In the mid-eighties the idea that the region-favoured crop strategy had outlived its life was not easily acceptable. The action plans for wheat or paddy or oilseeds were the policy staple. Here the agroecology had to adjust to technology. The thought that a land and water development strategy for the agroecological regime was needed and an appropriate cropping sequence could follow was considered risky. Whatabout food security, I was constantly asked.

Since I was responsible with others for the action plans, some companions almost felt let down. It took time to adjust to the thought that the farmer could be trusted to meet demand, if he had adequate support.

The original watershed development strategy provided for an investment in land and water of about Rs.10,000/hectare. Around 40 percent would be provided by the development agencies, the rest would be community labour and bank funds. I had produced a famous table which showed that with order of effort, in 8 actual field efforts grain and energy requirements were met.

The economic returns were high around 18 percent on the rupee, but initially there could be financial losses, so support was needed. The table was flatteringly repeated and I recently saw a 50-project example. The late eighties saw a flurry of activities around the concept. The national watershed programme was revamped by S.V. Giri, making it more flexible and community- and NGO-friendly.Watershed activity was integrated with the employment programmes. In particular land shaping activities and labour put in tree crops on private lands were made eligible. In the early nineties, a NABARD annual report was on agroclimatic planning. The agroclimatic plan was presented to parliament and had widespread political support.

Meanwhile, the field situation was changing. The World Development Report of 1992 on the environment advocated community-based watersheds. The examples they gave ran into costs per hectare in tens of thousand of dollars. Given the Indian model programmes this looked strange. Then in 1992 in my periodical trips to a cooperative where I am known as an arthashastri, I saw a well-worked out project to harvest another 19 inches of water, which cost around Rs. 30,000/hectare. Interestingly, it was a viable investment.

I worked out for them that even with the lowest value added crops that could be grown, the debt and equity contribution, if the project was on a company mode, couldmeet an interest cost of 12 percent annual and a pretax profit rate of 28 percent. I pushed the project but the banks will not bite. A loan of small sums for agriculture is alright but not investment-oriented agriculture.

The trouble in this great country is that if the public banks don't say yes you are sunk, since there is no other source of finance. The second example I saw was the Sardar Sarovar distribution systems. Here lined computer-controlled hydraulic systems up to the field levels gave high efficiencies but cost money. A well-known journalist attacks the project by putting its irrigation costs at Rs. 50,000/hectare. The cost is lower if you subtract the cost of hydel power from irrigation cost but a repayment capability of around Rs. 6,000/hectare can finance an investment cost of over Rs. 25,000/hectare.

At the margin, the productive use of water gives these order of yields already and there is no need to give national crop yields to trash the investment requirements of Indian agriculture.

Ihad seen that the cost of irrigating drylands through lift irrigation projects could in some areas be over Rs. 20,000/hectare and these could be financed viably. But last week in Ahmednagar district, Crispen Lobo, who is running an NGO named WOTER, takes me to the field. He takes me to a village with 15 inches of rainfall and I see the wonders of land shaping and water harvesting. The farmers protect the treated lands from grazing. They are growing trees and fodder and grain for self-consumption.

In the entire region dryland horticulture is picking up. WOTER is funded by an international source and so is my lift irrigation project in the Panchmahals. They both insist on a community contribution. In the case of WOTER, it is 20 percent. The rest they treat as a grant.

The bankable cases are easier, in some sense, because a restructured banking system would finance them. But if the costs are Rs. 2,000/hectare plus and grants are required, Crispen asks me the replicability question. I don't know the answerand tell him we will all have to work on it. He tells me that I am not a politician but we both know that's not the answer. If two NGO's, I know, have treated around half a million hectares, then there is promise. Remember, a few years ago these were only 8 examples. But while preparing for Rio, as a special advisor for UNCED, I had argued for targets of 2 million hectares annual if sustainability requirements are to be met.

So near and yet so far. As I said, let us all work on it together. There must be an answer.