Int. J. Plant Sci. 153(3):503-512. 1992. ?3 1992 by The University of Chicago.All rightsreserved. 1058-5893/92/5303-0026$02.00

A MORPHOLOGICALINVESTIGATION OFTHE UNUSUAL CRYPTOGEAL GERMINATION STRATEGYOF BUNYAPINE (ARAUCARIA BIDWILLII)-AN AUSTRALIANRAIN FOREST CONIFER

G. E. BURROWS,* T. S. BOAG,*AND R. A. STOCKEYt *Ron Potter Centre,Charles Sturt University-Riverina, P.O. Box 588, WaggaWagga, NSW 2650, Australia,and tDepartment of Botany, University of Alberta,Edmonton, AlbertaT6G 2E9, Canada

The morphologyof the cryptogealgermination sequence of Araucariabidwillii (bunya pine) was studied. Individualseeds were 5-6 cm in length and 2.5-3.5 cm in diameterand had an averagefresh weight of ca. 16 g. In mature seeds the embryo consisted of a cotyledon tube, hypocotyl, and root cap that were 3-5 mm in diameterand 25-30, 3-4, and 5-7 mm in length, respectively.The embryo was surrounded by, but not fused into, a massive megagametophyte.Germination was rapid as a largediameter pseudo- radicleemerged from the seed 48 h afterthe startof imbibition.At 1 wk the cotyledontube had elongated to 3.5 times its originallength. As cotyledongrowth was positivelygeotropic the root and shoot meristems were forced out of the seed and buried in the soil below the initial position of the seed. At week 3 the hypocotyl was 12-15 and 3-4 times its originaldiameter and length, respectively,and had developed into a parenchymatoustuber. It appearedthat starchstored in the megagametophytewas being mobilized, translocateddown the cotyledon tube, followed by resynthesisin the expandingtuber. At week 4 the tuber had developed a thin reticulateperiderm and had also initiated numerous lateral roots from its lower third. At this stage a distinct abscission zone had developed at the base of the cotyledon tube and after a further2-4 wk the cotyledon tube and megagametophytewere easily detached from the tuber at the abscission zone. In matureseed the epicotyl consisted of only a small apical dome, but at week 5 it was a small shoot that commencedits growthupward through the soil by either splittingor pushingaway the cotyledontube. Initially,the tuberspossessed 4-6 pairsof strandsof primaryvascular tissue embedded in a parenchymatousmatrix. With the commencementof secondarygrowth each pairof strandsdeveloped into a cylinderof secondarytissues, which are eventuallyincorporated into the normal secondarygrowth increments.While extemal evidence of the tuber is lost as secondarythickening of the root and stem progressesthe unusual natureof the hypocotyl can be ascertainedby intemal examination.This germi- nation sequence allows A. bidwilliito transforma large surface seed into a subterraneanstorage and perennationorgan. Bunya pine is a rain forest emergent,and it appearsthat the carbohydratereserve of the tuber allows seedlings to exist for several years under the low light levels of the forest floor, until a breakin the canopy allows for rapidgrowth. The buriedbud reservecreated as the plumulegrew through the soil would allow for regenerationif shoot damage occurred. Introduction closely related to section Columbea species, A. "Cryptogeal"is the term applied to germina- araucana from Argentina and Chile and A. an- tion mechanisms that carry the plumule below gustifolia from Brazil and Bolivia, than it is to the soil surface before shoot elongation, i.e., the its physically closer relatives from the sections new shoots arise from below the ground even Intermediaand Eutacta (Haines 1983a). though the seed germinated on the surface. It is Bunya pine has been grown to a limited extent relativelycommon in the monocotyledons and is in plantations, 1,100 ha having been planted by known in a number of taxonomically unrelated the QueenslandDepartment of Forestry.Its use- families of herbaceous and woody dicotyledons fulness as a plantation species is limited by its (Clarksonand Clifford 1987). It is also known in possession of persistent, sclerophyllous,pointed the cycads, e.g., Encephalartosand Stangeria,but leaves, althoughit has a greaterdrought and cold within the conifers only the three species in the resistance than the widely planted (44,000 ha) Bunya and Columbeasections of Araucariahave hoop pine (A. cunninghamii) (Doley 1990). The a similar germination strategy. seeds of bunya pine are considered difficult to Bunya pine (Araucariabidwillii Hooker) is a store (Doley 1990), which creates problems in conifer native to Queensland, Australia, and is plantation establishment. The trees grow to 30- the only member of the section Bunya. It has a 45 m in height (Boland et al. 1984) and carry limited natural distribution and occurs in the cones in the upper crown that can weigh up to Bunya Mountains-Blackall Range area in the 10 kg (Doley 1990). Because of bunya pine's dis- south of the state (latitude 270S)and the Mount tinctive domed appearanceand strict symmetry Molloy/Mount Lewis area in the north (latitude it was once widely planted in parks and larger 16030'-17040'S)(Boland et al. 1984). It is more gardens,but propertydamage and the risk of per- sonal injuryfrom falling cones has seen it decline in horticulturalfavor. IAuthor for correspondenceand reprints. Cones mature in mid to late summer and fall Manuscriptreceived January 1992; revised manuscriptre- to the ground intact, with the cone scales still ceivedMay 1992. green on the surface and fleshy and resinous in-

503 rXI <.2.

