III.  BACKGROUND DOCUMENTS

1.  THE LIMITS OF LIMNOLOGICAL THEORY AND APPROACHES AS APPLIED TO RIVER-FLOODPLAIN SYSTEMS AND THEIR FISH PRODUCTION (by P.B. Bayley)

INTRODUCTION

This paper attempts to provoke discussion on fundamental concepts of limnology insofar as they differ between true lakes, rivers and river-floodplain systems.

Traditionally, limnology has been concerned with true lakes which have limited water level fluctuations. Studies have been further restricted by emphasis on the pelagic subsystem, which has been relatively easier to sample and model. This bias has to some extent been compensated by Wetzel (1975 and references therein) who has maintained that littoral production plays the major part in organic input in most lakes.

His text is taken as a starting point in the following discussion, which is also influenced by the author's experience in the Amazon basin and a large African lake.

RIVERS VERSUS RIVER-FLOODPLAIN SYSTEMS

Rivers have for some time been neglected, partly because of a desire by limnologists to study relatively closed systems. In addition, most studies have concentrated on temperate systems which have negligible floodplains or have had them eliminated.

Generally acceptable divisions between lentic and lotic systems have been possible in temperate regions. This is not so easy or desirable for river-floodplain systems. Many production processes - in particular, those concerning emergent macrophytes - take place to a significant extent in both sub-systems and life cycles of fish species comprising most of the biomass regularly utilize both environments during different parts of their life cycles.

There are also considerable abiotic interactions underlying these processes such as the hydrology and associated nutrient distribution. Their separation into lentic and lotic components does not necessarily lead to a better understanding of the functional aspects of river-floodplain systems.

Henceforth, the discussion deals with the latter as a single dynamic unity.

LAKES VERSUS RIVER-FLOODPLAIN SYSTEMS

The closest limnological approaches which may assist in the understanding of river-floodplain production processes are those concerning the littoral/eulittoral zones in lakes. Wetzel (1975) has maintained that in most lakes, organic and nutrient input resulting from emergent macrophyte production usually exceeds that from other sources put together. This is an understatement as far as river-floodplains are concerned.

In both systems, most nutrients appear to be obtained directly from the bottom substrate where large concentrations typically exist in the upper layers. Associated with the submerged parts of macrophytes are large quantities of periphyton and perizoon, and the considerable “biomass” of detritus derived mainly from the plants contains large numbers of detritivorous micro-organisms.

Many river-floodplain systems have significant forested areas which are regularly inundated and contribute significantly to production of larger individuals of some fish species. This biotope is not found in true lakes, but the mechanisms involved may be broadly similar to those concerning allochthonous production in heavily-forested streams. Despite the significant production of fruits and leaves in the flooded forest biotope, the author considers that the total organic production per unit area is much less than in the open floodplain areas dominated by emergent macrophytes.

The more important quantitative differences between lake and river-floodplain production are included in the following discussion.

FUNCTIONAL DIFFERENCES BETWEEN SYSTEMS AND EFFECTS ON RIVER-FLOODPLAIN PRODUCTIVITY

The mechanisms involved in the production processes in the different systems considered are probably qualitatively the same or very similar, but they express themselves differently as outlined below:

(a)   Different sub-sets of mechanisms play a dominant role in each system

For an extreme example, in a large lake phytoplankton production may be more important than “littoral production”, but in floodplains reaching a similar or larger size the latter is certainly dominant. This is generally more due to a high rate of turbid water input typical of large tropical rivers which restricts phytoplankton production, rather than due to differences in dissolved nutrient concentrations.

Similarly, benthic algal production would be expected to be much less important compared with either tropical or temperate lakes.

Direct input of dissolved nutrients from the rivers in river-floodplain systems is probably more important in “topping up” the system for long-term benefits rather than in controlling year-to-year production. Localized decanting of solids allows phytoplankton/zooplankton production to flourish, which may be important for young fish of some species, but in terms of total production it is apparently small.

Production of floating macrophytes deriving nutrients from the water column appears to be much more significant (e.g., Eichhornia, Pistia and floating stages of Papsalum repens; see Junk (1970)) than phytoplankton production. However, the author considers that this production is seemingly minor compared with that of emergent macrophytes rooted in the substrate. In any case it is probable that more nutrients are released from the ground when it is newly-flooded, rather than are derived directly from the water. The benefits from this will also depend on the degree of flooding. Terrestrial macrophytes which are annually flooded also contribute significantly to organic input to the aquatic environment (in particular Gramineae) as well as providing a substrate for food and shelter for young fishes.

Turnover of the macrophyte-detritus cycle in the river-floodplain is much faster than in the littoral zone of a lake because the stranding, drying (sometimes with burn-off) and reflooding allow aerobic breakdown and mineralization mechanisms to be more dominant. When “burn-off” or appropriate terrestrial herbivores are lacking, much of the decomposition occurs in the water, producing high BOD and limiting the distribution of fishes (University of Idaho et al., 1971).

Access to allochthonous sources of food is less consistent than in lakes or rivers but is considerable when the flood invades pasture, scrub or forest. Input of soluble organic compounds during this process may be more important also. Stream-transported allochthonous material may be broken down to more refactory substances before reaching lakes or larger rivers but is in any case of relatively minor importance in river floodplains.

The above considerations suggest that in many cases different sets of mechanisms are dominant in river floodplain systems.

(b) Dynamic effects on the mechanism sets responsible for production are quite distinct between river-floodplains and either lakes or rivers

Flood cycles (normally annual) in tropical river-floodplain systems have a more drastic effect on production processes than seasonal temperature effects in the temperate zone. The popular concept of high, year-round production in the tropics is completely misleading when applied to river-floodplain systems. Most fish species stop feeding completely during their migrations, which can often take more than half a year during the drawdown and low water seasons.

This “physiological winter” is more severe than in temperate lakes and rivers, where production processes and feeding continue well into the colder months, albeit at a slower rate. Even for those river-floodplain species which are not migrating for reproductive, distributive or protective strategies, their potential food sources are mostly on dry land during this period - with the obvious exception of top piscivores and parasites.

Good arguments have been put forward (Wetzel, 1975) that the emergent macrophyte-detritus cycle in the littoral zone adds stability to the lacustrine system. This is possible in both the short and long term because of the physical stability of the biotope as well as the nature of the mechanisms concerned. Short-term stability in the littoral zone in a river-floodplain system is not possible because of the constant displacement of this zone, resulting in production mechanisms which are most of the time much further from equilibrium levels. The flooding season produces the aquatic equivalent of a “rat race” in which plants and fish alike strive to keep up with a rapidly-expanding, food-rich environment in a consistent state of short-term disequilibrium. The situation is more similar to that of a newly-flooded reservoir than to a lake or river.

Long-term stability in the river-floodplain system is mostly dependent on the predictability of the flood regime. Annual maxima and minima, and rate of flooding can vary considerably in South American (pers. observ.) and African systems (Welcomme, 1975) affecting macrophyte and fish production significantly. It is also probable that subtle changes in year-to-year water movements affect spawning strategies differentially, causing changes in fish species biomass distributions.

Despite these destabilizing effects, it should be added that the river-floodplain system would be more unstable but for the vast reservoir of organic matter and nutrients stored in the detritus. To a certain extent this smooths out the effects of annual fluctuations in macrophyte production, although its availability from year-to-year to the dominant detritivorous fish species depends again on the degree of flooding.

CONCLUSIONS AND SUGGESTED APPROACHES

Any of the above effects on mechanism sub-sets can produce very different relationships between the density of a functional part (e.g., mineral concentration, dissolved/particulate carbon or biomass of a plant or animal) and its flux or production in river-floodplain systems and lakes or rivers. Other factors not considered here, such as effects on fish mortality rates and their vulnerability, also have different feedback effects on the subsequent fish productivity.

It is clear the production processes in river-floodplain systems are even more complex than in littoral zones of lakes, because of the former's more dynamic nature and lower predictability. Also they are as yet little understood, and many of the suppositions presented above have no hard evidence to substantiate them. This means that attempting to model the the system by building up sets of mechanisms studied separately is even more absurd and very expensive.

General holistic and comparative approaches, relying heavily on statistical methods, are a necessary if somewhat inelegant first step. Identification of significant driving and controlling forces, such as degree of flooding, has already been undertaken (University of Michigan et al., 1971; Kapetsky, 1974; Welcomme, 1975). Comparisons between and, with appropriate safeguards, within systems can assist in developing models (e.g., Welcomme and Hagborg, 1977) which can at least define probable limits.

More accurate and comprehensive resource evaluation, such as by remote sensing coupled with ground truth information, is essential. When combined with limnological data and indices of fish production (such as good catch and effort data), comparisons between systems will allow important variables to be identified.

Limnological sampling programmes should concentrate on understanding the variabilities of system components, as well as their total quantities during different stages of the flood cycle. Sampling programmes which are devised to test hypotheses concerning isolated mechanisms tend to be biased toward preconceived ideas and significant interactions can be overlooked when the list of measured variables is kept limited. The author does not wish to decry the laudable aims of testing hypothetical mechanisms but suggests that it is too soon to attempt so much detail in such complex systems, until we have more statistical guides as to which sets of mechanisms are most significant.

REFERENCES

Junk, W.J., 1970 Investigations on the ecology and production biology of the floating meadows (Paspalo Echinochloetum) on the Middle Amazon. Part 1. The floating vegetation and its ecology. Amazoniana, 2(4):449–95

Kapetsky, J.M., 1974 Growth, mortality and production of five fish species of the Kafue River floodplain, Zambia. Ph.D. Thesis, University of Michigan, 194 p.

University of Idaho et al., 1971 Ecology of fishes in the Kafue River. Report prepared for FAO (acting as executing agency for the UNDP). Moscow, Idaho, University of Idaho, FI:SF/ZAM.11, Tech. Rep., (2):66 p.

University of Michigan et al., 1971 The fisheries of the Kafue River flats, Zambia, in relation to the Kafue Gorge Dam. Report prepared for FAO (acting as executing agency for the UNDP). Ann Arbor, University of Michigan, FI:SF/ZAM.11, Tech.Rep., (1):161 p.

Welcomme, R.L., 1975 The Fisheries ecology of African floodplains. CIFA Tech.Pap., (3):51 p.

Welcomme, R.L. and D. Hagborg, 1977 Towards a model of floodplain fish populations and its fishery. Environ.Biol.Fish., 2(1):7–24

Wetzel, T.G., 1975 Limnology. Philadelphia, W.B. Saunders, 743 p.

2.  ASSESSMENT OF FISH STOCK AND PREDICTION OF CATCH IN LARGE RIVERS (by J. Holčik)

INTRODUCTION

A wide spectrum of methods for the assessment of fish populations and for the prediction of future catch have been described. Only few of these seem to be applicable in large rivers. The main reason is that river ecosystems can be considered unstable in comparison with standing waters ecosystems. They are also much more sensitive to factors of the surrounding environment including the activities of man. Furthermore, sampling procedures in riverine conditions have not been adequately developed in comparison with those in lakes. Because of this there is an almost complete lack of papers dealing with estimations of fish population and yield with the exception of the assessment of some migratory species like salmonids or sturgeons. The following list of methods should be considered rather as an arbitrary choice than a complete enumeration, and is made with regard to simplicity and rapidity in their practical application.

ASSESSMENT OF FISH STOCK

The velocity of the current, the nature of the bottom and the banks, obstacles to flow and the intensity of inland navigation may be considered the main factors determining the choice of methods suitable for the estimation of fish populations. The essential is to know the efficiency of the gear used with respect to the number of fish caught, and how closely the composition of the catch agrees with the composition of the stock.

Area catch methods

These methods depend on the assumption that the number of fish caught in a defined area is in relation to the whole area of the water body or to the area occupied by a given species. This method can be applied in different ways:

Estimation of fish stock from deposited eggs:   In riverine conditions the use of this method is limited to species laying their eggs on a substrate and cannot be applied to pelagophilic species. If the number of eggs on some spawning grounds, the mean number of eggs per female, and the proportion of sexes in the spawning shoal are known, the number of fish can be calculated as:

where N = number of fish in the spawning shoal, n_e = number of eggs found in area investigated, n = mean fecundity of female and s = proportion of sexes, and:

where = mean number of eggs sampled, a = sampling area and A = total area of the spawning ground. Though the method has been originally used for fish with pelagic eggs (Hensen and Apstein, 1897) and has been applied mainly in sea fisheries, it can also be used in fresh waters as described by Rothschild (1961) and Bastl (1977). A review of the theory of this method and its application in practice is given in papers of English (1964) and Saville (1964).

