Springer Old Growth Forests - Chapter 10 doc

Table 10.1 Potential attributes of belowground structural complexity structural or fine roots Depth in metres Variation in maximum rooting depth between species Vertical distribution of

Rooting Patterns of Old-Growth Forests:

is Aboveground Structural and Functional

Diversity Mirrored Belowground?

Jurgen Bauhus

10.1 Introduction

When we think of old-growth forests, we generally imagine old forests with large trees and possibly a highly diverse forest structure resulting from the death of individual trees and the resulting gap-phase dynamics (Oliver and Larson 1996; Franklin et al 2002) This is also reflected in the definitions of old-growth forests (see Chap 2 by Wirth et al., this volume), which normally do not refer to belowground structures and processes This neglect of belowground aspects, although intriguing, is understandable since so little information is available

It is well known that the species richness, and often also the biomass, of inverte-brates, fungi, and bacteria is much higher belowground than aboveground (e.g Torsvik et al 1990), yet we know very little about how their diversity and abundance belowground is related to forest age or forest structure Carbon storage

is an important value of old-growth forests, and the carbon stored in soils, forest floor, and belowground biomass often approximates the quantities stored in aboveground biomass (e.g Trofymow and Blackwell 1998; see Chap 11 by Gleixner et al., this volume) The turnover of fine roots is believed to be an important driver of soil carbon accumulation, yet little is know about how this process changes with forest age or forest developmental stages found in old-growth forests

Many of the attributes and values of old-growth forests are related to their aboveground structural diversity (McElhinny et al 2005) It is therefore interest-ing to ask whether this structural diversity in old-growth forests is mirrored belowground, and to what extent belowground structural diversity may contribute

to functional diversity To approach these questions, I will ask firstly which attributes may comprise belowground structural diversity, whereby the term belowground encompasses the substrates colonised by roots, the soil and forest floor layers

10.2 What Comprises Belowground Structural Diversity?

Since belowground structural diversity has not been defined, I will first reflect on the attributes that commonly are used to quantify structural diversity aboveground Stand structural diversity is a measure of the number of different structural attri-butes present and the relative abundance of each of these attriattri-butes, as summarised

by McElhinny et al (2005) for some forest types These attributes are responsible for variation in the vertical and horizontal structure of forest stands and are thought

to be indicative of biodiversity, i.e related to the provision of faunal habitat The range of attributes comprises vertical layering of foliage; variation in canopy density (e.g caused by gaps); variation in the size and distribution of trees; the height and spacing of trees, and their species diversity and biomass; the cover, height and richness of the understorey including shrubs; and the number, volume and range in decay stages of standing and fallen dead wood (McElhinny et al 2005) The equivalent belowground attributes that may be important for below-ground biodiversity and ecosystem functioning are listed in Table 10.1 In general terms, these structural elements provide horizontal and vertical heterogeneity in belowground structures In the following, some of the analogies between below-ground and abovebelow-ground structural elements or parameters will be pointed out and discussed, focussing in particular on the question of whether the structural diversity created by these elements increases with forest age

Table 10.1 Potential attributes of belowground structural complexity

structural or fine roots

Depth in metres Variation in maximum

rooting depth between species Vertical distribution of fine

or coarse roots

No of layers, evenness

of root distribution

in layers/horizons over the profile depth

Quantity and size of belowground coarse woody detritus (stumps, old root channels, etc.)

