Burning and herbicide control of invasive fennel Foeniculum vulgare, and effects on native species recovery on Santa Cruz Island, California, USA

  • Published source details Erskine Ogden J.A. & Rejmánek M. (2005) Recovery of native plant communities after the control of a dominant invasive plant species, Foeniculum vulgare: Implications for management. Biological Conservation, 125, 427-439


More than 100 years of sheep and cattle grazing, with associated trampling and soil erosion, has detrimentally affected native plant communities on Santa Cruz Island (California, USA). In an attempt to redress the significant damage, over 36,000 feral sheep Ovis aries and 200–300 cattle Bos taurus were removed, but another introduced species, pig Sus scrofa, was not eradicated. After sheep and cattle removal, many remaining, partially intact coastal sage and perennial grassland communities, as well as habitats highly modified by cultivation, became dominated by fennel Foeniculum vulgare which out-competed many native plants. Fennel is a perennial herb introduced to USA from southern Europe. Although common on Santa Cruz since the late 1800s, it did not become a dominant invasive until after sheep and cattle removal. The spread of fennel prompted the initiation of several small scale fennel control trials. This led to a study conducted in two fennel-infested areas to evaluate the efficacy of a burn and herbicide treatment for fennel control, and subsequent responses of native species.

Study site: Santa Cruz Island, is located 30 km off the coast of southern California, and is the largest of the California 'Channel Islands' (249 km²). It has a wet season from November to April (average total rainfall 440 mm, mean daily temperature 13.1 °C) and a dry season from May to October (<50 mm rainfall, mean daily temperature 18.5 °C). It has eight endemic plants, with a further 37 endemic to the California Channel Islands as a whole.

Currently, many plant communities are dominated by fennel or annual grasses (e.g. slender oat Avena barbata, ripgut brome Bromus diandrus, wall barley Hordeum murinum) and non-native forbs (e.g. black mustard Brassica nigra, yellow starthistle Centaurea solstitialis and stork's-bill Erodium cicutarium). Patches of native perennial bunch grasses e.g. California brome Bromus carinatus, blue wildrye Elymus glaucus and purple tussock grass Nassella pulchra, and native annual forbs e.g. ranchers fiddleneck Amsinckia menziesii, tarweed Hemizonia fasciculate and lupin Lupinus bicolor occur amongst the non-native species.

Experimental design: Twenty-five, 30 m × 30 m plots were established in fennel or grass dominated areas. Ten were randomly placed in grasslands that contained no fennel and 15 in fennel-infested areas. Areas infested with fennel were previously grassland pastures, so the grassland plots were considered pre-fennel invasion communities. Five grassland and five fennel blocks were designated controls (untreated). The average native species and fennel cover in the plots prior to treatment is shown in Table 1 (attached).

Treatments: The pilot studies found the greatest decrease in fennel cover and a significant increase in native species richness and diversity with a stem removal treatment or prescribed burn, and a winter herbicide application. Therefore, in November 1997, 15 blocks were burned to remove dead standing fennel stems and plant litter to improve herbicide contact with live fennel the following spring. Burning alone, did not kill fennel plants. In May 1998 and May 1999, these blocks were aerially sprayed with a 2% solution of amine-based triclopyr (Garlon 3A), plus 0.025% v/v Pro-Spreader surfactant; (Triclopyr is a selective herbicide that kills/negatively affects growing broad-leaf dicotyledonous plants only). Herbicide application in late spring was undertaken as it was considered to have the least impact on native herbaceous dicots which set seed and begin to senesce by early May, whilst fennel is near its peak of growth.

Plant cover and other habitat variables: Three, 16 m² plots were randomly located within each block within which visual estimates of percent cover of each species, bare ground, litter and pig rooting were made. For most analyses, species were pooled into one of three categories:

i) non-native grasses
ii) non-native forbs (excluding fennel)
iii) fennel and native species

Data were collected between March and May from 1997 to 2001.

Soil analyses: Five soil samples were collected at 0–10 cm and 10–20 cm depth in each block. Soil texture (% sand, % silt, % clay), pH, total nitrogen, soluble potassium, total phosphorus, and percent organic matter were measured for each block to test for heterogeneity of soil characteristics. Significant differences were not found among treatments thus were not incorporated into subsequent analyses.

Effects of treatment on plant community structure: In 1997, before treatment, fennel blocks were dominated by fennel with little or no grass, and grassland blocks were dominated by non-native grasses with little or no fennel (see Table 1). Fennel cover decreased significantly in the treated fennel plots (60% to <3% average cover) and Mediterranean annual grasses became more frequent, however these were mostly non-natives. This was contrary to the results of the small-scale pilot studies where native species cover, richness and diversity increased significantly with fennel suppression. The explanation for this discrepancy probably includes the fact that the pilot study was performed in a smaller area with more diverse plant communities adjacent to the plots which acted as seed sources.

There was a non-significant trend for increasing non-native grass cover in treated fennel blocks, and for decreasing non-native grass cover in treated grassland blocks over time. Non-native forbs (excluding fennel) had significant year, treatment, and year by treatment interactions (Table 2). There was a significant decrease in non-native forb cover over time in treated grassland blocks.

Influence of weather: Differences in yearly rainfall affected vegetation. The second year of the study (1997/1998 season) was an El Niño year with over twice the yearly average rainfall (1,101 mm), conversly 1998/1999 had less than half the average rainfall (244 mm). This is reflected by higher average plant cover in 1998 and lower cover in 1999.

Changes in untreated blocks over time: There was no increase in native species cover over time, but there was an increase in native species richness in both the untreated fennel and the untreated grassland blocks. It was expected that there would be an increases in native species in treated fennel blocks but not in untreated fennel or untreated grassland blocks. Unexpectedly, fennel cover also decreased significantly in the untreated fennel blocks. An explanation is increased pig rooting as this was the variable that three untreated fennel blocks were most strongly associated. Further analysis implied that species composition, fennel cover and pig rooting were important factors in making these blocks clearly different from the others.

Overall success of treatments: The plant community structure in treated fennel areas became less similar to untreated fennel blocks and closer to the untreated grasslands. Although the small-scale pilot studies found increases in native species cover and richness, this did not occur at the landscape level. Much of the Central Valley of Santa Cruz Island was vineyards and pastures from the mid-1880s until 1939, vineyards were then removed and the area heavily ranched. Native California grassland species are not adapted to intense grazing, and diminish or disappear under a heavy grazing regime. The grassland treatment area was assumed to be an abandoned agricultural field/rangeland before fennel invasion, and the fennel plots were reverted to the pre-fennel invasion plant community. Dominant Mediterranean grass species included Avena spp., soft brome Bromus hordeaceus, Italian rye-grass Lolium multiflorum, H.murinum and B.diandrus. All of these species are palatable to cattle and sheep.

Unfortunately, these Mediterranean grasses and the litter they produce have been shown to prevent/reduce the germination and establishment of native coastal plant species that make up the desired plant communities.

Reasons for increases in native species in untreated fennel blocks: Fennel is tall and dead standing stems often persist for many months, thus forming a different community structure compared to short herb-dominated grasslands. This difference may provide an explanation for the native species recruitment found in the untreated fennel blocks and their absence from the treated areas and most of the untreated grasslands. Bird seed dispersal via droppings has been shown to be important in old-field succession. A variety of bird species visit the tall fennel stands as the stems provide perches and this may facilitate native seeds dissemination into these areas. Fennel stems can also trap wind dispersed seeds from native woody species allowing additional recruitment of species such as A.californica and B.pilularis. Seedlings of A.californica, B.pilularis, man-root Marah macrocarpa, sugar bush Rhus ovata, and island redberry Rhamnus pirifolia (the last three fleshy fruited and attractive to birds) were found in significantly higher numbers in the untreated fennel blocks (10 of 15 plots), than the treated fennel and treated grassland blocks (1 of 45 treated plots) or the untreated grassland blocks (3 of 15 plots). All three of the untreated grassland plots with woody species seedlings were found in the same block, which was located adjacent to a coastal sage scrub community which may have provided a seed source.

Conclusions: Plant communities in the fennel infested blocks became more similar in species composition to the grasslands that existed before fennel invasion. In this respect the control was a success. However, the lack of a native seedbank and the accumulation of non-native grass litter probably prevented the recovery of native species in treated areas.

Unexpected changes can arise when non-native species are removed, such as experienced on Santa Cruz Island upon removal of feral sheep and cattle. The spread of fennel and the increase in feral pig numbers are two obvious changes that occurred in response to the removal of the sheep and cattle. The removal happened to coincided with the end of a 7-year drought which undoubtedly contributed to the rapid expansion of both fennel and pig populations.

Note: If using or referring to this published study, please read and quote the original paper.

Output references

What Works in Conservation

What Works in Conservation provides expert assessments of the effectiveness of actions, based on summarised evidence, in synopses. Subjects covered so far include amphibians, birds, terrestrial mammals, forests, peatland and control of freshwater invasive species. More are in progress.

More about What Works in Conservation

Download free PDF or purchase
The Conservation Evidence Journal

The Conservation Evidence Journal

An online, free to publish in, open-access journal publishing results from research and projects that test the effectiveness of conservation actions.

Read the latest volume: Volume 17

Go to the CE Journal

Subscribe to our newsletter

Please add your details if you are interested in receiving updates from the Conservation Evidence team about new papers, synopses and opportunities.

Who uses Conservation Evidence?

Meet some of the evidence champions

Endangered Landscape Programme Red List Champion - Arc Kent Wildlife Trust The Rufford Foundation Save the Frogs - Ghana Bern wood Supporting Conservation Leaders National Biodiversity Network Sustainability Dashboard Frog Life The international journey of Conservation - Oryx British trust for ornithology Cool Farm Alliance UNEP AWFA Butterfly Conservation People trust for endangered species Vincet Wildlife Trust