Télécharger le fichier original (Mémoire de fin d’études)

Willershausen (Pliocene, -3.2 to -2.6 Ma)

The famous Lagerstätten Willershausen (3.2 – 2.6 Ma; MN 16/17; Hilgen, 1991; Mai, 1995) is an ancient clay mine, now inactive and classified as a natural monument in 1977. Today it is included in the Geopark Harz, Braunschweiger Land, Ostfalen (Germany). The outcrop is located in the village of the same name, in the western Harz Mountains (51.7839°N, 10.1133°E) (Figure 1). Fossils were discovered around 1920’s and researches were conducted during many years by paleontologists from University of Göttingen. More than 50 000 fossils were found including vertebrates (fishes, turtles, amphibians (frogs and giant salamanders)), mammals (including the proboscis Anancus (Mastodon) arvernensis) arthropods (grasshoppers, crickets, bees, etc.) and a huge amount of plants material including the fossil leaves studied here (www.willershausen-harz.de). The main works on the fossil leaves were done by Straus (1977) and Knobloch (1998). Willershausen was a paleoforest included at least 200 angiosperm species including typical taxa from European paleoforests such as Zelkova, Carpinus, Acer, Parrotia, Carya, Quercus, Pterocarya etc. These paleoforest grew under warmer climate than today (ca. +5°C than today). The paleoenvironment was globally warmer than today (c.a. +5°C than today) and recently, a study based on plant species and morphologies of the leaves estimated more precisely mean annual temperature (MAT), mean temperature of the coldest month (CMMT), mean temperature of the warmest month (WMMT) and mean annual precipitation (MAP) (Thiel et al., 2012; Uhl et al., 2007). An artistic illustration of Willershausen can be found in the Appendix A.

Berga (Pliocene, -3.2 to 2.6 Ma)

Berga (3.2 – 2.6 Ma; Mai and Walther, 1988) is an ancient clay pit close the village of the same name in Saxony-Anhalt (Germany). It is located at the edge of the village (Figure 1), 60km in the east of Willershausen (51.4604°N, 10.9817°E). This is a smaller outcrop with approximately 600 fossil leaves including 160 taxa, described by Mai and Walther (1988). Almost all the taxa in Berga paleoforest are similar to the ones from Willershausen and the major ones are Fagus attenuata Goeppert., Alnus gaudinii (Heer) Knobloch & Kvaček, Acer integerrimum Viviani. The forests grew under cooler temperature than the Willershausen, but the climate was still warmer than today (Thiel et al., 2012; Uhl et al., 2007).

Bernasso (Pleistocene, -2.16 to -1.96 Ma)

Bernasso (2.16 – 1.96 Ma; Hilgen, 1991) is located in Hérault (southern France) close to the city of Lunas (43.7261°N, 3.2569°E) (Figure 1). The outcrops include both fossil leaves and pollen material (Leroy and Roiron, 1996; Suc, 1978). The fossil leaf collection contains approximatively 800 specimens with about 20 taxa (Leroy and Roiron, 1996). Majority of angiosperms from Bernasso were also found in Willershausen. The main angiosperm taxa are Carpinus orientalis Mill., Parrotia persica, Zelkova ungeri Spach. The first climatic analysis estimated warmer and wetter condition than today (Leroy and Roiron, 1996). Recently, re-estimation of Bernasso climatic conditions confirmed the high amount of precipitations, but suggested lower temperatures (Girard et al., in review).

Iran (Holocene, today)

