The issue is important, she went on to say, “since managers have so little money with which to manage vast landscapes dominated by invasives. They have to prioritize [and understand] which landscapes might undergo succession naturally.”
D’Antonio said this was brought home to her clearly five years ago, when she spent time in New Zealand. “I was shown sites where gorse had dominated after the forest was cleared,” she said. “Managers were claiming that without further disturbance, the gorse had been replaced by native species.”
As then-and-now photos of once-gorse-infested New Zealand hills went up on the screen behind her, gasps of disbelief could be heard in the room from those resource managers who have watched helplessly the relentless march across the slopes of Mauna Kea of gorse, one of the most intractable invaders in the islands.
D’Antonio had their full attention.
The challenge for resource managers, she said, was to understand the persistence of invaders in a given ecosystem. On the one hand, there’s “classic succession,” in which the “invaders are part of a series of changes over time and may also facilitate other invaders,” setting off an “invasional meltdown,” D’Antonio said.
On the other, invaders can facilitate “alternative stable states,” in which the system is fixed in a persistent alternative state. This approach “describes communities that were changed but are now relatively stable,” she said. After prominent ecologists began to use this concept about 10 years ago, she added, “it’s come to dominate our thinking about how native species persist on the landscape.”
How are resource managers to know which system is the appropriate one for a given landscape?
“Watch the systems over a long time period,” she said. If possible, “experimentally tweak systems and see what happens. And maybe you can predict based on an information blitz – such as the potential species pool, species traits, their interactions, and expert knowledge.”
She drew on her work at submontane woodlands in the Volcanoes National Park as a case study. “We were looking at invasive perennial grases in seasonal dry forests that were dominated by `ohi`a,” she said. These areas were invaded by drought-tolerant grasses, especially tufted beardgrass (Schizachyrium condensatum). Between 1988 and 1998, this represented 79 percent of groundcover in the understory of the study area, she said.
“The grass was a transformer,” she said, causing changes in the ecosystem through promotion of fire, in this case. “There was a four-fold increase in the frequency of fires, and a large increase in their size as well.”
D’Antonio and her co-workers set up 20×20-meter-square plots in which the grasses were removed to see what the system would be like without the grasses. They found that “grasses reduced the growth of native woody species by 40 percent,” she said. “They are suppressing the understory community, and also suppressed recruitment.”
However, “they also reduced the recruitment of two even more aggressive invaders: molasses grass and faya,” Melinis minutiflora andMorella faya, respectively. “These are much more potent invaders, in terms of their impact on natives. And they were recruiting heavily in the removal plots.”
In the early 1990s, the beardgrass seemed to be in a steady-state relationship with the ecosystem, she said. “We believed that the system had had a little tweak, due to goat grazing, that allowed Schizachyrium to become dominant, and it was able to maintain itself. We predicted that this was an unchanging alternative stable state. And in 1998, it looked identical to the way it had looked in 1988.”
“It was a good thing, in a way. The system seemed resistant” to further invasion.
D’Antonio predicted that in the absence of fire, the beardgrass would resist invasion by molasses grass and faya. There would be a slow accumulation of dead natives, and also an accumulation of standing grass fuel – “a slow creep towards the precipice of fire.”
“Sooner or later, there would be an ignition, and fire is definitely a precipice” that can convert woodlands and shrublands to a savannah of molasses grass, she said.
‘Over the Precipice’
But the fire didn’t come. In 2012, D’Antonio found, the beardgrass had declined by more than 75 percent, with 62 percent of all the beardgrass plants dead. Live biomass, she said, had decreased by 80 percent.
In the meantime, the two invaders that beardgrass had fended off – molasses grass and faya – had increased, while `ohi`a had declined by 50 percent and pukiawe, the dominant shrub in the understory, had declined by 25 percent.
“What’s driving the change in the system? There’s been no fire in the entire study period,” D’Antonio said.
The answer: “a change in precipitation.”
Annual rainfall has trended down since the late 1980s, she pointed out. “And probably what’s more important is the dry-season rainfall, from May to September.” By comparing yearly dry-season rainfall against a 30-year average, D’Antonio found that before 1998, wet summers were more common. “Since then, we’re seeing much drier summers. Plants in the ecosystems are having to undergo much longer drought stress in summer than they experienced before.”
“We noticed a couple of summers back that there was just no rain,” she said. “Then we saw that that has been typical for the last 10 to 15 years.”
Along with the decline in rainfall has been a rise in temperatures. “We’re talking about a 5 degree Fahrenheit average increase in temperature over the same time period,” she said. “So not only is it drier, but it is hotter as well. Both spring and summer are warmer, and plants are facing more stress than before.”
The changes, she noted, are consistent with data generated by Tom Giambelluca, Hawai`i’s climate-change guru. “Over time, at high elevation stations, above 800 meters … we see increases in temperatures,” she said, adding that her study site was at an elevation of 900 meters. “No question, these habitats are getting warmer, and also drier.”
“Native species are also showing signs of stress… The canopy has decreased in live cover quite considerably,” she said.
D’Antonio looked for other possible reasons for the decline of the invasive beardgrass, but found no evidence of changes in the soil or damage from insects or other organisms.
“So,” she said posing the question for the crowd: “looking at the framework for evaluating the persistence of the invader – is it succession or an alternative stable state?”
In the case of the changing landscape at the park, she said, “it’s succession: from an uninvaded to an invaded initial state. Then drought, climate change, and eventually a changed ecosystem.”
“We did not predict the strong decline of Schizachyrium or of the native woody species,” she noted. “When we started in 1990, climate change was not on anybody’s radar screen.”
But once the decline began, “we could predict the trajectory – invasion by other species. We’ve now done experimental work that could have helped us predict that if something knocked out Schizachryium, it would go the wrong way.”
Looking back, she said, “in 1990, we should have recommended controlling the seed supply of future invaders. The park had been working to control faya until the 1980s, when they gave up. It’s too bad. Now we’re at a point where we’re over the precipice.”
Now, she said, “climate change is important to consider in looking at the persistence of invaders. Some decline, others benefit. We just don’t know the relative strengths of individual invaders to withstand climate change.”