extent to which the state of health ofthese communities may be related tochanges in water quality, broughtabout by the general effects of pollu-tion, was evaluated.

The River Rimae study areaThe River Rimac is 138km long anddrains an area of 3583km2. The catch-ment area is below the Uco peak at5 100m above sea-level. Rain and snowfrom the peak drain into the catchmentarea, which flows to the Pacific andprovides practically the only source ofwater to the six million inhabitants ofLima, the capital of Peru. Lima is onthe Pacific coastal desert and sincethere is virtually no rainfall, ground-water recharge is also ultimately de-pendent on the Rimac.

The river basin is mountainous, andhas two important sub-river basins, theSanta Eulalia and San Mateo rivers,which converge near the town ofChosica above Lima. The upper riverbasin has several lagoons, especiallyin the Santa Eulalia river. The SanMateo and Santa Eulalia rivers have4.34 per cent and 6.33 per centgradients respectively. Downstream ofthe confluence of the San Mateo andSanta Eulaliea rivers the gradient is 1.7per cent. At this point the valley iswide and the agricultural activity ishigh. Many small rivers drain to theRiver Santa Eulalia: the Pillihua, theYana y Potoga on the left bank, andthe Sacsa, Pacococha and Carpa riverson the right bank. Near San Mateo the

rivers in the UK. Sophisticated com-puting techniques are now employedby the river authorities to overcomemisleading results arising from thenatural variation of the physical andchemical changes in the river.

The biotic monitoring of riversusing these organisms has rarely beenattempted in developing countries,although it could provide an appro-priate, low-cost, and sensitive ap-proach to the detection of pollution.

The same indicator groups of organ-isms which are present in northernEurope are also present in the Andeanregion at high and low altitudes, andthey also disappear in response toindustrial and organic pollutants. Thebasic principle underpinning the meth-odology is that where biotic diversityis high the river is in a relativelyhealthy chemical state, i.e., well oxy-genated, low in polluting organics andlow in toxic inorganics. Converselylow biotic diversity indicates a gener-ally poor water quality.

Many Peruvian rivers have high-level organic and metallic contamina-tion from mining, industrial, and do-mestic activities. These same rivers arethe principal source of drinking-waterfor many towns and cities. By studyingthe fauna in the River Rimac, the

The domestic sewage which is discharged directly into the River Rimac is justone source of pollution.

Using biotic indicators to assess waterquality in Peruby Marlene Guerrero and Barry LloydThe contaminants that make polluted waterunsafe for humans are even more fatal to theorganisms that live in the river. Marlene Guerreroand Barry Lloyd explain how the presence orabsence of these crucial organisms can help usmonitor water quality.

TRADITIONALL Y, monitoring riverquality involves regularly samplingand transporting samples to labora-tories, where (mainly) chemical analy-ses are undertaken. The purpose ofthese analyses is to assess the safetyand suitability of the water for humanactivities, including domestic con-sumption. Monitoring programmesalso attempt to identify the sources andmagnitude of pollutants with a viewto controlling them.

During the last 40 years attemptshave been made in both Europe andNorth America to introduce comple-mentary biological monitoring to as-sess the effects of chemical and physi-cal changes on the life of rivers. It hasbeen demonstrated that organisms thatlive in river and stream beds are theorganisms most sensitive to suchchanges, and therefore the most usefulto monitor. Those used in this studywere all visible to the naked eye andeasily collected. They include a widevariety of groups such as fly larvae,worms, crustacea, molluscs, spiders,and flat worms. Whereas 'grab sam-pling' for chemical analysis is onlyrepresentative of the moment of sam-pling and often misses the intermittentpolluting discharge, sampling for bi-otic indicators (such as these organ-isms) is more informative and reliablebecause the organisms respond rapidlyand may be used to detect changesresulting from both intermittent andchronic pollution.

In the UK, biotic monitoring (themonitoring of living organisms) hasdeveloped to the point where there isa reasonable understanding of theecology of these groups. Biotic re-sponses, such as the loss of diversityand an increase in pollution-tolerantspecies, can now be related to chemicalpollution for at least 50 OOOkm of

Marlene Guerrero is a Peruvian biologist andlecturer in Trujillo University's EcologyDepartment, and Barry Lloyd is Head of theEnvironmental Health Unit of the RobensInstitute, University of Surrey, Guildford,Surrey, GU2 5XH, UK.

WATERLINES VOL.lO NO.3 JANUARY 1992 5

Stallona

CBS

1 2345678Figllre 2. Determiflotiofl of the Treflt Bintic Ifldex at each s/a/infl (meall"ollles for eight ampliflg occasions).

7 8core a/ each s/a/iOIl

-. . ..a Stations primarily influenced by

mining and related activities, butwell oxygenated (1,2,4).

o Stations primarily influenced bydomestic organic wastes, butoxygen depleted (5,7,8).

a Stations relatively unpolluted andwell oxygenated (3 and 6).

,Figures 2 and 3 demonstrate theagreement of the two indices for alleight sampling stations. Inspection ofthese figures reveals that the lowestindices are found at stations 1 and 2,i.e. those points principally affectedby mine wastes. Station 4 shows somerecovery, because it has received con-siderable dilution from the better qual-ity water received from the Rio Blancotributary represented by station 3.

The Rio Blanco is a tributary whichhas no mining in its catchment areaand little human colonization. As aresult it has relatively high indices, andthe diversity of these, and the relativeabundance, is summarized for foursample occasions in Figure 4.

The only superior site is station 6,which at first may seem strange sinceit is downstream of badly polluted siteson the main river. This apparentrecovery is readily explained, however,by the fact that most of the river flowbypasses station 6 in a man-madetunnel for hydroelectricity generation,and the water in the river at this pointis largely that from clean feeder springsin or near the main stream. Thus thetwo stations with the highest indicesare indeed those with relatively unpol-luted flows (3 and 6).

One species of Trichoptera wasfound at stations 3 and 6, and occasion-

2 3 4 5 6Figllre 3. De/ermifla/iOIl of /lre halldler Biotic(meafll'O/lIesjor eight sampling occasions).

500

7

6

5

4

3

2

o

ResultsChanges in the selected organisms'communities occur in response tochanges in available food, temperature,and erosional and siltration patterns,as well as to changes in the concentra-tion of agricultural, industrial, anddomestic pollution.

Changes in species composition andrelative abundance must therefore beinterpreted cautiously, taking into ac-count both natural as well as human-influenced factors. It was thereforeessential to characterize the river at theoutset according to previous conven-tional monitoring approaches.

For most of its length the flow of theupper River Rimac is rapid to torren"tial, with few slow, depositing areas.This means' not only that areas ofsiltation are limited, but that there arealso numerous zones with gravel, largestones and boulders.

Opportunities for the physical aeta-tion of the water to occur abound as aresult of cascading and turbulence, andtherefore oxygen is rarely'limiting. Asa consequence most attention may begiven to the direct effects of industrialdischatges in'the ifPi5ei catchment area:these are principally heavy metals andthe associated suspended solids., andpH changes ari'sing from mining. In thelower catchment area there are morevillages and towns discharging theirdomestic waste directly into the river.As a consequence the effects of organicpollution dominate, and oxygen dep-letion has resulted.

The sampling stations may thereforebe divided into three types:

The stations selected were asfollows:

Station 1 CasapalcaStation 2 ChiclaStation 3 Rio BlancoStation 4 San MateoStation 5 TamboraqueStation 6 Ricardo PalmaStation 7 ChosicaStation 8 NafiaThe locations of stations are shown

on the map in Figure 1. They werechosen because the sites are suffi-ciently shallow to permit the author toenter the river wearing waterproofboots. While facing downstream astandard 0.1m2 area of stream bed wasdisturbed continuously for three min-utes by shuffling over the gravel!pebble bed to release the sediment andorganisms into the flow, which werethen caught in a net with a fine(0.5mm) mesh. The catch was thensorted on site and the number of groupsand species were identified. The maingroups were sorted and the number ofgroups represented were counted ac-cording to the Trent Biotic Index.

Table I shows how the Trent BioticIndex is derived from identifying thediversity of the groups of indicatororganisms. The investigator counts thenumber of different groups present andidentifies the most sensitive grouppresent. If, for example, more than onespecies of Plecoptera is present, theinvestigator may enter the table at thetop level and move across that rowuntil reaching the column in which thenumber of groups identified is encoun-tered. The corresponding Index isshown at the top of that column. Thusthe highest Index (X) is found at thetop right of the table and the lowestindex (0) is found at the bottom leftof the table.

The main skill required for thismethod is the ability of the investigatorto identify and separate the key groupsand species. It is quite feasible for anenthusiastic school biology teacher totrain secondary school pupils to iden-tify the main groups in the course ofthree or four extended field excursions.Thus the Trent Biotic Index is a robust,economical, and straightforwardmethod to apply.

A slightly more time-consuming,skilled, and quantitative technique re-quires that as well as being sorted intospecies and groups, the actual numberof organisms in each group is counted(i.e. 20 snails, 30 larvae). This lattermethod, the Chandler Biotic Score,was also used in the study and theresults were compared.

