Harvesting Wildflower Seed Crops from Marginal Land

INTRODUCTION
Native plant seeds are harvested by hand or machine, with variations ranging from grabbing off seed heads to using a sickle, and from portable seed strippers to large self-propelled combines. The recent introduction of the portable seed stripper by Prairie Habitats Inc. represents an interesting possibility of increased harvest speed on rough terrain or on smaller plots where a combine would be too big. Basically the unit is a special seed tripping head and catcher bag mounted on the end of an ordinary line trimmer, and it retails for $750. It could make the harvest of native grass and wildflower seed from roadsides, wetlands, remnants, and small production plantings a more economically viable venture for agency personnel and farmers alike. With this in mind, I compared the efficiency of hand and portable stripper harvest for several species, and recorded the rate of hand harvest for several additional species.

PROJECT DESCRIPTION AND RESUTLS
The goal in all cases was to try and harvest wildflower and other native plant seeds at the peak of maturity so as to maximize the rate of harvest and the quality of the seeds. Hand harvest consisted of pulling heads off of Heliopsis helianthoides, or cutting off entire inflorescences with a hand pruner for the rest of the species. These materials were then placed into ordinary paper grocery bags and allowed to dry in my barn. The machine was operated according to the instructions and to my increasing experience, as the stripper heads to be used vary with the species and engine speeds must be varied with wind and plant conditions. Three replicate samples using each method were conducted, with the time per sampling varying according to how long it took the machine to fill its catch bag (table 1). In the case of the Heliopsis helianthoides, the plants harvested were in a production plot, so 6 m2 subplots were harvested and the time to harvest recorded. The machine harvested materials were also placed into paper bags and allowed to dry under the same conditions as the hand samples. The length of other hand harvests varied according to conditions and extent of the plants being collected.

Seeds were threshed out by hand (or foot) or with a small hammer mill depending on the species. These were than cleaned with an old Clipper fanning mill before being weighed with a spring scale. T-tests were used to compare the harvest rates of hand and machine methods. The state seed lab conducted tests of germination and purity. Economic returns of harvest are here reported as gross returns per minute of seed harvest, with the wholesale value calculated as 50% of current retail values in Minnesota.

Results and Discussion:
Four species, Scirpus cyperinus, Silphium integrifolium, Glyceria grandis, and Calamagrostis canadensis were studied but dropped form these results because the seeds were harvested too early, could not be reasonably separated, or complete data could not be collected. Several other species were also trialed in order to test out the machine in fall of 1995. However, these (Aster sericeus, Solidago rigida, Eupatorium perfoliatum, Zizia aurea, Ratibida pinnata) species were not used due to problems hitting maturity, finding enough plants, or cleaning the seed. Attempts with these species did demonstrate that there is a certain knack required to use the machine stripper effectively, and that it is awkward to use on very tall species and very short species. Also, many seeds are not captured by the brush and bag as it is currently designed.

Three species were harvested using both methods (table 2). The machine was significantly faster at collecting seeds of the two flowers, but did not differ from hand threshing for the Scirpus. I believe this is the case because the machine is very good at getting seed heads that are at a medium height on plants that have tight heads. With the Scirpus the machine seemed to blow away a good deal of seed and trash and also became tangled in the mass of seed heads that are presented in stands of Scirpus. The quality of seeds collected by the two methods did not vary much, and the higher amount of weed seeds in the hand collection of Scirpus atrovirens is primarily due to Glyceria grandis being accidentally co-collected (table 2).

With Heliopsis helianthoides collections were made on a real basis so that harvest efficiency could also be compared. In this case, hand harvest netted 23.8 g per m2 and the machine gathered 10 g per m2. The reason for this was readily observed in the field: seeds were flying all around regardless of the stripper engine speed. This might be controlled to some degree by harvesting while the plants are a bit greener to avoid shattering, though it appears that larger seeds are not easily collected by the stripper. However, it is also informative to note that this spring that patch of Heliopsis used has filled in with hundreds of new seedlings.

Gross returns per time spent harvesting (this does not include travel, set up, or seed handling) demonstrate the machine harvest can generate valuable seeds faster in some cases, though clearly the current market harvest can generate valuable seeds faster in some cases, though clearly the current market value of some seeds makes hand harvest perfectly reasonable (table 2). If a manager or seed collector has a short term interest in collecting, or little ready capital, hand harvest may be preferable. The machine stripper really shows its advantage on more easily harvested species and may well pay for itself if seed supplies are large and valuable to the collector, whether as a marketed crop or as cheaper seeds for wildlife management and restoration. Also, cheaper imitations could be created by any handy person, perhaps with some improved harvest efficiency for very large or very small seeds. The machine stripper is much louder than had harvesting, but it may be worth putting up with the noise on preserves which are being opened up to commercial harvest. The inefficiency of the machine on a real basis for some species could mean instant reseeding of the preserve while rewarding the commercial harvester with higher returns per unit time.

Table 1. Species harvested and harvest times.

Species, Sampling Time, Replicates
Heliopsis helianthoides, variable: see text, 3
Monarda fistulosa, 2 minutes, 3
Scirpus atrovirens, 7 minutes, 3
Scirpus cyperinus, 3.5 minutes, 3
Silpium integrifolium, 5 minutes, 3
Baptisia luecantha, 11 minutes, 1
Agastache foeniculum, 25 minutes, 1
Penstemon gracilis, 25 minutes, 1

Table 2. Harvest rates (grams per minute), gross return rates (dollars per minute), and seed quality for hand and machine stripper harvested native plant seeds.

Species, Method, Rate, Gross Return, % Purity, % Germination and Viable, % Weed, % Crop
Heliopsis helianthoides, hand, 11 g, $0.92, 99.79, 17.91, 0, 0
Stripper, 56 g, $4.68, 99.59, 8/96, 0, 0.01
Monarda fistulosa, hand, 11 g, $1.82, 77.14, 47/47, 0.02, 0.02
Stripper, 21 g, $3.47, 78.53, 45/45, 0.09, 0.50
Scirpus atrovirens, hand, 13 g, $1.72, 61.23, 0/88, .057, 0.07
Stripper, 13 g, $1.72, 75.61, 0/89, 0.07, 0.07
Baptisia leucantha, hand, 20 g, $3.96
Agastache foeniculum, hand, 7 g, $2.00
Penstemon gracilis, hand, 9 g, $3.96

* Please not that numbers in rate column are for the actual pure seed collected.