The trend to speak of water productivity and "crop per drop" in plant-water relationships, where before the 1990s the term water use efficiency was commonly used, has to be deplored. Whereas rain or irrigation water is always only one of the many inputs and growth conditions to ensure a certain output or yield, the term productivity is not appropriate and has to be reserved for genuine production factors such as labour, land and capital. Besides, expressing this water productivity in output per unit water input also is not always meaningful. Low values such as common in wetland rice cultivation, or in dryland farming in cycles including a fallow year may be very economic or wise forms of crop water management, whereas high values such as achieved in drip irrigation may be not. Where the concurrent opportunity costs or forgone values, e.g., the alternative potentials of the saved or lost water is not taken into account, high water productivity values may not be a suitable target. Irrigation efficiency and water use efficiency remain useful and meaningful parameters, provided they are well defined and used at the level of individual farmers or irrigation projects. For more holistic and agro-ecological perspectives, they are appropriate only if used as a "value-free" parameter in need of further evaluation against social, ecological, sustainable or other desiderata. Historical aspects of the differing concepts in plant-water relationships and agricultural water management are extensively dealt with. It is shown that the 20th century concepts and definitions of crop water relationship parameters, so much needed in agricultural water management, hinge more on the economic and geographical conditions than on mere scientific notions. Field cases to illustrate and defend my conservative stances are provided. It is concluded that a split-up of domains is needed to properly define water use efficiency for different perspectives: a universal or single menu cannot be applicable due to the technical and socioeconomic diversity of issues and challenges. (c) 2006 Elsevier B.V. All rights reserved.
