Ecologically Based Integrated Weed Management to Restore Plant Diversity

Ecologically Based Integrated Weed Management to Restore Plant Diversity

Summary

Investigations into the use of ecological theory and the principles of succession in combination with on-ranch resources to restore native diversity to spotted knapweed-infested rangeland was completed during 2006. Invasive species management treatments were applied and ecosystem assessments were collected on one ranch in western Montana. In 2006, project coordinator James Jacobs left Montana State University for a position as the Invasive Species Specialist for the Natural Resources and Conservation Service. Lisa Rew became the project coordinator.

Objectives/Performance Targets

Accomplishments/Milestones

Bitteroot Valley Sites:
Two sites (Lando and Danny) in the Bitterroot Valley were established in June and July 2003 to study integration of sheep grazing and biological control insects to reduce the performance of spotted knapweed. Desirable species were then sown into some of the grazed areas to determine if the light soil and vegetation disturbance caused by the sheep provided an increased number of safe sites for seed germination and survival.

Each site was divided into two whole plots for the biological control treatment. Three hundred adult Cyphocleonus achates, a biological control root feeding weevil for spotted knapweed, were released onto each biocontrol plot in late July – August in 2003, 2004, and 2005. Three 4.5 by 4.5 meter grazing exclosures were constructed at random within each of the whole plots (biocontrol and no biocontrol) and used to test the sheep grazing effect on plant community characteristics. Approximately 300 ewes with lambs grazed both sites in July 2003. Native hay was cut in July, and seeds from native forb species were collected throughout the summer and fall for subsequent application onto plots. Seed treatments (a no seeding control, perennial native grass species, native forb species, native grass and forb species mixed, and hay) were applied inside and outside the exclosures on 1 by 5 m plots in the second week of November 2003.

Baseline density and cover data by species were sampled from four randomly placed 0.2 by 0.5 m frames inside and outside each exclosure, at the time the exclosures were built in 2003 and before treatments were applied. Sheep utilization of spotted knapweed was sampled by clipping all plants within one 0.2 by 0.5 m frame placed inside and outside each exclosure after the grazing treatment was applied.

In 2004, the grazing treatments on the two sites in the Bitterroot Valley were modified. Site 1 could not be grazed because the cooperator was denied access to water. Site 2 was grazed by 12 rams from July through October.

Density and cover by species were sampled from one randomly placed 0.2 by 0.5 m frame from each of the five seeding treatments inside and outside each exclosure at both sites. The number of spotted knapweed flower heads in each of the sample frames was also counted. Sweep net sampling was used inside and outside each exclosure to estimate weevil establishment. Spotted knapweed plants were uprooted and the roots were searched for weevil larvae. Five soil samples were collected and pooled from inside and outside each of the exclosures, and assessed for nitrogen content and nitrogen mineralization. The grazing treatment was only applied and quantified at Site 2: sheep utilization of spotted knapweed and grass was sampled by clipping all plants within one 0.2 by 0.5 m frame placed inside and outside each exclosure.

An additional seeding treatment was applied at both sites in November 2004. The seeding treatments were: no seeding, broadcast seeding of bluebunch wheatgrass, annual sunflower, and blue flax at a rate of 30, 10, and 10 lbs/acre, respectively, and, disc tillage of the soil to a depth of 5 cm with broadcast seeding of bluebunch wheatgrass, annual sunflower, and blue flax at a rate of 30, 10, and 10 lbs/acre, respectively. The treatments were replicated three times with plot sizes of ½ acre.

In 2005, management treatments were continued in the Bitterroot Valley project; again, Site 1 (Lando) could not be grazed because the cooperator was denied access to water. Site 2 (Danny) was grazed by 12 rams from July through October. Sweep net sampling was used inside and outside each exclosure to estimate weevil establishment. Spotted knapweed plants were uprooted and the roots were searched for weevil larvae. The seeding treatments applied in 2004 were sampled. The cover and density of all species, and the cover of bare ground and litter were measured in randomly placed 0.2 by 0.5 m frames within each treatment plot.

In 2006, sampling of the Bitterroot Valley project was concluded with assessments of cover and density of all species, bare ground and litter in random 0.2 by 0.5 meter frames within each treatment and sub-treatment plot.

Impacts and Contributions/Outcomes

Impact and Contributions/Outcomes:
Summary:
This research and demonstration aimed to show how management actions affect species abundance and our ability to change a plant community from being dominated by an invasive forb species to being dominated by more desirable species. Understanding disturbance (in terms of grazing and tilling), species abundance, availability and performance helps landowners and managers design and predict the outcome in integrated weed management. This project also shows how resources available on ranches can be used, following ecological principles, to manage invasive species and improve productivity.

Results from the Bitterroot Valley sites have important implications for management of spotted knapweed. Grazing had no affect at the Lando site in 2006, which is not surprising as there had been no sheep grazing at that site since 2003 (Table 1b). Grazing did significantly reduce spotted knapweed density at the Danny site (Table 1c). The biocontrol treatment had no effect on the Lando site in 2006, but significantly increased spotted knapweed numbers at Danny site in 2006 (Table 2b). Spotted knapweed cover was not significantly affected by the biocontrol agent at either site but cover was higher in the grazing exclosures at both sites (data not shown). Establishment of seeded species was generally poor, with the exception of bluebunch wheatgrass in the second seeding experiment (Table 5). Spotted knapweed density or cover was not reduced by the first seed treatment experiment, although hay cutting did significantly reduce the density of spotted knapweed at the Lando site in 2006 (Table 1b). In the second seeding experiment, the seeding and tillage treatment did reduce spotted knapweed density in the first sampling year (2005) after disturbance but not statistically significantly in the subsequent season (2006) (Table 7).

