Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu

The Multiple Uses of Wood

The dominant direct utilization of natural habitats, however, is household use of wood. Wood has multiple uses for the CA household, and households' rate of use of wood is at the centre of the total valuation of natural habitats. The first and best-known of these multiple utilizations is fuelwood. It is well established that fuelwood as an energy resource is critical to rural households in Zimbabwe, since 95 to 100 per cent of these households use wood rather than any other fuel source for cooking and heating (du Toit et al 1984, MacGarry 1987). While cooking and heating is the single most important use, other major fuelwood uses include heating for special occasions, for beer brewing, and to fire brick kilns (Grundy et al 1993). Indeed, rural dependence on fuelwood as an energy source is well enough known for the imputation of fuelwood use values in household budget surveys (1985Z$70 pa, equivalent to 9 per cent of total cash expenditure) (CSO 1988). Note that, despite this dependence, successive researchers have found no general evidence either of wide-spread fuelwood scarcities or of fuelwood demands causing deforestation (Whitlow 1980, du Toit et al 1984, McGregor 1991, Grundy et al 1993).

Less well researched, but equally important, is the use by households of construction wood. The only thorough piece of research on this subject is reported in Grundy et al (1993). This paper, though reporting research conducted in a resettlement area which is likely to be better wooded than most Communal Areas and hence have higher values for wood use than the average household, demonstrates the central role wood plays in the household's creation of housing assets. This is on account of the range of structures for which households use wood - Grundy et al list thirteen separate structures as requiring construction wood. Though the annual household volume required (about 2.7 m3) is lower than fuelwood consumption, it is probably of higher value, as construction wood needs specific tree species as their source which will produce poles of adequate length, straightness, durability and resistance to termites (Grundy 1990). Indeed, scarcities in the availability of construction wood have led both to households searching intensively for good woods to use, and the emergence of markets (Dewees 1992).

Finally, wood offers the household a number of uses aside from fuelwood and construction material, for example in making agricultural implements, furniture, household utensils, musical instruments and weapons (Ellert 1984). Once again, use of these resources is widespread, with over 90 per cent of households using wood for these purposes (Campbell et al 1991b). Similarly, there is widespread use of other wild resources to make household goods - in particular, use of reeds, grasses, cane, bark and clay to make ubiquitous and essential items such as inter alla thatch, sleeping mats, winnowing baskets, rope and cooking pots. Indeed, high demand for all these items means that some households rely heavily on income from associated activities such as carpentry, thatching, weaving and pottery (Cavendish, forthcoming). However, accurate data on quantities used and values generated are not yet available.

The Indirect Input of Natural Habitats Into Production

Natural habitats also provide a range of indirect inputs into the agricultural production system. The role of leaf litter and termitaria in providing nutrients to Communal Area farmers has already been discussed. However, probably the most important indirect input is the provision of free fodder and browse to livestock. A detailed discussion of the woodland-livestock-crop production nexus is left until later, as it is in livestock herding strategies that the necessary adaptation of farmers' production strategies to the properties of the surrounding ecosystem are most clearly seen. For now, it suffices to note that it is the exploitation of natural habitats and their properties that enables households to maintain livestock stocking rates at their current - if economically inadequate - levels, and thus it is natural habitats that underpin the entire agropastoral production system.

Other indirect inputs exist, but their value is less clear. One such input is the shade offered by the existence of trees in fields, important to agricultural workers during the long hours worked in fields during the agricultural season. Although the density of trees needed to offer such shade is low (0.5 trees ha-1), Wilson (1989a) found that some trees in fields were preserved solely for shade. This is confirmed by Campbell et al (1991c), who found that amongst non-fruiting trees left in fields, the primary motivation was the benefit of shade, and by the data on reasons for tree-planting that is presented in the section on natural resource management, where it is reported that shade is one of the leading reasons given for households' tree-planting activities.

Finally, there is a range of benefits that can be classed under the generic heading 'ecological services'. Trees have a number of local ecological benefits: these include improving soil organic matter, soil structure, soil moisture status and soil nutrient status under or near to tree canopies, and reducing soil erosion on a broader scale (Ingram 1990). The result of these ecological benefits is improved yields and reduced costs of production, a fact recognized by farmers who have been found both to deliberately crop under tree canopies, and reduce their fertilizer applications accordingly. Farmers also benefit from reduced soil erosion due to leaving trees in their fields, especially if these trees act to stabilize field contours and thereby reduce the labour needed to repair damage (Dewees 1992). However, it is also the case that trees can have some negative impacts on crops, through inter alia competition for water, excessive shading of crops, and the harbouring of pest populations (Campbell et al 1991a). Hence, the net ecological effect of trees in fields on crop production is not always positive.

