Weed Science, 45:77-84. 1997

Broom sna keweed ( sarothrae) dispersal, viability, and germination

Ballard L. Wood Broom snakeweed achene dispersal was monitored by placing surface-level traps out- Department of Animal and Range Science, Nov wards in rhe cardinal directions from 12 and collecring the achenes weekly Smte Univeniry, Las Crucer, NM 88003 or bi-weekly from September 1993 until seeds were no longer retained by the planrs after 42 wk. About 50% of the achenes dispersed between October and December. Kirk C. McDaniel Especially high numbers of achenes were dislodged during periods of intense winter Corresponding aurhor. Department of Animal and winds and rains, with 78% of the seed placed into the east tray and 86% falling Pwngc Science, Box 30003, Deparrmenr 3-1, New wirhin 50 cm of the parent . Achene producrion averaged 3,928 (2 1,146) per Mex~caScare Universlry, Las Cruces, NM 88003 plant in 1993 and 2,036 (2 987) per plant in 1994. Achenes collected over time dirccrlv from the inflorescence and achenes stored in nvlon ~ackenon the soil surface Dennis Clason averaged 82% viability during fall and winter. Ache& viibility dedined rapidly in Experimental Srarisrics Department, late spring, and few remained viable before the next seed crop. Greenhouse experi- Srare University, Laa Cruces, NM 88003 ments compared the influence of water application intend and water amount on broom snakeweed aermination and seedline, survival. Treatments consisted of 4 water intervals: daily, 5-x, 10-d, and 15-d inrervals; and 4 water amounts: field capacity (111 fc), 314 fc, 112 fc, and 114 fc. Germination was 52% at daily 111 fc, and no seed germinated at daily 114 fc. Data suggest that optimum germination occurs when soils are maintained at a minimum soil mavic ('Pm) > - 180 kPa for at least 4 d. Optimum 'Pm for seedling survival appears ro range benveen -300 and -900 kPa, while seedling mortality would generally be expected with a Vm of > -1800 kPa.

Key words: Seed longevity, seed ecology, seed retention, seedling survival, soil seed reserve, soil matric potential, GUESA.

Broom snakeweed is indigenous to semiarid western 1985). The seeds from ray florets are brown, roughly cylin- from Mexico to Canada (Lane 1985). In New drical (from 0.9 to 1.6 mm long and from 0.2 to 0.7 mm Mexico, broom snakeweed is considered a major weed wide) and weigh about 0.15 mg per achene. The often in- problem because it interferes with productivity of desirable fertile disk achene is always smaller than the ray achene. forage and often becomes dominant on large tracts of blue The achenes are pubescent when rows of white trichomes grama , causing substantial economic losses to are oppressed to the seed coat in che direction of che pappus. livestock producers (Carpenter et al. 1990; Torell et al. The trichomes act to anchor che achene and enhance soil 1988). Herbicide and burning programs can provide good penetration; presumably, they draw and retain water next to control, but treatment life is variable because of che cyclic the pericarp (Mayeux 1989). The pappus consist of small nature of bmm snakeweed populations (McDaniel and Dun- whire or yellowish erose scales (< 1 mm length) aligned can 1987). with the axis of the achene. This highly reduced pappus is Broom snakeweed is a short-lived perennial that propa- unlike that found with most members of the tribe , gates by seed in years when optimal environmental condi- which usually have a well-developed pappus for wind dis- tions occur, then subsequently dies from 1 or more factors, persal (Lane 1985). Thus, broom snakeweed seeds are absent including drought (McDaniel 1989), insect infestation of any specialized structures to facilitate long-range dispersal, (Wangberg 1982), and old age Uameson 1970). The average and most seeds fall close to the parent plant when dislodged life expectancy of broom snakeweed surviving beyond the (Osman et al. 1987). Good seed crops may occur every year on some sites, but climatic factors, plant age, and insects 1st year is approximately 4 yr, but some plants may live t; generally cause wide fluctuations in seed production from longer than 15 yr (Dittberner 1971). Environmental mech- year to year (McDaniel 1989). - anisms chat trigger a mass germination of seeds or factors By studying broom snakeweed seed ecology, additional that result in a major die-off of mature plants are not fully insight into the plant's population dynamics can be gained, understood, but this information is needed to make sound thereby increasing the opportunity to develop more efficient decisions about broom snakeweed control and management management strategies. The objective of this study was to (Torell et al. 1989). investigate the production and dispersal of broom snake- The inflorescences of broom snakeweed are numerous, weed achenes and to determine che longevity of achenes with 2 to 5 very small heads borne in clusters on short, retained in the inflorescence and on the soil surface. We also panicled stalks arranged near the ends of upright stems. examined achene germination under different water appli- Broom snakeweed heads usually contain 2 to 7 ray flowers cations to determine soil water requirements for emergence with yellow corollas, and from 0 to 9 disk flowers (Lane and survival.

