Aust. J. Bot.

, 1983, 31, 277-81

Tree Rings in Tropical

Callitris macleayana F. Muell.

J. Ash
Biogeography and Geomorphology Department,
Research School of Pacific Studies,
Australian National University,
P.O. Box 4, Canberra, A.C.T. 2600; present address:
School of Natural Resources, The University of the South Pacific.
Laucala Bay, Suva, Fiji.

Abstract
Growth rings were examined in trees of Callitrzs macleayana from the Atherton Tablelands in
Queensland, Australia. The climate is tropical with a distinct dry season that induces regular annual
growth rings. A correlation of 0 74 was found between annual ring widths and the duration of the
wet seasons. The potential of this species for dendrochronological study is apparent.

Introduction
Like most gymnosperms, Callitris (Cupressaceae) forms relatively distinct rings of
contrasting tracheid anatomy. Callztrzs species are most abundant in semiarid regions and
several attempts have been made to interpret their rings (Lange 1965; Pearman 1971; Ogden
1978) in which late wood seems to be formed during cool and dry periods.
Callitris macleayana F. Muell. is a tree, with a maximum trunk diameter of about 1 m,
which grows in relatively humid climates of eastern Australia, north of latitude 33" south,
typically on infertile soils. C. macleayana occurs in either pure stands or in association
with various sclerophyll tree species, often at the margin of rainforests, but seems incapable
of regenerating within such closed canopy forests. Callitris is readily killed by intense fires
and the seedlings are destroyed by grazing (Walker 1976), so Callitris populations have
probably been affected by changes in land use with the transition from Aboriginal to
European practices.
This study was intended to determine the nature and periodicity of Callitris macleayana
growth rings and to determine their value as indicators of tree age, population dynamics
and past climates.

Study Area and Methods

The study area is a ridge of metamorphic schist near Christmas Creek on the Atherton
Tablelands (lat. 17"21rS., long. 145'45'E.; alt. 640 m above sea level). It is thought that
Aboriginal fires maintained patches of sclerophyll woodland, such as the study site, on
ridges and infertile substrates in areas that otherwise would support rainforest. With the
cessation of Aboriginal fires early this century, conditions may have become more suitable
for colonization and survival by Callitris and, if so, stands of trees should date from this
period. In addition to Callitris macleayana, there are trees of Tristania conferta R.Br.,
Eucalyptus intermedia R. T. Baker, Banksia integrifolia L., Casuarina torulosa Ait., Alstonia
0067-1924/83/030277$02.00
 J. Ash

muellerana Domin and Acacia rnelanoxylon R.Br., with an understorey including

Xanthorrhoea Sm. and Gahnia J. & G. Forst. On lower slopes and adjacent basalt substrates
there is a diverse rainforest cor,imunity. The study area has been logged at least once
during the past 50 years.
Thirteen Callitris macleayana trees were cored on their upslope and downslope sides
at breast height, and the cores, 4 m m wide, were sanded smooth to reveal their woody
anatomy under a reflecting light microscope. The wood anatomy was described by the
radial diameter of the tracheids, which distinguishes wide, fast-growing tissues and the
narrow, late wood cells characteristic of periods of slow growth.
Temperature, evaporation and sunshine records were available from Kairi, 30 km north-
west, and rainfall records from Malanda, 15 km west. The annual rainfall at Malanda
averages 1700 mm and probably increases to 2700 mm in the study area. An index of
water availability (Fitzpatrick and Nix 1970) was calculated from the rainfall records at
Malanda by the methods of Keig and McAlpine (1974). The duration of the growing season
was calculated as the period with >O 5 water availability index values from one spring
dry season to the next.

Results
The growth rings are distinct alternating bands of pale and dark tissues (Fig. 1 ) and
may be delimited by the narrowest radial diameter of the tracheids in the dark bands

Fig. 1. Photographs of two cores from different trees of Callitris macleayana. Tree A was
readily matched with other fast-growing trees and water availability records. Note the double
ring of late wood formed in 1974 when two dry periods occurred. In tree B only the more
recent growth rings are readily matched with other trees: this tree was not included in the
statistical analyses. The white lines traversing the cores are fractures and scratches on the
wood surface.

(Fig. 2a). In trees with ring widths of more than 1-2 mm, patterns of rings are easily
matched between cores from the same and different trees. The rings may be matched by
using either ring widths or the pattern of narrow late wood tissues. The late wood tissues
vary in minimum tracheid diameter, width of tissue and occasional narrow bands of wider
tracheids within a late wood zone.
The match between different trees indicates a dominating seasonal climatic component
in the trees' growth. Solar radiation varies seasonally from c. 1400 to 2300 J cm2 day-',
mean daily temperatures range from c. 15 to 25°C and rainfall is very variable with c.
1200 m m from January to March and only 120 mm from August to October. When the
Tree Rings in Callitris macleayana
 J. Ash

trees were cored in July 1981 the temperatures and solar radiation were at their annual
minima but the tracheid diameters at the cambium were still relatively wide; thus it seems
that low temperatures and radiation are not causing late wood formation. The dry spring
period is more likely to cause a reduction in growth rates and ring formation. The water
availability index for Malanda fluctuates greatly each year and in some years there are
additional wet periods during the dry season (Fig. 2b).
The tracheid diameter patterns and water availability pattern may be readily matched,
with atypical anatomical patterns corresponding to years with atypical patterns of water
availability such as 1953, 1959 and 1960, 1974 and 1975 (Fig. 2a, 2b). This matching
indicates that the late wood tissues correspond to periods of low water availability. Overall,
the matching is clear but not all the minor fluctuations in one pattern correspond to the
other pattern. The mean ring widths for both cores from the five fastest growing trees
were calculated (Fig. 2c), as were the duration of wet seasons (Fig. 2d) and the annual
rainfalls (Fig. 2e). The Pearson correlation between the mean ring widths and the duration
of the wet seasons was 0 . 7 4 (P<O.001), and for annual rainfalls was 0 . 5 2 (P<O.002).
These results indicate that the rings are almost certainly annual and that water availability
is a major determinant of growth. The duration of water availability has twice the predictive
power of the annual rainfall because so much rain is lost by run-off. Since the water
availability was calculated for Malanda, with a significantly lower rainfall. no attempt was
made to derive a precise relationship between growth and water availability.
Correlation coefficients were also calculated between mean ring widths and various
annual (October-September) climatic indices recorded at Kairi. Records were not available
for all years so the number of years (n) varied. Annual hours of sunshine had an inverse
correlation of -0 58 ( n = 15, P<O 02), which supports the hypothesis that the duration
of the cloudy (wet) season is controlling growth. The annual mean minimum temperature
had a correlation of 0 39 ( n = 19, P<O 05), indicating that ring widths are greater when
minimum temperatures are higher. Minimum temperatures are generally higher when there
is nocturnal cloud cover, so this correlation supports the hypothesis that a prolonged cloudy
(wet) season favours growth. Neither annual mean maximum temperatures ( r = -0 30,
n = 19), annual mean temperatures (r = 0 14, n = 19), the lowest monthly mean minimum
temperature (r=O 14, n = 20) nor the highest monthly mean maximum temperature
( r = 0 10, n = 19) showed significant relationships with ring widths.
On average, ring widths were about 2 . 5 mm but there was great variation from tree
to tree and within some trees. Growth rates were most variable in small trees, and fastest
in larger trees, with recent ring widths of 3-5 mm. Among the 13 trees, trunk wood
diameters ranged from 97 to 265 mm and ring counts ranged from 24 to 62, but the
correlation between these parameters was only 0 . 1 5 (n.s.). Trunk diameter is therefore a
poor index of tree age, though the relationship probably would improve if the range of
tree sizes were increased.
Discussion and Conclusions
The study of tree rings has virtually been restricted to temperate regions that experience
sufficient climatic seasonality to induce seasonal patterns of growth (Fritts 1976). Though
tropical regions do not experience such great variations in light or temperatures, the water
availability may vary seasonally in a more or less regular fashion. The summer rainfall
regions of northern Australia have seasonal climates and, as this study shows, this is
sufficient to induce seasonal, i.e. annual, growth rings.
The identification of annual growth rings enables the aging of individual trees and the
analysis of past growth in terms of regional, generally climatic, patterns and individual
deviations from these patterns, generally related to competition, disease or herbivores.
In this study, tree ages indicate seed germination shortly prior to the years 1917-1955.
The ages of the trees are consistent with the hypothesis that regeneration became more
successful with the cessation ofAboriginal burning practices in the period 1870-19 15. Older
Tree Rings in Callitris macleayana

trees may have been removed by loggers so this interpretation of the tree ages is not
conclusive.
Since the oldest tree is not older than the rainfall records at Malanda, no attempt was
made to reconstruct past patterns of water availability. The climatic information recorded
in the growth rings may be greater than the calculated correlation coefficients indicate
because these were based upon a climatic station some distance from the trees. If suppressed
trees are included, it may be necessary to remove growth trends from the ring width records,
but this masks climatic trends. It is doubtful whether suppressed trees would exhibit such
direct correlations with climate as are displayed by larger trees, certainly the matching of
ring width patterns is less obvious.
The analysis of growth rings in older stands of Callitris may provide otherwise
unobtainable records of past climate and vegetation change.

Acknowledgments
I thank the Queensland Department of Forests, Atherton, for permission to work at
the study site and the Department of Biogeography and Geomorphology, Australian
National University, for supporting this research.

References
Fitzpatrick, E. A,, and Nix, M. A. (1970). The climatic factor in Australian grassland ecology. In
'Australian Grasslands', ed. R. M. Moore, pp. 3-26. (Aust. Natl Univ. Press: Canberra.)
Fritts, H. C. (1976). 'Tree Rings and Climate.' (Academic Press: London.)
Keig, G., and McAlpine, J. R. (1974). WATBAL: a computer system for the estimation and analysis
of soil moisture regimes from simple climatic data. 2nd Edn. CSIRO Aust. Div. Land Use Res.
Tech. Memo No. 7414.
Lange, R. T. (1965). Growth ring characteristics in an arid zone conifer. Trans. R. Soc. S. Aust. 98,
133-7.
Ogden, J. (1978). On the dendrochronological potential of Australian trees. Aust. J. Ecol. 3, 339-56.
Pearman, G. I. (1971). An exploratory investigation of growth rings of Callitris preissii trees from
Garden Island Naval Base, West. Aust. Nut. 12, 1-7.
Walker, P. J. (1976). Growth and regeneration of trees and shrubs. I n 'Rehabilitation of Arid Lands:
10 Years of Research at Cobar, N.S.W., 1964-1974, ed. G. M. Cunningham, ch. 8, pp. 1-28. (Soil
Conservation Service, N.S.W.)

Manuscript received 18 June 1982, accepted 1 February 1983