The Slide Hammer Seeder: A Novel Tool forPlanting Small Seeds

Eric B. Brennan

ADDITIONAL INDEX WORDS. planter, intercropping, interseeding, vegetable, herb,insectary plant

SUMMARY. Many important herbs [e.g., mint (Mentha sp.), thyme (Thymus sp.)],underused and nutritious vegetables [e.g., purslane (Portulaca oleracea), amaranth(Amaranthus tricolor)], and important biological control plants [e.g.,sweet alyssum (Lobularia maritima)] have small seeds (£ 1.5-mm long) that aredifficult to plant with raw (i.e., nonpelleted) seed using existing seeders. A noveltool known as the slide hammer (SH) seederwas developed for the precise seeding ofraw seeds of small-seeded plants. The SH seeder is a jab-type planter made primarilyfrom electrical conduit tubing and other materials that are inexpensive and readilyavailable in a hardware store or on the Internet. The interchangeable seed hopper ismade from a plastic snap cap vial that has one or more holes of varying diameterdepending on the desired seeding rate and seed size. Seed forms a ‘‘bridge’’ above thehole in the vial until they are dislodged from the force of the SH that dischargesseeds to fall to the soil. Detailed plans are provided for how to make and use the SHseeder. The fabrication time is 2 to 4 hours with a material cost of �$32. Idetermined the seed vial hole specifications for the precise seeding of a variety ofsmall-seeded plants, including chives (Allium schoenoprasum), chinese chives(Allium tuberosum), basil (Ocimum basilicum), grain amaranth (Amaranthus sp.),sweet alyssum, purslane, creeping thyme (Thymus serpyllum), and spearmint(Mentha spicata) that ranged in size from �200 to 11,000 seeds per gram. Thediameter of the hole that was suitable for discharging the seed from the vial wasalways larger than the average seed length, and the ratio of hole diameter to seedlength ranged from 1.07 to 1.62. Seeding rate uniformity evaluations wereconducted for these species using vials with one vs. two holes and showed that theseeding rate was higher by an average of 58% to 173% from a vial with two holescomparedwith one hole. Formost plant species evaluated, the SH seeder was able todispense as few as one to three seeds consistently. Seed discharge increasedsomewhat with increasing SHweight for all species evaluated. The SH seeder can beuseful for interplanting sweet alyssum as an insectary plant for aphid (Aphidoidea)control between existing plants of organic lettuce (Lactuca sativa), and forintercropping cultivars of purslane as a novel vegetable in between transplantedorganic broccoli (Brassica oleracea Italica group) plants. This novel seeding tool hasmany potential uses for direct, hand seeding in vegetable and herb productionsystems and in weed research trials. The seeder could be automated and made witha variety of alternative materials.

Planting seeds by hand has beenimportant since the beginningof agriculturemore than 10,000

years ago, and hand seeding is still

important for major staple crops likecorn (Zea mays) in many parts of thedeveloping world (Aikins et al., 2010;Omara et al., 2016). Although jab-style hand seeders were widely used forcorn in the United States during theearly 1900s (Moore, 2003; Voelker,

2009), their use in the United Statestoday is primarily in research plots foragronomic crops with relatively largeseeds (Gonzalez et al., 2018; Leonet al., 2000). However, these seedersare unsuitable for precision planting ofsmall-seeded species that are of interestas cash crop or as ‘‘insectary plants’’ inmany horticultural systems. Insectaryplants are plants that are grown inhigh-value vegetable crops, such aslettuce and broccoli, to provide pollenand nectar for beneficial insects [i.e.,hoverflies (Syrphidae), parasitoid wasps(Hymenoptera)] that can help to con-trol aphids (Ambrosino et al., 2006;Araj and Wratten, 2015; Brennan,2013, 2014, 2016; Chaney, 1998;Hogg et al., 2011). One of the mostcommon insectary plants used in re-gions such as the Salinas Valley ofCalifornia is sweet alyssum, which hasthe rather unique ability to flowercontinuously once flowering com-mences. Although sweet alyssum andother common insectary plants areoften planted as transplants in thesevegetable cropping systems to ensurethe presence of flowers early in thedevelopment of the vegetables, thereare situations in which direct seedingof insectary plants may be useful andmore cost effective. For example,direct seeding insectary plants maybe appropriate where the vegetable isalso direct-seeded and thereforepresent in the field for longer thanwould occur with transplants. Fur-thermore, direct seeding may bemore cost effective than growingtransplants and may encourage theuse of insectary plants because it doesnot require advanced planning; seedsof insectary transplants typically needto be sown 30 to 50 d before they areready for transplanting. Direct seed-ing of insectary plants also may bea land-efficient way for farmers tomake use of gaps in seed lines result-ing from malfunction of the cash

UnitsTo convert U.S. to SI,multiply by U.S. unit SI unit

To convert SI to U.S.,multiply by

3.6967 dram(s) mL 0.270529.5735 fl oz mL 0.03380.3048 ft m 3.28083.7854 gal L 0.26422.54 inch(es) cm 0.3937

25.4 inch(es) mm 0.03940.1786 lb/inch kg�cm–1 5.5997

28.3495 oz g 0.03530.9144 yard(s) m 1.0936

U.S. Department of Agriculture, Agricultural Re-search Service, U.S. Agricultural Research Station,1636 E. Alisal Street, Salinas, CA 93905

I appreciate the helpful comments by JamesMcCreight and William Wintermantel that improvedthe manuscript. I am also grateful to the threeanonymous reviewers who provided critical sugges-tions to improve the manuscript.