Number, volume and depth

features

10.3 Root Gaps and Horizontal Variation in Rooting Density

in Old-Growth Forests

Old-growth forests are the result of the long-term absence of stand-replacing dis-turbances (see Chap 2 by Wirth et al., this volume) The old trees may comprise the cohort that developed following the previous stand-replacing disturbance or they may have developed subsequently in gaps created by the death of trees from the distur-bance cohort In any case, by the time an old-growth developmental phase is reached, the disturbance dynamics, until then, will have been dominated by the formation of gaps in most types of forests Therefore, gaps are an important feature of old-growth forests They may occupy 5 15% of the stand area in temperate forests (e.g Runkle 1982; Emborg et al 2000) These gaps contribute greatly to the vertical structural diversity of old-growth forests and also, through the successional processes trig-gered within them (e.g Busing and White 1997; Rebertus and Veblen 1993), to species diversity Some canopy species may not be able to regenerate in these gaps and are replaced by more shade-tolerant species (Oliver and Larson 1996; Gilbert 1959) However, the prevalence of gap dynamics in old-growth forests leads to a patchwork of developmental stages, creating horizontal variation in canopy height, tree biomass and necromass and/or species composition (e.g Emborg et al 2000; Korpel 1995; Franklin et al 2002) Whether or not the aboveground structural diversity of old-growth forests attributable to gap formation is mirrored below-ground depends on whether root gaps are created in the process, and on whether belowground structural elements vary with the developmental stage of the patches This would, for example, be the case if these gap phases have different levels of fine- or coarse-root biomass, or roots with functional traits or mycorrhizal associa-tions that differ from those found during other developmental stages

First, I will explore the question of whether aboveground gaps create below-ground gaps Then I will ask to what extent belowbelow-ground gaps contribute to structural and functional diversity

Aboveground gaps are created when one or more trees participating in the main canopy layer are removed or killed While the foliage of trees within the crown is confined to a reasonably small projected ground area around the trees, the same is not true for roots These are far more wide-reaching than the branches (Stone and Kalisz 1991), and so roots tend to overlap much more than crowns For example, Bu¨ttner and Leuschner (1994) found complete spatial overlap of the root systems of the co-occurring speciesQuercus petraea and Fagus sylvatica Consequently, when one tree or a small group of trees is removed or killed, the soil beneath them remains partially occupied by the root system of neighbouring trees Thus, in most cases,

‘‘root gaps’’ do not represent patches with no live roots, but may be characterised as zones of reduced root competition

Runkle (1982) and Brokaw (1982) have defined canopy gaps These have a certain minimum size, and extend from the top of the canopy through all vegetation layers to a certain height above the ground We can distinguish between the actual canopy gap, between the edges of crowns, and the expanded gap between tree

stems Since we usually have no zone without roots, it is very difficult and certainly impractical to define a root gap in the field Thus, attempts to identify root gaps have focussed on the ground areas within the perimeter of aboveground canopy gaps Also, when root gaps have been studied, the focus was usually on fine roots, which are responsible for belowground competition for soil resources

The picture that emerges from the few available studies of root gaps is far from clear This is certainly due, in part, to the fact that the results are from different gap sizes and that root measurements are from different soil depths and times since gap creation Many of these studies are from tropical forests, and, in most cases, only the general gap area was analysed rather than the fine root distribution in relation to distance to the gap perimeter When fine root biomass between root gaps and intact adjacent areas was compared, a reduction could be observed in most cases (Fig 10.1) This reduction in fine root biomass was usually not more than 60% (mostly between 20 and 40%) The reduction in live fine roots following canopy creation could be very fast, for example, a 40% reduction in a subtropical wet forest system (Silver and Vogt 1993) However, when medium-term fine root growth between these two areas was compared, root gaps more often showed an increase than a decrease, possibly indicating a reasonably fast recovery of fine root biomass Unfortunately, very few studies provide information on the process of root gap closure with time Bauhus and Bartsch (1996) used in-growth cores to compare the fine root growth ofFagus sylvatica at the centre and perimeter of 30 m diameter gaps within an undisturbed ca 160-year-old forest in the Solling area of Germany Fine root growth in the stand was 390 g m–2(0 30 cm soil depth) over a 12-month period (Fig 10.2)

0

1

2

3

4

5

Fine-root biomass or growth in gaps relative to intact

forest (%)

biomass growth

gaps associated with canopy gaps in different forest ecosystems, where fine root biomass was compared to that of intact adjacent forest Note that the time since gap creation differs between studies (Bauhus and Bartsch 1996; Silver and Vogt 1993; Ostertag 1998; Denslow et al 1998; Wilczynski and Pickett 1993; Cavelier et al 1996; Sanford 1990; Battles and Fahey 2000)