The Total Valuation of Natural Habitats to Communal Area Households

The above discussion is clearly plagued by a lack of adequate data which would enable a detailed comparison to be made between environmental utilizations and more 'conventional' activities. However, one attempt has been made to provide a valuation of some natural habitat utilizations (those associated with trees and woodlands), namely Campbell et al (1991a, 1991b), who tried two methods to elicit household valuations. The first assigned values to natural habitat utilizations on the basis of assumed consumption levels derived from the secondary literature, and used local market prices to convert quantities to values. The results of this exercise suggested that tree products and services have an annual imputed value of nearly Z$1,000 per household, a figure considerably higher than median cash income levels. The second study was methodologically more ambitious, in that it used a Contingent Valuation Methodology (CVM) to attempt a valuation of a broader set of utilizations than was covered in the first study (eg browse, shade, erosion control). While these second were (at reasonable discount rates) lower than those of the first study, it is likely that the CVM study underestimated the true willingness-to-pay of Communal Area households. However, whichever of the two studies is more accurate, the key point is that the value of natural habitat utilizations for just a subset of activities is of the same order of magnitude as that of income measured in conventional household surveys.

Conclusions

In conclusion, several points are noted. First, natural habitats offer services which generally have marketed substitutes: despite these substitutes' existence, though, households continue to depend heavily on natural habitats. Thus, this pattern of extensive natural habitat utilization is consistent with the picture presented earlier of highly-constrained Communal Area households, and in fact generalizes the pattern of choice seen in the agropastoral production function and fertilizer use. Second, natural habitat utilizations are important in almost all the major economic areas of the peasant household, viz. income, consumption, asset accumulation (housing; livestock), and agricultural production. Furthermore, studies show that they have a value to the household that is probably equal to the subset of activities that economists have traditionally focused on, such that their absence in past studies has rendered our knowledge of the type of peasant household featured in this paper flawed. Finally, the dependence of these rural households on environmental resources, and their inability to stabilize such dependence through financial or physical asset accumulation, implies that households' production and consumption levels are conditioned on the nature of the ecosystem in which they find themselves placed. It is these links that are explored next.

SAVANNA ECOLOGY AND PEASANT RESPONSES

Zimbabwe's Communal Areas are located in the savanna, and hence their ecology is best understood as a subsystem of savanna ecology. Generally speaking, while savannas are superficially highly diverse in terms of physical characteristics such as physiognomy, flora and fauna, it has increasingly been realized that they have a strong underlying regularity, principally relating to a division into two savanna biomas: the 'arid' savanna and the 'moist' savanna (Huntly 1982). As their names suggest,
the key underlying determinant of savanna type is the level of rainfall. Arid savannas receive less than 650 mm rainfall pa, most of which falls in four summer months (with drought therefore frequent and severe), while moist savannas receive more than 650 mm annual rainfall. This primary distinction produces a host of secondary differentiations between the two biomes: for example, these blames can be distinguished on the basis of floristic criteria, vegetation structure and faunal groups. With regard to Zimbabwe, it was seen above that the country has been classified into Natural Regions on the basis of rainfall level, precisely the primary determinant of the savanna.

However, there is an important secondary filter of savanna type, namely the availability of nutrients within soils. The availability of soil nutrients is obviously another key determinant of primary production (ie type and composition of herbage) in the savanna. As before, two broad soil types can be distinguished. The first are eutrophic soils, ie soils where there has been little leaching of important nutrients, so that the availability of calcium, magnesium, potassium and sodium is high. In contrast, the second are dystrophic, or highly leached soils, where nutrient availability is low. Again, this broad distinction has been widely applied to Zimbabwe, where there is a clear difference between the clayveld - soils with adequate nutrients - and the sandveld - soils which are nutrient poor.

