Suitability of the Weber-Gut risk assessment methodology used in Central Europe for determining invasive woody plant species in Estonian historical parks


 The main objective of this research was to assess the suitability of the risk assessment methodology developed by Weber & Gut for Central Europe in order to determine the invasive species in the dendroflora of this region's historical parks. In order to do so a region was chosen where the number of old parks is high (Estonia) and then the following characteristics were studied: 1) the distribution of invasive species; 2) the viability of invasive species. Research results showed that the risk assessment for Central Europe developed by Weber & Gut is not suitable for determining invasive tree and shrub species in the historical parks of Estonia. The research viewed 18 species (classified as a high-risk category) and only half of them had characteristics inherent to invasive behaviour (produces reproductive offspring, often in very large numbers, at a considerable distance from parent plants, and thus has the potential to spread over a considerable area). Even in a small territory (45,339 km2), the regional differences in the occurrence of invasive species were substantial. No independent mass reproduction of most of the invasive species was detected during the use of the Weber-Gut risk assessment system. This research showed that the dangers of invasive species cannot be assessed based on one methodology, but each species should be assessed according to their habitat and specifics of the species.


Introduction
The problem of invasive species Invasive alien species are the second largest reason after habitat destruction for the decrease in biodiversity all over the world (European Commission, 2013). The Euro-pean Union has taken several measures to preserve biodiversity by limiting the spread of invasive alien species. Compared to herbaceous plants the problem of invasive slow-growing and long-living woody plants has not received much attention. Invasion of non-native species is one of the many threats to biodiversity and it is considered to be a major component of global change (Mooney & Hobbs, 2000). The problem of invasive species increases constantly due to the increase in people travelling and goods being transported. Although for a long time, exotic species have been valuable and desirable souvenirs brought home from travels to distant countries but they have also become a problem which is being dealt with at a national level. The shorter the lifespan of a species, the faster it reproduces, hence the greater risk of invasion by alien species. Therefore, the invasion of herbaceous plants is a rapidly growing issue. There seems to be no serious danger when talking about species with a long lifespan, such as woody plants. Alien woody plants have not been considered to be dangerous invasive species (Holm et al., 1997) but the problem has become significant in many regions due to greater distribution and naturalisation of woody plants. Seven woody plants (DAISIE, 2017) are listed amongst one hundred most dangerous invasive species in Europe.

Invasive woody plants in Estonian parks based on the Weber-Gut risk assessment system
Invasive herbaceous plants have been studied more than woody plants (Ööpik et al., 2008;Sander et al., 2008;Raska, 2010;Purik, 2011;Ööpik et al., 2013), but research about woody plants (Elliku & Sander, 1996;Mölter, 2011), manor parks (Uustal, 2003;Nutt, 2008;Sinijärv, 2013;Nurme et al., 2014) and the dendroflora of manor parks (Nutt, 2013;Nutt et al., 2013) have also been carried out. However, the species invading the parks' dendroflora have not been studied. A Master's thesis titled Risk assessment of alien woody plant species recommended for restoration of parks was defended in 2013 at the Estonian University of Life Sciences (Purik, 2013). The thesis drew attention to parks as a hotbed of invasive tree and shrub species reproduction which was considered to be a problem in the restoration of parks. Fifty-three species of alien woody plants in parks were classified into three categories (Purik, 2013) according to their invasion potential in 2013 and on the basis of the Weber-Gut risk assessment system which was previously used in Central Europe (Weber & Gut, 2004). In Estonia, 18 woody plants (Purik & Ööpik, 2013) have been listed as invasive amongst other alien species. In order to evaluate the extent of the problem of invasive species that belong to the high-risk category, the distribution of alien invasive tree species in Estonian parks classified as high risk was analysed in 2013 (Purik, 2013).

