ASSOCIATION BETWEEN RAINFALL VARIABILITY AND CEREAL OUTPUT IN GOZAMEN WOREDA, NORTH CENTRAL ETHIOPIA (2005-2014)
DOI:10.30825/5.EJPAU.35.2017.20.4

Getie Fentie¹, Tsetadirgachew Legesse^2
1 Debre Markos preparatory school, Debremarkos, Ethiopia
² Adama Science and Technology University, Ethiopia

ABSTRACT

The study assessed the association between rainfall variability and cereal output in Gozamen Woreda. It employed mean, standard deviation, the coefficient of variation and precipitation concentration index measures in analyzing the variation of rainfall over time. Correlation analysis was used to establish the association between inter-seasonal rainfall variability on each of the six major cereals output for the last ten years (2005–2014). Rainfall data analysis revealed a general decreasing of spring, summer, and annual rains, while autumn rainfall showed increasing trend. “Teff” registered more or less constant performance. Sorghum and “engido” exhibited declining performance, however maize and wheat had an increasing performance. Teff and barley negatively correlated with summer rainfall whereas maize and sorghum revealed a strong positive correlation with spring rainfall. “Engido” recorded negative correlation with autumn rainfall. From the result it is suggested that having recognized the physical potentials and demands, efforts should be made in areas of irrigation, traditional soil fertility maintaining practices. General environmental resources, conservation works should be done  to realize sustainable development in the cereal production sector in particular and overall efforts of the inhabitants of the site in general. (182/250)

Key words: Rainfall Variability, Cereal Output, Gozamen Woreda.

INTRODUCTION

Rainfall variability is the most important climatic condition affecting the wellbeing of man, particularly in tropic areas. The degree to which rainfall conditions vary across an area or time is an important characteristic of the climate of the area. The study of Harvest Choice pointed out that there is the high difference in crop yield from the same amount of rainfall of a particular area if its seasonal distribution varies. Low availability of water during a critical stage has a greater impact on output [5]. The overall trend is that more reliable rainfall correlates with higher yield and low yield variability [1].

The impact of rainfall variability is more important in Sub-Sahara, including Ethiopia where traditional crop farming is predominant. It degree to [10] that 95 percent of the food production of Sub-Sahara is the result of rain-fed farming. Here, the nature of rainfall directly reflected on the number of crop yields. Therefore, increase in variability of conditions vary decreases the productivity of crops, since much of the rainfall climate of as run-off [6]. Africa uses only 15 to 30 percent of the rainfall.

The temporal rainfall variability is generally the single most important determinant of cereal output in Ethiopia [9]. According to [7] a, 10 percent decrease in seasonal rainfall from the long-term average causes 4 percent decrease in the country’s food production. To this fact [4] added that both negative and positive anomalies can affect the yield of cereal due to surplus and deficit of water in the soil respectively. Although it is a national problem the impact of rainfall variability is greater in the northernmost populated and degraded areas, and in semiarid and arid lowlands of the country [3]. The output of major cereals produced in the region showed a similar pattern of variation as the seasonal and annual rainfall variability. Different crops were cultivated by farmers in the study site in different agro-ecological zones because of strong orientation towards self-sufficiency [8]. Today shortage of farmland forced farmers to concentrate on cereal production.  Cereals are an important source of stable food and income for farmers. Source of income outside farming is insignificant for farmers in the study site. 

Some projections indicated that even in the future rainfall variability would be a developmental challenge in Ethiopia. According to [11] in the coming 50 years, climate variability is expected to reduce the average income of Ethiopia by 30 percent. This will worsen the poverty situation in the country in general and the inhabitants of Gozamen Woreda in particular. Thus, the understanding of the association between cereal output and rainfall variability is needed to cope with expected changes in the socio-economic and environmental areas (aspects). This is the main rationale for selection of the study topic.

The main objective of this study was to:

• assess the association between rainfall variability and cereal output in Gozamen Woreda, North Central Ethiopia;
• explain the impact of rainfall variability on cereal output in the area.

