Studies on the Growth of Red Mangrove (Rhizophora Mangle L.) 3

Home , Rhizophora mangle

Studies on the Growth of Red Mangrove (Rhizophora mangle L.) 3. Phenology of the Shoot Author(s): A. M. Gill and P. B. Tomlinson Source: Biotropica, Vol. 3, No. 2 (Dec., 1971), pp. 109-124 Published by: The Association for Tropical Biology and Conservation Stable URL: http://www.jstor.org/stable/2989815 . Accessed: 30/08/2011 15:40

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

The Association for Tropical Biology and Conservation is collaborating with JSTOR to digitize, preserve and extend access to Biotropica.

http://www.jstor.org Studies on the Growthof Red Mangrove(Rhizophora mangle L.) 3. Phenologyof the Shoot'

A. M. Gill2and P. B. Tomlinson3 FairchildTropical Garden,Miami, Florida 33156 U.S.A.

ABSTRACT

An analysisof shoot extensiongrowth of Rhizophoramangle is presented,based on the continuousobservation of a small population over a period of two years. This provides phenological informationfor a tree which fits the general category of "evergrowing"although in South Florida the climate is distinctlyseasonal. Environmenthas a stronggoverning influence, but the general progressionof events is mediated endogenously. Growth is morphologically non-articulate, without obvious flushes, so that the unit of growth is regarded as the single internode. The apical bud has a uni- form constructionthroughout the year, consistingof three leaf primordia and associated stipules, the expansion of the oldest pair of leaf primordia coinciding with the initiationof a new pair. Consequentlyenvironmental control of growth is effectedthrough developmental processes in the apical bud. The rate of leaf expansion and leaf fall is at a peak in summer when temperaturesand radiation levels are highest. Leaf fall is closelycorrelated with leaf expansion so that thereis a relativelyconstant number of leaves on a shoot. However, there is some evidence that propagule abscission also affectsleaf fall. All lateral organs (vegetative branches,inflorescences, and suppressed axillary buds) are developed within the apical bud, the types of organ differentiatedat any one node apparently depending on plastochron length. This is most evident with regard to precocious (sylleptic) branches which are produced periodicallyin groups of from one to three. Suppressed axillary buds are very reduced and subject to apical dominance. When released from this. by pruning for example, they can reach a size which permitsthem to grow out as normal apical buds. Under continued apical dominance the life span of their potential for furthergrowth is limited. Rhizophoramangle is, in a general sense, "everflowering"but expansion of flower buds is most frequentin summer,leading to a peak of fruitproduc- tion about 4-7 months later and the largest crop of propagules 4-6 months later still. These observationsare sum- marized in the form of a model which indicates how a single reproductiveshoot might appear during this cycle, and how the population, as a whole, behaves. A hypotheticalmodel of interrelationshipbetween phenological phenomena is also presentedas a basis for furtherwork. The models serve to illustratehow shoot growth occurs in this tropical species, and how this is in marked contrastto the more familiarpatterns of shoot growth in temperatetrees.

THE MOST DIVERSE florasof the world occur in Twig morphologyof northtemperate trees is the tropicswhere trees are usuallythe dominantlife characterizedby segmentsdelimited by seriesof bud form.Despite this the studyof the growthand scale scars. These articulationsoccur at the bases developmentof tropicaltrees has been neglected of branchesand at intervalsalong themin such a (Tomlinsonand Gill 1971) and presenttheories of waythat they reflect the seasonal cessation of growth tree behaviorare based almost entirelyon a few imposedby the severewinter climate. In contrastto relativelywell-studied north temperate examples. In- this,Rhizophora shows no obvious articulationin vestigationsof a wider range of tropicalspecies the shoot system,an observationwhich led Koriba shouldbe made in orderto extendour knowledge (1958) to place the Singaporespecies R. conjugata and add perspectiveto currenttheories of tree and R. macronatain his "evergrowing"category. growth.This presentstudy of shoot growthwas SouthFlorida has a seasonalclimate, in contrastto undertakenwith this ideal in mind using a wide- the uniformclimate of Singapore,and one might spreadtropical species Rhizophora mangle L. which not expect the term"evergrowing" to apply to a extendsinto subtropical Florida. The studycontinues speciesin a -non-uniformclimate. The abilityof an the seriesof papersdealing with various aspects of "evergrowing"species to flourishin a seasonalcli- the growthand anatomyof this species (Gill and mateis investigatedin thispresent article. Tomlinson1969, 1971). A numberof recentstudies of northtemperate treeshave relatedthe typeof extensiongrowth of 1 Contributionto a study of the growth of native woody plants of South Florida supportedby a grant to Dr. P. B. the tree to the compositionof its terminalbuds Tomlinson fromthe Michaux Fund of the American Philo- (e.g., Critchfield1960, Gill 1971). A few similar sophical Society. Diagrams are the work of Priscilla Faw- studieshave attempted to do thisfor individual tropi- cett,Botanical Illustrator,Fairchild Tropical Garden. ical tree species (Bond 1942, 1945) but not one 2 Present address: Division of Plant Industry,C.S.I.R.O., P.O. Box 109, Canberra A.C.T. Australia 2601. of thesestudies has involveda specieswhich has a non-articulatemorphology and is, therefore,apparent- 31Presentaddress: Harvard Forest,Petersham, Massachusetts 01366 U.S.A. ly "evergrowing."Known examplesof growthtypes

