The Study of Land Potential: an Open-Ended Inquiry

The study of land potential: an open-ended inquiry

$1. Latham and B. Denis =Centre ORSTOM Soumea yew Caledonia

In the initial design of the Unesco/UNFPA The 'talasiga formation - the most extreme Project in Fiji, one of the stated objectives of a range of pyrophytic formations supported was scientific description of the agrological by very degraded soils - was either formed or qualities of the land resources, together greatly enlarged as a result of this inter- with an estimate of their capacity to with- ference. Soil erosion, and the creation of stand and support cultivation. Added to this large areas of colluvium, may also be presumed

+ was an emphasis on the dynamics of change in to have arisen from, or been accelerated by, the soil-vegetation complex, viewed as the the interference of man with natural environ- expression of interaction between variable mental process2. environmental conditions and the impact of In so far as the team was concerned with - changing human occupance. Following the evaluating the suitability of land for agri- conclusion of a mutually welcomed agree- cultural use, which was its pragmatic object- - ment between Unesco and ORSTOM, and ap- ive, the context had to be a search for the proval of this agreement by the Govern- optimal production capacity of these environment of Fiji, the present authors were char- ments, consistent with preservation or amelio- - ged with primary-responsibility for this ration of ecosystems that are already in a . task, in collaboration with other members very unstable condition. It was decided to directly recruited by the project. Fieid base the work on the task of classification - work was carried out on five islands (Lakeba, using the 'system for evaluation of soils' Taveuni, Nairai, Batiki and Kabara) among proposed by FAO (1976). Three stages of which four were also studied from other work are therefore. involved: points of view. In terms of their ecology, .- definition of ecological units; none of these islands was previously well- - evaluation of the agrological qualities described. The major source on the soil and of land; \%:etation of Fiji (Twyford and Wright 1965) - evaluation of land-use potential, taking 2s concerned principally with the two main account of the consequences which any par- islands - Viti Levu and Vanua Levu - and ticular form of use might have on the en- discussion of the smaller islands is based vironment. only on reconnaissance. The islands chosen for study in the present project were selec- - ted on the basis of their representativeness h terms of the range of ecological conditions encountered in the eastern part of Fiji. 'Based on work still in progress at the University of %ceover, all had been heavily affected by Auckland, and by this project. , human occupation, and hence were particular- 'In Fijian, the term 'talasiga' means 'sun-burnt land' ly appropriate for study in a pilot project (Parham 1972). The term designates both a pyrophytic Hthin the Man and the Biosphere Programe vegetation complex dominated by ferns, often in assoc- Of Unesco. On present evidence, i; would iation with Cas&na equisetifoolia, and also the very appear that these islands were first popu- degraded and often eroded soils which support the com- lated by about 3000 years ago1. Such plex. It is used in what follows as a generic term man for a soil-vegetation complex which, while not limited occupation must be presumed to have initia- to the eastern islands of Fijl, is particularly-well ted a range of environmental transformations, exemplified on some of these islands, outstandingly by clearing of the forest and use of fire. Lakeba.

IJnesco, 1980. Papulation-enviromennent relations in tropical islands: the case ai eashm? Fiji. (MAB Technical Notes 13).

-- --I , The emphasis given to each of these stages to the purposes of the project, which was varied according to the practical possibili- designed as a pilot study in this area. '&e ties of access to islands, to knowledge of approach adopted is represented diagramatrc- the environmental conditions as a whole, and ally in Figure 1.

f ìEOMORPHOLOGY PEDOLOGY VEGETATION SOCIO- ECONOMIC CONDITIONS Definition of land- Study of soils and Description of lnvestmegt Crops grown or forms and their capable of being evqlu tion their characteristics vegetation types possibilities grown

I/____ Land use ecology . \ DEFINITION OF ECOLOGICAL UNITS Environmental - consequences of t land use I I

I r hrphodynamic constraints Edaphic constraints . Techn ícal Cropping - erosion -water / possibilities for possibilities - sedimentation - minerals terrain management

I l I r AGROLOGICAL QUALITIES Chances of success Types of OF ECOLOGICAL UNITS of land use types land use

Figure 1. FZm chart representing the method employed for the sf&y of land potential

