5.1 Loss-assessment methods
5.2 Causes of losses
5.3 Other losses within the post-harvest system
5.4 Economic environment of the farm- household
The quantitative assessment of weight losses is important for evaluating the current post-harvest techniques and appraising possible actions for improvements. Despite the importance of cassava as the fifth most important agricultural crop in the world, so far very limited data on post-harvest losses exist (Bild der Wissenschaft, 1996). This is related to the low commercial value cassava carries and difficulties measuring post-harvest losses on such a heterogeneous product. The present study paid particular attention to the comparative analysis of various methodological approaches and to the quantitative evaluation of losses on stored dried chips. Figure 50 illustrates that assessing post-harvest losses requires the understanding of the complex post-harvest system, and various issues presented will be addressed throughout the discussion.
Figure 50: Various interacting factors related to the assessment of post-harvest weight losses on dried cassava chips
Two existing methods of quantitative loss-assessment - GM and VGM - have gained some acceptance for assessing losses in the root and tuber field (Wright, 1991). The GM was chosen as a reference method against all others because of its precision in measuring weight losses. A very good correlation of r = 0.9 was measured for chip boreholes in the trials under controlled conditions and in the field study in Ghana. The method is very precise but also laborious and time-consuming in obtaining baseline data (Wood et al., 1980; Boxall et al., 1986). Furthermore, the precise measurement of weight losses depends on the exact measurement of the moisture content. This could lead to a slight over- or underestimation of weight losses because the m. c. based upon an average sample. The technical operations might limit the use of the GM to research teams and project staff.
The other quantitative method, the VGM, described by Wright (1991) as fairly good for the assessment of weight losses, has its great advantage in being able to assess losses without collection of baseline data from undamaged chips. But during our trials under controlled conditions and the field studies in Ghana, the VGM either overestimated weight losses at low ranges of post-harvest losses or underestimated weight losses at higher ranges of damage. Also, the correlation of the VGM with boreholes was only r = 0.04 to 0.4 for the laboratory and field studies. Compton et al. (1993) stated that it was difficult to measure the volume of chips at higher levels of damage and that the data were of poor quality.
The newly-developed MM overcame the shortcomings of the above-discussed conventional loss-assessment methods. The method was able to measure weight losses of chips accurately over a storage period, in part due to the non-usage of baseline data. The method produced precise figures which positively correlated with boreholes and r = 0.9 for field studies and studies under controlled conditions. The method overestimated weight losses only slightly at the beginning of the storage season when few boreholes were visible. The necessary apparatus to operate the MM is a vacuum device which might limit the usage of the method to researchers, trained project and field staff, but otherwise its procedures are easy to execute.
All three technical methods have at least one common source of error: hidden infestation of chips by larvae, pupae or adult insects will tend to mask actual weight loss (Dick, 1988). Losses may be underestimated when insects or flour are still in boreholes and block frass tunnels, thus making volume or weight measurements inaccurate. But, in our studies it was found that this only caused slight errors. Difficulty also arose when the chips were severely damaged so that the surface started becoming crumbly and chips absorbed. This made measurements of the outer and inner volume impossible, especially measurements by the VGM and MM. However, all technical methods are able to produce objective results which make comparisons with other studies in other regions possible.
Another method is a compromise of a qualitative and quantitative assessment of losses, namely the VDSM which was introduced by Compton (1991) during a study in Togo. The method has the advantages of requiring no sophisticated equipment and uses reference photographs or drawings for grading chips in different classes. Techniques of this nature are used routinely for many field crops and at pre-harvest stages, particularly for disease studies (Chiarappa, 1971; Wright, 1995 and Krall, 1996). The VDSM also uses the relationship of boreholes-to-weight-losses to assess damage scales on cassava chips, but this can result in inaccuracy because not every hole is the beginning of a frass tunnel and results in weight losses. For example, it was observed that one P. truncatus beetle was able to drill up to seven boreholes before it made its way into the cassava chip during investigations under controlled conditions. Wood et al. (1980) concluded that visible damage seems to be a poor indicator of weight loss. Wright (1991) noted that visual scoring gives a better indication of losses at a particular time of storage and is not as applicable over an entire loss-assessment study.