Mention of a trademark, proprietary product, orvendor does not constitute a guarantee or war-ranty of the product by the U.S. Department ofAgriculture and does not imply its approval to theexclusion of other products or vendors that also maybe suitable.

Corresponding author. E-mail eric.brennan@usda.gov

https://doi.org/10.21273/HORTTECH04122-18

764 • December 2018 28(6)

crop seeder. To address the need fora seeder that allows the preciseplanting of small-seeded plants suchas sweet alyssum, a simple and ef-fective seeder called the slide ham-mer (SH) seeder was developed(Fig. 1). The objectives of this arti-cle were 1) to describe how to makeand use the SH seeder, and 2) toevaluate the accuracy of the seederfor a variety of small-seeded plantspecies including insectary plants,herbs, and novel vegetables. An onlinevideo describing how to make theseeder, a field demonstration of itsuse, and other features is described atBrennan (2018).

Material and methods

CONSTRUCTION OF THE SHSEEDER.Once thematerials areobtained,the time to make the SH seeder is 2to 4 h, and the estimated cost ofmaterials is $32.50 (Table 1). Thehandle is made of 1/2-inch electricalmetallic tube (EMT) conduit that iscommonly used to protect and routeelectrical wiring in buildings (Fig. 1).Table 1 lists the specifications of theEMT and the other materials neededwith the typical price from hard-ware stores in Salinas, CA. The fol-lowing tools are needed: tape measure,hacksaw or pipe cutter, scissors, utilityknife, flat-head screw driver, socket

driver for adjusting the hose clamps,pliers with a wire cutter, drill witha range of small bits (i.e., 1/16 inchor smaller depending on the size of theseed to be planted), round file with anouter diameter of �0.5 inch, benchvice, hammer, and flat-ended punch.The length of the handle may be varieddepending on the height of the opera-tor, but the prototype was cut witha hacksaw to 5 ft long and worked wellfor the author who is �6 ft tall. Toform the seeder foot plate, the lower3.5-inch end of theEMT is compressedin the bench vice until the 90� cornerbrace can be slipped into the EMT,after which it is compressed further tosecure the brace in the EMT. Thecompressed area can then be ham-mered against the anvil of the vise tofurther secure the brace. At �39.5inches from the bottom of the EMT,a 2-inch section of the EMT is com-pressed in the vice to form a slightlyflattened section [�0.5-inch o.d. at thenarrowest point (Fig. 2A)]. The innerstop ridge at the center of the EMTcoupler is filed down until the couplercan slide over the top end of theEMT and down to stop at the flat-tened section. The set screws on theEMT coupler are then tightened tolock the coupler to the handle. Theinner hole of the hammer stopwasher is filed to increase the diam-eter until the washer is able to slideover the top of the EMT and down torest on the EMT coupler. To makethe SH, the 3/4-inch metal tees areattached by hand to the threadedpolyvinyl chloride (PVC) riser andthe 90� elbow is attached to thelower tee. The SH is then placedover the handle to rest on the stopwasher and the sliding hose clampis added above the SH. The com-pression spring is then secured tothe top of the EMT handle usinga screw hose clamp (i.e., worm gear)that is tightened with an appropri-ately sized socket driver. The di-ameter of the sliding hose clampjust above the SH assembly shouldthen be adjusted so that it can slidewith the SH but not slide over thecompression spring. This ensuresthat the SH will engage the springwhen the seeder is lifted off theground.

The extension tube is made froma 2-inch-high by 4-1/8-inch widepiece of clear high-density polyethyl-ene cut from the central wall section

Fig. 1. The slide hammer seeder illustrating the various components that are aboutto scale; 1 inch = 2.54 cm, 1 ft = 0.3048 m. See Supplemental Fig. 1 for enlarged,color version.

• December 2018 28(6) 765

Tab

le1.M

aterialsneeded

formak

ingthevariousco

mponen

tsoftheslideham

mer

(SH)seed

erin

ord

erfrom

thetopoftheseed

erto

thebottom,an

dtheestimated

cost.

Quan

tity

orlength

Componen

tan

ddescriptionz

Cost

($)

Han

dle

5ft

1/2-inch

electricalmetallictubing(E

MT)·10ft(0.62inch

i.d.,0.71inch

o.d.).Typicallysold

in10-ftlengthsfor�$

3.30each.Thus,asingle

length

canmaketw

oseed

ers.

1.65

1-3

to4-inch-longsteelcompressionspringwithan

i.d.o

fabout0.72to

0.75inch

ando.d.o

fabout0.83inch.T

hisshould

slidefreelyontheEMT.

Theidealspringrate

isabout0.93lb/inch.Asuitablespringisonewherethespaces

betweenspringcoilscompressabouthalfway

when

theSH

seed

erislifted

upbytheslideham

mer.F

orexam

ple,see

compressionspringno.3

185(C

entury

SpringCorp.,Commerce,C

A)withthefollowing

specifications:madefrom

0.042inch

diameter

steelw

irewithazincfinishwith12coils,3.53inches

long(freelength),0.744inch

i.d.,0.828o.d.,

springrate

of0.93lb

per

inch.

4.55

1Hose

clam

pSAE#8,effectiverange:

1/2to

29/32inch,13to

23mm.

1.00

Slideham

mer

andham

mer

stop

1Hose

clam

pSAE#6,effectiverange:7/16to

25/32inch,11to

20mm.Adjusted

toslidebetweenthebottom

ofthecompressionspringandthe

upper

slideham

mer

coupler.

