Report Number 5 October 1986 FEWS Country Report NIGER

Africa Bureau U.S, Agency for International Development Map 1 Nat ional Population At-Risk, Sim' r Numbers & Locations NIGER: Summary Map September, 1986 Nationally: Morbidity & Mortality rM,, f5 Within Normal Levels

Agadez

Shistocerca gregaria. Zone of General (Desert Locust) Grasshopper Infestati ns

/" : !-J"/Maradi" <

Reductio in Dosso Crop Yields Reduction in Generally Good Crop Development Crop Yields Famine Early Warning System Country Report NIGER Anticipating A Good Harvest

Prepared for thq Africa Bureau cf the U.S. Agency for !nternational Development

Prepared by Price, Williams & Associates, Inc. October 1986

Contents Page

i Introduction 1 Summary 1 Rainfall and Vegetation 2 Map 2 3 Map 3 4 Agriculture 8 Map 4 9 Grasshoppers 10 Map 5 11 Market Prices 12 Graph Series 1,2 15 Health and Nutrition 16 Population At-Risk 17 Map 6 18 Map 7 iNTRODUCTION Th-s is the fif'h of a series of monthly reporti issued by the Famine Early Warning System (FEWS) on Niger. It is designed to provide decisionmakers with current infor­ mation and analysis on existing and potential nutrition emergency situations. Each situation identified is described in terms of geographical extent and the number of people involved, or at-risk, and the proximate causes insofar as they have been discerned.

Use of the term "at-risk" to identify vulnerable popu­ lations is problematical since Po generally agreed upon definition exists. Yet it is necessany to identify or "target" populations in-need or "at-risk" in order to determine appropriate forms and levels of intervention. Thus for the present, until a better usage can be found, FEWS reports will employ the term "at-risk" to mean...

...those persons lacking sufficient food, or resources to acquire sufficient food, to avert a nutritional crisis, i.e., a progressive deterioration in their health or nutritional condition below the status quo and who, as a result, require specific intervention to avoid a life-threatening situation.

Perhaps of most importance to decisionmakers, the process underlying the deteriorating situation is highlighted by the FEWS effort, hopefully with enough specificity and forewarning to permit alternative intervention strategies to be examined and implemented. Food assistance strat­ egies are key to famine avoidance. However, other types of intervention can be of major importance both in the short-term and in the long-run, including medical, trans­ port, storage, economic development policy change, etc.

Where possible, food needs estimates are included in the FEV/S reports, it is important to understand, however, that no direct relation exists between numbers of persons at-risk and the quantity o' food assistance needed. This is because famines are the culmination of slow-onset dis­ aster processes which can be complex in the extreme. The food needs of individual populations at-risk depend upon when in the disaster process identification is made and the extent of its cumulative impact on the individ­ uals concerned. Further, the amount of food assistance required, whether from intcrnal or external sources, dcpcnds upon a host of considerations. Thus the food needs estimates presented periodically in FEWS reports should not be interpreted to mean food aid needs, e.g., as under PL480 or other donor programs.

FEWS is operated by AID's Office of Technical Resources in the Bureau for Africa in cooperation with numerous USG and other organizations. SUMMARY The agricultural scason should end relatively well for Niger. Most of the country's gross cereal needs will probably be met by this tear's harvest. These results will vary iegionally, with farming zones in the south faring better than those further to the north. Some farming areas in more northern, traditionally marginal areas, experienced extended dry periods and late-season attacks of grasshoppers and other pests which will reduce yields. Some of them will harest less than 30% of the cereals they need for a year. Mortality and morbidity data show a cilm situation, yet a lack of good data on the nutritional status of vulnerable populations may hide on-going problems. Populations at-risk are similar in number and location to those last year, at about 530,000 people, mostly in the north and the east. A program of food assistance at about the same levels as last year in those areas will still be necessary.

Issues o The late-season grasshoppers and locusts now found in Niger can not be effectively treated this year. Late­ season committments by donors for grasshopper control would, howevcr, be useful prior to, and during the next rainy season to control eggs just laid. Planning for prevention rather than cure should begin now.

Key October Events o Yield estimates based on local information will be available late in October. if they show a different picture than that based on satellite data, a re­ examination of the national food needs would be nccess ry.

o If rains continue into late September and the first three weeks of October, many of the areas which were late in planting, or where replantings occured due to pest damage, will not suffer from large crop yield reductions.

