Growth rings in rimu from south Westland terrace forest

Transcription

1 New Zealand Journal of Botany ISSN: X (Print) (Online) Journal homepage: Growth rings in rimu from south Westland terrace forest D. A. Franklin To cite this article: D. A. Franklin (1969) Growth rings in rimu from south Westland terrace forest, New Zealand Journal of Botany, 7:3, , DOI: / X To link to this article: Published online: 10 Feb Submit your article to this journal Article views: 99 View related articles Citing articles: 11 View citing articles Full Terms & Conditions of access and use can be found at

2 1969] 177 GRO\VTH RINGS IN RUIU FROM SOUTH \VESTLAND TERRACE FOREST D. A. FRANKLIN* Forest Research Institute. New Zealand Forest Service (Received for pllblication 26 March 1969) SUMMARY A study of growth rings in relatively fast-grown rimu trees from terrace forest in south Westland revealed a number with conspicuously dark latewood bands. These have been called marker rings. The regular occurrence of these marker rings in the same sequence in a number of trees indicates that growth rings are formed contemporaneously in all relatively fast growing trees. A good correlation between the occurrence of marker rings and cool autumn temperatures in the years in which they were formed is very strong evidence that the rings are formed annuaily. Discontinuous false rings were found in some fast-grown trees. but continuous false rings were not discovered. Discontinuous growth rings were found in strongly fluted trees. in leaning trees. and in very slow-growing trees. No evidence of missing rings was found. but it is possible that growth rings could be absent from slow-growing tree'!. INTRODUCTION Growth rings in conifers consist of alternate layers of earlywood and latewood. Earlywood is formed in the early part of the growing season. and it seems likely that it is formed in response to hormones produced by growing buds. whereas latewood is formed in the absence of. or with a reduced supply of. these hormones (Larson. 1963). In temperate r~gio~s where there is one gro~i?g season a year. normally. one growth rmg IS formed each year and It IS then termed an annual nng. Even in areas where it is known that growth rings are normally formed annually. individual trees often fail to form rings on the lower part of the bole in some years. or the rings that are formed are not continuous around the circumference of the tree (Harris. 1952; Bormann. 1965). In some apparently discontinuous rings. only the earlywood portion of the ring is discontinuous. the latewood being continuous and fused with the latewood of the previous ring. Where two or more growth rings are formed in the one growing season. all except the last-formed ring are called false rings. Where false rings can be recognised. they are much narrower than neighbouring rings * Stationed at Forest apd Range Experiment Station, Rangiora. N.Z. II Bot. 7: Published online 10 Feb 2012

3 178 NEW ZEALAND JOURNAL OF BOTANY [SEPTEMBER and the boundaries between their latewood bands and adjacent bands of earlywood are indistinct (Larson. 1956; Lloyd. 1963). However, in very slowly grown trees these criteria are of no use, and unless accurate aging of trees is possible or unless marker rings can be recognised. it is often impossible to detect false rings. GROWTH RINGS IN RIMlJ Several authors who have worked on growth rates or age determination of rimu (Dacrydium cupressinum Lamb.) have assumed that the growth rings are formed annually, e.g. Cameron (1959, 1960) and Wardle (I 963a. 1963b). This assumption is based on the fact that rimu normally grows in habitats where there is a single and continuous growing season from early spring until late autumn. However, Lloyd (1963) found that an exceptionally fast-grown rimu tree from Russell Forest (Northland) showed exactly the same features as a nearby radiata pine of known age which produced as many as three complete false rings in some seasons, the false rings being recognised because of the narrowness and ill-defined boundaries of their latewood bands. Referring to slower growing trees. he further states that "there is often a quite striking regularity of paired or triple rings, each 'set' being separated from the next by a wider band of springwood than occurs within the 'set' itself. It happens too regularly to be a coincidence and is found in tanekaha, kauri, rimu, totara. and kahikatea in Russell Forrest". The implication is that each set probably represents one year's growth. In a previous investigation (Franklin. 1966) a study was made of the growth rings of 50 rimu poles and small trees growing in terrace forest at Kakapotahi (Westland) and released by logging 25 years previously. The widths of the last 20 to 25 growth rings in most of the trees were much greater than those of previous growth rings in the same trees, and considerably greater than those of the corresponding rings in 30 trees that had not been released. i.e. if the growth rings were formed annually, the released trees had markedly increased their diameter growth within five years of being released. This is strong evidence that at least one growth ring was formed annually, although the possibility that occasionally two growth rings were formed in the one year was not excluded. Patel (1967) has recently described the wood anatomy of rimu but made no comment on growth rings other than that they are indistinct. T~e present study ~as made on rimu growing in.terr8:ce forest in south Westland to determine whether growth rings In f1mu are formed annually, and if this is generally true. what exceptions occur by way of false rings, discontinuous rings and missing rings. As there were no rimu trees of exactly known age growing in this area. marker rings that could be readily identified in ali trees were sought. METHODS Discs were taken at breast height and within the crown from each of six trees at Wanganui Forest, and six trees at Ianthe Forest nine miles to the north-east. The trees had been felled in the winters of

4 1969] FRANKLIN-GROWTH RINGS IN RIMU 179 and 1966; the sapwood of trees felled prior to 1964 was too decayed to allow identification of rings. The diameter at breast height of the sample trees ranged from 8 to 17 in. (20 to 43 cm). After sanding. the surface of each disc was divided into quadrants and within each quadrant. each growth ring with conspicuously dark latewood was marked. Such rings were conspicuous usually because their latewood was darker than average for the quadrant. but occasionally because the latewood of neighbouring rings was paler than average. The growth rings were then numbered from the outside of the disc. and the position of marked rings within the overall sequence of rings was determined. For trees felled in and the first growth ring was numbered 2. 1 and 0 respectively. The frequency with which each of the last 50 growth rings was marked was then calculated. The incidence of marker rings was compared with monthly weather records for Hokitika. 30 miles to the north-east of the study areas. In this meteorological station was moved to Hokitika South. but the monthly figures for all weather phenomena for the period were adjusted by a constant amount. to make the 20-year averages the same as the averages for the period at Hokitika. Once marker rings had been identified. the discs. including several from slower-growing trees. were examined for false rings. discontinuous rings. and missing rings. and transverse sections were cut to study these phenomena further. RESULTS MARKER RINGS Results from subjectively marking the most prominent latewood bands in each quadrant are detailed in the appendix. Rings and 42 were marked most often. while rings and 50 were also marked in at least two-thirds of the discs and 30% of the quadrants. Rings 28 and 50 were more often marked in discs from crowns than in discs from butts. If growth rings are formed annually. these 10 rings would have been completed in the autumns of and These rings are hereafter referred to as 'marker' rings. No disc from either forest had fewer than three prominent marker rings. and in many quadrants the marker rings were quite distinctive on close examination although not sufficiently prominent to have been mar~ed initially. Individual qu~drants contained from to to 45 rings per Inch (25 mm) but marker flngs were most often recorded as being prominent where there were between 20 and 30 rings per inch. Some rings were also conspicuous because their earlywood was much paler than normal. Fourteen different earlywood bands were recorded at least once as being conspicuously pale; 11 of these were recorded I to 6 times (out of 96 quadrants). and three. between rings 20 and and 22. and 27 and 28. were recorded and 34 times

5 180 NEW ZEALAND JOURNAL OF BOTANY [SEPTEMBER respectively. The discs in which these three earlywood bands were conspicuous have been marked with an asterisk in the appendix. CORRELATION BETWEEN ApPEARANCE OF GROWTH RINGS AND WEATIIER If growth rings are formed annually. then the appearance of growth rings at Harihari is not correlated with rainfall or hours of sunshine recorded at Hokitika. but it is correlated with temperature; in particular, conspicuous latewood formation is associated with cool autumn temperatures. Using Spearmans Rank Correlation Coeflicient, the average mean daily temperature given by (mean daily maximum + mean daily minimum) /2 ior both March and April. and for both months taken together, is definitely correlated with the frequency with which rings were recorded as being conspicuous ( R = p < I : R = p < I: R=0.56. p<o.oooi respectively). Although the average mean daily temperature for May is only poorly correlated with the appearance of the latewood (R =0.28. p<0.oo5). when incorporated into the average for the period March-April it helps correct the few serious anomalies. For this reason the three-monthly average has been used in Fig. I. The mean daily maximum and the mean daily minimum temperatures contribute about equally towards the correlati0ns. Pale earlywood bands appear to be associated with hot summers. The most conspicuous earlywood band (between rings 27 and 28) corresponds with the hottest summer recorded at Hokitika this century ( ) when the mcan daily temperature from November to May averaged 4,.. above normal. FALSE RINGS Several false rings were recognised but only in the fastest grown trees (less than 15 rings per inch). They were always discontinuous, not extending more than a quarter of the way around the circumference of the tree. and their latewood was never fused to the latewood of adjoining rings (Fig. 2). Usually they could have been recognised as being false in increment borings by their blurred appearance and the sudden diminution of apparent ring width, but one or two could have been mistaken for normal rings. If continuous false rings were present. they were present in all discs examined and were indistinguishable from normal rings. In some quadrants. occasional growth rings had indistinct latewood and resembled the false rings found in fast-grown trees. but there was no sudden diminution in ring width and they were usually found in trees with more than 30 rings per inch. FLUTING An irregular circumference is a common feature of the lower boles of rimu trees; low bumps and shallow hollows are found on most trees while many of the larger trees have conspicuous vertical ridges and g!ooves which are called }lutes. In strongly fluted trees. up to 200 growth rmgs. may for~ S or 6 m. (13 or 15 cm) of wood in the ridges while form 109 only! m. (12 mm) of wood at the base of the grooves (Fig. 3),