At a distance of 5 m from the edge trees into gaps fine root production over the same period declined to 15 130 g m–2, whereas in the centre of gaps it was negligible Similarly, Mu¨ller and Wagner (2003) found the greatest fine root growth

in gaps only 2.2 m from the edge of the gap in a 35-year-old spruce (Picea abies) forest, while no live fine roots were found beyond 7.4 m from the gap edge These studies show that root gaps can persist for a substantial period of time, if gaps are not, or only slowly, recolonised by other vegetation, as was the case in the two studies cited above However, fine root biomass may recover rapidly when gaps are large enough to be recolonised by fast-growing understorey or shrub species (Bauhus and Bartsch 1996) Jones et al (2003) showed that belowground gaps in Pinus palustris forests closed quickly because understorey vegetation compensated for the absence of pine fine roots, in particular in gaps with higher soil moisture and nitrate concentrations than in the surrounding forest Campbell et al (1998) also found a rapid recovery of non-tree roots in small experimental gaps in mixed boreal forests in Que´bec Whether the speed of recolonisation depends on the contrast in soil nutrient and water availability between the root gaps and surrounding soil, such that the root gaps represent rich patches, is not clear (e.g Ostertag 1998) Higher concentrations of nitrate and phosphate in soils of root gaps as compared to undisturbed areas might facilitate colonisation by pioneer species (Denslow et al 1998) Once saplings have established in gaps, fine root growth in them might be higher than in the undisturbed surrounding forest (Battles and Fahey 2000) The occurrence of fine root gaps is related to the horizontal distribution of fine root mass of individual trees Models of fine root distribution of single trees indicate that the biomass over the entire soil profile is greatest near the stem and declines with distance from the tree (Nielsen and Mackenthun 1991; Ammer and Wagner 2005) However, other studies have indicated that the spatial distribution of roots around stems does not follow such a symmetrical pattern but may be related more to soil nutrient availability (Mou et al 1995) Large-crowned trees have more fine root

0 100 200 300 400 500 600

Distance from egde (m)

2 )

Fig 10.2 Fine root biomass production over a 16 months period determined by the ingrowth core method in an undisturbed European beech forest (0 m) and at different positions (5 and 10 m from the edge) within 30 m diameter gaps (0 30 cm soil depth) (after Bauhus and Bartsch 1996)

biomass and a greater maximum extent of the root system than small trees with little foliage Therefore, the reduction in fine root biomass following the removal or death of individual trees will be greatest in the immediate vicinity of the stump Owing to the lack of studies on this topic, we know neither what size of above-ground gap is required to create a root gap in stands of different tree dimensions, nor what reduction in fine root density is required to have a situation one might call a root gap However, some studies have quantified the distance from the gap edge at which fine root growth ceases (Bauhus and Bartsch 1996; Jones et al 2003) In addition to a significant reduction in fine root density or growth, root gaps should be characterised by changes in the level of resources such as soil moisture and nutri-ents, which would allow colonisation of the gap area by vegetation that previously could not become established under competition from trees occupying the area

A number of studies have demonstrated that belowground ecosystem processes can change dramatically in gaps (Denslow et al 1998; Parsons et al 1994), supporting the notion that the structural diversity in old-growth forests also leads

to functional diversity For example, reduced root competition by mature trees and increased precipitation in gaps lead to higher levels of soil moisture in gaps (e.g Bauhus and Bartsch 1995; Ritter and Vesterdal 2006) Depending on the microcli-matic and soil conditions in gaps, this may or may not lead to increased decompo-sition of organic matter (e.g Bauhus et al 2004) Increased mineralisation of nutrients and, owing to reduced uptake, increased availability of nutrients can lead to high losses via leaching (Bartsch et al 2002; Ritter and Vesterdal 2006; Parsons et al 1994) or in gaseous forms (Brumme 1995) Both processes commonly are associated with changes in soil microflora (e.g Bauhus et al 1996), which is often reflected in increased nitrification rates (e.g Parsons et al 1994) Von Wilpert

