Introduction

Epilobium ciliatum is in the evening primrose family (Onagraceae). It is one of the most difficult weeds to control in container crops, and is among the top five weed species in Oregon nurseries. Its seeds are dispersed via wind, it germinates rapidly, it grows vigorously, and few herbicides provide effective control. This article will discuss the nomenclature, biology, and effective control of this weed.

Name confusion

Epilobium ciliatum gets its name from the Greek word epi which means upon, and lobos, which means pod. This is characteristic of the flower of the plant that surmounts the pod-like ovary. Ciliatum, also Greek in origin, means fringed with hairs and is characteristic of the pubescence that lines the stem and inflorescence.

Other Latin synonyms for E. ciliatum are E. watsonii and E. adenocaulon, but all refer to the same species. E. ciliatum is commonly, but wrongly, referred to as fireweed. Most authorities agree it should be called ‘northern-willowherb’. Common plant names vary from region to region and thus are notorious for causing confusion, this is no exception. E. ciliatum is also known by the common names hairy willowherb, slender willowherb, or fringed willowherb depending on the region where it is found. The common name ‘fireweed’ is more accurately in reference to E. angustifolium. It is also common in the Pacific Northwest, although not in nursery production. It is a familiar and beautiful plant along many Oregon roadsides and forests.

The focus of this article is the troublesome E. ciliatum or northern-willowherb.

Establishment from seed

The seeds of northern-willowherb develop in long pubescent capsules about 3 cm in length. The seeds are attached to a tuft of hair, which aids in wind dispersal. Seed dissemination occurs from June to September. Wind dispersal mandates excellent sanitation for controlling this weed. Eliminate any plants growing in non-crop areas in and around the nursery. Northern-willowherb establishes readily in coarse gravel beneath containers. Hand-weed or use postemergence herbicides to eliminate these plants, do not allow them to mature and disperse seed.

The container environment provides an ideal site for germination. Research by Myerscough and Whitehead (1966) has thoroughly documented the factors that affect germination and establishment of this weed. Seeds can germinate in low light conditions, making them well adapted to germination under the canopy of container crops. Seeds can germinate over a range of temperatures from 40 to 86 F, although germination is more rapid in warmer conditions. This allows germination to occur virtually year round in Oregon container nurseries.

We conducted a simple trial at the research station to determine how quickly seed germinate after sowing. We collected seeds that matured on plants growing in the nearby landscape. We immediately sowed them in containers under overhead irrigation located in either full sun, under shade, or within a retractable roof greenhouse. Germination occurred in just 4 days in full sun, and in less than 7 days in the other sites.
Recently potted containers are highly susceptible to infestation from this weed. Herbicides should be applied soon after potting (within 3 to 5 days). Waiting more than one week provides ample time for containers to become infested with wind-borne seed and for germination of those seeds to occur. Herbicide application soon after potting is especially important during mid to late summer when plants are actively disseminating seeds.

Plant structure and growth

Northern-willowherb is a perennial weed that is simple and upright. Mature plants grow 3 to 5 feet in height. Unlike many other container weeds, northern-willowherb’s size (height) allows it to quickly outgrow many container-grown perennials and shrubs. Overshadowing the ornamental crop leads to reduced size due shading of the ornamental. A lavender crop at the research station was devastated by an aggressive northern-willowherb population that outgrew the lavender, stunted its growth, and accelerated a foliar blight.

Northern-willowherb grows equally well in full sun and shaded conditions. Research has shown that shoot growth is similar in 0 to 57% shade.

E. ciliatum produces a rosette as an over wintering structure. It may also produce turions, a type of hibernation bud in the fall. It is common to see these over-wintering structures in gravel container yards during the winter. As soon as weather warms in the spring, these rosettes bolt quickly into a tall stem with a flush of lush green to purple foliage. Remove these plants during the slow winter months before they bolt and produce more seed.

Control with granular herbicides

Most herbicides come in granular and sprayable formulations. Granular formulations are more common in container production due to their broad range of safety. An experiment at the North Willamette Research & Extension Center evaluated eleven of the most popular granular herbicides for controlling northern-willowherb.

