Irrigation innovation: protecting cotton from drought

Sensemaking / Irrigation innovation: protecting cotton from drought

From simple solutions for smallholders to the latest 'smart' irrigation systems and GM strains, Jon Turney explores how cotton farmers are preparing for growing pressure on water resources.

By Jon Turney / 18 Jul 2013

The world’s major crops, and the folks who tend them, have to adapt to new demands in the 21st century, and cotton is no exception. Global warming means climate instability, with heat, drought and floods in unpredictable combinations.

How will cotton fare in this new, harsher world? Not as badly as some people think. Its main defence will be more careful management, especially of water. Cotton needs high soil moisture at critical periods, particularly early in development, and then likes long, hot spells when it reaches maturity. As Kater Hake of Cotton Incorporated’s Agricultural and Environmental Research Division puts it, “for most crops, the more water the better. But cotton needs not too much, especially at the start of growth and late in growing”. Some sub-tropical regions manage that combination without help, but many cotton fields depend on irrigation. The key, then, is how much and when.

Cotton’s history should be easy to improve on. Slave-worked plantations in the US and extraction of all the value of Indian cotton in British mills are in the past. But the dried-up salt flats where the Aral Sea once lapped the shores of Kazakhstan are more recent evidence that the cost of cultivating the fibre can be higher than anyone imagined.

However, the transformation of the 26,000-square-mile Aral Sea into a chemical-laden dustbowl has more to do with the way the Soviet-era command economy was run than with the thirst of Gossypium hirsute, the main cotton plant species. Yes, it was irrigation that caused the sea to shrink to a quarter of its original size and lose 90% of its volume, destroying a once flourishing fishing industry, but it was irrigation organised so that little water actually fed the plants.Tens of thousands of kilometres of unlined canals channelled water from the two rivers that fed the inland Sea to the fields in what was previously desert. Most of it sank into the soil or evaporated on the way, or from where it lay in the cotton fields. Officials didn’t care. Some even believed that removing the sea entirely would create more fertile land, not more dusty, wind-blown desert. By the time the disaster was apparent, it was too late to avert it. Modest improvements in use of water could perhaps have slowed it down. Now, there are many refinements in irrigation technique which could have prevented it altogether.

The key with cotton is how much water, and when

Smallholder farmers can reduce their water usage, by using bed and furrow irrigation, for instance, instead of flooding their land. Simple gated pipe systems (small evenly spaced gates in the side of the pipe are used to control and target the flow of water) can also cut how much water is needed for irrigation. And increasing the soil’s organic matter can improve the water infiltration and retention capacity of the land. Educational initiatives to help farmers better understand the causes of groundwater stress, such as the farmer-managed Groundwater Systems Project, run by the UN Food and Agriculture Organization in Andhra Pradesh, India, have also made a difference.

Aside from these relatively straightforward solutions, farmers can also take advantage of the latest spray irrigation systems, automation tools and moisture sensors. In fact, the current buzz term is ‘smart irrigation’. In the US, the development of more efficient irrigation systems can clearly be seen as you move west across the cotton belt, from Florida to Texas. Cotton in Texas is a billion-dollar annual industry, and is currently under intense pressure from an extended drought. Fully half the crop was lost in 2011: between November 2010 and the following July, Texas only received 1.2 inches of rain, around one-tenth the normal amount of rainfall for that period. Planting was down in 2012-13, as the rains continued to fall short.

Cotton in Texas is a billion-dollar annual industry

Even in kinder years, Texas is not plentifully supplied with water. So irrigation has evolved to make better use of what there is. Farmers have moved from flooding fields, to pivot sprays above the leaves (not all that effective), to modified designs which deliver spray lower down (much better). The next move is to install costlier, but still more efficient, soil-fed irrigation, which uses a network of small pipes to deposit water near the roots. This drip irrigation reduces evaporation loss and is also an effective way of applying fertiliser, as needed.