The Hyrcanian forest is located in the northern Iran and covered about 50 000 km2 (Akhani et al., 2010). It expands from the Golestan National Park to the border with the Azerbaijan and is contained between the Alborz mountain to the south and the Caspian Sea to the north (Miller, 1994) (Figure 1). The extinct taxa from Europe such as Parrotia or Pterocarya have some plant species retained only in this forest (Akhani et al., 2010; Maharramova et al., 2015). Besides, figure 2 presents the actual distribution of those extinct plant taxa in Europe paleoecosystems. The mean annual temperatures in the Hyrcanian region are around 16°C (Akhani et al., 2010). The climate in this region seems to be globally warmer than the paleoclimate from Willershausen, Berga and Bernasso. The mean annual precipitations are really heterogeneous from west to east, from ca. 1600 mm to 500 mm.
Figure 2: Actual distribution of plant taxa found in Europe there are 3 Ma, based on the statistical plant fossil recorded in Willershausen.
Mapping has been realized by Laure Paradis (ISEM, Montpellier) according to the datasets included in Coombes and Debreczy (2015). As suggested in the thesis and in many different studies, the Caucasian region, especially the Hyrcanian forest in Iran/Azerbaijan, includes the larger quantity of relictual plant taxa presented in Europe during the Pliocene (Arpe and Leroy, 2007; Noroozi et al., 2008; Probst, 1981; Tralau, 1963; Zohary, 1973). The map to the top has subdivided the superposed area distribution of plant taxa in five distinct groups, the map to the bottom has subdivided in 3 groups.

PLANT-INSECT INTERACTIONS

The species diversity and composition of plant communities may vary across different biogeographic zones, which correspond to different abiotic and biotic parameters (Gurevitch and Fox, 2006). Consequently, the structure of the forests based on plant-insect interactions, inter alia, also change (Agrawal et al., 2006). There are two kinds of these interactions.
The entomophile pollination, which is the relationship between insect and plants that that we could qualify as mutualisms (Bronstein et al., 2006). The concept is generally the fecundation of plants by means of insects, in exchange insect mostly feed flowers nectars (or other components in some cases) during the process (Faegri and Pijl, 2013; Gurevitch and Fox, 2006). These mutualistic interactions between plant and insect species has been molded by millions of years of coevolution and already have an accurate attention to the scientific community (Bronstein, 1994; Wappler et al., 2015).
Herbivory (or phytophagy) is the other plant-insect interaction. This is mostly binding for the plant, because the insects collect energy from the plants which lose this energy (Gurr et al., 2003; Singer et al., 2004). Unlike pollination, herbivory can provide a unique record of the plant-insect interactions in the past as their traces on leaf blade are observable on the plants (Labandeira, 2006, 2013). Up to now, arthropods/insects and plants have been often considered separately, this work is a rare opportunity to develop researches on biotic interactions through insect feeding traces.

Relationships and correlations between plant-insect interaction and environmental factors

Fossil leaves are known to be potential indicators of past climates (Wilf, 1997). Insect feeding observable on fossil leaves could be an important source of information (Labandeira, 2006). Thus, relation between abiotic factors (such as climate) and ecosystems structure could be partly estimated by investigations on plant-insect interaction (e.g., Labandeira, 2002; Lewinsohn and Roslin, 2008). Researches highlighted some possible association between climatic parameters and plant-insect interactions.
Fernandes and Price (1992) showed a negative correlation between amount of precipitations and galling as the fungi (more efficient in wetter conditions in most cases) are the main organisms responsible of gall-makers insects death (Carroll, 1988). On the contrary, leaf mining seems to be better adapted in wetter environments (e.g., Yarnes and Boecklen, 2005). The impacts of the temperature are diverse on plant-insect interaction. However, variation of temperature depends on many factors (such as CO2 atmospheric concentration), then the impacts of temperature on the plant-insect interactions are still unclear. Generally, warmer temperatures allow an increase in insect development and reproduction rate (Ayres and Lombardero, 2000). Moreover, Virtanen and Neuvonen (1999) measured on Macrolepidoptera in Finland that warmer temperatures lead to an increasing of insect species diversity. Consequently, it expected to observe higher insect feeding consumption on leaves (e.g., Lemoine et al., 2014; O’Connor et al., 2011; Zvereva and Kozlov, 2006). Furthermore, insect feeding is mostly related to different plant defense strategies based on chemical components that many abiotic factors such as temperatures and atmospheric CO2 concentration could disturb (e.g., Zavala et al., 2008). Indeed, studies have shown that elevated CO2 concentrations tend to decrease jasmonic concentration (plant defense hormone) on plants, then increasing exposure to insect feeding (in: DeLucia et al., 2012). On the contrary, the high CO2 concentration could also increase salicylic acid concentration potentially allowing greater defense against herbivores (in: DeLucia et al., 2012). Moreover, a combination of CO2 elevation with temperature elevation lead to decrease the nitrogen (N) concentration in the leaves (Bezemer and Jones, 1998; Stiling and Cornelissen, 2007). That is resulting a less palatability of the leaf (Stiling and Cornelissen, 2007) and causing insects to feed on more foliage to compensate this diminution of nutritional resources (Bezemer and Jones, 1998). Thus, consequences of high temperaturse combined with elevated atmospheric C02 concentration on the plant-insect interaction is hardly predictable. Furthermore, the high specialization of some herbivorous insects on plants reveals long time persistence of some of plant-insect interactions (Labandeira, 2006, 2002; Wilf, 2008). It could potentially disturb some interpretations in the fossil record between plant-insect interaction and abiotic factors such as climate. Investigations on paleontological studies could provide better indications on relations between insect feeding traces and climate (e.g., Currano et al., 2008; Wappler et al., 2009; Wilf et al., 2001; Wilf and Labandeira, 1999). For example, a study showed during the Paleocene-Eocene Thermal Maximum a positive correlation between herbivory and the combination of both high temperature and atmospheric CO2 concentration (Currano et al., 2008, 2016). However, most studies on the changes of plant-insect interaction according to the climate have not paid attention to how organisms could acclimatize on temperature and CO2 changes (Shaw and Etterson, 2012).
Furthermore, it is important to note that plant-insect interactions could also be related to some biotic factors. A high diversity of plants should involve a higher diversity of plant-insect interaction (e.g., Price, 2002). Fernandes and Price (1988) suggested that the richness of gall-making insect is often positively correlated with the plant species richness, a correlation later detected in South Africa by Wright and Samways (1998). It has also been shown that a richness of gall-inducing insect species correlates to the forest age and health (Fernandes et al., 2010).