When calibrating damage classes with different loss-assessment methods during our studies, the range of weight losses for each scale varied with the quantitative methods chosen for calibration. Visual scoring is generally inaccurate in determining weight losses and is better used for a general estimation of losses sustained by the farmer. The pre-determined five scales require the ability of farmers and traders to distinguish the scales for damaged chips without difficulty. An increase of visual scales as suggested by Compton et al. (1993) requires experienced staff who are able to differentiate such fine gradings for infested cassava chips. The possible usage of the method by different groups, i.e. farmers, traders and extension staff, suggests it will be widely used in the future.
The variability of local storage techniques indicates that a definitive loss-assessment methodology for all situations would be im-practicable. There can be no single best method for assessing losses; the methodology must be prescribed for each situation to meet local constraints. Moreover, it must be stressed that each method gives different values for post-harvest losses incurred (Boxall, 1986).
Cumulative weight losses
It is important to relate actual weight losses to the quantities of cassava chips originally stored and the pattern of family consumption to give an indication of overall losses throughout the year (Boxall, 1986). Our farm studies in Ghana indicated fairly low cumulative losses, about 2.5% at the end of the storage season compared to 8% weight loss measured with loss-assessment methods after eight months of storage. This is in contrast to findings of Nicol (1991) and weight loss estimations of 15 - 20% for chips stored for eight months. The station studies measured cumulative weight losses of about 5% compared to about 21% weight loss in the eight months of storage; that is, about three to four times higher losses than those of the farm study. This could be attributed to the different quality of cassava chips bought at different farm locations for our on-station trials, different stock management and pest pressure from a nearby maize store.
Obtaining a true representative sample from all parts of the stored batch was not always possible during the chronological field studies. The distribution of insects and losses among different layers of the stores has been often discussed. Wright et al. (1993) noted, "There is a trend for all the measures to increase towards the base of the store." This is true for high levels of infestation but less applicable for low levels of pests. In our study, 20 chips were selected from the surface layer and, if possible, from two to three layers below the surface of each store. In summary, due to the enormous variation of size, texture and outer volume, the sample size should be sufficient to keep these problems in controllable limits.
Qualitative loss analysis
Most of the commonly used loss-assessment methods measure physical weight losses on chips and do not address other measures, such as quality-nutritional loss, loss of seed viability or commercial loss. This is not because measuring weight losses is the most important aspect of loss but simply because it is easy to measure. The quality aspect is quite difficult to define and measure objectively, especially when no quality criteria exist in most of African countries. The Ghanaian authorities are so far not operating qualitative standards.
During the trials under controlled conditions, the starch level of the variety SORAD, was reduced by about 7% after 12 weeks of storage. It seems that the infestation of chips by insects resulted in the generation of large quantities of cassava flour. These results confirm the findings of Wright et al. (1993), who noted that P. truncatus infestation reduced starch levels by about 4% in the station trials in Togo and referred to it as a negligible loss. Kumar et al. (1991) found that in trials under controlled conditions, about 7% reduction of the starch level was measured on plain dried chips. In contrast are losses measured during our field studies in Ghana, with only a slight decrease of 0.3% of the starch level of local cassava varieties.
Other quality aspects worth considering relate to insects in that they produce uric acid, inoculate fungi and bacteria, and leave faecal matter and cast-off skins in and on the chips, creating a foul odour. The potential harmful effects of fungal and bacterial metabolites on consumers are so far not really appreciated in developing countries (Majumder, 1982; Coker, 1994). However, the assessment of such losses is very difficult and needs the incorporation of costly laboratory analysis.
Significant differences in the HCN-content were observed between the beginning and end of the storage period. During the trials under controlled conditions there was a pronounced decrease of about 57% HCN (variety SORAD), and a decrease of about 44% HCN (local cassava varieties), was noted during our field studies in Ghana. Post-harvest processing of chips is very important to decrease the toxicity of cassava by reducing its cyanogenic glycoside content; hence, the reduction of HCN-content is necessary prior to consumption of chips.
Newly dried cassava chips in the Northern region of Ghana contained a moisture content of 8 -10%. With the onset of the rainy season, there was inceased moisture absorption by the chips to 14 -16%. Higher levels of m. c. were obviously beneficial to insects and could have affected the susceptibility of dried chips to their damage (Shires 1979). Chittenden (1911) noted that P. truncatus is tolerant of dry conditions within the ranges of 22 - 35° C and 50 - 80% r.h.. Golob et al. (1990) observed P. truncatus infestation on maize at a moisture content of 9 to 11% in a field study.
