Decomposition of leaf litter in headwater streams.

Decomposition of leaf litter in
headwater streams.
Effects of changes in the environment and contribution
of microbial and shredder activity on litter
decomposition.
Johan Lidman
Degree thesis in ecology 60 hp
Master’s level
Report passed: 16 06 2015
Supervisor: Micael Jonsson
Decomposition of leaf litter in headwater streams.
Effects of changes in the environment and contribution of microbial and
shredder activity on litter decomposition.
Johan Lidman
Abstract
Headwaters, which are the most common stream order in the landscape, are mostly
dependent on energy produced in the terrestrial system, largely consisting of leaf litter from
riparian vegetation. The aim of this study was to investigate the decomposition in headwaters
of leaf litter from three native (alder, birch, spruce) and one non-native (lodgepole pine)
species and how decomposition responds to changes in the environment. Further, microbial
and shredder influences on leaf-litter decomposition and aquatic decomposer ability to adapt
to non-native species was investigated. By using field-data from this study, calculations were
made to assess if microbes and shredders are resource limited. Litterbags were placed in 20
headwater streams in northern Sweden that varied in water chemistry, stream physical
characteristics and riparian vegetation. The results revealed that species litter decomposition
of different plant species was affected differently by changes in environmental variables.
Alder and birch decomposition were positively associated, whereas lodgepole pine deviated
from the other species in decomposition and its relationship with important environmental
variables, indicating that the ability of the boreal aquatic systems to decompose litter differs
between introduced and native species. When including macroinvertebrates, shredder
fragmentation generally increased decomposition, but was not significant for all sites.
Resource availability for microbes and shredders was controlled by litter input, and no risk of
resource limitations was evident during the study period. These findings highlight a
complexity of the decomposition process that needs to be considered when predicting
changes due to human activities.
Key words: Leaf-litter decomposition, headwaters, microbial decomposition, shredders,
ecosystem adaptation
Contents
1 Introduction .................................................................................. 1
1.1 Headwaters ..................................................................................... 1
1.2 Decomposition of litter ................................................................... 1
1.3 Litter quality and stoichiometry ...................................................... 2
1.4 Macroinvertebrate production ........................................................ 2
1.5 Human disturbance......................................................................... 2
1.6 Study aims ...................................................................................... 3
2 Material and methods .............................................................. 3
2.1 Study site ........................................................................................ 3
2.2 Water chemistry and stream characteristics ................................... 4
2.3 Litter decomposition ...................................................................... 5
2.3.1 Study design ................................................................................ 5
2.3.2 Mass loss ..................................................................................... 5
2.3.3 Decomposition rate ...................................................................... 5
2.3.4 Nutrient content .......................................................................... 6
2.4 Litter input ..................................................................................... 6
2.5 Estimation of litter standing stock .................................................. 6
2.6 Macroinvertebrates ........................................................................ 6
2.7 Fungal biomass ............................................................................... 7
2.8 Statistical analyses.......................................................................... 7
3 Results ............................................................................................. 8
3.1 Water chemistry and stream physical characteristics ...................... 8
3.2 Litter decomposition ...................................................................... 9
3.3 Macroinvertebrates ...................................................................... 10
3.4 Ergosterol ...................................................................................... 11
3.5 Nutrient content ........................................................................... 12
3.6 Litter input ................................................................................... 12
3.7 Variables influencing microbial decomposition ............................ 14
3.8 Macroinvertebrates ...................................................................... 16
3.9 Estimation of litter standing stock .................................................17
4 Discussion ................................................................................... 18
4.1 Riparian vegetation ....................................................................... 18
4.2 Water chemistry and stream physical characteristics ................... 18
4.3 Adaptation to non-native species .................................................. 19
4.4 Variable associations .................................................................... 19
4.5 Decomposer community and litter decomposition ........................ 20
4.6 Litter standing stock ..................................................................... 21
4.7 Conclusion .................................................................................... 21
5 Acknowledgement .................................................................... 22
6 References ................................................................................... 23
Appendix 1
Appendix 2
Appendix 3
1 Introduction
1.1 Headwaters
Headwaters (catchments < 15km2) are the dominant stream order at the landscape scale,
representing up to 85% of the total stream length. Therefore, headwaters constitute a large
part of the land-water interface, and act as a buffer against inputs of pollutants from the
terrestrial environment for downstream lakes and streams by processing pollutants that
enter headwaters (Bernhard et al. 2003, Bishop et al. 2008). Furthermore, headwaters
within the same catchment can show a large variability in many important environmental
factors, including water chemistry, temperature, morphology, and food resources (Temnerud
and Bishop 2005, Meyer et al. 2007). This variability provides niches for a wide range of
organisms that contribute to landscape-scale biodiversity (Leroy and Marks 2006). Healthy
and functional headwaters are therefore important both for the function of the whole stream
network and as providers of habitats to organisms. These small systems are often limited in
algal production, since dense canopy cover commonly shades sunlight. With low algal
production, allochthonous (from terrestrial origin) produced carbon is the dominating
energy source, largely consisting of litter input from the riparian vegetation (Fisher and
Likens 1973, Bärlocher 1985).
1.2 Decomposition of litter
When terrestrial litter enters the aquatic system it will start to decompose. Initially the litter
will rapidly lose mass due to leaching, up to 30% within the first 24h (Petersen and Cummins
1974). Second, microorganisms, such as fungi and bacteria, colonize the litter and start to
decompose its structure and change its physical and chemical characteristics (i.e. microbial
conditioning). This makes the litter more palatable for detritivorous macroinvertebrates, i.e.
aquatic invertebrates defined as ‘shredders’, which fragments the remaining litter to fine
particulate organic matter (FPOM). In turn, FPOM becomes an energy source for organisms
specialized to filter feeding or gathering small deposited particles (Bärlocher 1985, Webster
and Benfield 1986).
The litter decomposition process is influenced by several environmental factors, and studies
have tried to disentangle what these factors are. Temperature has been shown to affect
decomposition mainly by increasing microorganism metabolic activity, leading to higher
feeding rates, and consequently higher decomposition rates (Webster and Benfield 1986).
Microbial decomposition of litter is also affected by the availability of dissolved organic
carbon (DOC) in the water, as DOC is used as an energy source by bacteria. With high DOC
levels, bacterial abundance increases and affects microbial decomposition positively.