Estimation based on the fish catch data:   This method is widely used in sea fisheries (Nikolsky, 1965) and only rarely in lakes (Jefimova, 1967; Turner, 1975), but it seems to be suitable in riverine conditions too (Allen, 1951; Williams, 1965). The way of calculating the number or the biomass of fish is essentially the same as in equation (1), however the efficiency of the gear used should be incorporated:

where p = population estimated, n = number of fish sampled, K = efficiency of gear used, A and a = total area and area sampled respectively. Most difficulties arise in defining the constant K, which depends not only on the type of fishing gear but also on the kind of species, length of fish, physiological condition, time of day, weather and other factors.

Catch and effort methods:   The relative density of fish stock, and also indirectly its number, can be measured by an “index of effectiveness of exploitation”, as suggested by Leopold and Dabrowski (1975). This index consists of the total fish catch of a given water body, for example, a river or any particular reach, divided by the sum of the fishing intensity of all types of fishing gear. The fishing or exploitation intensity of a fishing gear is the product of its standard fishing effort multiplied by the number of days during which it is used: the standard unit of fishing effort is a conventional fishing gear, catching 1 kg/day of fish. As emphasized by authors, this index can be used not only to derive the catch per unit of area or intensity per unit of area but also for the proper evaluation of the state of a fish stock in a given water body. Though originally developed for, and used in, lakes (Bonar, 1977), this method seems to be quite reasonable for rivers too. Naturally, a range of other methods based on the catch and effort data have been described in Beverton and Holt (1957), Robson and Regier (1967), Gulland (1969) and Ricker (1975).

Seber-Le Cren two-catch method:   Where two successive catches c₁ and c₂ are taken with the same effort from a population the estimate of the size of the population is given by:

and the variance:

from which the standard error is:

According to the authors (Seber and Le Cren, 1967) the above method depends upon the following conditions:

 (i) that p (p = (c₁ - c₂)/c₁) is large enough to have a significant effect upon n,

(ii) that p is constant, i.e., that the fishing effort is the same for the two catches and the fish remaining after the first fishing are as vulnerable to capture as those that were caught in the first fishing,

(iii) that there is no recruitment, mortality, immigration or emigration between the times of the two fishings, and

(iv) that the first catch is removed from the population, or, if returned alive, the individuals are marked so that they can be ignored in counting the second catch.

This method was used by Mann et al. (1972) for the estimation of young fish in the River Thames. The area enclosed by a net should be properly known and two samples, c₁ and c₂, should be taken in rapid succession.

Estimations based on catch statistics and biological data

In cases where developed fisheries exist, the estimation of the fish stock can be made by using the catch statistics combined with biological data on the fish populations. Only two methods have been designed for or utilized in river fisheries.

Backiel's method:   Backiel (1971) suggested a simple method based on catch statistics, data on mortality, and the ratio between production and biomass (P/B):

where B_c = annual catch, B_m = biomass removed through natural mortality causes and K= P/B coefficient.

If there is insufficient data to calculate natural mortality and the P/B coefficient, the following reasonable assumptions can be applied:

 (i) Catch and biomass of fish removed through natural mortality is supposed to be of the same order of magnitude thus, total mortality is the sum of both figures;

(ii) Mortality is balanced by recruitment and growth according to Russell's (1931) concept of an exploited fish population, i.e., the population is in steady state.

(iii) The P/B coefficient varies with a narrow range from 0.5 to 1.0 or slightly more.

The first two assumptions are generally used in practice. Concerning the P/B ratio, Huet (1964) supposed that in most riverine situations production is about 50% of the biomass, hence P/B = 0.5. Mann (1965) in evaluating his own data from the River Thames agreed with Huet, and Backiel (1971) also used this figure for his estimation of predatory fish production in the Vistula River. It seems, however, that this ratio may be higher, varying between 0.6 and 0.9 in rivers from temperate zones (Mann, 1965; Backiel, 1971; Holčik et al., 1975). Mann et al. (1972) later found that in the Thames the P/B coefficient in bleak, roach, dace and dudgeon, was 1.92, 1.12, 1.75 and 1.94, respectively, i.e., higher than those previously calculated. In this example, eggs and fry were incorporated into the calculation. According to the data from Cuban lakes (Holčik, 1970), Lake Kariba (Balon, 1974) and the Kafue River (Kapetsky, 1974), it seems that in tropical rivers the P/B coefficient varies from about 0.7 to 1.3. The P/B ratio is very sensitive to the earlier stages of the life history, as shown by the case from the Thames, and this should be closely defined when presenting such data.

According to Backiel (1971) the assessment of fish stocks made by this simple procedure may be very close to that carried out by more cumbersome and time-consuming procedures, even if only the catch statistics are available. One can expect, however, that where there are differences in fishing intensity during the high and low water years, the stock may be under or over-estimated, respectively.

Biostatistical method:   The biostatistical method can be used in developed fisheries where there are well organized ichthyologic surveys to evaluate the stock of commercially valuable species. The classical method of Derzhavin (1922) was later considerably refined by Boiko (1934 and 1964), Monastyrsky (1940) Chugunov (1935) and Fry (1949). The concept, explanation and examples can be found in Ricker (1971 and 1975).

Estimations based on marking

In large rivers where the status of fish populations fluctuates highly according to variations in water level, the assessment of the stock by mark-recapture methods seems to be applicable only rarely. The method might be used in floodplain water bodies such as lakes or side arms in time of stable or decreasing discharge within the main channel (Holčik and Bastl, 1975). Recently, only Williams (1965), Koops (1975) and Raymond and Collins (1975) have used this method for estimation of fish populations in the Thames, Elbe and Columbia Rivers, respectively, but these all have relatively stable water regimes. The principles, limitations and various aspects dealing with the sampling requirements and the practical application of mark-recapture methods are summarized in the papers of Robson and Regier (1964 and 1968) and the manual of Ricker (1975).

Huet's method

Though very simple and at first view very rough, the method described by Huet (1949 and 1964), can be used to assess the approximate ichthyomass of the rivers, according to the following formula:

K = B × L × K                                                                              ............... (5)

where K represents the annual productivity of the water in kilogrammes per kilometre of river, L represents the average width of the stream in metres, B is the “biogenic capacity” and k is the coefficient of productivity. Values of B are from 1 to 3 for waters with little fish food, from 4 to 6 for average waters and from 7 to 10 for rich waters. The coefficient k if the product of k₁ (annual average temperature), k₂ (acidity or alkalinity of the water) and k₃ (the type of fish population) (see Table 1 in Huet, 1964). The production value thus obtained, divided by the P/B coefficient gives an approximate value of ichthyomass in kg/km (Mann, 1965).

This method has been improved by Lassleben (1977) and was used for the assessment of fish catch in the German stretch of the Danube (Kölbing, 1978). Lassleben calculated the k₃ coefficient (type of fish community) according to the area occupied by limnophilic (bream) and rheophilic (barb) species, respectively, assigning the value 2 for the former and 1 for the latter. Where the two groups of species occur together, the following generalized equations can be used when experimental fishing results in catches with proportion P₁ of limnophilic and P_r of rheophilic species.

and

the joint k₃ value is then: k_31r = (% area₁ × k₃₁) + (% area_r × k_3r)

It has been found that the estimated and observed catches for the region of the Danube River examined were in close agreement.

This method, which can be regarded as useful for the estimation of potential fish catch from a river can be further improved as follows:

 (i) The extrapolation of the values for k₁ (temperature gradient) can be done according to equation:

k₁ = -0.6671 + 0.1667°C

(ii) Only the simple percentage of rheophilic and limnophilic species found is sufficient to establish the k₃ coefficient:

(iii)   The biogenic capacity (B) of river may be assessed using the biomass of benthic invertebrates instead of the quantity of aquatic vegetation. According to Albrecht (1953 and 1959) streams with a biomass of zoobenthos less than 60 kg/ha can be considered to be poor, those with a biomass from 60 to 300 kg/ha medium and streams with a biomass of between 300 and 700 kg/ha to be good. Dividing these values among particular B grades 1–3 (poor), 4–6 (medium) and 7–9 (good) and calculating the regression equation, the following equations were obtained for the estimation of B:

B = 0.00 + 0.05 B_b

for values of biomass of benthos (B_b) of 60 kg/ha and less, and:

B = 0.35158 + 0.45469 log B_b

for values of benthic biomass from 60–700 kg/ha.

The following example shows how the results obtained by this method correspond with the actual values. In the Zofin arm of the Danube below Bratislava (1.2–3.8 ha in area) the mean biomass of benthic invertebrates in years 1971–73 was 357 kg/ha (B = 6.443), mean annual temperature 11.3°C (k₁ = 1.217), the water was rich in minerals (k₂ = 1.5), the percentage of limnophilic and rheophilic species was 87.4 and 12.5 respectively (k₃ = 1.875). Hence:

k = 6.443 × (1.217 × 1.5 × 1.875) × 10 = 220.53 kg/ha

The actual production of this arm in this period was 244.83 kg/ha. The estimated standing crop of the stock was 220.53/0.68 = 324.31 kg/ha. The actual standing crop found was 353.97 kg/ha.

PREDICTION OF CATCH

Several methods are used in ichthyology to predict future catches. These can be grouped into three sets depending on the nature of data utilized (Nikolsky, 1965). Besides statistical methods based on the analysis of catch with respect to the population dynamics of particular species, biostatistical methods are also used, mainly in the U.S.S.R. (Malkin, 1978 and Protopopov, 1978). Such methods are cumbersome, however, and only one attempt has been made to predict the catch of the whole taxocene vulnerable to fishing in rivers. For simplicity and quickness it is better to use simple methods based on statistical correlations between some environmental parameter and the catch.

Prediction based on environmental parameters and catch

As shown by many authors, for example, Antipa (1912), Stanković and Janković (1971), Ivanov (1956), Holčik and Bastl (1973, 1976, 1976a), Welcomme (1975) and Welcomme and Hagborg (1977), the dynamics of fish populations as well as catch in large rivers is a close positive correlate of hydrological regime. Calculating regression equations based on some hydrological indices, for example, water level, discharge of annual extent and duration of spates, of the preceding year and the data on the catch of the next, one may be able to predict the catch of fish in the following year (Welcomme, 1975; Holčik and Bastl, 1977). Despite its simplicity this method may give fairly accurate results and when improved by the addition of other terms to the regressions it seems to be promising for fisheries.

Prediction based on relationship between the fraction of catch and the total catch

This method has been suggested by Leopold (1972) for the prediction of catches of Coregonus albula in Polish lakes, but it may also be applicable in rivers. Leopold found statistically significant correlations between certain fractions of C. albula catch in year n-1 and their total catch in year n. For prediction, he used the multiple regression:

P = a + bC_s + cC_m                                                                           ............... (6)

where P = predicted catch, C_s = catch in spawning season in the year n-1 and C_m = sum of catches during the season of maximum yield in the year n-1. In the case of C. albula, C_s = catch in November and C_m = sum of catches from May until September. He also found that the plain linear regression of catch from the spawning period in year n-1 could be used. It seems that combination of this with the previous “environmental” method may be applied in riverine fisheries.

To evaluate the quality (Q) of prediction Andreev et al. (1977) suggest the following:

where r² = coefficient of determination of the correlation between actual and predicted catch. The higher the value of r, the lower the calculated quality index and hence the better prediction. To assess the accuracy of prediction (A_p) the same authors suggest the formula:

to calculate the mean absolute value of A_p, where P = predicted catch and C = actual catch. The result multiplied by 100 gives the mean percentage difference between the predicted and the actual catch.

COLLECTION OF DATA

Most types of fishing gear can be used for sampling. In addition to such conventional gear as seines, gillnets, entangling nets, traps and trawls (Steinberg and Dahm, 1975), electrofishing (Penczak and Zalewski, 1973; Micha et al., 1975) and even Rotenoning may give valuable results (Hocutt et al., 1973).

Particular attention should be paid to fishery statistics, which can be used for both stock evaluation and catch prediction. Whenever possible, the statistics should give data on the catch, not only of particular species from the river as a whole, but also for particular sections of the river as distinguished by their abiotic conditions. It is equally important that the catch data be separated by area (river or floodplain) and also by time of year. Data should include catch by gear, number and kind of gear, number of fishermen and boats, and the number of fishing days. In rivers with developed sport fisheries it is necessary to have statistics on the catch of anglers.

Observations should also be made on the ecology of fishes, particularly on aspects of seasonal activity including longitudinal or lateral migrations, spawning time and breeding grounds, movements connected with changes of the water regime, etc.

Records of non-biological parameters such as temperature, water level, discharge, duration of floods, time of the onset and culmination of floods and the area inundated by them might be crucial in many problems connected with assessment of stocks and prediction of catch.