As already mentioned, insects are the main cause of weight losses on cassava chips in our study, whereas micro-organism and rodent attack are in general of minor importance. The majority of farmers (65%) referred to insects as their main problem by the end of the storage season. Rodents (17%) and mould (7%) accounted less for the damage to chips. Nevertheless, micro-organism and rodent attack may become more serious under changing climatic conditions or due to a sudden increase in impact of other factors.
In laboratory experiments with beetles of the Bostrichid family, these seemed to be the most destructive pest. Weight losses on chips measured about 39% to 57%, depending on the climatic conditions and start population for trials with single populations of P. truncatus This destructive potential is also confirmed by experimental work in Togo (Wright et al., 1993; Compton et al., 1993) and Tanzania (Hodges, 1985). Schulten (1988) noted that infestation by the Larger Grain Borer Prostephanus truncatus (Horn), a Tanzanian strain, caused losses three to five times higher than those caused by indigenous pests. The enormous destructive potential of P. truncatus could be attributed to (1) its enormous frass potential and (2) high reproductive rate in comparison to the other insect species. Single populations of R. dominica and D. bifoveolatus under the same climatic conditions measured weight losses of 2-6% after eight weeks. In single experiments with T castaneum, no development was measured; this could be due to the cannibalistic behaviour of this species and the fact that the beetle depends on already damaged chips as a secondary pest.
Further experiments with different insect species exposed to 100 g dried chips were set up in the laboratory because always a range of insect pests is found infesting dried chips in Africa. The weight losses for chips exposed to mixed infestation of two different insect species were between 19-37%, depending on the insect species associated with P. truncatus such as the Lesser Grain Borer, Rhyzopertha dominica (Fabricius), the Wood Borer, Dinoderus bifoveolatus (Wollaston) and Tribolium castaneum (Herbst) The development rates of P. truncatus were reduced in comparison to their single infestation of 100 g cassava chips but still outcompeted all other species in numbers, especially in experiments with T. castaneum. The reduction in numbers of P. truncatus in experiments with mixed infestation is referred to as an effect of interspecific competition (Giga and Canhao, 1993).
Insect species collected and specified during the field studies in Ghana were already known to thrive on cassava (Pingale et al., 1956; Ingram and Humphries 1972; Parker et al., 1979; McFarlane, 1982; Hodges 1985). Weight losses of cassava chips due to exposure to insects started becoming obvious with the onset of rains and by April/May in the second and third months of storage. After four to five months of storage, 58% to 78% of the sampled chips were damaged by insects. Weight loss levels during the on-farm studies were mainly associated with the levels of primary storage pests: D. minutus with 79%, S. zeamais with 10%, and R. dominica with 6% related to the total number of insects. It seems that the different compositions and numbers of insect species in farm and station studies also caused different weight losses on dried cassava chips, with 21.3% in the station compared to 8% on the farm.
In both surveys, the Wood Borer D. minutus accounted for about 74% on the farm and 83% in the station of the total insects between the second and eighth month of storage. This number dropped only in the station trials to 53% in the eighth month, and the maize weevil Sitophilus zeamais increased in number from 10% in the sixth month to about 42% in the eighth month in the station. It seems, especially after the station study, that S. zeamais is a strong competitor for food and space. These results fall in line with findings from Giga and Canhao (1993) and experiments of population dynamics in laboratory experiments with P. truncatus and S. zeamais. The maize weevil was a strong competitor towards P. truncatus but the competitive relationship was also dependent on climatic conditions and mode of presentation of food.
Storage at farm level is a dynamic system involving continual movement of the stored produce in which there is a changing insect population. Establishing relationships between numbers-of-insects and post-harvest-weight-losses are difficult because of intra- and interspecific competition of insect species and possible secondary damage due to mould infestation or rodent damage. Moreover, two species, Heterobostrychus brunneus and Sinoxylon sp., which accounted together for only 0.3% of the total quantity of insects, caused considerable losses on chips in a few stores but are rarely mentioned in the literature as pests on cassava chips in Africa (Frappa, 1938; Lepesme, 1944; Mangoendihardjo, 1981; Dobie et al., 1991 and Delobel, 1992).
