characteristics of the psidium cattleianum (myrtaceae) seed bank in hawaiian lowland wet forests ...

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors nonprofit publishers academic institutions researchlibraries and research funders in the common goal of maximizing access to critical research

Characteristics of the Psidium cattleianum (Myrtaceae) Seed Bank in HawaiianLowland Wet ForestsAuthor(s) Amanda L Uowolo and Julie S DenslowSource Pacific Science 62(1)129-135 2008Published By University of Hawaii PressDOI httpdxdoiorg1029841534-6188(2008)62[129COTPCM]20CO2URL httpwwwbiooneorgdoifull1029841534-618828200829625B1293ACOTPCM5D20CO3B2

BioOne (wwwbiooneorg) is a nonprofit online aggregation of core research in the biological ecological andenvironmental sciences BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies associations museums institutions and presses

Your use of this PDF the BioOne Web site and all posted and associated content indicates your acceptance ofBioOnersquos Terms of Use available at wwwbiooneorgpageterms_of_use

Usage of BioOne content is strictly limited to personal educational and non-commercial use Commercial inquiriesor rights and permissions requests should be directed to the individual publisher as copyright holder

129

Characteristics of the Psidium cattleianum (Myrtaceae) Seed Bankin Hawaiian Lowland Wet Forests1

Amanda L Uowolo23 and Julie S Denslow3

Abstract Psidium cattleianum Sabine (strawberry guava) is one of Hawailsquoirsquos mostdisruptive alien plants Dense stands can suppress growth and establishment ofnative species support high populations of crop-damaging fruit flies and pre-clude restoration or management of native forests Our research investigatedfactors affecting persistence of P cattleianum seeds in lowland wet forest soilsWe collected soil cores from four forested sites immediately after fruit fall and65 months later We found abundant germination of P cattleianum seeds imme-diately after fruit drop Soil collected under mature P cattleianum clumpsyielded 761 viable seedsm2 We found no viable seeds 65 months after fruitdrop We evaluated seed longevity using seed bags buried below the litter layerthat we retrieved after 28 56 196 and 365 days Seeds either germinated or de-teriorated rapidly after fruit drop after 28 days 223 of the buried seeds wereviable and there were no viable seeds at 196 days Predator effects were assessedusing trays with a known number of seeds with and without predator exclosuresAfter 28 days 37 of the seeds in the open trays were damaged by predatorsThe lack of a persistent seed bank likely is due to a combination of rapid highgermination rates postdispersal seed predation and seed mortality We suggestthat chemical or mechanical control efforts would be most efficient and effectiveif conducted at least 3 months after the fruiting season when the vast majorityof seeds have either germinated or died

Psidium cattleianum Sabine (Myrtaceae)also known as strawberry guava is consideredone of Hawailsquoirsquos most severe invasive weeds(Smith 1985 Wagner et al 1999 Motookaet al 2003) Psidium cattleianum is commonon all the major Hawaiian islands andthroughout the Pacific where it continues toincrease in density and impact Rapid growthrates ( JSD and ALU unpubl data)prolific fruit production and high rates ofsuckering produce dense stands quickly fromsmall populations (Huenneke and Vitousek1990) Psidium cattleianum is a serious threatto native forest ecosystems in Hawailsquoi and

elsewhere in the Pacific due to its ability toinvade even relatively undisturbed wet forestsand form dense thickets (Smith 1985 Huen-neke and Vitousek 1990) The fruits of P cat-tleianum support high populations of fruitflies (eg Dacus dorsalis Hendel) that are con-sidered one of the most important agricul-tural pests in Hawailsquoi (Vargas et al 1990)The efficacy of any control method for P cat-tleianum is influenced by the characteristics ofits seed bank because a persistent seed bankcontributes to high population growth ratesand reduces the likelihood of local eradica-tion We investigated longevity of seeds ofPsidium cattleianum in Hawaiian soils thecharacteristics of the seed bank and factorsaffecting the duration of seeds in lowlandwet forest soils in Hawailsquoi

materials and methods

Species Description

Psidium cattleianum has invaded native Hawai-ian ecosystems ranging from moist to very