REFERENCES

Albrecht, M.L., 1953 Ergebnissa quantitativer Untersuchungen an fliessenden Gewässern. Ber. Limnol. Flusstn. Freudenthal, 4:10–1

Albrecht, M.L., 1959 Die quantitative Untersuchung der Bodenfauna fliessender Gewässern. Z.Fisch., 8:481–550

Allen, K.R., 1951 The Horokiwi Stream: a study of a trout population. Fish.Bull.N.Z., (10): 238 p.

Andreev, N.N., A.I. Azvolinsky, M.Ya. Drapacky, 1977 Kriterij ocenki prognoza vozmoznogo ulova (Criterion of evaluation of the possible catch prediction). Rybn.Khoz.Mosk., 11:67–70 (in Russian)

Antipa, G., 1912 Uberschwemmungsgebiet der unteren Donau. Ann.Inst.Geol.Rom., 4(1910):1–172

Backiel, T., 1971 Production and food consumption of predatory fish in the Vistula River. J. Fish.Biol., 3:369–405

Balon, 1974 E.K., Fish production of a tropical ecosystem. Monogr.Biol., 24(2):497–523

Bastl, I., 1977 K reprodukcnej biologii belicky, Alburnus alburnus (Linnaeus, 1758) vo Vojcianskej systave dunajskych ramien (Notes on reproduction of the bleak Alburnus alburnus (Linnaeus, 1758) in the Vojka system of Danube arms. Biologia, Bratislava, 32(8): 591–8 (in Slovak)

Beverton, 1957 R.J.H. and S.J. Holt, On the dynamics of exploited fish populations. Fish.Invest., Minist.Agric.Fish.Food G.B. (2 Sea Fish.), London, ser. 2, (19):533 p.

Boiko, E.G., 1934 Ocenka zapasov kubanskogo sudaka (Estimation of the supply of Kuban zanders). Rab.Dono-Kuban-Nauchn.-Rybokhoz.Stn., 1:1–43 (in Russian)

Boiko, E.G., 1964 Prognozy zapasa i ulovov azovskogo sudaka (Forecasting supplies and catches of Azov zanders). Vses.Nauchno-Issled.Inst.Morsk.Rybn.Khoz.Okeanogr., 50:45–88 (in Russian)

Bonar, A., 1977 Relations between exploitation, yield and community structure in Polish pikeperch (Stizostedion lucioperca) lakes, 1966-71. J.Fish.Res.Board Can., 34(10): 1576–80

Chugunov, N.L., 1935 Opyt biostatisticheskogo opredelenia zapasaov ryb v Severnom Kaspii (An attempt at a biostatistical determination of the stocks of fishes in the North Caspian) Rybn.Khoz., Mosk., 15(6):24–9, (8):17–21 (in Russian)

Derzhavin, A.N., 1922 Sevriuga. Biologicheski ocherk (The stellate sturgeon (Acipenser stellatus Pallas), a biological sketch). Izv.Bakin.Ikhtiol.Lab., (1):393 p. (in Russian)

English, T.S., 1964 A theoretical model for estimating the abundance of planktonic fish eggs. Rapp.P.-V.Réun.CIEM, (155):164–70

Fry, F.E.J., 1949 Statistics of a lake trout fishery. Biometrics, 5:27–67

Gulland, J.A., 1969 Manual of methods for fish stock assessment. Part. 1. Fish population analysis. FAO Man.Fish.Sci., (4):154 p.

Hocutt, C.H., P.S. Hambrick and M.T. Masnik, 1973 Rotenone methods in a large river system. Arch.Hydrobiol., 72(2):245–52

Holčik, J., 1970 Standing crop, abundance, production and some ecological aspects of fish populations in some inland waters of Cuba. Vestn.Cesk.Spol.Zool., 33(3):184–201

Holčik, J. and I. Bastl, 1973 Ichtyocenczy dvoch dunajskych ramien so zretelom na zmeny v ich druhovom zlozeni a hustote vo vztahu ku kolisaniu hladiny v hlavnom toku (Ichthyocenoses of two arms of the Danube with regard to changes in species composition and population density in relation to the fluctuation of the water level in the main stream). Biol.Pr., 19(1):5–106 (in Slovak)

Holčik, J. and I. Bastl, 1975 Sampling and population estimation with small mesh seine as used during ichthyological research on Danubian backwaters. EIFAC Tech.Pap., (23) Suppl.1, vol.2:627–40

Holčik, J. and I. Bastl, 1976 Ecological effects of water level fluctuation upon the fish populations in the Danube river floodplain in Czechoslovakia. Acta Sci.Nat.Acad.Sci.Bohemoslov. Brno, 10(9):1–46

Holčik, J. and I. Bastl, 1976a Vplyv hydrologickych zmien na ichtyocenozy dunajskyen ramien (Influence of hydrological conditions upon the ichthyocenoses of Danubian arms). Zpr.Cesk. Zool.Spol., (7–9):55–76 (in Slovak)

Holčik, J. and I. Bastl, 1977 Predicting fish yield in the Czechoslovakian section of the Danube River based on the hydrological regime. Int.Rev.Ges.Hydrobiol., 62(4):523–32

Holčik, J. 1975 et al., Sledovanie produkcnych pomerov jednotlivych druhov ryb na jednom ramene ako modelovom objekte (Investigation of production of fishes in one Danubian arm). In Final report, Laboratory of Fishery Research and Hydrobiology, Bratislava, 24 p.(unpubl.) (in Slovak)

Huet, M., 1949 Appreciation de la valeur piscicole des eaux douces. Trav.Stn.Rech.Eaux For. Groenendaal-Hoeil, serie D, 10:1–55

Huet, M., 1964 The evaluation of the fish productivity in fresh waters. The coefficient of productivity. Verh.Int.Var.Theor.Limnol., 15:524–8

Ivanov, L.S., 1959 Uroven vody kak faktor povysenia zapasov i ulovov ryb v Dunaje (Water level as a factor influencing the increase of fish resources and catch in the Danube River). Dokl.Bolg.Akad.Nauk, 12(6):553–6 (in Russian)

Jefimova, A.I., 1967 Metod prognozirovania ulovov snetka Pskovsko-Chudskogo vodojema (The method of forecasting of catches of Osmerus eperlanus eperlanus m. spirinchus Pallas in Pskovsko-Chudskoy water body). Izn.Gos.Nauchno-Issled.Inst.Ozern.Rechn.Rybn.Khoz., 62:226–35 (in Russian)

Kapetsky, J.M., 1974 The Kafue River floodplain: an example of preimpoundment potential for fish production. Monogr.Biol., 24(2):497–523

Kölbing, A., 1978 The European method of fish harvest prediction in fluvial systems. Environ. Biol. Fish., 5(3):249–51

Koops, H., 1975 Tagging experiments for the assessment of the size of fish population in the rivers Elbe and Trave. EIFAC Tech.Pap.,(23) Suppl.1, vol.2:620–6

Lassleben, P., 1977 Las Schätzverfahren für Fischwässer nach Leger und Huet. Österr.Fisch., 28:53–64

Leopold, M., 1972 Metodyka prognozy i oceny polowow ryb o krotkim cyklu zyciowym w jeziorach polskich (na przyklade sielawy (Coregonus albula (Linnaeus)) (Methodology of forecasting and evaluating catches of fishes with short lifetime in the Polish lakes (on the example of Coregonus albula (Linnaeus)). Rocz.Nauk Roln. (H), 94(4):5–98 (in Polish)

Leopold, M. and B. Dabrowski, 1975 General premises and selected elements of a method of estimating fish stocks and populations in Polish lakes. EIFAC Tech. Pap., (23) Suppl.1, vol. 2:722–7

Malkin, E.M., 1978 Prognozy zapesov ryb na osnove biostatisticheskogo metoda, ioh opravdajemost i sovershenstvovanie (Fish stock forecast based on the biostatistical method, their reliability and improvement). Vopr.Ikhtiol., 18(2):229–42 (in Russian)

Mann, K.H., 1965 Energy transformations by a population of fish in the River Thames. J.Anim. Ecol., 34:253–75

Mann, K.H. 1972 et al., Productivity and energy flow at all trophic levels in the River Thames, England. In Productivity problems of freshwaters. Proceedings of the IBP-Unesco Symposium on productivity problems of freshwaters, held in Kazimierz Dolny, Poland, May 6–12, 1970, edited by Z. Kajak and A. Hillbricht-Ilkowska. Warsaw, PWN Polish Scientific Publishers, pp. 579–96

Micha, J.-C., J.-C. 1975 Philippart and B. de Moffarts, Echantilonnage piscicole en eau profonde dans deux affluents de la Meuse. EIFAC Tech. Pap., (23) Suppl.1, vol.1:201–8

Monastyrsky, G.N., 1940 Zapasy vobly Severnogo Kaspia i metody ich ocenki (The stock of vobla in the north Caspian Sea and methods of assessing it). Tr.Vses.Nauchno-Issled.Inst. Morsk.Rybn.Khoz.Okeanogr., 11:115–70 (in Russian)

Nikolskii, G.V., 1969 Theory of fish population dynamics as the biological background for rational exploitation and management of fishery resources. Edinburgh, Oliver and Boyd, 323 p. English translation of Teorija dinamiki stada ryb, published by Nauka, Moscow (1965)

Penczak, T. and M. Zalewski, 1973 The efficiency of electro-fishing with rectified pulsating current in the zones of a river of medium size, evaluated by the method of successive catches. Acte Hydrobiol., 15(4):343–55

Protopopov, N.K., 1978 Prognozirovanie vylova poluprochodnogo siga-pyzhiana r.Petchory (Catch forecast of semi-migrant coregonid Coregonus pidschian of the River Petchora). Rybn.Khoz., Mosk., 1978 (1):28–39 (in Russian)

Raymond, H.L. and G.B. Collins, 1975 Techniques for appraisal of migrating juvenile anadromous fish populations in the Columbia river basin. EIFAC Tech.Pap., (23) Suppl.1, vol.2:552–65

Regier, H.A. and D.S. Robson, 1967 Estimating population number and mortality rates. In The biological basis of freshwater fish production, edited by S.E. Gerking. Oxford, Blackwell Scientific Publications, pp. 31–66

Ricker, W.E., 1971 Derzhavin's biostatistical method of population analysis. J.Fish.Res.Board Can., 28( ):1666–72

Ricker, W.E., 1975 Computation and interpretation of biological statistics of fish populations. Bull.Fish.Res.Board Can., (191):382 p.

Robson, D.S. and H.A. Regier, 1964 Sample size in Petersen mark-recapture experiments. Trans. Am.Fish.Soc., 93(3):215–26

Robson, D.S. and H.A. Regier, 1967 Estimation of population number and mortality rates. IBP Handb., (3):124–58

Rothschild, B.J., 1961 Production and survival of eggs of American smelt, Osmerus mordax (Mitchill), in Maine. Trans.Am.Fish., Soc., 90(1):42–8

Russell, E.S., 1931 Some theoretical considerations on the overfishing problem. J.Cons.CIEM, 6 (1):3–27

Saville, A., 1964 Estimation of the abundance of a fish stock from egg and larval surveys. Rapp. P.-V. Réun. CIEM, 155:164–70

Seber, G.A.F. and E.D. LeCren, 1967 Estimating population parameters from catches large relative to the population. J.Anim.Ecol., 36(3):641–3

Stanković, S. and D. Janković, 1971 Mechanismus der Fischproduktion im Gebiet des mittleren Doneulaufe. Arch.Hydrobiol., 36(4):299–305

Steinberg, R. and E. Dahm, 1975 The use of two-boat bottom and midwater trawls in inland waters - experiences in the German fishery. EIFAC Tech. Pap., (23) Suppl.1, vol.1:23–35

Turner, J.L., 1975 Preliminary analysis of demersal trawl fishery of Lake Malawi. EIFAC Tech. Pap., (23), Suppl.1, vol.1:14–22

Welcomme,R.L., 1975 The fisheries ecology of African floodplains. CIFA Tech.Pap., (3):51 p.

Welcomme, R.L. and D. Hagborg, 1977 Towards a model of a floodplain fish population and its fishery. Environ.Biol.Fish., 2(1):7–24

Williams, W.P., 1965 The population density of four species of freshwater fish, roach (Rutilus rutilus (L.)), bleak Alburnus alburnus(L.)), dace (Leuciscus leuciscus (L.)), and perch (Perca fluviatilis (L.)) in the River Thames at Reading. J.Anim.Ecol., 34:173–85

3.   STATISTICS - SURVEY DESIGN (by G. Bazigos)

In developing countries, fisheries operating in inland waters are classified as small fishing unit economies. Fisheries operating in rivers and river basins are seasonal in nature and highly mobile in character. Specifically, mobility of fishing units can be divided into the following two categories:

(a) Peripheral mobility, i.e., movements of local fishing units which take place within the area of the surveyed body of water.

(b) Inflow/outflow movement of fishing units in the surveyed body(ies) of water during the main fishing season(s).