Micro-organisms, mainly fungi, can render produce unfit for human consumption. Especially in times of heavy rains high relative humidity damage due to micro-organisms seemed to be a problem in our studies in Ghana, where damage levels in the samples were measured between 4 - 5% on the farm and 11 - 34% in the station. The storage fungi, mainly Aspergillus spp. and Penicillium spp., produce mycotoxins that may carry a health risk for man and domestic animals (Coker, 1994). Farmers told us during the surveys that they simply dry the chips in the sun when mould infestation is visible and still use them for consumption. Moreover, they prefer the special taste of slightly mouldy chips and seem not to be worried about possible health risks.
Adequate data and appropriate survey and sampling techniques are scarce for measuring losses due to rodent attack. In the loss studies, rodents were of minor importance and only in a few cases did they seem to be a problem. Rodent damage in the samples was quite low, between 1-5% on the farm and 1-3% in the station.
Possibilities of reducing post-harvest losses
P. truncatus was not found in the samples in the Northern Region of Ghana, but we were told that the beetle has been detected in some parts of the Northern Region. So far, the beetle is mainly in the Volta Region (Compton, 1994). With a further spread of the beetle towards the Northern Region, the post-harvest management of chip storage could change, i.e. protection management would become necessary. Farmers stated in our study that useful insecticides are often unavailable, consultancy from the extension side is missing and they can rarely afford costly inputs.
Pest control using post-harvest protectants on dried chips should be the last option when physical and biological methods of control have already been in use. Identification of an effective chemical treatment for stored chips that is safe for the environment and the people is nowadays of international priority and has to consider several aspects. Toxic side effects for the consumers have to be considered; already, carcinogenic effects are related to pyrethroids which so far have been described as fairly safe for human beings (Der Spiegel, 1995). In this respect, chemical analysis of insecticides should not be limited to tests on the stability of an insecticide, but residues of the applied chemicals in the food should be also monitored. Thus far, insecticide trials with dried chips have been mainly conducted in the laboratories or under controlled field conditions (Golob et al., 1982 and 1990; Magoma, 1988; Wright et al., 1993; Reichmuth et al., 1995). Field trials on small farms are still pending, and hence, a recommendable insecticide application method for farmers has yet to be determined.
An "Integrated Pest Management" (IPM) strategy for stored chips based on economic damage thresholds with limited chemical control could be another option for reducing damage levels. Henckes et al. (1991) suggested for maize stored in Tanzania that, "Maize which is stored for on-farm consumption during the first months after harvest should preferably not be treated with insecticides. That means the total amount of maize will be divided into an untreated and a treated part. The quantity of produce which is treated with insecticides must be determined every year based on economic damage threshold."
The establishment of an economic threshold for numbers-of-insects-to-weight-losses is difficult due to intra- and interspecific competition. Furthermore, additional data on socio-economically relevant aspects are necessary, such as socio-economic behaviour of the farmer, judgement of risk and uncertainty. Current literature on threshold calculations emphasises that the calculation of a threshold for protection measurements requires data on other factors as well as on infestation damage relationships. In addition, farmers seem to be prepared to invest in adequate post-harvest initiatives only if an economic value is attached to chips and marketing of chips is a valid option (Freier et al., 1994; Bell and Mück, 1997).
Biological control of storage pests using predators and/or pheromone trapping could be another technique for reducing damage to chips. Biological control of P. truncatus is already an option in Africa to control the field and storage infestation of maize and to reduce the damage potential of this storage pest (Böge and Fischer, 1993).
Loss estimations for the post-harvest system revealed that losses appeared at all stages from the field to the consumers' table (Chinsman and Fiagan, 1986). Data gathered during the general questionnaire in 1994 (MOFA-GTZ) and the multiple visits to farmers in 1995 revealed that between harvest and processing of cassava, about 46 to 54% of the potentially harvested produced is lost even before storage. Jeon et al. (1991) estimated a loss of about 45% of the potential cassava yield during each operation within the post-harvest system in West Africa. Losses are estimated to be about 14% for harvesting, about 9% for handling and about 22% for processing. About 46 to 55% of the potentially harvested produce in the Northern Region is actually stored. Physical losses within the system ascertained during our surveys in Ghana could be also attributed to:
· Differences in the estimation of potential and attainable yield. The PPMED (1994) noted about 7.2 MT/ha or 2.9 MT/acre fresh cassava for the Northern Region might be addressed as potential yield, but the attainable yield could be much lower.
· Farmers do not harvest all cassava roots (Pessey, 1996; TechnoServe, 1994) and keep a certain amount in the ground as a reserve capacity.