Still smarter systems can incorporate either pivot spray or drip delivery. Variable rate irrigation moves away from wasteful blanket spraying and allows for the fact that fields and soils are not uniform. The farmer uses a detailed map of each field, compiled from knowledge on the ground and aerial photos. Mobile sprinklers linked to GPS to track their exact position can then be adjusted to deliver the right amount of water at each location they cover. Such systems cut water use by around 20%, and save fuel used for pumping in the same proportion.

Finally, there is a transition to ‘smart’ in the modern sense: not just where the water goes, but when it is applied. That can now be decided using continually updated local information. Computer models combine information about soil moisture in the field, the stage of crop growth, temperature, humidity, and local weather forecasts, and indicate when to water the crop. Sprinkler systems are already widely linked to control by sensors, and smart phone apps that define irrigation needs more precisely are on the way. 

This kind of innovation can also be combined with development of older methods. Farm ponds have long been used to store rainwater. Ed Barnes of Cotton Incorporated’s Agricultural and Environmental Research Division points to a feasibility study in Arkansas testing the idea that some ponds can be left unlined, allowing the water to seep into shallow aquifers for later recovery. This can take up less space than a larger pond and reduce evaporation losses, but needs careful surveying: build an unlined pond on sandy soil and the water is lost.

As recent years in Texas show, severe drought still undermines cotton harvests, but drought-resistant strains are also under development, both via conventional breeding and genetic manipulation. GM strains may be more acceptable for cotton than some other crops, as earlier GM varieties, offering pest resistance and herbicide tolerance, have already been adopted by most major cotton producers.

However, more conventional breeding is also becoming more powerful because breeders are now armed with the genome sequence for cotton, published just last December. In fact, that should be ‘genome sequences’, because the main cultivated variety has doubled up its genome: it has 44 chromosomes rather than the usual 22 found in plants. As Barnes’ colleague Hake explains, this means it has extra evolutionary flexibility, and should also make breeding for desired traits a little easier.

Sequencing a genome does not immediately tell you what all the genes do, but studies of the best understood plant, the weed Arabidopsis, have revealed what stress tolerance genes look like, according to Hake. Wild cotton has better stress tolerance than modern cultivated varieties, even though the genes are retained. Cotton also has relatively good heat and salt tolerance, so drought resistance is the main target for new strains. There are many variables to allow for, and Cotton Incorporated is experimenting with the blanket coverage of experimental plots with sensors which can be combined to give a complete picture of how trial plants are growing. In Arizona, some cotton plants are growing in fields where a mobile sensor rig measures plant height, temperature and leaf density every 10cm. Further refinements on the way include measuring chlorophyll concentration in the leaves, and integrating the on the ground measurements with a model of how sunlight is reflected off the canopy.

The challenge, as with all such sensor arrays, is dealing with the mass of data. “These machines operate seven days a week”, says Hake. Result: many terabytes of data which have to be related to the precise genetic make-up of the plants under surveillance. The goal is to develop an analysis which shows how a new strain to be offered to farmers will fare under a wide range of conditions, and especially to identify which ones will resist extreme conditions.

Meanwhile, cotton also offers opportunities to farmers struggling with saline soils. Cotton plants have around twice the salt tolerance of wheat or rice, for example, and could be a substitute crop in places where soil quality is threatening farms. This possibility has already been explored in Japan on land swamped by the 2011 tsunami. Tens of thousands of acres of farmland were left unfit for rice cultivation when the sea receded; the salt left behind cannot be washed away because the pumping equipment was destroyed.

In north-eastern Japan, the Tohoku project is providing some farmers with cotton seed to revive their land. First results from trials on a small plot were poor because planting was late, the weather was bad and the acreage was limited. But later sowings established that cotton can be harvested from what used to be rice fields. The first garments made from Tohoku cotton are now on sale in Tokyo stores.

As agriculture continues to evolve to meet changing conditions, the Japanese farmers whose land was inundated will not be the last who learn a new routine for growing an unfamiliar crop, but one which has been cultivated for thousands of years: cotton.


Jon Turney is a science writer, editor and lecturer, and author of ‘The Rough Guide to the Future’.

Photos: Flickr Calsidyrose

What might the implications of this be? What related articles have you seen?

Please register or log in to comment.