While there is clearly some relationship between rainfall level and soil nutrients, in fact there is sufficient independence of the two variables that it is possible to think of savannas in terms of a 2x2 matrix defined by moisture and nutrient availability (summing and Bond 1991). Relating this to the Zimbabwean case study, for example, it is possible to locate the main woodland zones as shown in Table 11.4.

Table 11.4 Principal woodland zones

Source: adapted from Frost et al (1986).

It is thus possible to classify the different ecological areas of the savanna, and hence also of Zimbabwe's Communal Areas, using rainfall levels and nutrient availabilities as the key stratifying variables. Most importantly, such a classification highlights two key characteristics of the local ecosystems in which Communal Area farmers find themselves located, characteristics which highly-constrained and resource-dependent peasants must respond to in their production strategies. These are the patchiness and extreme variability of savanna ecosystems. We discuss these attributes of the savanna in turn.

1. Each element in the savanna matrix - or 'patch' - offers the agropastoral household differing quantities and qualities of primary production for the various economic activities it is involved in (eg agricultural production, livestock maintenance, collection of wild foods, fuelwood collection and storage etc). For example, consider the availability of grass and browse for livestock in these different patches. In moist savannas, there will be a high and relatively stable level of primary production. However, if located on nutrient-poor soils, the herbage produced will be high in structural matter and low in nutrients, and is therefore of low value to herbivores. In contrast, arid savannas will have much lower levels of primary production, but if located on nutrient-rich soils, this herbage will be highly palatable, of great value to herbivores. Similar differentiations can be made by ecological niche for almost all the resource utilizations, with each patch offering a different vector of resource availabilities. Most important, though, is the fact that these different ecological zones are locally present in any given Communal Area, forming a 'small-scale vegetation mosaic' (Huntly 1982). 2. The characteristic of Communal Areas is their ecological variability. Savanna ecosystems generally are subject to high annual climatic variability (ie rainfall), as well as recurring shocks such as fire, drought and floods. Rainfall variability and such shocks are quickly reflected in changes in the amount of primary production (grass and trees): thus, savanna vegetation is characterized by sudden and unpredictable changes, leading also to explosions and crashes in the populations of dependent species (Caughley 1983). Despite this instability, though, savannas show impressive ecological resilience in that while a species' numbers might vary enormously, actual extinction is rare. Indeed - and this point will be important when discussing the issue of sustainability - it is argued by ecologists that such resilience is directly related to the flexibility of spatial, temporary and dietary niche brought about by the instability and diversity of savannas: not only have savanna components adapted to considerable and frequent disturbance by developing mechanisms which allow recovery from stress, but also they probably require such variability and shocks to maintain their resilience (Walker and Noy-Meir 1982).

These characteristics - patchiness and extreme variability - are the primary attributes of the environment in which Communal Area farmers find themselves located. How have peasants responded to these environmental characteristics? Given the agropastoral nature of Communal Area production, it seems clear that a major challenges facing peasants is the accumulation and maintenance of cattle (peasants' herbivores). As it is the draft and manure functions of cattle that
constitute their chief value, the farmers' main objective is to maximize over time the number of units of cattle available to the household (or agricultural unit). However, since farmers are in general too poor to purchase feedstuffs, a critical constraint facing agropastoralists is the highly variable quantity of fodder and browse (primary production) available to their cattle on an inter-annual basis. In this environment, as Sandford (1983) details, farmers can broadly adopt one of two stocking strategies: a conservative one, in which cattle numbers are kept consistent with the level of food available in natural habitats in drought years, or an opportunistic one, in which population levels are allowed to follow environmental fluctuations, implying cattle die-offs during droughts (or even other, less severe, low-rainfall periods). Each of these has its costs: the conservative, as environmental utilization by cattle is below its theoretical maximum for all but drought years; the opportunistic, as capital and labour invested in cattle in the period before a drought/low rainfall year has a very low return due to cattle deaths.

Communal Area farmers' desire to maximize cattle numbers means they adopt opportunistic stocking strategies, resulting in cattle build-ups during runs of good rain years, followed by drought-induced die-offs. This pattern was particularly clear in the 1980s and early 1990s, in which cattle numbers followed the pattern of drought and recovery. Thus, cattle numbers, which had reached 6.1 million by 1975, fell to 5.2 million following the 1981-4 drought, recovered to 6.5 million by 1989, and then collapsed to 4.3 million after the severe 1991/2 drought (summing and Bond 1991, Department of Veterinary Services pers comm). Many rangeland researchers have now charted and supported this opportunistic strategy as an economically optimal response to the instability of the savanna ecosystem surrounding Communal Area farmers (Sandford 1983, Abel and Blaikie 1990, Behnke and Scoones 1992).