· Peeling of roots involves a certain weight loss, but the peels are used for cattle feeding.
· Visibly damaged cassava roots are either consumed immediately or fed to cattle.
· Consumption of dried chips takes place before storage.
· A certain amount of harvested cassava chips are sold before storage.
Weight losses between harvest and storage of chips represented a substantial financial loss in 1995 of about 3.2 Mill. Cedis (about 4748 DM) for the 21 farmers involved in the intensive loss-assessment study. In addition, financial losses during storage amounted to about 170058 Cedis (about 252 DM) after 6 months of storage; weight losses are referred to as actual monthly losses of 4.6%. The calculations were based on the assumption that potential yield of cassava is about 7.2 MT/ha and the conversion rate for dried chips is 2.7 kg fresh cassava = 1 kg dried chips. If all weight losses are added up for 150 farmers involved in the MOFA-GTZ study in 1994, the financial loss is worth about 13.6 Mill. Cedis (about 25932 DM). Financial losses related to the storage of dried chips increase the figure by about 1 Mill. Cedis (about 1490 DM) for a storage period of 6 months. All calculations are based on producer prices; thus, the financial loss could be even higher when operating at wholesale prices. The physical and financial loss figures for the potential cassava yield within the post-harvest system are thus quite high, and it is worth focusing on this area in further studies.
Economic assessment of cassava chip storage
Economics plays an important part in post-harvest loss studies. It allows decisions to be made on the optimal use of resources. The immediate sale of dried chips after drying in February is less rewarding to the farmer than sale after six months of storage if the economic assessment is only based on gross returns, without taking into account costs for storage structures or interests. The farmers often calculate with lower capital costs if any and with nearly no operating costs. If the farmer decides to store until December when wholesale prices increase and the financial returns seem to improve, then cassava chips will be exposed to insects far longer and weight losses are likely to increase. The weight and financial losses then appear to be significant for the farmer and the survival of the farm family. Chips that are lost during storage have to be replaced at much higher costs at the end of the storage season. The storage of chips has also a survival function and must bridge the gap until the next harvest of other crops. It appears that farmers are better off storing chips for their own consumption as a security or strategic reserve.
So far, chip marketing is thus far more dependent on the farmers urgent need for cash for the family or farm inputs. Market access, economic stability or attractive prices could be favourable for the future marketing of chips, but in fact, market places are difficult to reach for the farmers and middle-men and women. The producers receive low prices for the chips and the price fluctuations over the storage season are not very attractive for the producers. In addition, an inflation rate of about 60% in Ghana in 1995 revealed no real price increases for producers in the long run.
Simple cost-benefit calculations are too static for a complex situation such as the post-harvest sector in developing countries. Semi-commercial farmers assess their profit, costs and benefits differently. Therefore, the socio-cultural environment of peasants has to be taken into consideration and especially their attitudes towards risk and uncertainty when assessing weight losses and marketing possibilities for cassava chips. The personal belief of the farmer is a major determinant of decision-making in the technical, economic and social decision framework of the post-harvest sector. The specific economic, technical and, especially, socio-cultural environments of farm-households have important implications for farmers' decision-making behaviour. Furthermore, decision-making in developing countries is more severely influenced by uncertainties such as unpredictable climate, imperfect market infrastructure and political instability. (Stumpf, 1989; Ellis, 1993).
Very few studies use the farmers' views on losses as an essential source of information. Although insects, mould or rodents may be the most important storage problem for cassava storage, this looks minute beside other constraints. Farmers stated that storing chips is a problem among many others related to poor health conditions and limited infrastructure. In the farmers' opinions insect pests are their most serious storage problem. The majority of the farmers involved in the farm study were able to detect infestation habitats, i.e. wooden material (67%), as one source of possible infestation. About 90% of the farmers believed that specific trees used for constructing drying- and storing-structures are more susceptible to insect attack, i.e. dawadawa (Parkia clapertiniana) and shea butter tree (Butyrospermum paradoxum). Despite their knowledge of possible infestation sources, a special wood treatment is not performed, i.e. proper hardening of wood with fire. Farmers also had a fairly good idea that the onset of rains, length of storage period and dryness of chips were factors that could increase the damage to chips.
