By Don Osmond, volunteer steward
Wild parsnip is a monocarpic perennial, usually taking 2-3 years to flower. I chose a 4 acre area in southwestern Wisconsin to evaluate my control efforts over time. Habitats ranged from dry degraded prairie to mesic lowland planted with forage. Spring burns occurred in 2021 & 2023. All the work was performed by me, so the methods are consistent year to year.
Herbiciding was performed using a calibrated backpack sprayer with 0.5% Garlon 4 (1.1 quarts/acre of product) & 0.5% MSO. This concentration was chosen based on an experiment consisting of a control plot & plots sprayed at 0.5, 1 and 1.5%, then monitored one year later. All 3 plots had similar very low densities of parsnip one year after spraying, compared to the high density control plot. It probably would be safer to settle on 1% concentration to account for variables, but I anecdotally had good results previously at 0.5%, so I stuck with that. The minimum concentration for broadleaf weeds on the Garlon 4 label is 1 quart/acre. Herbiciding was never done in the remnants. Spring spraying started in the first 2 weeks of May, depending on how cool it had been. Fall spraying started the second week of September (rosettes were in the process of emerging before then). I limited fall application to known, higher density patches since tall vegetation makes area searches unproductive.
For mechanical control, I use a sharpened Parsnip Predator shovel to sever the root a few inches below the soil. I place the shovel at the parsnip base to center the shovel notch on the stem, move the shovel back 2”, hold it about 30 degrees off vertical & slice the root with even force (no stomping or prying). If the plant doesn’t pull up easily, I move the shovel laterally & make another slice. If the plant isn’t swollen at the top of the stem, I throw it on the ground. If swollen, but flowers aren’t emerging, I throw it on the ground & cut the stem below the swelling with the shovel. Before June 15, if the plant is blooming or close to it, I throw it on the ground & cut the stem below the terminal flower & also below the lowest flower. Cutting the stem reduces the possibility that stem carbohydrates will be used to reflower, although I have no proof that is possible on a stem without a root. After June 15, I bag the cut plants & let them compost on site. I deposit the bags away from roads & trails since animals may break the bags open, allowing seed spread. My method is an attempt to keep the job manageable for one person. A 2 person team is ideal for digging & bagging. Keep in mind the plant is toxic to the skin, so long gauntlet waterproof gloves & a long sleeved shirt should be worn. Peak bloom for the early plants in my area occurs in late June. The first pass of mechanical control is in the first 2 weeks of June, with a second pass starting the 3rd week. The latter is to find late blooming plants, which are small & not in bloom during the first pass & thus easily missed. Many plants will have green seeds by July 9. Mechanical control can continue until July 15 (when seeds start to drop), but I try to be done by July 1 so I can switch to white sweet clover control. Despite trying to be meticulous, I always find a handful of plants each year that were missed & went to seed. For example, I did a second pass on June 19 & found a small but surprising amount of blooming plants July 1 to July 19. So a 3rd pass may be necessary. I’ve even occasionally found blooming plants, some quite tall, in September. If you only have limited time for one pass, shoot for July 4 in southern Wisconsin & stack the cut plants in as few piles as possible, away from roads & trails. This is well before seed drop, yet late enough to catch the early as well as many of the later blooming plants.
Here are the results:
Gallons of herbicide solution used
2020 (fall only): 9
2021: 30
2022: 9
2023: 9
2024: 6
2025: 2
Hours for mechanical control
2021: 19
2022: 12
2023: 6
2024: 7
2025: 8
Estimated plant density over this 4.5 year time period decreased from moderate-dense to light. Notice that mechanical control hours plateaued after 2 years, probably because at that point most of the time was spent walking around looking for plants.
The results roughly correlate with the following:
1) Cain, Nancy, et al. “The biology of Canadian weeds: Pastinaca sativa L.” Canadian Journal of Plant Science volume 90 (2) page 217 (2010). They report the seedbank is viable for 2 to 4 years.
2) A prairie expert plus the above reference indicate most parsnips produce flowers in the 2nd or 3rd year but some take 5 to 6 years.
Using my method, 2 factors can lengthen the time for adequate control.
1) Missed plants due to poor visibility in heavy vegetation, plants that bloom at very low heights or very late blooming plants (as late as September). It is very easy to miss short, blooming plants in tall vegetation. Especially since invasion was widespread at this site, requiring a search of the entire area, except for the dry remnants where parsnip success was poor.
2) Digging disrupts soil, bringing seeds to the surface where they are more likely to germinate. Proper use of the Parsnip Predator (no prying) minimizes this.
If practitioners are not meticulous, time for adequate control will be longer than what I experienced. For example, not making a second mechanical pass, thus missing a surprising amount of late blooming plants. Or missing a year, allowing substantial seed drop. That would severely set efforts back & illustrates the importance of careful site selection & control area size to match the predicted resources over the years needed.
Herbiciding is particularly effective after a spring burn since thatch removal makes rosettes quite visible. Herbiciding can be skipped if there are enough people to complete 2 thorough mechanical passes for the number of years needed. My results are a good example of how herbicides can be a force multiplier if resources are limited. Without them, there would be no hope of making a meaningful improvement at this site with just one worker.
grassland restoration network 
Thank you, Don, for this thorough treatment of the subject. I have two questions for you: 1) Why Garlon 4 instead of 3? Due to its effectiveness at lower temperatures? 2) Do you have anecdotal or other experience of swollen stem-ends/almost flowering plants/flowering plants producing viable seed if their reproductive parts are left attached after the stem is cut? My understanding was that as long as the plants are mechanically controlled at full flower or before, viable seed would not be produced. This is considerable incentive to target our work in May/early June (“one cut and done”), but if even plants in very early stages of flowering can produce viable seed, that is another point in favor of the herbicide approach or other early-season control
I like your point about matching the site to the long-term resources.