However, in order to avoid purely reactive strategies to savanna variability (ie temporal heterogeneity of resources), and hence in order to minimize cattle die-offs, Communal Area farmers actively exploit, through livestock migration during herding, the patchiness of local ecosystems (ie spatial heterogeneity of resources). In particular, herders capitalize on the non-covariance between rainfall levels and the primary production of the clayveld and sandveld zones, following on from their differing nutrient status. This non-covariance can be substantial, both intra- and inter-annually. For example, Scoones (1989b) shows that the wet year grass standing crop on the clayveld is a multiple of that on the sandveld, whereas the reverse is the case in dry years. Further, his range production data demonstrates substantial within-zone variability of primary production, and also highlights the important role 'key resources' (ie vleis, drainage lines, river banks and contour ridges) play in both zones in producing high quality and/or reliable foodstuffs.

Thus, with patches offering differing resource opportunities at different times, maximizing cattle numbers involves migration across patches. Researchers have charted both local and regional migrations, depending on the severity of climatic stress. For example, intra-annual rainfall variation poses the problem of surviving the late dry season: in response,
herders both migrate from local clayveld in the rainy season to local sandveld in the dry season and utilize heavily key resources (Scoones 1989b). During drought years, more ambitious migrations occur, with cattle moved up to 150 km from the homestead (Scoones 1989a). In general, the ability of herders to take advantage of the flexibility of dietary niches offered by local ecosystems, and their rights of access to these niches, is critical to the maintenance of livestock assets in the face of the rainfall variability noted above.

The focus thus far has been on the close relationship that exists between farmers' stocking strategies (opportunism and migration) and ecosystem properties. However, there are other ways as well in which the nature of the ecosystem and peasant strategies interact. For example, there is evidence that wild fruits - which suffer less during low rainfall periods than crops of the savanna - are used by households to smooth consumption. For example, Clarke (1983) found that one wild fruit alone (Grewin flavescens) comprised 25 per cent of the food items in the diet during the dry season after the 1981-82 drought, versus a much smaller proportion found by Campbell (1987) for wild fruit as a whole during better times. Furthermore, Campbell (1987) found fruit consumption predominated in the hot, dry season, which correlates well with shortages of cultivated foods but is out of phase with fruiting activity. However, it is not just wild fruits that have this role. During the severe 1991/92 drought, the price of wild meat actually fell, on account of the widespread adoption (despite its illegality) of hunting as a food-gathering and income generating activity (Cavendish, forthcoming).

As a final example, the patchiness of soils is exploited and responded to by farmers in their agricultural activities. For example, Communal Area farmers understand the different nutrient characteristics of soils, and hence the differing value of fertilizers in those soils (Swift et al 1989, Scoones 1992). Indeed, fertilizer applications vary widely, depending on the soil type being treated, even when different soil types coexist in the same field (Cavendish forthcoming). Similarly, planting decisions are made in the light both of crop type and of soil characteristics, with the coexistence of patch types offering farmers an opportunity to spread production risk.

So far, then, the pervasive role that natural habitat resources play in farmers' lives has been established, and evidence has been presented linking particular strategies adopted to the underlying determinants of the Communal Areas' environment. Despite this, it is often asserted that peasants are both ignorant about environmental resources and poor managers of those resources. In the next section, the opposite is suggested: on account of their value, environmental resources are both understood by peasants, and managed in many different ways.

PEASANT KNOWLEDGE AND ENVIRONMENTAL MANAGEMENT

The issue of peasant knowledge of soils - especially the clayveld/ sandveld distinction - has already been touched upon in the previous section. However, the full knowledge of farmers is much broader than this, with Scoones (1989b) claiming that farmer's environmental classification is dominated by soil differences (and according closely with ecological theory). Separate soils zones are clearly differentiated: not just sandveld and clayveld, but also red soils, sodic soils, rocky soils and so on, with each soil's properties (ie stability, nutrient status, water retention) known and understood. Furthermore, within this classification by soils, farmers distinguish separate habitats with reference inter alia to water movement, tree type and past use. For example, local woodlands are known by the prevalent tree type (eg Colophospermum, Acacia, Julbernadia, Uapaca); drainage pans, drainage lines, vleis and riverine areas are distinguished; separate categories are assigned to virgin land and abandoned fields (Wilson 1987).

