Thousands of surveys on rural livelihoods in developing countries are being done every year. Unfortunately, many suffer from weaknesses in methods and problems in implementation. Quantifying households' dependence on multiple environmental resources (forests, bush, grasslands and rivers) is particularly difficult and often simply ignored in the surveys. The results therefore do not reflect rural realities. In particular, 'the hidden harvest' from natural resources is generally too important to livelihoods for development research, policies and practice to ignore. Fieldwork using...