Because of the dynamic character of the fisheries and wide spatial distribution of the fishing units, statistical survey systems based on the “wholesale approach”¹ usually produce very poor results.

For the development of efficient statistical survey systems both the peculiarities of the surveyed populations and the practical problems encountered by developing countries have to be taken into account.

In the developed technology, current large-scale surveys are used for the collection of the required items of information. Specifically, the employed large-scale surveys can be grouped into the following two domains:

(a)   Frame Surveys (FSs) - aiming to provide information on the size, structure and spatial distribution of the surveyed fishing industry.
Depending on the size, organic structure and character, the survey methods used for FSs are either the aerial approach or road/water approach, or a combination of these two approaches.
Current Frame Surveys based on a well defined rotation system are also used for providing information on changes of the fishing industry over time.

(b)   Current Catch Assessment Sample Surveys are employed for providing current estimates on the input items (fishing effort) and output items (fish catch) of the surveyed fisheries.

The efficiency of the large-scale current surveys is, among other things, a function of the efficiency of the survey design of the surveys. It has been proved that the cost of the surveys decreases and the accuracy of the obtained estimates increases if a pre-stratification system of the surveyed area is employed. In the case of river basins, topographical maps of a scale 1:50 000 should preferably be used. By using proper criteria for stratification, the surveyed area can be divided into “hydrological strata” and each hydrological stratum can be further sub-divided into homogeneous hydrological zones.

Estimates of the level of localization of the fishing industry within the established hydrological zones, for the various distinct periods of a hydrological year, can be estimated by using the findings of a current Frame Survey. A further stratification of the established hydrological zones is employed by taking as a criterion for stratification the observed pattern of spatial concentration of fishing boats within them, here called hydrological sub-zones.

A multi-stage sample plan - with unequal probabilities - is used for the selection of the samples of units of the Catch Assessment Survey (CAS) within the established hydrological zones. Specifically, for the CAS, the method of sampling in space and time is used.

Estimates of the surveyed magnitudes are first calculated on a hydrological period basis. Estimates for the hydrological year as a whole are calculated by adding up the estimates of the individual hydrological periods.

¹  Wholesale approach: In this survey method, items of information of the surveyed magnitudes are collected by sending in the field hundreds of statistical recorders aiming to cover as much as possible of the surveyed area

In order to improve the accuracy of the obtained measurements the “landing approach”is used. The system also provides the basis for the calculation of recorder's variance andvalidation of their performance in the field.

4.   ON FISHING AND FISHERIES MANAGEMENT IN LARGE TROPICAL AFRICAN RIVERS WITH PARTICULARREFERENCE TO NIGERIA (by J.B.E.Awachie)

INTRODUCTION

It is now fairly well established that in tropical rivers, fishing and other fisheries activities are closely geared to the prevailing hydrological regimes. Specific and geographic differences in the extent and duration of longitudinal and lateral migrations of river and floodplain fishes in response to annual or biannual floods (as in the Niger system) have provided the basis for the development of thriving indigenous fisheries from ancient times.

Many effective methods for the exploitation of juvenile fish have also evolved. This is possibly in response to the particular characteristics of the ecology of floodplain fish populations, whereby the high water season is the period of most intense biological activity for fish (breeding, intense feeding and growth) with a consequent increase in ichthyomass due largely to the rapid growth of the young of the year. Furthermore, most floodplain fish species tend to have annual or short life cycles, maturing within one year and ready to spawn by the next flood season (Lowe-McConnell, 1977).

The above essentially classic picture of the fisheries in large tropical rivers and their floodplains is currently undergoing rapid changes due to a number of factors principal among which are:

(i)   increasing demand for fish protein and the consequent intensification of fishing with improved modern gear and techniques;

(ii)  the desiccation of both the river channels and their floodplains due to dam, irrigation and flood control developments. The future of the fisheries of large rivers will, therefore, depend on the extent to which the negative effects of the above factors on the productivity of the system are accommodated, thus ensuring the desired high level of sustainable yield.

In this paper, attention is drawn to aspects of the fishing and management components within the overall framework of the problems of fisheries in large rivers.

CRAFT AND GEAR

There is little doubt that the level of development of available fisheries technology and the effectiveness of its application can be quite important in determining the catch from river fisheries in both the main channel and floodplain. In Nigeria for instance, the introduction of new, modern and at first sight better craft and gear has not often led to increased catches, as has been observed not only in the Niger-Benue drainage system, but also in the Ogun and Oshun rivers to the southwest, as well as the Yedseran and Yobe/Hadejia complex to the northeastern parts of the country. The need for craft and gear to be developed and modified to suit environmental requirements (i.e., appropriate technology) in order to achieve desired goals is, therefore, to be stressed.

Craft

The most commonly used fishing craft even today is the traditional, manpowered, dug-out canoe (Awachie and Hare, 1978). Canoe sizes vary with the part of the river system in which they operate. In general, larger canoes are used in the main river channel while smaller and more manoeuvrable ones are used on the floodplains.

More recently, planked canoes have been introduced in the lower and wider reaches of large rivers. Indeed, currently, flat-bottomed plywood boats produced by FAO/UNDP technologists at Baga on Lake Chad designed to meet operational requirements of shallow floodplain swamps, pools, etc., are being tested in a number of large rivers in Nigeria.

The contribution of this craft toward increased fish production from the river-floodplainsis yet to be assessed. It is noteworthy that these boats can carry a number of cold boxes or coolers for the preservation of captured fish.

Gear types and distribution

One of the most striking features of river fisheries is the variety of gear used. Practically all the major methods of fish capture are represented. This is perhaps a reflection of the requirements of the equally wide variety of species and habitats presented by tropical rivers and their floodplains.

Passive gear, e.g., traps, weirs, setnets and longlines are predominant. In season and in some systems, active gear, e.g., castnets, beach seines, life and dip nets become very important in exploiting certain fish stocks (Stauch, 1966; FAO/UN, 1969; Welcomme, 1971; Awachie, 1976). Detailed descriptions of gear commonly used in the major river systems of Nigeria have been provided by Reed et al. (1967). On their general distribution and use in large rivers, Awachie and Hare (1978) have observed that as one moves up a river basin, fishing gear becomes less varied and sophisticated, tending toward the traditional traps and spears. Available information from other tropical African river systems indicates that in large rivers, traps, gillnets, beach seines and poisons top the list of methods used in exploiting their fisheries resources. This is in agreement with the more general findings of Lagler (1971) for large rivers.

Gear problems

Some of the more important problems which militate against the maximum utilization of the above gear types in the fisheries production of large rivers and their associated floodplains include:

 (i)  Financial constraints in acquiring the right quality and quantity of gear

(ii)  Rapid deterioration of certain gears in the hot tropical environment (Awachie and Walson, 1978)

(iii)  Gear selectivity and effectiveness in the exploitation of various stocks and habitats

(iv)  The problems of standardization of gear (types, sizes and meshes, etc.) found most suitable for specific species, habitats/environments.

The effectiveness of a gear is often overlooked in attempts to introduce new technology in order to increase production. Like craft, gear effectiveness in rivers is not always a function of its sophistication. Thus, in floodplain fisheries, various indigenous devices and practices are known to be more effective in certain habitats than more standard gear. This is exemplified by the performance of traps and foul-hook longlines in creeks and channels; and drain-in ponds or canals in swamps and pools (Welcomme, 1971; Awachie, 1976). Thus the development of efficient indigenous technology is indicated.

As the major components for gear effectiveness include those inherent in the gear structure and allied problem of selectivity, those due to the skill of operation and to environmental requirements, it seems clear that comparative data on gear and craft performance in the various river and floodplain habitats would be essential in developing management procedures for increased fish production.

FISHING ACTIVITY

Hydrology and seasonality

Seasonality is the dominant feature of the fisheries of large tropical rivers and it is in turn correlated with hydrological phenomena which give rise to high and low-water periods. The latter correspond to periods of low and high catch respectively, as demonstrated by field observations and experimental fishing (Awachie and Walson, 1978). Welcome and Hagborg (1977) have, however, established that the draw-down characteristics, especially the volume and area covered by water during the low-water period, may be of great importance in determining the ichthyomass and catch in the same year for a simple fishery such as is practised in most tropical rivers and floodplains.

The above generalizations have vital implications for developing appropriate management procedures as will be dealt with below.

Manpower and productivity

The exploitation of river and floodplain fisheries in tropical Africa is traditionally largely undertaken at the artisanal level. As noted above, a remarkable feature of artisanal fisheries is the use of a wide variety of gear thus maximizing the exploitation of the wide assortment of species and habitats of large rivers. Recent studies including those of Welcomme (1975, 1976), Welcomme and Hagborg (1977) have indicated that even from theoretical considerations, artisanal fisheries may well prove to be the best adapted to the particular demands of tropical floodplains.

From the foregoing, it can readily be appreciated that fisheries manpower is a very relevant problem of fish production of large rivers especially in the developing areas of the tropics. Fishermen and their support - extension staff are inadequate in a number of ways and these inadequacies reflect on productivity.

Most fishermen are part-time and, in season, they make significant contributions to fisheries activities. However, because of their main occupation, usually full-time farming, they are unable and often unwilling to make the required investment in terms of time, capital etc., on fisheries production.

Full-time hands are far fewer and their ratio to part-time fishermen during the flood season in the Lower Niger and Anambra systems may be as low as 1:10. Full-time fishermen in Nigeria come mainly from coastal areas and the Niger Delta and are largely migrant, moving further inland and upstream with the early flood and returning to their more permanent bases during the dry season. Because of the lack of technical support they receive from the available skeletal extention staff, they rely heavily on indigenous gear and expert knowledge of the rivers and floodplain terrain for their reasonably high productivity.

The need for developing fisheries manpower is clearly indicated. The target should be local people with detailed knowledge of local river systems, terrains as well as local and relevant channel and floodplain fisheries practices.

Fisheries statistics and productivity

Available information on productivity, standing stock and catch for some African floodplains - Niger, Benue, Sokoto, Senegal, Ouémé, Shire, Barotse, Kafue Flats, Massili and Okavango - has been reviewed by Welcomme (1975). As might be expected, the data show that productivity and catches are higher in the floodplain than in the main river channel. However, the picture given by commercial catch figures from 12 African floodplains while providing a useful guide may be far from accurate, at least for the Nigerian Niger and Benue, because of the present unreliable procedures for collecting and rendering catch statistics in the latter areas. As this drawback may exist to varying degrees in other tropical river systems, it is thought the gathering of reliable essential fisheries statistics over long periods, necessary for the development and management of river fisheries, should be high on the list of any effort to identify and provide solutions to the problems of large river fisheries. Here, the need for adequate training for all cadres of staff involved and coordination of procedures for collecting and making returns are spotlighted.

NOTES ON MANAGEMENT

Functional management procedures are yet to be identified in the large rivers of tropical Africa. The urgent necessity to employ scientific management techniques in the exploitation of the rich river fishery resources of the continent, as has been done for some of the large lakes of Africa, e.g., Victoria and Tanganyika, has been felt for some time. Unfortunately, unlike some of the above lakes, relevant basic scientific data on the biology and ecology of the fish and fisheries of large rivers in their undisturbed state are either unavailable or too scanty to be meaningfully employed in development/management planning. As Welcomme and Hagborg (1977) rightly pointed out, it is noteworthy that nearly all the studies that have so far been carried out have been done either to predict the effects of management and dam construction, or to try to explain and improve a deteriorating fishery.

However, on the basis of the present state of our knowledge, management of river and floodplain fisheries aimed at maximizing sustainable yield especially in developing countries with their large deficits in the animal protein requirements of hinterland peoples, should take full cognizance of the following:

 (i)  Environmental parameters operating in the productivity of rivers and floodplains especially the key position of hydrology. Field trials have shown that a single reliable measure for increasing fish production during the dry season is by maintaining a high water level (FAO/UN, 1969; Awachie, 1973 inter alia). Such cheap measures to achieve this in both the main channel and the floodplains as dams and small earthwork around floodplain swamps and pools or canalization of water from reservoirs to floodplains will no doubt prove profitable management tools.

(ii)   Stock dynamics and fisheries production and the need for regulation and protection of vulnerable stages in stock formation, reconstitution and growth cycle. Exploitation of fish at the beginning of the flood cycle can do positive damage to stocks by removing migrating gravid fish thus reducing the number of breeding females. Thus Soulsby (1959) and Cadwalladr (1965) have reported the collapse of Labeo stocks in two different rivers due to fisheries during the early flood. The imposition of closed seasons at the commencement of flood is indicated as a regulatory measure where applicable. Similarly, closed seasons can be used to prevent the massive destruction of the young of these species required for stocking floodplain ponds and pools, both natural or artificial, as discussed below. Suffice it also to mention in this connexion that although fisheries regulations, where they exist, are hardly enforced in African riverine environments, the value of floodplain lakes as fish sanctuaries from which to repopulate overexploited rivers or stretches, is gradually being appreciated. It is to be pointed out that this role has always been performed in the traditional setting by “sacred lakes” in which fishing is forbidden, even to-date in many riverine areas of southern Nigeria.