Following the research embodied in Wilson (1987), there can be little doubt about the extensiveness of peasant knowledge of trees. This undoubtedly follows from the importance of trees and tree products in the household economy of Communal Area farmers. As Wilson writes:

People living in the rural areas start to learn about trees from a very early age... Knowing the trees [about seventy species7 is seen as being a basic requisite of personhood... People not only learn to identify trees, but where they are found... This interest, and the many hours that people spend in the bush means that people develop a knowledge of tree ecology, and an acute awareness of the resource.

Thus, the outstanding feature of peasant knowledge of such key resources as soils and trees is not its paucity, but its breadth. This knowledge and high resource values results in an array of resource management strategies in the Communal Areas, at both individual and community level. One well known strategy (noted earlier) has been the preservation of certain highly valued fruit trees during the process of woodland clearance, such that even in severely deforested areas, fruit trees are the last to be removed. For example, Campbell (1987), working in Save, found that the availability of fruit was not statistically different between the most and the least forested parts of the study area. Selective cutting of the woodlands in his sample area meant that the three most frequently used fruit trees remained relatively constant in cover over the two different areas, while other, less highly valued species had disappeared in the deforested area. Similar results were found by Wilson (1989a). In this survey of trees left in fields, he found that all locally-valued fruit trees were found to be left in fields, and that in dioecious species, the non-fruiting sex had been removed from fields. Likewise, Gumbo et al (1990) reported in their survey of 24 villages that of all trees left in fields, 82 per cent were indigenous fruit trees. Most recently,

Grundy et al (1993) found that of 20 species left standing in newly-created fields, a majority were fruit trees, and this was the primary reason given by respondents for their selection. Fruit is not, however, the only reason respondents give for preserving trees. Trees in fields can also be preserved for spiritual reasons, for leaf litter, for shade and (less frequently) for their positive impact on crops (Wilson 1989a, Matose 1992), though in many cases these uses overlap. Cavendish (forthcoming) finds that, for established fields, all surviving trees are claimed by respondents to have been deliberately preserved.

Another common - and more active - resource management strategy is the planting of trees by households in and around the homestead or garden (areas which are the closest to having unequivocally private resource rights). Much has been written about tree-planting in the Communal Areas of Zimbabwe (eg see Campbell et al 1993, Bradley and Dewees 1993), and the following summarizes recent writings. The profitability of tree planting in some cases rests on the multiple use that trees offer households. An example of this point is Sclerocarya birrea (mupfura): its fruit is eaten or used to make wine (mutumbe); its nuts are used either to make a local butter and cooking oil or are sold; its shade improves agriculture; its bark and leaves are used as medicines; and its wood is used for stools, drums, household implements and fuel (Gumbo et al 1991). However, other trees - especially exotic fruit trees - can be profitable even with the single use of fruit consumption, as establishment costs are low. Though there is little economic data in this area, Dewees (1992) shows that under reasonable assumptions, the rate of return on individual fruit trees is greater than 10 per cent.

As a result, tree planting is widespread. Among the different surveys in which tree planting practices have been examined, the per centages of households engaged in tree planting were 61 per cent in du Toit et al (1984), 70 per cent in Beijer Institute (1985), 74 per cent in Grundy et al (1993) and 88 to 97 per cent in Campbell et al (1991a). Thus, tree planting by peasant households is clearly a ubiquitous activity. The four main reasons for tree planting are fruit, shade, fencing and construction woods: with fruit the most important, the great majority of trees planted are fruit trees, whether indigenous or (more frequently) exotics. However, reasons for tree planting vary with woodland availability, such that in the more deforested areas, meeting household construction and fuelwood demands are more the prime motivations. Household members also devote considerable time and resources to managing planted trees, especially those around the homestead, with trees being watered, fertilized, mulched and protected from livestock. Finally, this widespread planting of trees has occurred by and large without encouragement from central government, which instead has concentrated on an ambitious but mostly inappropriate programme of rural afforestation via centralized nurseries of eucalyptus seedlings. Tree planting, then, provides clear evidence of local people as responsible and dynamic resource managers.