However, not all DOC is available for bacteria to use. Meyer et al. (1987) showed that the
proportion of DOC used by bacteria changed with DOC molecule weight, indicating that not
only the amount, but also the quality of DOC is important for bacterial production and
consequently litter decomposition. Another important water chemistry parameter is water
nutrient concentrations. Both fungal and bacterial production can be limited by nitrogen (N)
and/or phosphorus (P). In comparisons of litter decomposition in nutrient rich and nutrient
poor systems, a majority of studies have found higher decomposition rates in nutrient
enriched systems, due to higher microbial productivity (Webster and Benfield 1986,
Suberkropp and Chauvet 1995, Gulis and Suberkropp 2003, Suberkropp et al. 2010).
As already mentioned, litter decomposition is also affected by fragmentation from
macroinvertebrates. However, unconditioned leaf litter contains lignin and tannins,
secondary compounds that shredders lack the ability to break down. Fungi and bacteria are
able to degrade secondary compounds into simpler compounds, and without this
conditioning, leaf litter is a poor food resource for shredders (Graça 2001). Fungi are a more
easily accessible nutrient source than leaf litter tissue to shredders, making litter colonized by
microbes a preferred substrate for shredders (Bärlocher 1985). Hence, litter fragmentation by
shredders is stimulated by microbial activity, and the shredder community. Although the
1
abundance of aquatic invertebrates affect litter decomposition rates, their community
composition is also important. For example, high decomposition rates have, in several
studies, been positively correlated with high species richness (Jonsson et al. 2001, Dangles
and Malmqvist 2004). One plausible mechanism behind this relationship is that species have
different niches, and are using litter resource slightly differently and perhaps together more
efficiently (Jonsson et al. 2001).
1.3 Litter quality and stoichiometry
Although several environmental factors affect how fast leaf litter decomposes, litter quality is
a strong determinant by affecting the colonization and decomposition of microbes and
macroinvertebrates and subsequent decomposition rates. Studies have shown that litter
decomposition is positively correlated with high N and P concentrations and negatively
correlated with high lignin and tannins levels in the litter (Gessner and Chauvet 1994, Aerts
1997, Leroy and Marks 2006). High nutrient concentrations positively affect fungal and
bacterial colonization, while lignin and tannins inhibit microbial growth (Webster 1986).
However, quantity of nutrients in the litter is not the only factor influencing decomposition
rates. Recent papers highlight the effect of litter stoichiometry, i.e. the carbon/nutrient
ratios, and its effect on shredder resource consumption (Fuller et al. 2014, Manning et al.
2015). Litter of a higher carbon/nutrient ratios typically decrease consumption and growth of
aquatic insects, and will therefore decompose slower than litter of low ratios. Litter
stoichiometry is also affected by the amount of dissolved nutrients in the water, via impacts
on fungal production. High levels of dissolved nutrients have been shown to decrease the
carbon/nutrient ratio in litter via microbial nutrient uptake (Rosemond et al. 2010).
Colonizing fungi can incorporate nutrients from both water and litter for production, thereby
making the litter more palatable for shredders (Manning et al. 2015).
1.4 Macroinvertebrate production
As already mentioned, shredders in headwater streams rely on input of terrestrial litter as an
energy source. The activity and composition of aquatic invertebrates (primarily shredders)
can therefore be reflected in riparian vegetation variation. For example, a quantitative
decrease in riparian vegetation input decreases production of shredders (Wallace et al. 1999).
However, changes in litter quality and composition have also been found to affect secondary
production in aquatic systems. Hence, streams with riparian vegetation that produce high
quality litter produce more aquatic invertebrates (Kominoski et al. 2011).
Temporal variation in stream macroinvertebrate abundance can follow the abrupt increase in
deciduous litter input to streams during fall, resulting in increasing growth due to increased
access to food (Haapala et al. 2001). However, as decomposition of deciduous leaves is
relatively fast, the majority will be consumed during fall and early winter. To avoid food
shortage during and after the winter, when litter input is generally low, input of more slowdecomposing species is also important to maintain macroinvertebrate production (Cummins
et al. 1989, Haapala et al. 2001). Without slow-decomposing, low-quality litter, detritivorous
aquatic invertebrates will likely be limited in growth and development up until emergence in
early spring to early summer, resulting in fewer macroinvertebrates that decompose litter the
next fall.
1.5 Human disturbance
The variability between headwaters can partly be ascribed to disturbance in the terrestrial
environment. Prior to human interference, natural disturbance such as fire, windstorms, and
dying trees formed the landscape, and continuously changed the conditions for both
terrestrial and aquatic systems. While dying trees only creates gaps in the forest, fire and
wind disturbance can affect larger areas, even whole catchments (Essen et al. 1997).
However, today almost 97% of the productive forested area in Sweden is affected by forestry
(Kempe 2014), which is the single most common disturbance in Swedish forests (Essen et al.
1997). To replace fossil energy sources, the demand to use forest products is increasing. In
2
Sweden, one attempt to increase the production of forest products was the introduction of
the North American tree species lodgepole pine (Pinus contorta). Today it occurs mainly in
the northern parts of Sweden where it covers approximately 600,000 ha, i.e. 4% of the
productive forested area. Lodgepole pine produces more litter but of lower quality, compared
to the native Scots pine (Engelmark et al. 2001). For aquatic ecosystems, the effects from
forestry practices can be seen in changed abiotic factors, such as increased temperature,
reduced canopy cover, higher stream flow, and changes in water chemistry (Allan 2004).
Harvesting of the riparian vegetation can also decrease litter input to the aquatic system
(Webster et al. 1992) and result in changed tree species composition. The present
management action to prevent damage from forestry on aquatic ecosystem is to create
riparian buffer zones along streams and rivers to maintain processes in the riparian zone,
including filtration of pollutants, sediment stabilization, shading, and water regulation, but
this is not practiced along headwaters (Kuglerova et al. 2014).
1.6 Study aims
Headwaters receive relatively little focus despite their contribution to stream networks and
importance for downstream aquatic systems (Bishop et al. 2008). Litter decomposition
processes in headwaters are important for both production of aquatic invertebrates and for
the uptake and recycling of nutrients and carbon in forests. Human disturbance transforms
landscapes, and it is increasingly important to understand how this affects vital ecosystem
processes. The majority of studies on decomposition in headwaters compares relatively few
locations (Gessner and Chauvet 1994, Suberkropp and Chauvet 1995, Haapala et al. 2001,
Lecerf et al. 2007), despite the innate heterogeneity of these systems, and seldom tries to
include all parameters that are known to affect litter decomposition. This study tries to
disentangle (I) which variables are important for decomposition of four leaf litter species,
using a wide range of sites and variables. Three of the species; alder (Alnus glatinosa), birch
(Betula pubecens) and spruce (Picea abies), are native in the area but differ in quality,
whereas the fourth species, lodgepole pine, is alien to the studied systems. Using both native
and non-native species makes it possible to also investigate (II) if decomposer in headwater
systems are adapted to current condition or can adapt to sudden change in litter input.