Wood et al.: Broom snakeweed . 77 Frcrln~1. Monthly (bar) and long-rime (line) averagc precipirarion, and minimum and maximum air remprarurer on rhe New Mexico Srare Univeriiry Corona Rescarch Ranch, NM. Time 0 is January 1993.

Materials and Methods Seed Dispersal Environmental Setting In midSeptember 1993, when broom snakeweed was in the early bloom stage, seed traps were installed around 6 This study was conducted on the New Mexico State Uni- similarly sized, healthy, mature plants in each exdosure. versity Corona Ranch about 25 km east of Corona, NM. Shallow trenches (I m long by 10 cm wide by 3.2 cm deep) Sample plants were selected from 2 study sites, each within were dug outwards from beneath the plant canopy in the 4 5-ha exclosures, located about 10 km from one another on cardinal directions. Collectors made of polyvinyl raingutter nearly level terrain at 1,870-m elevation. Rainfall is most were placed in the trenches so that the top edge was even common from July to September, and the 397-mm annual with the ground surface. To retain seed in the traps, an air average can range from 280 to 510 mm. foil was placed over each collector made of five 0.64-cm by Soils are comprised of the Taipa-Dean loam association, 91-cm wooden dowels secured to metal hardward cloth. In- which is shallow and underlain by a highly calcareous lime- side each collector, equal-sized (0.025 m2 surface area) re- stone bedrock. The Taipa loam is a fine-loamy mixed mesic movable compartments, also made of polyvinyl raingutter, Ustollic Haplargid. The Dean loam is a fine carbonate mesic divided collectors into 4 distance categories. Water drainage Ustollic Calciorrhid. Vegetation in both exdosures 1s com- was provided by using fine cloth netting at ends of the com- posed of a moderate stand of broom snakeweed (5 to 10 partments and by drilling several 0.5-mm holes in the col- plants rc2), with blue grama, wolftail (Lycurur phleoides lector bottom. H.B.K.), sand dropseed [Sporobolus qptandrur (Torr.)], Seeds were collected from the traps weekly or biweekly squirreltail [Elymur longifoliur (Smith) Gould], and several beginning in September 1993 until little flower material re- forb species in the interspaces. Winterfat [CeratoidPr Ianata mained on the plants the next summer, i.e., 1 flowering , (Pursh) Howell] and cholla [Opuntia in~bricara(Haw) DC.] season. The contents in each compartment were emptied are scattered throughout the area. into separate plastic bags and labeled by plant number, di- Weather in each exclosure was monitored throughout the rection, and distance. In the laboratory, flower and seed material study by automated stations1 recording air temperature, soil was separated from other debris using a No. 16 (0.1 168 cm) temperature, relative humidity, windspeed, wind direction and No. 50 (0.0297 cm) U.S.A. standard sieve and then and rainfall. Precipitation from January to September 1993 examined under a scope. Mature achenes rerained in the was about 30% below the long-term average and was near capitula were counted separately from rhose dispersed indi- or below normal every month unril December 1994 (Figure vidually. Lane (1985) reported that both ray and disk 1). The exception was in May 1994, when rainfall was 115 achenes were viable; however, disk achenes we examined were mm, or 392% above average. always sterile. Therefore, when we use the term 'achenes',

78 . Weed Science 45, January-February 1997 we are referring to ray achenes only. We nored signs of insect Laboratory-stored achenes were placed in 5- by 10-cm man- feeding on achenes, particularly during fall collections; thus, ila envelopes and maintained at room temperature (ca. 25 those obviously damaged by herbivory were counted sepa- C) before testing at 3, 6, 9, 12, and 18 mo (40 X 5 test). rately from undamaged achenes. Field-stored achenes were inserted into 5- by 5-cm mesh Dispersal data were first analyzed in a nested structure by nylon packets sealed with metal paper clips to keep out analysis of variance (SAS 1990), to compare achenes in traps herbivorous insects and returned to the Corona Ranch. The collected from the 2 study sites. Because there were no dif- packets were placed on the soil surface under a 1.75-un wire ferences between sites, data were combined and analyses run mesh cage until retrieved larer for testing after 3, 6, and 9 to compare seed numbers captured at different distances and mo (40 X 3 test) of storage. Because previous reports by direcrions from broonl snakeweed plants. When the analysis Mayeux and Leotta (1981) indicated broom snakeweed of variance indicated a significant difference, means were seeds are mosr likely to germinate on the soil surface rather compared by Fisher's Protected LSD test using the 5% prob- than at buried depths, seed was not buried. Laboratory and ability level. field viability data were analyzed separately as split-plot ex- periments, with site the main plot rreatment, and plants Seed Production and Retention within site the main plot sampling unit. Collection time by storage regime was the subplot treatment; seed packets were To determine seed production and retention within the rhe subplot sampling unit. Data were analyzed using SAS inflorescence, 6 randomly selected broom snakeweed planrs general linear model procedures (SAS 1990) to produce within each exclosure were harvested bi-weekly from Oc- analysis of variance tables and contrast rests ro compare field tober 23, 1993, to Augusr 10, 1994, and near mid-month and laboratory storage effects at the 5% level of probability. from October 20, 1994, to June 15, 1995. Plants were clipped carefully near the soil surface to avoid dislodging seed. Then each plant was bagged separately and transported Germination to rhe laboratory to be oven dried for 24 h at 50 C. The ~erminarionexperiments were conducted in a green- dried plants were hand threshed to loosen the achenes and house on the New Mexico State University campus in Las to remove the inflorescences from the stems. The remaining Cruces, NM. Soil obtained from the Corona Ranch was fine litter and capitula were subsequently pulsed twice in a sieved through a 5-mm screen and pasteurized for 24 h at seed scarifier to further loosen achenes. Final separarior~of 80 C. Plastic pots (16.5 cm dim) with filter paper placed achenes from chaff was made using 2 sieves (No. 7 clipper over the drainage holes contained 1,700 g of sieved soil. In screen and No. 120 seedburo) and a pneumatic seed cleanerz the 1st greenhouse experiment (April 2 to May 20, 1994), with the air blower set at 8.0 mm and turned on twice for laboratory-stored achenes collected in October 1993 were 10 s. The number of achenes in a 0.2 g subsample of the used. In a 2nd greenhouse experiment (November 10 to seed fraction was counted and extrapolated to estimate toral December 20, 1994), laboratory-stored achenes from both achene number per plant. When achene recovery from a October 1993 and October 1994 seed lots were used. Ger- plant was low (< 200 achenes per plant), an actual count mination results with 1993 seed were the same for both was made. Differences in achene number per plant over the greenhouse trials, so 1993 data were pooled for final anal- various collecrion dates were analyzed as a completely ran- yses. The experiment was analyzed as a completely random- domized design with site by sample date by plant as the ized design with a 2 (1993 and 1994 seed source) by 4 error term. Means were separated using Fisher's Protected (water amounts) by 4 (water intervals) factorial arrangement LSD rest at the 5% level of probability. of treatments. Pressure plate analysis (Rawls et al. 1982) conducted by the NMSU soils laboratory indicated that the Seed Viability water content was at field capacity (fc) when the soil nlarric Viability tests were conducted shortly after achenes were potential (Pm) was -30 Ma. Subsequent water amounts separated from rhe inflorescences (time 0) using tetrazolium were 314 fc (-71 kPa), 112 fc (-314 Ma), and 114 fc (TZ) analysis procedures similar to those described by Thill (-3967 kPa). Water inrervals were either 1 d, 5 d, 10 d, or er al. (1985). A random subsample of 40 achenes per plant 15 d. Six pots were randomly assigned to each treatment, (6 plants per site and sample date) were placed inside a 5-cm and 10 achenes from each year's seedlot were equally spaced p& dish on filter paper saturated with distilled water. Achenes on the soil surface with forceps. A wire was laid down the were allowed to imbibe for a minimum of 4 h; then, using middle of each pot to separate seedlots. Germinates were a dissecting scope, those containing an embryo were sepa- recorded daily; germination was considered successful when rated from those obviously without an embryo to determine cotyledons emerged. Pots were weighed before and after each the percentage of nondeteriorated achenes (achenes with water application and whenever a new germinant was noted. embryoltotal achenes X 100). Nondeteriorated achenes Using procedures described by Klute (1986), a moisture re- were dissected near the apical end, below the pappus, and lease curve was developed to determine Pm in each por placed in a 1% aqueous solution of TTC (2,3,5,-rriphenyl though time. A negative exponential regression analysis was tetrazolium chloride) for 8 h (Tetrazolium Conlmittee of used to predict achene germination (averaged over pots) ver- Association of Oflicial Seed Analysts 1970). Following the sus the Prn using the general form: soaking period, achenes with a red-stained embryo were used to calculate percentage ner viability (viable acheneltotal in TZ rest x 100). To examine seed longevity, the remain- The dependent variable (YJ was broom snakeweed germi- ing achenes from cach collecrion dare were stored eirher in nation. The independent variable (4 defines Pm a1 ger- the laboratory or in the field before testing by TZ analysis. mination. The parameter c defines the exponential rare at

Wood et al.: Broom snakeweed . 79 asoq~ro q2er2 [!os om! spjley) paas 'snqL .sasea13u! q~dap aql u!q~!m pasrads!p sauaq2v .a2urnsaropu! aql morj X1pn p!Inq se sasearmp uo!leu!urra% paama~usmoorq pallod -p!~!pu! paslads!p (%I 6) sden u! paraAoJar sauaqx isom -a1 (£861) xnaXem \ipunor2aql uo are spaas I~IJElrodsuer~ .a%po~s!poi sauaqx ralem pue pu!m apadur! Xem axjrns pos aql jo ssauuaAaun pasne2 smals pue sa?ualoUu! %uy!rls pq ro sdorpu~r30 prre rail![ 1ue1j .%u!Bpo1s!p I~IJEpuel Xaqi araqm mu uFm Xlpolan aql Xpuareddv '~661'~1 Xem uo %u!~durespapa2 -ar Xlqeqord sauaqx lsour 1(z861 I!!J :1861 ep!urqS pue -ard u!er asualu! ue ralje paraAoJa1 aram sauaqx $0 raqurnu ~au113)Xroq~alale s! a2uels!p psrads!p s'paamayeus uroorg q%!q Xla~!lelar v .(umoqs IOU elep) I-s m z.6 ueq~ssa1 aram '(6861) SmAg pue %U~OAXq palrodar sem lsea aql 01 spaads pu!m urnurpreur %u!pa?ard uaqm pue parmmo lua~a pamaqs %u!aq paas ysnrqaaes B!q q~!m uralled psrads!p nl! uo!lel!dpard ou uav sderl morj pal2alIo2 aram sauaqx -w!s v .isamqlnos aqi urorf spu!m %u!lear2 '1seo3jpl~)pue jo sraqmnu mol Xp~!lep~.sauaqx lelo~30 Xra~omr isa 3yped aql urorj u! aAom Xjp!dXl uor2a1 s!q~ u! smrols asne2 -q%!q aqr papa~ard(81 pue 6 raquraxa uaamaq m 1.91 -aq pal~adxasem ]sea aql spremo~sauaqx jo urailed psrad pm 61 pue 11 raqura~o~uaamaq ur ~ZIjo spaads -s!p aqL .(paluasard IOU eiep) ohqma padola~ape moql!m pu!~mnm!xem) spu!m q%!q Xq pa!uedmome (X~a~!~~adsar XIISOLU aram lueld e morj m> 5~ puoXaq paraAo2ar sauaq>v 'urur 1pue 01 1e 4~!1eplmm3e'MOUS XPSOUI)sluana uo!lei! .weld iuared aqljo m> 0s u!ql!m rnmo s%u!lpaas paamaqeus -dpard om^ .ralq pa%po[s!p an snerq prop lsu!e%e passard uroorq sour leql palrodar ow'(1861) .p la uewso q~!m sauaqx apqm is1 asrads!p 01 peaq aql jo ralua3 aql nau sauaq2e asne2 pue pu!m aql u! alerqp qp~s pro^ g!ls aqL saar%e umalled s!r[j, '(1 alqe~)u!21em Xdoue2 aqi reau 10 qleauaq Xp3anp [[aj sauaqx paamayeus moolq jo Xl!lo!em .uo!~el!dpard pue spu!m $!q se qms saxnngu! >!lem!p aql 'psrads!p aleulpej oi samlmrls paz!ppads InoquA 30 amasqe ro axasard aql pue armloAu! aql u! pau!elar Apn~saql inoq%norq~srol2a1lo2 aql u! pa~ouaram 'slue sauaqx arnlem jo raqurnu aql jo uo!l~~ye sem @srads!p se qms saroA!qraq %u!l1amp-puno% Inq '1pj aql u! sarnl jo ales aq.L .uoseas ixau aql spnq ramog mau jo luaurdo -eradura~mnm!u!w lapp X~%u!searxr!ql!m pappu!o2 Xro -lanap aql qr!m pappu!o2 X~q%nor~esrads!p jo uopessa3 .(z -~!qraq Xq pa%e.eurep sauaqx u! au!pap .paluerrem suraas arn%!j) palefialu!s!p Xlpuy sramop 01 paqxlle s~ooqspapp v am!) q3!qm le 'qm zp roj sder~u! pal3a[lo3 Xlpnu!iuo~ aram eare s!ql u! qxeasar inq 'Xpnis s!q~ u! paredm02 IOU sem sauaq3v y%aq 01 psrads!p pamol@ pue uasool 01 nxrq pun013 av uo oi pareduo2 a2uaxarogu! aql u! suo!~epard armlonu! morreu aql para%4!11 Xlqeqord srqr :(3 E'E-) paas u! axarag!a 'sralea paas pue lamo~se sl2asu! raylo Bu!warj mopq paddorp arnle~adma~r!e urnur!u!m a%era~e pue asaql paq!rxap (1861) .p la lalsoj .amg s!ql %u!mp aql '£661 'EZ raqo120 uo sderi aql u! sauaqx $0 Xra~om lueld aql q~eaurapunsluammdmo> uo!l2allo2 worj para IS[ av Bu!pa2ard qaam aqi %u!lna .(I amn%j) %u!zaarj reau -Aom uayo aram un~adup~dqo~~~~jo sapads pay!luap!un ~pjsarn~eradura~ r!e mnm!u!w uav 'raqouo Xlrea pun ue pm (sinus muoroywm sl!~aa~.(E arn%!j) Xro~!qraq Xq ~sn%nv-p!ururorj padola~apsramog aleu!mraiapu! 'mo[laX pa2emp pue arnleurur! se pa~ouaram sder~aql u! sauaqx s'paamayeus moorq jo Xl!ro!eur aql 'eare Xpnls rno u[ lsom uav 'uo!nallo3 jo qm p isry aql U! q2!q Xlppad -sa sem X~!A!I~ msu! 'rar\amoH ,parano2ar paas pi01 jo %L pamap Xron!qraq ~sasu~.paamayeus moorq u! X~ueur -lop paas alourord 01 ms!ueq2aur lmy!u%!s e s! '(1861) uo!ssnas!a pue salnsau elloq pue xnaXen Xq paisa%%nsse 'psradsp elnl!de~ ley1 Xlaq!~IOU s! I! 'snql !padop~apunpue arnieurur! X~pnsn .a%uerlue~alar aql 1noq2nolql u!%!ro aql oi X~AUOJs! aNn2 aram pue sl~erqXq paielnsde~uaXlploi Xlarer aram peaq a% 'uo!ieu!mra% paamaqeus woorq sassarddns m& qqqm

-a,nsopxa qxa u!q~!rnpamall03 sluald laqm 9 JOJ azep aldluer Xq ruaalu aql uo paseq an (IorIa pIepuelr ql!m qdr~2a"![) mumsa~ogu!2ql wolj pa~anom~sauarpv .po!~adlesrads!p 9661-~661 aql ~U!JR~SaInSOpxa z U! IIUP~~9 WOJ~ uo!na~!p pue a=uals!p ssome alep aldluen Xq $maw uo px~qare rden rr! sauaq3v .rueld lua~edalp wo~jruo!l=anp lau!ple, p u! ,no 8u!pualx:, pue Ldoue, aql qleaurapun pua 1 ql!~(mna p~epuejrqj!~ rqda12 ~uq)rd~n paar 8uol lu-l u! paxano3ar sauaq~paaMaFur mooq .Z moms FIGURE3. Herhiwry damage to broom snakewcid achenes recovered horn rraps during the 1993-1994 dispersal period covered by soil movement may be removed from the pool 1989). Besides a lack of soil water, flowering is negatively of potential progeny. We did not examine biological dis- influenced by old age, poor plant vigor, high shrub density, persal, but animals large enough to brush against a plant and insect feeding (McDaniel 1989). For example, Parker can feasibly atfect dispersal either by causing stems to eject (1985) nored rhc obligate feeding snakeweed grasshopper achenes by a whipping motion or by achene trichomes at- [H~pemtPniruiridir (Thomas)] completely defoliated threadleaf taching to the animal's coat and possibly transporting the snakeweed (Gutiemzin microcephafu) in localized areas, seed out of the area (Mayeux 1989). which resulted in no flowering or seed production in certain years. Seed Production Individual broom snakeweed plants are capable of pro- Seed Retention and Viability ducing thousands of seeds (Ragsdale 1969), but there is wn- The percentage of nondeteriorated achenes obtained from siderable variation in seed production among plants, pop- the inflorescence during the first G mo after dispersal began ulations, and years (Pieper and McDaniel 1989). Of 12 in October was nearly rwice as high for the 1993 seedlot as plants randomly sampled for seed production near the be- compared to the 1994 seedlot, or 74% versus 32%, respec- ginning of dispersal in October 1993, the number of tively (Table 2). The difference in percentage nondeterior- achenes counted ranged from 116 to 14,414 per plant and aced achenes between seedlots was probably related to the averaged 3,928 achenes per plant (+ 1,146) (dara not pre- relative vigor of broom snakeweed each year during flow- senred). A similar san~pleof 12 plants in Ocrober 1994 ering and seed maturation. During the summer of 1994, indicated a range from 25 to 8,921 achenes per plant and the highest average air temperatures on record occurred in a mean of 2,036 (+ 987). Broom snakeweed achene pro- Corona, which, coupled with low precipitation, resulted in duction may have been reduced in 1993 and 1994 in our less flower production than observed in 1993 (data not pre- study area because precipitation was below normal during sented). The differences in the quality of achenes produced flowering both years (Figure I). Under very dry conditions, during these 2 years was partially reflected in the relative broom snakeweed typically ceases to flower, whereas in wet- weights of nondeteriorated achenes, which were 41% higher ter years, the plant flowers profusely (Pieper and McDaniel (per 100 achenes) in October 1993 compared with October 1994. These data suggest that broom snakeweed achene de- TABLE1. Number of broom snakeweed achenes recovered from vclopmenr and the subsequent rate of deterioration may be seed traps located beneath the canopy edge ounvards to 1 m in 4 influenced by climatic conditions and inherent differences cardinal directions. Achenes were collected a1 2-wk intervals from berween seedlots. Ocrober 1993 rhrough August 1994. Each value in the cable is the Broom snakeweed achenes stored on the soiI surface in average using seed collections from 12 plants on heNew Mexico packets decayed little during fall and winter months (Oc- State University Corona Ranch. tober to March) for the 1993 and 1994 seedlots (Table 2). Achenen trapped by Achenes rrapped by Achenes retrieved from the packets after mid-May were of- cardinal direcrion disrance lion, plrnr ten seedless pericarps in various stages of decomposition, - -- Ear Norrh Sourh Wesr 0-25 25-50 50-75 75-100 regardless of storage time. The pericarp was often covered with fungi, but it is not known whether deterioration was No. of achenes per 0.025 m-2 a resulr of snakeweed seed physiology or the effects of ml- 11.2aa 1.7b 1.0b 0.7b 8.4a 3.9b 1.5bc 0.6~ crobial decay. The timing and rate of rapid deterioration of iMeans with the same lerrers by cardiud dircaion or by distance from achenes on the soil surface were similar to that observed planr do nor differ rignificanrly ar the 0.05 level. with achenes collected from the inflorescence both years (Ta-