~~~i .d' A %

- 0 Ws d# ';' s@ Bt -i- A.

20D Q 23 24 25 26

Figs.1-6 Various stages of seedling development in bunya pine. In figs. 4 and 6 the approximateinternal position of the shoot meristems is arrowed.Scales, where present, in cm and mm. Fig. 1, Proximal half of a mature seed cone showing its large size and the arrangementof the scales. Note the small diameter of the attachmentaxis (arrowed).Scale, 3 cm. Fig. 2, Seed that had been separatedfrom the cone scale unit that consisted of a large leatherywinged bract (b)and the ovuliferous scale (os). The membranouslayer (arrowed)holds the seed into the cone scale at maturity.Fig. 3, Seeds that have had the BURROWS ET AL.-CRYPTOGEAL GERMINATION OF BUNYA PINE 505 temally(Doley 1990).Their weight helps the cones pointing toward the middle of the pot to permit to roll down slopes and hence achieve an in- unimpeded pseudo-radicle emergence and sub- creased dispersal. The cones rapidly disintegrate sequent hypocotyl enlargement. on the ground as the structure dries (Haines The seeds were planted below the surface to 1983b). Further drying releases the large (5 cm promote uniformity of germinationand seedling length, 3 cm diameter), heavy (10-17 g) seeds morphology. As seed burial would be unusual from the cone scales, and they are then available under naturalconditions, a further 10 seeds were for dispersalby animals. Occasionallythe humid planted on the surfaceof the potting mix and the conditionsin the decomposingcone promote seed pots were covered with 70% shadeclothto main- germination(Haines 1983b), and germinationof tain a high humidity. All pots were maintained seeds within sealed plastic bags is also known in a shade house, under prevailing conditions. (Doley 1990). Seeds or seedlings were assessed on the day The seeds were an important food source for imbibition commenced, 3 and 7 d thereafter,and the local Aboriginals(Cherikoff and Isaacs 1989). then weekly for a further 6 wk. The hypocotyl However, while cones are usuallyproduced every regions of bunya pines growing in the field at year, they are producedin abundanceonly every Gympie, S.E. Queensland, were also studied. third year (Cherikoff and Isaacs 1989). Nutri- These plantswere 1-4 yr of age, secondarygrowth tional studies indicate that over 40% of the fresh was well advanced, and the hypocotyl region was weight of fresh seed is available carbohydrate up to 5.0 cm diameter. The hypocotyls were cut (Brand et al. 1985). Likewise, 60% of the dry into 6-8 mm thick discs with a radial saw, the weight of A. araucana seed is starch (Cardemil end grain polished and photographed. and Reinero 1982). In the latter half of the nineteenth century the Results first descriptions of bunya pine's unusual ger- WEEKS 0-8 mination sequencewere recordedby Diirr(1864), Mature seed cones were approximately22 cm Blanchard(1892), Heckel (1892), and Hemsley in diameter and were comprised of numerous (1901-1902). Since this time there have been overlapping scales (fig. 1). Most scales were fer- morphologicaland anatomicalstudies of seed and tile, although the scales at either end were either embryo formation (Wilde and Eames 1948; sterile or bore small, elongatedseeds. Cones were Haines 1983b) and of isolated stages of seedling held in trees by a relatively small diameter pe- development(Seward and Ford 1906; Shaw 1909; duncle, which indicates why the cone is shed Stockey and Taylor 1978; Rouane and Woltz whole, rather than the cone disintegratingwhile 1979; Stockeyet al. 1990), but no sequentialmor- attached to the tree (fig. 1). Unlike hoop pine, phological studies of germinationhave been un- where the seed is fused into the scale unit, in dertaken. The physiology of certain aspects of bunya pine the seeds were easily removed as in they cryptogeal germination Araucaria have also are held in by only a thin tissue layer (fig. 2). been investigated (Cardemil and Reinero 1982; Seeds were ca. 5.5 cm long x 2.5 cm at their Doley 1990). Given the continuinginterest in the largest diameter and weighed ca. 16 g. unusual germination sequence of bunya pine it The embryo and megagametophytewere pro- was considered useful to undertake a thorough tected by a thick (1.0-1.5 mm), woody sclerotesta morphologicalinvestigation. layer(fig. 3) that showedno obvious lines or points of weaknessfor Materialand methods seedlingemergence. The embryos were surroundedby megagametophytetissue, ex- Bunya pine cones