I. THE DEFINITION OF ECOLOGICAL UNITS

The term 'ecological unit' is a compromise term 'ecological unit' seems more appropriate. between the ' land unit ' of most land-evalua- Even so, the task of defining such units de- i tion studies and the 'ecosystem' favoured by mands a large input of new research, since i the Man and the Biosphere Programme. Many of these island-environments are little-known g the ecological units would in fact be con- scientifically, in view of their isolation .i*- gruent with natural ecosystems, being land and small size. Moreover, the environments > units of relatively uniform environmental of the different islands are very varied, and? conditions capable of separate analysis in - many elements are unique to particular islandgk terms of their life-support systems, but to Within-island variation is often as marked as-. speak of ecosystems involves questions of between-island variation. A short descriptict?? bounding and the import and export of energy of each islan< will introduce the disucssion i- and matter, which are not faced in this work. which follows. That discussion is based pri-; For an essentially reconnaissance study, the marily on Lakeba, the most complex of the - _. c *

fie study of land potential

i. islands studied from aq ecological viewpoint, KABARA but draws also on condítions elsewhere in the archipelago. Kabara is representative of a group of uplift- ed atolls, and occupies 52 km2. In the north- west, however, is an outcrop of basalt of (LAKEBa Pleistocene age. Three principal environments Although only 56 km2 in area, Lakeba is an may be distinguished. Most of the island is island of quite unusual ecological complex- formed of a limestone basin, sloping inward ity. Basically, the island constitutes an from the encircling ridge toward the centre, andesitic massif of Miocene age, partly and containing some pockets of bauxitic soils covered by Pliocene limestones, of which (humic f erralsols) ; this whole area is covered only fragments remain. In the present era, in dense forest. Second is the small volcanic it experiences an oceanic tropical climate, hill which carries the principal cultivated t* with a mean rainfall at coastal sites of area, but where soils (eutric cambisols) are around 2000 "/yr. A succession of somewhat badly eroded. Third are the coastal plain different palaeoclimates , together with areas of sand or rendzinas, which are mainly general uplift, has led to the formation of under coconuts. Kabara's environment is there- I a range of morphological units, including fore quite simple, but very different from

bauxitic plateau areas, eroded hills, large that of the other islands. \ colluvial zones in the valleys and lowlands, ! and alluvial coastal plains within which are TAVEUNI quite numerous valley and sub-coastal swamps. Awide range of soils has evolved on these Taveuni is much the largest of the islands morphological units, among which the most studied (264 kmz), and is not representative maturely developed are the 'talasiga' soils of any other island but itself. Geologically, (acric ferralsols and ferralic cambisols) , it is the most recent in formation. Major while the least mature are the soils of the eruptions have taken place within the past, plains (eutric fluyisols, rendzinas , histo- 3000 years, concluding perhaps within the $ást sols and humic gleysols). Great diversity 1500 years, and the whole volcanic mass is lof of vegetation is associated with these soil- upper Pleistocene and Holocene age. The cli- landform complexes: dense forest, reed thic- mate is unusually humid, mean annual rainfall ket, forb vegetation dominated by ferns (tal- exceeding 6000 mm in parts of the island. asiga vegetation) , and hydromorphic herbac- According to the age of the volcanic parent eous vegetation. Notwithstanding its small material, soils may be divided into an andic size, the range of ecological units encount- group and a ferralitic group with bauxitic , ered in Lakeba exceeds the range encountered tendencies. Except where cleared for coconuts on any of the other islands visited. or other cultivation, a dense. forest covers the whole island to the summits. Taveuni is &TI= AND NAIRA1 6 a young island, of fertile volcanic soils x characterized by a dense rain forest with These also are small volcanic islands, but of little or no evidence of degradation. hasaltic composition and Pliocene age. Batiki, the smallest island :tudied by the project, SUMMARY Covers only 9 km2, but Nairai covers 28 km2. - Both islands are hilly, and eutric cambisols, The environments of these islands are varied, ferralic cambisols and a few areas of rhodic and contain distinctive elements. Because of ferralsols are developed on the hills. Rend- its diversity, work on Lakeba alone occupied zinas are found on the coastal'plains. Reed half the time alloted to the ecological sur- thicket dominates the slopes, but the small very in the field. Maps have been prepared areas of plain, the valleyZbottoms and the for Lakeba at a scale of 1:25 O00 and of parts summit area of Nairai carry forestvegetation. of Taveuni at 1:50 000; sketch maps only were These two islands therefore constitute a prepared for Kabara, Nairai and Batiki, at Younger and less diversified environment than the 1:50 O00 scale. Table 1 summarizes the Lakeba. They differ from Lakeba also in the main environmental characteristics of each - relative importance of the ecological com- island. plexes. Eutric cambisols covered in reed The three essential elements in the defini- thicket dominate these islands very much as tion of ecological units are inquiries in geo- Lakeba is dominated by the talasiga format- morphology, soil science and the nature of the ion. biota. They are all indispensable to an under-

115 standing of environment, its natural fertil- tate international comparison, the FAO /Unesco ity and its dynamism. For each there are (1972) system was used for classification of problems of classification, but especially soils, and the Unesco (1973) system for vege- of soils and vegetation. In order to facili- tation.

1 Table 1. Ecological characteristics of the ?ûlands studied

GEOLOGY GEONORPHOLOCT SOILS VEGETATIOS

i LAKEBA 56 Andesites Bauxitic plateaux Acric ferralsols (?fiocene) Eroded slopes Humic and rhodic .. ferraisols Ferralic cambisols Eutric cambisols Colluvial Chromicf luvisols Forb vegetation zones Alluvial plains Eutric 'fluvisols Coconuts and food crops 5 and coastal Rendzinas Coconuts and food crops plains Gleysols and histosols Limes tone Karst landforms Humic ferralsols Lithosols

VAIRAI 28 Basalt Eroded slopes Humic ferralsols Dense forest (Pliocene) Ferralic qambisols Forb vegetation -- Eutric canibisols Scrub with reeds Coastal plain Rendzinas Coconuts and cultivation

(Pliocene) Alluvial and Rendzinas and Coconuts and coastal plains f luvisols cultivation

.I CABARA 52 Limestone Bauxitic plateaux' Humic ferralsols Dense forest -I Lithosols Dense forest Basalt Eroded slopes Eutric and chromic Cultivatlon (Pleistocene) cambisols Cultivation

Coastal plains Rendzinas Coconuts i

:AVEUNI 264 Basalt Cones, slopes and Humic ferralsols Dense forest and cultivation L low-angle flows Ferralic cambisols Dense forest and cultivation Humic andosols Dense forest and cultivation.I

EVALUATION OF THE -AGROLOGICAL CJURACTERISTICS OF ECOLOGICAL UNITS II. k- Having defined the ecological units, the next THE EDAPHIC CONSTRAINTS .> step in an applied study such as the present is to determine more precisely their agrolog- Without experimentation, or data from experi- Ai: ..a ical characteristics. Study of three princi- ments in closely comparable conditions else- - pal elements is involved: where, approach to the study of these con- ; - the edaphic constraints of the soil; straints has to follow,a process of logical 7 - the constraints of morphodynamic processes; synthesis based on observation. The princi- - the possibilities of rational management. pal characteristic of ehe soil being deter- 5'

116 The study of Zdpotentia2

mined, observation of the vegetation and of work in similar locations by Quantin (1975), existing agricultural practice is then com- and Colmet-Daage and Lagache (1965). , pared against these characteristics and evaluated. The biological activity of the soil Evidence of the biota md of was also tested, as an important measure of existing cuZtivation the quality of the soil-plant interface. In these islands, three pedological characteris- The above evaluation of chemical fertility is tics were deemed particularly significant in in large measure substantiated by observation of the natural vegetation and of present land evaluating the edaphic-- constraints of the soil: use. Both the density and the floristic com- - depth of the natural root zone; position of the natural vegetation were taken 1 - availability of water to the plants; into account.. Es timation of vegetation den- - the chemical fertility of the soil. sity makes possible an approximation of the biomass, and by inference of productivity. It Depth of the root zone. This is a basic needs, however, to be complemented by floris- variable in soil quality. The depth of the tic analysis of burned-over areas. Such flor- natural root zone determines the water and istic analysis often makes possible linkage nutrition supplies available to cultivated between a recently-burned assemblage and an plants. In these islands, shallow eroded un-burned assemblage to which it corresponds, , soils are abundant, and this circumstance and facilitates' correction of estimates con- places strict limitations on the possibility cerning the productivity of the land. of management. In all soil profiles this Observation of land use, and the informat- variable was observed with particular care. ion provided by cultivators are very valuable sources of information, even though subjective. AvaiZabiZitzj of water. The supply of water The project undertook field measurements of to plants is likewise a major determinant of taro yield (Haynes 1976) and also estimates vegetation growth. Even in these relatively of coconut yields from secondary data (Unesco/ humid climates, with mean annual rainfall UNFPA 1977; H.C. Brookfield, personal generally in excess of 2000 mm, drought is communication). These measures are not easy often a limiting factor of considerable-im- to interpret in view of uncertainties regard- portance. A major part of the rainfall ing cultivation methods, use of different var- occurs during heavy and drenching showers, ieties in the case of taro, and uncertain age and flows some distance across the surface of trees in the case of coconuts. They offer, - before being absorbed. Moreover, the dry however, the only available objective compari- season is often three or four months in sons of agricultural crop productivity between length, and long dry periods occur in the different parcels of land. rainfall records. Plants may be supplied A further method of judging the productiv- with water by concentration of sheet flow as ity of soils was therefore sought. The method vel1 as from water stored in the soil. The employed was to evaluate the biological activ- water reserve held in the soil is thus of ity of soils through measuremeat of the re- =for importance in drained situations, and lease of CO2 (Bachelier 1968). This method was estimated using the formula developed by is very sensitive to variations in soil hum- - hilaire (1961)l. Taking account of a mean idity, and comparison between sites is not evapotranspiration of around 100 "/month, possible except under equivalent humidity. it emerges that drought stress can become a . Measurements were therefore made in the . serious risk for plant life after only one months of February and March 1976, at the month without rain on the majority of these height of the rainy season when soil humidity soils (Table 2). Mineral deficiencies are is almost constant at field capacity. Very notoriously difficult to eyaluate in these environments, and their quantitative assess- ment demands experimental work such as has been carried out in the Cook Islands or the Q = Ch + l5- d (Ho - Hl)c Kingdom of Tonga (Widdowson and Blackmore O 10 1975; 1976). It would seem, however, that Where: Potassium deficiency is very likely to be a Q = potential water reserve in mm h = depth of the root zone in cm real constraint both on the ferralitic soils d = bulk density of the soil and the andosols. Moreover, it seems pro- z = depth bable that phosphorous assimilation would be Hg = humidity at field capacity a problem on the andosols, following the Hl - humidity at the wilting point. interesting results were obtained (Table 2) , were estimated principally from sedimentary corresponding well with other evidence. It deposition in swampy areas, especially on emerges that soil fertility in these islands Lakeba; only qualitative observations were generally exhibits a direct correlation with possible elsewhere.' The rate of sedimentat- organic matter content, the plant cover and ion could be measured through C14 dating of level of biological activity. In the absence samples obtained at different depths, and of data which could be extrapolated from erosion rates were then estimated from the 1 other situations, the project relied on area of the ,upstream catchment. This work these field and laboratory observations in was reported in gre2ter detail by M. Latham, its attempt to define a scale of soil fertil- P.J. Hughes and M. Brookfield in Brook- ity. field 1979, and also in the final report to be published by ORSTOM. EGosion rates of at l'lie morphodymie eonstraints least 100-1 25 t/h2/yr have been estimated e> -- both from forested and from talasiga catch- Erosion, sedimentation and the action of the ments. These rates are high, but are compar- - sea all have an important bearing on the able with those calculated from sites else- question of land capability. Erosion rates

Table 2. Comparison between ground cover, soil chcterkfics of the A horizon and bioZogica1 activity in Lakeba

ACRIC HUMIC FERRALIC CHROMIC EUTRIC HISTO-' HUMIC EUTRIC HUMIC FERRAL- FERRAL- CAMBISOLS LWISOLS FLWISOLS RENDZINAS .SOLS GLNSOLS CAMBI- CAELBI- SOLS SOLS .- SOLS SOLS

Ground cove1 Tala- Forest Talasiga Talasiga Coconuts Coconuts Wet Wet taro Reeds Foresi siga Taro

Mean cultivable depth (cm) 100 40-60 20-40 100 100 100 Hydromorphic 30-40 30-40

Potential water reserve (mm) 80-95 120-140 50-70 120 140-160 ( . Saturated '. ) i00 100

Physico-chemical characteristics (A1 horizon) : Texture SL A SL L AL S A A A A Structural instability O. 3 - 0.6 0.6 0.1 0.4 - - 0.4 0.4 Bulk density O. 65 - 0.70 0.87 0.83 0.86 0.52 0.57 0.82 0.80 Per cent carbon- 3.0 4.3 3.3 2.9 5.7 8.3 13.5 12.7 2.8 6.0 Per cent nitrogen 0.15 0.3 0.18 0.14 0.42 0.63 0.13 0.35 0.21 0.55 PH 5.3 5.6 4.9 5.3 6.5 7.1 4.4 5.6 5.8 6.0 Exch. Ca* (me/lOOg) 2.2 6.1 2.2 2.5 24.4 - 14.1 19.6 10.5 9.4 Exch. K+ (me/100g) O. 3 0.2 0.55 0.4 0.3 0.5 0.18 0.64 0.76 0.33 Cation exchange capacity (me/100g) 30.7 27.6 22.4 26.0 45.0 - 50 50 29.2 23.7 Base saturation (per cent) 16 48.9 29.1 40.4 73 - 50 50 76 75 Phosphorous (per thousand) 1.4 0.2 1.4 2.1 2.4 1.8 1.2 1.3- 2.1 0.7

Biological activity in mg of C02/m2/houc (wet season) 20-40 - 15- 80 35 184 120 240 235 108 138

118

.. .. - - y--- .-- -- . . .. ., ...... c .. ..

-- , ,

Pt &ere in the tropics (Trescases 1975; Roose far as it can contribute to extension of the 1975). Rapid erosion may also be observed cultivated area and the introduction of new qualitatively in the field, by the presence crops such as cereals, oil seeds, or improved of a number of gullies on the hillsides and pasture; slope remains a major constraint. .r the numerous examples of deposition of weath- The use of manure and fertilizers is certain- ered andesitic sands on the lower slopes. ly capable of more universal expansion. Both Material eroded from the slopes is distri- mineral fertilization and the use of green buted along the valley floors. Carried in manure would permit an increase in yields suspension in the streams, finer fractions and also would assist recuperation of infer- are deposited in the swamps and on flood tile soils. Mineral fertilization is present- plains. Other sediments result from sheet ly practised on a very small scale on Taveuni erosion on the slopes and are deposited on and Lakeba, but extension of these practices land of lesser gradient to increase the depth would demand both a programe of agricultural of the colluvial formations. Some of these education, and a large investment. I' deposits are beneficial to soils Situated in Very little has thus far been done to man- the lower part of forested catchments, or age the morphodynamic processes on these is- catchments with other upland soils of high lands. Work is essentially confined to an- fertility, and it is probably this supply of ciently-established drainage on Lakeba , and mineral fertility that explains the high[, to erosion control on the volcanic hill in value of the valley swamps of Lakeba, which Kabara. Modern drainage in Lakeba is con- have been regularly cultivated under taro for fined mainly to inland valley sites where the many generations. However, colluvial deposits slope is greater and depredation by livestock are less of ten valuable, being mainly derived less of a problem. Only a few of the former- from upslope talasiga soils of very low fer- ly-drained sub-coastal swamps are still in tility. production. It would seem that to bring the The effect of the sea on the coastal areas larger of these swamps back into use would is discussed elsewhere in this collection by now require major works. With the exception McLean, and is-not further developed here. of limited patches of wet land on Taveuni, none of which are still in production, and [PechnicaZ possibizities for small areas of Nairai and Batiki, the prob-

*errain management ~ - lems of drainage and irrigation are presently specific to Lakeba among the islands studied. The constraints discussed above are capable However, there are much larger areas of of some reduction by proper management. Such swamp on the main islands, some formerly man- control measures can only be on a modest scale aged and others never managed, especially in &I the short term, given-the level of invest- the Rewa delta near Suva. Zent for rural development proposed in the Erosion, on the other hand, is widespread current five-year plan for Fiji (Central Plan- throughout the eastern island region, yet the ning Office 1975), but a certain minimum level only example of effective anti-erosion meas- of investment is necessary if an increase in ures seen is on Kabara. On this raised atoll, T agricultural production is desired. Such most cultivation is concentrated on the slopes i; investments are possible both in improving and piedmont of the one volcanic hill. This - the edaphic qualities of the soils and in hill has been heavily cultivated for a very bringing the morphodynamic processes under long period, leading to severe erosion and control. truncation of the soil. The upper slopes are now partly circled by contour-line hedges of &ixrovement of soil quazity. Improvements Cyperaceae which check run-off and so reduce 2re possible both in the manner of working erosion. Similar practices need to be intro- the soils, and also in soil enrichment by duced throughout all the other islands of the fertilization. With only limited exceptions, archipelago, wherever annual food-crop cultiv- the soils of this region are presently worked ation is carried out on steep slopes. Present I aanually. Tractor ploughing has been intro- land management is of a very minor, order, and duced in Taveuni, and initiated in Lakeba, needs to be greatly extended in order to per- and considerable expansion is possible. How- mit the better utilization of the soil re- ever, mechanization is significant only in so sources of these islands.

1 1Y 'i - M. Lat1iarsz and B. enis is

III. DEFINITION OF LAND POTENTIAL

The definition .of land potential is an essen- severe degradation. This was especially the tial basis for rational agricultural planning. case in those areas now occupied by pyrophi- In order to be effective, the capabilities tic vegetation complexes such as the talasiga of each ecological unit need to be evaluated and reed-thicket. More recent examples in- in terms of the type of agriculture present- clude the almost uncontrolled introduction of ly employed, or of alternative types which weeds and plant disease in the late 19th " could easily be adopted. It is also necess- century, and moderg clearance of the forests. ary to estimate the impact of these land No new introduction should ever be proposed uses on the environment. without evaluation of its environmental impact , whether by comparison with known cases Evaluation of zand potential elsewhere, or by attempts at measurement on site. During the course of the project, For each ecological unit, there exists a three types of land use were examined from range of possible land uses compatible this point of view: with the edaphic qualities of the soil and - changes in soil under semi-continuous taro the morphodynamic constraints of the land. cultivation; Observation of existing cultivation practices - environmental changes under Pinus c&baea provides a useful guide, but there are also plantation; other cropping systems found elsewhere in - the effect of shifting cultivation on the comparable ecological conditions, which are environment. either not practised, or scarcely at all practised in present-day eastern Fiji. These , Soils under taro. Soils which had been under include cereal cultivation, oil seeds, coffee almost continuous cultivation for up to ten and oil palm cultivation. Such crops could years on Taveuni, and far longer on Lakeba, be established on certain islands to provide were examined by Denis; brief results were more diverse sources of income, but their presented in the project's first general development would generally require a com- report (Unesco/UNF'PA 1977) and will be plete restructuring o€ the rural landscape expanded in the final report to be published and economy. Such remodelling affects prin- by ORSTOM. The Taveuni soils were andosols cipally the dominant position presently occu- and humic ferralsols; in Lakeba the sites pied by the coconut palm. Estate and small- were on gleysols and histosols, under irri- < holder groves of coconuts presently cover gation and drainage. The object was to de- much the larger part of the best land on all termine whether or not 'the absence of a fallow the islands. The trees are often old, badly period had provoked soil degradation. Two maintained, and of low yield. In cash terms, pairs of sites were examined on each island; they offer a low rate of return per hectare, in each pair one plot had been continuously and their presence both inhibits the intro- cultivated while the other was under very old duction of more rewarding crops and constitut- regrowth. Physico-chemical characteristics es a brake slowing down moves toward diver- (porosity, structure, pF, organic matter, pH, sification (Unesco/UNFPA 1977). exchangeable complex, phosphorous) were ex- The resultant scheme is summarized in amined together with the biological activity Table 3, where seven land-use systems are of the soil, using the method described above. recognized, cross-classified within each eco- On the richer soils no significant differences logical unit by four levels of capability: good, average, mediocre and poor. The object tween the cultivated and uncultivated soils, is to permit the planner to make a rational a result which implies that intensification choice among the range of possibilities. of cultivation is entirely feasible in these environments. However, this conclusion can- Consequences of Zand-use systems not be extended to be applied to the less for the environment stable soils; intensification on the latter would require fertilization and green manuring Changes in the system of land use can often, however, lead to very unfavourable secondary Soils under Pinus caribaea. A similar study -i effects on fragile natural environments. It was carried out to test the effect of Pinus $ has been remarked above that clearing and the caribaea lplantations, which are becoming wide- use of fire, which followed the arrival of spread on talasiga and other degraded soils f man in these islands, led in some islands to in Fiji. The general view is that the*plant- ?,

! 120

u .. ..

L . .. --

I- t i fie skdy of laml potenticll

I 1 Table 3, An earmrple of kmd ctassification for agriculture and fores-: the case of Laakeba i POSSIBLE OF LAND USE GOLOGICAL UNITS CLASSIFIED TYPES fl AGROLOGICAL QUALIm FIELD CROPS IRRIGATED SUBSISTENCE LOW-DEMAND IMPROVED TREE FOREST MECHANIZED CROPS CROPS CROPS, LONG PASTURE CROPS PLANTATION - FALLOW

HIGH QUALITY, little erosion: -Alluvial and coastal plains Eutric fluvisols, now coconuts tt-t -- - Swamps, Humic Gleysols, now grass fallow and taro -- +I+

HEDIUM QUALITY, little erosion: Coastal plains, Rekdzinas -- -- now coconuts 'i Swamps, Histosols and Gleysols now under swamp vegetation -- si-

MEDIOCRE QUALITY, liable to erosion: - Colluvial/alluvial terraces, Chromic Luvisols, now talasiga si- -- , +t + +

MEDIOCRE QUALITY, serious liability to erosion: - Steep slopes, Humic Cambisols, now forested -- si- .si- + - Lower slopes, Eutric Cambisols, now reed thickgt -- -- si- +

VERY MEDIOCRE QUALITY, little erosion: - Plateaux, Acric Ferralsols. now talasiga -- -I. +

VERY MEDIOCRE QUALITY, serious liability to erosion: 1;-Hills and slopes, Ferralic Cambisols and Rhodic Ferralsols, now talasiga -- -- +t

LOW QUALITY. serious liability to erosion: - Very steep slopes, Humic Cambisols, now forest -- --