Wood et al.: Broom snakeweed . 81 TABLE2. Percentage of nondereriarated achenes obtained directly from the inflorescences (time 0) and aFter storage on rhe roil surface for 3, 6, and 9 mo. Data are averaged from 12 collecred from 2 locations on the New Mexico Stare Universiry Corona Ranch bcainnina in October 1993 and 1994. 1993 reedlor 1994 seedlor Time in nlonrhs Time in rnonrhr Sample date 0 3 6 9 Sample dare 0 3 6 9 -

Ocr. 11 Ocr. 22 Ocr. 29 Nov. 11 Nov. 5 Nov. 23 Nov. 24 Dec. 9 Dec. 22 Dec. 17 Jan. 7 Ian. 22 Jan. 28 Feh. 5 Feb. 19 Feb. 23 Mar. 4 Mar. 18 Mar. 9 Mar. 31 Apt. 15 Apr. 14 Apr. 30 May 17 May 15 Jun. 1 Jun. 14 Jun. 15 Jun. 28 Jul. 13 Jul. 15 Jd.24 Aug. 10 LSD (0.05) ble 2). Mosr of the 1993 achenes (98%) deteriorared within for 1 to 2 yr after a parent plant population is eliminated. the inflorescence and on the soil surface between April 30 This agrees with observations that once broom snakeweed to June 15, whereas the majority of 1994 achenes (96%) is controlled by herbicide spraying and seedlings do not es- deteriorated between April 15 to May 15. These results in- tablish within 2 yr, treatment life is likely long-lived dicare that the majority of broom snakeweed achenes are (McDaniel and Duncan 1987). However, if even a small nor long-lived and that only a small proportion of the orig- proportion of the original parent population survives and inal number of seeds persist into the next production year. continues to produce seed, the weed population may even- Other shrub species in the family, such tually reestablish. as bie sagebrush (Artenriria h'dentata Nutt.) (Youne and vans 1%9), also' produced prolific but shot;-live$ seed crops thar rarely persist for more than a year. Germination When percentage viability data (TZ test) were analyxd, When percenrage germination data were analyzed, no in- no interactions or main effects attributable to study area teractions or main effects due to seedlots were observed. were observed, but initial net viability was higher for the Achenes watered to field capacity had the highest percentage 1993 seedlot (95%) compared with the 1994 seedlot (83%) (data not presented). Net viability of achenes collected from the inflorescence or stored on the soil surface was similar TABLE3. The influence of warer amounr (fraction of field capaciry) through time when compared by recovery date within a and water inrenal (summed over seedlots and resting dares) on sample year (October to August) (Figure 4). During both percenraae eerminarion of broom snakeweed achenes. study years, achenes exhibited a pronounced periodicity of viability, as seeds were always most viable when recovered Fraction of field upziry during fall and winter (October to April) but declincd sharply late in spring (May to July). A similar periodicity of Water inrervd Ill 314 112 114 viability was not observed for achenes stored in the labora- Day - % tory for 2 yr, which agrees with Mayeux and Leotta (1981), 1 52a 38b 6d 0 who reported that viability of broom snakeweed achenes 5 35b 18c 1 0 remain unchanged after 4 yr of laboratory storage. 10 23c 3d 0 0 Results from the viability study suggest that it may be 15 23c 4d 1 0 possible to severely reduce broom snakeweed from rangeland 'Means followed by rhe rme lerrers are not significanrly different ar the with conventional control methods, provided the environ- 5% lewl of probabilicy according ra Fisher's Prorerrd LSD rest. Where mental conditions necessary for germination do not occur letters do nor appear, no significant ditferences wrrc found.