6 1969} FRANKLIN-GROWTH RINGS IN RIMU 181 In deep grooves, individuals growth rings cannot be distinguished by eye but in shallow grooves or hollows where there are less than about 150 rings per in., all rings are often present (Fig. 4). In some discs the pattern of growth rings shows that on a given radius, what was a hollow in previous decades may later be a bump and vice versa. In a few trees this is the result of a sudden change in growth & O ~O 4", I '~ I') " 2 40 O r---~~~----~------_r------~~~ 51 FlO. I-Number of times growth rings were marked as being conspicuous and the average mean daily temperature for Mar~h-April-May in the year that they were formed, assuming the rings to be annual.

7 FlO. 2-Disc from fast-grown tree felled in A well developed false ring occurs between rings 15 and 16, while other false rings occur between rings 7 and 8, 11 and 12. and 21 and 22. Ring 30 occasionally (as in the centre of this photo) resembles a false ring, but it is always continuous. Rings 9, 1 S, 22, 2S, 40, 42, 49 and 50 were recorded as being prominent, and the ea~lywood between rings 14 and 15, and 27 and 28 was recorded as being conspicuously pale. Photos-H. Hemming FIG. 3-Disc from a larae tree showina a deep groove. Growth rings were acc';lrately counted elsewhere on. the disc. The last 150 arowth rings form S.S In. of wood on the left-hand ndge, but only 0.2 in. of wood at the base of the roovc W C ntfl!ti 10.",,,"t.....,,_...!_~ ,...& I_...l

8 1969] FRANKLIN-GROWTH RINGS IN RIMU 183 pattern. but more often it is the result of a slight tangential shift in the position of ho!lows and bumps with each successive growth ring (Fig. 5). MISSING AND DISCONTINUOUS RINGS In trees where marker rings could be recognised. no rings were found to be completely missing around the circumference of a tree. although many rings appeared discontinuous to the naked eye. Discontinuous rings are found in strongly fluted trees as mentioned above. and also in very slow-growing and in leaning trees. Most very slow-growing trees are suppressed and lean towards the light. Growth rings in these are formed on the lower side and are absent or of the fused-iatewood type on the upper side. In some very slow-growing trees without appreciable lean. growth may be confined to one sector for several decades and then may be suddenly confined to another sector. later becoming confined to different sectors again (Fig. 6); the reasons for this are unknown. It is likely that in rimu. as in other species. some growth rings may not be represented at all in the lower part of the bole in very slow-growing trees. but it was not possible to establish this in the absence of discernible Photo-H. Hemming FIG. 4-Disc from tree felled in 1965 showing one of the six shallow grooves that occur around the circumference of the tree. All growth rings are continuous but those at the bllse of the groove cannot be recognised macroscopically. On the left-hand side. rings 16,20.22 and 25 have been marked. while on the right-hand side. rings and 39 have been marked. The pale earlywood between rings 27 and 28 can be seen on both sides of the photo.

9 184 NEW ZEAI.AND JOURNAL OF BOTANY [SEPTEMBER marker rings in slow-growing trees and without any knowledge of their age. MICROSCOPIC FEATURES OF GROWTH RINGS Microscopically, growth rings in rimu are indistinct for the following reasons (see Fig. 7): (I) The transition between eariywood cells and the succeeding latewood cells is very gradual. (2) Compared with other species, latewood cells do not differ greatly from earlywood cells. ( 3) The boundary between latewood and the succeeding earlywood is irregular because of the presence of frequent axial parenchyma. In this study, cells were considered to be latewood cells only if the radial diameter of their lumina was less than two-thirds that of the firstformed eariywood ce])s. Most latewood bands consist of two to four layers no matter how many layers of earlywood cells are present. Some examples are: Phul~H. Hemmlns FIG. 5-Disc from tree felled in The hollow in rina 40 on the right of the photo has gradually shifted to the left with succeedina growth. The bump in the centre of the photo remained on the same radius until about 1 S growth rings ago, since when growth on this radius has almost ceased.