BIOTROPICA 3(2): 109-124 1971 109 in treesas a wholehave been consideredin a recent are shownin figure1. In winter,maximum temper- reviewby Tomlinsonand Gill (1971) in a very aturesmay be high but minimumtemperatures can generalway where the suggestionis made thatthere fall close to freezing.Rainfall is spoiradicin winter are unlimitedways in which growthcan be ex- and long periodswithout rain are common.In con- pressedin tropicaltrees. trast,rain in summeris frequentand heavy.The An earlierpaper in this series on Rhizophora annualaverage is 58 in (153 cm). (Gill and Tomlinson1969) describedand illustrated Soils in the mangrovezone alwaysremain wet the generalmorphology of R. mwngleand thispro- and salinityvaries accordingto tide, rainfall,and vides a backgroundfor the presentstudy in which evaporation.The plants studiedwere 0.5-4 m tall extensiongrowth and reproductiveevents are con- and occurredin an area of mineralsoil inundated sideredin relationto the seasonal changesin the by brackishwater. environment.Radial growthhas not been investi- gated despiteits possiblerelationship to otherphe- All reproductivestages may be observedat any nologicalevents. time of the year. Consequentlyobservations were begun at differenttimes on threedifferent sets of MATERIAL AND METHODS shootsshowing a varietyof reproductivestages. One of these sets (shoots 22-41) was chosen because The source of materialwas a populationof red of the apparenthigh vigor of its shootsand their mangrovein the vicinityof FairchildTropical Gar- abundantbranching. Measurements were begun on den, Miami, Floridaon the edge of BiscayneBay. shootsof set one (1-21) in November1968; on Climatically,the area is characterizedby a warm shootsof set two (22-41) in December1968; on humidsummer and a relativelycool drywinter. Both shootsof set three(42-58) in March 1969. Obser- maximumand minimumtemperatures are high in vationson all shootscontinued until December 1970. summerwhen insolationreaches its highestvalue. The 58 shootsmeasured belonged to a total of 22 Representativevalues for the period of observation plants. .100?

707 'AGOo- 60~~~~GO

W600- _ ~ ~ ~ ~ ~ ~ ~ ~ P ~~~~~~~~~~~~~~~~J50~ ~ ~ 5;

Nt NX9NAWIJAS ONDFANB'TA5OXO D 1968 1991970

FIGURE 1. Changes in radiation levels and in average maximumand minimumtemperatures in South Florida (Miami InternationalAirport) during the observationalperiod. Data fromNational Climatic Center,Asheville, North Carolina.

110 Gill and Tomlinson On each occasionthat measurements were taken tions; the heightof the bar representsan average a diagramof each shoot was made depictingthe rate of some process (e.g., leaf production)for a position of each leaf, flower,branch, fruit, and 30-dayperiod. For thisreason only the height of each viviparousseedling. Measurements were made at bar is significantfor comparativepurposes. approximatemonthly intervals. By examiningspe- Survivalof suppressedaxillary buds is limited cificshoots at intervalsin thisway data relatingto as has been demonstratedby Gill and Tomlinson leaf production,leaf fall, branch and flowerap- (1969) so theirlongevity and fatewere determined pearancecould be assembledtogether with records experimentally.In one set of experimentsbuds were of germinationand seedlingdevelopment on thetree releasedfrom apical dominanceby pruningbranches itself.(It mustbe recalledthat Rhizophoora is vivi- back as faras thenode supportingthe pair of oldest parous.) All this informationprovided a picture leaves so that,essentially, the leaf-bearingzone of of the seasonalcycle of growthand reproductionof the shoot was removed.In anotherset of experi- the species in South Florida.The period of obser- mentsolder leaflessaxes were also prunedback to vation was sufficientlylong to enable the whole stumpdiameters of 1 cm, 2.5 cm, and 5 cm. Prun- reproductivecycle of about 16 months(from floral ing was done on 2 3 July 1970, and the various initiationto propagulefall) to be followedon in- treatmentswere replicatedfive times. dividualshoots. The progressionof eventsfollowed in detail in this way provideda complete,quanti- tativepicture of what could be observedonly in a APICAL BUD AND UNIT OF GROWTH generalway for much larger populations. The apical bud of Rhizophorais remarkablein For a numberof reasonsobservations could not thatleaves, flowers, and precociousbranches all grow be made at preciseintervals. Therefore, data have withinit. Suppressedaxillary buds are also formed been plottedas rates in the form of histograms but do not expand.Various combinationsof these (figs. 2-21) so that any seasonal change can be axillaryorgans can occur at the same node. The detected.The widthof each bar in thesehistograms bud is enclosedby a pair of interpetiolarstipules representsthe intervalbetween successive observa- belongingto the youngestpair of expandedleaves.