RAINFALL and VEGETATION Rainfall during the season has been generally less than that expected from the 30 year historical record. This is true for all areas except the southcrn-mo. t portions of the Niamey and Dosso Departments, the southern half of Maradi Department, and a west-central pocket in the Zinder Department (see Map 3). A look at the map shows how little of the country has had more than 80% of the levels expected compared to the historical record.

Despite this, rainfall has been fairly regular in spacing, and generally sufficient for crop needs. It vias Map2

NIGER: Reference Points

. \ Bi lim

/ lferotm\ "ua

BrAir u n t a i n s / 1 2 50kin 'S

, ; --- Desert "-

e Tchin- Tbrzrn - he " Tabaraden

16N ------­ - - - ­

15N---- n l lan TauousJ iaou Gor MOSE

13 ----Nt-- Rudi Maadi Tessaoua Lake Mad

1 2N ---­ - M - oure - - -G ya FE S W # 5 4E 6E 8E 1OE 12E Map 3 V

NIGER: Rainfall & GRIN OF"PASTURES-- -

Ve g e t a t io n T ah o u .n TO DR- 1 5 N

Maradi Zinder 14N September 30, 1986 13N CROP iLD U BIOASS GNERALLY SIMILAR TO BIOMASS LE S-DELOP 1 IDtrYIONS DUE TO S ige5 TrAN IN 1o6 12N EARLY IN SEASON

CUVi.ATIE RAINFALL AS PERCENT OF 30 YR RCR

RAINFALL STATIONS A. TILLABERY B. NIAMEY 72 C. GAYA 15N TAHOUA 7 D. 5 E. BIRNI NKOMI

B * __80_ 61 14N F. MARADI _0. ZINMR t8_____I13 H. MAGARIA o I. MAINE SOR A 2--'<9 80ox 12N J. N'omu

0 2E 4E 6E BE 10E 12E FEWS/FWA #5 somewhat late in arriving in many areas, and set the growing season back from 2-4 weeks. If rain continues into latc September and the first three weeks of Octobcr, most areas which were late in planting will not suffer.

One finds the major exceptions to minimal crop moisture needs being met in the northern half of Niamey Department and in Diffa Department, where crop yield reductions are likely. These two areas have also had somewhat more of a problem with pests than other areas. In northern Niamey Department, heavy rains came in the early part of September, yet they were too late to improve the crops. Pastoralists, however, will benefit from a late greening of pastures due to these rains.

Satellite images show that the vigor of the vegetation is generally similar to that of 1985. Since 1985 was a relatively good year, in terms of harvest and pastures, this tends to reinforce the positive forecast for the food crop and for the availability of animal fodder. Only a complete cessation of rainfall in October could reverse the generally positive look of things now.

AGRICULTURE The agricultural season comes to an end in October. Although there is still some risk of change for the worse in these final weeks, the harvest will likely be one of the better ones in thcss last few years.

This has not come without some amount of drama. Much of the millet and sorghum crop was planted I to 4 weeks late due to a late onset of rains, or to replantings made necessary by pest damage. Rainfall since then has been only adequate in most areas, sometimes with a mix of dry stretches and heavy downpours. Soil moisture amounts fell several times to minimum crop-sustaining levels in many areas.

Diffa Department, a marginal crop producing area in the best of years, has suffered throughout this season from the effects of low, and poorly-spaced rainfall. Another area of particular crop stress includes the northern half of Niamey Department. Areas with a generally positive look to the agricultural campaign include the southern half of Maradi Department, and scattered areas in Tahoua Department.

NOAA YIELD ESTINIATES - The initial 1986 National Oceanic and Atmospheric Administration (NOAA) forecast of millet and sorghum crop yields in Niger arc shown in Table I and 2. This forecast was made on the basis of satellite and

4 rainfall data and does not reflect potential crop pest damage.