10 1969] FRANKLIN-GROWTH RINGS IN RIMU 185 Number of rings per inch Number of earlywood cells Number of latewood cells Rings with very conspicuous bands of latewood had more layers than usual of latewood cells (up to eight) but the unusual paleness of some earlywood bands could not be explained. The smallest rings recognised consisted of only one layer of earlywood cells and two of latewood cells, equivalent to 400 rings per inch, and were found at the base of a shallow hollow. An indeterminate number of rings at the base of a deep groove were represented only by continuous rows of latewood cells. DISCUSSION Because all marker rings occurred in a consistent sequence in all discs, it is highly probable that all growth rings in a disc are contemporaneous with their numerical counterparts in other discs. Furthermore. it is probable that only one growth ring was formed in each of the summers and because the discs were cut from trees felled in three different years, and correspondence among their ring sequences was obtained by allowing one ring per year for this. Photo-H. Hemming FIG. 6-Disc from a very slow-grown, tree showing irregular diameter growth. More than 300 separate growth flngs are present but accurate counting is impossible.

11 186 NEW ZEALAND JOURNAL OF BoTANY [SEPTEMBER Photo-S. Downurd PIO. 7-T.S. of rimu abowin, an annual rin,. Note the aradual transition from earlywood to the aucceedina latewood and the frequent axial parenchyma (marked with c:roaaes) which c:onfuae the boundary between latewood and the aucceeclina earlywood.

12 1969] FRANKLIN-GROWTH RINGS IN RIMU 187 If growth rings are formed annually, the microscopic evidence, particularly from fast-grown trees, suggests that latewood cells are not formed until late in the season, i.e. late autumn. This agrees with what is known about the growth habits of rimu in this district; resting buds are apparently not formed, and terminal growth continues throughout the summer. The wide latewood bands characteristic of marker rings may have been initiated earlier than usual because of cool autumn temperatures. or may have resulted from a mild winter which permitted growth to continue for longer than normal. The good correlation between the frequency of prominent rings and early autumn temperatures suggests that cool autumn temperatures are primarily responsible. If growth rings are formed annually, nine out of the ten marker rings and six other rings were formed in years when the average mean daily temperatures for the period March to May was lower than 52.6 p (Fig. I). The anomalous marker ring, number 28. is adjacent to a conspicuously pale earlywood band and several inconspicuous latewood bands. and was thus marked quite often as being conspicuous although it does not have a wide latewood band. Three other rings, numbers 32, 44. and 45 were apparently formed in years with cool autumns and yet were marked as being conspicuous only five, two and three times respectively out of a possible 96. Rings 44 and 45 often have wide latewood bands but are not as conspicuous as rings 42 and 46; there is no apparent reason for the anomalous position of ring 32. The correlation between the appearance of growth rings and temperature is extremely good, particularly when it is realised that (a) only the average temperatures for whole months have been used in the analyses, (b) the climate station and the study areas are 30 miles apart, (c) growth rings were judged on their appearance relative to that of their neighbours and not specifically on the width of their latewood bands. and (d) all growth rings can occasionally appear conspicuous at some point in a disc. The last point suggests that although cool autumn temperatures are primarily responsible for the formation of prominent latewood bands. other factors are also involved. The evidence presented in this paper that growth rings are formed annually in rimu is circumstantial. However, now that marker rings have been identified. it should be possible in future to establish whether growth rings are consistently formed annually in this district. by sampling trees periodically. ACKNOWLEDGMENTS I would like to thank Mr G. R. Whiteside for his help in the field. Mr J. D. Coulter of the Meteorological Service for the meteorological data, and officers of the Forest Research Institute for their comments on the manuscript.

13 188 NEW ZEALAND JOURNAL OF BOTANY [SEPTEMHER REFERENCES BORMANN, F. H. 1965: Changes in the growth pattern of white pine trees undergoing suppression. Ecology 46 (3) : CAMERON, R. J. 1959: The management potential of the native forests of north Auckland. N.Z. II. For. 8 (1) : : Natural regeneration of podocarps in the Whirinaki River valley. N.Z. II. For. 8 (2): FRANKLIN, D. A. 1966: The response of rimu poles and small trees in terrace forest to release by logging. N.Z. Forest Service Silvicllltilral Report 53 (un published). HARRIS, J. M. 1952: Discontinuous growth layers in Pinus radiata. N.Z. For. Prod. Res. Notes I (4): LARSON, P. R. 1956: Discontinuous growth rings in suppressed slash pine. Trop. Woods 104: : Stem form development of forest trees. Forest Sci. Monogr. 5: ~4. LLOYD, R. C. 1963: Indigenous tree rings. N.Z. II. For. 8 (5): PATEL, R. N. 1967: Wood anatomy of Podocarpaceae indigenous to New Zealand. I. Dacrydillm. N.Z. II Bol. 5: WARDLE, P. 1963a: Vegetation studies on Secretary Island, Fiordland. 5. Popula. tion structure and growth of rimu. N.Z. II Bot. I: b: The regeneration gap of New Zealand gymnosperms. N.Z. II Bot. I: