A two-level unreplicated factorial experiment to determine the effect of organic and inorganic fertilizers on dry matter yield of permanent pasture

Abstract

In organic pasture systems, the balance of soil fertility is maintained by maximizing the cycling of on-farm nutrients. In this study, the effect of fertility management on dry matter yield in organic pasture systems was examined using an unreplicated 25 factorial design. The factors were nitrogen (N), phosphorus (P), potassium (K), compost and seaweed extract, each at two levels (with [+] and without [−]). The trials were conducted in 2000 and 2001 at two fields on an organic dairy farm in Nova Scotia, Canada. The pasture mix is predominantly perennial ryegrass, orchardgrass and white clover. The fertility treatments were applied in spring, and forage yield was collected from early-, mid- and late-summer harvests. The analysis revealed the significance of several interaction effects, which varied considerably from harvest to harvest. Both +N and +Compost treatments were associated with increased yield. High levels of P and K already present in the soil either diminished or varied the effects of P and K treatments. Seaweed extract enhanced the effectiveness of N and compost, but reduced yield when used alone. The results suggest that the commonly recommended use of seaweed extract for organic pasture management can be detrimental to biomass production unless sufficient nitrogen is provided. Despite the unusually dry weather in the 2 years, the study demonstrated the applicability of unreplicated two-level factorial experiments in agricultural field studies.

Analysis of Augmented Unreplicated Factorial Designs Repeated in Time

“…Despite of the widespread use of unreplicated designs, they are uncommon in agricultural experiments (exceptions include Thies and Fery; Astatkie, Joseph, and Martin; Martin et al ; Payne; Perret and Higgins; and Camara et al ). Machado and Petrie point out that many agricultural researchers consider unreplicated experiments to be unscientific and unacceptable for publication.…”

Experiment design to achieve desired statistical power

“…One may also exploit ''hidden replication'', through the use of factorial experiments, to increase power (e.g. Astatkie et al 2006).…”