560

t3O 36 l1 Q20t0 l0 ii' ~~~~~~010 Akal 16ukc( Vieb3w3b *a P06L*11CW*nfi-oM outsicse

FIGURE 2. Apical bud structure. The plan view on the left shows the positions of stipular primordia (lb-3b) by solid lines, those of the leaf primordia (2a-4a) by stippled arcs, and those of axillarystructures by hatched circles. The small, crossed circles outside the bud indicate the positionsof the youngestexpanded leaves. On the right,the lengths of the leaves and stipules are shown for a bud which has just begun to open, i.e., the outermostleaf (2a) is unfolding. Lengths of stipular primordia are shown as solid black bars; those of leaf primordiaas stippled bars. For illustration of actual bud see Gill and Tomlinson (1969).

Studies on the Growth of Red Mangrove 3. 111 fi0

0 ND F ZA O AY K Ti 968 1969 1970

>15

10~ ~ ~ ~ ~~~~~~17

110 19691-970

112 Gill and Tomlinson The stipulesinitially underwrap and overlap each congested(e.g., 12 nodes in a 5 cm lengthof othertightly but, as the bud grows,they become shoot formedin 1-2 years) so bthatmeasurements looser and eventuallyfall to reveal axillarystruc- of increasein lengthto estimategrowth rates are tures.The next pair of oppositeexpanding leaves meaningless.This is why in this studythe appear- also appearswhich, in turn,has interpetiolarstipules ance of leaves has been used as the basic unit surroundingan apical bud. The bud of Rhizophora to measurerates of growthrather than stemelon- is unusualbecause any one pair of stipulesencloses gation. not only youngernodes but also the lateralorgans Differentiationof lateralorgans, leading to the (flowers,vegetative branches, or suppressedbuds) productionof suppressedaxillary buds, inflor- at the node of insertionof the subtendingleaf pair escences,and branches,appears to beginin the axils (see illustrationsin Tomlinsonand Gill 1969, fig. of the thirdpair of leaves fromthe apex since this 8). Thus, althougha leaf pair may be expanded, is theyoungest pair of leavesin whichlateral organs lateralorgans in its axils remainenclosed by the can be recognizedby dissection(fig. 2). No de- stipulesfor a furtherplastochron. tailed attemptwas made to studypossible earlier The total contentof the apical bud was deter- stages;Rhizophora buds are difficultto microtome minedby dissectionunder the binocularmicroscope. but we hope to approachthe problemin the future. Examinationof budsat equivalentstages of develop- Althoughthe bud has only threepairs of leaves, ment,i.e., with a new pair of leaves just beginning two plastochronsmust pass before the structures to appear above the stipules,was made several in the axil of this thirdpair of leaves becomeex- timesduring the year (February,June, October, and posed. One plastochronis necessaryfor the expan- December). Bud structurewas verysimilar on each sion of its subtendingleaves but, as the axial struc- occasion.Buds varied from 3-7 cm long but ir- tures are still enclosedby the associatedstipules, respectiveof their size there were always three anotherplastochron must pass beforethe subtended pairs of leaf and associatedstipule primordia. This structuresare revealedby abscissionof thesestipules. arrangementis representedin figure2 where the The kindsof axillarystructure developed at any parts above the youngestexpanded leaf pair are one nodeare correlatedwith shoot vigor as measured shown all schematically, axillarystructures being by averageplastochron over the periodsof investi- shown a as singlehatched circle. The relativelengths gation.This is demonstratedby the figurescited in for successive organs in a representativebud are table 1. Plastochronlength is measuredsimply by shownto the rightin figure2 in the formof a histogram.This gives some indicationof relative TABLE 1. Average plastochronin relationto frequencyof growthrates. Bud burstis due to the growthof axillary structuresdeveloped over period of outerleaf primordiaexceeding that of the enclosing observation. stipuleswhich are graduallypushed apartuntil they Percent Percent absciseand fall.In any bud about to lose one pair Leaves Axils Total axils withvegetative Shoots of leaves (by expansion)and the associatedpair of produced formed branchesabranches branches sampled stipules (by abscission),new primordiabegin to Shoots 1-2 1 replacethe old even as the outermostleaves expand and the stipulesfall. In this way bud complement 14-16 136 3 2.2 0 9 18 108 14 12.9 0 6 is restored and maintainedthroughout the plasto- 20-26 136 36 26.5 2.8 6 chroncycle. Thus, the basic intervalof time in the extensiongrowth of this plant is the plastochron Shoots 22-41 itselfand the appearanceof successiveleaves can 18-22 164 16 9.8 62.5 8 be used as a measureof growth.An extended"flush" 24 144 24 16.6 62.5 6 involvinga close seriesof expandingleaves cannot 26-30 170 53 31.2 86.8 6 be recognizedhere. If the term "flush"is to be Shoots 42-58 used it only involvesone leaf pair and one inter- 16-20 168 50 29.8 4.0 9 node. We thus have a type of growthin marked 22-32 198 92 46.4 10.8 8 contrastto that of northtemperate woody plants. a Vegetativeplus reproductive. Internodesin manyshoots in Rhizophoraare often

FIGURES 3-5. The rate of leaf productionexpressed as the numberof leaves producedper 30-day intervalfor the period of time shown. Measurementsbegan at differenttimes as indicated by the positions of the firstbar on each histogram. Each histogramrepresents a differentset of shoots. Figure 3 representsshoots 1-21. -Figure 4 representsshoots 22- 41.-Figure 5 representsshoots 42-58.

