Aspen clones were traditionally identified based on similarities in several phenotypic traits including leaf shape. This required several visits of the stands, laborious measurements and subjective visual assessments. In this study, we investigated a novel approach to clone identification using digital morphometrics of leaf shape complemented with bark characteristics and spring phenology. Aspen clones were delineated based on molecular (microsatellite loci), morphological (leaf shape, bark colour and type) and phenological (when first fully expanded leaves appeared) characteristics. Leaves were scanned and images were analyzed using normalized elliptic Fourier descriptors and principal component analysis. Using microsatellite loci, 18 clones were identified among 60 aspen trees in three sites investigated in this study. When employing digital morphometrics, foliar types in two out of the three sites matched the clones defined by microsatellite markers. Many ramets from the third site were clustered erroneously into incorrect clones. The reclassification test indicated that leaf shape contains features according to which very similar clones can be differentiated with low error rates. However, because it was not possible to set a threshold for maximum distances within clones, application of digital morphometrics of complex leaf shape for clone identification in natural aspen stands with a high number of multi-ramet clones and many singletons is unfeasible. We judged spring phenology as the least reliable trait for clone recognition and suggested possible causes of heterogeneous leaf flushing among ramets of the same genotype.

Dans cette thèse, nous avons examiné une nouvelle approche pour l’identification des clones de tremble. Nous avons utilisé les liens racinaires clonaux pour l’exploration de leurs fonctions potentielles dans le maintien de la diversité clonale et de la communication entre les tiges interconnectées. En particulier, nous avons examiné les hypothèses suivantes: (1) les clones de tremble peuvent être identifiés sur la base de la morphométrie digitale des feuilles; (2) les liens racinaires clonaux facilitent le maintien de la diversité clonale en intégrant différents génotypes par les greffes racinaires; et (3) les liens racinaires permettent la propagation et déclenchement des mécanismes de défense contre une attaque d’herbivore entre les individus interconnectés avant l’arrivée des insectes. Pour vérifier ces hypothèses, nous avons excavés les systèmes racinaires de trois stations de tremble purs où les clones ont été identifiés à l’aide de marqueurs moléculaires (microsatellites), morphologiques (forme de feuille, couleur et type de l’écorce) et phénologiques (apparition des feuilles). Les feuilles ont été récoltées et scannées. Les résultats ont été analysés à l’aide des descripteurs Fourier elliptiques normalisés (nEFDs). Les analyses dendrochronologiques ont été utilisées pour reconstruire le développement du système racinaire et le greffage racinaire. L’expression temporelle des mécanismes de défense induits (inhibiteurs de la trypsine de type Kunitz et dihydroflavonol réductase) a été suivie par une PCR quantitative en temps réel. Nous avons montré que l'évaluation quantitative de la forme des feuilles est difficilement applicable à l’identification des clones dans des peuplements naturels de peupliers caractérisés par la présence d’un grand nombre de clones et de génotypes uniques. Les systèmes racinaires excavés ne contenaient pas de génotypes uniques et qui n’étaient pas présent dans les tiges des arbres vivants. Néanmoins, l’interconnectivité substantielle à l’intérieur des clones ainsi qu’entre les clones, due à la présence de nombreuses greffes, conduit à la formation de vastes réseaux racinaires génétiquement diversifiés. Nous avons constaté que ces systèmes racinaires n’entraînent pas automatiquement l’induction des mécanismes de défenses, tels qu’ils sont exprimés dans les plantes directement exposées aux attaquesd’insectes phytophages. Il semble plutôt que d’autres moyens de communication, comme les composés volatils, puissent servir de voies de transmission des signaux entre les individus interconnectés.

Trembling aspen (Populus tremuloides Michx.) as a clonal tree species possesses a complex root system through which trees of the same or different clones are connected. Root connections have been studied with respect to resource sharing, but the nature, quantities or extent of what is shared between trees is relatively unknown. In this study, we posed the hypothesis that systemic defense induction signals could also spread through these root networks and trigger defenses in neighboring ramets before arrival of pests. Temporal expression pattern of Kunitz trypsin inhibitor (KTI) and dihydroflavonol reductase (DFR) genes, two markers of poplar defense, was followed by quantitative real-time polymerase chain reaction. The expression was quantified in systemic leaves of wounded and healthy plants that shared the same parental root and in untreated controls grown in separate pots. Untreated interconnected plants did not show induced resistance upon herbivore-simulated attack. Although wound-treated ramets induced defense genes, untreated interconnected plants produced an expression pattern similar to non-connected controls. Root connections do not automatically lead to induction of defensive traits that are expressed in plants directly under damage thought to simulate herbivory. Rather, it seems that other communication means such as airborne volatiles can serve as signal transmission pathways among neighboring plants.

Trembling aspen ( Populus tremuloides ) is a clonal tree species, which regenerates mostly through root suckering. In spite of vegetative propagation, aspen maintains high levels of clonal diversity. We hypothesized that the maintenance of clonal diversity in this species can be facilitated by integrating different clones through natural root grafts into aspen ’ s communal root system. To verify this hypothesis, we analyzed root systems of three pure aspen stands where clones had been delineated with the help of molecular markers. Grafting between roots was frequent regardless of their genotypes. Root system excavations revealed that many roots were still living below trees that had been dead for several years. Some of these roots had no root connections other than grafts to living ramets of different clones. The uncovered root systems did not include any unique genotypes that would not occur among stems. Nevertheless, acquiring roots of dead trees helps to maintain extensive root systems, which increases the chances of clone survival. Substantial interconnectivity within clones as well as between clones via interclonal grafts results in formation of large genetically diverse physiological units. Such a clonal structure can signifi cantly affect interpretations of diverse ecophysiological processes in forests of trembling aspen.