Estonian Historical Parks
Estonia with a total area of 45,339 km 2 has about 1,100 manors (Rosenberg, 1994) and about 800 parks which have been preserved. The oldest manor parks were established already in the 17 th century (Maiste, 1996) and a majority of the parks were founded in the 18 th -19 th century .
The use of alien plant species in park design began in the middle of the 19 th century (Sinijärv, 2013). It is known that alien species can be bought from plant nurseries but also brought from trips abroad (Nutt, 2008). Although historical events have had a devastating effect on parks that once were magnificent and many exotic species susceptible to local climate have been destroyed, there are still resistant species which can reproduce without special care. The issue of invasive woody plants in Estonian parks was recognised in regard to the restoration of old manor parks (Purik & Ööpik, 2013).

Description of the area -climate, soil and dendroflora
The distribution, viability and invasiveness of species are largely dependent on natural conditions (Pyšek et al., 2008). Esto-nia is located in Northern Europe (580-600 N, 220-280 E). The territory of Estonia covers 45,339 km 2 and lies on the eastern coast of the Baltic Sea (the coastline is 3,794 km long). The average height is 50 m above sea level (max 317.4 m). Estonia is located in the transitional temperate climate zone and the climate is strongly influenced by the North Atlantic Current, the Baltic Sea and the geographical location of Estonia. Estonia has four seasons and the annual average temperature is about +5 °C (average temperature in February is -4…-5 °C and +18 °C in July). Temperature fluctuations from the average are frequent. Precipitation surpasses the evaporation rate. The annual precipitation rate is between 550-800 mm, on average, and the average relative humidity is 80-83%. In Estonia, the vegetation period is almost 200 days, on average (between 175-195 days), being the shortest in the north-eastern part (less than 180 days) and longest in South Estonia and islands (more than 185 days). (Estonica, 2020).
The characteristic features of soil in Estonia are as follows (Raukas, 1995): 1) abundance of wetland soils; 2) presence of limestone in the soils; 3) abundance of carbonate soils; 4) soils are rich in stones, especially gravel; 5) the soils are colourful due to their granulometric, mineral and chemical composition, and variable water regime. About half of the territory (45.6%) of Estonia is covered with forests (Adermann, 2009) and there are 81 indigenous woody plant species (Kull, 2009).
Several databases were used to assess the distribution of invasive tree species in parks as follows: 1. The analysis of the distribution of invasive species is based on Piret Palm's database which was compiled in 2009 (Palm, 2009). Data was collected from inventories which were prepared during the period of 1977 to 1997. The database contains 370 tree and shrub species in 304 parks all over Estonia (Figure 1). Data analysis was performed using the statistical programming language R. A distribution map was compiled about each species which helped to view the dangers of invasion region (county) by region. Comparing the distribution maps to each other enabled to assess the peculiarities and extent of the dangers of invasion. In order to characterise the frequency of distribution in each region, a percentage of the species' occurrence was calculated in all of these parks, county by county.

The viability of invasive species was
analysed from 2003 to 2009 on the basis of the dendrological inventory data of 17 parks (Table 1). The data used in the analysis included 13,994 trees. Their species and breast height diameter were determined and their age was calculated based on the breast height diameter methodology  which considers the growth speed of a tree species (either coniferous, deciduous hard or soft tree). All the trees were divided into two age groups: old (more than 100 years old) and young (less than 100 years old) . Based on the difference in the number of old and young individuals the species' viability or ability to reproduce was determined. The database included 1,836 coniferous and 12,108 deciduous trees. There were 11,902 native trees and 2,042 exotic trees. Deciduous trees prevailed among native species (11,218 trees), outnumbering coniferous trees (684). There were 1,152 coniferous and 890 deciduous trees among exotic species. Altogether, there were 571 invasive species among exotic species, 549 of which were coniferous and 22 deciduous. In order to evaluate the extent of the problem of high-risk invasive species the distribution of alien invasive tree species classified as high risk in Estonian parks was analysed in 2013 (Purik, 2013). Freeware R was used for data analysis.

