A geographical analysis of change in a Hawaiian sugarcane plantation

Abstract

Efficient use of irrigation water when water supplies are limited is crucial to sugarcane production. This project examines technological change in relation to water management at Hawaiian Commercial and Sugar Company (HC&S) of Puunene, Maui in the context of spatial information management. These include the change from furrow irrigation to drip irrigation, the discontinuation of the evaporation pan network, and managing with a computer water balance model.

A field history database of sugarcane harvests at HC&S provided data for a comparison of furrow irrigation and drip irrigation. A combination of statistical and mapping tools were used to evaluate soil, climate, and management variables over space and time for the entire plantation. The topic of greatest concern to HC&S management is "Where to put water when water is short". An objective of this study was to use a simple geographical information system (GIS) system to spatially organize soil and weather data needed for water allocation decisions.

SCS Soil Survey 7.5 minute quadrangles Wailuku, Maalaea, Paia, Puu o Kali, and Haiku were combined to form a continuous soil map for HC&S plantation. HC&S field boundaries were overlaid with the soil map and a database of soil types by field was created. Forty-five years of harvest information including yield, irrigation, and climate variables for almost 3,000 harvests were analyzed spatially using maps and over time using both maps and graphs. Multivariate analysis techniques were used to analyze relationships between variables with different spatial groups.

Sugarcane yields increased after the plantation converted to drip irrigation. The spatial pattern of yields also changed. With furrow irrigation the highest yields were in the Keahua division which had silty clay loam soils. With drip irrigation, coarser soils in the Maalaea division became the highest yielding. With the discontinuation of the network of evaporation pans in the late 1980s the scale of information collected on water demand was reduced. Water management by computer depends on representative data. Improved weather data will help direct where irrigation water is most needed.