Comparing microbial- and macroinvertebrate-mediated litter decomposition makes it
possible to assess (III) the relative importance of microbes and macroinvertebrates for litter
decomposition in headwater streams. Lastly, the collected data allows for estimations of leaf
litter quantities in the studied headwaters to (IV) detect risks of food limitations for
macroinvertebrates over the studied season.
2 Material and methods
2.1 Study site
The study was conducted in 20 first- to second-order streams in the Swedish province of
Västerbotten and Ångermanland (Figure 1). Mixed boreal coniferous forest is the main forest
type within the study area. The streambeds varied from being dominated by gravel and sand
to largely consisting of smaller boulders covered with aquatic mosses. Forestry practice is the
major disturbance factor in the area, with clear cutting followed by soil scarification and
plantation of coniferous trees as the main method (Essen et al. 1997). The studied streams
differ in type of land cover and surrounding forest structure (i.e. forest age and species
composition), from young forest dominated by alder and birch to old spruce forests, creating
a gradient in external influences on the headwater sites. The study was carried out in 2014,
from May to November. At each stream, 100 m that was dominated by riffles was chosen as
study site.
3
Figure 1. Map of study area where each point represents a site.
2.2 Water chemistry and stream characteristics
Water samples were taken four times during August to November for measures of
absorbance, DOC, reactive phosphorous (PO4), nitrate (NO3), ammonium (NH4), dissolved
organic nitrogen (DIN), total nitrogen (TN), conductivity and pH. Samples where filtered in
the field with a 0.45-µm nylon membrane filter (Sarstedt, Nümbrecht, Germany). All samples
were stored during fieldwork in a cooling box with ice packs. In the lab, pH and absorbance
samples were stored at +4 ºC. DOC and nutrient samples were stored at -25 ºC before being
analyzed. All water chemistry samples were analyzed at the Swedish University of
Agricultural Science (SLU) in Umeå. From the absorbance spectra SUVA254 was estimated.
SUVA254 is the UV absorbance at wavelength 254 and is used to estimate the content of
aromatic carbon (Weishaar et al. 2003). Compared to other carbon molecules, aromatic
carbons are less available for organisms to use (Perdue 1998). At the first water sampling
occasion, not enough water volume was collected for the SUVA254 measurements. Only
three replicates for SUVA254 were therefore used. Water temperature was continuously
monitored during the experiment with HOBO Pendant temp/light loggers that were
attached to an iron bar at each site. On occasion, water levels were below the logger, but
those readings were brief and could be removed from the temperature data. In total, 64 days
were included, after an exclusion of 23 days for all sites. Light condition was determined in
August before leaf abscission using a spherical densiometer. Stream depth, width and
velocity were measured at three points along the study site in each stream. For depth, the
measuring was repeated three times during the study period to get an average for the whole
period.
4
2.3 Litter decomposition
2.3.1 Study design
Lodgepole pine and Norway spruce needles where collected in early June by shaking
branches, which makes dry needles come off. Alder and birch leaves were collected at
abscission. The needles and leaves were air-dried indoors (20 ºC) to constant weight.
To measure microbial decomposition 15x15-cm fine-mesh (0.5 mm) bags were used to
exclude effects from invertebrates. Another set of 12x17-cm coarse-mesh (5 mm) bags was
used to measure invertebrate leaf litter decomposition. In each bag, 3.0 (±0.05) g of air-dried
leaves from one of the studied species was inserted. For the coarse mesh bags, only birch
leaves were used. The stalks on birch and alder leaves were removed before weighing and
insertion into the bags. Unlike the fine mesh bags, the coarse mesh bags was not closed by
folding the top over and secured with staples, but by pressing the sides together and closed
with plastic strings to create more space in the bag.
At each stream site, five replicates of each litter species and five coarse-mesh bags containing
birch (25 in total) were anchored randomly to a chain secured with an iron bar in the
streambed. The bags were placed some distance from each other to prevent overlapping. Bags
containing coniferous species where introduced to the streams in August (week 34) and bags
with deciduous species in September (week 39). The longer time in the streams for conifers is
due to their known slower decomposition rates. Therefore, more time is needed to see
potential difference among sites. The litterbags were retrieved in November (week 47),
rendering a total of 87-91 days for coniferous litter and 53-56 days for deciduous litter.
Before processing, the bags were stored at -18 ºC. This study design that performed one
litterbag collection, as opposed to sequential collections, was chosen to be able to include a
high number of sites in the experiment. With more sites, a wider range of variation in the
variables presumed to influence litter decomposition can be obtained.
2.3.2 Mass loss
The decomposed litter was rinsed before being oven dried at 60 ºC for 48h. The dried litter
was then combusted in 550 ºC for 40 min to obtain ash weight (Benfield 1996). Litter ashfree dry mass (AFDM), used in the statistical analyses, was obtained by subtracting ash
weight from dry-mass.
Following the same procedure as above, initial AFDM was obtained by combusting three 1 g
samples of each of the four litter species. The average from the tree samples was used to
calculate initial AFDM. Mass loss was then calculated by subtracting AFDM remaining from
initial AFDM (Eq. 1):
MLoss=AFDMInitial-AFDMRemaining
(1)
2.3.3 Decomposition rate
To calculate decomposition rate for each treatment and site, the decomposition constant (k)
was calculated, using the negative exponential decay model (Eq. 2):
k= ln(Mt/M0)/t
(2)
where Mt is the AFDM at time t and M0 is the initial AFDM. A decision was made to use the
same time of exposure for each species despite a maximum difference of three days between
sites. This was because when comparing the results of using different exposure times, a
difference in two days had a greater impact on k than two days longer in the field would have
had on litter mass loss since decomposition rate slows down with time as water-soluble
compounds and easily decomposed material is lost early in decomposition processes
(Webster and Benfield 1986).
5
There was a significant positive correlation between mass loss and k (Pearson’s productmoment correlation, p<0.001, r=0.94). Hence, in all analyses, k instead of mass loss was
used to enable comparison between species with different time in the field
2.3.4 Nutrient content
The initial litter quality of each species was measured by sending litter to SLU Umeå for
analyses of C and N content. The C/N ratio was used as a proxy for litter quality.
2.4 Litter input
To quantify amount and type of litter input to each stream, three litter traps, i.e. trays that
were 40x50 cm in size, were placed immediately adjacent to each stream channel at each
study site in late May. To be able to assess potentially different effects of summer and fall
litter input, trapped litter was collected at the end of summer (20-22 August) and in late fall
(late October), the latter date being after complete defoliation of deciduous trees. The
collected litter was sorted to six categories (i.e. birch leaf, alder leaf, pine needle, spruce
needle, twig, and ‘other’) and was oven dried at 60 ºC for 48h to obtain dry mass. The
category ‘other’ consisted mainly of grasses and herbs. The dry litter was then combusted to
obtain AFDM as above.