Analysis of the distribution of invasive species
Three species of coniferous trees categorised as high risk are abundant in parks.
The most numerous invasive coniferous species is European larch (Larix decidua) which grows in 64% of the parks. Two fir species are a little less abundant (Abies sibirica 55% and Abies balsamea 38%).
Although European larch (Larix decidua) was the most numerous coniferous tree in parks (in 64% of the parks) (Table 2), there were remarkable differences in its regional distribution. In West Estonia and the western islands, the occurrence of the species was less frequent. In addition, there was not a single Abies sibirica in the parks of the southwestern region, including the islands (Figure 2). The same pattern occurred with other coniferous species (Abies balsamea, Larix decidua). Compared to coniferous trees, there were fewer invasive deciduous trees or shrubs in parks (Table 2, Figure 1). The northern region was exceptional because Acer pseudoplatanus was found in over 70% of the parks. Populus alba was present in 80% of the parks on the islands and Amelanchier spicata in 60% of the parks in the southern region. There were only a few samples of Acer saccharinum and Robinia pseudoacacia found in all the parks (Figure 3). The most widely distributed species are mostly shrub species. Caragana arborescens, which belongs to the third risk category, is widespread (found in 80% of the researched parks) (Figure 4). Sambucus racemosa (63%), Sorbaria sorbifolia (56%), Symphoricarpos albus (55%) appear in about half of the studied parks. The most abundant invasive shrub species is Caragana arborescens which occurs in 90% of the parks in different regions. Symphocarpos albus and Sambucus racemosa are also present in over 40% of the parks, and Sorbaria sorbifolia is in 20% of the parks in the western region. Rubus odoratus is not present in the western region and on islands and in other regions can be found in less than 15% of the parks. The smallest number of invasive species is in the western part of Estonia where half of the invasive species are present, and on the islands where only 11 species can be found.
The difference in the presence of coniferous trees in the western region compared to the rest of Estonia is noticeable -two species cannot be found in any of the parks in the western region and only one species occurs in the parks of the islands. The deciduous species Acer pseudoplatanus occurred in the northern region (in 70% of the parks), Populus alba on southwestern islands (in 80% of the parks) and in the southern region. Amelanchier spicata was present in 60% of the parks. Robinia pseudoacacia and Acer saccharinum were present only in small numbers (0-20%). Distribution of the shrub species Caragana arborescens was very high. Also, Symphocarpos albus, Sambucus racemosa, Sorbaria sorbifolia were abundant. However, the distribution of Rubus odoratus and Cornus alba was very small.

Analysis of the viability of invasive species
In order to analyse the severity of the problem of invasion, the parks that had the highest number of species' individuals and invasive offspring present were analysed.
The proportion of young trees growing in parks that have reproduced naturally is very small, which means that many species do not reproduce independently and thus fail to pose an invasive threat ( Figure 5). The number of invasive coniferous and deciduous trees in sixteen different parks was generally under 40.

The number of invasive trees is bigger in Hummuli and Räpina manor park, and in Kukruse manor park which is an exception as there is a European larch (Larix decidua)
alley with 250 trees. The three coniferous trees categorised as invasive (Larix decidua, Abies sibirica, Abies balsamea) produced offspring, hence, had new individuals in these parks which shows that the species are able to reproduce. The same conclusion cannot be made about deciduous trees categorised as invasive because only one species had new individuals in one park (Table 3).