1.00

13/4-inch

·8-inch-longthreaded

PVCriser[PVC(1

inch

o.d.,0.82inch

i.d.)]thatwillslidefreelyovertheEMT.T

heserisersaretypicallysold

with

threads(14threads/inch)that

match

thethreadsontheirontees

notedbelow.

1.69

23/4-inch

threaded

,black,irontees.Theseattach

toboth

endsofthePVC

riser.

5.78

13/4-inch

threaded

90�ironelbow.

2.39

11–1

/2-inch

o.d.steelwasher,with�0

.69inch

i.d.Thei.d.willneedto

befiledslightlyso

that

itcanslidedownovertheEMT.Washer

thickness

should

be�0

.15inch.

0.15

11/2-inch

EMTscrew

coupling.Theinner

stopin

thecenterofthecouplerwillneedbefileddownso

that

thecouplercanslideovertheEMT.

0.38

Seeder

shoetubean

dseed

hopper

assembly

1Irrigationflag

with0.06-inch

o.d.,steelwire.

Thisisusedto

maketheJ-shaped

wirethat

securestheseed

vialin

theextensiontube.

0.10

1Hose

clam

p#6,effectiverange7/16to

25/32inch,11to

20mm.ClampstheJ-shaped

wireto

thehandle.

1.00

1ormore

Seedvial,7

drams,clear,polystyren

esnap

capvial.A

pproximately1.09inch

o.d.atthebase,1inch

i.d.,2inches

tallwithcapon.A

suitablevialisone

wheretheouterdiameter

atthevialbaseisaboutthesameorslightlysm

allerthan

theouterdiameter

ofthelarger

endofthecopper

reducing

couplerthat

thevialrestsonin

theSH

seed

er(Fig.5).Vialsareoften

sold

incasesof144vialsfor�$

40per

case;forexam

pleseeitem

no.PLC-

03732(Q

orpak,Bridgeville,PA).

0.28

1Smooth,clear,rectangularpiece

ofhigh-den

sity

polyethylen

e[H

DPE(2

inches

high·4-1/8inches

wide,

0.01-inch

thick)]cutfrom

thewallof

asingle-use

bottle

forwater

orsoda.

Thisisusedto

maketheextensiontube.

0.00y

3/4·3/4inch

Cellophanetape,

3/4-inch

wide.

0.00x

5·1inch

Ducttape.Thiscanbemadebycuttingortearinga1-inch-w

idestripfrom

a5-inch-longstripofducttapethatiscommonlysold

in1.89-inch-w

ide

rolls.

0.00x

11·1/2-inch

copper

reducingcoupler

4.59

1ft

1/2-inch

typeLcopper

pipe.

Typically

sold

in10-ftlengthsfor�$

14each.

1.40

3Hose

clam

pSAE#12,effective

range:11/16to

1-1/4inch,1

7to

32mm.Oneoftheseclam

pstheextensiontubeto

thecopper

reducingcoupler.

Theother

twoconnecttheseed

ershoeto

thebaseofthehandle.

3.00

1Hose

clam

pSAE#16,effectiverange:

13/16to

1-1/2inch,21to

38mm.Thisclam

psthereducingcouplerto

theEMT.

1.19

1Seeder

foot.3.5

·3/4inch

90�corner

brace.Both

armsofthebrace

are3.5

inches

long.

1.49

Total

31.64

z1ft=0.3048m,1inch

=2.54cm

,1inch

=25.4

mm,1lb/inch

=0.1786kg�cm

–1,1thread/inch

=0.3937thread/cm

,1dram

=3.6967mL.

yNoestimated

price

islisted

forthepiece

ofHDPEbecause

thiswillbefree

from

abottle

that

was

toberecycled

.xNoestimated

price

islisted

because

ofthesm

allam

ountoftaperequired

.A50-yard(45.7

m)rollofduct

tape1.89inch

widetypically

costs�$

5,anda300-ft-longrollof3/4-inch-w

idecellophanetapetypically

costs�$

3.

766 • December 2018 28(6)

RESEARCHREPORTS

of a 2-L plastic soda or water bottlefollowing the steps in Fig. 3. Theseeder shoe is fabricated from a 1-ft-long section of 0.5-inch type L cop-per tubing as described in Fig. 4. Theextension tube and attached copperreducing coupler (Fig. 3G) is thenslipped over the top of the seedershoe, and three hose clamps are usedto attach the seeder shoe and seedhopper assembly to the lower end ofthe handle by the seed foot plate(Fig. 5). The union of the seedershoe and coupler can be secured asneeded with duct tape, but normallythis is unnecessary. The J-shapedwire that holds the seeding vial inthe extension tube is made from a 6-inch-long piece of wire cut from anirrigation flag and bent into the J

shape using pliers. The J-shaped wireis then secured to the handle witha hose clamp so that it touches thetop of the seed vial sitting in theextension tube (Fig. 5). The positionof the seeder shoe and seed hopperassembly relative to the seeder footplate is adjusted up or down on thehandle as needed, depending on thedesired seeding depth.

Seed vials for dispensing seed aremade by drilling one or more holes inthe bottom of the vial depending onseed size and desired seeding rate.The holes should be drilled in the flatbottom area of the vial �3 to 5 mmaway from the outer edge of the vial(Fig. 2B). If multiple holes are used ina vial, it is best to position these onopposite sides of the vial bottom so

that they function independently (i.e.,if a piece of debris in the seed clogs onehole, the other hole is far enough awaythat it will remain open).

SEED VIAL HOLE DIAMETER

SELECTION, CALIBRATION, AND USE

OF THE SH SEEDER. The easiest wayto determine an appropriate hole di-ameter that will consistently dis-charge seed is to use a drill bit witha diameter that is at least as largeas the longest dimension of the seed.Standard drill sets available in hard-ware stores typically range from aminimum of 1/16-inch diameter(1.59 mm or�#52 to #53 wire gaugesize) to 0.5-inch diameter, butmay beavailable as individual bits in smallerwire gauge sizes or in jewelry drill bitsets that typically range from #80(0.34 mm) to #60 (1.02 mm). Mul-tiple holes can be added to oppositesides of the vial bottom to achievehigher seeding rates and can beplugged as needed by covering themwith adhesive tape.

To determine the minimum ap-propriate hole diameter, it is helpfulto place a representative seed sampleinside a folded piece of transparentadhesive tape so that it can be com-pared with the diameter of availabledrill bits or seed of another specieswhere the appropriate hole diameterhas already been chosen. A hand lensor dissecting microscope can help tocompare seed size with drill bit di-ameter. A good starting point is totest holes that are slightly larger thanthe seed length.

Once a candidate hole diameterhas been selected and the hole hasbeen made, the vial can be tested inthe seeder. To do this, add seed to thevial so that it is �1/4 to 3/4 full, andload the vial into the extension tube.If the vial has insufficient seed, theseeds may not be dispersed evenlyenough over the holes to allow fora uniform seeding rate. With onehand, hold the seeder handle perpen-dicular to the ground with the seederfoot plate resting on a block of woodso that the seeder shoe is just abovethe rim of a glass or plastic jar (e.g., 16fl oz, �4 to 5 inches tall) that willcatch discharged seed. With the otherhand, raise the SH up to the spring onthe top of the handle, and then releasethe SH so that it drops and hits thehammer stop washer. Alternatively,the seeder can be attached temporar-ily with screws through the seeder

Fig. 2. Detailed illustrations of the hammer stop assembly (A), and the position ofthe holes in seed vials of the slide hammer seeder (B); 1 inch = 2.54 cm, 1 dram =3.6967 mL. EMT = electrical metallic tube; PVC = polyvinyl chloride. SeeSupplemental Fig. 2 for enlarged, color version.

• December 2018 28(6) 767

foot plate to a large wooden block orshipping pallet, so that the seeder issecured in an upright position anddoes not need to be held vertically byhand during testing. Count the num-ber of seeds discharged with each hitof the SH and repeat at least 10 timesas needed to test for seeding rateuniformity with a single hole open inthe seed vial. For some species, a vialwith two or more smaller holes mayprovide a more uniform minimumseeding rate than a vial with a singlelarger hole.

Once an appropriate hole diam-eter has been determined that consis-tently delivers a minimum number ofseeds with each hit, and relatively fewor no hits with no seed discharged,additional holes can be added to the

vial to achieve the desired seedingrate. If one has a limited selection ofdrill bit diameters, a round needle filecan be used to gradually increase thediameter of a hole, but this may notbe ideal because it may be difficult toreplicate. As noted in the followingPractical Implications section, theseeding rate also can be adjusted by‘‘diluting’’ the target seed with in-expensive seed of another species thathas been killed.

To adjust the seeder to the ap-propriate seeding depth, it is helpfulto use a light-colored test seed, suchas grain amaranth, that is relativelyeasy to see in dark-colored soil. Theseeder shoe tube can then be movedup or down on the seeder handle toadjust the seeding depth for the site-

specific soil conditions. After the ap-propriate seeding depth has been de-termined with the test seed, the vialcan be replaced with the appropriatevial of the desired seed species. It maybe helpful to try using several seedingdepth and seed vial settings in a smallarea of a field, and count seedlingemergence to help to select the bestseeder settings.

The procedure for using the SHseeder for field planting involvesholding the seeder in one hand, plac-ing the seeder foot plate on the soilsurface, striking the SH against thehammer stop washer, and then liftingthe seeder to the next planting loca-tion. It is important to stress that theprimary force striking the SH stopwasher comes from the weight ofthe SH that is held relatively looselyin the hand of the operator as the SHfalls to the stop washer. If an appro-priate vial hole was chosen, the SHdoes not need to be struck hardagainst the hammer stop to dischargeseed. After the hit, the seeder can beleft to rest on the soil and leanedforward as the operator steps forwardas if the seeder is a walking stick that isthen lifted and moved to the soilsurface at the next planting location.Using this motion, it is relatively easyto sow seeds in 30 to 40 target loca-tions in 1min as is shown in the onlinevideo (Brennan, 2018). The compres-sion spring on the top of the handleacts as a cushion to minimize thechance of dispensing seed when mov-ing from one target seeding locationto the next. The seed vial hole may betoo large if the vial consistently dis-charges seed when the seeder is liftedbetween planting locations.

HOW THE SH SEEDER WORKS.Seed is discharged in small quantitiesfrom the seeder when the force of theSH hits the hammer stop washer andthe force is transmitted down thehandle to dislodge seed in the vial(Brennan, 2018). This jarring forcetemporarily disrupts the ‘‘bridge’’ or‘‘arch’’ of seed directly above the holein the vial (Fig. 6) and allows a smallamount of seed to flow through thevial hole and down through theseeder shoe to the soil. After the seedhas been discharged, the seed directlyabove the vial hole reforms a bridgethat prevents additional seed frombeing discharged until the next hitof the SH. Although bridging orarching is a common problem that

Fig. 3.Materials needed and steps for fabricating the extension tube that holds theseed vial on top of the seeder shoe tube; 1 inch = 25.4mm. See Supplemental Fig. 3for enlarged, color version.

768 • December 2018 28(6)

RESEARCHREPORTS

can hamper the flow of dry mate-rials, such as seed and granular fertil-izers, from hoppers or silos, bridgingwith the SH seeder is advantageousbecause it prevents seed dischargeuntil the bridge is disrupted by theforce of the SH striking the hammerstop washer.

EX P T . 1 : S E ED I NG R AT E

UNIFORMITY EVALUATION. Seedingrate uniformity of the SH seeder witheach hit of the SH was tested withplant species with a range of seed sizes(Table 2) using vials with one and twoholes. Grain amaranth seedwasobtainedfrom the bulk food section of a gro-cery store in Monterey, CA (WholeFoods Market, Austin, TX). Sweetalyssumseedwas producedbyKamprath

Seeds Inc. (Manteca, CA). All otherseeds were purchased from Johnny’sSelected Seeds (Winslow, ME). Grainamaranth is produced by three spe-cies with light-colored seeds [foxtailamaranth (Amaranthus caudatus),bloodamaranth (Amaranthus cruentus),prince’s-feather (Amaranthus hypochon-dracus)], but it is not known whichspecies produced the seed described inthis article.

For the eight species tested, theaverage seed length was determinedby randomly selecting four seeds andmeasuring the seed length with theeyepiece graticule on a dissecting mi-croscope thatwas calibratedwith a stagemicrometer. The SH seeder calibrationprocedure noted previouslywas used to

determine the minimum, appropriatevial hole diameter for seed of eachspecies with the seeder temporarilyattached to a large wooden block thatheld the seeder vertically. Theweight ofthe SHused in this evaluationwas 642gand it dropped 10 to 11 cm beforestriking the hammer stop washer. Twoholes of the selected diameter for eachspecies were drilled into the vial bottomon opposite sides and approximatelyone-fourth of the distance from the vialside (Fig. 2B). With one hole in the vialbottom closed with adhesive tape, thevial was loaded with seeds and thenumber of seeds discharged into a glassjar for each of 20 consecutive hits wasrecorded; the vial was approximatelyone-half full for all species except forthe smallest-seeded species (spearmintand creeping thyme), where it wasapproximately one-quarter full. Theseed vial was then emptied, the tapewas removed to open both holes, andthe vial was reloaded to test it with twoholes with 20 consecutive hits. Explor-atory Software for Confidence Intervals[ESCI (Cumming and Calin-Jageman,2017)] was used to calculate 95% con-fidence intervals (CIs) for number ofseeds discharged from one vs. twoholes, and the mean and 95% CI ofthe difference assuming that the one-and two-hole comparisons were inde-pendent groups. The one- vs. two-holemeasurements were considered differ-ent if the 95% CI of the difference didnot include zero.

E X P T . 2 : S H W E I G H T

EVALUATION. With spearmint, basil,and chinese chives, the effect of SHweight on seeding rate was evaluatedusing hammers with the followingweights: 532, 686, 816, 946, 1082,and 1217 g. Thus, the percentageincrease from the lowest-weight ham-mer was 29%, 54%, 78%, 103%, and129%, respectively. These weight in-crements were achieved by adding 3/4-inch fittings to the lower tee of theSH where the lowest-weight hammeronly had the PVC riser and the upperand lower tees attached, the 686-ghammer had the 90� elbow added (asin Fig. 1), and the subsequent in-crements had one, two, or three 3/4-inch PVC risers (1-1/2 inches long)each addedwith a 3/4-inch black, ironcoupling. Within each species thesehammer weight increments were eval-uated in a randomized complete blockexperiment with six replicates usingvials with two holes of the diameter

Fig. 4. Photographs of the steps to fabricate the seeder shoe of the slide hammerseeder; 1 inch = 2.54 cm. See Supplemental Fig. 4 for enlarged, color version.

• December 2018 28(6) 769

used in Expt. 1. For each hammerweight and replicate, the seed dis-charged from each of four hits of thehammer was counted and used tocalculate the average seed dischargecount. The seed vial was emptied andrefilled for each replicate and the vialwas inverted between treatments withineach replicate to remove any potentialseed settling effect. SigmaPlot (version13; Systat Software, San Jose, CA) andPROC REG in SAS (SAS Institute,Cary, NC) were used to determineand plot the best-fit regression re-lationships between seeding rate and

SH weight. Regression model selec-tion was done to maximize the ad-justedR2 for the best-fit first-, second-, or third-order model using the Selec-tion=ADJRSQ option in PROCREG.

Results and discussion

SEED VIAL HOLE DIAMETER

SELECTION, CALIBRATION, AND USE OF

THE SEEDER. The selected, minimumhole diameter for dispensing seedfrom the vials was always larger thanthe average seed length, although thisrelationship differed considerably by

species (Fig. 7; Table 2). For example,the selected hole diameter for spear-mint (0.71 mm) was similar to theaverage seed length (0.67 mm) witha ratio of 1.07, whereas the holediameter for chinese chives (5.16 mm)was nearly 2 mm larger than theaverage seed length (3.18 mm) witha ratio of 1.62. Consequently, vialhole diameter needed to be muchlarger than the seed length for chi-nese chives than spearmint, becausechinese chives seeds are angularwhereas spearmint seeds are almostspherical. Angular seeds dischargeless readily through the holes thanround seed; therefore, attempts touse smaller vial holes for chinesechives often resulted in more hitswith no seed dispensed.

EXP T . 1 : S E ED I NG R A T E

UNIFORMITY EVALUATION.As expected,with all species evaluated, the aver-age number of seeds dispensed witheach hit of the SH was greater withtwo holes in the vial bottom thanwith one hole (Figs. 8 and 9). Withfour of the eight evaluated species,the selected hole diameter resultedin some hits with no seed dischargedwith a single hole (spearmint, purs-lane, sweet alyssum, chinese chives),although among these four speciesthis usually represented 15% or lessof the hits; purslane was the excep-tion, with 30% of the single-hole hitsresulting in zero discharge. Thegeneral trend of wider CI in thetwo-hole than one-hole seeding rateindicates more variability with thetwo holes. With the smallest speciesevaluated (spearmint), the singlehole discharged a single seed in 13of 20 hits, and two holes dischargedtwo seeds in 10 of the 20 hits. Seedsingulation is desirable where seed islimited or expensive, or where themarket demands harvested productsof a uniform size, such as withlettuce that is sold by the head.

E X P T . 2 : S H W E I G H T

EVALUATION. SH weight influencedseed discharge from the SH seederfor the three species evaluated andoverall tended to increase gradually asSH weight increased (Fig. 10). Forexample, the average increase in seeddischarge (seeds per hit) from thelightest (532 g) to the heaviest SH(1217 g) was 1.5 (mint), 2.7 (basil),and 1.5 (chinese chives). Thus, al-though the SH weight had increasedby 129% from the lightest to heaviest

Fig. 5. Detailed illustration of the seeder shoe tube and seed hopper assembly ofthe slide hammer seeder; 1 inch = 2.54 cm. See Supplemental Fig. 5 for enlarged,color version.

770 • December 2018 28(6)

RESEARCHREPORTS

SH, the average increase in seed dis-charge per hit was 80% (spearmint),44% (basil), and 17% (chinese chives).It is important to note that althoughthe average number of seeds dischargedper hit was always greater from the1217-g than 532-g SH for all threespecies evaluated, the 95% CI of thepaired difference between the lightestand heaviest SH, shown in [ ] waspositive for spearmint [0.3, 2.8] andbasil [1.1, 4.2] but not for chinesechives [–3.5, 6.5]. This indicates morevariability in the effect of SHweight onseed discharge with chinese chives thanfor spearmint and basil.

PRACTICAL IMPLICATIONS AND

POTENTIAL MODIFICATION OF THE

SH SEEDER. The SH seeder is a novel

seeder design that in its current formor modification may be useful ina variety of settings from the homegarden to transplant nurseries andlarge-scale vegetable operations whereone wants to plant small amounts ofseed at a fairly uniform seeding rate.For example, in the home garden orsmall-scale nursery where a limitedamount of seed needs to be plantedin a flat for transplanting, a seed vialcould be held by hand and movedslowly over the soil, while tappingthe vial cap with the other hand todispense the seed on the soil’s sur-face. The seed could then be coveredwith a shallow layer of soil. This vialtapping method of dispensing seedresults in more accurate and even

seeding than if the vial was shakenlike a salt or pepper shaker over thesoil surface (no data presented).

In large-scale vegetable opera-tions in the Salinas Valley of Califor-nia, I have found the SH seeder tobe a rapid and efficient way to inter-seed raw sweet alyssum seeds as in-sectary plants at specific locationsbetween lettuce seedlings. Thisinterseeding can be done after let-tuce emergence or thinning, so thatthe interplanted sweet alyssum canbe placed at specific locations relativeto the lettuce (i.e., equally spacedbetween the lettuce, or where thereare gaps due to unintended skips bythe lettuce seeder). A 5/64-inch-di-ameter drill (1.98 mm) was used tomake the holes in the seed vial for theinterplanting of sweet alyssum inlettuce field trials (E.B. Brennan,unpublished data); this hole diame-ter is slightly larger than what wasused for the one- vs. two-hole com-parison [1.85 mm (Fig. 8A)] andworked well particularly where thevial had two holes. To ensure at leastone or more vigorous plants of theinsectary species like sweet alyssum,it is best to use a vial with multipleholes even if this results in severalemerged seedlings at each seedinglocation. More vigorous plants typi-cally will compete successfully withothers.

The SH seeder has also workedwell for interseeding cultivars of purs-lane as a novel and potential earliermaturing cash crop between trans-planted broccoli (E.B. Brennan, un-published data). Furthermore, inweed research, the SH seeder couldbe a useful tool for seeding smallamounts of weed seed at specificlocations within a cash crop. In anyof these cases, an additional way tocontrol the seeding rate other thanvarying the number or diameter of theseed vial holes is to ‘‘dilute’’ the seedof interest by mixing it with killedseed of an inexpensive species such asgrain amaranth. The diluting seedcan be killed by heating it for severalminutes in a microwave. Grain am-aranth may be ideal for dilutingother seeds because it is readilyavailable, inexpensive, and has rela-tively round seeds that flow wellthrough holes that can be made witha commonly available, thin drill bit[1/16 inch (1.59 mm)]. This drillbit diameter is slightly larger than

Fig. 6. Photograph of the ‘‘bridge’’ or ‘‘arch’’ of grain amaranth seeds formed overthe 1.51-mm (0.06-inch) diameter seed discharge hole drilled in the bottom of theseed vial using a #53-gauge drill bit. The five seeds that form the bridge areoutlined and prevent seed from moving through the hole until the bridge isbroken by the force of the slide hammer striking the slide hammer stop washer(Fig. 2A). For purposes of illustration, the hole in the vial bottomwas made at theouter edge of the vial bottom adjacent to the tubular wall of the vial and rim thatgoes around the edges of the vial bottom, but normally the hole would be madeaway from vial wall (Fig. 2B). See Supplemental Fig. 6 for enlarged, color version.

• December 2018 28(6) 771

Fig. 7. Relationship between the selected seed vial hole diameter vs. seed length. Seed length refers to the longest dimension ofthe seed and is shown for each species as a 95% confidence interval (CI) error bar with the mean at the horizontal line on the CI(n = 4). The reference line shows where there is a 1:1 relationship between seed length and vial hole diameter. Means and 95%CIs that are further to the right from the reference line indicate species in which the selected hole diameter was proportionallylarger than the average seed length and thus have a larger hole:seed length ratio. See Table 2 for the vial hole diameters ininches, millimeters, and wire gauge size; 1 mm = 0.0394 inch.

Table 2. Seed characteristics and vial hole diameters used to evaluate the slide hammer is of the handle is made out of electricalconduit tubing discusses the different components to the handles made of electrical conduit was used for a half-inch electricalconduit seeder.

Common name,cultivarz

Seed size(no./oz)y

Avg seedlength (mm)x

Vial holediam (mm)

Drill bit diamto make vial hole[gauge (#) or

inches]w

Spearmint 311,700 0.67 0.71 #70Creeping thyme 210,000 0.76 0.91 #64Sweet alyssum 71,000 1.53 1.85 #49Purslane, ‘Gruner Red’ 61,088 1.12 1.70 #51Grain amaranth 37,500 1.18 1.51 #53Chives, ‘Staro’ 20,100 2.12 2.78 7/64 inchBasil, ‘Genovese’ 17,900 2.80 3.97 5/32 inchChinese chives 5,900 3.18 5.16 13/64 inchzCultivar name is shown in single quotes where available. Listed in order of seed size.ySeeds reported at the Johnny’s Selected Seeds (Winslow, ME) website for all plants except for grain amaranth that was calculated from the approximate number per pound(600,000) reported in Putnam et al., (1989); these are provided as a rough guide because seed size is often thought of in seeds per unit of weight; 1 seed/oz = 0.0353 seed/g.xMean of four seeds; 1 mm = 0.0394 inch.wWire gauge size (no.) was the original unit of drill bit diameter measurement for many of plants but can be converted to inches where needed from the vial hole diametercolumn; 1 inch = 25.4 mm.

772 • December 2018 28(6)

RESEARCHREPORTS

the 1.51-mm hole used for the one-vs. two-hole comparison with grainamaranth (Fig. 7D), but has workedwell for seeding purslane mixed withkilled grain amaranth seed. Grainamaranth may even work to dilutesmaller seed as long as the small seeddoes not segregate to the bottom ofthe vial over time. Other inert ma-terials (i.e., sand) may also work asa diluting material for some speciesif the seed does segregate in themixture and the mixture dischargesfrom the seeder only when the SH ishit.

The force that discharges seedfrom the vial when the SH hits thehammer stop washer potentiallycomes from the weight of the SHand how fast and hard the hammer

is hit against the stop washer. Toachieve the greatest uniformity inseeding, the person operating theSH seeder should rely primarily onthe weight of the SH to dischargethe seed. In my experience using theSH seeder, a SH with a weight of500 to 700 g provides adequate forceto discharge seed and is not overlyheavy.

Although the SH seeder de-scribed here is hand-operated, thereare several ways that an automatedversion of it could be developed fordispensing small amounts of seedor other granular material at spe-cific locations in agricultural settings.For example, an automated SHseeder could be added to a precisionvegetable seeder to allow growers

to interplant small amounts of rawseed of sweet alyssum as insectaryplants at specific locations alongwith the vegetable seed. Poten-tially, this technique would bemore cost effective than using pel-leted sweet alyssum seed as describedin Brennan (2015). A similar ap-proach also could be used to inte-grate an SH seeder into a mechanicalvegetable transplanter or a grain drillfor cover crop seeding.

Although the SH seeder describedhere is made with materials that arereadily available in hardware stores indeveloped countries or through theInternet, alternative materials couldbe used to make the seeder in less-developed regions of the world. Forexample, a straight, strong, thin-culmed

Fig. 8. The number of seeds discharged from vials with one or two holes in the vial bottom with each hit of the slide hammerseeder for spearmint (A), creeping thyme (B), purslane (C), and grain amaranth (D). Plots A toD are in order of increasing vialhole diameter. Seed length refers to the longest dimension of the seed. The drill bit diameter in wire gauge size (#) or inchesthat was used to make the holes is included with each plot label; the vial hole diameter in millimeters can be used to convert thedrill bit diameter to inches if needed. The dots are the raw data (gray for one hole, black for two holes) of 20 consecutive hits(from left to right) that are clustered around a vertical error bar that is 95% confidence interval (CI) with themean at the centralhorizontal line on the bar. The mean difference between the one- vs. two-hole measurements and the 95% CI of the differenceare shown on the right of each plot with the percentage change from one to two holes noted above the difference; 1 mm =0.0394 inch.

• December 2018 28(6) 773

bamboo [e.g., golden bamboo (Phyl-lostachys aurea), black bamboo (Phyl-lostachys nigra), or weaver’s bamboo(Bambosa textilis)] may work well formany of the components (handle,seeder shoe, and SH). In addition, wireor strips of tire inner tube or otherwrapping material could be used in-stead of the hose clamps to attach thevarious components. Even the springon the top of the seeder could be madeby hand with wire or with some othercushioning material (Brennan, 2018).Furthermore, a plastic pill vial withholes drilled in the bottom may workwell as the seed vial.

Literature citedAikins, S.H.M.,A.Bart-Plange, andS.Opoku-Baffour. 2010. Performance evaluation of jab

planters for maize planting and inorganicfertilizer application. ARPN J. Agr. BiolSci. 5:29–33.

Ambrosino,M.D., J.M. Luna, P.C. Jepson,and S.D. Wratten. 2006. Relative frequen-cies of visits to selected insectary plants bypredatory hoverflies (Diptera: Syrphidae),other beneficial insects, and herbivores.Environ. Entomol. 35:394–400.

Araj, S.E. and S.D. Wratten. 2015. Com-paring existing weeds and commonly usedinsectary plants as floral resources for a par-asitoid. Biol. Control 81:15–20.

Brennan, E.B. 2013. Agronomic aspectsof strip intercropping lettuce with alyssumfor biological control of aphids. Biol.Control 65:302–311.

Brennan, E.B. 2014. Efficient intercrop-ping for biological control of aphids intransplanted organic lettuce. 26 June

2018. <https://www.youtube.com/watch?v=KVLgt2_J1Wk>.

Brennan, E.B. 2015. A biological controlbuffet in the salad bowl of America. 26June 2018. <https://www.youtube.com/watch?v=zLvJLHERYJI>.

Brennan, E.B. 2016. Agronomy of stripintercropping broccoli with alyssum forbiological control of aphids. Biol. Control97:109–119.

Brennan, E.B. 2018. How to make theslide hammer seeder. 14 Dec. 2018.<https://youtu.be/olO9zX1ggs8>.

Chaney, W.E. 1998. Biological control ofaphids in lettuce using in-field insectaries,p. 73–83. In: C.H. Pickett and R.L. Bugg(eds.). Enhancing biological control:Habitat management to promote naturalenemies of arthropod pests. Univ. Cal-ifornia Press, Berkeley, CA.

Fig. 9. The number of seeds discharged from vials with one or two holes in the vial bottom with each hit of the slide hammerseeder for sweet alyssum (A), basil (B), chives (C), and chinese chives (D). Plots A to D are in order of increasing vial holediameter. Seed length refers to the longest dimension of the seed. The drill bit diameter in wire gauge size (#) or inch that wasused to make the holes is included with each plot label. The dots are the raw data (gray for one hole, black for two holes) of 20consecutive hits (from left to right) that are clustered around a vertical error bar that is 95% confidence interval (CI) with themean at the central horizontal line on the bar. The mean difference between the one- vs. two-hole measurements and the 95%CI of the difference are shown on the right of each plot with the percentage change from one to two holes noted above thedifference; 1 mm = 0.0394 inch.

774 • December 2018 28(6)

RESEARCHREPORTS

Cumming, G. and R. Calin-Jageman.2017. Introduction to the new statistics:Estimation, open science, and beyond.Routledge, New York, NY.

Gonzalez, V.H.,M. Tollenaar, A. Bowman,B. Good, and E.A. Lee. 2018. Maize yieldpotential and density tolerance. Crop Sci.58:472–485.

Hogg, B.N., R.L. Bugg, and K.M. Daane.2011. Attractiveness of common insectaryand harvestable floral resources to bene-ficial insects. Biol. Control 56:76–84.

Leon, A.J., F.H. Andrade, and M. Lee.2000. Genetic mapping of factors affect-ing quantitative variation for flowering insunflower. Crop Sci. 40:404–407.

Moore, S. 2003. A history of corn: Hand-held corn planters and checkrows. 26 June2018. <https://www.farmcollector.com/equipment/history-of-corn-corn-planters>.

Omara, P., L. Aula, B. Raun, R. Taylor, A.Koller, E. Lam, J. Ringer, J. Mullock, S.Dhital, and N. Macnack. 2016. Handplanter for maize (Zea mays L.) in thedeveloping world. J. Plant Nutr. 39:1233–1239.

Putnam, D.H., E.S. Oplinger, J.D. Doll,and E.M. Schulte. 1989. Amaranth. 26June 2018. <https://hort.purdue.edu/newcrop/afcm/amaranth.html>.

Voelker, D. 2009. Stepping forward:Hand-held corn planters. 26 June 2018.<https://www.farmcollector.com/equipment/hand-held-corn-planters>.

Fig. 10. Relationship between slide hammer weight and the number of seedsdischarged from vials with two holes for spearmint (A), basil (B), and chinesechives (C). The weights of the slide hammers evaluated were 532, 686, 816, 946,1082, and 1217 g representing a 29%, 54%, 78%, 103%, and 129% increase,respectively, relative to the lightest hammer (532 g). Dots are the average numberof seeds discharged from four hits with the slide hammer for six replications ofeach slide hammer weight. The numbers to the left of a dot indicate the number ofreplicates (i.e., 3x = 3 replicates) wheremore than one replicate had the same value;1 g = 0.0353 oz.

• December 2018 28(6) 775