The destruction of a thriving forest may not seem like a fitting pursuit for restorationists, but that was our obsession when we began work on San Onofre Beach in 1994. The forest was a single-species stand of exotic Black Mustard (Brassica nigra). Normally only waist-high, this "forest" was two meters in height, with occasional " forest veterans" nearly three meters tall. This widespread southern California weed had a nearly unbreakable hold on the open ground between remnant patches of coastal sage scrub on the coastal bluff. The high fertility, both natural and from agricultural chemicals, brought dense stands of mustard exploding out of the ground every year. It was apparently impossible for the few natives to spread into the openings, where mustard dried the soil too quickly to allow their seeds to germinate. Our objective was to beat back the mustard and establish a functional coastal sage scrub ecosystem on 33 acres of this rare coastal bluff vegetation. This project was carried out under the California Department of Parks and Recreation's Natural Heritage Stewardship Program.
The three miles of bluff-top terrace sits between I-5 and the Pacific Ocean in the northwest corner of San Diego County. The patches of coastal sage scrub that occupied the site had volunteered after the land passed from military use to the California State Parks system. The terrace is crossed by compacted roads and includes failed attempts at restoration dating back to the 1980's.
Most coastal sage scrub restoration in California is done with hydroseeding, numerous container plants, and irrigation systems at a cost that can reach $40,000 to $60,000 per acre. Even so, many of these expensive projects result in open stands of the two or three "weediest" natives, with annual grasses, mustard, and other exotics filling in between them. Our budget was less than 5% of that amount, we were short on personnel, and had deadlines imposed by California's short winter and spring planting season. We resolved to build a functional ecosystem that would start small but steadily improve on its own, depending on plant and soil ecology rather than traditional landscaping techniques. The key methods would be inoculation with mycorrhizal fungi, missing from the mustard forests, and placement of seed with a sophisticated planting method called land imprinting. Mycorrhizal inoculation can establish a network of beneficial fungi that helps establish natives, build soil structure, promote a healthy soil microflora, and suppress growth of exotic weeds. Land imprinting shapes the soil and presses the seed into contact with the soil, allowing the native seed to germinate nearly as quickly as the weeds. With mycorrhizal inoculum in the soil, the native seedlings could be in the unusual position of competing successfully against the weeds.
Controlling Black Mustard
Black mustard has all the attributes of a successful weed, including rapid growth, copious seed production, and independence from mycorrhizal fungi. Under the mustard canopy was an understory of more mustard, and the supply of seeds was nearly infinite. The fertile soil was of little use to native plants, which are adapted for the low-nutrient conditions of a natural ecosystem. The mustard, on the other hand, was able to take full advantage of the fertility. Any hapless native that might germinate under the mustard was hopelessly outmatched in growth rate and competition for water and sunlight.
We decided that the project would be, to some extent, an experiment in which we would test a variety of techniques to manage the mustard. We established replicated plots throughout the length of the terrace to test a variety of chemical and mechanical methods of control. We used repeated herbicide applications in an attempt to exhaust the seed bank, but made no detectable progress. We flail-mowed the weeds, followed by either raking into windrows or leaving the stems in place. We followed this with herbicide from sprayers and from a tractor-mounted wick applicator.
Raking the stems produced some of our finest stands of new mustard. The most promising variation was cutting and leaving the dead material in place and imprinting over these dead stems, which had a suppressive effect on the weeds. Organic mulch is often credited with suppressing weed growth by shading, but we are convinced that the main effect is removal of soluble nitrogen (nitrate and ammonium) from the soil solution through microbial immobilization. That is, the carbon compounds in the dead mustard biomass enable soil microbes to grow and take up soluble nitrogen into their own biomass. The nitrogen is later re-released when the microbes exhaust the supply of carbon and become dormant. This well-documented process deprives the weeds of soluble nitrogen, and reduces sufficiently their rapid growth rate. Many species of weeds (more properly called "ruderals") are so dependent upon this predictable supply of soluble nitrogen that they do not even germinate unless nitrate is available in the soil solution.
As predicted by this view, many native species were able to germinate under the mulch of dead mustard stems. Among the species making a strong showing was California brome grass (Bromus carinatus), a native perennial that proved to be the key to controlling the mustard, propagating the mycorrhizal network, and establishing a functional ecosystem.
Land imprinting was a key part of the process. Imprinting is done with a heavy roller that first places inoculum below ground, then presses seed into a pattern of small watershed divots, each about one per square foot in area. This traps water and allows it to infiltrate even if a crust has formed on the soil. The imprints also promote air exchange between the soil and the atmosphere. It is important that the method presses the pattern into the soil. Other methods of forming a pattern depend on roughening the surface or just scooping soil out of scattered pits. While these are clearly an improvement over no surface treatment, they usually result in the soil melting into a smooth crusted pile in the first serious rain.
Establishing the mycorrhizal network
Above ground portions of plants photosynthesize, reproduce, and feed herbivores. The rest of the action in a functional ecosystem takes place in the soil. A key component of native ecosystems is the network of mycorrhizal fungi, linking the roots of plants, even those of different species. The network absorbs nutrients in a sort of communal uptake system, it holds the soil into aggregates, and favors important groups of soil bacteria and free-living fungi. It inoculates new seedlings early in life, allowing them to establish a place in the plant community. The network thus dominates both microbial and plant ecology. The network makes the difference between an ecosystem and a weed patch, and defines the difference between soil and dirt. The network is the defining component of ecosystem function, and establishing the network is the most important goal of land restoration.
Once established, the roots and mycorrhizal network take over the original role of the organic mulch, quickly absorbing soluble nutrients and passing them to the native plants where they are safely out of reach of the ruderals.
Some of the most dramatic evidence for the central role of mycorrhizal inoculation was provided by the uninoculated control plots. Early evaluation of three of those plots showed, on average, five times the survival rate of native seedlings on the inoculated plots. Species diversity was double, suggesting that half the species in the seed mix had no survival at all without inoculum in the soil. Four years later, two of the control plots are still marked, and there is a sharp line at the plot boundaries with weeds on the uninoculated side and native vegetation on the inoculated side. Any native seedlings apparently died before the fungi could migrate even a few feet into the control plots.
Successful establishment of the network requires a carefully planned succession of events. Mycorrhizal inoculum must go into the ground at the right depth, followed quickly by germination of the right set of mycorrhizal host plants. Fungi without host plants soon die out, and host plants without mycorrhizal fungi languish while exotic weeds overtake the site. The fungi must be waiting below ground for the roots of the new seedlings. Viable propagules must be in the root zone and must be close enough together that every root has a good chance of finding them quickly. Once the fungus is drawing sugars from the host plant, its external phase can rapidly spread through the soil. It proliferates especially near decomposing roots or other microsites that are rich in nutrients. Centers of newly formed symbiosis spread outward and overlap. These short segments soon form a massive network of several plant and fungal species, and dominate nutrient movement, soil aggregation, and microbial ecology throughout the below-ground ecosystem.
An important priority is choice of the right host plants. While most coastal sage scrub natives are able to become mycorrhizal, some contribute more energy to the mycorrhizal network than others. These "net builders" should have highest priority in the seed mix. Net builders are found in almost every flora, mostly among short-lived perennial grasses and composites. To distinguish the net builders from other members of these families, look for the second wave of invaders on disturbed sites. The net builders need some inoculum already on site, and so rarely invade freshly disturbed sites. In coastal southern California the net builders include bush sunflower (Encelia california), coast goldenbush (Isocoma menziesii), golden yarrow (Eriophyllum confertiflorum) and of course, California brome.
At San Onofre there was a range in establishing the network. The best results, the most rapid exclusion of exotics, and the highest density of germinating seeds and surviving seedlings occurred next to existing patches of native vegetation. We believe the established plants of the native coastal sage scrub are able to support the fungal network while the seedlings get started. Further away from the existing patches of natives, the seedlings would have to feed a growing network while also producing their first leaves and trying to out-pace the weeds. Away from the established native vegetation, the network apparently developed slowly, gaining strength the second year. It appears that the first year seedlings cultivate the mycorrhizal fungi, which later proliferate as the shrubs grow. On these plots, the spreading network remarkably diminished the vigor and size of the weeds which once completely dominated the bluffs.
Prescription for success
We have developed a prescription of sorts for building a mycorrhizae/ native plant network and suppressing exotic weeds. First, choose several plants that are vigorous growers and vigorous net builders. We used California brome, bush sunflower, golden yarrow, and coast goldenbush. Shrubs that are always used in restoration, the weediest of our natives, have been inadvertently chosen for their ability to survive without mycorrhizae, not for their ability to promote it. We included such species as California buckwheat (Eriogonum fasciculatium), monkey flower (Mimulus aurantiacus), and bladder pod (Isomeris arborea), but not as prescribed net builders. They turned out to be the best performers on compacted roads and trails.
We also established islands of native bunchgrass, starting with small forestry-type containers of pre-inoculated mycorrhizal plants. These closely-spaced bunchgrasses quickly began to exclude mustard seedlings within the island, even though mustard grew freely all around them.
Now four years into the project, we see a transformed site. Some of the former roads have disappeared, indistinguishable from the patches of original coastal sage scrub. Most of the former mustard forests have been reduced or completely replaced by coastal sage scrub, or by thriving stands of California brome, with shrubs coming up between the grasses. The patches of mustard that persist are on the sites where we made our mistakes: i.e., we sacrificed natives to kill the weeds, we imprinted when soil conditions were not right, or we used one of our less successful mustard control methods. Even those sites are shrinking year by year. Perhaps mycorrhizal coastal sage scrub will make black mustard a locally endangered species.