Pacific Science (2008) vol 62 no 1129ndash135Work of the US GovernmentNot under copyright

1 Manuscript accepted 26 April 20072 Corresponding author (e-mail auowolofsfedus)3 Institute of Pacific Islands Forestry US Depart-

ment of Agriculture Forest Service 60 Nowelo StreetHilo Hawailsquoi 96720

wet conditions from 15 to 1300 m elevation( Jacobi and Warshauer 1992 Wagner et al1999) All forms of P cattleianum grow asshrubs or small trees typically 2ndash6 m tall inmesic to wet forests in Hawailsquoi (Wagner etal 1999) We used seeds from the commonround yellow-fruited form (P cattleianum flucidum Degener) in this study Fruits contain15ndash70 angular seeds (Diong 1982 Huennekeand Vitousek 1990) of 25ndash5 mm in length(Morton 1987 Wagner et al 1999) Peakfruit fall typically occurs between June andDecember (Diong 1982 Huenneke andVitousek 1990) Psidium cattleianum seeds arereadily spread by humans and feral pigs (Susscrofa) (Diong 1982 Jacobi and Warshauer1992) as well as by a variety of birds (Me-deiros 2004)

Study Area

Study sites were established within the UpperWaiakea Forest Reserve (19 35 0 N 155 12 0

W ) the Pulsquou Makalsquoala Natural Area Reserve(19 34 0 N 155 12 0 W ) the lsquoOlalsquoa ForestReserve (19 27 0 N 155 11 0 W ) and the Ka-haualelsquoa Natural Area Reserve (19 26 0 N155 10 0 W ) on the island of Hawailsquoi Allsites are on windward Hawailsquoi Island at ap-proximately 900 m elevation Estimated an-nual rainfall is 3000ndash4000 mm at lsquoOlalsquoa andKahaualelsquoa and 4000ndash5000 mm at UpperWaiakea and Pulsquou Makalsquoala (Giambelluca etal 1986) Projected mean annual temperaturebased on adiabatic lapse rates is 17ndash175Cfor the elevation range of the four study sites(Giambelluca and Schroeder 1998) All of thestudy sites are on relatively young tholeiiticbasalt lava flows that formed 200ndash1500 yrBP (Wolfe and Morris 1996) The soils atUpper Waiakea and Pulsquou Makalsquoala are deepand well drained (Typic Udifolists US De-partment of Agriculture Natural ResourceConservation Service [USDA-NRCS] un-publ data) over weathered lsquoalsquoa lava The soilsat Kahaualelsquoa (Lithic Hapludands USDA-NRCS unpubl data) and lsquoOlalsquoa (Lithic Udi-folists USDA-NRCS unpubl data) are veryshallow to shallow moderately well-drainedsoils formed in volcanic ash deposited overpahoehoe lava bedrock The forests are classi-fied as native lowland wet forests with

an lsquoohilsquoa (Metrosideros polymorpha Gaud) low-land wet forest community type (Gagne andCuddihy 1999)

Seed Longevity

Seed longevity of P cattleianum in soil was as-sessed by tracking the fates of known num-bers of seeds placed in bags and insertedunder the litter at each of the four study sitesFully ripened and intact P cattleianum fruitswere collected from and beneath trees in Oc-tober and pooled from multiple P cattleianumgroups across lsquoOlalsquoa Forest Reserve Seedswere immediately cleaned air dried andstored for 1 week in an air-conditioned envi-ronment Percentage seed viability was as-sessed on a subsample of the pooled seeds(10 replicates of 50 seeds each) 1 week afterthe seeds were collected and on the same daythe seed bags were placed in the field At eachstudy site one set of four seed bags wasplaced under each of five groups of large Pcattleianum trees Each sealed bag (6 by 75cm 86-mesh silkscreen fabric) contained 20seeds and 5 g of sieved autoclaved soil Thebags were placed beneath the litter layer(generally 25ndash5 cm deep) in contact withthe soil surface and were placed in the fieldduring the fruiting season in late OctoberOne seed bag from each of the five sets ofbags placed at each site was retrieved after28 56 196 and 365 days When the buriedbags were retrieved the seeds were evaluatedand seeds that had germinated or decayedwere separated from those that were intactIntact seeds were tested for viability by plant-ing in sterile growing medium in a greenhouseunder irrigation and approximately 53 sunGermination of seeds was monitored over a16-week period

The buried bags provided informationon changes in the numbers of viable persis-tent seeds and seed decay and senescenceover time We modeled the decline in num-bers of viable seeds as an exponential decayfunction (SigmaPlot 802 [SPSS Inc 1986ndash2001])

y frac14 y0ekt

in which y is the number of the original seedsremaining and persisting as viable seeds in the

130 PACIFIC SCIENCE January 2008

soil at time t y0 is the number of seeds attime 0 t is the elapsed time in days and k isthe decay constant The mean residence time(MRT) is calculated

MRT frac14 1k

The seed bags excluded mammalian and manyinvertebrate predators from access to theseeds but did not prevent infestation by soilfungi therefore k and mean residence timerepresent longevity of the seeds in the soil inthe absence of predators

Seed Bank

We collected four soil cores (500 cm3 to 5cm deep) from beneath each of five clustersof mature P cattleianum trees from each ofthe four study sites on each collection dateWe selected P cattleianum tree clusters withcanopies exposed to full sun to maximize theprobability of sampling where trees had gen-erated dense seed shadows Soils were col-lected twice immediately after fruit fall (lateNovember) and approximately 65 monthsafter fruit fall (early June) The soil fromeach core was spread to a depth of 1 cm oversterile medium and placed in the greenhouseunder irrigation and approximately 53 sunPsidium cattleianum germinants were countedand removed weekly over a 16-week period

Seed Predation Rates

We compared rates of seed damage to pro-tected and unprotected seeds spread on traysplaced in the forest Twenty trays (43 by 43by 5 cm deep) containing 50 P cattleianumseeds each were placed at each of the fourstudy sites half of these trays were coveredwith hardware cloth (635-mm mesh) to ex-clude large predators and half were leftopen Open and protected trays were placedadjacent to one another but 10 m away fromthe canopy of P cattleianum trees to avoid ad-ditional seed input from fruit fall The seedsplaced in the trays were from the same com-posite sample of seeds used in the buried seedbags and were cleaned air dried and storedfor 1 month in an air-conditioned environ-ment until the seeds were placed in the fieldForest floor litter material was placed in the

trays with the seeds and the trays were linedwith 2 mm nylon mesh to prevent washoutTrays were placed in the field in late Novem-ber at the end of the fruiting season at allstudy sites After 28 days the trays were col-lected seeds and litter material were sepa-rated and seeds were counted and evaluatedfor evidence of predation We assumed thatsecondary seed dispersal did not occur andby extension that missing seeds had beenconsumed and not dispersed as viable seedsSeeds with evidence of animal activity (ieteeth marks seed fragmentation) were con-sidered damaged and were assumed to benonviable as well Data from one subse-quently overturned tray were excluded

Data Analysis and Synthesis

We used the data from the seed longevity andseed predation experiments to develop a pro-jection of the fate of P cattleianum seeds 65months after fruit fall The baseline germina-tion trial was used to estimate the percentageof nonviable seeds in the seed rain Data fromthe buried seed bags provided information onpersistent viable seeds in the seed bankSeeds recovered intact in the seed bags wereconsidered persistent or viable if they subse-quently germinated in the greenhouse Seedsthat did not germinate in the seed bags whilein the field nor in the greenhouse when re-moved from the bags were considered to havedied The exclosure experiment providedthe rate of predation we assumed that allpredation occurred during the 28-day periodfollowing seed dispersal We assumed thatnongerminating seeds from the seed bags in-cluded those that were nonviable in the seedrain and that a certain percentage of seedsthat subsequently germinated in the seed bagsin the field would have been lost to mamma-lian and invertebrate predators in the firstmonth

results

Seed Longevity

Results from the buried seed bags showedthat the decrease in viable seeds with timefollowed an exponential decay curve ethy frac148641e00463t r2 frac14 099 P frac14 lt0001THORN (Figure

Psidium cattleianum Seed Bank Uowolo and Denslow 131

1) Mean residence time for the seeds was216 days In the absence of predation 95of the seeds had either germinated in the fieldor lost viability in 66 days 99 of the seedswere nonviable after 96 days

Seed Bank

High numbers of viable P cattleianum seedswere found in soil collected during fruit fallat our sites (761 viable seedsm2 n frac14 4 sitesSE frac14 309) yet no viable seeds were foundfrom soil collected 65 months after peakfruiting season Results from the soil samplescollected 65 months after fruit drop wereconsistent with results from the seed bags re-trieved after 65 months in neither case wasthere evidence of a persistent seed bank


A mean of 369 (n frac14 4 sites SE frac14 71) ofthe seeds in open trays were missing or dam-aged by predators after 28 days in the fieldThe amount of seeds missing or damaged bypredators ranged from 0 to 90 in the opentrays Less than 01 of seeds (n frac14 4 sitesSE frac14 005) were missing from the protectedtrays after the same time period

Seed Fates

Figure 2 summarizes the fates of P cattleia-num seeds in lowland wet forests 65 monthsafter dispersal At 196 days 28 seeds out of100 placed in the bags were dead when recov-ered having failed to germinate when placed

Figure 1 Mean number of P cattleianum seeds (per 100 seeds) germinating after being buried in seed bags Baselineviability was assessed when the bags were buried (day 0) and bags were collected after 28 56 196 and 365 days Errorbars represent the standard error among four study sites There were no viable seeds in the bags collected after 196 and365 days


in the greenhouse We assumed that 13 ofthose seeds had been nonviable in the seedrain and estimated therefore that 15 seedshad died after being placed in the field Theremaining 72 seeds germinated in the fieldbefore the bags were recovered We assumedthat 369 or 27 of those would have beenlost to predation in the first month followingdispersal leaving an estimated 45 seeds ger-minating in the field Dead and nonviableseeds also lost to predation (369 of 13 non-viable and 15 dead seeds) are not shown inFigure 2 We estimate that most P cattleia-num seeds were lost from the seed bank dueto predation and germination with smallerpercentages nonviable from pre- or post-dispersal processes We found no evidenceof viable seeds 65 months after fruit fall

discussion

Seeds of tropical rain forests may have theshortest ecological longevity of seeds ofany plant community (Vasquez-Yanes and

Orozco-Segovia 1993) The absence ofpersistent seed banks among trees in tropicalsystems has been attributed to predation(Thompson 1987) and to simultaneous andrapid germination of the seeds immediatelyafter dispersal (Vasquez-Yanes and Orozco-Segovia 1993) or in response to favorableephemeral conditions (Skoglund 1992) Al-though many seeds of tropical and subtropi-cal rain forest trees are desiccation tolerant(orthodox) and thus have the potential toform persistent seed banks in the soil(Vasquez-Yanes and Orozco-Segovia 1996Tweedle et al 2003 Dalling 2005) mosttropical rain forest plants do not do so (Hop-kins and Graham 1987 Thompson 1987Skoglund 1992 Vasquez-Yanes and Orozco-Segovia 1993) Psidium cattleianum with or-thodox seeds (Becwar et al 1983) appears tobe typical of most tropical trees Desiccationtolerance in the laboratory is not necessarilycorrelated with the persistence of the seedsin the field (Tweedle et al 2003) becauseseeds in the natural environment are subjectto predation invertebrate and microbial in-fection and conditions conducive to germi-nation Germination and predation appearto account for the majority of the loss of Pcattleianum seeds from soils of Hawaiianlowland wet forests

Although our study did not identify theseed predator(s) damage was typical of thatby rodents Black rats (Rattus rattus) housemice (Mus musculus) and Polynesian rats(Rattus exulans) are known seed predators inHawaiian forests and are an important con-straint on reproduction of native plants (To-mich 1986 Cuddihy and Stone 1990 Cabinet al 2000) Rodents have been shown to beeffective predators of exotic plant seeds inHawaiian forests as well (Medeiros 2004)and P cattleianum seeds may be a readilyavailable food source for alien rodents Preda-tion rates are likely to be spatially variable aprevious study found no evidence of P cat-tleianum seeds in rodent droppings at anothersite (Medeiros 2004)

Results from the buried seed bags suggestthat even in the absence of predation P cat-tleianum seeds do not survive burial in naturalconditions for an extended period of time

Figure 2 Projected fates of P cattleianum seeds 65months after dispersal into lowland Hawaiian forest soils


The persistence of seeds has been shown toincrease with an increased depth of burial(Toole and Brown 1946 Lonsdale et al1988) and we might have found P cattleia-num seeds to have been viable longer if ourseeds had been buried more deeply Despitethe many combinations of variables that mayhave affected the outcome of the buried bagexperiment the lack of any viable seeds inthe soil samples we collected 65 months afterfruit drop suggests that seeds do not naturallypersist from one fruiting season to the next inthe soil environment In two other studieswhere soil cores were collected from forestswith mature P cattleianum plants present via-ble P cattleianum seeds were not encounteredeither (Huenneke and Vitousek 1990 Drake1998)

An understanding of local P cattleianumphenology and the characteristics of the seedbank is of value in the selection and schedul-ing of control methods for this invasive plantBecause most P cattleianum seeds do not livebeyond 3 months in the soil chemical ormechanical control efforts would be mostefficient and effective if conducted at least3 months after the fruiting season Thebiological control agent Tectococcus ovatus iscurrently proposed for the control of P cat-tleianum in Hawailsquoi and has been documentedto reduce fruit and seed production of plantsin Brazil ( Johnson 2005) Our results suggestthat a biocontrol agent that reduced P cat-tleianum seed production would also rapidlydeplete soil seed stocks therefore increasingeffectiveness of chemical and mechanicalcontrol

acknowledgments

We thank the Division of Forestry and Wild-life of the State of Hawailsquoi for access to fieldsites We are grateful to Cheyenne PerryRoddy Nagata Wendell Sato and Kim Buz-dygon for their assistance in the field green-house andor laboratory

Literature Cited

Becwar M R P S Stanwood and K WLeonhardt 1983 Dehydration effects

on freezing characteristics and survival inliquid nitrogen of desiccation tolerantand desiccation-sensitive seeds J AmSoc Hortic Sci 108613ndash618

Cabin R J S G Weller D H LorenceT W Flynn A K Sakai D Sandquistand L J Hadway 2000 Effects of long-term ungulate exclusion and recent alienspecies control on the preservation andrestoration of a Hawaiian tropical dryforest Conserv Biol 14439ndash453

Cuddihy L W and C P Stone 1990Alteration of native Hawaiian vegetationEffects of humans their activities and in-troductions University of Hawailsquoi Coop-erative National Parks Resources StudiesUnit University of Hawailsquoi Press Hono-lulu

Dalling J W 2005 The fate of seed banksFactors influencing seed survival for light-demanding species in moist tropical for-ests Pages 31ndash44 in P M Forget J ELambert P E Hulme and S B VanderWall eds Seed fate Predation dispersaland seedling establishment CABI Publish-ing Cambridge United Kingdom

Diong C H 1982 Population biology andmanagement of the feral pig (Sus scrofa) inKipahulu Valley Maui PhD diss Uni-versity of Hawailsquoi at Manoa Honolulu

Drake D R 1998 Relationships among theseed rain seed bank and vegetation of aHawaiian forest J Veg Sci 9103ndash112

Gagne W C and L W Cuddihy 1999Vegetation Pages 45ndash114 in W L Wag-ner D R Herbst and S H SohmerManual of the flowering plants of HawailsquoiBishop Museum Special Publication 87University of Hawailsquoi Press and BishopMuseum Press Honolulu

Giambelluca T W M A Nullet and T ASchroeder 1986 Rainfall atlas of HawailsquoiR 76 Department of Land and NaturalResources Division of Water and LandDevelopment State of Hawailsquoi Honolulu

