Assessing the degree of variation in stable carbon isotope ratios (δ13C), abaxial stomatal index (SI) and epidermal-cuticular morphology in mature populations of extant species from one plant family and their response to environmental influences, provides a good indication of the reliability of these foliar characters for use in palaeoclimate reconstructions, including estimations of palaeoatmospheric CO2 concentrations.
These foliar characters were determined in:
(1) Six species from two Myrtaceae genera grown in seven common garden experiments in four locations in Queensland, Australia. The climate of these locations ranged from semi-arid to wet tropical.
(2) 70 species from 18 Myrtaceae genera in natural plant communities. The climate of
these sites ranged from wet tropical to temperate.
(3) Cuticle fragments of fossil Lauraceae species from the Early Eocene of Tasmania,
Australia (5 C only), and of a single fossil Myrtaceae morphotype from the Late
Oligocene to Mid Miocene of Southern New Zealand (SI and cuticle morphology
C values in extant Myrtaceae species from natural populations and common garden experiments strongly reflected the mean annual temperature range, mean annual moisture balance and mean daily 9am relative humidity of the site. In contrast, foliar δ13
C values were found to be a poor indicator of mean daily maximum temperature.
Abaxial stomatal index in evergreen myrtaceous species was significantly decreased under shaded conditions within the forest canopy. Significant differences in stomatal index were also noted between populations of some species, which was related to different site climate parameters in separate fiinctional groupings of species. This does not support the widely accepted assumption that the measurement of stomatal index standardises environmental effects on stomatal density for all species.
Abaxial oil gland density of Myrtaceae species was generally not related to environmental parameters. The ecological function of myrtaceous foliar oil glands may instead relate to the activity of the oil's constituents against herbivores and pathogens. Alterations in oil composition between Myrtaceae species and its relative activity may thus determine absolute numbers of foliar oil glands.
Abaxial giant stomata were generally more abundant in Myrtaceae species occurring in rainforest and wet sclerophyll forests compared to species from open woodland and grassland communities. However, no consistent response in giant stomate density to climate parameters was observed between populations of individual species, implying that they may be an ontogenetic anomaly in many Myrtaceae species, without an ecological fianction.
Other noteworthy cuticular features include the striations observed in myrtaceous species from rainforest and wet sclerophyll forest, likely to aid in the shedding of excess leaf surface moisture. Domed cuticular thickenings were also a common feature of Eucalyptus and Corymbia species occurring in open woodland and grassland communities. The degree of prominence of these thickenings varied between populations of individual species and probably acts to reflect the high levels of incident radiation experienced in these plant communities.
Implications for interpreting the plant fossil record
The possibility of diagenetic alterations to leaf δ13C values must be considered. In addition, information regarding the origin of the deposited leaf material must be considered, to discount any influence of physiological stress on plant δ13C values. Sun leaf morphotypes must be distinguished from and used in preference to shade morphotypes to avoid the influence of microclimate effects (e.g. respired CO2 and decreased irradiance) on leaf δ13C. The use of fossil leaf cuticle tissue for δ13C analysis instead of whole leaf tissue removes the complications arising from alterations in leaf δ13C due to leaf age (senescent vs live material from the canopy) and the selective preservation of macromolecules in whole leaf tissue. The influence of changes in δ13CO2 over time can be removed from fossil cuticle δ13C values through conversion to Δ13C (discrimination) values, which may then be useful in identifying fluctuations in moisture balance (and therefore precipitation regimes) over time.
The canopy structure of southern Hemisphere evergreen forests implies that effects of microclimate on stomatal initiation must be considered. Long leaf life spans of species in evergreen forests may also result in considerable variation in stomatal index values encountered in a single species from an evergreen forest at any point in time, due to leaf flushing occurring throughout the year in contrasting seasons and retention of leaves formed in both drought and above average rainfall years (of particular relevance in Australia). The use of SI values for palaeoclimate estimations from fossil dispersed (fragmented) cuticle material is not recommended since it represents an unstandardised sampling and reflects the large inherent variation in stomatal index known to exist over the leaf surface.