Individual and combined effects of tree stand attributes and skid trail area on ground flora diversity at stand scale

Get Complete Project Material File(s) Now! »

The effects of tree stand on ground flora diversity

In managed forests, the choice of tree species is one of the forester’s fundamental acts. Tree species richness and dominance is considered as a biodiversity indicator (MCPFE, 2003; Barbier et al., 2008). Forest age influences the chemical and structural properties of soil, and consequently affects understory vegetation (Honnay et al., 1999; Dupouey et al., 2002). Tree abundance (represented by basal area, stem density) is shown in numerical studies to have negative effect on understory vegetation diversity (Alaback and Herman, 1988; Thomas et al., 1999). The preservation or absence of a subcanopy layer greatly modifies understory vegetation, especially by modifying understory light (Kwiatkowska, 1994; Nagaike et al., 1999). Tree species composition also influences ground flora diversity, though the results are inconsistent. For example, in comparison of understory diversity in coniferous trees with deciduous trees, there were 10 results with higher understory richness under hardwoods and 4 results with higher SR under conifers (Barbier et al., 2008). Besides, fewer studies compared hardwood (or coniferous) species internally.
It is difficult to make generalizations on the effect of tree species on understory diversity, the effects of tree species on ground flora diversity vary greatly and some results are even conflicting (Barbier et al., 2008). This may be partly due to other factors not taken into account in most studies, especially those related to site characteristics and management practices (Barbier et al., 2008). More precise indications on past land use (especially former agricultural land or forest land), forest history (tree age and past tree composition), tree regeneration method (natural or plantation, e.g. Fahy and Gormally, 1998), thinning intensity in the last decades (e.g. Nagaike, 2002) would be useful for clarifying how a tree species acts through specific management practices (Barbier et al., 2008). Given that silviculture and the growth of overstory can profoundly affect the composition and development of understory species (Alaback and Herman 1988; Stewart 1988; Bailey et al., 1998; Thomas et al., 1999), it is important to understand what the relationships between overstory and undertory species are and, more importantly, how silviculture regimes would affect these relationships (He, 1999).

Are the dominant factors affecting the ground flora diversity different among ecological groups and among individual species?

The classification of the ground flora into ecological groups is a basic and important step to better document biodiversity responses. Relationships between ecological or functional groupings of plant species and environmental gradients can provide evidence for environmental filtering, particularly when the traits suggest an advantage in the associated environment (McGill et al., 2006; Burton et al., 2011).
In this study, we hypothesized that the response of ground flora to the ecological variables can be affected by the species traits as follows: life form, seed bank strategy, light and moisture requirements and successional status (data source: Julve, 2007; Hodgson et al., 1995). Tree regeneration on skid trails was often investigated in previous studies. Seed bank was considered to be an important potential seed source for the restoration of plant communities (Bakker & Berendse 1999), which has been shown to be related to the ground vegetation response to skid trails (Roovers et al., 2004; Godefroid and Koedam, 2004). Light and moisture requirements as well as successional status are also basic plant traits widely used in studies of ground flora diversity (Brockerhoff et al., 2003; Jennife et al., 2005; Fierke and Boone Kauffman, 2005).

Is the influence of skid trails different at stand scale and at fine scale?

It is useful to understand the different role of skid trails at two scales, which could provide useful information for skid trail design and forest management such as: a) at stand scale, distinguish the sensitive stand to skid trail system, and estimate the appropriate area or density for different stand types; b) at local scale, find out the suitable width, spacing of skid trails under different types of forest. In addition, it could help identify the scale that was most affected by skid trail disturbance.

The best indicators: stand type and basal area

In our study, stand type was the best indicator of ground flora richness (Table 3.2). For abundance, either stand type or basal area of tree species was best, depending on the ecological traits of the ground flora (Table 3.3). For example, stand type best indicated peri-forest species while the basal area of the main tree species best indicated forest and non-forest species abundance. In a similar vein but with a slightly different species mix, Barbier et al., (2009) found that basal area was a better model in French oak-hornbeam lowland forests than models incorporating tree species richness or evenness at the of 400-m2 plot scale. In previous studies, age, stand type and basal area or other tree stand variables have been directly or indirectly found to be important factors impacting understory diversity (Nagaike et al., 2005; Nilsson et al., 2008; Sciama et al., 2009; Skov 1997) in managed forest, but few (Barbier, et al., 2009) compared those important variables to detect which one might be the best indicator under the multiple hypotheses framework (Chamberlin, 1965). The effects of skid trails were weak, and the associated models were far from the tree stand attributes models.