MATERIAL AND METHODS

The study was carried out in Gozamen Woreda, which is astronomically located between 10° 00' 18'' N to 10° 39' 00'' N and 37° 24' 00'' E to 37° 53' 00'' E in the northern part of Ethiopia about 299 km in the Northwest of Addis Ababa, the capital of  Ethiopia (see Map 1).

Data collected, organized and analyzed by both qualitative and quantitative techniques. The long-term rainfall and cereal output analysis was the function of descriptive statistics. Means, standard deviation, and coefficient of variation are used to analyze the trend of rainfall over the study period and assessed output of major cereals cultivated in the study area. Precipitation Concentration Index (PCI) also used to assess the interannual, seasonal and monthly rainfall variations of the study area. To analyze the association between rainfall condition and the output of major cereals (teff, wheat, barley, “engido” and sorghum) correlation computation is made the Precipitation Concentration Index (PCI) values calculated as follows as indicated in [9b].

Where:
PCI – Precipitation Concentration Index
Pi – the total rainfall on the ith period
∑ – summation over the period

In this analysis, PCI value lessens 10 indicates uniform distribution while values between 12 to  20 indicate high concentration and those above 20 indicate very high rainfall deviation.

Rainfall variability over a period of time l also analyzed with the help of descriptive statistics. The standard deviation (SD) and coefficient of variation (CV) are computed to analyze the nature of inter-annual and seasonal variability of rainfall as follows              

Where
∑ – summation over the period,
Xi – rainfall of the i-th period
 – mean rainfall of the ith period
n – the number of years the rainfall data computed

In short, the coefficient of variation (CV) is computed by dividing standard deviation of the respective period by its mean. In this analysis, the coefficient variation value closed to one indicates highly variable distribution while zero indicates perfectly uniform distribution.To assess the association between rainfall variation and the output of measure cereals cultivated in the site, long-term inter-annual and seasonal total average rainfall are correlated with the output of each cereal. Two variables, rainfall totals, and outputs of cereals statistically computed and tabulated then geographically inter prated.

Factors having an influence on cereal output other than rainfall variability anomalies can be described. The general biophysical features of the study area, socioeconomic situation, the experience and perception of the community towards the impact of rainfall variability on cereal output generated from key informants and focus groups are analyzed qualitatively. The result of the analysis is presented in tables and graphs. The general description is written under each presentation. Qualitative information has been directly quoted and paraphrased and followed by relevant decrease in of the ideas and views of the respondents.

RESULTS AND DISCUSSION

The Trend of Rainfall in Gozamen Woreda
The result of analyzed data revealed that the annual average total rainfall of Gozamen Woreda is 1315.9 mm with the standard deviation of 97 mm and coefficient of variation of 7.02 percent. The general trend of the annual and seasonal rainfalls of the study area are shown in Figures 1 and 2.

Concerning the temporary rainfall trend of Gozamen Woreda, the respondents of the study area explained that although the interannual rainfall variability is not clearly recognized in the past thirty years, the vivid variation has been observed in the onsite and offsite of rainfall. The decreasing trend of both spring and summer rains becomes obvious. Spring was as the season receiving the second heaviest rainfall of the thirty years under investigation, 16 years showed rainfall above the long-term average annual total rainfall, while the remaining fourteen years showed a record annual rain below the long-term mean. Between 1992 and 2010 the annual rainfall had been above the long-term mean excepting the years of 1995, 1998, 2003, 2005 and 2009. The highest annual total rain was recorded within this period (1992–2010). The wettest year of the thirty years period (1985 to 2014) in Gozamen Woreda is 1997 followed by 2006 and 2001. The year 1991 exhibited the lowest annual rainfall record. It is found to be the second worst drought year in the recent past in the Woreda after 1984.

