Can carbon-13 in large herbivores reflect the canopy effect in temperate and boreal ecosystems? Evidence from modern and ancient ungulates

Abstract

Local environmental conditions under dense canopy are known to result in depletion in ¹³C abundance in plants compared to an open land context. This canopy effect has been observed in tropical as well as in mid-latitude forest ecosystems. However, the impact of the canopy effect on tissue ¹³C abundance of temperate and boreal forest-dwelling herbivores has not been thoroughly explored. Nevertheless, the canopy effect has been suggested to explain a decrease of about 3‰ in collagen δ¹³C values in ancient large herbivores from western Europe during the forest expansion of the Late-Glacial–Early Holocene period (ca. 15,000–6000 cal BP). Some papers have considered the ¹³C decrease in large herbivore as the main result of global change in atmospheric CO₂ content. A detailed review of δ¹³C values of large herbivores (reindeer, red deer, roe deer, and bison) from open and closed environments from high and mid-latitudes confirm that the canopy effect observed in plants is passed on to their consumers. In the Paris Basin, the decline in δ¹³C values of large herbivores at the Late-Glacial/Early Holocene transition around 10,000 years BP appears to be different according to the considered species, namely red deer, roe deer, and large bovines (bison and aurochs). Moreover, differences in the pattern of decrease in δ¹³C values are observed in red deer between French northern Alps and French Jura. These differences among species in their isotopic response through time for a given geographical location, and within species from different locations, suggest variance in ecological responses of species that are associated with the relative use of forested habitat. As a result, ¹³C abundances in collagen can be considered as a direct tracker of the degree of closure of the habitat of ancient herbivores.

Introduction

Habitat use and dietary preferences of ancient herbivores is an important question in palaeobiology. In tropical contexts, the use of carbon isotopic signatures (δ¹³C) of modern and fossil herbivore tissues is nowadays routinely used to document these issues, since herbaceous plants in open environments and herbaceous and arboreal plants from forested contexts use different types of photosynthesis, C₄ versus C₃, which exhibit very distinctive carbon isotopic signatures (e.g. Ambrose and DeNiro, 1986, van der Merwe et al., 1990, Vogel et al., 1990, Cerling et al., 1998). In boreal and temperate ecosystems, where C₄ plants are practically absent, C₃ plants exhibit isotopic contrasts according to environmental conditions (see reviews in Tieszen, 1991, Heaton, 1999, Dawson et al., 2002). Variation in δ¹³C values in C₃ plants has been observed in response to soil moisture, low humidity, irradiance, temperature, nitrogen availability, salinity, and atmospheric CO₂ concentration. Despite these numerous factors influencing ¹³C assimilation by C₃-plants, some systematic patterns are observed between different plant assemblages. One of those patterns is the so called “canopy effect” that corresponds to a vertical gradient in the δ¹³C values of forest trees, with high δ¹³C values at the top of the canopy and low values at the bottom, and to a ¹³C depletion in plants growing under forested closed canopy relative to the same plant type subject to more open growing conditions (e.g., reviews in Broadmeadow and Griffiths, 1993, Heaton, 1999). This ¹³C depletion in plants growing under forested closed canopies may be linked to the combination of two factors: (1) atmospheric CO₂ available to plants in poorly ventilated understory is ¹³C depleted relative to the general atmosphere as the result of CO₂ recycling from leaf litter (e.g. Schleser and Jayasekera, 1985, Gebauer and Schulze, 1991, van der Merwe and Medina, 1991), (2) a CO₂ concentration gradient and light attenuation under the forest canopy leads to depleted ¹³C abundances in understory plants due to change in photosynthetic activity and stomatal conductance (e.g. Francey et al., 1985, Ehleringer et al., 1986, Gebauer and Schulze, 1991, Broadmeadow et al., 1992). Some authors also invoke a higher water availability for plants growing under canopies relative to plants growing in more exposed sites (Broadmeadow et al., 1992, Brooks et al., 1997). Consequently, the intensity of canopy effect is expected to depend on the characteristics of the canopy. General trends are that the more complex and denser the canopy, the greater the extent of light reduction and the degree of recycling of respired forest floor CO₂ within the understory layers (France, 1996). For a significant ¹³C depletion to be observed in arboreal plant formations, two conditions need to be fulfilled: CO₂ produced by plant respiration is confined and the gradient of CO₂ is not dissipated by wind turbulence (e.g., Broadmeadow and Griffiths, 1993, France, 1996, Roche, 1999), and the light attenuation is sufficient thanks to a high Leaf Area Index (Broadmeadow and Griffiths, 1993, Buchmann et al., 1997).

