Final Report: Abiotic Controls on Invasive Species and Biodiversity: Comparison of Forest and Shrubland

EPA Grant Number: R828901
Title: Abiotic Controls on Invasive Species and Biodiversity: Comparison of Forest and Shrubland
Investigators: Meixner, Thomas , Allen, Edith B. , Fenn, Mark
Institution: University of California - Riverside
EPA Project Officer: Hiscock, Michael
Project Period: July 15, 2001 through July 14, 2004 (Extended to June 14, 2005)
Project Amount: $448,122
RFA: Exploratory Research to Anticipate Future Environmental Issues (2000) RFA Text |  Recipients Lists
Research Category: Ecological Indicators/Assessment/Restoration , Water , Ecosystems


Plant invasion is subject to biotic as well as abiotic controls. The objective of this research project was to investigate the abiotic controls on plant invasion in coastal sage scrub (CSS) and conifer forest ecosystems of southern California. Specifically, we focused on the controls on mineral nitrogen availability and the impacts of this availability on the presence, absence, and dominance of invasive plant species. Our starting hypothesis was that CSS ecosystems would be more susceptible to increases in atmospheric nitrogen deposition because of the relatively closed nitrogen cycle in these ecosystems compared to the more open nitrogen cycle of conifer forests in southern California.

Summary/Accomplishments (Outputs/Outcomes):

Our project has three focus areas. First, we measure deposition. Second, we detailed the movement of water and nitrogen in the CSS and forest ecosystems of southern California. This work is meant to determine how closed or open the system is to atmospheric deposition inputs. Additionally, we investigated the influence the invasive grasses and soil nitrogen concentrations have on biogeochemical turnover in the CSS. This biogeochemical work has a direct impact on how this ecosystem is influenced by plant invasion. The third area defined the biotic impact of increases in nitrogen deposition and the impact on biodiversity across the CSS and forest gradient we studied.


The deposition work conducted as part of this project and in coordination with other projects can be divided into two categories. First, we used passive sampling of throughfall to estimate deposition, and second, we used isotopic characterization of deposition, leachate, and runoff from field sites to define the fate of deposition. In the area of passive sampling, unfortunately sampling in CSS turned out to be a technical failure. The resin samplers that have worked very well in open and forested canopy situations proved susceptible to animal damage and most importantly sediment filling from wet splash of sediment. Although this result is discouraging, the failure in this project has led to several incremental steps. The first one we discussed during the course of this project. The use of surrogate surfaces such as small sterile soil samples used to measure deposition over a period of time. This idea was discarded because of wind deposition and erosion of the sample. Subsequently, the idea became to use filter paper with soil glued to the surface. This idea was developed subsequent to this project, but the developments would not have been possible without the failures involved in this project. Because of the lack of passive deposition measurements in the CSS environment, we were not able to assess differences between CSS and coniferous forest as we initially intended. However, through this study in cooperation with a nearby study in chaparral ecosystems, we were able to demonstrate that the very high deposition values observed by Fenn and Poth (2004) appear to be highly local and related to the specific habit of coniferous vegetation and the large exposure of the field site they studied to fog deposition events, which appear to double if not increase deposition by more than 300 percent. By comparison, results in Meixner and Fenn (2004) indicate that deposition, although generally similar in low habit chaparral vegetation, was somewhat higher under Canyon Live Oak (quercus chrysolepis), which generally stands as an island species in the generally low growing chaparral. Combined these results indicate that CSS should have lower deposition rates than coniferous forest because of the general low growth profile that the species has. However, this hypothesis is still waiting to be confirmed by field tests with the new deposition monitoring approach.

The second area of investigation involving deposition and tracing its impact on coniferous and CSS ecosystems involved collaboration with isotopic chemists at the University of California at San Diego. This collaboration resulted in one publication (Michalski, et al., 2004) that demonstrated several important facts about atmospheric deposition of nitrate in xeric ecosystems. First, amounts of atmospheric deposition in the litter and soil were much greater in the soils of the coniferous forest, indicating much greater rates of deposition. Soils in CSS still had some amount of atmospheric deposition in them, however.

Fate and Transport of Deposition

The 3 years of field results from this study record the soil system with depth under extreme drought conditions, mild drought conditions, and under a year with average precipitation. These 3 years provide a strong comparison between our various locations and between the 3 years for comparing the movement of the measured anions, chloride, and nitrate in wetter years.

During Year 1 of the project (October 2001 to September 2002), there was minor translocation of these anions beneath the 25 cm depth of the two CSS sites. Overall, lower concentrations of anions were measured at the wetter, more pristine CSS site of Lake Skinner (clean and relatively uninvaded). This probably reflects both some leaching of these anions, as well as their low initial input. Less leaching to depth at the Botanic Gardens (polluted and invaded) field site was observed. The second year of data shows a similar pattern at the invaded site with little infiltration to depth, whereas the shrub dominated site showed infiltration to depth. The third year of data appears to confirm these conclusions. These results suggest that a shift in the hydrologic response of the degraded Botanic Gardens CSS ecosystem may be occurring as shrubs, and their deeper root conduits, decrease in number. The top 25 cm of the soil may be becoming the dominate zone of water movement, with increasing amounts of exogenous nitrogen being concentrated in a decreasing depth of the soil. This closing off of deeper soil may actually tighten the hydrologic and nutrient cycles in the CSS and ecosystem and further foster plant invasion (Figure 1).

Figure 1. Schematic of the Positive Feedback Between Nitrogen Deposition, Fire, and Grass Replacement of Shrubs and the Hydrologic and Biogeochemical Changes That Ensue

Figure 1. Schematic of the Positive Feedback Between Nitrogen Deposition, Fire, and Grass Replacement of Shrubs and the Hydrologic and Biogeochemical Changes That Ensue

At the forest sites, steep mountainous terrain is expected to control solute transport through soils during precipitation and snowmelt events. At the “dirty” Camp Pavaika site, the movement of water and nitrate through soils on a south facing 30° slope in the San Bernardino Mountains were studied. Here, soils 1.75 m thick overlay a 2-m thick zone of weathered bedrock. Two rock layers approximately 40 cm thick occur at soil depths of 35 and 130 cm. Neutron probe (NP) measurements indicate that snowmelt infiltrates to the buried rock layers and then travels downslope. Soil solution chloride and nitrate concentration depth plots reflect this control. Nitrate was observed to increase downslope during snowmelt along the sharp contact between the soil and weathered bedrock as well as within rock layers. Here roots are concentrated. This work illustrates the need to quantify differences in solute transport because of geomorphic position, especially when studying natural systems with complex topography. In addition, the results from this work suggest that the deleterious effects of anthropogenic nitrogen additions to this ecosystem may be concentrated in the top 1.5 to 2 m. We are in the process of analyzing soil samples for the effects of acidification to see if the effects are relegated to select sections of the soil profile. Such spatial partitioning with soil depth could lead to a rapid flux of nitrate into mountain streams with snowmelt. Additionally, the more shallowly rooted vegetation, such as shrubs and annuals, could be more negatively impacted by over fertilization and acidification than the overstory Ponderosa Pines (Figure 2). Also, during all years of data collection, NP access tubes were established at each site (at least four for each site). Soil coring till point of refusal seasonally was conducted for each site.

Figure 2. Chloride Profiles for Barton Flats and Camp Paivika Field Sites

Figure 2. Chloride Profiles for Barton Flats and Camp Paivika Field Sites

Figure 2. Chloride Profiles for Barton Flats and Camp Paivika Field Sites. Notable bulge in the drought year at ~75 cm depth is coincident with a rock layer that seems to influence water and solute movement in the old surface system of Camp Paivika

Vegetation-Biogeochemistry Feedbacks

Several studies were completed regarding the biogeochemistry of invasive grass communities in the semiarid sites in Riverside County. In particular, the results of a decomposition study indicate that litter nitrogen is the key to decomposition rates. This study examined nitrogen deposition in exotic annual grasslands that occur in areas historically dominated by CSS in southern California. The objective was to study the effects of nitrogen deposition on levels of plant tissue nitrogen and litter decomposition. The study was conducted at sites that occur at each end of an air pollution gradient in southern California. Bromus diandrus litter that was harvested from the high nitrogen deposition site had higher percent total nitrogen than litter from the low deposition site. Decomposition rates of high and low nitrogen litter were measured at the high and low nitrogen deposition sites and in experimentally fertilized plots at the low nitrogen deposition site. Results suggest nitrogen deposition results increased decomposition rates of B. diandrus. Furthermore, in southern California, because nitrogen deposition occurs during the dry season and soil nitrogen concentrations are lower during the rainy than the dry season, litter nitrogen may have a greater effect on decomposition than soil nitrogen levels during the rainy season.

Figure 3. Illustration of Positive Feedback Between Grass Invasion and N Deposition

Figure 3. Illustration of Positive Feedback Between Grass Invasion and N Deposition

Figure 3. Illustration of Positive Feedback Between Grass Invasion and N Deposition. (a) In a shrub dominated landscape, nitrogen mineralization and decomposition processes are slower and lead to a brake on invasion when deposition is low. (b) As grasses dominance increases, mineralization and decomposition processes increase their rate. This increase leads to more nitrogen being available and the grasses building an advantage of the shrubs.


We have investigated the effects of nitrogen deposition on biodiversity in two ways. First, a nitrogen fertilization experiment was initiated in 1994 near the north shore of Lake Skinner in the Western Riverside County Multispecies Preserve. The objective of the research was to determine the effects of nitrogen on CSS vegetation. The nitrogen fertilizer was applied to mimic the effects of anthropogenic nitrogen deposition. Nitrogen deposition is increasing worldwide, especially in urbanized and agricultural regions such as southern California (Padgett, et al., 1999; Allen, et al., 2006). The exotic grasses responded with elevated biomass within the first few years after fertilization began, but the native forbs did not begin to respond with decreased cover until 11 growing seasons had passed (Figure 4). Although fertilization caused a slow and gradual response in vegetation change, this vegetation type is changing rapidly because of fire. The fine fuel threshold for fire is 0.5 T/ha (Fenn, et al., 2003), which occurred for exotic grass biomass during most years with nitrogen fertilization but in fewer years in unfertilized control plots. Thus, elevated fuel loads driven by nitrogen deposition are promoting more frequent fire.

Second, biological surveys have been completed at the CSS and forested sites. These surveys indicate that biomass and presence of invasive species is increased in the semiarid CSS sites but not at the forested sites. At the CSS sites species richness declined from 67 to 16 for native forbes species. Soil nitrogen concentration was inversely related to forb richness (R2 = 0.83, P = 0.005). Exotic grass cover increased from 1 percent to 69 percent with increasing nitrogen (R2 = 0.68, P = 0.02). Conversely, native shrub cover decreased from 35 percent to 1 percent with increasing nitrogen (R2 = 0.70, P = 0.02). Overall these results support our original hypothesis that nitrogen deposition may lead to decreases in biodiversity and that semiarid systems would be more susceptible to these impacts than forested sites.

Take Home Messages

This project accomplished several specific things that are significant scientifically and to managers of natural resources.

What Was Accomplished? Through our studies of atmospheric deposition, its fate, cycling, and effects on biodiversity we have demonstrated three major things. First, hydrologic shifts in CSS caused by species dominance change to annual grasses encourage and sustain grass invasion in part. This statement is not meant to imply that this hydrologic shift is the only contributing factor for the relatively permanent replacement of shrubs by grasses but that this hydrologic shift is one component of enforcing the shift. Second, increased cycling of nitrogen in high nitrogen deposition region grass likely increases nitrogen availability and again reinforces the grass invasion. Third, nitrogen fertilization both in plots and along the gradient was demonstrated to decrease biodiversity, and thus, nitrogen deposition is an exacerbating environmental variable to the grass invasion.

Figure 4. Top Shows Percent Cover of Native Forbs at a Fertilized (+N) and Unfertilized Set of Plots (-N)

Figure 4. Top Shows Percent Cover of Native Forbs at a Fertilized (+N) and Unfertilized Set of Plots (-N). Bottom shows same plots but exotic biomass. Percent of native cover seems most effected in wet years; exotic biomass differences appear to have become larger in more recent years. +N are plots fertilized with 60 kg N/ha annually; - N are unfertilized controls.

Why is This Significant? These results are significant because they demonstrate two things. Plant invasions cause shifts in ecosystem processes related to hydrologic fluxes and biogeochemical cycling. Second, these ecosystem process shifts in the case of grass invasion of CSS seem to have encouraged the sustaining of the invasion.

Who Will Be Able To Use Results/Products? Land managers of semiarid scrub habitats that are dealing with invasions should take note of the importance of hydrological and biogeochemical processes and how they can effect the ability of a plant invasion to be sustained.

The U.S. Environmental Protection Agency and air quality agencies need to know, at least in the case of CSS habitats and possibly other habitats, that degradations in air quality can exacerbate or encourage plant invasions, and thus, biodiversity markers need to be incorporated into the planning and management of air quality and atmospheric deposition standards.

How Will it Further Science/Management of Resources? The work is another contribution in the area of the interaction between biotic and abiotic processes and how they interact to influence ecosystem processes. Managers should benefit from insight into how certain ecosystems are susceptible to resource allocation shifts.

Supplemental Keywords:

invasive species, atmospheric deposition, nitrogen biogeochemistry, hydrologic flushing, leaching rate, ecological effects, nitrogen oxides, terrestrial ecosystem, restoration,, RFA, Scientific Discipline, Geographic Area, Ecosystem Protection/Environmental Exposure & Risk, Ecosystem/Assessment/Indicators, State, Forestry, Monitoring/Modeling, Ecological Effects - Environmental Exposure & Risk, Biology, Exp. Research/future, Futures, emerging environmental problems, extinction risk, atmospheric nitrogen, ecological exposure, biodiversity, endangered species, forest, biopollution, runoff, shrubland, exploratory research, hydrology, forests, invasive species, irrigation, California (CA), rainfall, ecological dynamics, acid deposition, futures research

Progress and Final Reports:

Original Abstract
  • 2002 Progress Report
  • 2003 Progress Report
  • 2004