Vegetation above a feral pig barrier fence in rain forests of Kilauea's East Rift, Hawaii Volcanoes National Park

Reports were scanned in black and white at a resolution of 600 dots per inch and were converted to text using Adobe Paper Capture Plug-in.Immediately after the 1993 construction of a barrier fence to block the movements of feral pigs in forests of Kilauea's East Rift within Hawaii Volcanoes National Park (HAVO), a systematic framework of transects and plots was established for collection of baseline vegetation data upslope of the fence. Distribution and estimated abundance of the most invasive alien plant species were determined. The most widespread alien grass species was Hilo grass (Paspalum conjugatum); although it typically had low estimated cover values, this grass was almost ubiquitous. The most common invasive alien tree species was strawberry guava (Psidium cattleianum); it was found along transects in the western half of the study area, where its cover was estimated as 5-25% or 25-50%. Firetree (Myrica faya) occurred less frequently and had lower cover values than strawberry guava. Two other invasive woody species were found at low frequency or in limited areas; yellow Himalayan raspberry (Rubus ellipticus) was restricted to the slopes of Kane Nui o Hamo, and cane tibouchina (Tibouchina herbacea) occurred infrequently at widely scattered localities. Alien plant frequencies from the current survey were compared with those from a previous plant survey in 1988. A third of the alien plant species along transects, including firetree, yellow raspberry, and strawberry guava, had very similar frequencies on both surveys. Frequency of Hilo grass and scaly swordfern (Nephrolepis multiflora) increased greatly in the study area between the two surveys. Some of the observed changes in vegetation may have been influenced by recent disturbance to the forests by cinder deposits from Pu'u '0'o, in an earlier phase of the current eruption. The locations of rare native plants were mapped along transects, and numbers were compared with those from the previous survey of 1988. The endangered pendent kihi fern (Adenophorus periens) was not relocated on Park transects; this species may have disappeared from the slopes of Kane Nui o Hamo in the last five years. Koli'i (Trematolobelia grandifolia), a "species of concern" has persisted on Kane Nui o Hamo, and its current size class structure indicates a stable population. The 12 rare plant species that were observed on East Rift transects were concentrated in several sites, including Kane Nui o Hamo, forests south and west of Napau crater, relatively open forest southeast of the 1840 flow, and the southwest corner of the study area near the Naulu Trail. Frequencies of 'oha (Clermontia spp.), indicator species for pig damage in Hawai'i, were relatively high overall in the study area, although the impact of pig predation was indicated by the paucity of terrestrial plants and a low freqeuncy of large Clermontia. Remonitoring a subset of transects after 1.5 years revealed that terrestrial Clermontia declined in frequency, while epiphytic plants increased over the same period. Tree fern density in the study area was 38/100m2, and the trunk height class of 1-2 m was well represented in East Rift forests. A higher density of tree ferns was observed in the western half of the study area, primarily due to the greater number of tree ferns >1m. There were fewer tree ferns on the lower halves of three main transects than were found on the upper reaches, farthest away from the barrier fence and upslope from uluhe-dominated forest. 'Olapa (Cheirodendron trigynum) appeared to be an inconsistent indicator species for pig activity. This important rain forest tree was maintaining a stable population in East Rift forests, despite the long-term presence of feral pigs there. Differences in 'olapa density were noted in the western versus the eastern, unprotected part of the study area, where lower numbers of terrestrial 'olapa saplings were seen. When compared with the upper transect reaches, fewer 'olapa saplings were found along the lower portions of the three main transects, in the area near the open end of the barrier fence. Vegetation cover and woody plant density of recently pig-disturbed sites were examined in 1994 using 39 vegetation plots, and a subset of 26 plots was remonitored 1.5-2 years later. The vegetation of disturbed East Rift forests was found to be poor in native woody plant species, and ground cover was very sparse. In the interval between monitoring, the cover of two alien species (Hilo grass and scaly swordfern) increased dramatically. Early succession indicates that these two plants will become dominant components of ground cover in pig-disturbed areas. Little change was noted in native woody plant density overall, but several native species, including tree ferns, displayed increases over the 1.5 year monitoring period. Feral pig density, estimated from frequency of pig activity, was low in the East Rift study area (averaging 1.9 pig/km2), even before systematic control efforts began. Quarterly pig activity monitoring revealed an increase in activity along Park transects until the end of 1993, followed by a steady decline to 2.4 pig/km2 by the end of the study in January 1996. Data from this study suggest that the park's snaring efforts reduced the feral pig population in the most protected, interior part of the forest upslope of the barrier fence within two years of the project initiation. Success in lowering pig density was not observed on transects at or beyond the open terminus of the barrier fence, indicating that pig ingress continued throughout the study. The current survey may only be considered a baseline study of vegetation and pig activity in managed East Rift forests. Future remonitoring may be needed to evaulate the success of this management effort in promoting long-term recovery of native vegetation.National Park Service Cooperative Agreement 8010 2 900