MONITORING EROSION ON MONITORING EROSION ON KAHO‘OLAWEKAHO‘OLAWE

U.S. Department of the InteriorU.S. Geological Survey

By Scot Izuka, USGSBy Scot Izuka, USGSPresented at the Presented at the Hawaii Water Quality Hawaii Water Quality Conference Conference March 25, 2008March 25, 2008

A USGS study in cooperation with the Kaho‘olawe Island Reserve CommissionA USGS study in cooperation with the Kaho‘olawe Island Reserve Commission

Area = 45 mi2

Uninhabited Rainfall ~ 25 in/yrNo perennial streams

400 AD – Earliest human habitation 400 AD – Earliest human habitation (archaeological evidence)(archaeological evidence)

Early times – grass and trees grew in Early times – grass and trees grew in thick soil on most of islandthick soil on most of island

Wild goats (~1800 to 1990s)Wild goats (~1800 to 1990s)

Sheep and cattle (1858 to 1952)Sheep and cattle (1858 to 1952)

Target bombing by U.S. military (1941 Target bombing by U.S. military (1941 to 1993)to 1993)

UXO clearing (1998-2003) 74% of UXO clearing (1998-2003) 74% of island cleared; 9% to depth of 4 feetisland cleared; 9% to depth of 4 feet

Restoration projects, including Restoration projects, including revegetation (2001 – present)revegetation (2001 – present)

(Sources: Kaho‘olawe Island Reserve Commission; (Sources: Kaho‘olawe Island Reserve Commission; Macdonald and others, 1983)Macdonald and others, 1983)

History History

HakioawaHakioawaDenudation by grazing

Upper elevations bare

Restoration EffortsRestoration Efforts

Motivation for StudyMotivation for Study

Need to assess effectiveness of restoration Need to assess effectiveness of restoration efforts in reducing erosion on Kaho‘olaweefforts in reducing erosion on Kaho‘olawe

ObjectiveObjective

Monitor erosion and sediment transport in Monitor erosion and sediment transport in Hakioawa and Kaulana watersheds, which are Hakioawa and Kaulana watersheds, which are currently undergoing restorationcurrently undergoing restoration

ApproachApproach

Periodic measurements of soil erosion at Periodic measurements of soil erosion at selected locations (both in restoration and selected locations (both in restoration and non-restoration areas)non-restoration areas)

Monitor streamflow and suspended-sediment Monitor streamflow and suspended-sediment discharge at mouths of streams discharge at mouths of streams

Helicopter View of Bare AreaHelicopter View of Bare Area

BadlandsBadlands

Badlands Erosion in Soft MaterialBadlands Erosion in Soft Material

Pedestal RocksPedestal Rocks

Upper Horizon is Removed in Many Upper Horizon is Removed in Many AreasAreas

Remnant Remnant hummock of hummock of upper soil upper soil horizonhorizon

Soil stripped, leaving Soil stripped, leaving underlying hardpanunderlying hardpan

Erosion Monitoring MethodErosion Monitoring Method

Install transects (“pins”)Install transects (“pins”)

Measure ~ every 6 Measure ~ every 6 months for 3 yearsmonths for 3 years

HakioawaHakioawaKaulanaKaulana

HakioawaHakioawa

Restoration Restoration startedstarted

No restoration No restoration

RillRill InterfluveInterfluve

HummockHummock

PRELIMINARY RESULTSPRELIMINARY RESULTS

FIELD-MEASUREMENT FIELD-MEASUREMENT DATESDATES

1.1. January 2007January 2007

2.2. September 2007September 2007

3.3. March 2008March 2008

Rainfall Between MeasurementsRainfall Between Measurements

Dec. 2007

NWS Gage 499.6 Kaho‘olawe

Data from National Climatic Data CenterData from National Climatic Data Centerand National Weather Serviceand National Weather Service

Change Between January 2007 and Change Between January 2007 and March 2008March 2008

(negative = erosion)(negative = erosion)

Rill 7 -2 0Interfluve 0 -1 -1Hummock NA -6 -6All 2 -2 -1

Average Change (mm/yr)Restoration

startedNo

RestorationAllFeature

Statistical Significance?Statistical Significance?

Erosion rate of restoration sites was Erosion rate of restoration sites was statistically less than that of non-statistically less than that of non-restoration sites (one-tailed test, 95% restoration sites (one-tailed test, 95% confidence interval)confidence interval)

Erosion rate of rills was not Erosion rate of rills was not significantly different from that of significantly different from that of interfluves (two-tailed test, 95% interfluves (two-tailed test, 95% confidence interval)confidence interval)

Even So, Rills Show Extremes of Even So, Rills Show Extremes of Erosion/Deposition Erosion/Deposition

H15H15

H35H35

H37H37

Where Are the Biggest Changes? Where Are the Biggest Changes?

H25H25

H14H14 H19H19

H17H17

H41H41

Some Variation in Rills and Some Variation in Rills and Interfluves not Accounted for in Interfluves not Accounted for in

Sampling DesignSampling Design

Are the transects representative of Are the transects representative of the entire watershed?the entire watershed?

Assume that we have enough Assume that we have enough transects to eliminate bias in transects to eliminate bias in computed statisticscomputed statistics

Restoration Effect Likely to Mature over Restoration Effect Likely to Mature over TimeTime

Plants are small at Plants are small at presentpresent

Erosion rate likely to Erosion rate likely to change in the future change in the future as plants growas plants grow

Erosion data Erosion data collected today can collected today can be used basis for be used basis for comparison in futurecomparison in future

No Conclusions, Just ObservationsNo Conclusions, Just Observations

Rills show extreme variability in deposition Rills show extreme variability in deposition and erosionand erosion

Erosion rate less extreme in interfluves but Erosion rate less extreme in interfluves but mean is not statistically different from rillsmean is not statistically different from rills

Erosion rate high in hummocksErosion rate high in hummocks

Erosion rate statistically lower in restoration Erosion rate statistically lower in restoration than in non-restoration areas (but there is than in non-restoration areas (but there is question of whether transect data are question of whether transect data are representative of the whole watershed)representative of the whole watershed)

Erosion rate is likely to change as plants Erosion rate is likely to change as plants grow; data collected today can be basis for grow; data collected today can be basis for comparison with a future studycomparison with a future study

Stream Gage and Automatic Stream Gage and Automatic Sediment SamplerSediment Sampler

SummarySummary

All preliminary – after only one year of All preliminary – after only one year of erosion monitoring erosion monitoring

Rills show extremes for both deposition Rills show extremes for both deposition and erosion, but the mean change was and erosion, but the mean change was not statistically different from interfluvesnot statistically different from interfluves

Erosion also high in hummocksErosion also high in hummocks

At this stage, erosion rate in restoration At this stage, erosion rate in restoration areas is statistically lower than in non-areas is statistically lower than in non-restored areasrestored areas

More monitoring to come!More monitoring to come!