were collected on the day of cept at their radicle end, and there was no phys- fall in late summer from specimen trees growing ical attachment between the embryo and the at WaggaWagga, N.S.W., and approximately65 megagametophyte tissue (fig. 4). Longitudinal fully formed seeds were extracted.The seeds were sections of the embryos showed that in a total planted in sterile potting mix in 15 cm diameter embryo length of ca. 4 cm the root and shoot pots. Two seeds per pot were planted 1 cm below meristems were only 3-4 mm apart, while the the surface, with the long axis positioned hori- bulk of the embryowas made up of the cotyledons zontally and with the radicle end of the seed and the "root cap." The cotyledons were fused

middle sclerenchymatousintegumentary layer partiallyremoved, exposing the inner papery integumentarylayer that covers the megagametophyte.Fig. 4, Partial removal of the megagametophyte(m) revealingthe torpedo-shapedembryo. Note the root cap that gives the lower third of the embryo a matt appearance.ct, cotyledon tube. Fig. 5, Germinatingseeds 2-3 d after imbibition, showingthat an increasein internalvolume has split the middle integumentarylayer sufficientlyto permit radicle extension.Note the mucilaginoussheath that covers the radicle.Note also the strandsof the inner integumentarylayer (arrow) in the right-hand-sideembryo. Fig. 6, Germinatingseeds, week 1. Note that the mucilaginoussheath covers the cotyledon tube, as shown by the trappedparticles of potting mix, but the sheath does not cover the rapidlyelongating hypocotyl (h). 1 -

.e t

_ ,'*_ .. i; 4 -~~~~~~~~~ 2~~~- 1

L '~~~~~~~2~~O BURROWS ET AL.-CRYPTOGEAL GERMINATION OF BUNYA PINE 507

Table1

DIMENSIONS (cm) OF VARIOUS COMPONENTS OF BUNYA PINE SEED OR SEEDLINGS DURING THE FIRST 6 wk OF GERMINATION

CT length Outside CT Hypocotyl/tuber Root Week Total seed diameter Length Diameter length 0 .. 2.5-3.0 0 .3 x .5 .3-.4 .3-.5 0 1.. 8.5-9.5 4.0-4.5 .3 x .5 1.5-2.0 .3-.5 0 2 .. 8.5-9.5 4.0-5.0 .3 x .4 4.0-4.5 .5-.8 .5-1.0 3 .. 9.0-10.0 4.0-5.0 .2 x .3 4.5-5.0 1.1-1.4 2.0-3.5 4 .. 9.0-10.0 4.0-5.0 .2 x .25 4.5-5.0 1.5-1.7 3.5-4.0 5 .. 9.0-10.0 4.0-5.5 .2 x .25 4.5-5.0 1.8-2.0 4.0-6.0 6 .. 9.0-10.0 4.0-5.0 .2 x .25 4.5-5.5 1.9-2.3 7.0-9.0

Note. CT, cotyledon tube. into a tube; however, at this stage the central At week 2 the hypocotyl had doubled or trebled hollow was only 0.1 x 2.0 mm in cross section in length during the previous 7 d and was 85%- (C.S.) and was not an obvious feature. The cot- 95% of its final length (table 1). Between weeks yledon tube was oblong in C.S. (3 x 5 mm) and 1 and 2 the hypocotyl began to increase in di- had a smooth glossy appearance,while the lower ameter, and consequently a clear external dis- third of the embryo was covered by relatively tinction between the cotyledon tube and the hy- loose cells that gave a matt finish (fig. 4). pocotyl began to develop (fig. 7). Seeds exhibited no dormancyand germination At week 3 elongation of the hypocotyl had was rapid as pseudo-radicleemergence occurred ceased, and it was approximately 12.5 times its within 48 h of imbibition. Water uptake during original length. Rapid increases in the diameter imbibition caused an increasein internalvolume of the hypocotyl were still occurring, and con- that consequentlysplit the integumentalong lines sequently it could now be termed a tuber (fig. 8). of relative weakness (fig. 5). Radicle extension The cotyledon tube began to reduce in diameter was alwayspositively geotropicregardless of seed and consequentlyhad a grooved appearancethat orientation.When the radicle and hypocotyl first gradually became more prominent (fig. 10). It emerged they were covered by a loose, fibrous, appearedthat the ridges correspondedto the po- mucilaginous mat that could be easily removed sition of the vascular bundles within the tube, from the underlyingtissues (fig. 5). while the furrowscorresponded to the parenchy- At 1 wk elongation of the cotyledon tube was matous tissues between the bundles. The outer largely complete, and it was three to four times surface of the cotyledon tube began to darken its original length (table 1; fig. 6). In the same from what appeared to be tannin or phenolic periodthe hypocotylelongated by five to six times, compound deposition in the surface cell layers, althoughits diameterremained unchanged (table not from periderm formation. The tuber began 1). The mucilaginous sheath now covered the to develop a thin reticulateperiderm, while small length of the cotyledonarytube, but not the hy- white pustules formed on the lower third of the pocotyl. At 1 wk therewas little externalevidence tuber.An obvious root, as distinct from the tuber, to indicate that each bunya pine seed had not had also formed (fig. 9). produced a single, unbranched,rapidly growing, At week 4 most seedlings had developed a dis- large diametertaproot. Dissection was needed to tinct abscission zone at the base of the cotyledon show otherwise. tube adjacent to the tuber (figs. 9, 10), although