TABLE 1: NOAA MILLET YIELD FORECAST

NOAA DEPT. AVG 1984 1985 1986 86to 86/85 86/84 Yield Yield Yield Yield AVG % % KGS KGS KGS KGS Yield Yield Yield

Niamey 380 250 440 320 84 73 128 Dosso 400 270 450 380 95 84 141 Tahoua 390 230 450 510 131 113 222 Maradi 400 270 480 450 112 94 167 Zindcr 430 280 480 460 107 96 164 Diffa 320 2 480 300 94 63 1600 Agadcz NA NA NA NA NA NA NA

TABLE 2: NOAA SORGHUM YIELD FORECAST

DEPT. AVG 1984 1985 1986 86 to 86/85 86/84 Yield Yield Yield Yield AVG % 0 KGS KGS KGS KGS Yield Yield Yield

Niamey 380 330 340 380 100 Ill 128 Dosso 390 340 440 410 105 93 120 Tahoua 450 350 370 570 127 154 162 Maradi 190 150 200 270 142 135 180 Zinder 260 160 300 330 127 110 206

Source: NOAA 9/15/86 Yield Forecast

The sixth column in both tables shows the estimated 1986 yields as a x.rcentage of the average 1981 to 1985 yields. It shows a generally positive trend in 1986 yields compared to previous years, with only Niamey, Dosso, and Diffa falling below the previous years' average in millet yield. Sorghum yields, on the other hand, are all better than the average. Compared to 1985 (see column 7), which was a very good crop year, most of this year's yields are down slightly in millet, and better in sorghum.

Particularly noteworthy in these forecasts are the Niamey millet estimates, which show yield reductions in this important grain producing department. What this depart­ ment-level forecast does not show is the wide intra­ departmental variation in yields from the north;., rn half of the dcpartmcnt to the southern half. Northern yields

5 are normally much less than southern ones. Most of the crop stress this year has been felt in the northern half.

Gross millet production (using these yield figures) will probably be greater than the 84% of average for the department as a whole, while yields will probably be less than the 84% of average in the northern half of the department.

Estimate of 1986 Millet and Sorghum Production - Using the NOAA forecast of estimated yields and FAO average crop area figures for both these crops, a preliminary assessment of 1986 millet and sorghum production can be made. Table 3 shows the results of this analysis.

TABLE 3: Estimated 1986 Millet and Sorghum Production

DEPT. MILLET SORGHUM TOTAL 85 86 85 86 85 86 (MT) (MT) (MT) (MT) (MT) (MT)

Niamey 353 257 33 36 386 293 Dosso 272 228 25 24 297 252 Tahoua 196 222 88 135 285 357 Maradi 284 267 90 122 374 389 Zinder 280 269 100 110 380 379 Diffa 26 16 NA NA 26 16 Agadcz NA NA NA NA NA NA

TOTAL 1,411 1,259 336 427 1,748 1,686

An interesting picture emerges from these figures. While the millet crop should produce 152,000 MT less millet than in 1985, the sorghum crop may be 91,000 MT larger than 1985's crop. The combined 1986 millet and sorghum crop would only be 62,000 MT less than in 1985.

Also particularly noteworthy from these figures is the extra importance oi Tahoua Department's production this year. While it ranked fifth in terms of millet and sorghum production in 1985, it will, according to this forecast, be ranked third this year, and very close to Maradi and Zinder in total production. This goes contrary to many perceptions about the Department of Tahoua's contribution to the national crop.

Per Capita Grain Production Forecast - Given the above figures, and using a loss rate of 15% for millet and sorghum, a rough initial estimate of per capi i grain production can be made for each department, and on a national basis (see Map 4). A net rice production figure of 21,000 MT (same as actual production in 1985) was added to the Niamey Department cercal production total for this analysis.

TABLE 4: NET PER CAPITA CEREAL PRODUCTION

DEPT 1986 NET Per Capita Population Production Production (MT) (KG)

Niamey 1,829,779 314 147 Dosso 893,937 214 239 Tahoua 1,174,946 303 258 Maradi 1,282,863 331 258 Zinder 1,374,523 322 234 Diffa 197,484 14 7 Agadez 211,196 0 0

TOTAL 6,965,628 1,498 215

Source: Population Estimate from FEWS, September, 1986

As the above table shows, four of the departments produce more than the commonly used minimum standard of 167 kg per person. Estimates of the Nigerien diet usually show that actual normal consumption of cereals like these usually is more than 220 kg per capita per year. On this basis, and on a national scale, the per capita cereal needs would largely be met by this year's production.

The regional variance in per capita production is very striking. Agadcz and Diffa are always food production deficit areas, and this year's figures only confirm that fact. Niamey Department's figures should be viewed in light of the comments made above about the yield estimates. The per capita food deficit in this area might be slightly less than shown here, but great differences will be found within this Department.

Other factors which affect the per capita food situation presented here are important. The production of njebe, or cowpeas, is ;ignificant in Niger (net production of 102,000 MT in 1985) and helps to cover a good part of dietary needs for calories and protein where it is consumed. Illegal exports, and sometimes imports, of grain to Nigeria and elsewhere have historically been significant factors in reducing or increasing the amount of grain available nationally. Security stocks, and food

7 Map 4 NIGER: Estimated 1986 Cereal Production

(Millet, Sorghum, Rice)

NET PRODUCT ION

'000 wT 0" O0TO 200 200 TO 300 300 TO 400 303

2 331 322

PER-CAP ITA PRODUCTION

KG Per Capita ~0 TO 100J 100 TO 200 25 200 TO 300

239 258 234

Source: NOAA Yield Estimates 9/15/86 8 reserves held privately and by the government, are not included in this analysis. Transportation and distribut­ ion of food supplies are not without problem or cost, and tend to limit interdepartmental transfers. All of these factors aside, the food supply made available by this year's grain harvest should be capable of meeting most national and sub-national needs.

GRASSHOPPERS For so much of this rainy season, Niger stood out from its neighbors by its lhck of a serious grasshopper problem. While hundreds of thousands of hectares were either being threatened, infested, and/or treated in Mali, Mauritania, Senegal, Burkina, and Chad, Niger's monthly agricultural reports talked about small hundreds or thousands of hectares, most of which were quickly treated by aerial spraying.

The situation changed in September. As some grasshopper experts using computerized simulation models had earlier predicted, hatching picked up in September (see Map 5). A large number of previously uninfested areas reported large numbers or grasshoppers in wild vegetation and in fields. Increased problems were also reported in areas around those which had earlier been treated, for example in Ouallam and Diffa.

A good-sized population of desert locust (Shistoccrca 9renaria) was reported to be in a traditional breeding ground in the area between Iferouane and Bilma. A large advancing group of migratory locust (Locusta migratoria) was reported somewhere in "an undetermined area near Lake Chad". Grasshoppers (Oedalcus scnealensis), in greater than previously reported densities, were reported all along the northern limit of cultivation, which runs roughly between the J5th and 16th degrees of latitude.

The Government of Niger (GON) made a quick reassessmeni of the resources available for combatting this renewed threat. It found that if these threats were confirmed, present resources would be insufficient to cope. Donors were convened by the GON and responded relatively quickly with promises of increased aid. These promises include support for an increase in aerial spraying by small planes.

Early in October, however, reports from field teams investigating the size and nature of the threat have not confirmed that it is imminent, nor that it will cause ecoiromically important damage this year. There could still be some amount of damage to late-planted crops, and to niebc (cowpeas) and groundnut fields. Additionally,

9 Map 5

NIGER: Grasshopper & Locust Problems September 31, 1986

~Shistocerca gre |aria

D serl Locust)

Low Dens it y Infestation Zon

August, 1986 45-50,000 ha Treated

Areas Treated All Locations Approximate YEAR-TO-DATE Surface Areas Not Exact 160,000+ ha Treated

0. the report of locusts coming from Lakc Chad secms not to have any basis.

The threat now appears to be greatest next year, as eggs laid this year hatch. The presence of large numbers of eggs in the northern cropping zones, and of desert locusts in the north will require that preparations be made for protection next year. If donor committments made hastily and in a crisis mode late this season can be reprogrammed for next year, the nature of the threat could be greatly reduced.

MARKET PRICES Millet, sorghum, rind rice are the primary components of the diet for all but the most nomadic of Nigeriens. As such, market prices for them are critical indicators of the national well-being.

There is a pronounced seasonal c:,;le in prices. Normally, prices will be at their lowest just after the harvest, or from November to Jnnuary or February. They will be at their highest in the period called the "soudure", or pre­ harvest period, usually running from June to September.

A deviation from, or exaggeration of these general patterns can be both a confirming post-harvest indicator of the actual size of the harvest, as well as an early pre-harvest indicator of perceptions about the size of the coming harvest.