Giambelluca T W and T A Schroeder1998 Climate Pages 49ndash59 in S P Juvikand J O Juvik eds Atlas of Hawailsquoi 3rded University of Hawailsquoi Press Honolulu

Hopkins M S and A W Graham 1987The viability of seeds of rainforest species


after experimental soil burials under tropi-cal wet lowland forest in north-easternAustralia Aust J Ecol 1297ndash108

Huenneke L F and P M Vitousek 1990Seedling and clonal recruitment of theinvasive tree Psidium cattleianum Implica-tions for management of native Hawaiianforests Biol Conserv 53199ndash211

Jacobi J D and F R Warshauer 1992 Dis-tribution of six alien plant species in up-land habitats on the island of HawailsquoiPages 155ndash188 in C P Stone C WSmith and T Tunnison eds Alien plantinvasions in native ecosystems of HawailsquoiUniversity of Hawailsquoi Press Honolulu

Johnson M T 2005 Petition for fieldrelease of Tectococcus ovatus (HomopteraEriococcidae) for classical biological con-trol of strawberry guava Psidium cattleia-num Sabine (Myrtaceae) in Hawailsquoihttpwwwfsfeduspswtopicsecosystemprocessestropicalinvasive (accessed 11January 2007)

Lonsdale W M L S Harley and J D Gil-lett 1988 Seed bank dynamics in Mimosapigra an invasive shrub J Appl Ecol25963ndash976

Medeiros A C 2004 Phenology reproduc-tive potential seed dispersal and predationand seedling establishment of three inva-sive plant species in a Hawaiian rainforestPhD diss University of Hawailsquoi atManoa Honolulu

Morton J 1987 Cattley guava Pages 363ndash364 in Fruits of warm climates Julia FMorton Miami Florida

Motooka P L Castro D Nelson G Nagaiand L Ching 2003 Weeds of Hawailsquoirsquospastures and natural areas College ofTropical Agriculture and Human Re-sources University of Hawailsquoi at ManoaHonolulu

Skoglund J 1992 The role of seed banks invegetation dynamics and restoration of drytropical ecosystems J Veg Sci 3357ndash360

Smith C W 1985 Impact of alien plants on

Hawailsquoirsquos native biota Pages 180ndash250 inC P Stone and J M Scott eds Hawailsquoirsquosterrestrial ecosystems Preservation andmanagement University of Hawailsquoi Coop-erative National Parks Resources StudiesUnit University of Hawailsquoi Press Hono-lulu

SPSS Inc 1986ndash2001 SigmaPlot version802 for Windows Point Richmond Cali-fornia

Thompson K 1987 Seeds and seed banksNew Phytol 10623ndash34

Tomich P Q 1986 Mammals in Hawailsquoi2nd ed Bishop Museum Honolulu

Toole E H and E Brown 1946 Final re-sults of the Duvel buried seed experimentJ Agric Res 72201ndash210

Tweedle J C J B Dickie C C Baskinand J M Baskin 2003 Ecological aspectsof seed desiccation sensitivity J Ecol91294ndash304

Vargas R I J D Stark and T Nishida1990 Population dynamics habitat prefer-ence and seasonal distribution patterns ofOriental fruit fly and melon fly (DipteraTephritidae) in an agricultural area Envi-ron Entomol 191820ndash1828

Vasquez-Yanes C and A Orozco-Segovia1993 Patterns of seed longevity and ger-mination in the tropical rainforest AnnuRev Ecol Syst 2469ndash87

mdashmdashmdash 1996 Comparative longevity ofseeds of five tropical rain forest woodyspecies stored under different moistureconditions Can J Bot 741635ndash1639

Wagner W L D R Herbst and S HSohmer 1999 Manual of the floweringplants of Hawailsquoi Bishop Museum SpecialPublication 87 University of HawailsquoiPress and Bishop Museum Press Hono-lulu

Wolfe E W and J Morris 1996 Geologicmap of the island of Hawailsquoi Map 1-2524A US Department of the InteriorUS Geologic Survey Miscellaneous In-vestigations Series


129

Characteristics of the Psidium cattleianum (Myrtaceae) Seed Bankin Hawaiian Lowland Wet Forests1