Figs.7-13 Various stages of seedling development in bunya pine. In figs. 7-9 arrowsindicate the approximateposition of the shoot apicalmeristems. Fig. 7, Germinatingseeds, week 2. Note the continuedlengthening of the hypocotyland its increased diameter. Fig. 8, Germinatingseeds, week 3. Note the continued increase in hypocotyl/tuber(t) diameter, the reduction in cotyledon tube diameter,and the developmentof a root (small arrows)as distinct from the tuber. Fig. 9, Germinatingseeds, week 4. Note the developmentof a thin reticulateperiderm on the tubersand the formationof small pustules (small arrows) on the lower third of the tuber. Fig. 10, Detail of the right-hand-sideseedling in fig. 9, showingperiderm development on the tuber, development of a distinct abscission zone (arrowed),and reductionin cotyledon tube diameterresulting in a grooved appearance.Fig. 1, Seedling6-8 wk of age showing separationof the megagametophyte,integumentary layer, and cotyledon tube from the hypocotyl/tuberwith its root and shoot meristems.Note the numerouslateral roots that have developed from the pustulesshown in fig. 9. Fig. 12, Seedling3 mo of age showingincomplete separation of the cotyledontube from the tuber. Note that the plumule lacks chlorophyllbut is not excessively etiolated. Fig. 13, One-year-oldseedling. While the hypocotyl/ tuber swellingis still prominentwith subsequentsecondary growth in the shoot and root, it will not be externallyobvious. l;zr* D o : P

Ni.~~~~~~sl

641 1

Figs.14-17 Successive cross sections down throughthe former tuber region of a bunya pine 4 yr of age. Note the unusual patternsof xylem formation and the large pith areas. Scales, 3 mm. Fig. 14, Stem wood immediately above the previously tuberousregion. Fig. 15, Section from near the "tuber"apex. Figs. 16, 17, Mid-regionsof the "hypocotyl"/"tuber." BURROWS ET AL.-CRYPTOGEAL GERMINATION OF BUNYA PINE 509 separationdid not occur even with gentle force. strandsmerged together again toward the base of At this stage the cotyledon tube had contracted the tuber. Serial examination of successive cut to approximately50% of its originaldiameter and surfacesindicated that the strandswere indepen- had a leatherytexture compared with its previous dent of each other for most of their lengththrough firm and turgid nature (fig. 10). the tuber; i.e., there were no anastomoses. By weeks 6-8 the pustules observed at week 3 Secondarygrowth proceededin a typical man- had developed into lateral roots that formed in ner in the stem and root tissues above (fig. 14) longitudinalrows down the sides of the tuber (fig. and below the tuber/hypocotyl.In the mid-region 1 1). In this area of root initiation the outer tuber of the tuber the vascular strands graduallycom- sloughed away to reveal another periderm cov- menced normal secondarygrowth, but after sev- ered surface beneath. The upper tuber surface eral months each pair of strandsjoined together, remainedintact. Between weeks 5 and 6 the first forming a small diameter ring. It appeared the detachmentsof the tuberfrom the cotyledon tube vascular cambia were joined by short arcs of in- occurred,although this would not have happened terfascicularcambium that differentiatedacross had the seedlings not been lifted; i.e., separation the intervening parenchyma.This occurred first would not have occurredthis early under natural across the smaller gap to the outside and later conditions. Separationwas usually complete at a across the largerinner gap, and thus four to six well-formed abscission layer, although a hinge cylinders of secondaryvascular tissue developed sometimes formed (fig. 12). Under natural con- within the tuber. The tracheids of the secondary ditions the epicotyl grows up the cotyledon tube xylem were formed in rows, but very few of these for a shortdistance until it splits the tube or push- rows had