Assessment of Agricultural Biomass Residues for Anaerobic Digestion in Rural Vakinankaratra Region of Madagascar

This study estimates agricultural residue biomass available for biogas generation in smallholder farming systems in the rural Vakinankaratra region of Madagascar, during 2017–2018. Estimations of biomass were done using a combination of agricultural household surveys, literature models, and publicly available data. Manure from four types of farm animals and 17 residue types from ten crops were assessed. In the studied period, gross biomass produced from animal manure and crop residue were 19.4 ± 7.41 and 7.3 ± 1.08 tonnes fresh weight per year per agricultural household, respectively, of which up to 54% and 83% are estimated as recoverable for the production of bioenergy in the studied area, respectively. Estimations indicate that available animal manure and crop residue have the potential to generate 291 ± 92 and 745 ± 122 Nm3 of methane per year per agricultural household respectively, equivalent to 10.5 ± 3.34 and 26.8 ± 4.28 GJ of heat energy from manure and residues, respectively. Theoretically, the average estimated energy potential can result in the complete substitution of domestic fuels in agricultural households. Approximately 0.12 tonnes of nitrogen per household per year can be recovered from the estimated digestate (using all residue types) after energy recovery, which can be employed for crop fertilization. The recovered nitrogen corresponds to 0.26 tonnes urea fertilizer per household per year. The investigation shows that anaerobic digestion based on crop residue and manure has the potential to meet a significant portion of energy needs of smallholder farmers in the Vakinankaratra region of Madagascar and can make an important contribution to providing fertilizer for on-farm use.


Introduction
Biomass plays a significant role globally as a renewable energy source, providing approximately 10% of the world's primary energy supply [1]. To meet increasing demands on modern energy access, biomass must be utilized more efficiently by, for example, biogas production as opposed to combustion of raw biomass [2]. The lack of energy access is one of the most serious challenges in Madagascar [3]. Cooking by electricity and gas (including biogas) remains a luxury energy source for most of the population in the rural area of the country [4]. Electricity access remains low at about 15% of the population, and only 4% in rural areas have access to electricity as of 2015 [5]. Around 80% of the population work in the agricultural sector [3], and the majority of households rely on traditional sources of energy for cooking and heating, and kerosene and candles for lighting. Firewood and charcoal have been used by the majority of the population (95% of households) as their basic energy source leading to increasing concerns regarding local deforestation [4]. Poor indoor air quality is associated with premature deaths and contribute to a broad range of child and adult diseases [6], which in turn negatively impacts the production capacity of agricultural households.
Biomass can replace traditional fuels and reduce energy poverty and greenhouse gas (GHG) emissions, and contribute to rural development [7], if properly utilized with appropriate technology. Agricultural residues from crops (carbonrich) and animal production (nitrogen-rich) are good sources of bioenergy and can contribute significantly to bioenergy generation, particularly through anaerobic digestion (AD) [8]. Biogas generation through AD is one of the most promising technologies for decentralized rural energy production as it not only generates clean energy (biogas) but can also generate organic fertilizer (digestate) for farming applications [9]. Compared to direct burning and composting, AD offers both clean fuel and organic fertilizer, rather than simply one or the other. However, its implementation and usage are still in the early stages in most developing African countries, especially in smallholder farming systems where it was anticipated to have an impact. This is partly linked to a poor understanding of the biomass resources potential and/or inefficient utilization thereof for other purposes in smallholder farming systems. Data on agricultural residue yields remain limited, while data on crop yields are readily available, as the main objective of agricultural production was always to maximize yields, whereas the total biomass yield was not considered important [10]. Furthermore, there is poor understanding to what degree other uses would possibly compete for the available biomass, for example, the demand for use as animal feed, and as use for recycling nutrients via composting or direct combination as fertilizer [11].
Anaerobic digestion technology implementation not only produces biogas but also produces a stabilized digested slurry (digestate) that can serve as a source of plant nutrients [12]. The digested matter can be utilized as a biofertilizer to improve soil fertility and biological quality, and thereby improve crop productivity or to grow fodder for animal feed [12][13][14]. Anaerobic digestion transforms nitrogen into an immediately available form, offering a quick fertilizer response that can be applied when crops display signs of deficiency. Small-scale farmers represent most of the rural population in developing African countries, and the use of mineral fertilizers is very low; consequently, there is an opportunity for high-quality biofertilizers to help farmers to improve productivity. Its implementation also provides a good opportunity for mitigation of GHG and reducing global warming via (i) substituting fuelwood for cooking, (ii) substituting kerosene for lighting and cooking, (iii) substituting mineral fertilizers, and (iv) reducing deforestation [15].
So far, the potential availability of agricultural residues for bioenergy in Madagascan smallholder farming systems has not been reported. Thus, this study aims to determine whether the agricultural residue biomass resources in the farming systems in the rural Vakinankaratra highlands of Madagascar is adequate to make AD viable and sustainable. The study objectives are (a) to estimate the potential availability of gross agricultural residue biomass feedstock resources through surveys, (b) to quantify the portion of gross agricultural residue biomass resources that is recoverable and available for the production of bioenergy, (c) to estimate the available biomethane energy potential of the available residue biomass resources, and (d) to estimate the digestate potential of the residue biomass available.