Key to symbols: +i-+highly suitable - ft suitable + low chance of success -- land use type not suitable

ing of resinous trees degrades the environ- This study, carried out on plantations ment, and several studies in Eyrope have from five to 15 years in age, yielded sig- demonstated the role of resinous trees in nificant evidence of net amelioration of producing podzolization. As in the taro the soil in the planted areas; moreover, study, paired samples were measured, in each there was also an enrichment of the spontan- case one site being selected from a planted eous ground flora. Positive consequences of area, the other from a nearby area under its pine planting on soil permeability have 'natural' cover. In addition to the physico- earlier been demonstrated by Bayliss-Smith chemical and CO measurements described above, (1977). It perhaps follows that the establish- 2 observations were also made of the spontaneous ment of a continuous ground cover under Pinus ground flora. c&baza is beneficial both to the soil and L- M. Lattiain and B. Denis

to the hydrological regime. At this stage, range of field studies was therefore under- no significant tendency toward degradation taken, calling not only on pedology, but al- of soil under the resinous trees could be so on geomorphology, sedimentology, agronomy discerned. and botany. Limitations of competence were encountered, for while project members were The effect of shifting cuZtivation. In view selected for their interdisciplinary interests of the widespread development of impermanent l the primary expertise of the ecological team field plots on steep slopes, it seemed also was in the field of pedology, and a full suite necessary to evaluate the effect of forest of specialists could pot be made available to clearance for this type of agriculture. Shif- advise the team in the field. Existing docu- ting cult,ivation is probably of ancient ori- mentation, in particular unpublished material gin 2n these islands, but there has been provided by the geologists of the Mineral Re- considerable recent extension due to the new sources Division, the pedological studies of marketing opportunities for taro and yaqona. Twyford and Wright (1965) , the botanical Observation suggested that loss of soil from studies of Parham (1972) and a good aerial the slopes was a more significant consequence photography coverage from high, middle and of these developments than degradation of the low altitudes for Lakeba, and middle and high soil itself or of the vegetation. In Lakeba, altitudes for other islands, had to substitute it was possible to measure this erosion from as the base .of inquiry. 'Further information the sediment deposited in a swampy area of was obtained from the local professional ser- the base of a partly forested hill, which has vices - meteorology, agriculture and forests long been cleared and cultivated (Waitabu) . - and also from unskilled but well-informed This was not a completely satisfactory method, guides in the field. The larger part of the and sediment traps would have been preferable. botanical work, in particular, was achieved - Not only is it impossible to separate sedi- by relating vernacular names obtained in the ment derived from the gardened areas and the field to the data provided by Parham (1972) uncultivated slopes, but the amount of eros- and specimens in the Suva herbarium. ion is also under-estimated, since quantities Inevitably, more questions were posed than- of material in suspension are carried further were answered. The problems raised in the downstream in floods. None the less, some course of the ecological survey demanded indication is possible, and on the basis of solution through research in geomorphology, preliminary results we estimate erosion in a botany and palaeobotany: in particular, these partially-cleared basin at 1.24 t/ha/yr , or problems required the establishment of a 8.3 cm/lOOO yr. If this estimate is compared quaternary chronology for the Fijian region., with the mean rate of deepening of tropical an explanation of the evolution of bauxitic soils by weathering, estimated at between 0.5 soils in Lakeba, and of the origin of the and 5 cm/lOOO yr (Leneuf 1959; Hervieu 1968; talasiga formation. Partial answers could be Trescases 1975), we would seem to have only suggested, but each of these questions de- a moderate rate of soil truncation. Any in- mands a whole research project to itself alone. crease in the clearing of land for cultivat- Even the edaphic study of soils had to rest ion would, however, accelerate this rate of at a level of unsatisfactory generality. I soil loss, as seems to have happened on the Foliar analysis would be necessary in order volcanic hill in Kabara. to detect deficiencies. Pot trialk would al- The three examples cited above demonstrate so have been desirable. For all this the pro- the value of measurement of environmental ject would have required more specialists, change under human use of the land; simple more time, and more money. I application of conventional wisdom is not Finally, study of the all-important impact sufficient. However, it was not possible to of man on environment has to remain at an in- ! go beyond the reconnaissance work described dicative rather than conclusive level. The above in view of limitations of time and re- survey could do no more than estimate the sources in this pilot project. effect of forest clearance on slopes by shifting cultivation. Agronomic systems, and the CONCLUSION problem of fallow periods and their success- ional vegetation had to be discussed without The study of the soil-vegetation complex with- full analysis. Working in an untouched area, in the Fiji project was designed to explore the project could do no more with its res0urce-C the dynamics of change, as well as to establish the basis for an evaluation of land re- (Editorial footnote) The principal author of the pre- sources in terms of their rational use: a sent paper in particular was selected for this reason. t i 122