The effect of tree stand variables

In the Montargis forest, the majority of current standard-with-coppice stands have not been managed as SWC for decades and are gradually being converted to high forest systems. Generally, the SWC undergo two stages to become a mature high forest: 1) every ten years, low intensity improvement cuts are carried out to maintain 30 mature oaks per hectare; 2) during a final 10-year period, three to four regeneration fellings remove the shelter of mature trees to trigger natural oak regeneration (Jarret, 2004). After that, the oak saplings gradually develop into mature high forest. A high-forest rotation is typically 180 to 200 years (Jarret, 2004).
Previous studies have indicated that conversion from SWC to high forest causes high species loss and a decline in ground flora abundance (Baeten et al., 2009; Brewer, 1980; Rooney & Dress, 1997; Van Calster et al., 2007; Van Calster et al., 2008). The SWC forest can provide a wide variety of environmental conditions (e.g. light, temperature, soil acidity) due to regular harvesting or different rotation cycles among stands; this leads to high species diversity in SWC forests (Ash and Barkham, 1976; Packham et al., 1992). In our study, though the SWC stands in the Montargis forest are no longer regularly cut, they still tend to maintain more species than the mature even-aged stands. This higher ground flora diversity is mainly due to the high richness and abundance of the peri-forest and non-forest successional groups or the intermediate-light species group. Though it was not the case in our study, some studies comparing diversity between typical SWC forests (with regular cutting) and high forests found more heliophilous species in SWC stands, such as Lonicera periclymenum L., Rubus fruticosus agg. and Ajuga reptans L. (Baeten et al., 2009; Brown and Oosterhuis, 1981). Other ecological groups which were not included in our research such as vernal species and seed banking species have also been found to prefer the SWC disturbance regime (Ash and Barkham, 1976; Peterken 1981; Rackham 1975).


Fine-scale variation of environmental factors

PR was significantly greater on wheel tracks (TR) than on controls in STP50 and STP63 and between the two wheel tracks (BE) in STP63 (Fig. 4.3). Nsam and BD were also significantly higher on TR in STP63. PR, Nsam and BD on the subplot location of BE and TR increased with the increasing age of high stands. No significant variation in MaxD among subplot locations was detected (Fig. 4.3). BD, Nsam and MaxD significantly associated to PR (P<0.001). Light and soil moisture did not vary among subplot locations whatever the stand type (Fig. 4.4), but varied among stand types (P<0.001).

Ecological group level

The best models fell into two broad categories (Table 4.3): models related to subplot location that indirectly represent the disturbance gradient and models related to micro-site factors of soil compaction degree, soil moisture or light. Among the total 10 ecological groups, for richness, 2 groups (tree and short-term seed bank species) had their best models related to the interactive effects of subplot location and stand type, 4 groups (herb, shrub, long-term seed-bank and heliophilous species) were best related to soil compaction indices (PR, MaxD or BD) and 4 groups (transient seed bank, high-humidity, low-humidity and shade-tolerant species) to soil moisture (Table 4.4). For abundance data, 3 groups (long-term seed bank, high-humidity, shade-tolerant) had their best models that related to subplot location, 3 groups (tree, shrub, and heliophilous) related to the interactive of subplot location and basal area, 2 group related to PR (shrub, transient seed-bank) and 2 groups related to soil moisture (herb and short-term seed bank) (Table 4.5). Light-only (L) models were preferred to alternative models for the richness of heliophilous species and the abundance of long-term seed-bank species (Tables 4.6 and 4.7). Models combining light and stand type (linear and quadratic models) were the best for the richness of herb species.
For richness data, the magnitude and negligibility of the effects estimated from the best models are shown in Tables 4.8 and 4.10. Subplot locations of BE and TR had positive effects on tree and short-term seed bank species in STP50, which was the same case for the subplots of BE, TR and BO on tree and short-term seed bank species in STP63. Soil moisture showed positive effects on the richness of all the ecological groups which had best models related to soil moisture (transient seed bank, low-humidity, high-humidity and shade-tolerant species). For compaction indicators, BD had a positive effect on heliophilous species richness. The effect of MaxD was either weak (long-term seed bank species) or uncertain (shrub). PR effect was also weak (herb). The effect of light was either weak (herb) or uncertain (heliophilous).
The transition of stand type from STP30 to STP50 had negative effect on shrubs in the additive models of MaxD and STP, while it had positive effect on herb richness in the additive models of PR and STP. The transition of stand type from STP50 to STP63 had positive effect on herbs in the model combining light and stand type, while it was weak for herb and uncertain for shrub.

Table of contents :

Chapter I Introduction
1.1 European temperate forest management, mechanisation and ground flora diversity
1.2 Skid trails
1.3 The effects of tree stand on ground flora diversity
Chapter II Research objectives
2.1 General research objective
2.2 What are the respective single and combined effects of tree stand attributes and skid trail area on understory diversity at stand scale?
2.3 What is the relative importance of subplot location, soil moisture, soil compaction light and stand type on ground flora diversity at fine scale?
2.4 Are the dominant factors affecting the ground flora diversity different among ecological groups and among individual species?
2.5 Is the influence of skid trails different at stand scale and at fine scale?
Chapter III Individual and combined effects of tree stand attributes and skid trail area on ground flora diversity at stand scale
3.1 Abstract
3.2 Introduction
3.3 Materials and methods
3.4 Results
3.5 Discussion
3.6 Conclusions
3.7 Supplementary Material
Chapter IV Plant diversity patterns on skid trails in high forest oak stands of different ages and links to soil moisture, soil compaction and light 
4.1 Abstract
4.2 Introduction
4.3 Material and methods
4.4 Results
4.5 Discussion
4.6 Conclusion
4.7 Appendix
Chapter V Discussion
Chapter VI Conclusion


Related Posts