Gear control measures in large rivers and floodplains in general are unlikely to be effective because of the characteristics of the riverine environment coupled with their mode and level of exploitation. However, the protection of certain species or young fish from overexploitation with intensively operated gear such as the atalla liftnet in the Niger-Benué system, is possible because of the short duration of the fishery (Awachie and Walson, 1978).

(iii)  Production phases, “high and low water” and gear/manpower requirements and servicing. The average holding of gillnets and longlines which are very effective in the above situation is currently grossly inadequate. Management inputs, in the form of improved gear, craft and better educated manpower will improve output from the main channel and large floodplain lake or lagoon fisheries.

(iv)  Multi-purpose use of river and floodplains and measures to offset their adverse effects. A concensus appears to be emerging that damage to fisheries can be minimized, if not controlled, by informed discussions at the planning stage between policy formulators, policy executors, especially engineers, and fisheries biologists.

AQUACULTURE

Floodplain

Aquaculture on the floodplains of large rivers is considered here as a management measure aimed at maximum exploitation of their fisheries resources. Fish culture can be carried out either in excavated ponds or in natural floodplain lentic waters. Aquacultural practice in the latter may, indeed, be regarded as a natural extension of the procedures for keeping the largest possible area of the floodplain under as high a level of water as possible during the dry season in order to increase fish production (Awachie, 1968, 1973; Reed et al., 1967). The techniques for setting up and managing conventional artificial pond farms are fairly well known but their wide development in tropical Africa has been hampered by lack of adequate local experience, lack of fish feeds or inadequacy of feed supply, problems of regular fish seed supply due to the absence of effective hatcheries, lack of or unwillingness to stake the relatively high initial capital required for the enterprise, to name a few.

Because of the low cost of their establishment, natural floodplain lakes and ponds have recently been attracting the attention of local fishing communities. The evolution of floodplain lentic waters into farm ponds, their present level of development, procedures for their management and future prospects for enhancing the fisheries production of the floodplains of the Niger-Benue system, have been discussed by Awachie (1976).

Depending on their size, life span and location on the floodplain, these natural lakes and pools may be used as either seasonal or perennial farm ponds while swamps, shallow lagoons with minimally suitable depth of water at the height of the dry season can be put under semi-culture to increase their yield.

The proposed procedures for managing natural floodplain fish culture, in the various types of available lentic waters are summarized in Table 1. Also some of the suggested schemes for converting these natural water types into various culture categories (Awachie, 1976) are given in Appendix 1. As can be seen from the above, management procedures are, as in capture fisheries, based on the manipulation of (i) the hydrological cycle, (ii) features of the floodplain morphology and (iii) the biological characteristics of available local species relevant to fish culture.

There are very obvious advantages of this type of culture over conventional farm ponds especially because of the small number of minor constructions needed which can be done with cheap local materials, stocking with suitable local species, the seed of which can be readily provided in season with the many available and efficient juvenile fisheries. It is thought that increased fish production from culture is more likely to be achieved by developing the aquacultural potential of natural floodplain ponds and lakes. This would appear more so because of the general caution and unwillingness of local fishermen and entrepreneurs to invest in conventional artificial pond farms because of their usually poor returns. In Nigeria this attitude persists in spite of the government policy to make loans specifically for fish culture available on very soft terms through special Agricultural Development Banks.

Main channel

The indigenous limited culture of fish in kraals or pens fenced off cheaply from main river channels (and large lakes and lagoons as well) exists among the riverine peoples of the lower Niger. Estimated yields of 900–2 060 kg/ha/year from such experimental fish pens stocked with various combinations of Clarias, Tilapia and Parophiocephalus in 1968/69 were reported. The result also indicated the potentials of polyculture involving local tilapias and highly priced carnivores such as ophiocephalids. It is considered that kraal and pen culture if developed and encouraged where the morphology of rivers permit, would aid maximum exploitation of the fisheries potential of tropical rivers.

Pilot trials

The necessity for well planned and executed production trials in various ecological and geographic zones of tropical rivers, for the successful development of culture fisheries in main channels and natural floodplain waters cannot be over-emphasized. So also is the organizational set-up to undertake these pilot schemes. It is noted here that in Nigeria the envisaged feasibility studies to determine potential areas for this type of development have not been possible since 1974 because of the political restructuring of the country from 12 to 19 States and the policy to set up new multi-purpose river and lake basin authorities to handle the integrated development of all land and water resources including fisheries, in their areas of authority.

The role of pilot projects should include among others:

 (i)  Testing the feasibility of transfer of available culture technology to floodplain aquaculture in tropical Africa.

(ii)  Developing acceptable procedures for the use of wastes of various types in river and floodplain culture taking into consideration all relevant parameters especially human water supply and public health aspects.

(iii) Undertaking in-depth studies of the comparative economics of various culture methods with a view to determining the most economically productive and viable procedures within the various social conditions. Such relevant factors in tropical environments as the reduction of initial capital and feed costs in intensive culture, the role of local mills and local agricultural wastes in producing low cost protein feeds, the comparative efficiency of performance of local herbivore and carnivore fish species in mono- and poly-culture, etc., should be given adequate attention.

(iv) Assessing requirements for supporting aquacultural research and extension services vis-à-vis the current shortage of trained and experienced manpower and best training arrangements in the prevailing circumstances.

(v) Reappraising those factors which have militated against the development of fishermen's cooperatives in tropical Africa because of the great potential of such groups for increased production if they can be put on sound foundations. It is clear that virile cooperatives will help to improve the social welfare of the predominantly small-scale and part-time operators as well as removing other constraints in fisheries technology, distribution and financing.

CONCLUSIONS AND RECOMMENDATIONS

From the foregoing outline, treatment of problems of fish production and management in large tropical rivers, the following conclusions and recommendations are presented for consideration:

 (i)  The need for the integrated approach in planning development strategies for increased exploitation, processing and distribution of the fisheries resources, is strongly indicated in order to achieve improved welfare and quality of life of hinterland fishermen in their rural setting.

(ii) The emergence of an appropriate fisheries technology for increased production will involve colateral activities of updating traditional gear and craft which have been found to be operationally effective in the environment and the adaptation and adoption of new ones to meet the demands of effective usage in the light of the high variability factor in the morphology of rivers and floodplains. Manoeuvrability of craft and gear is a major requirement in order to minimize costs, e.g., new flat-bottom planked boats with shallow draughts are proving effective not only in moving across rivers and swamps but also in transporting considerable quantities of the catch in cold boxes/coolers without frequent grounding.

(iii) Documented management procedures for large rivers are unavailable in tropical Africa. Their development has been hampered by lack of essential basic data on the biology of the undisturbed environment, fish and fisheries. Meaningful attempts to formulate appropriate techniques for managing the fisheries must consider the key parameters of hydrology, fish stocks and their complex dynamics of species composition and population, as well as the traditional local approach to regulatory controls in fishing. It is indicated that such regulations as closed seasons and setting up of fish sanctuaries which are already in customary use will prove effective tools for managing impaired fisheries. Gear control measures are not expected to succeed until the artisanal level of fisheries exploitation is significantly changed.

(iv) The floodplains of tropical rivers with their extensive network of swamps, lagoons, lakes and ponds have great potentials for aquaculture. In addition to conventional culture in excavated ponds which is capital intensive, the natural ponds and pools offer challenging opportunities for developing cheap fish culture. Full realization of the latter will hinge on the evolution of appropriate technology for their cheap conversion and management based on a through appreciation of relevant environmental and social parameters as indicated above.

(v) Pilot projects are shown to be a sine qua non for the establishment of the requisite appropriate technology and management procedures for both wild and culture fisheries. They will also help to remove other identified major constraints to fisheries production especially the prevailing low educational and economic status of the fishing community.

REFERENCES

Awachie, J.B.E., 1968 Fish culture trials with natural stock and ponds on the lower Niger floodplain near Atani. Umudike Report of the Fisheries Investigation Unit (unpubl.)

Awachie, J.B.E., 1973 On conservation and management of inland water resources of Nigeria. 1. Natural lakes and ponds with special reference to their utilization for fishery development. In First symposium on Environmental Resource Management in Nigeria. Ile-Ife, University of Ife Press

Awachie, J.B.E., 1976 Fish culture possibilities on the floodplain of the Niger-Benué drainage system. CIFA Tech.Pap., (4) Suppl.1:256–81

Awachie, J.B.E. and L. Hare, 1978 The fisheries of the Anamabra, Ogun and Oshun river systems in southern Nigeria. CIFA Tech.Pap., (5):170–84

Awachie, J.B.E. and E.C. Walson, 1978 The Atalla fishery of the lower Niger, Nigeria. CIFA Tech. Pap., (5):296–311

Cadwalladr, D.A., 1965 The decline in the Labeo victorianus Blgr. (Pisces: Cyprinidae) fishery of Lake Victoria and an associated deterioration in some indigenous fishing methods in the Nzoia river, Kenya. East Afr. Agric. For.J., 30(3):249–56

FAO/UN, 1969 Report to the Government of Nigeria on fishing technology relating to river and swamp fisheries of Northern Nigeria. Based on the work of William Reed, FAO/TA Fishery Technologist. Rep. FAO/UNDP(TA), (2711):90 p.

Lagler, K.F., 1971 Capture, sampling and examination of fishes. IBP Handb., 3:7–44

Lowe-McConnell, R.H., 1977 Ecology of fishes in tropical waters. London, Edward Arnold, 64 p.

Reed, W. et al., 1967 Fish and fisheries of Northern Nigeria. Zaria, Northern Nigeria, Ministry of Agriculture, 226 p.

Soulsby, J.J., 1959 Status of the Lake Mweru fishery. Rep. Jt. Fish. Res. Organ., (8), 1958:30–8

Stauch A., 1966 Le bassin camerounais de la Bénoué et sa pêche. Mém. ORSTOM Paris, (15):152 p.

Welcomme, R.L., 1971 A description of certain indigenous fishing methods from southern Dahomey. Afr.J.Trop.Hydrobiol.Fish., 1(2): 128–40

Welcomme, R.L., 1975 The fisheries ecology of African floodplains. CIFA Tech.Pap., (3):51 p. Issued also in French

Welcomme, R.L., 1976 Extensive aquaculture practices in African floodplains. CIFA Tech.Pap., (4) Suppl.1:248–55

Welcomme, R.L. and D. Hagborg, 1977 Towards a model of a floodplain fish population and its fishery. Environ.Biol.Fish., 2(1) :7–24

APPENDIX 1
Suggested Schedules for Converting Floodplain Ponds into Fishculture Ponds
(from Awachie, 1975)

SEASONAL CULTURE ACTIVITIES

(a)  Seasonal Nursery Farms (natural stock)

To achieve an extension of the natural nursery role of the fadama, the following sequential activities may be undertaken:

(b)  Seasonal Harchery/Nursery (chosen species)

(c)  Rearing Ponds (natural stock)

(d)  Rearing Ponds (chosen species)

PERENNIAL CULTURE ACTIVITIES

(e)  Hatchery/Nursery (chosen species)

(f)  Rearing Farms (chosen species)

The above conversion schemes are to serve as guidelines and may be modified according to the requirements of limiting local cultural parameters.

5.   FISHERY PROBLEMS ASSOCIATED WITH MULTIPLE USES OF LARGE RIVERS (by V.R. Pantulu)

INTRODUCTION

Large rivers are valuable natural resources; they are the sites for generation of hydroelectric power, milieu for a variety of exploitable organisms, sources of water for diverse human uses including irrigated agriculture, repositories of wastes, and channels for navigation. The potential uses of large rivers have always attracted human settlements on their shores resulting in growing population pressures and conflicts arising out of multiple demands and uses. This paper proposes to analyse some of the major uses of large rivers and the fishery problems arising therefrom, with particular reference to the tropical zone. The ultimate objective of all resource use enterprises would be to improve the lives of the people. Therefore, an ideal water use plan should take into consideration the need for balanced development, envision conflicts arising out of multiple uses, and resolve them in an environmentally compatible manner. This emphasizes the need for projecting future problems of an as yet unrealized multiple water use plan. In the present state of our knowledge a forecast of fishery problems arising out of multiple uses would necessarily be largely a matter of inference and conjecture rather than of conclusion. The unusual nutritional and economic importance of fisheries of large rivers makes it imperative that fisheries scientists gather adequate background information through intensive investigations in rivers proposed to be “developed”.

Fishery problems arising out of the different uses of rivers would obviously vary with the type of use and the nature of changes it brings about in the riverine habitat, fish food supply, spawning sites, nursery grounds and lebensraum in general. These changes, in turn, will have differential effects on different species of fishes; some may benefit, others may not be affected at all, and yet others may be harmed, depending on their life history, ecological requirements, behaviour and adaptability. Therefore, in the ultimate analysis, the nature of influence of a particular type of use on the vitality of individual fish species is important in gauging fishery problems of multiple use of rivers. Also to be taken into account in this context is synergism of different uses, where one type of use may exacerbate or alleviate the effects of another.

Fishery problems created by a particular type of river use would obviously depend on the actions underlying that use. These actions are the potential causative factors of changes that would eventually lead to impacts on fisheries. Therefore, the linkages between these causative actions and their consequences for the natural fisheries of a geographical entity must be traced systematically for each of the uses. The potentially significant causal relationships should be identified and, where possible, the directions and magnitudes of potential impacts inferred. Armed with such knowledge, one would be in a position to alert planners to the implications of multiple water use programmes on aquasystems and fisheries.

MULTIPLE USES - THEIR GOALS; OBJECTIVES AND CONSEQUENCES

Among the more important goals for uses of large rivers that generally conflict with fisheries interests are:

–    agricultural development (including irrigation)
–    industrial development (including hydroelectricity)
–    flood control
–    navigation improvement, and increasing urban water supplies.

Of these various goals, agricultural and industrial development can be identified as the primary ones since historically these have provided the major benefits to humans.

Each of these two primary goals has a specific objective. The objective for agricultural development might be to increase field crop production by a given percentage per unit of time throughout the area. The objective for industrial development could be to initiate or increase the electro-process, agriculture-related, and water-dependent (e.g., river) transportation industry in the area. These objectives would be achieved by a combination of water management, and ancillary actions.

From the goals, objectives and means presented, six major clusters of activities that can significantly affect fishery productivity can be identified. Each cluster includes both construction activities and ongoing management operations, and can generate its own set of direct and indirect impacts upon the fisheries. The clusters of activities are as follows: (i) dam construction and implacement; (ii) agricultural development; (iii) industrial development; (iv) flood control; (v) navigation improvement, and (vi) provision for adequate domestic water supply.

  (i)   Dam construction and implacement

Although dam building itself is not a goal, but is a means for satisfying some or all of the five major goals associated with multiple water use development, many water use fishery problems result from the direct and indirect effects of constructing dams and other physical structures for water management schemes.

Construction activities typically require large-scale earth moving to construct earthfilled dams and to provide new port facilities, the recruitment or importation of a construction labour force, the creation of at least temporary communities for this labour force, and temporary diversion and disruption of river flows to permit construction. Related activities may include the clearing of vegetation from some or all of the lands to be flooded above the dams, the resettlement elsewhere of populations presently living in those lands, and the actual flooding of lands above the dams up to the maximum water levels. Subsequently, activities may involve the construction of new communities, industries and port facilities.

 (ii)   Agricultural development

A principal objective of water resource development is to increase field crop production by the implementation of intensive agricultural techniques. This may require the control of downstream flooding of potentially productive land and the irrigation of land seasonally for year-round or “multiple cropping” cultivation.

Development of irrigation agriculture requires the construction of canals and/or other aquaducts, and the diversion through these canals of water from a reservoir or other source. To realize the multi-cropping potential of irrigation agriculture may require additional actions such as application of fertilizer, mechanizing cultivation, greatly increasing livestock production, diversifying field crops and probably draining large areas of natural fish- producing lakes, ponds and pools.

Agricultural development plans also include action to control the salinity of agricultural lands in delta regions to increase their potential for agricultural field crops. This will include diking of both sea and river banks to prevent inundation, draining brackish waters from these lands into stream courses by gravity ditches and pumping, augmenting seasonal low flows in the river to prevent upstream intrusion of brackish waters, and irrigating poldered lands by freshwater canals from upstream. Ancillary activities include intensive production of field crops in presently undeveloped land, which will require additional fertilizers and pesticides, etc.

(iii)  Industrial development

Dams primarily designed to produce hydroelectric power involve using penstocks and turbins for electric power generation, constructing transmission lines to major urban centres, and managing river flows to ensure year-round minimum “heads” at downriver hydropower dams and for navigation.

Typical industries that emerge from large dams are electro-processing, agriculture-related and river transport-oriented. Electro-processing industries will use hydroelectric power potential to produce iron and steel, calcium carbide, caustic soda, chlorides, ferro-alloys, copper, phosphoric acid, tin and particularly, aluminium. Agriculture-related industries will accompany planned agricultural development and will include input industries to produce pesticides, fertilizers and other chemicals, livestock and fish feed, veterinary medicines, implements, and output industries for food processing and oil extraction. River transport-oriented industries will use navigation improvements to exploit mineral and lumber resources - for mineral processing (copper, iron ore, coal, rock salt and tin concentrate) and wood processing (lumber, plywood, pulp and paper, and shipyards) - and to export products of, and import supplies for other industries.

Primary activities associated with industrial development will include constructing industrial plants, forming urban communities, with related urban public health services; consuming water resources; and increasing riverine commercial vessel traffic. Secondary activities include producing industrial waste materials, both airborne and waterborne; generating increased concentrations of urban and municipal wastes; and accelerating development of land resources, both directly for industries and their related services and indirectly for the use of urban industrial populations. Diverting water to urban and industrial uses will increase with time, will require budgeting, and will ultimately limit development.

(iv)  Flood control

As a corollary to agricultural development and/or urbanization, downstream seasonal flood control is part of many water use plans. The intention is to reduce or eliminate seasonal inundation of alluvial lands, particularly in the low-lying river-mouth and coastal areas, so that these lands can be developed more intensively. Current maximum downstream flood levels are lowered by redistributing the flow pattern more evenly between the wet and dry seasons, with the excess wet season waters being stored in reservoirs. Dikes eliminate flooding by wind and tide in coastal zones and, accompanied by freshwater flushing, poldering and pumping, reclaim shallow sea bed for agriculture.

Downstream flood control includes both primary and ancillary activities that may affect fishery productivity. Primary actions include constructing dikes to reduce overbank flooding and seasonally storing water in various types of impoundments. Ancillary actions include such activities as substituting single-transplant for floating rice, and other agricultural and industrial efforts.

(v)   Navigation improvement

Navigation facilities are improved in some water use schemes to permit commercial shipping upriver, which requires a minimum river depth of 3 m. The required primary activities to do this include constructing navigation locks over intervening waterfalls, rapids or dams, inundating existing waterfalls and rapids by damming and thus limiting the locks to dams, regular dredging of river channels and of sea channels at the river mouth, and augmenting seasonal low flows from upstream impoundments to maintain navigable depths. Ancillary activities include increasing riverine port facilities and vessel traffic, developing facilities to handle new industrial production along with future maintenance dredging, and disposing of dredged soil, waste oil, sewage, and other materials from the transport vessels.

(vi)  Urban water supply

Population grows and is continually accompanied by a drift of people toward urban areas. This drift is accelerated by those aspects of water resource development schemes that foster industrialization and urbanization. Domestic, service, manufacturing and industrial needs for water in cities thus grow. Demands may be met by storing surface waters, by tapping underground supplies, by bringing in water through aquaducts, and locally and increasingly by still costly desalination. Often, primary actions such as the foregoing need to be supplemented by major ecosystem alterations such as constructing remote, large water storage reservoirs, by diverting all or parts of the flow of nearby river systems, or by other means such as recycling water for re-use in any of a variety of ways. Associated actions include such activities as water treatment to improve quality and sewage and wastewater treatment to enable re-use.

FISHERY PROBLEMS

Fishery problems contingent on multiple uses of rivers would, as mentioned earlier, differ not only from one use to another but would also depend on the geographic distribution, abundance, behavioural and physiological adaptations of the species of fishes concerned. Therefore, when the implications of multiple uses are understood - in so far as the state of knowledge will allow us to infer - the possible impacts on the amounts of standing crop and catch can be explored and potential economic consequences derived. An important tool in this connexion therefore would be a knowledge of the possible effects on specific species of fish.

From the standpoint of adaptations of fishes to the environment, two broad sectors of riverine fisheries can be identified: the mainstream and tributary fisheries. These can be further subdivided into:

 (i) upland rapid stream fisheries

(ii) floodwater fisheries of plains

(iii) estuarine and coastal fisheries.

Again, within these fisheries, the impacts would be different on migratory and sedentary species, fishes with limited and wide distribution, mainstream spawners, inundation zone spawners, egg scatterers, nest builders, and so on.

Mainstream fishes

It is to be expected that the mainstream association of fishes will be adversely affected by agricultural, industrial and community development. The initial effect of impounding the mainstream could be very critical to certain species and beneficial to others in this aggregation. The reasons are many, but for the most part the effects can clearly be expected to result from: (i) general reduction or increase, often drastic, of preferred habitat; (ii) diminuition or enhancement of natural food supplies (riverine insects, molluscs and other invertebrates, plankton algae, and higher plants); (iii) reduction or increase of suitable spawning sites (clear exposed bedrock surfaces with plenty of moving oxygenated water for some species or large areas of slow-moving to quiet waters for others, such as the egg scatterers); and (iv) alteration in extent of shelter areas (which might then become territory for predators).

Important factors which relate to these changes are those of hydrology, sediment load, nutrient content and biocide and toxicant loads.

(a)   Hydrology:  Dam construction and operation will change the hydrology of the river downstream. Seasonal peak flows will be reduced to keep them within the river banks. Inundation zone areas that provide fish spawning and feeding habitat will be grossly reduced, if not eliminated entirely, in some river stretches, and significantly reduced in others - wherever flood control is effective. Water quality characteristics of floodwater flows may be altered, including changes that are important for the biological signalling of fishes. The timing of peak flows may influence both the onset of longitudinal and lateral spawning migrations and their duration.

For species that migrate longitudinally in the river systems, dams will create obvious problems. However, migratory organisms are a special case, so their problems should not be confused with the general inability of mainstream species to cope with life in or below reservoirs. For example, at least initially, concentrations of upstream migrators at the bases of dams will make them unusually vulnerable to overexploitation thereby compounding the negative effect of the dams as barriers to spawning grounds. Flow regulation also could reduce the growing period for hatchlings in nursery waters, mostly in the residual inundation zone. The velocities of water flows downstream will be reduced and, associated with this, average water temperature will rise. These changes should affect fish species sensitive to changes in temperature and current. Seasonal low flows will be increased to benefit navigation and salinity control downstream, however, and these increased flows will also support higher levels of dry-season fish stocks than currently exist and will help to prevent harmful concentrations of waterborne toxic substances discharged with wastes.

(b)   Sediment load:  Construction of impoundments also will change sediment flows, particularly in the main channels of the river. Sediment loads affect the aquatic biological production systems because they limit light penetration and thereby regulate photosynthesis. They contribute to potentially productive habitats in shallow waters and seasonally inundated areas, and they transport some nutrients to fisheries downstream. Thus the reduction in sediment load will significantly alter the existing downstream ecosystems. Also, the scouring effects of the river in picking up sediment below the dam will tend to eliminate habitats and foraging areas of river bottom species. Clearer waters will also favour some predator fishes.

Dredging and channelling a river for navigation purposes will complement the impacts of changes in the sediment load rate. Not only will dredging remove considerable amounts of habitat and forage base, it will also stir large amounts of sediments back into suspension, offsetting to some extent the decrease in the nutrient load and increase in predator activity. However, cropping of fish may be facilitated if large spaces are opened up for haul seining or drift gillnetting. Some problems of soil disposal from dredging operations might also occur for example, if the dredged material is dumped on nearby shallow spawning or nursery grounds.

(c)  Nutrients:  Nutrients, like other sediments, also tend to be trapped in impoundments rather than transported downstream. While this increases the productive potential of the reservoir fishery, it diminishes the fishery potential downstream. Some additional nutrients, however, will be made available by runoff of increased amounts of fertilizers and organic debris from intensified and expanded agricultural development downstream, by the discharge of organic waste from urban areas and by discharge of nutrient-laden waters from the lower reaches of the reservoir. Elevating the nutrient load by such means can have positive effects on fisheries by increasing the overall productivity of the river and its carrying capacity for fish, if the load is not excessive. Animal wastes will increase productivity, except where the concentration is too strong - near large poultry production operations, cattle shipping yards, and slaughterhouses. Over- loads of nutrients such as the foregoing can cause extreme eutrophication and a build up of biochemical oxygen demand (BOD), thus rendering the affected stretches of the river uninhabitable to fish. The total nutrient flow to the oceans might be increased significantly from the above sources. The effects on estuarine fishes and fisheries should merit close investigation.

(d)   Biocides and other toxic materials:  Perhaps more important than the changes in nutrient load will be the increase in biocide loads. Industrial activities - pulp and paper mills, textile mills, plating plants, chemical factories - on the mainstream or tributaries can adversely affect mainstream fish stocks, as is well known. Other mainstream effects might result from an increase in shipping with concomitant pollution from spillage of petroleum products and ship-waste effluent from bilge water and boat sewage. Water-dependent industries are located on flowing water expressly for the convenient discharge of waste chemicals as well as for the supply of fresh water. The discharge therefore will include toxic wastes as well as heated waste water, with the increased temperature considered as a pollutant from the perspective of many riverine fish species. Increased pesticide runoff, especially from intensively cultivated high yield crops which require large amounts of pesticide due to the inherent sensitivities of monoculture and to reduced genetic ability of high-yield crops to withstand insect pests, will offset some of this increased productive potential. Pesticide use will more than directly affect paddy, canal, pond and mainstream fish; it will affect organisms of the higher trophic levels of the aquatic production system throughout the water resource system and in turn eventually lower the reproduction rate and increase the mortality rates of the fishes themselves. Transfer of pesticides to humans through body contact, food chains, or via drinking water poses a special set of problems.

The decrease in sediment load mentioned earlier also may serve to alter distribution of toxic materials in the river flow, in that sediments tend to pick up toxic substances by absorption.

Tributary fishes

Fishes of tributaries will probably be least affected by the construction of mainstream impoundments. For the most part, tributary species with extensive ranges of distribution in small streams are very broadly adapted and have wide environmental tolerances because of the variety of habitats and the accentuated seasonal changes in conditions of life in tributaries. Many of these species can even prosper in sluggish or quiet waters and may indeed profit from the formation of reservoirs and impoundments. However, the factors of industrial and agricultural development which can adversely influence mainstream species can also influence tributary species in the same manner, but more dramatically, because tributary channels have a smaller dilution capacity than large mainstream channels. Tributary species with large ranges of distribution, should not be confused with rapid-water upland tributary forms, as the latter have much more stringent biological requirements, such as strong currents and high dissolved oxygen in the water, and may be markedly influenced by reservoir submergence of stream areas critical for their survival. These species lack the broad adaptability to adjust to living in an impoundment. Thus some species will be eliminated from the standing water, although some populations may persist in tributaries upstream and possibly downstream from the impoundments because they generally lack migratory habits. As such, they do not have to move through impounded areas to spawning habitats. Some headwater tributary species will benefit from agriculture (e.g., from fertilizers) but several may be hurt. Where there are hill-stream fishes (such as Homaloptera spp., Glyptothorax spp., Scaphiodonichthys acanthopterus in Southeast Asia) there can be adverse effects from an increase in turbidity and organic content of the water in agricultural areas. An excess of nutrients and decomposing material can reduce the dissolved oxygen levels substantially from those to which these particular species have become adapted. This group will not profit from development in the way that the tributary species of extensive range will because this group as a whole appears to lack the broad environmental tolerances that the wide-range forms have. Adaptability and/or broad tolerance are necessary prerequisites to a wide distribution throughout the varied habitats prevailing in the waters of a river basin.

CONCLUSION

It should be obvious from the foregoing that fishery problems of multiple uses of rivers are varied and complex. A capability to predict possible fishery impacts of planned river uses would go a long way in maximizing benefits from all the contemplated uses, otherwise the very benefits to be derived from a given use can be negated by resultant losses in fisheries, which could be considerable. For instance, in the Mekong basin, where a fairly comprehensive study of the possible effects of integrated development of the river basin on the fisheries was conducted, it was estimated that in an 85-km stretch of the river alone, fishery losses would be about 2 000 metric tons of annual catch with a market value of U.S.$ 1.05 million. Such studies, attempting to evaluate fishery problems of multiple river use, are unfortunately few and far between. But the example cited should serve to highlight the magnitude of losses that could be sustained and the need to plan a balanced development programme for multiple use of rivers, based on intensive studies, such as those on the Mekong river, so that the resultant benefits could be enhanced and losses offset by compensatory actions.

6.  AN ALTERNATIVE SCENARIO FOR RIVER BASIN DEVELOPMENT IN AFRICAN WOODLAND SAVANNAS¹ (by T. Scudder)

INTRODUCTION

The purpose of this brief commentary is to emphasize the extent to which major river basin development projects in Africa are biased toward the development of the urban-industrial sector at the expense of the rural areas in which dams and other regulatory devices are constructed. Within these rural areas, impacts on the economic systems of local residents, including their agriculture and fisheries, are far more negative than they need be. For this reason, suggestions are included for making river basin development projects more effective mechanisms for both rural and urban development.

THE IMPORTANCE OF RIVERINE AND LACUSTRINE HABITATS FOR SAVANNA POPULATIONS IN AFRICA

Within Africa, savanna woodland covers from 40 to 80 percent of the continent depending on how one defines this environment. The largest single spread of savanna in the world, it is traversed by a number of great rivers, including the Senegal, Niger, Nile and Zambezi - which are undergoing major modification as a result of the construction of large-scale dams, canals, and in some cases irrigation works. Many of the countries involved are dominated by a single river system which more often than not constitutes the heartland of the nation. How that system is utilized will influence the future of the human population, and of the ecosystem as a whole for generations to come. Along with the Orange, the Limpopo, and the Chari-Logone, these rivers flow for much of their length through savanna environments. At the same time their annual regime is influenced by the patterns of seasonal rainfall which are associated with savanna woodland vegetation. Though there are partial exceptions where river flows are evened out below major swamps and wetlands, these regimes are characterized by major annual fluctuations, with a significant annual flood toward the end of, or after, the rainy season, and flow ratios that exceed 1:10. Under such circumstances, dams constructed to control river flows must be backed up by huge reservoirs - which is the case, for example, with the Kariba, Volta and Aswan High Dams on the Zambezi, Volta and Nile Rivers. In evening out annual flows, such dams effectively terminate the annual flood, requiring major modifications in downriver life-styles since the lives of riverine people are intricately inter-related with the annual rise and fall of water levels.

The negative impacts of river basin development on downstream populations are illustrated by the Kainji Dam Project where Adeniyi (1973) found that termination of the annual flood seriously cut the dry season harvests of downstream farmers and the yields of fishermen. No longer flooded, thousands of hectares went out of cultivation, with over 20 000 households among the Nupe alone affected. As for fishing, the income of fishermen from Adeniyi's three study villages decreased after the cessation of annual flooding 73, 60 and 47 percent, respectively. Actual catches monitored at one landing point by the Kainji Lake Research Project dropped over 50 percent.

Adverse impacts extended even further downstream reaching to the apex of the Niger Delta. According to Awachie (this report), “by 1970 the Lower Anambra Basin which hitherto had been responsible for 70 percent of freshwater fish and yam production in Eastern Nigeria, had lost 60 percent of its fish output and yam production running into 100 000 tonnes”.

¹  This paper is excerpted almost entirely from “African River Basin Development and Local Initiative in Savanna Environments”. Forthcoming in Human Ecology in Savanna Environments, David R. Harris, editor (London and New York, Academic Press)

THE INLAND DELTA OF THE NIGER: A CONTEMPORARY EXAMPLE OF A COMPLEX AND PRODUCTIVE LOCAL LAND AND WATER USE SYSTEM

In the Inland Delta of the Niger, as in the Lake Chad Basin, a number of ethnic groups through time have worked out a complicated set of relationships which have maintained environmental quality, a substantial human population, and a reasonable degree of peace between pastoralists, sedentary agriculturalists (who increasingly own livestock) and professional fishermen. The overall system is both complex and dynamic, with each population of producers using a repertoire of diverse techniques which are modified through time according to labour resources, management needs and local environmental and socio-political conditions.

Though the numbers of people and livestock involved vary, according to Gallais (1967), there were in the fifties over 300 000 Peul (Fulani) with over 3 million stock, 70 000 farmers and 80 000 fishermen using the Inland Delta. The number of Peul and their livestock build up gradually during the dry season, the herdsmen moving inland from the river system during the rains. As for the farmers, in years of “normal” flood, they cultivate some 100 000 ha of seasonally flooded land within the Inland Delta. In September, just before flooding begins, floating varieties of African rice are planted which adjust to the rising waters by lengthening their stems. Later non-floating varieties are also sown, but in shallower waters on higher land.

After the annual floods peak, the flood recession or décrue cultivation cycle begins. While the cultivation of rice on the rising (crue) flood has a relatively narrow distribution in West Africa, recessional cultivation is common throughout the desert and savanna areas of the continent. Perhaps best known is basin irrigation in Egypt, décrue cultivation is also important along the Senegal River (where perhaps 120 000 ha are cultivated following a “normal” flood), the Niger (from Guinea to Nigeria), and within the Lake Chad Basin. It also was important to the Gwembe Tonga prior to their relocation in connexion with the Kariba Dam scheme. Though cereals, pulses and cucurbits vary from area to area, within the Inland Delta bulrush millet (Pennisetum) and dura (Sorghum) are the principal crops, with the first sown near the top of the recession area since it is more drought-resistant. As for the fishermen, who are mainly Bozo and Somono, their annual catch has been recorded as high as 100 000 tons.

Such types of local food production systems have supported millions of people for millenia without adverse ecological impacts. During periods of drought, the importance of the associate driver and lake basins is magnified especially for pastoral peoples. Though they then suffer significantly reduced water flows or water levels which increase the pressure on the remaining grasslands and the potential for conflict, it would be hard to over-emphasize the importance of the Senegal and Niger River Valleys, and the Lake Chad Basin, to pastoral peoples during the most recent Sahelian drought.

Such low-cost, labour-intensive systems combining natural irrigation with livestock management and fisheries deserve far more research and development attention than they have received to date. For their practitioners, capital-intensive large-scale irrigation projects are not the answer for raising living standards since in Africa they tend to involve a relatively small number of tenants and labourers - often at not too profitable levels (Barnett, 1977; Sørbø, 1977). In addition, the financial costs for such projects are increasing rapidly because of inflationary pressures, and because of a growing awareness “of the need for far better drainage facilities to reduce the omnipresent dangers of water-logging and salinity. Here it is not inappropriate to note that recent FAO data show that at least 50 percent of the irrigated land in the world is saline, with several hundreds of thousands of hectares going out of cultivation each year because of salinity alone” (Brokensha et al., 1977). Under such circumstances, the enhancement of existing systems of land and water use based on natural fluctuations of tropical river systems makes increasing sense.

THE SOCIO-ECONOMIC IMPACTS OF RIVER BASIN DEVELOPMENT ON LOCAL POPULATIONS

The short-term impacts of large-scale river basin development projects on the life-styles and economic systems of people like the multi-ethnic population of the Inland Delta of the Niger to date have been largely negative. In connexion with the Kariba, Volta (Ghana), Kainji (Nigeria) and Kossou (Ivory Coast) Dam projects, over 200 000 savanna residents were forced to move from the future lake basins. While compulsory relocation in connexion with river basin development will continue to pose problems for people and governments alike (Scudder, 1976), the major West African projects currently being planned for the Niger, Senegal and Chari-Logone systems will adversely affect millions of downstream users unless project design is radically altered. As currently designed, the Manatali Dam on the Senegal River will drastically reduce downriver recessional cultivation which covers over 100 000 ha in a good year. While planners intend to replace this with 400 000 ha of capital-intensive irrigation, it will take generations to bring this land under cultivation. Meanwhile, with annual flows evened out, local crop yields will drop and dry season grazing will decrease, as will fish landings. Even assuming that the local people will be incorporated within the irrigation projects (and there is no guarantee that this will be the case), what are they to do in the interim? Upstream dams on the Niger could have an even more serious effect on the Inland Delta, while high dams on the Chari-Logone system present very serious implications for the farmers, herdsmen and fishermen who utilize the Lake Chad Basin. Though phase one of the Jonglei Canal Project in the Sudan could be the first major river basin development project to have major positive impacts on a local population, even there the ultimate outcome will depend on decisions which have yet to be made concerning the alignment of the canal and the nature of phase two.

The fact of the matter is that in most cases the interests of the local river and lake basin populations are not taken into consideration by river basin planners. This applies both to those currently occupying future reservoir basins and to downriver residents. The impact of large-scale dams and canals on the physical and biotic components of the ecosystem has received even less attention. There are a variety of reasons behind this most unfortunate state of affairs, the most important being the following: national policies which favour the urban consumer; a primary focus on water resource development as opposed to the integrated development of human, land and water resources throughout basin areas; and “the development from above” syndrome coupled with “the myth of the conservative peasant”.

