Koa (Acacia koa Gray) is an important endemic tree species in Hawai`i for ecological, cultural, and economic reasons. Previous research has shown that koa growth is limited by low plant available soil phosphorus (P), but its nutrient dynamics are poorly understood. An ongoing nutrient cycling study was established in 2002 in a 24 year-old koa forest on the island of Hawai`i on 1,500 to 3,000 year old Hapludands derived from volcanic ash over `a`ā lava. The objective of this study was to develop a mechanistic understanding of how P cycling is altered by silvicultural treatments, specifically crop tree release, grass competition control (CC), and CC + P amendments (CC+P) (750 kg P ha-1 over 18 months). Various phosphorus pools were monitored to determine short and long term P fertilisation effects and to identify immediate sources of available P. Between January 2004 and May 2005, plant available P was measured using ion-exchange resin membrane stakes (PRS-probes) in all treatments. Competition control had no effect on P availability. Available P in CC+P was 50 times greater at 5 cm and 20 times greater at 15 cm than in the control. Litter P content was greater as well: 21 kg P ha-1 in CC+P as compared to 10 kg P ha-1 in the control. Results from Hedley fractionation and phosphatase activity are pending. These preliminary results indicate that P fertilisation increased P availability for at least 14 months after the last fertiliser application. This calls into question the classical assumption that the high P sorption capacity of young Andisols leads to rapid declines in soil P availability after fertilization. Because litter P cycling is likely insufficient to account for the magnitude of this sustained increase, other mechanisms must be at work to maintain the high availability of added mineral P.