Image from Levia, D. F., and S. Germer (2015)
A primary goal of the Honolulu Board of Water Supply (BWS) and its Watershed Program is to ensure an adequate supply of fresh water for current and future generations.
In this regard, the capacity of O`ahu’s watersheds to capture and store precipitation is critical: it is the sole natural source of fresh water supply for the island.
Our BWS Canopy Interception Study attempts to quantify differences in rainfall capture between native and invasive forests found in the Wai`anae Mountains.
This aspect of the water budget of O`ahu’s watersheds (and of watersheds in general) has not been well-studied. Until recently, the only similar research in Hawai`i was located in high-elevation “cloud forest” terrain on the island of Hawai`i, where fog drip is a primary component of precipitation. Because that research was not representative of the majority of forested areas on O`ahu, this study was established as a first step to address O`ahu's forests.
Canopy Interception
RF = TF + SF + CI → CI = RF – TF - SF
where
RF = gross rainfall TF = throughfall
SF = stemflow CI = canopy interception
By measuring gross rainfall, throughfall and stemflow in the field at native and invaded forest sites, the canopy interceptions of the native and invaded sites can be calculated and compared.
Image based on Rainfall Atlas of Hawaii (University of Hawaii, 2011)
For this study, the native and invaded forest test plots are located in the Mākaha watershed of O`ahu, part of the Wai`anae Mountains. This is considered a high-priority watershed by BWS in terms of watershed protection and restoration efforts, ranking high with respect to groundwater recharge and groundwater production.
The Mākaha watershed is largely undeveloped; land cover is generally shrubs and evergreen forest. Elevation across the watershed ranges from sea level to 4,025 feet at Mount Ka`ala, the highest point on O`ahu.
Historical rainfall patterns across the Mākaha watershed are shown above. As expected, average annual rainfall is greater along the mountain slopes compared to lower elevations, reflecting the capacity of the higher mountain slopes to capture the predominant northeast tradewind-borne precipitation. The rainy season on O`ahu extends roughly from October through April.
Example test plot layout, for native plot
Gross Rainfall Equipment
Because the native and invaded test plots are located within close proximity to each other, a single tipping bucket rain gauge (standard 8-inch diameter, 0.01-inch / tip) installed near the plots collects gross rainfall data to represent both the plots.
Throughfall Equipment
For each test plot, a total of three troughs lead to a fabricated tipping bucket (150 mL / tip) rain gauge. The rain gauge is nested in a stainless steel can, levelled and anchored to the ground for stability.
Troughs were installed under trees typical of the test plots (species, diameter, height) in configurations that represented the average canopy and gap proportions and minimized overlap of throughfall collection between troughs.
Stemflow Equipment
Stemflow collectors connected to a fabricated tipping bucket (150 mL / tip) rain gauge were used for each test plot.
Trees selected for stemflow data collection represented the range of typical species, diameter, and height found commonly on the test plots.
Stemflow collectors were supported through lengths of 7/8-inch diameter poly tubing protected in 1-inch diameter high density polyethylene (HDPE) pipe, leading toward the rain gauge, nested in a stainless steel can, levelled and anchored.
The native and invaded forest test plots were selected based on elevation, aspect, gross rainfall, and vegetation type. An effort was made to select test plots having similar characteristics except for vegetation type, to minimize any influences on canopy interception other than vegetation type.
Elevation: About 80 percent of the Mākaha watershed lies below the elevation considered to be the lower boundary of “cloud forest”, where fog drip becomes a primary component of precipitation (approximately 2,460 feet MSL). Therefore, the native and invaded test plot locations were selected below this elevation to represent the majority of the watershed area (at about 1,640 feet and 1,605 feet MSL, respectively).
Aspect: In physical geography, aspect refers to the horizontal direction to which a mountain slope faces. The aspect of a slope can significantly influence its local climate (microclimate). To minimize this influence, test plot locations having relatively minor slope (average of about 10 degrees) and similar aspect (approximately northwest) were selected.
Gross rainfall: Test plot locations were selected close to one another (approximately 200 feet apart), within the same range of historical rainfall (greater than 65 inches annually). Also, for practical purposes, the historical rainfall amount is on the higher end for the watershed. Given the higher anticipated gross rainfall, differences in throughfall, stemflow and canopy interception may be more readily detected.
Vegetation type: The native forest test plot was selected with an array of native trees and plants expected for a relatively intact forest at this elevation in the Wai`anae Mountain range. In contrast, the invaded forest test plot was selected with a classic monotypic character; in this case, the test plot contained almost exclusively Psidium cattleianum (strawberry guava). Strawberry guava is considered to be one of the most destructive invasive species in Hawaii by the U.S. Department of Agriculture, the State Department of Land and Natural Resources, the Hawaii Invasive Species Council, and the watershed conservation community in general.
Each of the two test plots is square-shaped, and 40 feet by 40 feet in extent. The plots were temporarily gridded into 10 foot by 10 foot sectors, and plant surveys were conducted to inventory all species within the test plots. The native test plot contained alahee, kukui, koa, lama, maile, olopua, ti, and palapalai fern. Minor amounts of nonnative species such as coffee, Christmas berry, and basket grass were also present. The invaded test plot contained almost exclusively strawberry guava. Minor amounts of native species, particularly maile, were also present.
Preliminary data for April 2017 through April 2018 show that native forest throughfall is significantly greater than invaded forest throughfall, while invaded forest stemflow is significantly greater than native forest stemflow.
Long-term data collection and quantification is expected, as well as ongoing collaboration with University of Hawaii, studying related water budget characteristics (evapotranspiration, soil moisture). Finally, a CI Study is being established near the boundary of the Nuuanu / Manoa watersheds, to assess native versus invaded rainfall capture in the Ko`olau Mountains.