Various events over the last two centuries have destroyed the vegetation and caused rapid soil erosion on large areas of the small, arid, windy tropical shield-volcano island of Kaho`olawe, Hawai`i. These activities were largely halted in the 1990s, and efforts have been made to restore the island's vegetation in order to stem erosion. In 2003, the Kaho`olawe Island Reserve Commission (KIRC) began restoration efforts using native vegetation. In 2006 to 2010, the U.S. Geological Survey (USGS), in cooperation with the KIRC, monitored streamflow, fluvial suspended-sediment transport, and erosion rates in the Hakioawa and Kaulana watersheds on northeastern Kaho`olawe to provide information needed to assess the effectiveness of restoration efforts. This report presents the results from this monitoring.
Results.-Hakioawa and Kaulana gulches were dry about 90 percent of the time during the monitoring period; mean annual flow was 0.06 ft3/s at Hakioawa Gulch gage and 0.01 ft3/s at the Kaulana Gulch gage. For the period when the sediment gages on both gulches were operating concurrently (October 2007 to September 2009), sediment discharge was higher from Hakioawa Gulch than from Kaulana Gulch. The annual suspended-sediment loads for the concurrent period averaged 1,880 tons at the Hakioawa Gulch gage and 276 tons at the Kaulana Gulch gage.
Of the 77 erosion-monitoring sites in the Hakioawa and Kaulana watersheds, 50 had overall rates of change indicating erosion for the monitoring period, ranging from -1 to -10 mm/yr and averaging -3 mm/yr. Seven sites had rates of change indicating overall deposition, ranging from 1 to 15 mm/yr and averaging 5 mm/yr. Twenty had rates of change below detection (less than ?1 mm/yr).
The average rate of change for the 26 sites in areas that have undergone restoration by the KIRC was below the detection limit of the erosion-monitoring method. In comparison, the 51 sites in nonrestoration areas averaged -2 mm/y. Both of these averages, however, include sites that showed overall erosion as well as sites that showed overall deposition.
The average rate of change was -1 mm/yr for both the 32 sites on rills and the 42 sites on interfluves; both categories include sites that showed deposition as well as sites that showed erosion. All three sites on hummocks showed overall erosion, with an average rate of -8 mm/yr. Both the Hakioawa and Kaulana watersheds showed an average rate of change of -1 mm/yr, and both included sites that showed erosion and sites that showed deposition.
For sites with negative rates of change indicating erosion, the average rate of change during the monitoring period was -2 mm/yr in restoration areas and -3 mm/yr in nonrestoration areas. For sites with positive rates of change indicating deposition, the average rate of change was 5 mm/yr in restoration areas and 6 mm/yr in nonrestoration sites. The average rate of change for rills was 1 mm/yr in restoration areas and -2 mm/yr in nonrestoration sites. The average rate of change for interfluves was below detection in restoration areas and -1 mm/yr in nonrestoration areas.
Potential Use and Limitation of Data.-Additional statistical comparisons of various subsets of erosion data can be used to assess the effectiveness of restoration efforts or how existing landforms, vegetation, climate, and other physical basin characteristics affect erosion and fluvial sediment transport in the watersheds. Further investigation to identify what factors cause the Kaulana watershed to have much lower runoff and sediment loads than the Hakioawa watershed may yield valuable information for developing and modifying restoration strategies. Continued monitoring of streamflow, sediment transport, and erosion is key to assessing the long-term effectiveness of restoration and can provide insight to the island's recovery since the eradication of feral goats and cessation of use as a military bombing range; the results of this study provide the