a radial orientation (figs. 15-17), as in es it off (fig. 12). Dissection of fresh seeds showed a normalgrowth increment (fig. 14). As secondary that the megagametophytehad a firm solid tex- growth proceeded, the parenchyma core within ture, but by week 6 it had a spongy nature. each ring was crushed, as was the pith area be- In matureembryos the shoot apex consisted of tween the rings (figs. 16, 17). only a small apical dome, with no primordial With furthersecondary growth it appearedthat leaves present.At week 6, well-developed shoots, as the cambial rings increased in diameter they 0.2-1 cm in length,with numerousleaf primordia eventually fused or grafted into each other and were present.In older seedlingswhere the shoots thus formed a single large convoluted circle. The had pushed through2-3 cm of overlying soil, the inner segments of cambium graduallyceased to leaves were small and achlorophyllous,although function, while the outer arcs joined together to the intemodal distances were small (fig. 12); i.e., functionas a typicalvascular cambium producing there was no obvious etiolation when compared normal increments of secondary xylem with the with shoots that had developed above ground. tracheids in radial rows (figs. 16-17). While the The tubersreached a maximum diameterof 2.0- external distinction between the tuber and the 2.3 cm at ca. 2 mo of age, but because of the remainderof the axis is lost with continued sec- collapse of the parenchyma from utilization of ondary growth, the unusual nature of the hypo- stored carbohydratethey shrink to be only 1.0- cotyl regioncan alwaysbe ascertainedby internal 1.5 cm in diameter at 12 mo of age (fig. 13). examination. The seeds placed on the potting mix surface displayedthree developmentalresponses: (i) 50% Discussion developed a typical large tuber, but as the coty- ledon tube extended outside the seed coat by an COMPARISON WITH OTHER CRYPTOGEAL SPECIES averageof only 2.5 cm, the shoot apex was buried Clarksonand Clifford(1987) noted two main to only 1-2 cm below the soil surface;(ii) another characters associated with cryptogeal germina- 30% also had 2.5 cm of cotyledon tube beyond tion: (1) some fusion of the cotyledons is present, the seed coat, but the radicle did not penetrate and usually the greaterthe fusion, the greateris the surfaceand a much smaller (0.8 cm diameter the depth at which the short apex is buried; and x 3.0 cm length)tuber was formed;and (iii) 20% (2) large-seededness.Bunya pine, a species not did not germinate, as compared with 100%ger- consideredby Clarksonand Clifford,adds further mination when seeds were planted below the soil support to these generalizations. surface. In cryptogeal species, varying levels of coty- ledon fusion have been recorded.Fusion may be WEEK 8-YEAR 4 (i) limited, consisting only of the margins of the In young, full-size tubers the primaryvascular cotyledon petioles, as in Marah oreganus(Schli- tissues were arrangedin a continuous ring im- sing 1969) and Elephantorrhizaelephantina (van mediately below the epicotyl, but in the midsec- der Schijffand Snyman 1970); (ii) complete, ex- tion of the tuber they had split into four to six cept for a tube extendingfrom above the plumule pairs of long, narrow (0.2 x 0.9 mm) strands to the cotyledon tips as in Butyrospermumpara- embedded in a largevolume of parenchyma.The doxum (Jackson 1968); or (iii) complete so that 510 INTERNATIONAL JOURNAL OF PLANT SCIENCES the cotyledons are solid except for an air space be apoplastic, in contrast to the entirely sym- above the plumule, as in Jedda multicaulis plastic transportpathway that would be possible (Clarksonand Clifford 1987). Bunya pine is sim- in Jedda and Marah. ilar to B. paradoxum in the degree of cotyledon The mucilaginous mat observed at days 2-7 fusion. may be a hydratedform of the outer layers of the In J. multicaulis,E. elephantina,M. oreganus, massive "root cap," i.e., the cells that gave the and several of the species described by Jackson lower third of the embryo a matt finish. It could (1974) the developing shoot must burst through allow low-friction penetration of the soil by the the base of the fused cotyledons. From published pseudo-radicle. In M. oreganus epidermal root information, bunya pine appears to be unique hair-like cells were found on the cotyledon tube among cryptogeal species in the formation of a but were not present in bunya pine. distinct abscission zone