We created detailed classification maps of native and non-native plants on Lāna‘i by combining high-resolution WorldView-2 satellite imagery with ground-collected data. We used the high-resolution maps and historical satellite and aerial imagery to analyze changes in ‘ua‘u—Hawaiian petrel, Pterodroma sandwichensis—nesting habitat. For the first time at island scales in Hawai’i, this species-level mapping used high resolution—two-meter resolution—satellite imagery with machine learning classification methods. This work can serve as a template for mapping other islands.
Overview
There is an ongoing landscape-scale effort to restore a portion of important ecological and cultural features on Lāna‘i. However, there was no reliable baseline data about the current state of vegetation, which is essential to monitor effects of restoration efforts and to fine-tune plans along the way. To address this gap, researchers employed a two-fold approach: (1) creating the species-level vegetation maps and (2) using those maps to analyze habitat change over time. Results are now informing conservation work within the Project Area—the eastern portion of the island spanning from the Keomoku coast up to the island’s highest area above Lāna‘ihale.
Highlights:
USGS, in partnership with University of Hawai‘i at Mānoa and Pūlama Lāna‘i, developed high-resolution vegetation maps for the Project Area to identify recent habitat loss and incipient invasive plant populations. The maps characterize vegetation communities across the landscape to serve as the foundation for developing resource management strategies. Using these datasets, the team created fine-scale vegetation models to track changes in the uluhe (Dicranopteris linearis), a native fern that provides nesting habitat for ‘ua‘u at Lāna‘ihale.
Other potential uses:
Vegetation maps can be used by to managers to track the impacts of planned ungulate (hooved animal) suppression within the Project Area. Changes in wildfire fuel loads after ungulate removal aren’t assessable without these vegetation datasets.
These high-resolution maps can aid in erosion and sediment control planning by quantifying changes in bare soil exposure and characteristics of areas vulnerable to and sources of erosion.
Vegetation models provide a tool for land managers to understand baseline vegetation cover and track changes in vegetation over time in response to on-the-ground efforts, non-native plant invasion, and climatic changes.
Below are data releases related to the topic.
High-Resolution Land Cover Maps of Lāna‘i, Hawai‘i, 2020 High-Resolution Land Cover Maps of Lāna‘i, Hawai‘i, 2020
Below are publications related to this topic.
Fine-resolution land cover mapping over large and mountainous areas for Lāna‘i, Hawaii using posterior probabilities, and expert knowledge Fine-resolution land cover mapping over large and mountainous areas for Lāna‘i, Hawaii using posterior probabilities, and expert knowledge
We created detailed classification maps of native and non-native plants on Lāna‘i by combining high-resolution WorldView-2 satellite imagery with ground-collected data. We used the high-resolution maps and historical satellite and aerial imagery to analyze changes in ‘ua‘u—Hawaiian petrel, Pterodroma sandwichensis—nesting habitat. For the first time at island scales in Hawai’i, this species-level mapping used high resolution—two-meter resolution—satellite imagery with machine learning classification methods. This work can serve as a template for mapping other islands.
Overview
There is an ongoing landscape-scale effort to restore a portion of important ecological and cultural features on Lāna‘i. However, there was no reliable baseline data about the current state of vegetation, which is essential to monitor effects of restoration efforts and to fine-tune plans along the way. To address this gap, researchers employed a two-fold approach: (1) creating the species-level vegetation maps and (2) using those maps to analyze habitat change over time. Results are now informing conservation work within the Project Area—the eastern portion of the island spanning from the Keomoku coast up to the island’s highest area above Lāna‘ihale.
Highlights:
USGS, in partnership with University of Hawai‘i at Mānoa and Pūlama Lāna‘i, developed high-resolution vegetation maps for the Project Area to identify recent habitat loss and incipient invasive plant populations. The maps characterize vegetation communities across the landscape to serve as the foundation for developing resource management strategies. Using these datasets, the team created fine-scale vegetation models to track changes in the uluhe (Dicranopteris linearis), a native fern that provides nesting habitat for ‘ua‘u at Lāna‘ihale.
Other potential uses:
Vegetation maps can be used by to managers to track the impacts of planned ungulate (hooved animal) suppression within the Project Area. Changes in wildfire fuel loads after ungulate removal aren’t assessable without these vegetation datasets.
These high-resolution maps can aid in erosion and sediment control planning by quantifying changes in bare soil exposure and characteristics of areas vulnerable to and sources of erosion.
Vegetation models provide a tool for land managers to understand baseline vegetation cover and track changes in vegetation over time in response to on-the-ground efforts, non-native plant invasion, and climatic changes.
Below are data releases related to the topic.
High-Resolution Land Cover Maps of Lāna‘i, Hawai‘i, 2020 High-Resolution Land Cover Maps of Lāna‘i, Hawai‘i, 2020
Below are publications related to this topic.