Agricultural Research Management Information System

Home
Research Summary
Search
About Us
- ARMIS
- Brochure
Contact Us
Report
User Request
Data Input
Help
- Operation Manual
  - PDF
  - Video
- Program Area & Commodity
We have reached 37600 number of research entries at this moment.

- Logout

Research Detail

Home
Research
Detail

Zoning of Bangladesh — an Exercise Based on Existing Cropping Practices

Sajjad Zohir
Research Fellow
Bangladesh Institute of Development Studies

Abstract/ Executive Summary:

Current practices in zoning for the purpose of crop sector analysis in Bangladesh are often constrained by non-availability of data. The paper reviews some of these exercises and presents an alternative set of zones based upon observed similarity in cropping patterns across 64 new districts .The exercise essentially involves finding vector distance and grouping vectors with low (intra-group) vector distance. Some data are presented to suggest of variability across the newly defined zones; and comparisons are also made with zones that are currently in vogue among development practitioners in Bangladesh .Since data pertaining to the newly defined zones may be availed on a regular basis, it is hoped that the present exercise will be of greater use for policy planners and makers in the country.

Keywords: Zoning, Cropping practices

Location:

Start Date:

End Date:

Program Area: Farming System

Non-commodity: Classification

Objectives:

To review zoning for the purpose of crop sector analysis and present an alternative set of zones based upon observed similarity in cropping patterns across 64 new districts.

Methodology:

The three broad approaches to zoning in Bangladesh agriculture have been outlined. The availability of data being the binding constraint, any attempt to aggregate data in terms of an alternative classification needs to be related with the administrative units. This is exemplified in the MPO exercise where each planning area had to be identified in terms of upazilas (either fractions or wholes). Ideally, one should also be able to reconstruct the BBS data in terms of the AEZs. The reliability would however diminish since one would have to work with smaller fractions (of upazilas) and aggregate over too many of such fractions. More importantly, such an exercise could only highlight on variability across agro ecological zones and may have little use for planning purposes. With the objective to account for variations in agro ecological characteristics (that are likely to reflect in current cropping practices) as well as to be able to relate the zones with BBS units (for which data are generally available), we have carried out a separate zoning exercise. The details on methodology and the outcomes of the exercise are outlined in this section. It is assumed that prevailing "cropping practices" in various areas largely reflect their agro ecological characteristics. Disaggregated information on cropping practices are found from BBS sources in terms of acre ages under various crops. We use such 1983/84 agriculture census data to identify zones that are relatively more homogeneous within, in terms of the cropping practices. We purposively confined to (new) district level data in order to limit the size of the exercise.2 For each district, a normalized vector was obtained, each element of which effectively represented the percentage of gross cropped area allocated to a particular crop. Thus, for 64 districts, we had 64 such vectors. The purpose was to measure the distance between two vectors for various pairs of vectors,3 and group these vectors in such a way that vector distances for various pairs within each group were minimal.

It may be noted that normalizing the vectors was essential for our exercise. One may argue that the scalar differences, reflecting the differences in cropping intensities across districts, is lost in the process. It is however, necessary to recognize that cropping intensities are highly correlated with the mix of crops that are cultivated over a year. The loss in information due to focus sing on normalized vectors may, therefore,, be of not much significance; especially when detail distinctions across crop varieties have been made. Based on the earlier mentioned method, bilateral differences for all 64 districts were calculated. Districts among which bilateral differences were found to be minimal, were grouped into a zone.For purpose of identification in this paper, we call them Cropping Pattern Zones (CPZ).

Source of Information: The Bangladesh Development Studies,Vol. 21, No. 2 (June 1993), pp. 93-105

Article Link (Online Journal):

Funding Source:

Total Budget (Tk.) :

Achievement/Findings:

The association of the newly defined zones with the MPO planning areas is not quite smooth except in the case of one zone, all other zones have fractions of some planning areas. This is quite expected since the catchment areas and therefore the planning areas include fraction of upazilas in many cases. The CPZs, presented in this paper, account for administrative units on which data are collected on a regular basis. While AEZs and Planning Areas of MPO are more rigorously defined in technical terms; their applications on time-series data are not feasible. One strategy may therefore be to generate future data in terms of the more rigorous classifications. Alternatively, one may choose to apply the suggested CPZ classifications on existing data.

Knowledge Management: Journal