that allows for the pro- grammed detachment of the depleted megaga- LARGE-SEEDEDNESS metophyte from the tuber. As a consequence,the Seed sizes and weights for some of the better epicotyl is not impeded by the cotyledon tube in described cryptogeal species are as follows: J. its upwardgrowth, although in other species the multicaulis-fruit with one seed: 4-6 x 6-7 cm/ tube may provide the epicotyl with some protec- 75-100 g (Clarksonand Clifford 1987); M. ore- tion from abrasivesoil particles.While Araucaria ganus-one seed: 0.8-1.0 x 1.5-2.5 cm/1.l g bidwilliiis closely related to A. angustifolia and (Schlising 1969); and E. elephantina-one seed: A. araucana,the seeds of the two South American 2 x 2.5 cm/(no weight given) (van der Schijff species possess two separate cotyledons rather and Snyman 1970). than a cotyledon tube (Haines 1983a). In A. ar- These measurementsand those of bunya pine aucana (Cardemiland Reinero 1982, fig. 10) and (one seed: 2-3 x 5-6 cm/16 g) support the gen- A. angustifolia(Hill and de Fraine 1909, pl. 15, eralization(Clarkson and Clifford1987) that large- fig. 5; Stockey and Taylor 1978, fig. 9; Rouane seededness is associated with cryptogeal germi- and Woltz 1979, fig. 20) the epicotyl is able to nation. and thus grow upward between the cotyledons, ECOLOGY no advantageexists in the formationof abscission zones at the base of each cotyledon. Cryptogealgermination is generallyinterpreted A comparisonof cryptogealspecies shows that as an adaptationto fire (Jackson 1974; Clarkson the shoot apical meristem is buried to the follow- and Clifford1987), to heavy grazingor trampling ing degrees:J. multicaulis:2-8 cm, E. elephan- (Clarkson and Clifford 1987), or as a drought- tina: up to 4 cm, M. oreganus: 5-25 cm, Com- evasion mechanism (Schlising 1969). bretum binderanum (Jackson 1974): 4-6 cm. Bunya pine is a tropical and subtropical rain Bunya pine, with 2-5 cm of cotyledon tube ex- forestemergent. Seedlings may spendseveral years tension, is in the normal rangedisplayed by cryp- achievinglittle or no growthin the low light levels togeal species. Seedlingsof A. angustifoliaand A. of the forest floor before a break in the canopy araucana, as illustrated by Hill and de Fraine allows for rapid upward growth. Thus, the car- (1909), Stockey and Taylor (1978), Rouane and bohydratestored in the tuber permits bunya pine Woltz (1979), and Cardemil and Reinero (1982) to survive under light levels that may be too low would appear to have a more limited (< 1 cm) for appreciable carbon gain through photosyn- buryingof the shoot apex. thesis. In J. multicaulisthe seedling remains attached The timing of cone fall and the sequence of to the seed/fruit for an extended period and the seedling development correlatewell with the cli- fruitcan also remain attachedto the parentplant, matic patterns of S.E. Queensland. Cone fall co- while in E. elephantinaattachment between seed- incides with the wettest period of the year, which ling and seed occurs for sufficient time that the allows for imbibition of the seeds, and the radicle vascular bundles in the cotyledon tube are sec- would encounter the least resistance as it pene- ondarilythickened. This would indicate that nu- trated the moist leaf litter and soil of the forest trient transferfrom the seed to the tuber can con- floor. The seedlings would be able to establish a tinue over an extended period. In contrast, the well-developed root and tuber system before late formation of an abscission zone by bunya pine autumn/early winter. The tubers of bunya pine means that nutrient transfer must be relatively areknown to be resistantto desiccation(Hemsley rapid and efficient. In J. multicaulis and M. or- 1901-1902), and although bunya pine occurs in eganus the seeds lack endosperm and the entire areas with an annual rainfall of 900-2,000 mm volume of each seed is embryo, with the bulk of (Boland et al. 1984), winter and early spring in the nutrientsbeing stored in the largecotyledons. S.E. Queensland can be relatively dry (189 mm In bunya pine most nutrient and carbohydrate average rainfall in Kingaroy May-September). storageis in the megagametophyteand thus trans- Thus the tubersmay also have a drought-evasion port from this tissue to the cotyledon tube must function. BURROWS ET AL.-CRYPTOGEAL GERMINATION OF BUNYA PINE 511