et al (2000) documented a rapid and very high increase in seepage water nitrate concentrations at a soil depth of 180 cm following the removal of a singlePicea abies tree in a pole-sized stand Their results showed that such dramatic changes in nutrient transformation processes, equivalent in magnitude to changes following clearfelling, can occur even in very small gaps, probably as a result of the rapid turnover of mycorrhizal fungi Therefore, although most gaps in old-growth forests are small, and only a few gaps are large (Runkle 1982; Butler-Manning 2007), it can

be assumed that most, if not all, gaps leave a belowground signal in the form of root gaps, which contribute to the functional diversity of old-growth ecosystems These patches, with their higher soil moisture and greater nutrient availability, can be colonised by species that were either absent or present in low abundance in undisturbed parts of old-growth stands These different species might in turn support a different soil microfauna and microflora than would dominate under undisturbed conditions, thus contributing to belowground biodiversity

While it is likely that roots gaps are formed when aboveground gaps are created,

it is not clear whether the vegetation patches of different ages that developed subsequently in gaps also differ in their belowground biomass There are indica-tions that fine root growth of saplings in gaps exceeds that of the surrounding stand matrix (Battles and Fahey 2000) For fine roots, the biomass of even-aged stands may be indicative of that found for similar stand developmental phases in patches of

old-growth forests However, owing to the onerous nature of such a task, few studies have compared different age classes of the same species on comparable sites and at the same soil depth Where this has been done for the same ecosystem (Idol et al 2000), it was shown that fine root growth, mortality and decomposition

in 4-year-old stands was as high, if not higher, than in 10-, 29-, or 80- to 100-year-old stands of an oak-hickory forest Where data on fine root biomass from the literature have been compiled, such analyses do not reveal clear temporal patterns of fine root biomass with stand age, and even show contradictory trends for different species In their analysis of published studies, Leuschner and Hertel (2003) showed that the fine root biomass ofFagus sylvatica appeared to decline with age, whereas fine root biomass inPicea abies stands increased with stand age

It is conceivable that in spruce stands, owing to the low quality of litter and associated accumulation of forest floor mass with age (Meiwes et al 2002), nutrients become more growth-limiting with age, hence necessitating the mainte-nance of more fine roots for nutrient capture in older stands compared to young stands This age-dependent increase in fine root biomass has also been observed in Abies amabilis forests in the Washington Cascades in the United States (Grier et al 1981), or for a chronosequence of a tropical montaneQuercus forest in Costa Rica, ranging from an early-successional stage to old-growth (Hertel et al 2003) In both cases, this increase in fine root biomass also coincided with an increase in forest floor thickness and greater fine root biomass in the forest floor layer but not in the mineral soil, lending support to the idea that the increase may be linked to the immobilisation of nutrients Cavelier et al (1996) reported that fine root biomass in the surface mineral soil of early-successional stages from a tropical montane cloud forest did not differ from that in a mature forest However, the fine root biomass (<5 mm) growing in a root mat above the mineral soil surface on logs and branches was higher in the 20-year-old and the mature forest than in the young secondary forest They assumed that, in this extremely moist environment, fine root density saturation and oxygen limitation in the surface mineral soil leads to an expansion

of the root system ‘‘aboveground’’ The increasing heterogeneity of the micro-topography with stand age facilitated this process

However, this pattern of increasing fine root biomass with age cannot be generalised, as demonstrated for fine root biomass in chronosequences of the species Fagus sylvatica, Picea abies, and Quercus cerris (Claus and George 2005) In these cases, fine root biomass peaked at around 20 30% of the mature age of each species, and at levels of 150 200% of the fine root biomass at stand maturity In their study, however, two of three sites had almost no forest floor, whereas at the third site, differences in fine root biomass resulted from differences

in forest floor thickness The pattern observed in the latter study mirrors the temporal dynamics commonly observed for foliage biomass or leaf area index, suggesting that, for these stands, the ratio of foliage biomass to fine root biomass is maintained at a similar level with stand age This may be a more general pattern in situations where nutrient and water availability do not decrease with age However, Uselman et al (2007) found that the fine root production and turnover increases with forest age from 77 years to over 850 years in a successional sequence at