National policies which favour the urban consumer

The urban bias in third world development analysed by Michael Lipton (1977) is especially blatant in tropical Africa, in spite of the fact that rural populations constitute the majority in all countries, with their numbers ranging from approximately 60 percent of the total (in Zambia, for example) to over 90 percent of the total in some of the poorer Sahelian countries. With the possible exception of Tanzania and the Ivory Coast, price structures and other national policies favour the urban consumer at the expense of the rural producer, a situation which has been documented for a number of countries including the Sudan, Kenya and the six Sahelian countries. Big dams are a classic example of this bias since they are primarily unipurpose schemes for the generation of hydro-electric power which is almost exclusively for the urban-industrial sector. Indeed, no appreciable rural electrification is associated with any of the existing or planned major dam projects in African savanna woodland, although Kossou may become an exception to this generalization. Rural residents, of course, realize this bias which is one reason why so many migrants are flocking to urban areas throughout Africa.

The primacy of water resource management and hydropower generation

Except in the Sudan and South Africa, the economic justification of big dams previously constructed in African savanna woodlands was almost exclusively based on the generation of hydro-electric power for the urban-industrial sector. Examples include Kariba, Volta and Kainji. As for the exceptions, water management is still the primary consideration, with capital intensive irrigation projects added to hydropower generation, or in the Sudanese case, taking precedence. In other words, current design and construction has locked governments into a particular type of development that precludes pursuit of other alternatives. More specifically, a strategy which maximizes power generation may be at the expense of more productive means for obtaining energy, including crop agriculture, livestock and fisheries development.

A case in point is the Kafue Hydro-Electric Scheme in Zambia. According to Williams (1977), “The main problem has been that the development of the hydro-electric potential appears to have been planned without regard to the need for water for irrigation on more than a minor scale”. Furthermore, little was done to exploit the lower reservoir's potential for fisheries development in spite of the fact that technical recommendations had been made in ample time to launch a compatible mound building programme which would have increased landings by spreading fishing effort over a much wider area. In their assessment of the situation, members of the Kafue Basin Research Committee of the University of Zambia stated their conviction “that development plans should recognize that biological and agricultural necessities may ultimately overwhelm other factors: in particular that the biological productivity of an area such as the Kafue Flats may in the long run prove to be vastly more important to mankind than its short-term value as a modified water storage for the generation of electrical power” (Williams and Howard, 1977).

“The development from above” syndrome coupled with “The myth of the conservative peasant”

River basin development is a classic example of “the development from above” syndrome, a sort of economic “laying on of hands” whereby national planners backed up by international expertise and finance superimpose development strategies on local populations who have virtually no say in their planning, implementation, management or evaluation - in spite of the fact that they are the principal risk-takers. In general, both national and international planners tend to believe “the myth of the conservative peasant”, which presents local farmers and fishermen as a homogeneous mass of non-innovating, custom-bound conservatives - in spite of the fact that social science research since the end of the second world war presents the exact opposite picture (see for example, Bates, 1978 and Hill, 1970).

THE SOCIO-ECONOMIC RATIONALITY AND DYNAMISM OF LOCAL SYSTEMS OF LAND AND WATER USE

The socio-economic rationality and increasing dynamism of socio-cultural systems in Africa is well illustrated by the Nilotic Dinka and Nuer of the Sudd region, both of whom (but especially the Nuer) are still considered by most planners and academics with Sudan experience to be among the most conservative people in the world. Though this image may have accurately reflected the situation prior to Sudanese Independence in 1956, the floods of the early sixties and the civil disturbances that finally came to an end with the Addis Ababa agreement in 1972 have played a major role in opening up the societies of both people to external stimuli.

Because of herd size reduction during the sixties and early seventies owing to flooding, fighting and removal of veterinary services, increased numbers of Dinka and Nuer began to rely on fishing for subsistence purposes and to meet their growing cash needs. Though formerly a despised full-time activity among cattle-owning Nilots, today fishing has not only been commercialized but may well be the most important single source of capital for other commercial activities. In 1976, when I travelled through one small eastern Nuer settlement close to the Jonglei Canal alignment, there were less than ten village stores - all Nuer-owned and all made from local materials. On my return in 1978, I found over 20, approximately three-fourths of which were capitalized through fishing. West of the Nile, the same applies in the Adok Ador area. There also, the more successful fishermen attempt to diversify their business operations as soon as possible by starting up one or more stores and entering the cattle trade. Though no longer fishing themselves, they continue the fishing operation by recruiting a head fisherman and a number of labourers, using the profits to improve the structure and inventory of their stores and to build up their herds for domestic use and trade. Western Nuer cattle drovers now cover substantial distances and are well aware of price differentials, some travelling 400 km to Juba, while others go still further to Yambio, close to the Sudan-Zaire frontier.

A POSSIBLE DEVELOPMENT SCENARIO FOR DAM CONSTRUCTION BENEFICIAL TO LOCAL POPULATIONS

Since the generation of urban-consumed hydropower pays the bills in most cases for river basin development, do options actually exist which can benefit local populations without undermining the financial viability of the project? In the absence of the necessary benefit-cost analyses, this question cannot be answered one way or the other. I would like, however, to discuss briefly one option which should sufficiently increase biological productivity in some projects so as to justify a reduction in hydropower generation. This option would involve drawdown regularization within the reservoir backed up behind the dam synchronized with simulation of a downstream flood. Properly executed, it would increase significantly the production of crops, livestock and fish by lake basin and downstream residents. It would also have positive social equity and environmental impacts, although the major justification would have to be economic, with the trade-off between reduced power generation and increased agricultural production (including fisheries) favouring the latter. Since no such analyses have been carried out in connexion with river basin development in tropical Africa, a major need for the future is to widen the range of water use alternatives to include drawdown regularization and flood simulation.

Though the extent of the drawdown varies annually primarily in response to reservoir intakes, as do high and low water marks and the dates each year on which the drawdown begins and ends, the area which is annually exposed in the larger African reservoirs amounts to tens of thousands of hectares. At Lake Volta, for example, experts associated with the UNDP/FAO/Ghanian Lake Volta Research Programme estimate a total drawdown area of over 80 000 ha, of which at least half is potentially cultivable.

Drawdown cultivation is a form of labour-intensive, low-cost recessional cultivation similar to the systems of natural irrigation already practised by local populations in many river basins. For this reason, few problems are anticipated in promoting systematic cultivation of drawdown areas following dam construction. Although extension and marketing services would have to be improved, along with other inputs, both the human resources and, frequently, the necessary trunk roads are there. For all their negative impacts on local populations, the construction of big dams does accelerate the incorporation of local populations into a regional or national system. Among the best in Africa, all-weather roads to the dam sites are an important factor in this regard.

At Kariba as elsewhere, local farmers have attempted to replicate their previous system of recessional cultivation within the drawdown area but increased risks preclude extensive cultivation (Scudder, 1972). Since the same applies to other major reservoirs, once reservoir levels begin to draw down, prospective farmers have no way of knowing whether or not the cultivation season will be long enough to allow them to harvest the crops sown. After farmers lose their crops on a number of occasions, further lake basin cultivation during the rains and the dry season tends to be restricted to land above the high-water mark, with farmers hand-carrying water to their crops should water levels drop too rapidly during the dry season. Aside from isolated gardens, the crop potential of the extensive drawdown areas is wasted.

The solution to this problem is to regularize the annual drawdown during the latter portion of the dry season and the beginning of the rainy season so as to guarantee farmers a minimum number of water-free days before reservoir levels would be allowed to rise. Not only would farmers be told when they could plant, but they would also be advised as to what vegetable, cereal and fodder crops could be expected to mature within the growing season. The benefits in terms of production, social equity and employment could be considerable. At Kossou, for example, 1-2 ha plots could support 5 000-10 000 families. Cultivation would not necessarily be restricted to the dry season and early portion of the rainy season (although prices would be better at that time), since varieties of floating rice could be grown when reservoir levels rise. Benefits could also be increased by a range of other productive activities including livestock management and fish ponding (as a supplement to existing reservoir fisheries).

Critical constraints on livestock management in most African savanna environments are water and food during the height of the dry season, which is, of course, why many pastoralists and farmers move their herds down to the rivers and wetlands at that time. With their greater shorelines, the major African reservoirs not only increase the dry season water supply but also the food supply. Though perhaps too oligotrophic for the cage rearing of fish (Coche, 1977), savanna reservoirs may also have some potential for fish farming. While experimentation with supplemental feeding would be needed, one approach would be to build small wiers across some of the many inlets that occur along the reservoir edge. Equipped with gates, these would allow water and fish to pass inside when the water level rises. When the reservoir subsequently draws down, water could be retained within the “fish ponds” until the time came for harvesting the yield.

The annual drawdown would also be used to simulate and regularize a restricted downstream flood, this being an intentional management strategy. Unlike the situation prior to dam construction, such a simulated flood would eliminate the previous extremes which either flooded out downstream farmers or drastically cut their total acreage when river flows were seriously reduced during periods of drought. In largely removing the major constraints for existing production systems, improved water control and flood simulation would increase productivity, whereas the current management of existing African reservoirs significantly lowers the productivity of downstream users. Though elimination of the extreme floods could well have negative ecological impacts, nonetheless these would be far less than those which can be expected where the annual flood is evened out. While fish productivity still might drop off to an extent, the drastic reductions that occur when the annual flood is virtually eliminated would be avoided.

The major cost of the proposed scenario will be a reduction in power generation since simulation of a downstream flood will require some reservoir waters to be passed through the sluice gates rather than through the turbines. This will reduce the volume of water stored as will drawdown regularization during years when inflows from exceptionally heavy early rains will have to be passed through the sluices so that reservoir crops are not flooded out before the end of the period of guaranteed drawdown. While the magnitude of this loss may be too great to change the regime of existing dams and perhaps too great in terms of the design of selected future dams, such decisions should be based on careful optimization studies which broaden the scope of river basin development projects to consider a wider range of alternatives relating to the integrated development of the human, land and water resources of the entire river basin and inter-basin region. Here it is important to recall that Africa does have a significant share (40 percent) of the world's hydro-electric potential. Already in southern, central and western Africa, this potential is being distributed through international grids so that the continent's wealth in this regard should enable planners to consider a wider range of uses (and especially those that enhance the biological production of energy through crop agriculture, livestock management and fisheries) even though generating capacity may be reduced at selected sites. As for major wetlands like the Inland Delta of the Niger and the Nilotic Sudd, far more emphasis should be paid to their management rather than to their drainage.

REFERENCES

Adeniyi, E.O., 1973 Downstream impact of the Kainji Dam. In Kainji: a Nigerian man-made lake. Socio-economic conditions. Kainji Lake studies, edited by A.L. Mabogunje. Ibadan, University Press, Nigerian Institute of Social and Economic Research, Vol. 2: 169–77

Barnett, T., 1977 The Gezira scheme: an illusion of development. London, Frank Cass

Bates, R.H., People in villages: micro-level studies in political economy. Submitted forpublication in World Polit., 1978

Brokensha, D.W., M.M. Horowitz and T. Scudder, 1977 The anthropology of rural development in the Sahel: proposals for research. Binghampton, N.Y., Institute for Development Anthropology, Paper 4

Coche, A.G., 1977 Premiers résultats de l'élévage en cages de Tilapia nilotica (L) dans le Lac Kossou, Côte-d'Ivoire. Aquaculture, 10:109–40

Gallais, J., 1967 Le delta intérieur du Niger: étude de géographic régionale. Dakar, Institute Fondemental de l'Afrique Noire, 2 vols.

Hill, P., 1970 Studies in rural capitalism in West Africa. Cambridge, Cambridge University Press

Lipton, M., 1977 Why poor people stay poor: urban bias in world development. Cambridge, Mass.,Harvard University Press

Scudder, T., 1972 Ecological bottlenecks and the development of the Kariba Lake basin. In The careless technology: ecology and the international development, edited by M. Taghi Farvar and J.P. Milton. Garden City, New York, Natural History Press, pp. 206–35

Scudder, T., 1976 Social impacts of river basin development on local populations. In River basin development: politics and planning. Proceedings of the UN Interregional Seminar on River Basin and Interbasin Development. Budapest, Institute for Hydraulic Documentation and Education, Vol. 1:45–52

Sørbø, G.M., 1977 How to survive development: the story of New Halfa. Monogr. Ser. Dev. Stud. Res. Cent.Univ.Khartown, (6)

Williams, G.J., 1977 Transferred water use in the Kafue Flats. In Development and ecology in the Lower Kafue Basin in the nineteen-seventies. Lusaka, University of Zambia, Papers from the Kafue Basin Research Committee

Williams, G.J. and G.W. Howard (eds.), 1977 Development and ecology in the Lower Kafue Basin in the nineteen-seventies. Lusaka, University of Zambia, Papers from the Kafue Basin Research Committee