Reconstruction of the plant-insect interactions in the fossil record

In order to identify, compare and discuss difference of plant-insect interactions, it is indispensable to adopt a standardized approach, regardless of the locality studied. The preservation of an insect trace represents one behavior of an insect (Genise, 2017). It reflects how insect interacted with its environment that can correspond, for example, to traces of movement, of reproduction and feeding (Genise, 2017). In the fossil record, these traces, and especially the insect feeding traces, can be preserved and observable on the fossil leaves (Genise, 2017). These plant-insect interactions can provide the sole evidence for the paleoecological interpretation or they can be used to supplement other fossil resources (such as pollen, wood charcoal, etc.). Nevertheless, until the late twentieth century there was no formal icnotaxonomic system to designate traces of insects on fossil leaves. Previously, some authors named taxonomically certain traces but these were not grouped into a system of hierarchical structures (Amerom, van H.W.J., 1966; Friĉ, 1901; Kinzelbach, 1970; Straus, 1977).
Nowadays, three classification systems exist to label fossil traces with different approaches, and even classified under different systems.
The first was established by Vialov (1975). He decided to include all the traces that showed the consumption of plant matter made by insects in the group Phagophytichnidea, which in turn was further divided into two subgroups Phagophytichnida (with three families) and Phagophylignichnida, to differentiate those damages produced in leaves of those produced in fossil woods.
Another method was proposed by Krassilov and Rasnitsyn (2008), which includes 72 icnoespecies that include oviposition, galls, mines, buds and exceptional examples of non-marginal excisions. The last, and most common method, corresponds to an informal classification system proposed by Labandeira (2002), that has been enhanced later (Labandeira et al., 2007). In this case, the classification is based on the concept of a guild (in the sense of a functional grouping) and has proved to be a valuable tool in ecology (e.g.,Simberloff and Dayan, 1991). In this classification, there are categories that group the traces according to the feeding behavior of the insect (Labandeira et al., 2007). These damage types (DT) can be classified into seven functional feeding groups: hole feeding (HF), margin feeding (MF), skeletonization (S), surface feeding (SF), galling (G), mining (M), and piercing-and-sucking (PS). All of these interactions are considered to be herbivory because they use plant tissue as a ressource (Labandeira et al., 2007). The DTs can also be classified as most likely to be specialized (made by monophagous and/or olyphagous insects that consume only one or a few closely related plant species) or generalized (typically made by polyphagous insects, e.g., most of external damages such as HF or MF) (Labandeira et al., 2007). Specialized feeding is delineated by evidence from extant analogue feeders, morphologically stereotyped damage patterns, and restricted occurrences confined to particular host-plant species or tissue types in either fossil or extant host taxa (Labandeira, 2002). Moreover, today 70% of insect damage results from monophagous or oliphagous insect (Bernays and Chapman, 1994) that making their consideration particularly important. This system also allows to make the distinction in the dataset between the damage richness (i.e. the number of different DT represented in a dataset) and the damage frequency (i.e. the relative quantity of DT per dataset). This system, because of its completeness and of its simplicity, was used in the complete study presented in this manuscript.