The “canopy effect” causing changes in ¹³C abundances of plants is expected to be passed on to herbivores feeding on understory vegetation. Indeed, some studies tentatively established links between ¹³C abundances in animals and the canopy effect. For instance, depleted ¹³C abundance has been shown in tissues from elephants living in tropical forest compared to those living in savannah (van der Merwe et al., 1990, Vogel et al., 1990), but the occurrence of C₄ plants with δ¹³C values as high as − 12‰ in savanna grass may contribute to an exaggeration of the difference between the δ¹³C values of both populations. Also, the more negative δ¹³C values measured in South American monkeys living in closed canopy forest compared to those living in more open environments presented by Schoeninger et al. (1997) may be partly due to differences in dietary specialisation between the considered species. One recent attempt to document canopy effect in red deer collagen δ¹³C values based on the comparative study of five European populations from C₃ contexts remains unconvincing (Stevens et al., 2006), but this outcome is more due to the difficulty of finding red deer populations unaffected by anthropogenic interferences rather than on the absence of the canopy effect (see discussion in the present paper). Finally, there is no ideal example showing a significant difference in δ¹³C values for two populations of the same species, one living under a closed-canopy forest and the other one living in a more open C₃ environment.

The potential of ¹³C measurement of herbivore tissue for tracking the degree of closure of habitat is an important question to address. Indeed, climatic fluctuations in the past has led to major vegetation changes, as it was the case of the transition from Late-Glacial to Early Holocene (ca 15,000–6000 years cal BP) in western Europe. As a result of global climatic warming, the vegetation composition changed drastically from steppe–tundra dominated by grass and herbaceous dicotyledones to temperate dense deciduous forest through intermediate stages of open boreal-like forests as reconstructed from pollen records (e.g. Amman and Lotter, 1989, Beaulieu et al., 1994a, Beaulieu et al., 1994b). Among game species, arctic-steppe species like reindeer (Rangifer tarandus) and bison (Bison priscus) were gradually replaced by temperate species like aurochs (Bos primigenius), red deer (Cervus elaphus) and roe deer (Capreolus capreolus) after a period of co-existence during the Late Glacial Interstadial. Late-glacial and Early Holocene major change of landscape provoked a change in the subsistence strategies of prehistoric populations linked to the change in the habitat of their game. Therefore, it is of high interest to define the degree of opening of the habitat of ancient herbivores using their δ¹³C values. In this context, the decrease in herbivore ¹³C abundances during Late-Glacial–Early Holocene transition in western Europe has been attributed to the increasing forest cover at this time (Drucker et al., 2003, Noe-Nygaard et al., 2005). In contrast, some papers consider that the decrease in ¹³C abundance of herbivores during this period essentially results from an increase in CO₂ concentration in the atmosphere along with a decrease in plant ¹³C abundance (Richards and Hedges, 2003, Stevens and Hedges, 2004).

We wanted to determine whether the canopy effect is a convincing explanation for the observed decrease in carbon-13 abundance in herbivores during the Late-Glacial/Early Holocene transition. For this purpose, we address the following two questions. First, in modern boreal and temperate environments, can the canopy effect result in depletion of ¹³C abundance in tissues of large herbivores regardless of their forage preferences? Second, during the Late-Glacial and Early Holocene, did the pattern and timing of the decline in ¹³C abundance in herbivores vary according to species and geographical location? Such variation is expected if a local parameter, such as canopy effect, rather than a global parameter, such as atmospheric CO₂ content, is primarily involved in the herbivore ¹³C depletion.