Mt Shasta, California, while the aboveground litterfall declines Owing to the paucity of published data on fine root biomass in chronosequences, it remains speculative whether age-related increases in fine root biomass or production are related to decreasing nutrient or water availability Therefore, it is also not possible

to state whether the spatial heterogeneity in fine root density associated with patches of different ages is higher in old-growth forests than in younger, regrowth stands At sites where an accumulation of forest floor material with stand age occurs, it is likely that gaps, and subsequent development of even-aged groups in these gaps, contribute to a diversification of forest floor conditions and thus rooting patterns Ponge and Delhaye (1995) demonstrate how relationships between stand developmental phases and forest floor development in an old-growth European beech forests influence earthworm communities The presence of different age classes or age-related vegetation communities may also influence the distribution

of soil microflora and microfauna, which can be age- and species-specific The coarse root biomass of individual trees is closely related to their size As trees become older and taller they need to be firmly anchored, and they need to support an increasingly far-reaching network of fine roots Coarse root biomass of individual trees is therefore commonly predicted through allometric relationships to measures such as tree diameter at breast height (e.g Bolte et al 2004) or by using root-shoot ratios (Mokany et al 2006) The relationship between aboveground and root biomass declines with tree biomass and stand age (Mokany et al 2006) However, the declining trend in these relationships is confined largely to young ages and small values for stand biomass Above 30 years or 100 t ha–1, root:shoot ratios remain surprisingly constant This means that, for most situations, below-ground biomass is closely related to the age or biomass of stands and patches Therefore, old-growth forests, which are often characterised by a high spatial heterogeneity of aboveground biomass (Korpel 1995; Butler-Manning 2007), should also have a high spatial heterogeneity of belowground biomass However,

it is not yet clear what relevance this spatial heterogeneity of belowground biomass has for ecosystem function (cf Chap 11, Gleixner et al., this volume) The below-ground equivalent of large old trees, which are an important feature of old-growth, is their large root systems Analogous to dead wood in the crowns of old living trees, with its particular importance to species such as woodpeckers or xylobiotic arthro-pods, the large root systems may generate large coarse dead roots, which provide belowground habitat and substrate Eventually, when the aboveground part dies, stumps and structural roots will constitute a large input of woody material The function of these large dead roots will be discussed further below

Trees also act as conduits for the input of precipitation and chemical elements into forest ecosystems Depending on the architecture of tree crowns, different proportions of these factors enter the system in the form of crown drip and stem flow For smooth-barked species with mostly steep-angled branches such asFagus sylvatica, the proportion of stem flow is so high that the soil at the base of trees exhibits significantly different chemical characteristics than soil further away from the stem Usually it is substantially more acidified (Koch and Matzner 1993) The quantity of stem flow increases with tree crown size so that the channelling effect

for inputs and the resulting changes in soil chemical conditions is more pronounced for soil at the base of large trees as compared to small trees Consequently, this phenomenon will also lead to higher spatial belowground heterogeneity in old-growth forests comprising species, where stem flow forms an important part of their input fluxes When tree species have different influences on nutrient cycles and soils (for the underlying mechanisms see Binkley and Giardina 1998), this can lead

to long-lasting spatial patterns in soils (e.g Boettcher and Kalisz 1990; Fujinuma

et al 2005) We can assume that these patterns are more profound in old-growth forests, where trees are long-lived and tend to maintain these different influences for long periods of time

