The original file was created by the Hawaii Natural Heritage Program. Please see Entity Attribute Overview Description for more information.
Sam Gon III, Director of Science The Nature Conservancy of Hawai'i
Basis for ecoregional subunits: The hierarchical nature of the natural community classification currently in use by The Nature Conservancy of Hawai'i is based on divisions of three major parameters: Elevation, Moisture, and Physiognomy (see Atlas of Hawai'i, Chapter "terrestrial Ecosystems" describing in more detail the classification hierarchy). The subregional units defined below and provided via GIS layers are based on the current and presumed distributions of units based largely on a middle level of the hierarchy. The units that emerged correspond to what seem to be the most consistent and discernible Elevation-Moisture-Physiognomy triplets. With some lumped in the interest of keeping the number of units reasonably small and manageable.
C COASTAL (all lumped for the present)
LDS LOWLAND DRY SHRUBLAND/GRASSLAND
LDF LOWLAND DRY FOREST/SHRUBLAND
LMF LOWLAND MESIC FOREST/SHRUBLAND
LWF LOWLAND WET FOREST/SHRUBLAND
MDF MONTANE DRY FOREST/SHRUBLAND
MMF MONTANE MESIC/FOREST/SHRUBLAND
MWF MONTANE WET FOREST/SHRUBLAND
SDF SUBALPINE DRY FOREST/SHRUBLAND/GRASSLAND
A ALPINE (all lumped for the present)
WC WET CLIFFS (lowland and montane lumped)
DC DRY CLIFFS (all elevations lumped)
Lumping decisions: The 12 units that emerge result in a manageable set of units that have conservation meaning and are usually readily distinguishable from one another. Similar physiognomic units (e.g., adjacent lowland mesic forest types would be treated together as a unit) as well as those that typically form complex spatial mosaics in an area were lumped (e.g., mosaics of forest and shrubland, or of shrubland and grassland). This is not arbitrary. Native forests run from closed canopy to sparsely distributed trees growing among shrubs and grasses. Dry native shrublands and grasslands also very typically occur together in the landscape. The following table discusses briefly these and some of the other major considerations:
C generally consistent set of salt-tolerant taxa and conditions
LDS driest native vegetation types (e.g., Heteropogon, Sida, etc., too dry for persistence of native tree species, extremely fire prone
LDF typically dominated by one or more of Erythrina, Diospyros, Sapindus, etc., manageable for dry forest maintenance, greater plant diversity, fire prone
LMF many potential dominant taxa, great jump in plant diversity, highest tree diversity, less fire prone than LD units
LWF typically dominated by Metrosideros, with a variety of codominant taxa, good plant diversity, different understory taxa (e.g., Cyrtandra, lobeliads, ferns) than in montane wet systems, especially on older high islands (Kaua'i, O'ahu, Moloka'i, Läna'i).
MDF restricted to Maui/Hawai'i, typically dominated by Metrosideros, but other dominant taxa possible, and several diagnostic communities present (e.g., Chamaesyce MDF); sometimes difficult to discern transition with lower SDF
MMF typically Acacia and/or Metrosideros dominated, but with diagnostic constituent species, and wet indicators lacking/inconspicuous; important forest bird habitat, prone to cattle ranching, logging, several diagnostic communities (e.g., Acacia/Sapindus-Metrosideros MMF, Nestegis MMF)
MWF consistently Acacia and/or Metrosideros dominated, typical codominants of Cheirodendron, Cibotium, and/or Dicranopteris/Sticherus, includes bogs
SDF typically Sophora and/or Myoporum, but other diagnostic communities present, including a number of shrublands and grasslands, no abrupt transition with MDF
A relatively few vegetated units here,
WC typically marked by near-vertical, wind-swept, wet, slopes covered with shrubs and ferns, waterfalls-seeps, meaningful to conservation because of difficult management access
DC consistent set of dominants, such as Eragrostis, Artemisia, Bidens, etc. prevalence of lichens and bare rock between vegetated areas, same management difficulties as WC
Sources of information: The units depicting those native-dominated areas that remain to us today were derived from a combination of sources:
1. field surveys (with direct mapping of physiognomic units and dominant canopies), 2. helicopter overflights (where dominant canopies and sometimes understory conditions could be identified and mapped while in flight), 3. satellite/remote imagery (SPOT output for all islands, and C-130 multispectral passes over E Maui) 4. aerial photographs (e.g., USGS orthophotoquads, 1970s B&W, and color false infrared aerials from the 70s, 80s, and early 90s). 5. preexisting vegetation maps (especially those of Jacobi 1985+)
Defining native-dominated: Native-dominated was defined relatively crudely as an area showing 50% or more native canopy and occupying an area of greater than 100 ha (about 250 acres). A few exceptions were made, in order to depict some isolated and significant units, such as remnant native dry vegetation at Pu'u o Kalï, East Maui (about 90 ha). There are many areas that sit very close to the 50% native cutoff, especially in the lowlands of Hawai'i Island. Metrosideros/Pandanus Lowland Wet Forest stands in the vicinity of the Hilo Airport are among these marginal examples.
Correspondence with the "real world": Patterns that emerged from the current distribution of remaining native units (e.g., the distribution of dry and mesic forest remnants on O'ahu) were compared against maps depicting elevation and general rainfall conditions, such as isohyetal maps (e.g., Hawai'i Water Authority 1959, Daniels, in Armstrong 1983) for correspondence. There was good general correspondence, with exceptions that could be explained by factors such as substrate, slope, aspect, presence of groundwater, etc., that might make prevailing soil conditions wetter or drier than isohyetal gradients alone might suggest. However, in general, the conclusions settled on by the advisors that created the natural community classification held true. If there was a conflict between the distribution of a known physiognomic unit and an arbitrary division (e.g., Metrosideros montane wet forest extending above what would normally be the montane-subalpine boundary at 2000 m elevation) the physiognomic unit boundary took precedence. Many examples of this kind of adjustment in the elevation boundaries are apparent on Maui and Hawai'i maps, where, thanks to Jacobi mapped units, we could make the necessary overriding decisions. These adjustments need to be made by those most familiar with the current ground conditions, and we expect some adjustments will be needed on several of the maps. For example, the dry-mesic boundary in the vicinity of Kïlauea on Hawai'i Island should be adjusted to include more mesic areas where dry forest is now depicted.
Reconstructing the past: Prehuman distribution of subregional units were generated by expanding the current patterns into those areas devoid of native ecosystems. Infrequently there were sources that described vegetation in historical times (e.g., Degener, Rock, Hillebrand, Bryan, Munro, etc.) that allowed for some partial corroboration of the reconstructions. At other times (e.g., for Ni'ihau), there were no such sources, and current isohyetal patterns were overlaid on the island, and general moisture and presumed physiognomic conditions were assigned within the pre-defined elevation and moisture zones.
Vegetation and moisture adjustments for reconstructions: Being aware that existing vegetation can ameliorate prevailing moisture and other conditions, we presumed that areas now considered dry, but near-mesic (e.g., from 40 to 50 inches annual rain) probably supported mesic vegetation in prehuman times, when vegetation canopies were more intact. Extension of mesic vegetation downslope for prehuman Läna'i is one example. We also considered areas subject to consistent cloud-interception, assigning them wetter conditions than rainfall figures might suggest.
Why assign cliff communities? Cliff communities were assigned on the basis of shared physiognomy (slopes too steep to support trees), repeatable sets of communities (generally dominated or characterized by a variety of dwarf-shrubs, grasses, ferns, etc.), manageability challenges (dangerous and difficult access), and cliff-related conditions (frequent landslide disturbance, extreme drainage conditions, often windswept, rapidly drying unless predominantly very wet). There were no mesic cliffs assigned, because cliffs occurring in otherwise mesic situations tend to be dry. It is very interesting that cliffs tend to be natural dividers between lowland and montane units. For example, the wet lowlands east of Wai'ale'ale lie well below 1000 m elevation, and are separated from the montane plateau of the Alaka'i Swamp by wet cliffs. Other salient examples of this phenomenon are seen in such plateaus as Nämolokama on Kaua'i, Oloku'i on Moloka'i, Lïhau and Helu in the West Maui Mountains, etc. These cliffs have at least until relatively recently provided natural protection to these montane remnant native ecoregional units.