In growingthrough the overlyingsoil the bunya Nishida 1981, figs. 2-4, 6). These structuresshow pine epicotyl buries many leaves. It is highly no resemblanceto normal bunya pine tubers at probable that in the axil of each of these leaves any sectioninglevel or at any developmentalstage. thereis an axillarymeristem (Burrows 1986, 1987) Stockeyet al. (1990, figs. 40,42) illustratea bunya capableof developing into a new shoot (Burrows pine seedlingthat developed a corm-typeswelling 1989). These relatively undifferentiated meri- after the root tip had been damaged. They sug- stems are exogenous in origin but are covered by gested that in past periods many seedlingswould specializedperiderms and are thus shielded from have grown in volcanic soil subject to frequent the soil environment. Thus the bunya pine seed- ash fall, and thus root apex damage and the sub- ling develops a bud reserve,protected by the soil. sequentformation of turbinateswellings may have In Queenslandrain forests fire is rare, and there been a common occurrence. are few herbivoresthat would attackbunya pine's The top-shaped structures shown by Stockey pungent, sclerophyllous leaves. In either situa- and Taylor (1978, figs. 1, 4, 5) have a greater tion, the bud reserve should permit the reestab- externalresemblance to bunya pine tubers, but it lishment of an above-groundshoot system. appears that they may not be araucarian,or at In M. oreganus the hypocotyl eventually de- least not closely associated with bunya pine, for velops into an immense, perennial underground the following reasons: tuberthat initiates annualvines afterthe wet sea- i) There do not appearto be any vascularbun- son, to replace those that died in the preceding dles from the cotyledon tube in the cortex (Stock- dry season (Schlising1969). Jedda multicaulishas ey and Taylor 1978, fig. 3), as could be expected a multistemmed habit with a complex crown of from their figure13; however, tissue preservation stem tissue (Clarksonand Clifford 1987), while of these specimens is poor; and in E. elephantinathe above-groundstems die back ii) Stockey (1975, 1978) indicates that for the each winter (van der Schijffand Snyman 1970). extinct species A. mirabilis(placed in the section Thus in these species the tuber is a perennating Bunya) the maturecones were a maximum of 10 organ for the life of the plant. In bunya pine the x 10 cm and individual seeds were 0.2-0.6 cm initial differencein diameter between the tuber wide x 0.8-1.3 cm in length, and this appearsto and the stem is reduced as (i) the stem and root be too small to produce, via cryptogeal germi- increase in diameter through secondary thick- nation, a largelyparenchymatous structure of 2.3 ening and (ii) the tuber decreases in diameter cm diameter x 3.5 cm length. through the utilization of the stored carbohy- As per the present study, the papers of Shaw drate. After approximately4-5 yr growthno dif- (1909), Stockey and Taylor (1978, fig. 14), and ferencein diameterexists between the hypocotyl/ Stockey et al. (1990, fig. 41) show some initial tuber region and the adjoining shoot and root, stagesof secondarygrowth in bunya pine and the and anatomical study is needed to reveal the un- unusualarrangement of the tracheids.This study usual natureof this region (figs. 14-17). Thus, in indicates that, for a fossil structure to be posi- bunya pine the tuber is a perennatingorgan while tively identified as a seedling of bunya pine, or a the plant is small, but not subsequently. close relative, it should possess four to six arcs or cylinders of secondary xylem. PALEOBOTANICAL CONSIDERATIONS The studies of Sewardand Ford (1906, fig. 16), Wieland(1935), Gothan (1950), Calder(1953), Hill and de Fraine (1909), Stockey and Taylor Stockey and Taylor (1978), Nishida (1981), and (1978, fig. 17), and Ferreira(1981, fig. 5) indicate Stockeyet al. (1990) have interpretedvarious fos- that the tubersof A. araucanaand A. angustifolia silized plant remains, rangingfrom corm to top would also produce some tracheids that are not or turbinatein shape,as possible araucarianseed- in radial files, even if their early stages of sec- ling structures. ondary growth are not as unusual as in bunya When cut in C.S. the turbinateforms are prin- pine. If suspected fossil seedlings were more cipally composed of tracheids in radial files and closely related to the South American species, a small centralpith of uncompressedparenchyma there should still be some evidence of tracheids cells (Gothan 1950, pl. 1, fig. 5; Calder 1953, figs. in nonradial rows. 20, 22-25; Stockey and Taylor 1978, fig. 33; Literaturecited Blanchard,J. 1892. Curieuse germinationde l'Araucaria foods. 