Abstract :
[en] Many exotic tree species have been introduced worldwide for ornamental or wood production purposes, and some have become invasive in the introduction area, representing a substantial threat to native biodiversity. Forestry is a major pathway of introduction of non-native trees (NNT), and this trend is increasing in Europe due to current afforestation programs promoting the use of NNTs to diversify and improve resilience of planted forests in the face of climate change. However, species selected for wood production are often fast-growing species massively planted on suitable sites, with silvicultural practices enhancing survival rates. These factors increase the probability of these species escaping from cultivation and invading natural habitats. We can therefore assume that an invasion debt exists in Europe regarding NNTs, concealed by the long lag-phase needed by trees to go through the introduction-invasion continuum.
Time since introduction and propagule pressure have been proven to increase the probability of naturalization of an NNT. Functional traits and dispersal capacities play a role in later stages of the invasion process. The process of invasion by NNTs has been extensively studied in the Southern hemisphere and with pioneer, light-demanding species such as pines, eucalyptus or acacias, yet there is a lack of understanding of the invasion process of NNTs in dense temperate forests. If these ecosystems have long been considered more resistant to invasions, there is an increasing number of records of shade-tolerant species invading native natural or semi-natural forests. It is therefore crucial to unveil the mechanisms underlying forest invasion by NNTs and to identify potentially invasive species before they become widely planted.
I used a network of eight old forest arboreta in Southern Belgium as sentinel sites to detect potentially invasive tree species. A systematic monitoring of the sites allowed me to gather data on the density, distance and size structure of the natural regeneration of NNTs. Abiotic characteristics of the habitat were also measured. Several NNTs displayed an abundant natural regeneration in the arboreta, which was further enhanced by planting intensity. Some of these species were already known to be invasive, such as Quercus rubra, Prunus serotina and Robinia pseudoacacia. Maple species were also found in dense regeneration patches, especially Acer rufinerve, which is already listed as invasive in Belgium. Most importantly, almost 20% of the frequently planted conifers displayed important regeneration and dispersal potential, and tolerated a wide range of environmental conditions, including shaded understorey, which could lead to the invasion of mature forests. Tsuga heterophylla was particularly prolific, and created dense, impenetrable stands. These maple and conifer species could be part of the invasion debt threatening European forests.
To further investigate the dispersal potential of exotic conifers, the realized dispersal of Tsuga heterophylla, Abies grandis and Thuja plicata was quantified from isolated forest trials. The monitoring of recruitment curves of three of these conifer species confirmed the high invasive potential of Tsuga heterophylla and Abies grandis if planted in favorable sites, especially under coniferous cover. However, Thuja plicata encountered more dispersal and regeneration limitations.
A strategy of fast resource acquisition through high relative growth rate (RGR) and specific leaf area (SLA) has been highlighted in numerous studies comparing native to invasive species, or non-invasive to invasive species. However, this hypothesis has almost always been tested on light-demanding species. The relationship between seedlings developmental traits and invasiveness was tested for the two groups of emerging invasive trees identified in the old arboreta, i.e. Acer and conifer species. To allow a finer analysis, invasiveness was quantified on continuous gradient instead of the usual non-invasive/invasive dichotomy. Global invasiveness was calculated based on proxies extracted from the GBIF Database (number of regions and countries invaded) and the Global Compendium of Weeds (number of citations and risk score). Local invasiveness combined values of regeneration densities and dispersal distances measured in the eight forest arboreta. For maples, invasiveness was positively correlated to growth rates in biomass and height, SLA and number of leaves. For conifers, more invasive species displayed faster height increment, presumably in a strategy of fast light acquisition. A strategy of fast resource acquisition is therefore a key component of the invasion process, even in shaded forest ecosystems.
The results of this PhD indicate that temperate forests are not immune to invasion by non-native trees, and that several tree species might still be in a lag-phase preceding invasion. Combining the results from the monitoring of forest arboreta and the growth experiments, a “forest invasion syndrome” emerges, combining shade-tolerance, high growth rate enhancing competitiveness in canopy gaps and long-distance dispersal ability. High planting intensity and enhanced habitat invasibility through frequent disturbance or silvicultural practices (e.g. abundant coniferous cover) may amplify this syndrome. This is consistent with recent studies indicating invasive tree species in forest adopt a “sit-and-wait” strategy, combining shade tolerance and high response to light after disturbances allowing them to outcompete native trees. Non-native trees, especially those alien to continental Europe, should not be promoted in afforestation program without a thorough risk assessment. Old forest trials and arboreta can act as sentinel sites and provide useful information for a wiser species selection and smarter management practices of temperate forests.