10.4 Pit-and-Mound Topography in Old-Growth Forest

In addition to root gaps, small scale gap-phase disturbance in old-growth forests may also lead to heterogeneity in soil conditions, when trees are blown over and their root plates are tipped up (Liechty et al 1997) Since the risk of windthrow increases with tree height, this phenomenon is more common in taller, older forests than in young forests In addition, the volume of the pits created by uprooting and the size of the mound resulting from the decay of the root plate are strongly related

to tree size (Putz 1983; Clinton and Baker 2000) The importance of this phenome-non differs with forest type, depending on the typical root development of the dominant species, which is often influenced by soil thickness and the soil water regime In mixed species forests in the Carpathian Mountains and boreal forests in Central Russia, the surface area covered by pits and mounds ranged from 6.5 to 25% (Ulanova 2000) The latter case is obviously the result of a catastrophic windthrow event In northern hardwood, and hemlock-hardwood old-growth forests

in Michigan, the surface area in pit-and-mound topography shaped by recent windthrow events was 27% and 33%, respectively (Liechty et al 1997)

Both pits and mounds can be associated with important ecosystem functions The uprooting of trees bares mineral soil, which may be required for the regenera-tion of species with small seeds or that otherwise have difficulty in germinating and becoming established on thick forest floors (Bazzaz 1996) In addition, the root plates or mounds constitute microsites with higher light availability or reduced competition For example, Nakashizuka (1989) showed for an old-growth mixed temperate forest with a bamboo understorey that the mounds created through tree fall were more important for tree regeneration particularly for species with small seeds than the gaps created at the same time Often, mounds are characterised by favourable conditions for root growth, which result from the forest floor mixing with mineral soil, increased soil temperature and favourable soil moisture regimes, especially when the surrounding soils are not free-draining (Clinton and Baker 2000) Mounds may also be sites of increased soil faunal activity (Troedsson and Lyford 1973) At the same time, pits represent microsites that accumulate forest floor material and are often moister than their surroundings (Beatty and Stone 1986;

Liechty et al 1997) Mound disturbance is also very important in boreal or alpine forests, in which nutrients become locked up in the forest floor; the mixing of forest floor with mineral soil remobilises these nutrients For Picea sitchensis-Tsuga heterophylla forests in south-east Alaska, Bormann et al (1995) calculated that windthrow and associated soil mixing is required every 200 to 400 years to maintain soil fertility (see Chap 9 by Wardle, this volume)

Although the creation of pit-and-mound micro-topography has some important ecosystem functions and clearly contributes to soil heterogeneity, particularly in old forests, there is no information on its effect on rooting patterns, and little information on its effect on soil biodiversity However, pits and mounds as well

as coarse woody debris (CWD) appear to be important structures in maintaining the diversity of vascular plants in old-growth forests, especially where these have to compete with an understorey of shade-tolerant tree seedlings (Miller et al 2002) Whether the maintenance of aboveground plant species diversity, through specific associations with belowground micro-flora and fauna, provides positive feedback

to belowground biodiversity is uncertain The diversity of mycorrhizal fungi does not follow patterns of plant diversity (Allen et al 1995) If mounds are a preferred substrate for tree growth, and hence colonisation by roots, this may also create positive feedback leading to spatial variation in the accumulation of organic matter

10.5 Old-Growth Structures Harbouring Roots

It has been recognised that so-called ‘‘residual structural elements’’ play an impor-tant role in the conservation of biodiversity in managed forests (Franklin et al 1997) The term describes structures such as standing dead trees (snags), logs on the ground, residual live trees, and undisturbed vegetation patches, which can be found following natural disturbance events in old forests, but which are usually removed

or are much reduced when forests are harvested regularly Franklin et al (1997) created the term ‘‘life-boating’’ for the retention of these structures in managed forest ecosystems to illustrate the function of these structures as refuges for different taxonomic groups during the re-organisation and aggradation phase of ecosystem development Interestingly, this concept has been applied so far only to aboveground structures Therefore, we might ask whether old-growth forests also have belowground structural elements that provide important habitat or are impor-tant for ecosystem functions following disturbance Obvious candidates for such residual structures are root channels Are there more, deeper root channels in old-growth forests, and what are their functions?

Root channels can be created by soil cracks or by old, decaying roots Such channels constitute preferential flow paths for seepage water, act as dispersion pathways for microorganisms and invertebrates, and facilitate root movement through the soil and possibly the deep rooting of young trees Based on measure-ments of soil microbial biomass and activity, Bundt et al (2001a) have suggested