AIM OF THE THESIS

The general purpose of this thesis is to better understand differences in plant-insect interactions in three past environments. This study allows for a better understanding of the insect damage patterns observable in paleoenvironments. For this goal, the study focuses on European paleoenvironments from the Late Pliocene and the Early Pleistocene. It allows accurate comparisons with the plant-insect interactions knowledges on both current and fossil record. This thesis is built according to three geographical areas which were linked to three different time period observations.
“Chapter 2” (Adroit et al., 2016) is focusing in Bernasso (southern France). Patterns of interactions in this outcrop was correlated to climatic conditions inferred for plants (Leroy and Roiron, 1996; Girard et al., in review). The results provide first analyses of plant-insect interaction pattern for the early Pleistocene. In parallel, a part of this chapter debates on different fossil leaves preparations and the consequences of them.
“Chapter 3” (Adroit et al., submitted a) focuses on relative interpretations of plant-insect interactions in the outcrops of Willershausen and Berga, both from the Late Pliocene of Germany). Results on Bernasso (Chapter 2) were also used for comparison. These outcomes better highlight interpretations of abiotic parameters and taphonomic bias are also discussed.
“Chapter 4” (Adroit et al., submitted) focuses on the analyze of plant-insect interactions on modern and fossil leaves of Parrotia persica. Modern leaves of P. persica were collected in the Hyrcanian forest (northern Iran). Results from this study allow the comparison of interaction pattern on Parrotia leaves under different current and past climatic conditions.
Additionally, a complementary study (“Chapter 5”) about the discovery of a new DT exclusively found on P. persica in the Willershausen paleoforest and the Hyrcanian forest is in progress (Adroit et al., in prep).
This thesis ends by a Chapter 6 that try to mixed the information obtained in the 3 first chapters and to answer to the thesis problematic.