Trade gallon containers were filled with 100% Douglas fir bark amended with Osmocote and Micromax. Containers were treated with the herbicides and rates listed in Table 1. Rates were based on the maximum labeled rate for each product. Herbicides were incorporated with ½ inch of water immediately after application. Containers were irrigated daily thereafter with ½ inch irrigation per day using overhead sprinklers. Seeds were applied to one group of containers immediately after herbicide application and to another group of containers 30 days after herbicide application. Applying weed seeds 30 days after herbicide application was done to demonstrate how effective herbicides are in preventing weed establishment long after herbicide application.

Control was evaluated 60 days after herbicide application and reported in Table 1. Control was measured by counting the number of weeds, and then by measuring the dry weight of weeds in each container (hereafter termed shoot dry weight, or SDW).

When seeds were applied immediately after herbicide application, Ronstar and PrePair provided the most effective control in terms of reducing weed number and SDW. Non-treated control pots had fewer weed numbers than most herbicides; however, this is likely because the weeds grew so large (SDW) that several of the largest weeds outgrew the others. Despite relatively high germination numbers, Pendulum and Regal O-O greatly reduced SDW. In this situation, the seeds germinated but were so poorly stunted that they failed to grow beyond the cotyledon stage. Regardless of high germination numbers, the apparent level of control provided by Pendulum and Regal O-O was considered to be high.

When seeds were applied 30 days after herbicide application, Ronstar and PrePair again provided the most effective control. Regal O-O and Pendulum also provided good control at this phase of the study.
Control with sprayed herbicides

Sprayable herbicides are becoming increasingly popular in container production because they are easy to apply with large boom sprayers, they can be applied more uniformly, and they are often less expensive. This experiment was conducted similarly to the first experiment except that the sprayable herbicides Gallery, Devrinol, Factor, and Surflan were compared to the granular herbicides RegalKade, Ronstar, and Rout. The sprayable herbicides were applied either alone or in combination with another product, as listed in Table 2.

When seeds were applied at the time of herbicide treatment, Devrinol and Surflan did not dramatically reduce weed numbers. However, weeds that germinated were stunted and did not grow. Thus Devrinol and Surflan provided effective control whereas Factor provided poor control and Gallery provided no control. Among the granular herbicides, RegalKade and Rout provided poor control. Ronstar provided excellent control, which is consistent with the previous experiment.

When seeds were applied 30 days after herbicide treatment, weeds were overall smaller throughout all treatments primarily because it was later in the summer and weeds were growing more slowly. Among all herbicides, sprayed or granular, only Ronstar reduced weed numbers. Ronstar and Surflan+Gallery reduced weed growth more than other products.

Gallery provided no control throughout the experiment when applied alone, and it did not provide any additional control when tank mixed with other products.

Practical applications of these results

Ronstar or other granular herbicides that contain high rates of oxadiazon provide the most effective preemergence northern-willowherb control. However, one cannot rely on Ronstar alone. Another of Oregon’s major container weeds is pearlwort (Sagina procumbens), and it is very tolerant of Ronstar. Over-reliance on Ronstar may lead to a weed shift from northern-willowherb to pearlwort.
Gallery provided no meaningful control in this experiment. However, it is a good herbicide for controlling other weed species, and can be safely applied over the top of many woody ornamental crops. Gallery is a valuable tool for controlling weeds with sprayable herbicides in container production. Do not discount this product solely because of its poor control of this one weed species. Surflan was the most effective of the sprayed herbicides in this trial, although it was not as effective as granular Ronstar in long-term northern-willowherb control.

Summary

Northern-willowherb is well adapted for growing in the container environment. It is wind-dispersed, grows well in wet habitats, and often out-competes smaller container ornamentals. Based on results from 2 years and 9 experiments with this weed (we only presented 2 experiments here), herbicides alone will not provide satisfactory control. Sanitation is critical for reducing seed numbers and improving overall control success. Use a combination of sanitation and effective herbicide management to control this species.

Disclaimer: This article is for educational purposes only. Mention of a specific product should not be interpreted as an endorsement, nor should failure to mention a product be considered a criticism. Always read the product label prior to using any herbicide.

Additional reading

Myerscough, P.J and F.H. Whitehead. 1966. New Phytologist Vol. 65, No.2 (April) pp192-210.