Amanda L Uowolo23 and Julie S Denslow3

Abstract Psidium cattleianum Sabine (strawberry guava) is one of Hawailsquoirsquos mostdisruptive alien plants Dense stands can suppress growth and establishment ofnative species support high populations of crop-damaging fruit flies and pre-clude restoration or management of native forests Our research investigatedfactors affecting persistence of P cattleianum seeds in lowland wet forest soilsWe collected soil cores from four forested sites immediately after fruit fall and65 months later We found abundant germination of P cattleianum seeds imme-diately after fruit drop Soil collected under mature P cattleianum clumpsyielded 761 viable seedsm2 We found no viable seeds 65 months after fruitdrop We evaluated seed longevity using seed bags buried below the litter layerthat we retrieved after 28 56 196 and 365 days Seeds either germinated or de-teriorated rapidly after fruit drop after 28 days 223 of the buried seeds wereviable and there were no viable seeds at 196 days Predator effects were assessedusing trays with a known number of seeds with and without predator exclosuresAfter 28 days 37 of the seeds in the open trays were damaged by predatorsThe lack of a persistent seed bank likely is due to a combination of rapid highgermination rates postdispersal seed predation and seed mortality We suggestthat chemical or mechanical control efforts would be most efficient and effectiveif conducted at least 3 months after the fruiting season when the vast majorityof seeds have either germinated or died

Psidium cattleianum Sabine (Myrtaceae)also known as strawberry guava is consideredone of Hawailsquoirsquos most severe invasive weeds(Smith 1985 Wagner et al 1999 Motookaet al 2003) Psidium cattleianum is commonon all the major Hawaiian islands andthroughout the Pacific where it continues toincrease in density and impact Rapid growthrates ( JSD and ALU unpubl data)prolific fruit production and high rates ofsuckering produce dense stands quickly fromsmall populations (Huenneke and Vitousek1990) Psidium cattleianum is a serious threatto native forest ecosystems in Hawailsquoi and

elsewhere in the Pacific due to its ability toinvade even relatively undisturbed wet forestsand form dense thickets (Smith 1985 Huen-neke and Vitousek 1990) The fruits of P cat-tleianum support high populations of fruitflies (eg Dacus dorsalis Hendel) that are con-sidered one of the most important agricul-tural pests in Hawailsquoi (Vargas et al 1990)The efficacy of any control method for P cat-tleianum is influenced by the characteristics ofits seed bank because a persistent seed bankcontributes to high population growth ratesand reduces the likelihood of local eradica-tion We investigated longevity of seeds ofPsidium cattleianum in Hawaiian soils thecharacteristics of the seed bank and factorsaffecting the duration of seeds in lowlandwet forest soils in Hawailsquoi

materials and methods

Species Description

Psidium cattleianum has invaded native Hawai-ian ecosystems ranging from moist to very

Pacific Science (2008) vol 62 no 1129ndash135Work of the US GovernmentNot under copyright

1 Manuscript accepted 26 April 20072 Corresponding author (e-mail auowolofsfedus)3 Institute of Pacific Islands Forestry US Depart-

ment of Agriculture Forest Service 60 Nowelo StreetHilo Hawailsquoi 96720


Study Area




Seed Longevity



y frac14 y0ekt




MRT frac14 1k


Seed Bank







results

Seed Longevity




Seed Bank




Seed Fates





discussion









acknowledgments


Literature Cited




































Study Area




Seed Longevity



y frac14 y0ekt




MRT frac14 1k


Seed Bank







results

Seed Longevity




Seed Bank




Seed Fates





discussion









acknowledgments


Literature Cited




































MRT frac14 1k


Seed Bank







results

Seed Longevity




Seed Bank




Seed Fates





discussion









acknowledgments


Literature Cited




































Seed Bank




Seed Fates





discussion









acknowledgments


Literature Cited




































discussion









acknowledgments


Literature Cited





































acknowledgments


Literature Cited



































characteristics of the psidium cattleianum (myrtaceae) seed bank in hawaiian lowland wet forests ...

Documents