This paper presents a method to associate land use/cover with productivity in 16 Agrotech Observatories (AOTs) in Mexico. Compact agricultural areas in Mexico have been identified, which are monitored as to their behavior concerning production and rural productivity in a network of AOTs, which is a compact agricultural area representative of agro-ecological, technological and social conditions in the country. To optimize production and agricultural productivity in compact areas, a multidisciplinary and holistic approach with four lines of activity (agro-ecological, technological, economic, and social), and ten actions are used. The objective of this work was to obtain the land use/cover and productivity of sixteen compact agricultural areas (AOTs) in the Mexican Republic, using panchromatic and multispectral SPOT 5 imagery, in order to provide information to the agricultural sector of the country, and to support decision making contributing to the optimization of production in areas with high actual and potential productivity. As an example, in this paper the land use/cover and productivity “AOT 20 Hidalgo” were described. Currently, it is important to have updated and accurate information to support actions and programs of federal, state and local government for farmers, particularly in compact areas with high agricultural production potential.
Land use is the human modification of the Earth’s surface, which has strongly affected and will increasingly shape planetary functions [
Land-use science, also called land-change science [
For the Mexican government, the gradual and sustainable increase in productivity and competitiveness in compact agricultural areas of the country is important; this initiative is supported by the application of knowledge generated in agricultural research institutions and universities. Pursuing this initiative at the national level is a complex task due to the diversity of agro-ecological and socio-economic factor combinations. Compact agricultural areas in Mexico have been identified by a group of experts; they are monitored as to their behavior concerning production and rural productivity in a network of Agrotech Observatories, also named Compact Agricultural Areas (AOTs). So, an AOT is a compact agricultural area representative of agro-ecological, technological and social conditions in the country. A multidisciplinary team of scientists and researchers analyze and define the best production options for the different types of producers in these areas in order to maximize profits and minimize risks. To optimize production and agricultural productivity in compact areas, a multidisciplinary and holistic approach with four lines of activity (agro-ecological, technological, economic, and social), and ten actions are established. The objective of this work was to obtain the land use/cover and productivity of sixteen AOTs in the Mexican Republic, using panchromatic and multispectral SPOT 5 imagery, in order to provide information to the agricultural sector of the country, and to support decision making contributing to the optimization of production in areas with high actual and potential productivity.
This work was carried out during 2011. The criteria used for selecting the 16 compact agricultural areas of the country were: 1) Use the cartography of the agricultural borderline; 2) Generate a national coverage mesh with cells, each one of 10,000 hectare; 3) Select cells with greater than or equal to 6000 ha of agricultural area and delineate AOTs to contain ten or more selected cells (
All compact agricultural areas were characterized according to geographical location, climate and soil conditions, actual and historical crops yields. As shown in
AOT number | Area (hectare) | States involved |
---|---|---|
26 | 200,000 | State of Mexico |
28 | 150,000 | Puebla |
03 | 160,000 | Sonora |
11 | 160,000 | Aguascalientes-Zacatecas-Jalisco |
10 | 200,000 | Zacatecas |
22 | 120,000 | State of Mexico-Hidalgo |
19 | 160,000 | Queretaro-State of Mexico-Hidalgo |
16 | 150,000 | Guanajuato |
32 | 150,000 | Oaxaca |
39 | 120,000 | Quintana-Roo |
27 | 150,000 | Morelos-State of Mexico-Puebla |
30 | 120,000 | Veracruz-Oaxaca |
34 | 120,000 | Tabasco-Chiapas |
35 | 240,000 | Chiapas |
13 | 150,000 | Jalisco |
20 | 120,000 | Hidalgo |
Total: | 2,470,000 | 14 |
To determine the land use/cover in all AOTs, we used 31 panchromatic scenes and 31 multispectral scenes of SPOT-5, with 1A correction level, which were obtained from the receiving station SPOT-Mexico, supported by the Ministry of Agriculture. The spatial resolution of the panchromatic and multispectral images was 2.5 and 10 m respectively. A process of geometric correction and ortho-rectification was applied to the satellite imagery [
By visual image analysis and photointerpretation, the agricultural borderline was established; the approach consisted of involving the use of patterns, texture, size, color and topological features among objects. A mosaic of images was generated to facilitate management and interpretation, thus obtaining a continuous area to define and identify the agricultural areas and perform the scanning process in an efficient manner. In this procedure, RGB band combinations were used to highlight the vegetation cover and attain greater accuracy in line drawing in digitization. For this case study, three SPOT-5 panchromatic and three multispectral images covering the “AOT 20 Hidalgo” were used (
The level of detail of the information layer of the agricultural borderline plays an important role in the estimation of the agricultural surfaces and for the determination of land use; this layer supports an initial estimation of the extent of agricultural cover and provides a framework for sampling and for the generation of work units to be used in the classification and information extraction processes. Using the ArcGIS V-10.0 geographic information system (GIS) and a reference mesh, zooms were used on the RGB screen compounds at a 1:10,000 scale to preserve homogeneity in digitization (
The segments are geographical objects over which on-the-field sampling for collecting information about land use/cover is applied. These segments are delineated to fall within the limits of the agricultural borderline. To generate sampling segments within each AOTs, the [
・ Segmentation percentage: 3.
・ Segment Size: 49 hectares.
・ Segment width: 700.
・ Distance (threshold): 0.4.
・ Number of replicas: 2.
・ Number of segments: 55.
Sensor | K/J | Making scene |
---|---|---|
Spot 5 panchromatic | 588-309 | 16/11/2011 |
Spot 5 panchromatic | 588-310 | 23/10/2011 |
Spot 5 panchromatic | 589-310 | 04/11/2011 |
Spot 5 multispectral | 588-309 | 16/11/2011 |
Spot 5 multispectral | 588-310 | 23/10/2011 |
Spot 5 multispectral | 589-310 | 04/11/2011 |
Previous to fieldwork the parcels within each segment were digitized over ortho-rectified panchromatic images to obtain the parcel distribution, which was used in a training field context (
In studies of land use/cover, fieldwork is very important to generate training points with GPS equipment and to assess the quality of the results. The fieldwork consisted of identify the different land use/cover within the segments. The composite RGBs for each segment were used to locate training sites; for this, the coordinates of the center points of each segment were captured with map-mobile computers and satellite navigators to locate the plots on the ground.
The information captured on the field consisted of establishing the type and condition of the crop, its genotype and soil moisture regime, and of sampling actual yields to determine the productivity of different agricultural land covers. Furthermore, to know the surfaces and the potential crop yields, a yield forecasting technique based on an existing methodology was used [
Each work crew had a digital camera to get pictures of the sampled points. In
A buffer for all segments is generated so that the spectral signature of a given crop is more accurate. Within the
segments, the different land uses are defined and the data are transformed to a raster format in order to associate it with the image to be classified. Training fields for each spectral class are generated, and a digital classification is performed for discrimination of crops in the work units of the AOTs.
The classification of images is one of the most important parts of digital image analysis. Supervised classification algorithm with maximum likelihood classification was used in this study [
For the digital classification the Erdas Imagine® processor was used. In the analysis of training fields only segments of replica # 1 corresponding to 50% of the total segments visited were considered. The other 50% corresponding to replica #2 were used for validation of the digital classification [
The classification results were collected in a new data-image, similar to the originals with respect to structure and size, but where the digital number (DN) of each pixel does not correspond to a reflectance value but to an assigned category. This new image is the final result of the work, with two types of products: cartographic and statistical.
In all compact agricultural areas, the agricultural borderline cartography was updated by digitizing SPOT-5 Panchromatic images [
Continuing with the example of “AOT 20 Hidalgo”, the agricultural area is shown in
Continuing with the example of the AOT 20 Hidalgo, the actual land use/cover was as follows: corn grain 22.4%, alfalfa 21.9%, oats 2.9%, corn forage 2.8%, beans 1.3%, Opuntia 0.4% among others covers (
Other crops of economic and social importance are beans and maguey (Agave americana). The first crop is also a common staple of the daily diet in Mexico. Maguey, of which only 312.8 ha are reported, is used for the extraction of natural syrup, an input to the production of an indigenous fermented drink called pulque.
Concerning the AOT 20 Hidalgo exemplified here, for all the crops that were analyzed the potential productivity is surpassed by the actual one. In the case of the corn-grain crop, the current yield of 7.6 ton∙ha−1 can potentially reach 12.0 ton∙ha−1, with production volumes of 114972.8 ton to 232081.2 ton respectively.
Land use | Land cover | Surface (ha) | (%) |
---|---|---|---|
Agricultural | Corn-grain | 15128.0 | 22.4 |
Agricultural | Alfalfa | 14750.8 | 21.9 |
Agricultural | Oats | 1974.1 | 2.9 |
Agricultural | Corn-forage | 1910.9 | 2.8 |
Agricultural | Bean | 901.2 | 1.3 |
Agricultural | Opuntia | 312.8 | 0.4 |
Agricultural | Bare soil | 2707.1 | 4.0 |
Agricultural | fallow soil | 13106.2 | 19.4 |
Urban | Urban area | 1228.7 | 1.8 |
Infrastructure | Roads | 755.1 | 1.1 |
Other uses | Other uses | 14581.2 | 21.6 |
Total area | 67356.6 | 100.0 |
Land cover | Actual productivity | Potential productivity | ||||
---|---|---|---|---|---|---|
Surface (ha) | Yield (ton∙ha−1) | Production volume (ton) | Surface (ha) | Yield (ton∙ha−1) | Production volume (ton) | |
Corn-grain | 15128.0 | 7.6 | 114972.8 | 19340.1 | 12.0 | 232081.2 |
Alfalfa | 14750.8 | 12.0 | 177009.6 | 18439.2 | 17.0 | 313466.4 |
Oats | 1974.1 | 10.6 | 20925.5 | 4435.5 | 15.0 | 66532.5 |
Corn-forage | 1910.9 | 17.8 | 34014.0 | 8267.3 | 25.0 | 206682.5 |
Bean | 901.2 | 2.3 | 2072.8 | 3241.7 | 3.8 | 12318.5 |
The potential productivity can be achieved by taking into account: 1) the climatic and soil requirements of each crop to increase acreage, in particular for corn as illustrated; 2) the application of technological packages recommended by agricultural research institutions.
This requires a comprehensive and detailed knowledge of the climatic and edaphic resources and of the land use patterns within the compact agricultural areas as defined for the country. Therefore, it is very important to accurately know the land use/cover and geo-spatial distribution of crops, particularly in areas with a high productivity potential. Thus, land cover is defined as the observed (bio)-physical cover on the specific area surface. It includes vegetation and man-made features as well as corn crop, alfalfa, etc. On a fundamental level, land cover is the most important element for description and study of the environment [
It is important to have current and accurate information to support federal government programs aimed at the rural sector in Mexico, particularly in areas of high productivity potential. Information of land use/cover is very important for the proposed productive restructuring actions of the Mexican agriculture. It is possible to improve production and productivity of the countryside, with studies of potential soil use. What is also needed is a more efficient national policy which incorporates technological recommendations generated by research institutions and supports decision makers within the Ministry of Agriculture as well as the country’s farmers.