Of the thirteen years registered total annual rainfall below the total mean for the thirty years period only four were in the last ten years (2005–2014) time. In other words, the period 2005–2014 accounted for only about 26.7 percent of the years with less than total means in the study site. Generally, the rainfall, of the last year's period showed a slightly increasing trend. For the matter of comparision, the first two ten year's time span – 1985–1994 and 1995–2004 recorded 32.5 and 33.6 percent respectively while the period 2005 to 2014 exhibited 33.9 percent of the total mean rainfall of the thirty years.

The statistical data indicates that the annual distribution of rainfall in Gozamen Woreda is more or less unimodal. It begins in spring and continuous usually up to end of September reaching its apex in August.

As shown in table 1 much of the annual rainfall of the study area falls in summer (June, July, and August), which accounted for about 57.1 percent of the average annual total rainfall in the thirty years under investigation. Of months of the year, August is found to be the wettest accounted for about 22 percent of the total annual rainfall of the study area followed by July which shared about 20.8 percent of the average total annual rainfall. February showed the lowest share followed by January and December. December, January and February represented only 2.7 percent of the average total annual rainfall. The four months such as June, July, August, and September recorded about 75 percent.

The result of the rainfall variability statistical data analysis prevails that for the ten years period (2005–2014), for which cereal output data obtained the precipitation concentration index (annual) showed 10.1 percent value. As computed based on mean seasonal rainfall data depicted in table 1 seasonal PCI is found to be 42 % indicating the presence of to extreme rainfall seasons.

According to Table 2, the result of mean annual and seasonal rainfall variability analysis of the period under observation (2005–2014) indicates that five of the ten years recorded positive anomalies from the mean rainfall of the ten years period. 2006 showed the highest total annual rainfall positive deviation from the mean followed by 2010 (refer Tab. 2).

The year 2009 emerged as the driest out of the other years showing the highest negative deviation from the average mean of the total rainfall. 2012 was registered as the second lowest rainfall. Spring showed the highest negative anomaly of the four seasons of the years. Summer rainfall showed decreasing trend recently. The latter five years (2010–2014) registered 414 mm rainfall less from the first five years (2005–2009). Autumn rainfall showed increasing trend. It revealed 307 mm more for the last five years. Although the statistical data did not reveal clearly the decreasing trend for spring average mean annual rains, the discussants notified the significant variability on the onset that observed in spring rainfall in Gozamen Woreda

The impact of rainfall variability on cereal output in Gozamen Woreda
The results of statistical analysis revealed that the annual total rainfall of Gozamen Woreda in the thirty years of period ranges from 1,004.7 to 1,517.0 mm.

The ten years rainfall for which cereal output data obtained (2005–2014) showed the very low annual variability in the study area. The balance between the positive and negative deviations from the average total mean rainfall of these years recorded zero. The highest positive deviation from the average total mean rainfall of the ten years is only 18 percent while the highest negative deviation computed at 14 percent (see tab. 2). Even for spring rainfall the variation in timing and distribution is more important than gross total annual records. This high variability of rainfall during spring creates a negative impact on farmland preparation, farming patterns and sowing-time especially of warm weather and long cycle crops, maize and sorghum. In years when spring rainfall is timely and adequate, farmers prepare their farms properly and get a chance to produce the appropriate cereal on the appropriate field on time. 

In some years rain occurs on time even beginning at the end of February and beginning of March. But its continuity becomes abnormal. In this case, farmers saw maize and sorghum. But, their growths were interrupted due to discontinuity of rainfall. That causes lose of seeds and other agricultural inputs. Moreover, it negatively affects the correct farming pattern preference of farmers. However, today, due to the introduction of new seeds and technology the sowing period of maize is prolonged up to the end of May especially in the southern and warmer part of the study area. The autumn rainfall shows increasing trend. be described results of statistical analysis related to the association between rainfall variability and cereal output in Gozamen Woreda, the study area, may not be that much significant for interpretation. This is due mainly to the short period of time (only ten years) of the output data used as well as the non-linear nature of the association between rainfall and cereal output. Besides, rainfall is not the only factor that determines the output of cereals. Above all, it should be taken into consideration that Gozamen Woreda is not water stressed. This can be observed from the climate data of the period (2005–2014) under investigation. The driest year of the ten years period, 2009 recorded an annual total rainfall of 1,156 mm (source). Therefore, what is most important in the study site is onset and offset of rainfall than mean amounts. The key informants notified that very heavy summer and autumn rainfalls cause loss of yield by interrupting the proper germination and growth of cereals. Extended heavy rainfall during autumn is considered as destructive to cereals production. This is clearly observed on the output of “engido”. Most of the years with better “endigo” output are those with negative autumn rainfall anomaly.

The performance of cereal production in Gozamen woreda
Teff, wheat, barley, “engido”, maize, and sorghum are major cereals cultivated in Gozamen woreda. Teff, which is the most important staple food in the study area, showed the lowest variations in every aspect of major cereals production. It is widely cultivated over the larger part of the study area except for the Dega(temperate) and Wurch (alpine) agro-ecological zones. As shown in Table 3 Teff recorded annual output above its mean value in five of the ten years under study. The largest is recorded in 2014 (+23 percent above the mean) followed by 2013 (+12 percent). The largest negative deviation from the mean was in 2007 (-35 percent). The size of land cultivated was positive in six of the ten years. It was largest in 2012 (+16 percent) followed by 2014 (+9 percent). The largest negative deviation from the mean of cultivated land for teff was in 2005 (-16 percent) followed by 2007 (-15). The productivity of teff is the lowest of major cereal cultivated in the site. Its high labor (both human and animal) and chemical fertilizer requirement diverted the interest of farmers more to other cereals especially wheat and maize which give more yields with lower expenses. However, its market value is greatest of all.

As Table 3 shows wheat and maize exhibited higher deviations from the means more or less with the same manner due to their increasing acceptance over a larger part of the woreda. Moreover, the discussants also production wheat argued that the production of wheat is significantly increasing.

Barley is slightly higher in its deviations than teff. Respondents revealed that due to several factors, especially loss of its varieties barely is losing its emphasis, in the area. Now a day, the cultivation of barley is restricted in the Dega and Wurch agro-ecological zones of the study area. “Engido” is becoming less acceptable. Its poor grade is the main reason for its losing importance as the respondents and experience of the researcher. In addition to the decline of the per capita farmland share of farmers in the study site, the increasing trend of September and October rains, as well as unknown disease which cause frequent loss of “engido” output, are causes of the decreasing trend of “engido” production. Sorghum revealed the highest negative deviations both in cultivated land and total output. Generally, for years to which cereal output obtained, 2007 shows the poorest performance in cereal production in the site. In this year positive anomaly for output is observed only to sorghum. Except for sorghum, “engido” and maize other cereals showed negative anomalies to their cultivated land area in 2007–2011 had been the second poorest year. Again sorghum registered positive value above the mean of output 2011. The information realizes that 2014 and 2013 showed better performance in the production of the major six cereals cultivated in Gozamen Woreda.

As Table 4 indicates wheat becomes the most widely produced both in cultivated land and total output. It is found to be the largest both in cultivated land (33.2 percent) and total output (42.6 percent). Maize is the second most important cereal in terms of total output and third important in the average cultivated land after wheat and teff. Teff revealed more or less constant percent share both in cultivated land and output throughout the period. It stands second in the cultivated land after 2008 and third on average total output after wheat and maize.

Generally Sorghum, “engido” and barley showed the lowest share in that order both in cultivated land output in the ten years . 2007 found to be the year of low percentage shares for mainly cool weather cereals especially for teff and wheat whereas it was the year of positive anomalies for maize, “engido” and sorghum. 2006 also showed low performance for teff and barley while it exhibited good results for maize, “engido” and sorghum. Refer the table in the text as (tab. 1).

Table 5 shows a correlation between cereals output and seasonal and annual rainfalls in Gozamen Woreda (2005–2014). The major cereals which are important sources of food, as well as income to producers, show variations with the association to rainfall conditions at different degree and manner. Teff more or less shows a negative correlation with summer rainfall owing to the heaviness of rainfall during this period over the larger part of the Woreda. It is positively coroneted to autumn rainfall. Barley weakly negatively correlated with summer rainfall. Maze showed a positive correlation. “Engido” is strongly correlated with summer rainfall. But show negative strong correlation with autumn rainfall. Its association with spring and annual rains also recorded negative. Because in years of better spring rainfall conditions farmers motivated more to prepare and sow wheat, barley, and teff than “engido”. As it is explained in its output analysis, the importance of “engido” decrease due partly to the increasing trend of autumn rainfall in the study area. Sorghum which is cultivated in the hotter part of Gozamen woreda showed strong negative correlations with rainfalls except annual.

Generally, “engido” and sorghum are cereals losing their significance in the study area. “Engido” was cultivated even on infertile lands. Its production cost was also very low compared to major cereals. Nowadays, in addition to the poor grade increasing autumn rainfall as well as unknown plant disease had caused frequent yield failure in “Engido”. Hence, some respondents explained the disadvantages of cultivating “engido”.

CONCLUSIONS AND RECOMMENDATIONS

The main objective of this study was to assess the association between rainfall variability and cereal output in Gozamen Woreda, North Central Ethiopia. So basic variables assessed are rainfall variability and the output of major cereals produced in the study area- teff, wheat, barley, “engido”, maize and sorghum. Mathematical calculation and descriptive statistics analysis results revealed that Gozamen Woreda is not water stressed. What is important is the temporal distribution timing and its continuity. Summer (June, July, and August) is the wettest season of Gozamen followed by Autumn (September, October, and November) unlike the large parts of Ethiopia where spring is the second wettest season.

August was found to be the wettest month on the site whereas February is the driest month of Gozamen Woreda. Temporal variability of rainfall is generally found to be the most important determinant of cereal output. The declining trend of spring rainfall creates stress on warm weather crops in particular and the farmland preparation even for others. However, the introduction of new technology, chemical fertilizers, and “better seeds” could overcome the problem (reduce the stress) particularly in maize production by extending its sowing period up to early June in the warmer parts of the site.

The performance of major cereal produced in the past ten years period (2005–2014) for which cereal output data obtained, exhibited variation though at a different magnitude. Teff which is the most favorite food. It showed the lowest variation in its performance for the last ten years compared to other major cereals. Generally, the highest deviation from the long-term means observed in sorghum followed by “engido”. Correlations result in analysis revealed association between rainfalls and cereal but at different degree. Teff which should more or less constant performance in the period under investigation exhibited strong negative correlation with summer rainfall. “Engido” is highly negatively correlated with autumn rainfall. Barely is positively correlated with all rains except summer to which it forms a high negative correlation.

To the general fact, the increasing trend of autumn rainfall is not appreciated in the study area. Owing to the decreasing growing period of many cereals resulting from the application of new agricultural inputs, autumn rain found to be wasteful.

Having concluded the overall findings presented in the study it is found worth full to forward the following recommendations which would have policy implications and predictive roles in socio-economic development tasks of the study area:

• Due emphasis should be given to the physical potentials that of the study area. Physical potentials should be maintained, developed, and restored include biodiversity – resulting from different agro-ecological zones, huge and industrious agricultural labor, huge livestock resource potential and irrigation potentials.
• Having recognized the physical potentials and the contemporary local and national demands efforts should be made by agricultural experts and the community on areas of irrigation, traditional soil maintaining practices, proper livestock rearing services, and general environmental resources conservation works.
• As to the view of farmers and significance of natural cereal species, the complete shift towards results of genetically engineering costly seeds should be properly analyzed by actors of the field.

ACKNOWLEDGEMENTS

We acknowledge Adama Science and Technology University for funding and supporting this research work.

REFERENCES

Accepted for print: 18.12.2017

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.