82 . Weed Science 45, January-February 1997 Surface Laboratory Inflorescence Storage Storage

Oct Jan Apr Jul Oct Jan Apr Jul

F~cua~4. Percenrage ner viability of broom snakeweed achener recovered from inflorescences and aher storage on rhe soil rurfacc and in rhe laboratory during rhr 1993-1994 and 1994-1995 dispersal periods. of germination (33%) when averaged over seedlots, trial tervals, and amounts, 72% of the variability in percentage dates, and water intervals (Tablc 3). No achenes germinated germination was attributed to variation in soil 'Pm. As ex- when water was applied at 114 fc, and few achenes germi- pected, the estimated curve exhibits a rapid decline in ger- nated when pots were watered to 112 fc. A higher percentage mination as soils become drier The curve is steepest near 0 of achenes germinated (24%) when watered daily than when when germination is highest and flattens when germination watered at 5-, lo-, or 15-d intervals, regardless of water declines below about 10%. Averaged across all experiments, amount. Germinarion generally decreased proportionally as 91°h of achenes germinated when soil 'Pm remained > water interval lengthened and water amount decreased. -180 kPa for at least 4 d (data not shown). Achenes watered daily to 111 fc had the highest percentage The majority of propagules survived the greenhouse ex- of germination (52%). Achenes watered daily to 314 Fc had periments in the daily 111 fc (76%), the daily 314 fc (58%), the same percentage of germination as achenes watered at and the 5-d 111 fc treatments (64%). Few survived under 5-d intervals to 111 fc (average 37%). Percentage of ger- the other water amounts (data not presented). Seedlings that mination was not different for achenes watered to 314 fc at died under rhe daily 111 fc regime died mostly from damp- 5-d intervals and achenes watered to 111 fc at 10-d and ing off fungus. This may partially explain why low-lying 15-d intervals (average 21%). areas that periodically retain standing water in snakeweed Broom snakeweed percentage of germination was nega- communities are usually void of the shrub. Seedling mor- tively and inversely related to the soil 'Pm at germination talicy in treatments other than when watered daily 111 fc (Figure 5). When data were combined over years, water in- occurred when soil 'Pm averaged 1853(5 364) kPa. Seed-

> 0 .- 0 200 400 600 800 1.000 1,200 Soil Water Potential (-kPa) FIGURE5. Acrual (symbolj and (line) percenrage gcrminacion of broom snakeweed under changing soil manic potentials. Predicred gerrninarion values were determined wing Equarion 1 (0= 1164.14 - 0.90 r~'..4 = 0.72, and r = 30)~