2.5 Estimation of litter standing stock
Calculations on litter standing stock were made to assess if shredders could be limited in
growth by lack of food at the end of the study. However, since both litter input and
decomposition data only was available for alder, birch and spruce, estimations were only
done for those species. Incoming litter to the stream, based on the litter trap data, was set to
decompose over 53 days according to the observed decomposition rates from the litterbag
experiment. Only litter input from the fall period was used, since decomposition rates were
based on that time period. I assumed that fall litter input followed the same temporal pattern
across all sites, and neglected downstream litter transport, as it likely would be compensated
for by upstream transport into the study site. Therefore, in the calculations, all litter was
entering the stream at the same time and no net loss of standing stock due to transport was
assumed. For Birch, the effect of macroinvertebrate feeding on litter standing stock could
also be estimated. The calculations were based on the negative exponential decay model,
where each species was calculated separately (Eq. 3) and the summarized (Eq. 4):
M53=M0*e-kt
(3)
Mtot=MAlder+MBirch+MSpruce
(4)
where Mt is the litter mass remaining at time t and k is the decomposition rate. MAlder is the
mass remaining for alder at time t, and the same for MBirch and MSpruce.
2.6 Macroinvertebrates
The coarse-mesh bags were retrieved at the same time as the fine mesh bags using a hand net
to prevent loss of invertebrates when removing the bags. By rinsing the leaves over a 35-mm
mesh sieve, the invertebrates were removed, picked and preserved in 70% ethanol before
being sent to a taxonomist for identification to species/genera. The species were then
categorized to functional feeding groups using previous classification of the species (SchmidtKloiber and Hering 2012). In later analyses, only shredders were considered, whereas other
functional feeding groups were not used in the study. Shredder invertebrates were dried at
60 ºC for 48 h to estimate dry weight.
6
2.7 Fungal biomass
In the laboratory, samples of approximate 30 mg of litter were taken from the fine-mesh
bags. The samples where frozen and stored at -18 ºC, before being analyzed for ergosterol
content at SLU Uppsala. Measuring ergosterol levels is a common method for estimation of
fungal biomass, since ergosterol is found exclusively in cell membrane of living fungi
(Gessner and Chauvet 1994). Since leaf mass was taken from the litterbags for the ergosterol
analyses before they were processed and weighed, the proportion of AFDM of the remaining
litter was multiplied with the dry weight from the ergosterol sample and added to the AFDM
of the fine-mesh bag it originated from.
2.8 Statistical analyses
Differences in decomposition were analyzed using one-way analysis of variance (ANOVA),
with litter species as a fixed factor and decomposition rate as the response variable. Tukey’s
HSD for pair-wise comparisons was used to determine significant difference among litter
species. The same analysis was performed with ergosterol content as response variable. Two
two-way ANOVAs were conducted to test for significant differences in litter input between
the summer and fall period, with time period and litter categories as fixed factors and the
different categories proportional or biomass input as response variable. Invertebrate effects
on decomposition were determent in a paired t-test, comparing decomposition rates for birch
in fine- and coarse-mesh litterbags. To assess if all sites followed the same pattern of
differences in decomposition rates between fine- and course-mesh bags, a two-way ANOVA
was conducted, with site and treatment as fixed factors and decomposition rate as response
variable, followed by Tukey’s HSD.
To explore relationships between abiotic factors and riparian vegetation at each stream site, a
principal component analysis (PCA) was conducted. Two PCAs, one for summer and one for
fall, were also conducted to visually describe differences in the composition of litter input
between streams. Component 1 scores from these two PCAs were later used as single
measures of litter input composition. Partial least squared (PLS) models were used to
determine which variables were the most influential for determining decomposition for each
litter species. For the coarse-mesh bags, invertebrate variables were also included (Table 1). A
PLS analysis is comparable to a multiple regression analysis, but is preferred when dealing
with large numbers of independent variables and relatively low numbers of observations (i.e.
sites). Additionally, a PLS does not assume normally distributed data, which in field studies
often is difficult to obtain (Hulland et al 2010). The evaluation of the PLS models was based
on the level of variance explained (R2), loadings of the independent variables, and the
variable influence on projection (VIP). The independent variable loading describes the
relative strength and direction of the relationship between independent and response
variable. The VIP value summarizes the importance of each variable (Stenroth et al. 2015). In
the models, the limit for a variable to be included in the final model was a VIP value at 0.7. To
visually describe difference in variable relationships between litter species, a PCA bi-plot was
conducted on the fine-mesh bags and the most important variables from the PLS analyses.
Litter availability for shredders can either depend on amount of litter input from the riparian
vegetation or the decomposition rate of standing stock. Hence, two PLS models were
conducted to assess the relative importance of litter input and decomposition rate for food
availability to shredders. The first model included abiotic variables combined with microbial
decomposition and litter input of alder, birch, and spruce. The second model also included
macroinvertebrate decomposition, but was only done for birch litter. All analyses were
performed in version 2.15.1 of R (R Core Team 2012).
7
Table 1. Variables used in the PLS models on litter
decomposition. For litter input, summer and fall values were
kept separate. For the ergosterol variable, only ergosterol for
the analyzed species was used.
Variables
Litter input
Alder
Birch
Pine
Spruce
Twigs
Other
Total litter input
Water chemistry
Conductivity
pH
DOC
SUVA254
NH4
NO3
DIN
TN
PO4
Stream characteristics
Water depth
Chanel width
Stream velocity
Canopy openness
Temperature
Invertebrates
Species richness
No. individuals
Dry weight
Other
Ergosterol
Litter input summer PCA
Litter input fall PCA
Analyze used
All
“
“
“
“
“
“
“
“
“
“
“
“
“
“
“
“
“
“
“
“
“
Coarse
“
“
All
“
“
3 Results
3.1 Water chemistry and stream physical characteristics
Water chemistry variables varied among sites and over the study period (Appendix 1). Mean
temperature during the study ranged from 3.81 0C at site G1 to 4.85 0C at site G2 (Table 2).
Site Bcc stood out in canopy cover with a canopy openness of 87.4%, while the other sites
ranged from 1.4 to 22.1%. Both stream channel width and mean depth was highly variable
among sites. The widest channel was found at site V1 (185.0 cm), while the deepest site was
site KL with mean depth of 26.6 cm. Site B1 had the lowest stream flow at 0.025 m/s, while
the highest velocity was found at site BF (0.307m/s). All site characteristics can be viewed in
table 2.