Studies on the Growth of Red Mangrove 3. 113 thetotal number of leavesproduced during the peri- numbersof leaves per shoot stay constantwithin od of observation;the more leaves produced,the fairlyprecise limits. Some overalltrends which pro- shorterthe average plastochron. With a progressivelyduce changes in total leaf numberare discussed increasingvigor, measured by shorterplastochrons, later with regard to propaguleloss. there is an average tendencyamong a group of shootsto changethe type of lateralstructure develop- PRECOCIOUS (SYLLEPTIC) BRANCHING.-Under ed at thenode. With shortplastochrons the tendency naturalcircumstances branches of Rhizophoraform is to develop a vegetativebranch; with somewhat withinthe apical bud and have no basal prophyllar longer plastochronsthe tendencyis to develop a scalesor othermodified basal leaves.Since theyare reproductivebranch; whereas, with the longestplas- formedprecociously and have these morphological tochronthe tendencyis to develop a suppressed peculiaritiesthey may be termed"sylleptic" accord- ing to the axillarybud alone. The situationis complicatedby terminologyproposed by Tomlinsonand Gill the frequentassociation of a suppressedbud (a sup- (1971). ernumerarybud) with both vegetativeand repro- Vegetativebranches can appear at any time of the ductivebranches. However, this processwill only yearbut mostappear in summerwhen growth is be betterunderstood when the timesof initiation mostactive. Measurements which demonstrate this of axillarystructures can be pinpointedmore pre- are shownin figure15 whichrepresents the most ciselyby anatomicalmethods and the activitywithin prolificset of shoots (22-41). Branchescan occur the bud can be more accuratelyassessed. This will eitheron axes withcongested internodes or on those be thesubject of a futurestudy. with long interrnodes.On erectshoots with longer internodesgroups of branchesusually form but these groupsnever have morethan three component pairs VEGETATIVE GROWTH of branchesas describedelsewhere (Gill and Tom- LEAF APPEARANCE,LONGEVITY, AND FALL.-The linson 1969). These groupsform over a period of seasonalityof leaf appearanceis apparentin the severalmonths and can develop duringany season set of data for each populationof shoots (figs. 3, although,as with formationof individualbranches, 4, 5). There is a low in productivityduring winter theirgeneral appearance seems to be morefrequent and a high in summer.This demonstratesthat the in summer.On distal parts of horizontalshoots plastochronvaries in lengthfrom short in summer branchingis of the 'Terminalia-type"illustrated by to long in winter.No more detailedobservations Gill and Tomlinson(1969). Branchingin thispart were made regardingthis but the data do suggest of the shootsystem has not been studiedin further a minimumplastochron of the orderof threeweeks detail. and a maximumof at least 3-4 months.During AXILLARY BUDS.-Suppressed buds in the axils the longerplastochrons it is difficultto say, from of newlyexposed leaves are small (0.5-1.5 mm in externalexamination, whether the shoot is "grow- height) and simple,consisting of a pair of prophyl- ing" (by primordialgrowth) or not. lar scales, a pair of stipules,and a pair of leaf Leaf age varies widely.Some leaves live only primordia.The prophyllsdie earlythrough desicca- a fewmonths while others survive up to 17 months. tionleaving only the stipular-enclosedstructure. The In general,however, leaf age is of the order of stipulesare fleshyand forma canoe-shapedstructure 6-12 months,the average apparentlyvarying ac- which envelops the remainderof the bud. Bark cordingto the time of initiationand perhapsthe formedover the bud eventuallyobscures it. The vigorof the shoot.Data whichshow this are plotted longevityof these buds was determinedby the for the threesets of shootsin figures6-8. Leaves pruningexperiments described under materialand formedearly in the year tend to fall duringthe methods. period of peak leaf fall in summer,while those In viable buds the firstobserved response to formedduring the period of peak productionalso pruningwas swellingbeneath the bark in nodal in summertend to persistfor a longerinterval. positions.This was followedby splittingof the The species is evergreenand thereare usually barkand the appearanceof the bud. As growthcon- about 4-5 pairs of leaves on each shoot with the tinued outer primordiaaborted. Many of these lowermostpair beginningto senesce.Leaf fall does aborted primordiaresembled stipules rather than not proceedat a continuousrate but again has a foliage leaves. When the buds reach a heightof markedlyseasonal trend as is shownin the data for about 5 mm they have the same componentsas the threesets of shootswhich make up figures12, apical buds of regularshoots but in miniature.On 13, and 14. Since leaf fall is mostrapid in summer, expansionof theirfirst pair of leaves these buds correspondingto a maximumof leaf production, have the same structureas, but are less thana tenth