Table des matières

CHAPITRE I: GENERAL INTRODUCTION 
1.1 STUDIED AREAS
1.1.1 Willershausen (Pliocene, -3.2 to 2.6 Ma)
1.1.2 Berga (Pliocene, -3.2 to 2.6 Ma)
1.1.3 Bernasso (Pleistocene, -2.16 to -1.96 Ma)
1.1.4 Iran (Holocene, today)
1.2 PLANT-INSECT INTERACTIONS
1.2.1 Relationships and correlations between plant-insect interaction and environmental factors
1.2.2 Reconstruction of the plant-insect interactions in the fossil record
1.3 AIM OF THE THESIS
CHAPITRE II: BERNASSO, A PALEOFOREST FROM THE EARLY PLEISTOCENE: NEW INPUT FROM PLANT-INSECT INTERACTIONS (HÉRAULT, FRANCE) 
2.1 ABSTRACT
2.2 INTRODUCTION
2.3 MATERIAL & METHODS
2.3.1 Study area
2.3.2 Fossil leaves collection
2.3.3 Fossil analysis and classification of damages
2.3.4 Statistic analysis
2.4 RESULTS
2.4.1 Damage diversity and distribution on the leaf species
2.4.2 Conservation bias
2.5 DISCUSSION
2.5.1 Representative and taphonomic bias
2.5.2 Damage abundance on the paleoforest
2.5.3 Community structure of the paleoforest: ecological parameters
2.6 CONCLUSION
2.7 REFERENCES
CHAPITRE III: PLANT-INSECT INTERACTIONS PATTERNS IN THREE EUROPEAN PALEOFORESTS OF THE LATE-NEOGENE – EARLYQUATERNARY
3.1 ABSTRACT
3.2 INTRODUCTION
3.3 STUDIED AREA
3.3.1 Willershausen, Lower-Saxony, Germany
3.3.2 Berga, Thüringen, Germany
3.3.3 Bernasso, France
3.4 DATA ANALYSES
3.4.1 Plant-insect interaction identifications
3.4.2 Leaf Mass per Area
3.4.3 Statistical analyses
3.5 RESULTS
3.5.1 Comparisons of plant-insect interactions
3.5.2 LMA per species at Willershausen
3.5.3 Structure of the paleoforests with the damage distribution on plant species
3.6 DISCUSSION
3.6.1 Floristic richness and herbivory representativeness
3.6.2 Absence of correlation between LMA and plant-insect interaction
3.6.3 Parameters involved in insect damages differences
3.6.4 More precisions provided by proportion of
generalized/specialized damages
3.7 CONCLUSION
3.8 REFERENCES
CHAPITRE IV: CHANGES ON HERBIVORY PATTERN ON THE PERSIAN IRONWOOD (PARROTIA PERSICA) OVER THE LAST 3 MILLION YEARS 
4.1 ABSTRACT
4.2 INTRODUCTION
4.3 MATERIAL AND METHODS
4.3.1 An overview to the study area
4.3.2 Characteristics of the Persian ironwood (P. persica)
4.3.3 Sites and sampling method
4.3.4 Damage types
4.3.5 Leaf characteristics
4.3.6 Statistical tests
4.4 RESULTS
4.4.1 Measurements of Persian Ironwood leaves
4.4.2 Overall patterns of herbivory
4.4.3 Distribution of the plant-insect interaction
4.4.3.1 Margin feeding
4.4.3.2 Hole feeding
4.4.3.3 Skeletonization
4.4.3.4 Surface feeding
4.4.3.5 Galling
4.4.3.6 Mining
4.4.4 Patterns of plant-insect interaction among the locations
4.4.5 Detailed variation of plant-insect interaction along both altitudinal and longitudinal gradients in Iran
4.4.5.1 Frequency according to the gradients (altitude, longitude)
4.4.5.2 Richness according to the gradients (altitude, longitude)
4.5 DISCUSSION
4.5.1 P. persica similarities between the locations
4.5.2 Leaf-removed area
4.5.3 Variation of plant-insect interaction along abiotic gradients
4.5.4 Comparison of herbivory pattern though time
4.6 CONCLUSION
4.7 REFERENCES
CHAPITRE V: A new damage type on fossil and modern leaves of Parrotia persica 
5.1 PREAMBLE
5.2 INTRODUCTION
5.3 MATERIAL & METHODS
5.3.1 Leaf materials
5.3.2 Measurements
5.3.3 Statistics analyzes
5.4 RESULTS
5.4.1 Overview of the sampling
5.4.2 Morphological description of the new Damage Type
5.4.3 Measurements of DT297 in fossil and modern records
5.4.4 Enlarger version of DT297foA in the fossil record: DT297foB
5.5 PRELIMINARY DISCUSSION AND CONCLUSIONS
5.6 REFERENCES
CHAPITRE VI: DISCUSSIONS AND CONCLUSIONS 
6.1 EVIDENT CAUSES ON THE DIFFERENCES OF PLANT-INSECT INTERACTION IN THE EUROPEAN PALEOFORESTS
6.1.1 Potential consequences of climate variations on fossil leaves
6.1.2 Consequences of coldest temperatures on herbivory rates
6.1.3 Consequences of thermic and hydric seasonality on specialized plant-insect interactions
6.2 LIMITS FOR INTERPRETATIONS
6.2.1 Biotic parameters causing changes on insect damage rates
6.2.2 Taphonomic and methodological biases
6.2.3 Plants physiology
6.3 NEW PERSPECTIVES
6.3.1 Increasing knowledge on the history of the European paleoforests
6.3.2 A better understanding of climatic impact
6.3.3 Improving comparison of plant-insect interactions among paleoforests
6.3.4 Outlooks
GENERAL REFERENCES 
INDEX FOR FIGURES & TABLES 

Télécharger le rapport complet