Wood et al.: Broom snakeweed . 83 ling mortality under drying soil conditions should be ex- Mcrhodr of Soil Analysis. Part 1, 2nd ed., Volurnr 9. Madison, WI: pected to vary depending on the age and condition of the Sail Scicncc Society of America, pp. 635-660. fiuse, W. H. 1970. Temperacure and moisrure rrress affecr germinarion of seedling at the time of stress. Gurierezin rnrothrnr. J. Range Manage. 23: 14h144. These results suggest that optimal broom snakeweed ger- Lane, M. A. 1985. of Gutiemziiz (Composirae: Astereae) in minarion occurs when water is applied frequently and sat- Norrh America. Syrt. But. 10:7-28. urates the soil. Field observations of broom snakeweed ger- Mayeux, H. S. 1983. Effects of sail rexrure and seed placement on emer- gence of four rubshrubs Weed Sci. 31:380-384. mination from 1990 rhrough 1995 on rhe Corona Ranch Mayeux, H. S. 1989. Snakeweed seed characteristics and germination rr- closely correspond with these findings (Carroll 1994). Dur- quiremenrs. In E. W Huddlesron and R. D. Pieper edr. Snakrweed: ing this 5-yr period, precipitation during the 2nd quarter Pmblpmr and Perspecrivei. N. M. Srace Univ. Agrk. Exp. Sm. Bull. (April through June) was near or below normal every year 751:3949. Mayeu*, H. S. and L. Leotta. 1981. Gprminarion of broom snakrwrcd and except 1992 and 1994, when rainfall was above normal and chreadleat snakeweed seed. Weed Sri. 31:380-384. 63% of broom snakeweed seedlings eounted during this McDanicl, K. C. 1989. Snakeweed popularions in New Menico, 1979- study emerged (Carroll 1994). Under field conditions, low 19RL). In E. W. Huddlesron and R. D. Pieper, eds. Snakeweed: Pmb- daily precipitation (< 25 mm) for 1 or several days is prob- lems and Perspeccks. N. M. Stare Univ Agric: Exp. Stn. Bull. 751: .*12-75 -,. ably insufficient to result in successful germination. We sus- McDaniel, K. C. and K. W. Duncan. 1987. Broom snakeweed (Gurirrrrzio pect a higher amounr of precipitation (> 25 mm) that keeps mrorhrae) conrrol wirh picloram and menulturon. Weed Sci. 355337- the soil surface wet for several days is necessary for germi- 841. nation. Besides soil water, condirions conducive to optimal Osman, A,, R D. Plepcr, and K. C. McDaniel. 1787. Sod sced banks associated wirh individual broom snakeweed planrs. 1. Ranae Manaw. germination include seed placement near the soil surface " - 40:44143. (Mayeux 1983) and an alternating air temperature regime Parkrr. M. A. 1985. Sizr-dependanr herbivore attack and the demography between 10 to 20 C (Kruse 1970). For mass germination of an arid grasslaud shrub. Ecology 66:850-860. on rangeland, it appears these conditions in central New Pieper, R. D. and K. C. McDaniel. 1989. Ecology and management of Mexico must culminate before mid-May, after which time broom snaknveed. In E. W. Huddlrsron and R. D. Picper. eds. Snake- weed: Problems and Perspecriver. N. M. State Univ. Agric. Exp. Sm. seed viability declines dramatically The optimal culmination Rflll.-.. . 751.1-17, .. . .-. of these events elsewhere must be adjusted to localized con- Pijl, 1. van der. 1982. Principles otdirperral in higher plants. 3rd ed. Nrw ditions. These generalizations need to be tested over a wide York: Springer-Verlag. 215 p. range of environmental conditions where broom snakeweed Ragsdale, B. I. 1969. Ecological and phenological characrerisricr of peren- nial broomwccd. Ph.D. disserrarion. A&M University, College is a dominant species. Srarion. TX. 142 p. Rawls, W I., D. L. Braken~iek,and K. E. Sarron. 1982. Esrimarion ofsoil waicr .propcrries. . Transactions of the American Socicw of Agricultural Sources of Materials Engineers. 1316-1320. ISASI Srarisrical Analysis Systems. 1990. SAS Ploiedurer Guidr. Veriiorz 6 ' Automated Stations. Cam~bellScientific. lnc.. 815 W 1800 3rd ed. Cary, NC: SAS Insrirure. N, Logan. UT 84321-1784. Terrazolium Commirrre of Association of Official Seed Analysrs. 1970. In Pneumatic seed cleaner, E. L. Erickson Products, Brookines.- D. F. Grabe, ed. T~rrmliurnTesring, Handbook for Agricultural Seed SD. No. 29, p. 62. Thill, D. C., D. L. Zamora, and D. L. Kambitsch. 1985. Germination and viability of common crupina (Cnqinn mrlgarir) achenrs buripd in the field. Weed Sci. 33344-348. Literature Cited Tordl. L. A.. H. W. Gorrlon, K. C. McDaniel, and A. McGinty. 1988. Economic impacts of perennial snakeweed inferrarionr. ht I.. F James, Carpenter, B. D., D. E. Erhridge, aud R. E. Sorrbee. 1'390. Eonomiclosser M. H. Ralph$, and D. B. Nidsen, eds. The Ecology and Economic from broom snakewed poisoning in carrle. Rangelands 12:206-208. lmpacr of Poisonous Planrs on Livrstock Producrion. Boulder. CO: Carroll, D. B. 1994. Broom snakweed [Gutierezin rnrorhrne (Pursh) Brirr. Wercview Press, pp. 57-69. & Rrr~bylseedling response co spring and summer burning in cenrral Torell, L. A.. K. Williams, and K. C. McDaniel. 1989. Probability of snake- New Mexico. M.S. rheri,. New Mexico Srare Univerriry, Las Cruces, weed die-of and invasion on rangeland. In E. W. Huddlesron and R. NM. 99 p. D. Pieper, eds. Snakeweed: Problems and Perspecrives. N. M. Sratr Dirtbernrr, P L. 1971. A demographic scudy of some remidererr Univ. Agrir. Exp. Stn. Bull. 751:'153. plants. M.S. rheris. New Mexico State University, Lar Cruces, NM. Wangberg, J. K. 1982. Dcsrructive and porenrially desrrucrive insects of 67 p. snakeweed in wrsrcrn Tenas and eartern New Morio and a diorisric Ulnrr, S. and A. Shmida. 1981. Why arc adaptations for long-range reed model of their biaric inreracrions. J. Range Manage. 35235-238. dispcrral rare in desert plann! Oceologia 51:133-144. Young, J. A. and R. A. Evans 1989. Dispersal and grrminarion of big Fosrrr, D. E., D. N. Ueckert, and C. J. DeLoach. 1981, lnreccr associared sagebrush (Arrem;sia nidmtar) seeds. Weed Sci. 37:ZOl-206. with brt~om snakeweed (X?nrhocephizlurn rarorhme) and threadlcaf snakeweed (Xanrhocrphnlr'rn mi,.roi?phizio) in west Texas and easrcrn Nomenclature: Broom snakeweed, Gutiemzin iarothrae (Pursh) New Mexico. I. Range Manage. 34:446454. Britt. & Rusby GUESA; Blue grama, Bolrttloua grmilir (H.B.K.) Jamaon, A. D. 1970. Valuc of broom snakeweed as a range condirion Lag. A- Griffiths. indicator J. Rangc Manage. 23:302-304. Klure, A. 1986. Warcr recenrion: laboratory merhods. hr A. Klute, ed. RereiuedJanuary 3L 1996 and approved September 24, 1996

. 84 . Weed Science 45, January-February 1997