There are, of course, factors wnich modify these general trends besides simple internal supply and demand, such as food aid, illegal small-farmer exports of grain to neighboring countries, and government pricing policies. The impact of these other factors is reflected, but not quantified, in the price data shown below (see Graph Series I & 2).

Graph Series #1 shows millet, sorghum, and rice prices over the last nine years for Niamey, Dosso, Maradi, and Zinder departments. Prices in the other three departments are generally higher than these because they are less favorable farming areas, but the general pattern of price movement is similar.

Graph Series #2 extracts only the millet price data for 1st and 3rd quarters. It more clearly shows price trends and how they relate to the pre-harvest (3rd quarter) and post-harvest (1st quarter) periods.

Notable in all the graphs is a general peaking of prices from mid 1978 to late 1978, mid 1981 to mid 1982, and II Graph Series 1 MILLET PRICES, 1978-86 Niger: Cereal Prices, 1978- 2IAM.,.-o.R., ,°­ 86 2

10

" u 180

140

120

S 100

80

40

RD ­

78 79 80 81 82 83 84 8 86 1078-86 by QUARTER 0 NIAMKY + DOSSO 0 MARJAI A ZINDER

RICE PRICES, 1978-86 SORGHUM PRICES, 1978-86 2L&MIY. DOSO. MARDI, ZINDER 23,IAMET, DOSSO, MAR.DI, ZINDER 280- 230­

210 240 -D20 0 220 - 190 180 20 - 170 l + 180 18+ 150 uU 10 140 / 130 140 12 120 N 110 | IG 100c A. 10 A. 90 so 80 70 80 80­ 40 50 40

78 79 82 83 84 85 88 78 79 so f1 82 83 45 Is

NIAMY 1978-88 by QUARTER 1078-85 by QUARTER 0 + Doso 0 MARADI a ZINDER 0 NIAMEY + DossO Q ARADI AZIND­ Graph Series 2 MILLET PRICES, Ist & 3rd QTRS, AGADEZ

Niger: 1ST & 3RD Quarter 280-8 Millet Prices, 280 1978-86 18loo Oot,

140 "

U 120 O

40 20 78 79 80 81 82 83 4 85 8 lot 3rd QTR AVG PRICUS. I978-86 S O*t QTR + ad QTR

.4..

z ." MILLET PRICES, 1st & 3rd QTRS, DIFFA MILLET PRICES, 1st & 3rd QTRS, DOSSO 300 - 300 986

280 -280

250 -280

200 240 240 a I I a i I a a I I a I[ I I .

2o 22

200 * " "( ....

K Le + 18 0t 0 860 ­ to80 ++.

80 80

00 40

78 79 80 81 82 83 84 85 8s 1,t h 3rd QTR AVC PRIMS. 197e-86 l kt 3rd QTR AVG PUIC15. 1978-88 IQTR0o +. 2nd QT 0 lot QTR + 2nd Q1R MILLET PRICES, 1st & 3rd QTRS, MARADI MILLET PRICES, 1st & 3rd QTRS, NIAMEY 10'78-88 1078-88 30 3O

280 280 4­

240 t4­

220 " 2no

200 - - *

140 " 140 0-U" 4. ... U 120 U It " 100 - 100 .

80 - . ­ - 80 - "3...

80 " *" 60 ; 04 40 40

2. a a , 2 ,,,,,.,i ., p 78.a 79 80 81 82 83 84 86 78 70 8T 81 82 83 84 if 86 lt, & 3rd QTR AVC PRIC23. 1078-86 lt & 3rd QT2 AVG PRICIS. I878-86 1 1st QTR + 2nd QT2 a lot QTR + Rnd QI from late 1984 to mid 1985. The prices for millet and sorghum in these periods are from two to three times what they are at other periods.

The cause of the rise in prices in 1984-85 can easily be attributed to the extremely poor 1984 harvest, where only 748,000 MT of cereals were produced (see Table 5). The millet and sorgum price graphs show a rise in prices which becomes clearly accentuated several months before the harvest in the fourth quarter, accurately foreshadow­ ing the later reality of a poor harvest.