3. Seeds and nuts. Food Technol. Aust. 37:275- bidwillii.Rev. Hortic. (Paris),pp. 295-297. 279. Boland,D. J., M. I. H. Brooker,G. M. Chippendale,N. Hall, Burrows,G. E. 1986. Axillary meristem ontogeny in Ar- B. P. M. Hyland, R. D. Johnston,D. A. Kleinig, and D. J. aucaria cunninghamiiAiton ex D. Don. Aust. J. Bot. 34: Turner. 1984. Forest trees of Australia.Nelson-CSIRO, 357-375. Melbourne. 1987. Leafaxil anatomyin the Araucariaceae.Aust. Brand,J. C., V. Cherikoff,and A. S. Truswell. 1985. The J. Bot. 35:631-640. nutritional composition of Australian Aboriginal bush- 1989. Developmental anatomy of axillary meri- 512 INTERNATIONAL JOURNAL OF PLANT SCIENCES

stemsofAraucaria cunninghamii released from apical dom- 1974. Cryptogealgermination and other seedling inance following shoot apex decapitationin vitro and in adaptionsto the burningof vegetationin savannaregions: vivo. Bot. Gaz. 150:369-377. the origin of the pyrophytichabit. New Phytol. 73:771- Calder,M. G. 1953. A coniferouspetrified forest in Pata- 780. gonia. Bull. Br. Mus. (Nat. Hist.) Geol. 2:99-138. Nishida, M. 1981. A corm-like hypocotyle of araucarian Cardemil,L., and A. Reinero. 1982. Changesof Araucaria seedling from the Upper Cretaceousof Hokkaido. J. Jpn. araucanaseed reservesduring germination and early seed- Bot. 56:111-116. ling growth.Can. J. Bot. 60:1629-1638. Rouane, M. L., and P. Woltz. 1979. Apport de l'etude des Cherikoff,V., and J. Isaacs. 1989. The bush food handbook. plantules pour la taxonomie et l'evolution des Araucari- Ti Tree Press, Sydney. acees. Bull. Soc. Bot. Fr. Actual. Bot. 126:67-76. Clarkson,J. R., and H. T. Clifford. 1987. Germinationof Schlising,R. A. 1969. Seedlingmorphology in Marah (Cu- Jedda multicaulisJ. R. Clarkson(Thymelaeaceae): an ex- curbitaceae)related to the CalifornianMediterranean cli- ample of cryptogealgermination in the Australianflora. mate. Am. J. Bot. 56:552-561. Aust. J. Bot. 35:715-720. Seward,A. C., and S.0. Ford. 1906. The Araucarieae,recent Doley, D. 1990. Utilisation of intrinsic water in the ger- and extinct. Phil. Trans. R. Soc. London B 198:305-41 1. mination of Araucariabidwillii seeds. Seed Sci. Technol. Shaw, F. J. F. 1909. The seedling structureof Araucaria 18:33-42. bidwillii.Ann. Bot. 23:321-334. Diirr, F. J. 1864. Ueber das Keimen und die Vermehrung Stockey, R. A. 1975. Seeds and embryos of Araucariami- der Araucariabidwilli. Gartenflora 14:103-105. rabilis.Am. J. Bot. 62:856-868. Ferreira,A. G. 1981. Aspectos estructuralesde las semillas 1978. Reproductivebiology of CerroCuadrado fos- de Araucariaangustifolia (Bert.) 0. Ktze. IheringiaSer. Bot. sil conifers:ontogeny and reproductivestrategies in Arau- Porto Alegre 26:3-7. caria mirabilis(Spegazzini) Windhausen. Palaeontographi- Gothan, W. 1950. Uber die merkwiirdigenfeigenartigen ca 166:1-15. Kieselknollchenaus dem versteinertenWald des Cerro Stockey,R. A., M. Nishida, and H. Nishida. 1990. Structure Cuadradoin Patagonien.Misc. Acad. Berolinensia1: 149- and diversity of the woody conifer seedling-likestructures 154. from the upper Cretaceousof Hokkaido,Japan. Bot. Gaz. Haines, R. J. 1983a. Embryodevelopment and anatomyin 151:252-262. AraucariaJuss. Aust. J. Bot. 31:125-140. Stockey, R. A., and T. N. Taylor. 1978. On the structure 1983b. Seed developmentin AraucariaJuss. Aust. and evolutionaryrelationships of the CerroCuadrado fossil J. Bot. 31:255-267. conifer seedlings.Bot. J. Linn. Soc. 76:161-176. Heckel,E. 1892. Surla germinationdes grainesd'Araucaria van der Schijff, H. P., and L. Snyman. 1970. The mor- bidwilliHook. Ann. Fac. Sci. Mars. 2, fasc. 6. phology and germinationof the seed of Elephantorrhiza Hemsley, W. B. 1901-1902. Germinatingseeds of Arau- elephantina.J. Arnold Arbor. 51:114-127. caria bidwillii.Proc. Linn. Soc. Lond. 114:1-2. Wieland,G. R. 1935. The CerroCuadrado petrified forest. Hill, T. G., and E. de Fraine. 1909. On the seedlingstructure CarnegieInstitution of WashingtonPublication 449. of gymnosperms.II. Ann. Bot. 23:189-227. Wilde, M. H., and A. J. Eames. 1948. The ovule and "seed" Jackson, G. 1968. Notes on West African vegetation. III. of Araucariabidwilli with discussion of the taxonomy of The seedling morphologyof Butyrospermumparadoxum the genus. I. Morphology.Ann. Bot. 12:311-326. (Gaertn.f.) Hepper.J. West Afr. Sci. Assoc. 13:215-222.