RECENT IMPROVEMENTS IN THE ENERGY EFFICIENCY OF AGRICULTURE - CASE-STUDIES FROM ONTARIO, CANADA

Energy used (GJ) per tonne of crop produced defines energy efficiency. Low energy use/tonne of crop produced indicates higher energy efficie ncy. Because of improved crop breeding (more stress tolerance, genetic gain) crop yields have increased. Concurrently, energy used has decre ased. Based on an analysis of energy use/crop produced for Ontario's m ajor row crops (corn and soybean), energy efficiency of crop productio n increased from 1975 to 1991. In corn, energy efficiency increased 49 .8% from 87.95 l of diesel fuel equivalent (LDFE)/t of crop produced t o 44.16 LDFE/t from 1975 to 1991. Total corn production (t) and yield (t/ha) increased by 60.0% and 20.1%, respectively, whereas energy use/ ha (LDFE/ha) and total energy use (LDFE) decreased by 39.7% and 19.7%, respectively. For soybean, energy use/ha (LDFE/ha) decreased by 46.3% but total energy use (LDFE) increased by 93.8%. The increase in total energy use was caused by a 260.8% increase in soybean acreage (ha). B ecause the increase in soybean production (t) of 278.2% was much great er than the increase in total energy use (LDFE), energy efficiency inc reased 48.7% from 75.46 LDFE/t to 38.67 LDFE/t. The proportion that fe rtilizer manufacturing contributes to total annual energy use in corn decreased slightly (74.2-71.0%) between 1975 and 1991; however, energy use in fertilizer manufacturing declined by 42.4% because of the incr eased efficiency of manufacturing nitrogen fertilizers. Increased effi ciency of fertilizer application also decreased energy use, although a pplication is a much smaller proportion of the total energy use budget . Similar decreases occurred in soybean production. Our analysis of no -till systems using various herbicides and application rates indicated that the systems using the highest herbicide application rates requir ed the least energy, because decreased herbicide application rates may require increased cultivation. This, however, depends on the timing ( pre-plant, planting, post-emergent) and type (banded broadcast) of app lication and is tore only for newer herbicides with low application ra tes. Copyright (C) 1996 Elsevier Science Ltd