Sampling and Data Collection
A survey of a sample of agricultural households in the Highlands area of the Vakinankaratra region was carried out by The National Center for Applied Research on Rural Development (FOFIFA) and The French Agricultural Research Centre for International Development (CIRAD) as part of the Project "Ecological intensification pathways for the future of crop-livestock integration in African agriculture" (EcoAfrica). The survey was carried out with two teams of specifically trained investigators/pollsters from a sample of 405 agricultural households drawn at random from 15 fokontany (smallest administrative division) belonging to five municipalities chosen to represent the diversity of production systems of this agroecological zone (see Fig. 1 and Table 1). The questionnaires made it possible to identify the productive resources, the allocation of these resources according to agricultural activities, and practices and performances obtained for the entire agricultural year 2017/2018 (October 2017-September 2018). The farm surveys were conducted with paper questionnaires, and the information collected was entered into a database using Microsoft Access. The verification and the auditing were done using Microsoft Access, and the data were exported to Statistical Packages for Social Sciences (SPSS) to carry out the statistical analysis.

Availability of Agricultural Residue Biomass Resources for AD
Agricultural residue biomass availability analysis comprises the estimation of biomass potentially available from either crop residue or animal manure categories. From agricultural production in smallholder farming systems, the energy potential of the available residue resources was estimated according to the steps illustrated in Fig. 2 [17].

Crop Residue Biomass Availability Estimation
Crop residues are generated from agricultural activities as by-products of crop production systems, and usually, its quantity depends on the crop yields. Biomass from crop residues is generally classified into two different categories: process residues and field residues. Field residues are defined as the residues which remain in the fields as a by-product of post-harvesting activities of the crop, whereas process crop residues are those generated during the processing of crops [9]. Field residue availability for energy uses is normally low, due to practical challenges associated with the collection of residues and the fact that all residues cannot be removed without influencing soil fertility adversely. On the other hand, process residues are usually obtainable in greater amounts as a result of the processing of the crop and may be utilized as an energy source [18]. The technical limitations (methods of harvesting, processing, and transporting) and possibility of destruction by uncontrolled fires are also a factor for residue availability.
The biomass residue availability potentials can be classified into gross residue potential, which includes the total quantity of biomass residue generated, and the recoverable residue potential which only constitutes the technically recoverable residues, or that proportion of the residues that remain once residues have been employed for other competing uses, e.g., employed for heating and cooking fuel, soil fertility, animal feeding and bedding, and surface mulching [10,19,20]. The potential of gross and recoverable crop residue can be estimated using Eqs. 1 and 2 [19][20][21][22]. The potentials of both the gross residue and the recoverable residue are assessed in this study.  Table 1 and supplementary material (Table S1)

Gross Residue Biomass Resource Potential
The gross residue potential estimation of a particular agricultural crop relies on the area cultivated, the crop yield, and the residue-to-product ratio (RPR). The yields of crop residue vary even more than the yields of the crop and are thus difficult to take into consideration, as it relies on location, plant variety, climate conditions, agricultural practices, and other factors [10]. Due to this reason, the residue-to-product ratio (RPR) values determined in the relevant literature at crop level were compiled for different crops and the average value for each crop residue type was used as given in Table 2. The potential of gross crop residue is estimated using Eq. (1): where GRP is the gross residue potential generated from "n" numbers of crops in tonnes (t); A (i) is the area under the i th crop in hectare (ha); CY (i) is the average crop yield of the i th crop in t·ha −1 , and RPR (i) is the residue-to-product ratio of the i th crop.

Recoverable Residue Biomass Potential
As described above, it is assumed that not all crop residue biomass will be available for the production of bioenergy The field-based biomass residue amounts that can be collected realistically are estimated via the recoverability fraction (also called surplus availability factor) of the crop residue biomass [20,26]. The recoverability factor (RF) is the fraction of residues that are available realistically for the production of bioenergy after part of it is utilized elsewhere [2,20,37]. The RF values for residue biomass were compiled from similar previous studies in different developing countries and the average value for each crop residue type was used due to lack of data specific to Madagascar, in order to estimate the recoverable residue potential as presented in (Table 2). The recoverable residue potential is estimated using Eq. (2): where, RRP is the recoverable residue potential from "n" number of crops in (t); GRP (i) is the gross residue potential generated from the i th crop in (t); and RF (i) is recoverability factor of i th crop.

Methane and Energy Potential Estimation from Crop Residue
The potential of the biomethane of the crop residues was estimated using Eq. (3). The values of the specific methane yield and the mean total solids concentration for the crop residues were obtained from literature as given in Table 2 [17].
where, MP is the potential methane production of crop residue (Nm 3 CH 4 ·year −1 ); MY is the methane yield from literature (Nm 3 CH 4 ·kgTS −1 ), and cTS is the concentration of total solid (%).
The potential amounts of energy available from the recoverable methane have been transformed from Nm 3 methane to gigajoule (GJ) of heat energy, applying the factor 0.036 GJ·m −3 methane [2].

Animal Manure Availability Estimation
Animal manure is an important input in the production of biogas. The quantities of animal manure ( AM ) potentially generated and recovered are estimated using the number of animals ( P live , head), mean annual manure production per animal ( M , kg·year −1 ·head −1 ), and recoverable fraction ( RF ). The animal manure biomass quantity which can be collected for energy application is calculated using Eq. (4) [26].

Methane Potential Estimation from Animal Manure
To estimate the potential methane production from animal manure, the main parameters needed comprise estimated manure per head per day and the concentration of total solids  [24,34] in the manure, as well as the methane yield per unit of total solids. To estimate the quantity of biomethane that can be generated by each animal category, parameter values are derived from similar studies carried out in Ghana and Tanzania [2,38], as described in Table 3. Equation (5) can be used to estimate the potential of biomethane from animal manure that can be generated from recoverable manures [17]: where MP is the potential methane production from recoverable manure (Nm 3 CH 4 ·year −1 ), cTS is total solid concentration (%), and MY is the methane yield from literature (Nm 3 CH 4 ·kgTS −1 ).

Energy Potential Estimation from Animal Manure
To estimate the potential amount of energy available from animal effluents, manure generated by cattle, pigs, poultry, goats, and sheep are considered.

Fuel Equivalent Estimation
In developing countries, the generated biogas can be used for the replacement of most commonly used traditional fuels such as firewood and kerosene. Biogas equivalent fuels were estimated based on the assumption that 80% of the produced biogas would be utilized for substituting firewood and the remaining 20% for substituting kerosene used in the households [15]. Firewood and kerosene equivalents of the generated biogas were then computed applying the calorific values of these fuels. All values of the coefficients were derived from literature and utilized in the estimation, as summarized in the supplementary material (Table S2).

Digestate Potential Estimation
Digestate is a high-quality organic fertilizer for crops with significant contents of nitrogen (N), phosphorus (P), and , micronutrients, and organic matter. It is usually utilized as fertilizer to crops without any further processing. In this study, the total quantity of digestate potential and its fertilizer equivalent was estimated. For the estimation, the mass of the digestate generated was calculated by subtracting the biogas mass (the quantity of substrate transformed into biogas) from the substrate/feedstock mass, as presented in the supplementary material (Tables S2-4). The mass of the biogas was derived based on the specific biogas yield and biogas density, by assuming the composition of biogas (average 60% CH 4 and 40% CO 2 ), and component densities (CH 4 0.72 kg·m −3 and CO 2 1.96 kg·m −3 ) [39]. The nutrient contents (N, P, and K) in the digestate were calculated by assuming that the digestate comprises a mean value of 52 g N, 42 g P 2 O 5 , and 43 g K 2 O per kg digestate on a dry weight basis [41].

Results
The biomass resource availability analysis of agricultural residue was conducted based on the number of agricultural households that practice mixed farming (crop and livestock production). Figure 3 shows that rice, maize, and potatoes were the most common crops, cultivated by 96% (irrigated 91% and rain-fed 36%), 87%, and 82% of all farmers, respectively, but additional crops were also cultivated by smallholder farmers, including beans, sweet potatoes, soybean, taro, cassava, tobacco, and others. Furthermore, several farmers had small numbers of animals mainly poultry, cattle, pigs, and very few small ruminants (sheep and goats).
The gross and recoverable biomass potential, available biomethane and its energy potential, and digestate potential of agricultural residues biomass (including crop residues and animal manure) were estimated for farming households in rural Vakinankaratra highlands in 2017/2018 in Madagascar. Biomass estimations are conducted based on the percentage of the household that farmed with a particular crop or animal species.

Crop Residue Biomass Resource and Energy Potential
In the agricultural system of rural Vakinankaratra highlands, the main crop residues during the 2017/2018 production year are from rice, maize, potato, cabbage, soybean, sweet potato, cassava, beans, taro, and tobacco. Residual biomass from these crops that are relevant to the production of biogas consists of the straw, husks, stalk, cobs, leaves, stems, peels, and shells/pods following harvesting and/or processing. The annual estimates of the gross residues in the agricultural system are based on the production of crop and residue-to-product ratio. Table 4 presents the mean crop production data and the generated residue potential from these crops during the 2017/2018 production year. The potential methane production and its equivalent total amount of energy from The recoverable residues for energy production are obtained by subtracting the amount used for other purposes from the gross crop residue generated. The estimates display that the average total amount of potentially available residues from crop production for the production of biogas is approximately 6 t per year per smallholder farming household cultivated with a respective crop, i.e., 83% of gross residues are available as recoverable in rural Vakinankaratra. At an individual level, potato contributed the maximum quantity of recoverable residue at approximately 1.33 t (22%), followed by rice and maize residues at approximately 0.96 and 0.52 t (16 and 9%) to the total recoverable residue per farming household farmed with a respective crop. At crop level, cassava residue has the highest competing uses, and only 39% are considered available for energy generation purposes as the peels are fed to animals (only 20% recoverable) or dumped into solid waste [2]. On the other hand, tobacco stems and stalks have the lowest competing uses for energy generation among the considered crop residues.
The potentially recoverable residue resources can be exploited for anaerobic digestion to generate biogas. The overall methane potential estimated from recoverable crop residues is 745 Nm 3 methane per year, equivalent to 26.8 GJ per year of heat energy per smallholder farming household in the study area. The available total methane potentials from crop residues are mainly from rice, potato, and maize with 213, 137, and 125 Nm 3 per annum of methane, respectively, equivalent to 29, 18, and 17% of the total crop residue considered, as shown in Fig. 4. This corresponds to 7.7, 4.9, and 4.5 GJ per annum of heat energy from rice, potato, and maize residues, respectively, as shown in Fig. 5.

Animal Manure Biomass and Energy Potential
The most common livestock in farming systems of rural Vakinankaratra highlands are poultry, cattle, and pigs, owned by 78%, 66%, and 60% of all households, respectively. Among those that farmed with poultry, the mean number owned per farming household was 13, followed by cattle (3), and pigs (3). Table 5 presents the mean number of livestock per smallholder farming household, the estimated manure generation, and the methane equivalent and its energy potentials. The mean number of animals per farming household is small. From the manure produced by the four animal categories listed above, the gross animal manure produced is estimated to be 19.4 t on a fresh weight basis per year per smallholder farming household as presented in Table 6. The highest gross quantity of manure contributed is by cattle, 13.14 t per smallholder household per year at an individual level.
With regard to total recoverable livestock manure, 10.5 t on a fresh weight basis per smallholder farm household per year is estimated, i.e., 54% of gross manure generated are available as recoverable. Livestock (usually cattle, goats, and sheep) are allowed free range during the day as most family farms are situated in rural areas. Thus, it is assumed that for half the day, the amount of produced manure from most cattle, sheep, and goats is not recoverable. However, animals are mostly kept close to the house throughout the day to prevent animal theft, which provides a good opportunity to facilitate manure recovery. Pigs are commonly kept in agricultural system enclosures that facilitate the easy recovery of animal manure. Poultry was found in the highest numbers in the studied area; however, it generates the smallest amount of manure per head because they are free-roaming by day and only spend the night in an enclosure.
From the recoverable fraction of animal manure, approximately 291 Nm 3 of methane can be produced: this is equivalent to a total of 10.5 GJ per year of heat energy in the rural Vakinankaratra region, as presented in Table 6. The total methane potential available from the recoverable livestock manure is largely from cattle, about 163 Nm 3 of methane (56%) as presented in Fig. 6. This high potential is because of the relatively large quantity of manure produced by cattle (approximately 12 kg FM per head per day on average), leading to a large recovery of cattle manure. Furthermore, the majority of surveyed households owned cattle as part of their production system, with an average of three heads of cattle per farmer.  (Tables S5-7). The estimated total digestate contains nutrient contents that were estimated to be 121, 98, and 101 kg of total-N, total-P 2 O 5 , and total-K 2 O per year, respectively.

Discussion
Animal manure and crop residue availability estimate for AD purposes is crucial for biomass supply sustainability. The estimated results display an important residue biomass potential for the generation of biogas in small-scale farming systems, which can substantially improve energy access and minimize biomass use in conventional ways. The estimated gross animal manure and crop residue biomass resources potentials were approximately 19.4 and 7.3 t on a fresh weight basis per smallholder household, of which 10.5 t (54% of gross) and 6 t (83% of gross) are available as recoverable for the generation of biogas, respectively, as summarized in Table 6. This indicates that crop residues have lower competing uses than animal manure. Manure  is usually applied to the farming fields to act as fertilizer. Anaerobic digestion could be a good alternative to provide both high-quality organic fertilizer and bioenergy, thereby decreasing the competing uses of manure. From the available crop residue and animal manure biomass resources, the total methane potentials are 745 and 291 Nm 3 of methane per annum per farming household, equivalent to 26.8 and 10.5 GJ of heat energy, respectively. The energy potentials of each agricultural biomass residues per year per household, for the different types of residues available, are presented in Fig. 5. Crop residues show both higher methane potential and energy potential from available agricultural residues than animal manure. This is a result of the low numbers of animal types and the low quantities of each type owned by the surveyed farming households, and of the higher competing uses for manure in the farming systems. Overall, the residues that have the highest biogas potential are those from cattle among animals and from rice, potato, and maize among crops. Biomass production is important during the rainy season (November to April). With irrigation, there is some production during the dry season, but this is much lower. Thus, the seasonal availability of the crop residues should be carefully considered for farmers in areas relying on a dryland production system as all crop residues are not available all year round. Additionally, other factors that could affect agricultural biomass resource availability, for example, animal breed types (manure) or changes in crop cultivars (crop residue), were not considered.
This study revealed that anaerobic digestion based on animal manure and crop residues can make a substantial contribution to meeting the energy demands of agricultural households in the rural highlands of the Vakinankaratra region of Madagascar. It is estimated that the methane generated has the potential to replace 4.6 t of firewood and 0.26 m 3 of kerosene per year. These findings revealed that if more priority will be given to bioenergy production from animal manure and crop residues, the existing problem of energy access to several regions of Madagascar can be eliminated so long as the quantity of residues required for the animal feed and soil fertility is maintained. Moreover, proper utilization of animal manure and crop residue biomass resources for the production of bioenergy in rural areas of Vakinankaratra can also replace a high percentage of traditional cooking fuel that is represented by firewood and straw (about 83.3% of households in the rural area) [42]. Switching from these conventional solid fuels (such as wood, straw, charcoal) to more efficient modern fuels like biogas can lead to substantial  Theoretical Available reductions in household air pollution as well as reduce pressure on natural resources. Biogas is currently not being utilized to a significant extent by any households in the rural areas of Madagascar. In Madagascar, the electricity supply does not cover the entire territory: there are three interconnected networks, around the towns of Antananarivo-Antsirabe (RIA), Fianarantsoa (RIF), and Toamasina (RIT). The total length of the current transmission lines is approximately 1000 km. However, a large part of these transmission and distribution networks are obsolete and are increasingly causing incidents. Most of the lines and equipment are overloaded. In the Vakinankaratra region, only 9% of urban dwellers and less than 2.5% of rural dwellers have access to electricity through the grid in 2019 according to the Ministry of Energy, Water and Hydrocarbons (MEEH). So, it seems difficult with so few rural people having access to electricity and such an underdeveloped grid to consider connections from biogas production. The energy produced should be consumed on-site.
The estimated energy potential from the AD of estimated agricultural residues in this study is higher than the energy demand estimated for cooking. To satisfy the cooking needs per capita per day, 0.33 m 3 of biogas is required as it has been estimated for Nepal [43]. Haladová et al. [3] estimated 1 m 3 of biogas for meal preparation only for the average Malagasy family per day. The estimation was done based on 200 L of biogas (with 60% of methane content) needed for cooking three meals for one person per day, with a mean Malagasy household size of 4.9 persons. This biogas quantity is lower than the one reported for Nepal. For Vietnam, it has been estimated that 0.8 to 1 m 3 of biogas is needed for a typical farming household of six people per day, which is comparable [44]. However, Bond and Templeton [45] stated that the production of biogas to provide a five-member family with two cooked meals a day is about 1.5-2.4 Nm 3 biogas in developing countries. Similarly, Tolessa et al. [17] reported 0.5 Nm 3 per day per person of biogas for cooking and 1.25 Nm 3 per day per person of biogas for complete replacement of traditional fuels for low-income South African households. Based on these reports, biogas generation to meet the energy demands for cooking with biogas is approximately 0.8-2.5 Nm 3 biogas (0.44-1.5 Nm 3 methane) per day per average household in developing countries. This is equivalent to 5.8-19.7 GJ heat energy per year (which corresponds to 3.2-10.8 GJ·year −1 available energy demand for cooking with biogas by assuming an average of 55% biogas thermal efficiency).
After cooking, lighting is the second most usual enduse of biogas, particularly in the areas that lack connection to the electrical grid. Biogas is used for lighting with the aid of exclusive gas mantle lamps which consume around 0.07-0.14 Nm 3 of biogas per hour [43]. Biogas lamps are less efficient than electric-powered lamps but more efficient than kerosene-powered lamps [44]. Farming households normally prefer biogas for cooking instead of lighting. Thus, the use of other lighting technologies is recommended, for example, home photovoltaic systems consisting of solar panels, light-emitting diode lights, and battery and charger for a cell phone.
Digestate is a high-quality fertilizer for crops with significant contents of nitrogen, phosphorus, and potassium. As described in the "Methodology" section, the mean nitrogen content of digestate is 5.2% (4.2% P and 4.3% K) on a dry weight basis. After AD, ammonium (NH 4 + ) accounts for about 65-80% of the total nitrogen in the digestate, which is highly bioavailable and immediately available for crop uptake [12]. In digestate, the content of ammonium is directly associated with the total N content in the substrate. The higher the NH 4 -N contents in the digestate, the higher the efficiency of the digestate as nitrogen fertilizer. However, during the AD process, the increases in the concentration of NH 4 + promote losses of gaseous N after digestate is applied to the soil and increased the short-term availability of plant N. Thus, to improve the use efficiency of N and promote digestate with a greater NH 4 + share of total N, caution should be given to vegetation periods, application methods, and season to minimize losses of N such as by mixing the digestate instantly with soil. One ton of dry digestate contains approximately 52 kg N on average, which corresponds to 115 kg urea fertilizer. From the estimated 2.3 t total quantities of digestate generated from all residue types, N content is 121 kg per year per agricultural household in the rural Vakinankaratra region of Madagascar. This corresponds to 268 kg urea fertilizer per year for those farmers who have all the residues, and which can then be used for crop fertilization.
Digestate can be utilized as organic fertilizer and is indicated to be more appropriate than raw agricultural residues (e.g., manure, slurry) for fertilizer application. It has greater organic carbon retention and increased bioavailability of nitrogen due to decreased losses of N during decomposition. Degradable organic matter is easily transformed into methane and carbon dioxide during anaerobic digestion, whereas complex organic matter, for example, lignin remains in the digestate, thus increasing its quantity of effective organic carbon, which remains in the soil for at least 1 year. Therefore, this contributes to the humus build-up in soil (in digestate 33.7 kg per ton on average vs. in pig manure 20.0 kg per ton on fresh weight) [12]. High humus status enhances the infiltration of soil microorganisms and the holding capacity of water is sufficient for rain-fed crops [46].
The utilization of residue biomass resources to provide clean renewable energy to smallholder farmers can alleviate the problems that arise from the absence of modern clean energy facilities. It also enhances the farming households' productive capacity. Livestock manure is either applied directly to crops as manure, left unused, or utilized for cooking in dried form, whereas the unused portions of crop residues are left or burned in the fields, which leads to loss of nutrients present in the residues as well as air pollution.
If utilized properly with appropriate processing technology, animal manure and crop residues can be a source of biogas. The production of biogas through anaerobic digestion technology is well suited to small-scale farmers as both an alternative energy source and to recycle nutrients, as well as a management practice for residues. Biomass resources are the main factors that affect the development of small-scale anaerobic digestion technology in rural parts of developing African countries. The development of smallscale anaerobic digestion technology is influenced not only by biomass resources but also by the level of rural social economy and other factors [47]. The income of consumers and their location's energy status are vital in deciding whether to select biogas as an energy source or not. In some poor rural areas, farmers do not have available capital to fund digester installation. This may require government support during the initial phase so that the cost of installing and maintaining a biogas system does not prevent farmers from using this source of energy. A smaller household type design digester with a volume of 4 to 12 m 3 might be more affordable for a family farm with two or three cows (or other equivalent substrates) [48,49], such as in rural Vakinankaratra areas. The household-scale type digester appears more valuable for small farmers to produce biogas and organic fertilizers with lower investment and maintenance costs [50]. A household-scale type digester might be more attractive for a small-scale farm as feedstock are available at a certain distance without transport and storage issues. Moreover, the farms are small, so the quantities to be stored are not very important. However, this constitutes a point of improvement: to reinforce the storage capacities so that this one is done in good conditions, it is besides one of the techniques popularized for the improvement of the quality of the manure. Therefore, investments must be planned in addition to the digester.
Finally, this study proposes that small-scale anaerobic digestion technology should be adapted or developed according to local situations (including biomass and water availability). This could be done to explore possible strategies that will identify ways to enhance biomass potential and utilize the resources feasibly (e.g., investigation on whether the use of a smaller digester or a community digester shared by households is more feasible in local situations). Further study is also recommended on the adverse effect of utilizing digestate as a fertilizer as well as about digestate storage and utilization mechanisms after the AD process, considering a local context; however, this did not form part of this study on biomass estimation. Future studies would be of great value to future project assessments relating to the effective use of the available resources for AD application in rural communities, particularly future investigations relating to alternative usages and approaches such as natural biogas/bio-natural gas for vehicles, BioNGV (e.g., for replacement of conventional fuels) in the circular economy and economic development context to realize the full potential of AD and help farmers to better use their farm residues. This could also allow farmers to diversify their income and improve their socioeconomic development.

Conclusion
Using data from agricultural household surveys combined with literature models and publicly available data enabled us to estimate biomass resources availability for anaerobic digestion (AD) in small farms in the Vakinankaratra region of Madagascar. The estimation of agricultural biomass residue has revealed that there is an important potential of biomass feedstock for AD application in the region. Small-scale biogas technologies could be utilized near the source of feedstock in smallholder farming systems and supply energy off the grid. This can make an important contribution in meeting the energy need in rural areas of the country, where the government has been unable to supply modern energy applications. Besides, a stabilized organic fertilizer produced from livestock manure and crop residues through AD can improve the soil quality and fertility and improve crop productivity within an agroecological small production system. The established baseline for biomass resource estimation at the small farming system level can be used for other provinces in Madagascar and elsewhere, to establish a similar baseline and quantify the potential for bioenergy generation from locally available biomass resources. Further studies are recommended to assess the ecological and economic value of biomass feedstock collection and conversion at a national level to integrate into the mainstream energy sector. Data Availability All relevant data are within the paper.

Declarations
Competing Interests The authors declare no competing interests.