114 Gill and Tomlinson of the lengthof, apical buds at the same stage. In the experimentin whichshoots were pruned back for varyingdistances on July23, 1970 bud a swell was evidentafter six weeks. Leaf expansion 1* had begun afteranother month. Regrowth did not occur in branchescut back furtherthan 2.5 cm, at least not within six monthsafter pruning. It o ~ ~ t seems unlikelythat growthcould still occur after this lapse of time.Where onlythe leafyzone of a shootwas removedor wherethe prunedaxes were 0 less than 1 cm in diameterbuds on the distalpart 12 0i200tZL-ZL 2G, ".A32Ox of the remainingstump could grow out. It 'was determinedthat these axes on whichoutgrowth of buds could occurwere up to about threeyears old and the buds were up to nine nodes behind the originalterminal bud. The conclusion,therefore, is thaton axes morethan 1 cm in diameterand more than about three years old suppressedbuds are 'Tine of cxj. nzvomn. incapableof growth. is~~~~~~~~~~~" 1015i hok 22-tt e REPRODUCTION 7 FLOWERING (figs. 16, 19) .-The most vigorously vegetativegroup of shoots (21-41) had the weak- est reproductivebehavior and this group is not consideredfurther. Attention is thus focused on the remainingtwo groupsof shoots(1-21 and 42- 58). These were tagged in November 1968 and in March1969, respectively. Information with refer- ence to them is presentedin figures16-18 and figures19-21, respectively.At both times,shoots with a wide range of reproductivestructures were chosenwith bias towardshoots with flowerbuds Time of expansZio. or developingpropagules. Flowering abundance in both groupsconfirms the peak readilyobserved in 1Z summerin the whole populationbut in shoots 42-58 sporadicpropagule-fall (fig. 21) in the first fj yearis not verytypical. This unusualsituation did 20 not persistinto the second cycle of flowerswhen timingof reproductionharmonized with that evi- dentin the populationat large.The resultsfor the set of shoots1-20 (fig. 18) are,therefore, probably moretypical. Althoughflowers may be foundat any time of the year in the populationat large, quantitative measurementreveals that flower bud productionwas mostabundant in May-July(figs. 16, 19). Flowers open 1-2 monthsafter flower buds appear so that Ttmnof . the time of most abundantflowering is July-Sep- exrao*u tember. Petals and anthersare lost withina few days FIGURES 6-8. The average leaf age for each of the three sets of shoots shown as a functionof the time of leaf ex- of floweropening and fertilizationseems to take pansion. Figure 6 representsshoots 1-21.-Figure 7 repre- place duringthis period. In general,the percentage sents shoots 22-41.-Figure 8 representsshoots 42-58. is low. Values for The values within each bar refer to the total number of of flowerswhich are fertilized leaves fromwhich the average was computed. two groupsover two seasonswere 5.7, 7.2, 6.8, and

Studies on the Growth of Red Mangrove 3. 115 20

200 16?

190~~~~~~~~~~~~ r)

~i8o 170 IC

t50

1-4~~~~~~~~~~~~~~~~~~~~~~1

60 *0~1 t

190 2t2t to

16070 It3n.SO .10i 1701D 0 96 17 >1608

116 Gill and Tomlinson O percent.These valuesare incorporatedinto figures scissedalso variesand may be affectedby the en- 16 and 19 (percentagevalues to leftof peaks). vironment.This statementis based on detailedmeas- FRUIT AND HYPOCOTYL GROWTH (figs. 17, 20). urementswhich are not consideredrelevant to the -By comparingpeak timesfor flower bud appear- presentarticle but which will be publishedlater ance, blossoming,hypocotyl emergence from the in a discussionof the biologyof Rhizophoradis- fruit,and propaguleloss, the averageintervals be- persal. Despite the variationthe major period of tweenthese variousevents for a numberof small propaguleloss mostconsistently occurs around Sep- sampleshave been ascertained.Seasonal variations tember(e.g., fig. 18; the data in fig.21 maynot be have been revealedwhich are, to some extent,com- verytypical). pensatorybut this will be dealt with later when The dispersiveunit is the viviparousseedling, someaspects of thegrowth of thefruit and hypocotyl consistinglargely of its hypocotyl.When it falls have been discussed. fromthe treethe seedlingis usuallybetween 8 and Fruitstake fromfour to seven monthsto "ma- 24 cm in length,of which90 percentis hypocotyl. ture"after fertilization, where maturity of the fruit On a weightbasis the hypocotylaryproportion of is measuredby thefirst appearance of thehypocotyl. the propaguleis much largersince the epicotylis It mustbe rememberedthat the seedlingdevelop- rathernarrow. ment is viviparousand the stage selectedis arbi- traryand representsa maximumvalue. At maturity,DISCUSSION fruitsare usually about 2.5 cm long and weigh This studyhas demonstratedthat in the seasonal about 1 gm dry. climate of South Florida phenological changes The firstexternal evidence of germinationis the throughoutthe yeardo occurin a treespecies which appearanceof the hypocotylat the distalend of the has no obviousperiodicity of growthprocesses since fruit.Internally, germination, as revealedby elon- thereare no morphologicalarticulations in the shoot gationof theembryo, may be observedapproximate- system.Rhizophora may be describedas everflower- ly six weeks beforethis, at least as recordedin ingin thesense that any stage of reproductionmay be Fiji by Guppy (1906). In Florida,germination is observedat any timeof the yearbut, quantitatively, apparentlymost common early in the yearas hypo- markedfluctuations do occur. A model depicting cotyl protrusionreaches a peak frequencyaround some of these changesin a single shoot (below) March (e.g., fig. 17). As notedfor otherreproduc- and in the populationsat large (above) is shown tive events,however, this is only a trendand the as figure22. Stronginternal control of manyphen- ruptureof the fruitby the emerginghypocotyl may ologicalevents is evidentbut thiscontrol is modified be seen at any timeof the year.Such seasonalvari- by environmentin such a way thatshoots are out ation enablesan assessmentof the effectof climate of phase with each otherand only trendscan be on fruitdevelopment to be made. In brief,shorter recognized.These interrelationshipsare depictedin periodsof development(e.g., fourmonths) are evi- the formof a flow chartin figure23. The more dentin summer;longer periods are usual in winter doubtfulinterrelationships are indicatedby dotted (e.g., seven months). lines. From theseaccumulated data we can analyze The growthof the propagule,from first appear- growthprocesses in Rhizophorain a hypothetical ance to abscission,takes fromfour to six months. way. As withfruit maturation, shorter times of develop- Leaf production(shown in figs. 3-5) is most mentare associatedwith the higherradiation levels commonin summerwhen radiation and temperature and sustainedhigh temperaturesof the summer levels are highest.The internalcontrol mechanism months,and longertimes are associatedwith the is such that the outermostpair of leaves of the lowerradiation levels of winter.Early growth of the apical bud expandswhenever three pairs of foliar propagulesis morestrongly affected by climatethan and stipularprimordia are present.Bud composition are laterstages of development.A pronouncedad- appearsto be underendogenous control. The season- verse effectof unfavorableweather is thus most al trendin leaf expansionmay thus be viewed as noticeableon youngseedlings. a responseto environmentsimply through changes The lengthof timebefore the propaguleis ab- in the rate of productionof primordiaalone and