TABLE 5: Millet & Sorghum Harvests, 1978-85 ('000 MT)

1978 1,493 1979 1,606 1980 1,731 1981 1,635 1982 1,649 1983 1,687 1984 748 1985 1,783 1986 est. 1,686

The peak in prices in 1978 and 1981-82 also predicts, if less strikingly so, downward dips in harvests in those years. Rice, much of which is imported and distributed through government contolled channels, shows less volatility in price, while still mirroring the general patterns shown in other grain prices.

The other striking feature or these graphs is the downward trend of prices into 1986. In most cases, 1986 prices are similar to those in the late 1970's, and two to three times less than in 1981-82. If this year's fall in prices is as accurate an indicator as it has been in past years, this year's harvest will be relatively good.

HEALTH and NUTRITION Current, regular morbidity reports show a very calm surface of normal levels of disease and sickness. As is normal, there are a great number of cases of malaria, diarrhea, measles, and a great number of other assorted health problems, some of which result in a number of deaths. For example, there were 109 malaria-related deaths reported during a six-week period this summer. The number of cases of diarrhea, largely in children, which are reported is only a small fraction of those which

15 occur. Yet there were over 32,000 cases treated during this same six-week period.

Despite this unfortunate normality, there is a disturbing gap in current, trustworthy nutrition data which might belie the surface calm of other health indicators. Early results from a study being completed by CARITAS, a relief agency, show relatively high rates of malnutrition in the areas surveyed. More information on the results of this survey will be available next month.

POPULATION AT-RISK The GON estimate of Population At-Risk defines those populations as being from villages having a food deficit of 70 per cent or greater, based on millet and sorghum production, and those nomadic and/or displaced persons having no adequate animal or other resource able to provide adequate food supply. The following estimate is current as of mid-September, and uses population figures supplied by FEWS/Niger.

There will be little change from last year in terms of number of people at risk, and their locations. In Niger as a whole, (see Map 6 and 7) approximately 7.6 per cent of the population, or 530,370 people, qualifies as being "at-risk" under this criterion. Of this total, approximately 244,321 (3.51 %) are sedentary farmers, and 286,049 (4.11%) tire herders with insufficient food resources, or arc displaced persons.

Geographically, the dispersion of sedentary populations at-risk is restricted to a few areas, most of them found above the 14th and and below the 16th degrees of lati­ tude, or the northern-most areas of cultivation. Princi­ pal arrondissmcnts and sedentary populations at-risk among these include Ouallam (36,608) and Tillaberi (53,083) in Niamey Department, Illela (47,581) and Tahoun (33,481) in Tahoua Department, and Dakoro (6,500), Guidan Roumdji (15,850), Mayahi (13,086), and Tcssaoua (5,877) in Maradi Department.

The nomadic populations at-risk are found in primarily in areas north of those enumerated above. They include Ouallam (12,500) in Niamey Department, Dakoro (8,000) in Maradi Department, Goure (20,000), Magaria (18,780), and Tanout (48,000) in Zinder Department, Tchin Tabaradcn (124,389) in Tahoua Department, Maine-Soroa (36,700) and N'Guigmi (24,000) in Diffa Department, and Arlit (30,000) and Tchirozerine (90,000) in Agadez Department. The displaced populations arc found primarily in, or around larger urban centers like Maradi, Zinder, and Agadcz.

16 Map 6

NIGER: Location of Major Populations At-Risk

SEDENTARY Arrondissement 1. Tillabery 2. Oual lam 3. Tahoua 4. Illela 5. Tchin­ 2 aTabaraden 6. Dakoru 7. Guidan Roumdj 8. Mayahi 9. Tessaoua 10. Magaria 11. Tanout 12. Goure .D...... 13. Maine- DISPLACED .Soroa 14. N'Guigmi ...... 15 . Ar I it 16. Tchiro­ zer ine

17

FEWS/PWA 05 Map 7 NIGER: Population At-Risk

PER CENT POPULATION AT-RISK

100 TOTO 10 40 40 TO 80 80 TO 100 80

NUMBER POPULAT ION AT-RISK --­

0 TO 5000 5000 TO 2500 7~~5000 TO 75000 ' "-'-'­ ''l~l~l 75000 TO 124380

rxw~ftM 18 Most of these areas will probably b- found on this list no matter the year, or the size or the national harvest. Because the geographic pattern of harvest results this year closely mirrors the results of 1985, it is relatively safe to predict that food aid programs will again be needed in the same areas as last year, and at about the same levels.

19