( . -- !%e study of land potential

:han outline the dimensions of answers to its ies, who provided support both in field and questions. However, it has been possible to in Suva, and gave invaluable aid with labora- shed light on a number of aspects of the is- tory analysis of samples. They also wish to land environment, its natural qualities and thank the Chief Technical Adviser for his also its vulnerability. enthusiasm. and constant support. Finally, they gratefully acknowledge the help of A. ACKNOWLEDGEMENTS Combeau and P. Quantin, who read earlier drafts of this paper and offered valuable ne authors are particularly grateful to all critiques. Translation into English was chose who aided their missions in Fiji, es- carried out by H.C. Brookfield, who suggested pecially Unesco, the Ministry of Fijian a number of amendments in order to relate Affairs, who provided transport, and the the paper more closely to others in thgs !histry of Agriculture, Forests and Fisher- collection.

> BIBLIOGRAPHY

BACHELIER, G. 1968. Problèmes relatifs 1 l'atmosphère LENEUF, N. 1959. L'altdration des granites calco- du sol et utilisat'.on possible d'un détecteur de akazins en Côte d'Ivoire. Thèse ORSTOM, Paris. gaz pour la mesurebe sa teneur en gaz carbonique. PARHAM, J.W. 1972. Plants of the Fiji islands. Cah. ORSTOM, sdr. pécbl., 6 (I), p. 95-104. Government Printer, Suva. BAYJXSS-SMITH, T.P. 1977. Hurricane Val in north QUANTIN, P. 1975. Arehip,eZ des #ouveZles-B6brides: Lakeba: the view fsom 1975. In: The hurricane sols et quelques donndes du milieu naturel (Ambrym, hazard: natural di@ster and small populations. Aoba, Mamo, Pentecôte). ORSTOM, Paris. H.C. Brookfield (Ed,), p. 65-97. Unesco/UNFPA ROOSE, E. 1975. Erosion et ruisse&ment en Afrique de Island Reports 1. Australian National University l'Ouest: vingt anndes de memes en petites parcelles for Unesco, Canberra. expdmhentates. ORSTOM, Abidjan. (Mimeographed BROOKFIELD, H.C. {Ed.). 1979. kkeba: Environmental ' document). change, population dynamics and resource use. TRESCASES, J.J. 1975. L'évolution géochimique supergène Unesco/UNFPA Island Reports 5. AustralianNational des roches ultrabasiques en zone tropicale et la University for Unesco, Canberra. foxmation des gisements nickélifères en Nouvelle- CENTRAL PLANNING OFFICE. 1975. Fiji's seventh develop- Calédonie. Mém. ORSTOM No. 78. ORSTOM, Paris. ment plan, 1976-80. Central Planning Office, Suva. TWYFORD, J.T.; WRIGHT, A.C.S. 1965. The soi2 resmes COLMET-DAAGE, F.; LAGACHE, P. 1965. Caractéristiques of the Fiji islands. Government Printer, Suva. de quelques groupes de sols dérivés de roches vol- , UNESCO. 1973. Internationat ckssification and mapping caniques aux Antilles Françaises. Cah. ORSTOM sdr. of vegetation. Ecology and Conservation 6. pddol., 3 (Z), p. 91-122. Unesco, Paris. fAQ. 1976. A framework for land evaluation. Soils * UNESCO/UNFPA. 1977. Popuktion, resources and develop- hlletin No. 32. FAO, Rome. ment in the eastern islands of Fiji: information fop EAO/UNESCO. 1972. Soil map of the world. Legend, Vol. 1. deeiswn-making. Unesco/UNFPA General Report No. 1. Unesco, Paris. Australiatl National University for Unesco, Canberra. %LAIRE, M. 1961. Irrigation et utilisation des WIDDOWSON, J.P.; BLACKMORE, L.C. 1975. Fertility of réserves naturelles. Ann. agronomiques, 12, p. 87-97. Cook Islands soils - interhi report. N.Z. Soil Bureau, UMS, P. 1976. Some aspects of agmhzkure in Taueuni Wellington. (Mimeographed document). and kkeba. Unesco/UNFPA Working Paper No. 4. WIDDOWSON, J.P.; BLACKMORE, L.C. 1976. Fertility of the Australian National University for Unesco, Canberra. soils of Tonga - interim report. N.Z. Soil Bureau, "EU, J. 1968. Contribution 1 l'étude de l'alluvion- Wellington. (Mimeographed document). nement en milieu tropical. Mh. ORSTOM No. 24. ORSTOM, Paris. opulation-environment relations in tropical4*4 islands: 4 the case of eastern Fiji

Based on the findings of the Unesco/UNFPA pilot project ‘Studies on population- environment relationships in the eastern islands of Fiji’ f

Edited by H.C. Brookfield c