FIGURES 9-11. The change in total number of leaves presenton each set of shoots is shown by a solid line. A dotted line in figures 10 and 11 representsthe long-termtrend of change in leaf number.The height of the bars indicatesthe number of propagules lost from the shoots over the same time period. This propagule loss is expressed as the number lost per 30-day interval. Figure 9 representsshoots 1-21.-Figure 10 representsshoots 22-41.-Figure 11 represents shoots 42-58.

Studies on the Growth of Red Mangrove 3. 117 30 I -21

0 f0 j r 1t5 1968 1969 1970

W0 .

$33Lsl fi;E!t j hl 5~~~~~5ioou zt?- I 10

ss Figurt1 0 .

7'-s

o1968 9917

0- 19GB ~ ~~19ro 91970

FIGUR-ES 12-14. The rate of leaf fall expressedas the number of leaves falling per 30-day intervalfor the period of time shown. Measurementsfor each set begin with the firstbar. Figure 12 representsshoots 1-21.-Figure 13 represents shoots 22-41.-Figure 14 representsshoots 42-58.

118 Gill and Tomlinson Zfwats ezt+I

196S 1969 1970

FIGURE 15. Branch productionin the set of shoots 22-41 expressedas the number of branchesproduced per 30-day interval during the period December 1968 to December 1970. withoutinvolving any bud dormancywhich would pressedaxillary buds borne singlyin the leaf axil allow a buildupof leaveswithin a bud. Consequent- are moreusually formed in winterwhen leaf pro- ly, whenbuds are disposedto elongate(by having ductionis at a low level. The situationdescribed sufficientprimordia), they do so. Therefore,by the does not implya simple substitutionat the same definitionof Doorenbos (1953) thereis no dor- site of one type of lateralorgan for anotherbe- mancyof apical buds. This is the rationalebehind cause the node is complexin that it oftenbears the use of the term"suppressed" rather than "dor- morethan one typeof structure.In particular,sup- mant"to describelateral buds whichdo not expand pressedbuds are usuallyassociated with branches immediatelythey are formedwithin a terminalbud and flowers(see above). The correlationssuggested (Romberger1963). Expansionof leaves may be in figure23 are made on the basis of a population placed under partialenvironmental control simply sample ratherthan individualshoots. Clearly the by changesin the rateat whichassimilates are pro- situationin any one shootis complicatedby growth duced in expandedleaves. Similarly,growing prop- correlationssince many factorscontrol leaf pro- agules and senescingleaves may also affectapical ductionas figure23 shows.Some of thesecorrela- vigor,serving as additionalsinks (propagules) or tionsmay be discussedfurther. sources (old leaves) of nutrients,respectively. All When branchesare formed,feedback through theserelationships are shownby the arrowspointing assimilatedemand may decreaseapical vigorto the towards"leaf production" in figure23. extentthat branchesare no longerproduced. Evi- The level of apical vigor,as measuredby rate dencefor this is thaton any one shootthe branches of leaf production,appears to affectthe differentia-occur in groupsand thatthe maximumnumber of tion of axillarystructures as is indicatedby the successivebranch pairs is three(Gill and Tomlinson measurementsincluded in table 1. These relation- 1969), a numbercorresponding to the numberof shipsare suggestedin figure23 by the arrowslead- pairs of leaf primordiain the bud. Thus, when a ing from"leaf production."At maximumlevels of shoot is at a level of vigor conduciveto branch vigor,precocious branches are formed;at medium formation,all availablesites are utilizedin such a levelsinflorescences are formed;and at lowestlevels waythat subsequently developed nodes lack branches. suppressedaxillary buds are formed.Both branches When apical vigoris moderateflower buds are and flowerbuds are most commonlyformed in formedand a sequenceof eventsfollows which can summerwhen shootvigor is greatest,whereas sup- influencesubsequent vegetative development. Flow-