Discussion
This article gives an overview of the invasion problem of slow-growing and long-living woody plant species in the parks of Estonia. When using the Weber-Gut risk assessment system used in Central Europe, the number of alien woody plant species was 53 in 2013, which were divided into three risk categories (Purik & Ööpik, 2013). In order to evaluate the severity of the invasion problem, the plants' species and age were analysed which enabled to determine their distribution and viability or reproduction ability.
The study showed that based on the Weber-Gut risk assessment system used in Central Europe all the species categorised as high risk (total of 18) had not reproduced or produced an offspring in these parks. Therefore, adding them to the list and classifying them as high risk is not justified in Estonia. The species Robinia pseudoacacia, Rubus odoratus, Acer saccharinum, Populus balsamifera, Populus alba, Acer negundo, Acer pseudoplatanus were the only deciduous trees that had new individuals. All the coniferous trees on the third risk category list (Larix decidua, Abies sibirica, Abies balsamea) produced offspring.
Although invasive species may cause big problems on a wide scale, this cannot be observed within the tree and shrub species of old manor parks in Estonia, because most exotic species growing there do not compete with native species. There are species that are called invasive but their offspring do not exceed the new individuals of native species and therefore, they do not dominate the native species.
Bioinvasion goes with other processes triggered by man, such as global changes in the environment and climate, and this is considered to be an equal danger to biota next to over-exploitation, change and fragmentation of habitats and pollution (Kangur et al., 2005). This is a complicated issue that needs more complex research in order to study the impact of local environmental conditions (climate, soil, etc.) on invasive species. Foreign species have been intentionally introduced into the parks when they were built (18 th -19 th century) and therefore, the distribution of species depended on the owner's personal preferences and if they were able to get the foreign species. The first scientific descriptions of Baltic governorates (Pistohlkors, 1797;Friebe, 1805) highlight the non-local tree species and this marks the beginning of the introduction of foreign species into Baltic governorates (Sander & Meikar, 2004).
However, there are still some introduced species (Larix decidua, Abies sibirica) which have been adopted in Estonia and used, for example, in forestry (Keskkonnaminister, 2006). It would be incorrect to claim that avoiding the use of the mentioned species would solve the problem.
The issue of invasive species has to be solved in a complex manner, not based on single parks.
The methodology used to determine the viability of trees cannot be applied to shrubs and therefore, a methodology that considers the specific characteristics of shrub reproduction should be used. The distribution of shrubs in parks has to be evaluated. The database that was available did not include the necessary data. A high percentage of young trees in the parks indicates that they are able to reproduce independently. Due to the fact that there is no diameter data on shrubs, it is not possible to assess the natural renewal of shrubs according to their age and hence, assessment of their reproductive abilities should be based on the shrub population and distribution data.
The severity of invasion cannot be evaluated based on the total number of trees. Although the tree species are categorised as high risk, their viability and offspring have to be analysed. Therefore, a comparison of old and young trees was carried out, which showed that in Kukruse park, there are many invasive tree species, although the number of new individuals is small. However, there are still parks (Hummuli, Kiidjärve, Kuremaa, Räpina, Riidaja, and Saku) where the number of new individuals exceeds the number of old trees ( Figure  5), which represents a species' ability to reproduce naturally.
From the point of view of the aim of this study, it is clear that the problem of invasiveness cannot be addressed the same way everywhere. It has clear regional peculiarities (Figure 2, 3, 4) and the species categorised as high risk are not distributed similarly in all the parks of the studied regions. Also, all the species do not produce new individuals. Corrections have to be made in the invasive species list: 4 tree species (Larix decidua, Abies sibirica, Abies balsamea, Acer pseudoplatanus) and 4 shrub species (Caragana arborescens, Symphocarpos albus, Sambucus racemosa, Sorbaria sorbifolia, Physocarpus opulifolius) need to be classified as high-risk category species.
Undoubtedly, current results are not final as the growth and distribution of woody plants is a long-term process. The first alien woody plants were brought to Estonia in the 17 th century and from then on the number of introduced species has been growing. The results of this research confirmed the assumption that there are regional variations in the distribution of species. The mild climate of northern Estonia might be the reason why these species can be found in large numbers in this area compared to the inland. But without further research this is not certain because the reason may be historical, i.e. the mentioned species have never been introduced into these parks.
Another aspect is connected with the maintenance of parks. However, in the case of invasive species it may be suggested that maintenance does not play an important role because they should be strong competitors for native species. Still, we estimated that the presence of the so-called invasive species should be viewed in connection with the maintenance of parks.

Conclusion
-Our research showed that the risk assessment for Central Europe developed by Weber & Gut is not suitable for determining the invasiveness of tree and shrub species in the historical parks of Estonia. -Only half of the 18 species (determined to fall in the high-risk category) have characteristics of invasive behaviour (producing reproductive offspring, often in very large numbers, at a considerable distance from parent plants, and thus having the potential to spread over a considerably wide area). -Even in the relatively small territory of Estonia covered in our research (45,339 km 2 ), regional differences in the occur-rence of invasive species are noticeable. -Based on the Weber-Gut risk assessment system we did not recognise mass-independent reproduction of most of the tree species determined as invasive. -Our research showed that the dangers of invasion cannot be assessed with a single uniform methodology, because each habitat and species needs to be analysed individually and according to their specifics.