Long before the last embers had cooled from the Broomsedge Fire at Hawai`i Volcanoes National Park in July 2000, Rhonda Loh had begun plotting ways to defeat the next one. By August, Loh, at the time a botanist with the Hawai`i Volcanoes National Park, had mapped out her plan for revegetating the 1,000-acre burn area in a way intended to minimize fire damage as well as reduce its risk and had secured funds for the work.
Over the next three years, Loh, other National Park Service employees and an army of volunteers labored to rebuild the area with native plants. By the time their work ended, they had planted more than 19,000 seedlings raised in park nurseries and had sown more than 3 million seeds.
On a recent day, Loh, now acting chief of natural resources management for the park, gives a tour of the burn area to a visitor. She takes obvious delight in pointing out the new green growth peeking above and through the ubiquitous broomsedge and beardgrass. At times, the slightly built Loh is herself barely visible in the dry grasses, which can attain heights of five feet or more.
Almost 70 percent of the fire-damaged `ohi`a trees, the backbone of this seasonally dry woodland community, survived, Loh points out. Mamane trees planted after the fire are thriving. The understory is thick with `a`ali`i. Tall young koa trees dot the landscape, towering above all but the oldest `ohi`a snags. In between head-high bunches of alien grasses, small pukiawe and ohelo can be seen.
Further in from the Belt Highway that bisects the park is the mesic koa forest. When the fire reached this area, its progress was slowed by the moist environment provided by the dense canopy and deep carpet of green meadowrice grass. Today, scorch signs are barely visible. The koa have grown suckers, and the native trees and shrubs planted after the fire are thriving.
Carefully, Loh strides through the thick grass mat that hides the treacherous `a`a lava, making a beeline for a small, glossy-leafed tree on a nearby rise. It’s an alani, one of the rarer plants in this part of the island. She tugs off a leaf, breaks it apart, and drinks in its lemony fragrance. This, Loh says with a contented sigh, is the very best part of her job.
Until recently, fires were relatively infrequent in the national park. From the time records began to be kept and for the next 40 years, on average, less than a fire a year occurred. From 1924 to 1963, the total area burned amounted to fewer than 215 acres, and the average size of a fire was about 6 acres. Most fires occurred in areas frequented by visitors, in pastures that were leased to cattle ranchers, or in areas where Civilian Conservation Corps crews were working.
The next three decades saw a marked increase in fire frequency as well as size. Between 1964 and 1995, 97 fires occurred, burning almost 35,000 acres. According to a paper co-authored by Tim Tunison (Loh’s predecessor as natural resources manager), and Carla D’Antonio of the University of California-Berkeley, and Loh, “the approximately three-fold increase in fire frequency and nearly 60-fold increase in fire size followed the introduction and spread of alien fire-promoting grasses starting during the 1960s.”
The change resulted not so much from any rise in the natural or human sources of ignition, but more from an invasion of tall, dry grasses that provided abundant fuel for the fires. In 1959, a botanical survey of the park found no broomsedge or beardgrass, but just seven years later, they were the dominant species in two of the park’s major ecosystems – coastal grasslands-scrub and the seasonally dry, `ohi`a-dominated woodlands. Today, the two species of alien grasses – beardgrass from tropical America, broomsedge from the southeastern United States – form nearly a third of the biomass in the understory of the seasonally dry woodland, and two-thirds or more of the ground cover.
Before the grasses moved in, ground cover in dry, `ohi`a-dominated woodlands was patchy and consisted of native shrubs, lichens, sedges, and occasionally a native grass. Fires were rare. But from the mid-1960s to the mid-1990s, 81 percent of the areas invaded by the grasses had burned.
The burned areas gave researchers at the park, including Loh, ample opportunity to study natural recovery rates. The two dominant species in the dry `ohi`a woodland before fire — `ohi`a and pukiawe – suffered the most, with 55 percent of `ohi`a killed and pukiawe suffering a 10-fold decline in terms of the percentage of ground cover it made up, and a three-fold loss in density. Seedlings came up, but failed to mature in the presence of fast-growing alien grasses. On the other hand, several native species tolerated the fire well. `A`ali`i was killed by fire, but grew back to its pre-fire abundance. Mamane resprouted even after severe fires.
The researchers studied the impacts of fire in all areas of the park where alien grasses had become established – coastal grasslands, rain forest, and montane mesic forest ecosystems. Of all the areas studied, they found that the fires had the greatest impact on the coastal grasslands and the `ohi`a woodlands. In 1993, when they began experimenting with ways in which the park’s natural resources could be strengthened to stand up better to fire’s effects, they focused on these two ecosystems.
The operative word for this enterprise is not restoration, but rehabilitation. Restoration, involving replanting of the full suite of native species found in the dry woodland area before the alien grasses invaded, was simply not an option, Loh explains.
“You couldn’t begin to treat all the grass,” she says. Even if you removed it from the area you were trying to rehabilitate, “it’d come in from the outside. So that’s not really feasible. So if you accept the premise that you can’t get rid of the grasses, then you accept recurring fire in the future, and really can’t go back to the original mix of species. So we’re trying to focus on restoring elements of the original that can persist in this new alien grass/fire regime.”
In the long term, Loh says, the hope is that the more fire-tolerant woody native species will create a microclimate that will allow the less fire-tolerant plants to return.
“There is a study by Freifelder, Vitousek, and D’Antonio where they measured microclimate in a woodland with grasses versus open savannah,” she says. “In the woodland there was more shade and less wind penetration through the trees and shrubs. Conditions for fire were ameliorated. In the long term, building native woody communities decreases fire potential and that will allow species that are less fire-resistant to come in, either naturally, if seed is still there, or by planting.”
In 1993, “we started doing small research burns to identify species that are fire-tolerant – plants that, one way or another, are able to survive after the fire by resprouting or are able to establish seedlings,” Loh says. The search for fire-tolerant natives led to mamane and `a`ali`i. Other native plants tested for their fire-tolerance and identified as suitable species for the rehabilitation of dry woodlands include the Hawaiian poppy (pua kala), ko`oko`olau, and sandalwood. Altogether, researchers discovered more than a dozen species that could withstand fire and were adapted to grow in the dry woodland ecosystem. “All of these plants occur in the dry woodland ecosystem, but, with the exception of `a`ali`i, are relatively uncommon – possibly because of a history of browsing by feral goats,” Loh says. “Goats were removed from the park by the early 1980s. Now these species can be returned to the area.”
When fire broke out in late June of 2000 in the grass-covered woodland that lies between the Bird Park picnic area and the highway bisecting the park, park researchers were handed a golden opportunity to test their theories.
Almost as soon as the fire was brought under control, Loh and her colleagues began planning the start of their first large-scale experiment in rehabilitation. The burn scorched 923 acres of dry `ohi`a woodland and 85 ares of mesic koa forest. The researchers laid out 52 transects across the burned `ohi`a-dominated woodland and set up almost 700 circular plots, 50 feet in diameter, along the transects. Volunteers were enlisted to sow between 2,000 and 3,000 `a`ali`i seeds and 400-600 mamane seeds in each plot. In addition, some plots received seeds for koa and ko`oko`olau.
Four hundred and fifty plots received seedlings representing 15 native species. Most were selected for their fire-tolerant qualities, but a few `ohi`a and pukiawe were planted as well, simply because they were so numerous in the pre-burn community.
In the koa forest, which separated the `ohi`a woodland from Kipuka Puaulu – the most biologically diverse native community in the park and one of the park’s designated Special Ecological Areas – the researchers established 75 plots along a strip that was roughly 1300 feet by 170 feet. Into these they sowed `a`ali`i, mamane, and manele and planted seedlings of 18 native plants. In addition, they sprayed the meadowrice grass with herbicide to suppress it while the native seedlings became established.
Finally, to increase the fire resistance of a stand of young koa that separated the dry woodland community from Volcano Golf Course subdivision, another 1300×170 foot strip was laid out, planted with seedlings of koa and 10 other natives, and seeded with koa, mamane, and `a`ali`i. Throughout the area, the work crews eradicated more than 7,400 invasive plants of four species – strawberry guava, faya tree, Prickly Florida blackberry and Yellow Himalayan raspberry.
The work, involving hundreds of volunteers and dozens of Park Service employees, took three and a half years.
Finally, in December 2003, planting ended and the wait began.
Watching and Waiting
Since the 2000 Broomsedge Burn, fire has not struck the area again. But Loh and others have had other opportunities to test their theories on rehabilitation of grass-invaded ecosystems. In February 2003, 2,000 acres, including a rare lama forest, burned in a fire ignited by a lava flow. Three months later, lava from the so-called Mother’s Day flow ignited three separate fires. Nearly 4,500 acres were burned, including areas of rain forest that had not seen fire before.
In December 2004, the Kipuka Pepeiau fire consumed 656 acres of seasonally dry woodland community – an ecosystem similar to the area hit by the Broomsedge Burn but even drier, says Loh, with much lower expected rates of survival of `ohi`a. Rehabilitation has begun there, she says, with seeding of 2 million seeds in plots scattered throughout the burn. The greatest effort is given to planting 8,000 fire-tolerant plants in the more severely burned areas, where native species were completely taken out by the fire.
In addition to the outplanting at Pepeiau, Loh and her crews are taking out selected alien species. “We are targeting aggressive woody species, such as silk oak and faya tree,” Loh says. “There’s not a lot of them there, but because of the opening created by the fire, we expect to see recruitment of these woody species. So we are going in once a year for the next three years to remove them as they try to establish themselves.”
Work to rehabilitate coastal grasslands, especially those dominated by native pili grass, “is a different story,” Loh says. “With the pili grassland communities, there was a history of Hawaiian management of pili by fire. That practice ended in the 1800s. The park actually contains some of the best patches of pili remaining in the state, but these areas are being invaded by alien shrubs and grasses.
“We’re conducting research burns to examine the use of fire as a tool to manage pili grass lands, in order to determine under what conditions we might use fire to maintain pili. There’s a really big risk, because there are all these fire-adapted alien grasses growing with pili. Some may be more successful than pili at re-establishing after a fire. So far, what we have seen is that fire helps keep down alien lantana shrub, but other grasses, such as thatching grass, are very responsive to fire. We need to do a lot more of these experiments in a lot more areas of the park.”
One argument Loh hears frequently from park visitors is that the fires, especially those ignited by lava flows, are natural and should be allowed to run their course without human interference.
“A lot of people from the mainland will say it’s natural ignition, let it burn. Well, the ignition is natural, but the fires are carried by alien grasses. If the grasses weren’t there, the fires would not have been carried very far.
“The recovery of natives after fire is jeopardized by the presence of alien species that compete for resources, so we do have to actively manage fire, and we do this in two ways: with a policy of fire suppression, and then active reintroduction of native species.”
Elsewhere in the islands, resource managers have at times used ungulates to control fire fuel, especially in former ranch areas. Loh disagrees with this approach.
“Using ungulates is a tricky thing. If your only goal is to stop fire and you don’t care about restoring the native communities, the diversity of the plants and all the things that rely on them – if, in other words, your primary goal is just to avoid fires, then I suppose having livestock to keep fuel down is a good solution.
“But if you’re going to restore these systems to a semblance of the native communities they once were, I don’t see how you can do that with livestock present. Native ecosystems did not evolve with livestock and they don’t have the ability to persist with livestock. We’ve learned that. That’s why we’ve put up so many fences in the park.”
The park still isn’t entirely free of grazing animals, Loh says. “From sea level to just below the 10,000-foot elevation, we’ve removed mouflon, goats, and cattle. But pigs are more difficult to take out. We can’t control them everywhere, so we have to prioritize management areas. We focus most of our control in the rain forest, mesic forest, and upper-elevation communities, areas that are relatively intact and contain some of the richest biodiversity in the park.”
At the end of the day, Loh acknowledges that the current management approaches to rehabilitation may result in a woodland that is not a perfect match to the communities that existed prior to the arrival of invasive species. Loh calls these slightly-shifted native communities “tangential.” The mix of native plants may not be identical to what it was before the arrival of browsing animals and aggressive, fire-adapted grasses.
“Imagine a continuum,” she says. “At one end you have a pristine, all-native community. At the other, you have a damaged community dominated by invasive species. As you work back from the damaged end to the natural one, you may have to go slightly off course. You may take an arc, rather than a straight line.”
— Patricia Tummons
Volume 16, Number 9 March 2006