nutrient contrasts between Huambo haplic Ferralsols and Luanda eutric Cambisols (Table 5) and can be attributed to underlying parent materials [27,28]. Considering the detail of these analyses for Huambo Province, the ANCOVA analyses indicate that, with the exception of Nitrogen, all soil nutrient elements show statistically significant differences in relation to site (Table 6). However, pairwise comparisons with Tukey and Bonferroni tests indicates that only iron of the nutrients considered shows a statistically significant between two of the Huambo sites; all other nutrient elements have no significant difference between sites (Table 5). Furthermore, no significant correlations were observed between nitrogen, phosphorus, and potassium elements in Huambo province, with the exception of Mungo (Table 7). These statistical findings suggest variance in soil nutrients is evident across all sites but less so between sites and points to a spatial scaling of mineral diversity in the underlying Precambrian rock parent material, overlain in places by Quaternary deposits [29, 30]. Nitrogen levels can be attributed to biological input to soils independent from geology [31]. Contrast in nutrient concentrations resulting from slope position are less pronounced in their probability values but still significant, testifying to the action of slope processes in influencing distribution of elements [32]. Elements showing contrast across catenas are calcium, potassium, phosphorus, sulphur, iron and molybdenum (Table 6) and assessment of box-plot graphs indicates soil nutrient values have a tendency to be slightly higher in the mid-slope profiles at Huambo (Figure 4). Within profile variances with depth are much less pronounced in their differences with only potassium, manganese and zinc showing contrasts with depth (Table 6), indicating limited and similar pedogenesis influences on both macro- and micro- nutrients within the profiles. In Luanda province all soil nutrient elements without exception show statistical differences in relation to site (Table 8). Further detail of site contrasts is given by the pairwise Tukey and Bonferroni tests showing the same significances and demonstrate statistically significant variances in all nutrients between the four catena sites. Out of the six possible combinations, there are two significant pairs for Zn, three significant pairs for N, Ca, Mg, Cu, Mo, Al, four significant pairs for K, P, Fe, S and five significant pairs for Mn (Table 5). Similarly, and as at Huambo there were no significant correlations between nitrogen, phosphorus, and potassium nutrient elements with the exception of the Talelo catena where there is significant positive and strong correlation between nitrogen and potassium (Table 9). The observations from Luanda point to mineral diversity in the geological recent Neogene and Palaeogene parent material together with the association of nitrogen and biological process [29, 31]. As at Huambo, contrasts in soil nutrient concentrations across the catena’s are not strongly evident but are still statistically significant. Nitrogen, magnesium, sulphur, iron, zinc, copper, and molybdenum all show statistical contrasts at this spatial scale as does pH (Figure 4; Table 8). Assessment of box-plot graphs indicates that the upper-slope profiles hold the greatest nutrient reserves within the Luanda catenas. Statistically significant in-profile American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 248 variances with depth are limited to only Nitrogen, reflecting biological contributions to the profile and the limited influence of pedogenesis in redistributing soil nutrient [33]. Our findings from both Huambo - haplic Ferralsols and Luanda -eutric Cambisols suggest that underlying parent material has the biggest influence on element concentrations, further modified by slope processes. While soil profile weathering undoubtedly contributes to the soil nutrient elements within profiles, pedogenesis has been consistent across profiles and has had a minimal contribution to the variances and nutrient contrasts observed. Table 5: Post-hoc pairwise comparisons of nutrient elements in soil catena sites, Huambo and Luanda Provinces, Angola, with p-values for Tukey and Bonferroni tests. Comparisons based on estimated marginal means American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 249 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 250 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 251 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 252 Table 6: ANCOVA Statistical comparisons between elements within sites, profiles, and depths, Huambo Province haplic Ferralsols - Table 7: Statistical relationships between nitrogen, phosphorus and potassium, Huambo Province haplic Ferralsols - Spearmans Rank analyses American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 253 Table 8: ANCOVA Statistical comparisons between elements within sites, profiles, and depths, Luanda Province - eutric Cambisols Table 9: Statistical relationships between nitrogen, phosphorus and potassium, Luanda Province eutric Cambisols, Spearmans Rank Analyses American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 254 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 255 Figure 4: Box-plots of soil nutrient concentration along hillslope profiles, Huambo Province (haplic Ferralsols) and Luanda Province (eutric Cambisols) 4. Conclusions Up-to-date knowledge and understanding of the nutrient status of soils is of vital importance in giving a baseline for land capabilities and how best to intervene with effective fertiliser and land management practices that enhance productive sustainable agriculture [34]. While we acknowledge the spatial limitations in our study - our sample transects and profiles are a tiny fraction of the soil landscape we are considering, there has been no attempt to model or extrapolate to a wider area, and no consideration of climate-sequences - nevertheless our findings do highlight a wide range of nutrient element concentrations within and between haplic Ferralsols - a common soil class in Angola - and eutric Cambisols, managed for subsistence agriculture in Huambo and Luanda Provinces, Angola. This range included absence, deficiency and even excesses in certain localities. Comparatively, nutrient status in the two provinces is different; in general, the macro and micronutrients concentrations are higher in Luanda Province (Bom Jesus, Funda, Talelo and Ramiro) than in Huambo Province American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 256 (Bailundo, NGongoinga, Mungo and Lepi). Locally, soil nutrient concentrations vary with slope location. These variances can best be explained by differences in underlying parent material influenced by long-term mineral weathering and slope processes including human induced erosion of soils [35]. Land management practices in modifying organic input to soils may explain variances in nitrogen concentrations. Profile pedogenic factors has a more limited influence than parent material and slope on nutrient element variances. Our findings correspond with similar studies from elsewhere in sub-Saharan Africa [34, 36, 37, 38, 39) although at odds with some studies of the nutritional status of African soils that have high rates of nutrient depletions resulting from high population density, continuous cultivation and rugged and mountainous terrain [40, 41]. As yet, such findings are slow to find their way into policy making for subsistence agriculture and the related agricultural extension advisory services. At the very least, our findings open the question of which localities are more suitable for agriculture, or more directly, which soils will be best able to produce food in a sustainable way, enhance soil health maintain and improve human health, be self-sufficient and regenerative, and produce sufficient food to a growing population [30, 42, 43]. Our analyses can also provide a foundation for determining where subsistence agriculture with local historical and ecological knowledge can transform infertile, carbon poor, tropical soils into durable fertile rich and productive soils that can support agriculture ecologically and in a socially sustainable way [44, 45, 46, 47, 48]. In recognising the diversity of soil nutrient reserves, even within distinct soil classes, and to support and encourage subsistence farming, we highlight the need for detailed local analyses so that supportive intervention can be designed and planned for maximum effectiveness. This may involve identifying appropriate levels of introduced macro- and micro- nutrients as well as offering guidance on crops suitable for the nutrient reserves of a locality, and on crop rotations that are effective in conserving nutrients. We also recognise that, at least in Huambo Province, slope processes are influencing the distribution of nutrients and that steps are needed to reduce these erosive effects. Supporting subsistence agriculture in a changing environment and on which so many of sub-Saharan Africa’s populations relies is a major challenge for agricultural policy makers and planners. Understanding soil nutrient levels and distributions is an important part of this vital endeavor. Acknowledgments Lídia Teixeira is grateful to her family for all their support and for encouraging her to continue to believe. We gratefully acknowledge the Ministry of Higher Education, Science and Technology of Angola for supporting this research together with the financial support of INAGBE. We are grateful for the invaluable laboratory technical support of Ian Washbourne, University of Stirling. References [1] D.B. Smith, L.G. Woodruff, R.M. O'Leary, W.F. Cannon, R.G. Garett, J.E. Kilburn and M.B. Goldhaber. "Pilot studies for the North American soil geochemical landscapes project - site selection, sampling, protocols, analytical methods and quality control protocols" Applied Geochemistry, vol. 24, pp. 1357-1368, 2009. American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 257 [2] E.K. Towett, K.D. Shepherd and G. Cadisch. "Quantification of total element concentrations in soils using total X-ray fluorescence spectroscopy (TXRF)". Science of the Total Environment, vol. 463, pp. 374-388, 2013. [3] J. Kihara, P. Bolo, M. Kinyua, J. Rurinda and K. Piikki. "Micronutrient deficiencies in African soils and the human nutritional nexus: opportunities with staple crops." Environmental Geochemistry and Health, vol. 42, pp. 3015-3033, 2020. [4] E.K. Towett, K.D. Shepherd, J.E. Tondoh, L.A. Winowiecki, T. Lulseged, M. Nyambura, A. Sila, T.G. Vågen and G. Cadisch. "Total elemental composition of soils in Sub-Saharan Africa and relationship with soil forming factors." Geoderma Regional, vol. 5, pp. 157-168, 2015. [5] L.P. de S. Teixeira. “Geochemical, textural and micromorphological properties of Angolan agroecosystem soils in relation to region, landscape position and land management”. PhD Thesis, University of Stirling, Scotland, UK, 2022. [6] IUSS Working Group WRB. World Reference Base for Soil Resources 2014. International soil classification system for naming soils and creating legends for soil maps. update 2015. Rome: World Soil Resources Reports No. 106, FAO, 2015, pp. 192. [7] IUSS Working Group WRB. World Reference Base for Soil Resources. International soil classification system for naming soils and creating legends for soil maps. 4th edition. Vienna: International Union of Soil Sciences (IUSS), 2022, pp. 236. [8] A. Kabata-Pendias and A.B. Mukherjee. “Soils” in Trace elements from soil to human. Berlin, New York: Springer Science & Business Media, 2007, pp. 9-38. [9] T. Hengl, J.G. Leenaars, K.D. Shepherd, M.G. Walsh, G.B. Heuvelink, T. Mamo, H. Tilahun, E. Berkhout, M. Cooper, E. Fegraus, E. and I. Wheeler, I. "Soil nutrient maps of Sub-Saharan Africa: assessment of soil nutrient content at 250 m spatial resolution using machine learning." Nutrient Cycling in Agroecosystems vol. 109, pp. 77-102, 2017. [10] P.M. Kiala and I.A. Simpson. “Evidencing Local Climate Change and Its Implications for Subsistence Agriculture Planning in Huambo Province Angola, Southern Africa.” American Academic Scientific Research Journal for Engineering, Technology, and Sciences, vol. 92, pp. 135–152, 2023. [11] S. Bullock, N. Fedoroff, A. Jongerius, G. Stoops and T. Turisna, T. Handbook for soil thin section description. Wolverhampton: Waine Research Publications, 1985, pp. 152. [12] G. Stoops. Guidelines for analysis and description of soil and regolith thin sections. 2nd Edition. Madison, Wisconsin: Soil Science Society of America, 2021, pp. 240. American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 258 [13] A. Jones, H. Breuning-Madsen, M. Brossard, A. Dampha, J. Deckers, O. Dewitte, T. Gallali, S. Hallett, R. Jones, M. Kilasara, P. Le Roux, E. Micheli, L. Montanarella, O. Spaargaren, L. Thiombiano, E. Van Ranst, M. Yemefack, R. Zougmoré (eds.). Soil Atlas of Africa. Luxembourg: European Commission, Publications Office of the European Union, 2013, pp. 176. [14] B.J. Huntley. "Soil, water and nutrients". In: Ecology of Angola: Terrestrial Biomes and Ecoregions. (pp. 127-147). Cham: Springer International Publishing, 2023, pp. 127-147. [15] M.B. Sharu, M. Yakubu, S.S. Noma and A.I. Tsafe. "Characterization and classification of soils on an agricultural landscape in Dingyadi District, Sokoto State, Nigeria". Nigerian Journal of Basic and Applied Sciences, vol. 21, pp. 137-147, 2013. [16] J. Boyer. Propriedades dos solos e fertilidade. Salvador: Universidade Federal da Bahia, 1971, pp. 196. [17] J.N. Aranibar, I.C. Anderson, H.E. Epstein, C.J.W. Feral, R.J. Swap, J. Ramontsho, J. and S.A. Macko. 2008. “Nitrogen isotope composition of soils, C3 and C4 plants along land use gradients in southern Africa”. Journal of Arid Environments, vol. 72, pp. 326-337, 2008. [18] P. Gupta, and I. Rorison. "Seasonal differences in the availability of nutrients down a podzolic profile." Journal of Ecology, vol. 63, pp. 521-534, 1975. [19] S.R. Cavichiolo. “Perdas de solo e nutrientes por erosão hídrica em diferentes métodos de preparo do solo em plantio de Pinus taeda” Doctoral programme Thesis, Universidade Federal do Paraná, Curitiba, 2005. [20] P. Musinguzi, J.S. Tenywa, P. Ebanyat, T.A. Basamba, M.M. Tenywa, D.N. Mubiru and Y.L. Zinn. "Soil organic fractions in cultivated and uncultivated Ferralsols in Uganda". Geoderma Regional, vol. 4, pp.108-113, 2015. [21] A.E. Hartemink and A.J. Van Kekem. "Nutrient depletion in Ferralsols under hybrid sisal cultivation in Tanzania". Soil Use and Management, vol. 10, pp. 103-107, 1994. [22] F.A. de.Oliveira, Q.A.D.C. Carmello and H.A.A. Mascarenhas. "Disponibilidade de potássio e suas relações com cálcio e magnésio em soja cultivada em casa-de-vegetação". Scientia Agricola vol. 58, pp. 329-335, 2001. [23] S.T. Spera, H.P.D. Santos, R.S. Fontaneli and G.O. Tomm. “Effect of crop-livestock under no-tillage on some soil physical attributes after ten years”. Bragantia, vol. 69, pp. 695-704, 2010. [24] A. Melese, H. Gebrekidan, M. Yli-Halla and B. Yitaferu. "Phosphorus status, inorganic phosphorus forms, and other physicochemical properties of acid soils of Farta district, northwestern highlands of American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 259 Ethiopia". Applied and Environmental Soil Science, vol. 2015, pp. 1-11, 2015. [25] G. Nziguheba, S. Zingore, J. Kihara, R. Merckx, S. Njoroge, A. Otinga, E. Vandamme and B. Vanlauwe "Phosphorus in smallholder farming systems of sub-Saharan Africa: implications for agricultural intensification". Nutrient Cycling in Agroecosystems, vol. 104, pp.321-340, 2016. [26] P.R.S. Vendrame, O.R. Brito, C. Quantin and T. Becquer. "Availability of copper, iron, manganese and zinc in soils under pastures in the Brazilian Cerrado". Pesquisa Agropecuária Brasileira, vol. 42, pp. 859-864, 2007. [27] R.J.A.B. da Silva, Y.J.A.B. da Silva, P. van Straaten, C.W.A. do Nascimento, C.M. Biondi, C.M. Y.J.A.B da Silva, and J.C. de Araújo Filho. "Influence of parent material on soil chemical characteristics in a semi-arid tropical region of Northeast Brazil". Environmental Monitoring and Assessment, vol. 194, 331, 2022. [28] D. Dortzbach, M.G. Pereira, L.H.C.D. Anjos, A. Fontana and E.D.C. Silva. "Genesis and classification of soils from subtropical mountain regions of southern Brazil". Revista Brasileira de Ciência do solo, vol. 40, e0150503, 2016. [29] B.J. Huntley. “Landscapes: Geology, Geomorphology and Hydrology”. In: Ecology of Angola. Springer, 2023. http://doi.org/10.1007/978-3-031-18923-4_4. [30] J.G. Cobo, G. Dercon, G. and Cadisch, G. "Nutrient balances in African land use systems across different spatial scales: A review of approaches, challenges and progress." Agriculture, Ecosystems & Environment, vol. 136, pp. 1-15, 2010. [31] L. Augusto, D. Achat, M. Jonard, D.F. Vidal, B. Ringeval. "Soil parent material - a major driver of plant nutrient limitations in terrestrial ecosystems". Global Change Biology vol. 23, pp. 3808-3824, 2017. [32] K.D. Chadwick and G.P. Asner. "Tropical soil nutrient distributions determined by biotic and hillslope processes". Biogeochemistry, vol. 127, pp. 273-289, 2016. [33] M.M. Stone, J. Kan and A.F. Plante. 'Parent material and vegetation influence bacterial community structure and nitrogen functional genes along deep tropical soil profiles at the Luquillo Critical Zone Observatory". Soil Biology and Biochemistry, vol. 80, pp.273-282, 2015. [34] B.M. Butler, J. Palarea-Albaladejo, K.D. Shepherd, K.M. Nyambura, E.K. Towett, A.M. Sila and S. Hillier. "Mineral–nutrient relationships in African soils assessed using cluster analysis of X-ray powder diffraction patterns and compositional methods.". Geoderma, vol. 375 114474, 2020. [35] O. Olatunji, O. Ogunkunle, O. and F.O. Tabi. "Influence of parent material and topography on some http://doi.org/10.1007/978-3-031-18923-4_4 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 260 soil properties in southwestern Nigeria". Nigerian Journal of Soil and Environmental Research, vol. 7, pp. 1-6, 2007. [36] C. Aponte, T. Marañón, and L.V. García. "Microbial C, N and P in soils of Mediterranean oak forests: influence of season, canopy cover and soil depth". Biogeochemistry, vol. 101, pp. 77-92, 2010. [37] S. Uroz, C. Calvaruso, M.P. Turpault, A. Sarniguet, W. de Boer, J.H.J. Leveau and P. Frey-Klett. "Efficient mineral weathering is a distinctive functional trait of the bacterial genus Collimonas". Soil Biology and Biochemistry, vol. 41, pp. 2178-2186, 2009. [38] R. Finlay, H. Wallander, M. Smits, S. Holmstrom, P. Van Hees, B. Lian and A. Rosling, A. "The role of fungi in biogenic weathering in boreal forest soils". Fungal Biology Reviews, vol. 23, pp. 101-106, 2009. [39] M. Lego, D. Singh and S. Tsanglao. “Effect of differnt levels of NPK on growth, yeild and economic of Capsicum (Capsicum annuum L.) cv, Asha under shade nethouse cultivation”. International Journal of Agricultural Science and Research (IJASR), vol. 6, pp. 5-8, 2016. [40] M. Bekunda, E. Nkonya, D. Mugendi and J.J. Msaky. "Soil fertility status, management, and research in East Africa". East African Journal of Rural Development, vol. 20, pp. 94-112, 2002. [41] H. Hailu, T. Mamo, R. Keskinen, E. Karltun, H. Gebrekidan and T. Bekele. "Soil fertility status and wheat nutrient content in Vertisol cropping systems of central highlands of Ethiopia". Agriculture & Food Security, vol 4, pp.1-10, 2015. [42] T. Higa and J. F. Parr. Beneficial and effective microorganisms for a sustainable agriculture and environment. Atami: International Nature Farming Research Center, 1994. https://api.semanticscholar.org/CorpusID:16595978}. [43] L.K. Abbott and D.A. Manning. “Soil health and related ecosystem services in organic agriculture’. Sustainable Agriculture Research, vol. 4 (526-2016-37946), 2015. [44] D. Solomon, J. Lehmann, J.A. Fraser, M. Leach, K.Amanor, V. Frausin, S.M.Kristiansen, D. Millimouno and J. Fairhead. “Indigenous African soil enrichment as a climate-smart sustainable agricultural alternative’. Frontiers in Ecology and the Environment, vol. 14, pp. 71-76, 2016. [45] R. Chikowo, S. Zingore, S. Snapp and A. Johnston. “Farm typologies, soil fertility variability and nutrient management in smallholder farming in sub-Saharan Africa”. Nutrient cycling in agroecosystems, vol. 100, pp. 1-18, 2014. [46] P. Chase and O. Singh. “Soil nutrient and fertility in three traditional land use systems of Khonoma, Nagaland, India”. Resources and Environment, vol. 4, pp. 181-189, 2014. https://api.semanticscholar.org/CorpusID:16595978 American Academic Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) - Volume 97, No 1, pp 232-261 261 [47] P. Tittonell, A. Muriuki, C.J. Klapwijk, K.D. Shepherd, R.Coe and B. Vanlauwe. “Soil heterogeneity and soil fertility gradients in smallholder farms of the East African highlands”. Soil Science Society of America Journal, vol. 77, pp. 525-538, 2013. [48] S. Zingore, H.K. Murira, R.J. Delve and K.E. Giller. “Influence of nutrient management strategies on variability of soil fertility, crop yields and nutrient balances on smallholder farms in Zimbabwe”. Agriculture, Ecosystems and Environment, vol. 119, pp. 112-126, 2017.