Tropical Dry Forest Restoration in Hawaii



Allen, W. 2000. Restoring Hawaiis Dry Forests. BioScience 50(12):1037-1041


Work done on upper part of Kaupulehu

North and upslope from Kona

Fenced in 1956

20 rain per year

On lava flow (1800s)

Originally native dry forest


Causes of damage

Cattle grazing

Feral ungulates (cattle, pigs, goats)

Non-native grasses

West (dry) side:

fountain grass Pennisetum setaceum

East (wet) side:

kikuyu grass Pennisetum clandestinum

Grass-fed fires


Extent of damage

90% of dry forest is gone

40% of rain forest is gone


Initial restoration and investigation

Kaupulehu is fenced and buffered from fire by lava flow

Endangered spp. raised from seeds collected on site, outplanted

Rodents suspected because of lack of seedlings

Bait traps used

Fountain grass cut in 1995


Initial restoration and investigation, 2

Fountain grass down to 10% by 1996.

By 1997, canopy tree seedlings had increased from 0 to 838 in 53 plots

Diospyros sandwicensis (lama) (persimmon)

Subsequent drought slowed regeneration

Sub-canopy microhabitat still not there

Shading and drip irrigation proposed


Cabin, R.J., S.G. Weller, D.H. Lorence, S. Cordell, L.J. Hadway, R. Montgomery, D. Goo, A. Urakami. 2002. Effects of light, alien grass and native species additions on Hawaiian dry forest restoration. Ecological Applications 12 (6): 1595-1610.


Experiment done on lower part of Kaupulehu

Fenced in 1997

Same problems: heavy grazing, fountain grass


Experimental design

Two blocks

50% shade

Full sun

Fountain grass control



Plastic mulch


Species additions





Herbicide was 1% Roundup as grass was cut in January and as it grew back in April

Mulch site was cut to ground and covered with black plastic (for 120 d), which was removed at the start of the experiment

Trim: fountain grass was cut down to 60 cm.


Outplanting: 44 plants (7 spp) put into 1.8 x 1.8 m plot (32 cm spacing)

Seeding: mix of 12 native spp, pre-soaked, (numbers per spp varied)

Watered more heavily at first, then once a week

Planting done on 17 May

Experiment conducted for 20 months


Fountain grass cover greatest in trimmed sites, lowest in bulldozed sites

Native cover the reverse

Native cover high in shaded plots

Most natives in outplanted plots, next seeded, then control

Native and grass cover negatively correlated

Greatest outplant survival in shade, herbicided plots, plastic mulch


Dry forest restoration should focus on:

Reduction or removal of non-native ungulates

Control of dominant introduced grasses*

Use or creation of favorable microsites

Re-introduction of selected native species

Fire must be controlled*


In addition, seed dispersal is poor

Little wind dispersal

Almost no animal dispersal

Loss of native bird population in Hawaii

Seeds short-lived