WATERLINES VOL. 10 NO.3 JANUARY 1992 7

10JULY1990 17JULY1990

Figure 4. Percentages of sensitive species collected at River Blanco.

tion, and were therefore found at allstations which supported some life.They were also found in increasednumbers in the stations 6, 7 and 8.

Overall the abundance of the se-lected organisms varied enormouslyfrom station to station. The lowestmean value recorded was one at station2. The highest value of 310 specimenswas found at station 3. All counts werefrom a O.lm2 area of bottom deposit.

In this preliminary study we havenot attempted an advanced analysis totake account of natural physical andchemical variability in the river. Nev-ertheless, the general pattern of loss ofdiversity of sensitive species and lossof numbers is in agreement with theresponses to pollution found in north-ern Europe. Most of the work in thepast has focused on the response of theindicator groups to organic pollutionand oxygen depletion. Although morework is required, it would appear fromthis study that the same key indicatorgroups also respond to, and are there-fore valuable for, assessing the effectsof mine pollution. We therefore pro-pose that these methods should beapplied widely for the evaluation ofriver quality in developing countries,because they provide a potentiallyrobust and economical method formonitoring the effects of pollution.

New Publication: Liverpool Planning Manual 3

Services for ShelterANDREW COTTON and RICHARD FRANCEYSWater and Engineering Development Centre,Loughborough University of Technology

People who live in the rapidly developing urban areas of the ThirdWorld need housing. And housing needs the supportinginfrastructure of water supply, sanitation, drainage, access, powersupply and solid waste removal.

Services for Shelter describes, with extensive illustrations, alternativetechnologies for the provision of infrastructure to low-income housingwhich are affordable and which house-holders themselves canimplement and maintain. It describes techniques which may notalways suit conventionally accepted standards but which do meet theobjectives of infrastructure for health, access and security. The bookdescribes ways of involving low-income communities and localenterprises in services development and management. It reports onresearch that demonstrates the interaction between services on high-density sites. Particular attention is paid to the key elements ofdrainage and sanitation: drainage because the overall acceptabilityof a site is so influenced by the extent of flooding and standing water;sanitation because it is vital for health and because its costs are sosignificant. Dramatic reductions in the cost of services can beachieved through the use of on-site sanitation even in housing plotsas small as 30m2.£17.50, 160pp., A4paper, 1991,/SBNO 85323 287 3

LIVERPOOL UNIVERSITY PRESSPO Box 147, Liverpool, L69 3BX, UK

tary (station 3), however, both chiro-nomid and Simuliium larvae weresignificant components of the totalinvertebrate fauna. Simuliium spp,(black flies) are examples of fly larvaewhich are dependent on the high levelof oxygenation normally associatedwith fast-flowing streams. By contrastthe chironomids are more tolerant oforganic enrichment and oxygen deple-

1:. ~r~\.::,:d:::;~1"~m~4'V';' 13% _Simuliium13% Eph.Excel.Baelis 18%

17MAY 1990 21 JUNE 1990

ally at stations I and 4. The presenceof Trichoptera is associated with fastflow, adequate oxygen, eroding sub-strate and submerged vegetation, andof course the absence of significantamounts of toxic substances.

Plecoptera (a type of fly whoselarvae live in well-oxygenated water)were restricted to station 6 and absentfrom all other sampling stations. Ple-coptera species are the indicator groupfor the most sensitive species, and assuch are intolerant of any pollutioncausing the oxygen concentration to·be reduced.

Species of Mollusca (includingwater snails) were only abundant atstation 6. They are associated withincreasing calcium carbonate and saltlevels, and also with the erodingsubstrate. There was no evidence ofmolluscs at any other station.

Ephemeroptera (mayflies) were pre-sent, but poorly represented. Only twospecies were found at stations 3 and6. At all other stations the only mayflyfound was the pollution-tolerant Baetisrhodani. B. rhodani was present occa-sionally in the upper reaches (stationsI, 2, 4 and 5) and in high densities inthe lower reaches (stations 7 and 8).This mayfly has been shown by workworldwide to have some resistance tolow levels of organic pollution. The,Work in Peru shows that it is sensitiveto inorganic contamination though,such as that from mining activity.

A similar distribution to that ob-served for Ephemeroptera was alsofound for the Diptera (a group ofinsects). Again the groups representedwere limited and eliminated from theupper reaches where mine pollutionwas dominant. In the unpolluted tribu-

8 WATERLINES VOL. 10 NO.3 JANUARY 1992