Preservation of trees in fields (rather than woodlands) and tree planting in homesteads represent individual resource management strategies over largely private resources. The situation with regard to the management of the communal grazing lands, from where most wood is obtained, is less clear. It is true that researchers have discovered many traditional resource management practices or rules. Some of these have already been noted: for example, the preservation of fruit trees in the woodlands or, with regard to fuelwood, the fact that acquisition occurs largely by gathering dead wood, or by using the wood left over from other uses (ie. construction, making handicrafts, fencing), thereby alleviating pressure on woodlands by avoiding unnecessary cutting. Another mechanism regulating tree cutting is the belief in the sacredness of certain trees (which are prohibited from being cut) and, more generally, the connection between trees and Shona culture. For example, Grundy (1990) reports respondents naming 18 species of tree as sacred in her research area, while Campbell et al (1991a) found that 64 per cent of their survey households stated that trees were important for the spiritual wellbeing of their households. In some areas, woodland blocks have been set aside (often around burial sites) where all tree cutting is forbidden (du Toit et al 1984). Finally, when tree cutting does take place, it should be spaced adequately (Grundy 1990).

Other communal rules concern specific resources. To regulate fruit use, and hence reduce damage to the tree of early or indiscriminate harvest, many wild fruits are allowed to be consumed by the individual, but not sold. Communities also try to regulate hunting and fishing by restricting hunting periods, banning the killing of young animals, banning the use of (tree-derived) toxins in fishing, and banning post-drought fishing to allow stock recovery (Wilson 1989b, Cavendish, forthcoming). Harvesting techniques which minimize damage to resources also exist for rope from branches (Grundy et al 1993), for the wild vegetable derere and in particular for wild medicines, with bark taken from alternate sides of a tree, roots removed in small quantities and the remaining roots covered over, and overlap of plant collection areas avoided by harvesters (Campbell et al 1993).

However, many of these rules derive from a time of resource abundance. This situation is changing as a result of past government policy, continued land clearance for agriculture, rising demands and a confusion of authority over resource management in the rural areas. With rising resource scarcities, many of the rules documented above are under pressure. Tree cutting for fuel is now more common; sacred trees are now cut; valuable large trees are being cut down completely; preserved woodland blocks are shrinking or disappearing; communities have failed to stop the commercialization of the wild fruit trade; hunting regulations are ignored; the stealing of wood resources has been documented; and the pressure of urban demand appears to have led to the diminution of traditional medicine-gathering practices (du Toit et al 1984, Wilson 1990, Campbell and Grundy 1991, Matose 1992, Campbell et al 1993).

In response, however, new resource management strategies are developing. First, communities themselves are instituting new rules concerning the use of valued resources. A common example of this is the institution of fines for illegal cutting of trees in the grazing areas and of a management systems whereby permission to cut must be sought from the headman or VIDCO (Wilson 1990, Fortmann and Nhira 1992, Cavendish, forthcoming). Second, resource privatization is occurring apace, particularly over resources in fields. Whereas in the past these were regarded as community resources, trees, tree products, grasses from contour ridges and other borders and other such resources are increasingly regarded as private property, and permission must be sought from the field 'owner' before using a resource. Resource privatization is also occurring through the creeping annexation of communal land and woodlands (Scoones and Wilson 1989). Finally, resource domestication has been observed, with Wilson (1989b) noting the domestication by women of three 'wild' vegetables that were important to diets.

In conclusion, this discussion of peasant knowledge and resource management does not pretend that a rural utopia exists. Problems in resource management exist and in some cases are serious. However, this section contradicts the common view of the ecologically-ignorant and environmentally degrading peasant. Instead, Zimbabwean Communal Area farmers emerge as aware of their surroundings and responding to various resource challenges. Another quote from Wilson (1987) neatly captures this view:

... the studies of deforestation were really an experience of how locals did not see the process as the simple, rather hopeless, march of population pressure.... They saw it as a phenomenally complex pattern of distinct ecological processes with many different specific causal factors, and one that was weaving new, often unfortunate, patterns, but not a situation that should be abandoned. On the contrary, they have the technical ability to manipulate it. [emphasis as original]

Continue

Contents - Previous - Next