Studies on the Growth of Red Mangrove 3. 119 ersopen and maybe fertilizedabout 1-2 monthsaft- thegrowth of thepropagule affects the vigor of the er bud appearance.Fruits mature in 4-7 months. apex is not knownbut it seems likelyin view of Germinationof the fruittakes place on the tree resultsobtained for other species (Wardlaw 1968). and a large propaguleis formedwhich matures in This influenceis suggestedby the arrowconnecting 4-6 months.These developmentalsteps are always "tpropagule"with "leaf production"in figure23. longerin winterthan in summer.Whether or not Reductionin vigor may thus in turn lead to a

o 030 F C wot t

to)

10 ~ ~ ~ ~ ~ ~ ~~0

?Wf ~~~FnlMA'S T.TAff O D'T "F f O N'D

to-.

1968 ~~~~~1969 *970 FIGURES 16-18. Stages in reproductionevident in shoots 1-20 during theperiod of observationfrom November 1968 to December 1970. Rates are expressedas numbersof organs produced or lost per 30-day interval. Dotted lines indicate continuityof developmentbetween organs. Figure 16. Rate of appearance of flower buds. The values to the left of each peak referto the percentagesof flowersproducing fruits.-Figure 17. Rate of protrusionof hypocotylfrom the ends of the fruits (or "germination"rate). Germinationin factbegins earlier as the embryoenlarges before it is evident externally.-Figure 18. Rate of propagule fall.

120 Gill and Tomlinson temporaryfailure of flowerbud productionon the August-September(figs. 18, 21). Late-developing same shootat a laterstage. propagulesalso fall at this timesuggesting that ab- Abscissionof propagulesis most marked in scissionis affecteddirectly by environment.Indirect-

50at to k> >5

? lol T1 0 2O0C0 --- 40

io fiW nAzo D

151

)4:96 1970 NL'A IXZASO 'Ox'oM DY'TX

FIGURES 19-21. Stages in reproductionevident in the set of shoots42-58 duringthe period of observationfrom March 1969 to December 1970. Rates are expressedas numbersof organs produced or lost per 30-day interval. Dotted lines indicate continuityof development between organs. Figure 19. Rate of appearance of flower buds. The values to the left of each peak referto the percentagesof flowersproducing fruits.-Figure 20. Rate of protrusionof hypocotyl from the ends of the fruits (or "germination"rate). Germinationin fact begins earlier as the embryoenlarges beforeit is evident externally.-Figure 21. Rate of propagule fall.

Studies on the Growth of Red Mangrove 3. 121 The ?-oj,-tit bdk t1 n 7~~ A * OT 1^ ftovr. bud$ flotwt ,

r' A ZAr*2r~Afle , 5ooVO- ~ ypoco,yt&prfar.

teiof ir ' M4;;;; +k H W h 1 - 's o; ~~A-1@ Z, A&:X -= CCq> 1

Thontht A1 JA O0 DA?1S1A 0

FIGURE 22. Diagrammatic representationof sequence of events in reproduction. Upper: the plant population as a whole. Dotted lines indicate continuityof development. Lower: a single reproductiveshoot (actual example). The appearance of the shoot at monthlyintervals is representedsymbolically. Reproductionhad not been successfulin the year leading up to the period shown, nor was it successfulfor the period beginningin June of the second year.

ly,the stimulusappears to affectleaf abscissionalso, less constantbut it does not indicatethe factors as totalleaf number on shootsseems to be correlated controllingthe totalleaf numberlikely to be found closelywith propagule abscission (figs. 9, 11). It is on anyone shoot.This is usually8-10, i.e.,4-5 pairs. unlikelythat this drop in leaf numberis due to a Leaf numberis thus affectedby both leaf pro- change in leaf production.Perhaps substancesre- ductionand fall. In turn,the numberof leaves on sponsiblefor propagule loss are transportedto, and the shootmay affect the productionof futureleaves cause loss of, old leaves on the same shoot.This is throughthe supplyof assimilates,and it may affect suggestedby the arrowconnecting these two factors leaffall through shade-induced senescence. The num- in figure23. ber actuallypresent in any circumstanceis a reflec- Leaf senescenceand fall can be hastenedby tionof the balanceamong all theseprocesses. These frost,drought, and high winds but none of these relationshipsare suggestedby the threelarge circles eventscomplicated the presentstudy. Nevertheless, in the upperpart of figure23. a strongenvironmental influence on organabscission When few leaves are producedper year,the is possible,as indicatedin figure23. Where there leaves subtendonly suppressed axillary buds. In the were no propaguleson the shootto modifygrowth intactplant these are subjectto apical dominance correlationsleaf fall and expansionwere parallel. as has been demonstratedby the pruningexperi- This suggeststhat there is an internalcontrol of leaf ment. They grow upon release from this apical numberon the shoot. Controlwas not mediated dominance.When theyachieve the same threeleaf- throughleaf age per se as this varied seasonally pair compositionas regularapical buds, whatever (figs. 6-8). It is moreprobable that leaf expansion theirsize, then leaves are expanded.This suggests drains the resourcesof lower leaves on the same that all Rhizophoaramangle buds are disposedto shoot causingthem to abscise.Such a mechanism elongatewhen this composition is achieved.If thisis could explain why leaf numberremains more or so, then,by definition(Doorenbos 1953), bothaxil-

122 Gill and Tomlinson \ \ 2\Nttmber./

fO ef -CcS

{ Lewf re~i'oixczoixspply Le ?rodxdton. 7 '1enes ecE <

%<~~~~~~tofrPoppts

Alonmz.I

FIGURE 23. Hypotheticalinterrelationships between various organs, processes,and environmentalfactors suggested dia- grammatically. The influenceof environmentis indicatedby "E" and a shortarrow. Where other arrows are shown as a solid line supportingevidence for the relationshipis provided in the text.Dotted lines indicate possible relationshipsfor which little or no directevidence is yet available. lary and apical buds have no dormancy.Subject conceptsof treegrowth based on temperatespecies. to continuedsuppression in the intactplant they However,no detailsfor tropicalspecies have been becomeovergrown with bark and die withina few presentedbefore. It is an "evergrowing"type in the years. morphologicalsense of Koriba (1958) but it can As has beenindicated earlier (p. 111), correlation be difficultto say fromexternal examination wheth- betweenvigor and branchingis complicatedbecause er or not the shoot is actuallygrowing, especially suppressedbuds are also foundat the same node duringwinter. At the momentso littleinvestigation of insertionas inflorescencesand branches.The po- of growthprocesses in tropicaltrees has been carried sitionof suppressedbuds in relationto branchtype out thatit would be dangerousto make generaliza- is quite constant.The supernumerarybud always tions.It may be noted,however, that Camelliahas lies above the insertionof an inflorescencebut an articulatedmorphology suggesting intermittent below the insertionof a branch(Gill and Tomlin- growthalthough production of leaf primordiaby son 1969). Since suppressedbuds always develop shootapices is continuous(Bond 1942, 1945) and as axillarybranches, never belated inflorescences, po- the shootsare "evergrowing"in anothersense. This sitionalone does not determinethe typeof branch. is only one exampleof the variationin typesof All one can say is that reproductivebranches if shoot growthin tropicaltrees and the reader is presentare alwaysin the lowerposition. referredto a reviewby Tomlinsonand Gill (1971) The methodof growthdescribed for this plant where it is suggestedthat typesof shoot growth appears unusual at firstsight in relationto our in tropicaltrees are infinitelyvaried as compared

Studies on the Growth of Red Mangrove 3. 123 with the situationin temperatetrees. It is hoped, studieson tropicaltrees which are a prerequisite therefore,that these observationson Rhizophora for an understandingof tree growthon a cosmo- can be acceptedas a preliminarycontribution to politanscale.

LITERATURE CITED

BOND, T. E. T. 1942. Studies on the vegetativegrowth and anatomy of the tea plant (Camellia thea Link) with special referenceto the phloem I. The flush shoot. Ann. Bot. (London) n.s. 6: 607-629. 1945. Studies on the vegetativegrowth and anatomy of the tea plant (Camellia thea Link) with special referenceto the phloem. II. Furtheranalysis of flushingbehavior. Ann. Bot. (London) n.s. 9: 183-216. CRITCHFIELD, W. B. 1960. Leaf dimorphism in Populus trichocarpa. Amer. J. Bot. 47: 699-711. DOORENBOS, J. 1953. Review of the literatureon dormancyin buds of woody plants. Meded. Landbouwhogeschool. 53: 1-23. GILL, A. M. 1971. The formation,growth and fate of buds of Fraxinus americana L. in Central Massachusetts.Har- vard Forest Pap. 20: 1-16. , AND P. B. TOMLINSON. 1969. Studies on the growth of red mangrove (Rhizophora mangle L.) 1. Habit and general morphology. Biotropica 1: 1-9. AND . 1971. Studies on the growthof red mangrove (Rhizophora mangle L.) 2. Anatomyand development of aerial roots. Biotropica 3: 63-77. GuppY, H. B. 1906. Observationsof a naturalistin the Pacific between 1896 and 1899. Volume 2. Plant dispersal. Macmillan. London. KORIBA, K. 1958. On the periodicityof tree-growthin the tropics,with referenceto the mode of branching,the leaffall, and the formationof the resting bud. Gard. Bull. StraitsSettlem. 17(1): 11-81. ROMBERGER, J. A. 1963. Meristems,growth and developmentin woody plants. Techn. Bull. U.S.D.A. 1293. Wash- ington,D.C. TOMLINSON, P. B., AND A. M. GILL. 1972. Growth habits of tropicaltrees: Some guiding principles. In B. J. Meg- gers (Editor). Tropical forestecosystems in Africa and South America: A comparativereview. SmithsonianIn- stitutionPress. Washington,D.C. (in press). WARDLAW, I. F. 1968. The control and patternof movementof carbohydratesin plants. Bot. Rev. (Lancaster) 34 (1): 79-105.

124 Gill and Tomlinson