Soybean Yield Improvement — Contemplating the Past for Clues to the Future.

In a 1981 CSSA Symposium on genetic contributions to crop yield gains, I reported that USA soybean yields had increased from 1924 to 1980 at a linear rate of 20.9 kg/ha per year. At a 1998 CSSA symposium, 17 years later, I reported a somewhat higher linear rate (22.6) from 1924 to 1997, but also noted that an exponential equation could be fit (with a nearly equivalent R2) to the data. At this (2009) CSSA symposium, 28 years later, I will be reporting that the linear rate is now 23.1 from 1924 to 2008, and that an exponential equation can again be fit to the the data at about the same R2. Is this slight increase in the linear rate (over 28 years) reflecting an accelerating rate of on-farm soybean yield improvement that is arising from ever-greater on-farm use of genetic/agronomic technologies, or is it just statistical noise? Many past claims of this crop species reaching a yield plateau have (to date) proven to be illusions that were inappropriately drawn from clusters of sequential years when unfavorable weather prevailed. Still, in theory, it must be kept in mind that there is a finite soybean yield limit imposed by light interception potential, availability of water, and crop nitrogen requirements. Given this limit, on-farm soybean yield improvement must ultimately follow a logistic pattern over time that might involve acceleration upon approach to the logistic inflection point, but then subsequent deceleration to some yield plateau (which would be less that the theoretical yield limit itself). In my presentation, I will review the available data and published studies on the long-term nature of past on-farm soybean yield improvement and explore what that information might imply with respect to both pessimistic and optimistic claims about future soybean yields.