Despite our findings for the weight losses sustained by the farmers' crops (actual weight losses of 8% and cumulative weight losses of about 2.5%), the farmers perceived their losses differently and much higher. Most farmers perceived their losses as medium-severe to severe, between 5-20% depending on the family-size and the amount stored. This could be due to several factors:
· Cassava is a very important security crop and farmers depend for their survival on a sufficient amount stored insect-free in the lean season to bridge the gap until the next harvest.
· Interviews were mainly carried out only in the lean season and towards the end of the storage season, when farmers might have perceived chip losses much higher than actual.
· Other agricultural products have not been sufficiently stored.
· Farmers expected assistance from the project and exacerbated the situation.
However, farmers also accept a certain loss level as unavoidable and God's will. Cassava chips that are very badly attacked and visually classified into damage class V are still considered fit for consumption by many farm families. Also, farmers stated that infested chips are pounded into flour and the insects are usually sieved off.
There is still need for precise measurements of weight losses at all levels within the post-harvest system appraising others' decisions, but these data are location - and culture-specific. It should therefore be apparent that a simple weight loss figure will not always reflect the importance of a loss in any given situation. The importance of a crop, whether it is considered as an important staple food or a minor food crop, has to be emphasised and accordingly investigated. Expensive loss studies with no relevant importance are not only costly but also untimely in times of limited budgets and prioritising of activities. The choice of loss-assessment method depends very much on the objectives and resources available in a given situation.
A multi-disciplinary approach with emphasis on socio-cultural conditions is needed, but it should also be noted that a more holistic view requires more time and different techniques and methods for assessing qualitative data. If the rationale behind loss-assessment studies is to eventually reduce post-harvest losses through some form of change, the adoption of loss reduction strategies will depend largely on the financial resource and time availability of farmers. It appears that in times of limited financial resources, the reduction of weight losses is in fact not necessary. In the future, easier access to markets, better infrastructure and improved marketing potential for chips might in turn encourage the adoption of protective measurements.
Marketing of cassava in Ghana is quite common, especially for gari, a processed form of fresh cassava roots. The marketing of dried cassava chips is a fairly new practice for farmers and also wholesalers. In the North of Ghana, dried chips are mainly used for home consumption, but the market orientation of farmers suggests that they would be willing to produce more if an increase in the agricultural production of cassava and land area were possible and more attractive prices were paid. Already cassava production has increased by about 17% in Ghana due to soil fertility problems and costly inputs for other agricultural crops. Future market possibilities for chips might increase the production potential of cassava, i.e. as fodder for cattle within the country or even for exportation to Europe. The business of exporting chips to Europe for use by its cattle feeding industry is a result of the recent EEC quota (140000 MT in 1994) given to African countries. Some companies in Ghana have started exporting cassava chips to Europe, and the demand will increase in the near future for well-dried cassava chips. Already these companies reckon with a strategic reserve of two million tonnes of cassava roots in the ground which is not needed by farmers and can be easily exported (TechnoServe, 1994; Pessey, 1996). Companies are already introducing quality standards, such as 12-14% m.c., minimum 70% starch, and composition free of pests, extraneous material and aflatoxin. All this could lead to a more commercial attitude of farmers and perhaps also to a year-round availability of cassava chips. The market could also help to change some storage practices in able to offer smaller chips that dry faster, store better and keep for a longer time (Laryea, 1995).
Having drawn these conclusions, it is now possible to make a number of useful recommendations. It is not necessary to develop more methods but to elaborate relevant loss-assessment methods such as MM, GM and VDSM and to incorporate other valid indicators to consider qualitative and economic aspects of the chip, such as texture, colour, nutritive quality, organoleptic criteria or economic value. But, it should also be noted that quality issues are less relevant in situations of limited financial and nutritional resources.
Usage of pheromone traps and relating infestation pressure to a possible damage of chips is another option to measure weight losses; however, these relationships are influenced by intra- and interspecific competition of insect species, threat of predator and micro-organism damage and climatic conditions.
The socio-economic environment of the farmer and farmers' active participation should be emphasised. The incorporation of indicators on the socio-economic and socio-cultural environment of post-harvest systems is urgently needed, as is gender-specific analysis. The understanding of the separate roles and responsibilities of men and women is necessary because both might evaluate the same constraints differently. Furthermore, the already existing and future marketing possibilities of cassava and cassava chips should be considered in future studies, because the commercial value of chips decides in the end whether farmers adopt certain protective storage procedures.