Results in more detail:
It was anticipated that sheep grazing would sufficiently disturb the soil and vegetation to increase site availability for establishment of sown desirable species as part of the five seeding treatments imposed in 2003. The only sub-treatment to show a statistical difference was the hay plots for the 2003 experiment; in both 2004 and 2006 these plots had lower spotted knapweed densities, and litter was higher (Table 1a, b and c (non-grazed)). At the Danny sites, where grazing was imposed, the amount of bare ground was significantly higher in the grazed plots, where there was no biocontrol but there was no difference in the amount of bare ground in the seeding treatments (Table 1c). This may suggest that while grazing did increase the amount of bare ground, which could be linked to safe sites, this did not assist with seedlings establishment, possibly due to seeding time or seasonal rainfall.

The availability of spotted knapweed as measured by the number of flowering plants depended on site, year, grazing treatment, and biological control treatment. At the Lando site that had only one year of grazing, density increased from 2003 to 2004 regardless of the grazing treatment where no biological control weevils were released (Table 2a). Where the insects were released, density increased from 2003 to 2004 only where sheep grazed (Table 2a). However, by 2006, there was no significant statistical difference in the density of flowering spotted knapweed in all plots regardless of grazing or biological control treatment (Table 2b), but seeding treatment was important (Table 1b).

At the Danny site, flowering spotted knapweed density was significantly affected by grazing and biocontrol but not seed treatment in 2006. Mean density in the exclosures was 19/m2 compared with 6/m2 where sheep were grazing. However, the number of flowering spotted knapweed was significantly higher in the biocontrol plots (18/m2 compared with 8/m2 in the no biocontrol plots) suggesting a lack of success of the agent. Seed treatment was not significantly different with the exception of hay treatment (Table 1c). The predominant grass species at the Danny site were non-native pasture grasses whereas the predominant grasses at the Lando site were annual non-native grasses (data not shown). Surrounding plant community composition will play a large role in succession of spotted knapweed-infested communities under grazing, biological control insect, and revegetation management.

The rosette demographic provides an indication of the future population of the weed. Biocontrol did significantly reduce the number of spotted knapweed rosettes at the Lando site, but, strangely, increased the number at the Danny site (Table 3a). Seeding treatment affected the density of spotted knapweed rosettes at Lando but not the Danny Bitterroot Valley sites (Table 3b). Spotted knapweed rosettes densities increased dramatically at both the Lando and Danny sites from 2004 to 2006 in all seeding treatment except for the hay seeding treatment where the density of rosettes was similar from 2004 to 2006 (Table 3b). Hay mulch increases the percent litter (Table 1b and 1c), which most likely reduces the availability of safe sites for knapweed propagule establishment.

There was no significant effect of grazing on soil nitrogen dynamics, however, the trends are interesting and justify further exploration. Total available nitrogen (NO3 + NH4) tended to be lower, and potentially mineralizable nitrogen (PMN) tended to be higher, in the grazed plots compared to the plots excluded from grazing at the Danny site after two seasons of grazing (Table 4). These trends suggest that available nitrogen is being used more efficiently and that microbial production of available nitrogen is greater where there is grazing compared to where grazing is excluded. This indicates a more active nitrogen cycle where sheep graze compared to where there is no grazing. The nitrogen cycle is a critical ecosystem function that affects productivity. There are many unanswered questions concerning invasive species, management, and the nitrogen cycle that may provide insight into invasions and their management.

An additional disturbance and seeding study was initiated in 2004 because the disturbance of sheep and their grazing did not affect establishment of native seeded species. Thus, three seeding treatments were applied to the grazed plus biocontrol plots at the Lando and Danny sites. The treatments were no seeding, seeding, and tilling plus seeding. The species in the seeding mixture were bluebunch wheatgrass at 30 lbs per acre, annual sunflower at 10 lbs per acre, and blue flax at 10 lbs per acre. Of all species established, bluebunch density increased from 2005 to 2006 at both sites, flax densities remained low in both years, and, unsurprisingly, annual sunflower was only observed in the first year at very low levels (Table 5).

The density and cover of spotted knapweed plants and cheatgrass were slightly higher where biological control weevils were released compared to where they were not released for both sites in 2005 (Table 6). This could suggest that in the short term, the disturbance associated with biological control, in this case disturbance specific to knapweed flowering plants, increased the density of knapweed and also non-native cheatgrass. This is not an unusual result. Spurge and knapweed densities have increased in the first year after sheep grazing and decreased in following years. Dr. Jacobs expects bluebunch wheatgrass to increase in the long term under this type of disturbance. The comparison between the bluebunch densities in control and biocontrol plots could not be made for this experiment in 2006 as control plots at both sites had been compromised (a barn had been built on one site, and other plots could not be located). However, the overall bluebunch wheatgrass density (39/m2) on biocontrol plots at both sites increased relative to 2005 (14/m2).

Seeding treatment had a significant effect on spotted knapweed plant density at both sites in 2005, but there was no significant difference in 2006 (Table 7). Bluebunch wheatgrass densities were significantly different at both sites in 2005 and Lando in 2006 with highest densities in the till and seed treatment (Table 7). This result indicates that high seeding rates alone will not improve establishment of desired native species, and a pretty heavy disturbance such as shallow tilling is needed to create safe sites for establishment. Cheatgrass density did not differ statistically significantly between the different treatments (data not shown). It is anticipated that in the long term, the establishment of a long-lived perennial bunchgrass combined with the reduction in performance of spotted knapweed by sheep grazing and aggressive seeding will change the plant community to a native bunch grass community. There are data available for 2005 and 2006 concerning native and non-native grass and forb species. These will be analyzed for the final report in an attempt to better answer the preceding statement.

Collaborators: