Sustainable Management of Non-Native Thistles on Colorado’s Western Slope

by John Coyle* and Scott Nissen [Edited by Celestine Duncan; Photos by John Coyle]

Ranching and farming are key economic drivers on Colorado’s western slope. Protecting these natural resources from invasive plants is a priority for ranchers to preserve their rural lifestyle and ranching heritage.

Musk thistle (Carduus nutans) and Canada thistle (Cirsium arvense) are well established on the western slope and are a concern to agricultural producers. Biological control agents have not provided effective musk thistle control in this area. Both weeds are highly competitive, invading disturbed sites, pastures, rangeland and forestland.

Figure 1. Location of the research project in southwestern Colorado.

Figure 1. Location of the research project in southwestern Colorado.

A research project was initiated in 2014 on a 10,670-acre ranch near Cimmarron, Colorado (Figure 1) to develop sustainable management strategies for musk and Canada thistle. Objectives of the study were three-fold: 1) Investigate invasive thistle distribution through geospatial analysis^; 2) determine effects of musk thistle management on forage quality and native plant diversity; and 3) develop a sustainable invasive plant management plan for the ranch.


The ranch is located in the foothills and mountains of the San Juan Mountain Range with elevations ranging from 6,800 to more than 10,200 feet. About 30 miles of dirt roads and two-track trails traverse the ranch, with Highway 50 bisecting the northern portion of the property. Rainfall averages 13.8 inches with some local variation depending on elevation and aspect. Western portions of the ranch and lower foothills are generally hotter and drier than eastern portions and higher elevation ranges (Figure 2).

Figure 2. Rainfall varies on the ranch based on elevation and aspect, ranging from relatively dry sagebrush dominated western foothills (left) to higher elevation mountains (right).

Figure 2. Rainfall varies on the ranch based on elevation and aspect, ranging from relatively dry sagebrush dominated western foothills (left) to higher elevation mountains (right).

Permanent transects were established at 32 locations on the ranch to measure thistle abundance and density. Maps were generated showing density, location, and invasion patterns of both musk and Canada thistle.

Herbicide trials were applied in September 2014 at two locations on the ranch (east and west). Individual treatments were applied to the entire plot, and the same herbicide and rate reapplied on half the plot in 2015 to compare effects of single and consecutive applications. Vegetative cover was measured in each plot, and above ground vegetation collected, dried and weighed. Dried samples were subsequently analyzed for forage quality [See MATERIALS AND METHODS, right].


Results of the study suggest that long-distance seed dispersal of musk and Canada thistle is primarily by wind, livestock, and human and livestock related disturbance. Musk thistle is most abundant on lower elevation foothills, disturbed sites, and areas of concentrated livestock use. Higher elevation rangeland has reduced thistle density compared to lower elevation sites (Figure 3). However, Canada thistle is more abundant on higher elevation rangeland and irrigated pastures than musk thistle. The degree of livestock use and level of soil disturbance appear to be important factors influencing invasive thistle abundance and density.

Figure 3. Distribution of musk thistle (left) and Canada thistle (right) within the ranch based on data collected from transects and prediction based on inverse distance weighting (IDW).

All herbicide treatments significantly reduced musk thistle cover compared to non-treated plots across the two locations (Figure 4 and 5). The application of two consecutive treatments in 2015 and 2016 reduced musk thistle cover more than a single application (Figure 6). Milestone® herbicide at 3 fl oz/A provided similar control as Milestone at 7 fl oz/A or Tordon® 22K herbicide at 16 fl oz/A.

Figure 4. Percent musk thistle cover one and two years after a single herbicide treatment (YAT) at the east study location. Herbicides shown in this figure are labeled for use on livestock grazed lands.

Figure 5. Musk thistle is well established on disturbed sites such as this stock pond. Musk thistle prior to Milestone® application at 7 fl oz/A (left) and the same area one year following application (right)

Figure 6. Percent musk thistle cover two years after a single herbicide application (YAT) and one year after two consecutive herbicide applications at the west study location. Herbicides shown in this figure are labeled for use on livestock grazed lands.

Figure 7. Percent cover of native forbs one and two years after a single herbicide treatment (YAT) at the east study location. Herbicides shown in this figure are labeled for use on livestock grazed lands.

Cover of native forbs was significantly reduced by all herbicide treatments compared to non-treated plots. However, forb cover increased the second growing season following a single application of Milestone at 3 or 7 fl oz/A at the east location (Figure 7).

There was no significant difference in species richness in herbicide treated compared to non-treated plots at the east location one growing season following a single herbicide application. However, species richness was reduced by all herbicide treatments at the west site and at both sites following consecutive herbicide treatments.

There was no significant difference in total biomass of grasses and forbs (excluding musk thistle) in herbicide treated compared to non-treated plots. Forage quality was similar between treated and non-treated plots.

Management Implications

This study was undertaken exclusively for the benefit of a private cattle ranch to determine the best course of action in reducing the musk thistle problem. In concert with that goal is preservation of the productivity of the cattle operation while at the same time maintaining the overall health and vigor of the native plant community. Management priorities on a private cattle ranch are dictated by budgets and goals for the ranch, rather than outside stakeholder influence common on public lands. Operating without these constraints facilitates management decisions.

 This study successfully identified several different treatment regimens that can accomplish many of the objectives of the ranch by controlling invasive thistle and minimizing impact on the overall forage and species diversity. Results from field trials showed that regardless of herbicide choice, thistle can be controlled while preserving forage quality and species richness. This gives the weed manager greater flexibility in considering cost, availability, and label restrictions when deciding on an herbicide treatment.

Published February 2017; revised June 2019

®™Trademarks of Dow AgroSciences, DuPont or Pioneer and their affiliated companies or respective owners.

Milestone is not registered for sale or use in all states. Contact your state pesticide regulatory agency to determine if a product is registered for sale or use in your state. When treating areas in and around roadside or utility rights-of-way that are or will be grazed, hayed or planted to forage, important label precautions apply regarding harvesting hay from treated sites, using manure from animals grazing on treated areas or rotating the treated area to sensitive crops. See the product label for details. State restrictions on the sale and use of Milestone apply. Consult the label before purchase or use for full details.

Tordon 22K is a federally Restricted Use Pesticide. Always read and follow label directions.©2019 Corteva

Materials and Methods

Thistle density was determined along 32 permanent transect. Each transect was 0.5 to 1 kilometer (1,640 to 3,280 feet) long, about 10 meters (33 feet) wide, and located perpendicular to roads (Figure 8). Thistle data within each transect were recorded and data entered into ArcGIS. Inverse distance weighting (IDW) was used to predict the location of invasive thistle on sites that were not measured. More Information on IDW

Herbicide trials were established in September 2014 at two locations, one on the east portion of the ranch (mesic moisture regime) and a second site on the west portion (xeric moisture regime). Plots were arranged in a randomized complete block design with three replications per treatment. Individual plot size was 20 by 100 feet. Milestone® herbicide at 3 or 7 fluid ounces per acre (fl oz/A), and Tordon® 22K herbicide at 16 fl oz/A were applied with a research sprayer in a total application volume of 20 gallons per acre. A consecutive treatment of the same herbicide and rate was applied to half the plot (10 by 100 feet) in September 2015.

Figure 8. Permanent vegetation transects were placed at 32 locations on the ranch to collect baseline vegetation data and allow for long-term monitoring. Transects were located perpendicular to roads with one end beginning on the roadside.

Relative cover of musk thistle, forbs and grasses in treated plots were compared to non-treated plots. Cover data were collected using the Extended Daubenmire Method with six frames (0.5 meter squared [m2]) per replication in 2015 and 2016 (Figure 9).

Figure 9. An Extended Daubenmire Method (0.5 meter squared) was used to collect cover data within herbicide treated plots and the non-treated control.

Vegetative biomass was collected from six 0.25 m2 subplots within each treatment. Grasses were combined from each subplot, and forbs were separated by species. Vegetation was bagged, dried and weighed. Dried vegetation from each plot was sent to Ward Lab in Nebraska for wet chemistry feed analysis (i.e. crude protein, acid detergent fiber, neutral detergent fiber, total digestible nutrients, calcium, phosphorous, potassium and magnesium). Individual forb species were combined for the analysis with for a total of 30 grass samples and 26 forb samples across the two locations. All data were compared by Analysis of Variance (p<0.1).

*John Coyle is a former graduate student at Colorado State University and currently owner of Arc Valley Weed Management and Consulting in Cañon City, Colorado. He can be reached at Scott Nissen is Professor of Weed Science, Colorado State University.

^Geographic Information System (GIS) is the umbrella term for the system used to collect, manage and present geographic data. Geospatial analysis is the application of analytical techniques to the GIS data. It can be as simple as adding a buffer zone around a ranch road or as complex as interpolating musk thistle data to form a ranch-wide gradient of thistle densities.