8
Table 2. Mean values for stream physical characteristics for each site. Values in parentheses show ± 1 standard
error.
Site
B1
Bcc
BF
G1
G2
G3
K1
K5
K8
KL
Kr1
Kr6
Kr7
Rod
S2
S26
S3
S6
V1
V2
Depth
(cm)
22.8 (1.9)
7 (0.4)
13.4 (0.6)
16.3 (4.4)
9.4 (0.7)
20.8 (2.8)
8.3 (0.7)
10.1 (0.8)
10.1 (0.9)
26.6 (3.4)
6.2 (0.6)
9.9 (1.3)
6.9 (0.5)
16.5 (1.9)
9.4 (0.9)
12.1 (0.5)
5.7 (1.2)
5.8 (0.5)
14.3 (1.4)
20.3 (1.6)
Width
(cm)
121.8 (18.3)
43.8 (5.8)
101.7 (8.6)
81.7 (7.0)
83.2 (7.4)
73.3 (7.2)
39.7 (3.9)
65.3 (8.8)
67.5 (9.1)
144.3 (21.8)
59 (6.9)
84.7 (20.2)
85.2 (9.7)
139.2 (4.2)
40.7 (2.9)
50.3 (5.3)
73.8 (3.1)
56.2 (8.9)
185 (6.3)
104.2 (10.2)
Velocity
(m/s)
0.025 (0.011)
0.126 (0.032)
0.307 (0.060)
0.133 (0.038)
0.181 (0.049)
0.114 (0.038)
0.276 (0.053)
0.074 (0.025)
0.229 (0.037)
0.25 (0.035)
0.097 (0.016)
0.101 (0.047)
0.287 (0.184)
0.227 (0.054)
0.065 (0.019)
0.102 (0.026)
0.047 (0.027)
0.065 (0.009)
0.233 (0.028)
0.298 (0.085)
Temperature
(°C)
4.42 (0.055)
4.36 (0.045)
4.35 (0.051)
3.81 (0.049)
4.85 (0.072)
4 (0.054)
4.23 (0.032)
4.65 (0.049)
4.44 (0.053)
4.13 (0.055)
4.25 (0.048)
4.5 (0.064)
4.33 (0.048)
4.04 (0.059)
4.37 (0.052)
3.93 (0.056)
3.87 (0.053)
4.17 (0.057)
4 (0.054)
4 (0.056)
Canopy openness
(%)
5.72 (2.08)
87.36 (8.61)
9.88 (0.69)
22.1 (10.95)
6.15 (1.25)
15.43 (4.21)
11.53 (2.65)
16.55 (2.62)
1.47 (1.09)
3.21 (2.30)
3.64 (1.48)
8.58 (3.90)
5.2 (2.80)
8.93 (5.04)
1.39 (0.38)
9.36 (1.30)
1.82 (0.60)
3.81 (1.96)
15.34 (10.68)
5.55 (2.69)
3.2 Litter decomposition
There was a significant difference in decomposition rate for fine-mesh bags between litter
species (df=3, F=227, p<0.001). Tukey’s HSD showed that there was significant difference
between all species (Figure 2). Alder had the highest mean decomposition rate (k=0.0077)
and lodge pine the lowest (k=0.0025).
Figure 2. Decomposition rate (d-1) for studied litter species. Different letter
indicate significant difference at p=0.05. Error bars indicate ± 1 SE.
9
Decomposition rates for the coarse-mesh bags ranged between 0.009-0.024 d-1, with highest
decomposition at site V1 and the lowest at site G3 (Table 3). In general, coarse-mesh bags
(i.e. shredders present) showed a higher average birch decomposition rate than fine-mesh
bags (i.e. microbes) (df = 19, t = -10.0946, p<0.01). However, within sites only 11 of the 20
sites showed a significant higher birch decomposition rate due to shredder presence (Table
2).
Table 3. Mean decomposition rate of birch in fine-mesh
(Microbes) and coarse-mesh (Microbes+Shredders) bags at each
stream. Values in parentheses show ± 1 standard error. p shows
the result from a two-way ANOVA comparing decomposition of
birch with and without influence of shredders. Significant
differences (p<0.05) are bolded, and close to significant
difference are presented in cursive.
Site
B1
Bcc
BF
G1
G2
G3
K1
K5
K8
KL
Kr1
Kr6
Kr7
Rod
S2
S26
S3
S6
V1
V2
Decomposition rate (d-1)
Microbes
Microbes+Shredders
0.0055 (0.0002)
0.013 (0.002)
0.0087 (0.0008)
0.021 (0.001)
0.0068 (0.0003)
0.017 (0.001)
0.0067 (0.0003)
0.015 (0.002)
0.0070 (0.0002)
0.013 (0.002)
0.0066 (0.0008)
0.009 (0.001)
0.0081 (0.0008)
0.016 (0.001)
0.0054 (0.0003)
0.012 (0.002)
0.0068 (0.0003)
0.012 (0.001)
0.0073 (0.0005)
0.017 (0.001)
0.0061 (0.0004)
0.011 (0.001)
0.0065 (0.0004)
0.014 (0.001)
0.0079 (0.0004)
0.012 (0.001)
0.0066 (0.0003)
0.012 (0.001)
0.0070 (0.0005)
0.011 (0.001)
0.0064 (0.0001)
0.018 (0.001)
0.0058 (0.0004)
0.010 (0.002)
0.0069 (0.0004)
0.017 (0.001)
0.0079 (0.0007)
0.024 (0.004)
0.0066 (0.0005)
0.013 (0.001)
p
0.006
< 0.001
< 0.001
< 0.001
0.062
0.999
0.004
0.037
0.3
< 0.001
0.308
0.002
0.764
0.357
0.894
< 0.001
0.793
< 0.001
< 0.001
0.085
3.3 Macroinvertebrates
In total, 20 shredder species were found in the coarse-mesh bags (Appendix 2). Mean species
richness ranged from 0.6 to 7.6 species per litterbag and was highest at site Rod (Table 4).
Mean abundance was highly variable, ranging from 6 at site G3 to 113.2 at site Bcc. Mean
macroinvertebrate biomass (dry weight) ranged from 1.2 to 16.3 mg.
10
Table 4. Mean invertebrate species richness, abundance, and
biomass from the coarse-mesh bags. Values in parentheses
show ± 1 standard error.
Site
B1
Bcc
BF
G1
G2
G3
K1
K5
K8
KL
Kr1
Kr6
Kr7
Rod
S2
S26
S3
S6
V1
V2
Species richness
2.2 (0.4)
3 (0.3)
4.6 (0.5)
2.8 (0.4)
4.6 (0.7)
2.4 (0.5)
4.8 (0.6)
2.2 (0.4)
3.2 (1.0)
5.6 (0.6)
3.2 (0.4)
2.8 (0.4)
1.6 (0.5)
7.6 (0.4)
3 (0.3)
2.6 (0.2)
0.6 (0.2)
1.8 (0.2)
4.2 (1.1)
4 (0.5)
Abundance
(ind/bag)
63.4 (16.1)
113.2 (26.2)
22.4 (2.4)
24 (5.9)
16.8 (4.7)
6 (1.4)
79.6 (15.0)
9.4 (2.2)
30.2 (9.3)
47 (17.7)
26.2 (5.0)
55.8 (22.5)
26.8 (8.2)
33.8 (2.4)
103.8 (30.0)
7 (1.3)
35.8 (16.0)
46 (10.2)
15.8 (3.5)
14 (3.8)
Dry weight
(mg)
4.2 (1.64)
13.1 (2.61)
4.1 (0.25)
1.2 (0.56)
3.5 (0.42)
2 (0.11)
4.2 (1.49)
4.3 (0.19)
3.7 (0.85)
7.4 (1.77)
2.3 (0.49)
5.4 (2.23)
1.9 (0.79)
11.7 (0.25)
7.3 (2.98)
16.3 (0.14)
2.1 (1.58)
9.2 (1.04)
3.5 (0.29)
6.1 (0.34)
3.4 Ergosterol
There was a significant difference in ergosterol content among litter species (df=3, F=41.18,
p< 0.001). However, Tukey’s HSD analysis revealed that only birch was significantly different
from the other species (Figure 3).
Figure 3. Mean ergosterol content in different leaf litter species after the
end of the study. Different letters indicate significant differences. Error
bars represent ± 1 SE.
11
3.5 Nutrient content
The analyses of initial nutrient content showed that alder contained the lowest C/N ratio
(16.6), followed by birch (36.2), spruce (45.1) and lodgepole pine with the highest (51.3).
3.6 Litter input
Litter input varied among sites (Appendix 3), and was significantly higher in fall (t-test, df =
19, p< 0.001). The Tukey’s HDS showed that, in fall, the proportion of birch input was
significantly higher (p< 0.001), and the proportion of pine (p< 0.001) and spruce (p<0.001)
was significantly lower than during the summer (Figure 4a). Total birch biomass input was
also significantly higher in fall compared to in summer (p< 0.001) (Figure 4b). Since biomass
does not describe riparian vegetation input, the proportion of total input was used in the
analyses.
Figure 4. Difference in (a) percent of total and (b) total biomass input for different litter categories between
summer and fall. * indicates significant differences between the periods at p = 0.05. Errors bars represent ± 1 SE
Both PCAs for litter input showed negative relationships between amount of spruce needles
and birch litter (Figure 5). The same negative relationship was seen between alder and pine.
For litter input composition in summer, positive component 1 scores represent domination of
spruce and negative values represent domination of birch. Litter input composition during
fall showed the opposite pattern.
12
Figure 5. Principal component analysis (PCA) bi-plot of litter input composition during (a) summer and (b) fall.
The PCA on abiotic factors and riparian vegetation input showed that sites with high input of
birch were associated with high SUVA254 values, high input of other vegetation than needles
and leaves (i.e. ‘other’), high canopy openness, and high stream velocity (Figure 6). Spruce
input was negatively correlated with birch input, but positively related to twig and pine input,
conductivity, and total litter input during summer. Alder showed a positive correlation with
total input during fall and negative correlation with canopy openness, NO3, and DIN.
Figure 6. Principal component analysis (PCA) bi-plot of environmental variables,
including litter input from alder, birch, pine, spruce twig and other materials. PC1 and
PC2 together explained 40.8% of the variation (23.5% and 17.24%, respectively).
13
3.7 Variables influencing microbial decomposition
The PLS model for alder decomposition identified stream velocity as the most important
variable, with a positive relationship (Figure 7a). Other important variables were total litter
input during summer, canopy openness, SUVA254, and proportion of ‘other’ in summer litter
input. Of the five most important variables, only total litter input during summer was
negatively correlated to the decomposition of alder. For birch decomposition, proportion of
‘other’ in the summer litter input was the most important variable, followed by canopy
openness, levels of NO3 in the water, litter input composition in fall, and amount of twigs in
litter input during fall (Figure 7b). The positive correlation with litter input composition
means that decomposition of birch litter is faster at sites where birch dominates fall input.
Only proportion of twigs in litter input during fall was negatively related to decomposition.
Litter input composition during summer was the most important variable for explaining
decomposition of lodgepole pine. However, also proportion of spruce in both summer and
fall litter input explained large amounts of variation in lodgepole pine decomposition. Not
surprisingly, amount of birch in summer litter input was negatively correlated with lodgepole
pine decomposition (Figure 7c). For spruce decomposition, DOC was the most important
variable, with a negative relationship (Figure 7d). Also total litter input during summer and
TN concentration showed negative correlations to spruce decomposition, whereas there were
positive relationships with litter input composition in fall (i.e. domination of birch in litter
input) and pH. Because of higher decomposition rate, alder and birch decomposition had
greater impact on total litter mass loss than lodgepole pine and spruce. The PLS for total
mass loss was therefore influenced by the same variables as decomposition of alder and birch
(data not shown).
14
Figure 7. Results from partial least square (PLS) models on microbial decomposition of (a) alder, (b) birch, (c) lodgepole pine, and (d) spruce. The loading value is the strength
of each variable on the model and positive/negative values indicate direction of influence. The percentage in the parentheses is percent variation explained by the first
component of the model.
15
The PCA on explanatory variables and decomposition rates of different litter species showed
that decomposition rates of the deciduous species, i.e. alder and birch, were positively related
to each other (Figure 7). Decomposition rates of conifers were however not correlated, and
only lodgepole pine showed a negative relationship with decomposition of both alder and
birch.
Figure 8. Principal component analysis (PCA) bi-plot showing relationships
between explanatory variables (grey) and microbial litter decomposition rates for
the four litter species (bold). Only the five most influential variables from the PLS
models (Figure 6a-d) are included. PC1 and PC2 axis together explained 34.18% of
the variance in the data (19.64% and 14.52%, respectively).
3.8 Macroinvertebrates
Invertebrate species richness and biomass (both positive) were less important for birch litter
decomposition than total summer litter input and litter input composition in summer (both
negative). Also, canopy openness and litter input composition in fall, both with positive
relation to decomposition, showed higher influence (Figure 9). Since input of spruce needles
dominates litter input during summer, the negative relation to total summer input and litter
input composition in summer means that decomposition is lower where spruce dominates
the riparian vegetation. Positive relation to litter input composition in fall means that
decomposition is higher at sites where birch leaves dominate litter input in fall.
16
Figure 9. Result from PLS analysis on birch decomposition by both microbes and
shredders. The loading value is the strength of each variable on the model and
positive/negative values indicate direction of influence. The percentage in
parentheses is the variance explained by the first component of the PLS.
3.9 Estimation of litter standing stock
Litter loss of standing stock (i.e. alder, birch and spruce litter input in fall) due to microbial
decomposition (and leaching) ranged from 18.6% at site Kr1 to 36.3% at site Bcc. For birch
litter decomposition, an inclusion of shredder fragmentation in the calculations significantly
increased mass loss (t-test, df = 19, t = 6.79, p <0.001) with mass loss ranging from 39% to
72%. The PLS models showed that amount of litter input was the most influential variable for
predicting availability of litter standing stock at the end of the study period (Figure 9a).
When including macroinvertebrate consumption in the analysis, decomposition rate was also
an important determinant (Figure 9b). However, results from both analyses showed that
litter-standing stock primarily is determined by litter input and less so by microbial and
shredder decomposition rates.
Figure 10. Result from PLS analyses on estimation of litter standing stock after the study period and the influence
of different variables. (a) includes litter input from alder, birch and spruce, and only considers microbial
decomposition, whereas (b) only includes input from birch but includes macroinvertebrate consumption as well.
The loading value is the strength of each variable on the model and positive/negative values indicate direction of
influence. The percentage in parentheses is the variance explained by the first component of the PLS.
17
4 Discussion
Previous studies have shown that several factors, such as nutrient availability and
temperature, affect litter decomposition (Webster and Benfield 1986). My results are similar
to those previously found, but I also found that the effect of these factors are much more
complex because litter decomposition of different species were affected differently.
Additionally, I found a high influence of riparian vegetation composition on the in-stream
litter decomposition processes.
4.1 Riparian vegetation
The results from this study show that variation in riparian vegetation, and thus the quantity
and quality of litter input to streams, can have large effects on the in-stream litter
decomposition processes (i.e. carbon and nutrient cycling). Such an influence has only before
been shown for decomposition of high-quality litter (e.g. Kominoski et al. 2011), while I
found that all types of leaf litter, including poor-quality litter, were affected. A possible
explanation for increased litter decomposition in streams surrounded by high-quality litter
species – in this case birch – is that these streams have a greater microbial biomass (Webster
1986), leading to higher decomposition of both high- and low-quality leaf litter. However,
this argument does not hold for lodgepole pine needles that were of lowest quality (and
exhibited the overall slowest decomposition) but decomposed the fastest in streams where
poor-quality litter (i.e. spruce needle) dominated. In this case, an adaptation of the microbial
community to a natural input of low-quality litter could be an explanation (Strickland et al.
2008, Freshet et al. 2012). Besides influencing the microbial community, riparian vegetation
may affect the macroinvertebrate assemblage, if shedder species composition is adapted to
amounts and types of litter in the standing stock (Murphy and Giller 2000, Leroy and Marks
2006). This is in line with the results showing the fastest shredder-mediated litter
decomposition at sites dominated by birch leaf input (Figure 9). Further, results on the
influence of litter input during summer and fall suggest that microbes and shredders are
affected by the variation in quality of availability litter during the whole year.
4.2 Water chemistry and stream physical characteristics
The variation in litter decomposition in response to water chemistry and stream
characteristics, and that the responses differed among litter species, further show the
complexity in understanding drivers of the decomposition processes. Nutrient levels have
often been shown to increase decomposition (Suberkropp and Cauvert 1995, Suberkropp et
al. 2010). My results support this, and that the decomposition processes in boreal headwaters
is NO3 limited (Burrows et al. 2015), but only for alder and birch. The negative influence of
the other major nutrient, phosphorus (PO4), on spruce decomposition can be attributed to
PO4’s positive correlation with DOC, which therefore was negatively associated with spruce
decomposition. Microbes (primarily bacteria) are the main user of DOC in the water (Hall
and Meyer 1998). The observed negative relationship between DOC and spruce
decomposition may therefore be explained by microbes using C from the water rather than
from the litter. Correspondingly, when poor-quality DOC (high SUVA254) increases,
microbes start using leaf litter instead of DOC as energy source, resulting in increased
microbial litter decomposition. Hence, food selection in bacteria might change from litter to
DOC and vice versa, depending on the quality of leaf litter and DOC. This argument is
supported by the fact that DOC only was negatively associated with decomposition of lowquality litter (i.e. spruce), suggesting that DOC is a more preferable energy source than lowquality litter. Overall, these results suggest that the magnitude and effect of increased
nutrient input to headwater streams are dependent on type of system, type of litter, quality
(and amount) of DOC, and type of nutrient that is increased.
The positive influence of stream velocity on litter decomposition can be due increased litter
fragmentation, which not only increases mass loss directly but also indirectly by promoting
microbial colonization, and thus decomposition, due to increased surface area (Spänhoff et
al. 2007). With higher stream velocity sediment transport increases, further increasing
18
physical fragmentation of litter (Spänhoff et al. 2007). High toughness in needles could
explain why velocity did not affect decomposition of lodgepole pine needles. With the
predicted higher precipitation during wet season due to climate change (Whitehead et al.
2009), stream velocity is believed to increase, which may ultimately lead to increased
fragmentation and a likely increased litter decomposition.
Lodgepole pine decomposition deviates from the other species, as it was the only species that
was influenced by temperature, despite several studies arguing for the importance of
temperature for controlling microbial production and, hence, litter decomposition. In an
experiment investigating temperature effects on decomposition of litter of different qualities,
Fierer et al. (2005) showed that decomposition of low-quality litter was more sensitive to
changes in temperature than decomposition of high-quality litter. My results confirm these
findings, as lodgepole pine had the lowest litter quality of the four studied species. However,
the low variation in mean temperature (0.073 °C) between the streams I studied might not be
enough to significantly detect changes in the decomposition of alder, birch and spruce. The
effect of increased temperature on litter decomposition is also greatest at low temperatures
(Fierer et al. 2005), which is why effects of a small change in temperature may have larger
effect depending on what range of temperatures that are investigated. Therefore, in context
of climate change, it is important to not only consider a potential increase in mean
temperature, but also the temperature range. Nevertheless, my results suggest that increased
temperature will increase litter decomposition rates for lodgepole pine litter.
4.3 Adaptation to non-native species
Exotic species often differ in timing, quality and quantity of litter input to aquatic systems
compared to native species (Casas et al. 2013). My results show that aquatic systems’ ability
to decompose non-native species differ compared to native species, even compared to species
of similar litter quality (i.e. spruce). Additionally, lodgepole pine had the lowest
decomposition rate, which implies a low secondary production and production of FPOM in
stream where lodgepole pine needles dominates litter input (i.e. source availability to stream
detritivores). Whether the observed differences in litter decomposition of lodgepole pine
when compared to native species is due to the fact that it is alien to the streams (and their
organisms) or just because the litter is of low quality is hard to tell, because exotic species
often produce litter of low quality (Casas et al. 2013). However, results from Canhoto and
Graça (1996) showed no difference in decomposition when comparing native and non-native
species, but found a significant effect of litter quality, suggesting that litter quality is the main
driver. While my results do not refute these findings, the fact that lodgepole pine
decomposition was differently associated with the main influencing variables than
decomposition of the other low-quality species (i.e. spruce) suggests that other variables than
just quality can be important. The continuous use of lodgepole pine in the Swedish forestry,
sector may have negative effects on the productivity of headwater ecosystems but also on
downstream freshwater and near-shore ecosystems, such as lakes and wetlands
4.4 Variable associations
The positive association between birch input and several variables that in turn were
positively related to decomposition of alder, birch, and spruce decomposition shows that
riparian vegetation, stream characteristics and water chemistry change simultaneously as the
systems undergoes succession. Supporting this is the fact that, after disturbance in the
catchment, birch is common in an early-successional forest (Essen et al. 1997), as are
increased water velocity, decreased canopy cover, and high discharge of DOC and nutrients
(Swank et al. 2001). The positive relationship of SUVA254 to birch litter input, and negative
to spruce litter input, indicates that DOC quality may also follow forest succession, with poorquality DOC in an early-successional forests and high-quality DOC in late succession forests.
This is supported by the fact that, with exception of wetlands, the largest proportion of DOC
in streams originates from the terrestrial environment from either soil or leached from leaf
litter. During fall, after defoliation of leaves, the largest part of DOC originates from leaf litter
19
(Hongve 1999). Depending on tree species, both amount and content of leached DOC will
differ, which affects the water chemistry in the aquatic system (Hongve 1999). This change in
amount and content of DOC due to changes in tree species composition further shows how
the succession in the terrestrial and aquatic environment is associated.
These relationships between riparian vegetation, water chemistry, and stream characteristics
give valuable information for optimization of management methods for riparian zones, to
prevent adverse forestry effects on small streams. If the aim is to create buffer zones that
minimize negative effects and mimic natural systems after disturbance, riparian vegetation
should consist of early-successional tree species. This will create systems that represent water
chemistry and riparian vegetation that are needed in order to maintain landscape-scale
heterogeneity and habitats with rapid in-stream litter decomposition, high production of
FPOM, and high secondary production, which current management often fails to do
(Kuglerova et al. 2014).
4.5 Decomposer community and litter decomposition
The higher decomposition rate of deciduous species, compared to coniferous species, was
expected, based on result from previous studies (Petersen and Cummins 1974). Since no
previous study has compared the exact same species composition and used the same study
design in a boreal system, the specific decomposition rates (i.e. k) are hard to compare
between studies. Nevertheless, the negative relation between litter C/N ratio and
decomposition rates confirms the influence by litter stoichiometry. In addition, secondary
compounds in litter, which were not measured in this study, could be important.
Levels of ergosterol in birch and alder leaves were similar to those reported by Haapala et al.
(2001), but more surprisingly, fungal biomass did not correlate with either litter C/N ratio or
decomposition rates. Since fungi are the main contributor to stream microbial composition of
litter (Gulis and Suberkrop 2003), it is unlikely that a difference in bacterial biomass can
explain why ergosterol levels and decomposition rates are not related. Leaching of soluble
compounds from the litter during the first days in the decomposition process has large effects
on subsequent biomass loss and quality of remaining litter biomass (Petersen and Cummins
1974). Further, the leaching rate can vary significantly between species (McDowell and Fisher
1976). Therefore, a higher mass loss due to leaching in alder, and subsequent rapid decrease
in quality of remaining litter biomass, could explain why alder, despite its low ergosterol
content, had decomposed faster than birch. Further, since fungal colonization patterns differ
between species (Gessner and Chauvet 1994), the peak in fungal mass occurs at different
times, with litter of low quality (i.e. lodgepole pine and spruce) reaching its peak in fungal
biomass later. With only one measurement for the whole study period, the risk of
underestimating ergosterol levels for species that peak in ergosterol earlier (i.e. primarily
alder) is pronounced, and may explain the lack of relationship between ergosterol and
decomposition rates in this study. The temporal difference in peak biomass could also
explain why only decomposition of lodgepole pine was highly influenced by fungal activity,
that presumably show a late peak in ergosterol levels due to its low-quality litter. This is
despite the fact that previous studies have shown the importance of fungal activity in litter
decomposition.
The generally much higher decomposition rate in coarse- than in fine-mesh bags
demonstrate the importance of shredder fragmentation in the decomposition process, as
would be expected at high latitudes and hence low water temperatures (e.g. Iron et al. 1994).
Previous studies have shown that variation in abundance of shredders is the main
explanation for variation in litter decomposition rates (Graca 2001). However, the result
from this study showed that shredder species richness and biomass is more important than
abundance in decomposition of birch leaves, but less important than some other
environmental factors, such as litter input and water chemistry (e.g. Jonsson et al. 2001). The
influence of shredders species richness on litter decomposition shows the importance of
biodiversity for maintaining the function of ecosystem processes, even if specific shredder
species might be more important than diversity per se (Robinson et al. 1998). Hence, even if
20
environmental variables such as litter input and water chemistry are the most important for
litter decomposition, loss of species will potentially affect the whole stream network, since it
decreases litter decomposition, with consequences for both nutrient cycling and production
of FPOM, and subsequent water quality for downstream systems. Maintaining a high
biodiversity of shredders should therefore be of great interest to maintain good health and
function of headwaters.
Although shredder biomass and species richness were influential, birch litter decomposition
was most influenced by variables affecting microbial decomposition. Increased microbial
colonization can change litter stoichiometry and make litter more palatable for shredders and
their feeding rate more efficient (e.g. McKie and Malmqvist 2009, Fuller et al. 2014), which
would explain why variables affecting microbial decomposition is more influential than
shredder biomass and species richness when predicting decomposition.
4.6 Litter standing stock
The strong influence by litter input for predicting litter-standing stock showed that litter
input is the most important factor for litter standing stock in the streams (Figure 10).
Although my calculations predicted that biomass was still left after the study period, a loss of
72% during fall indicates that food shortage may occur during winter or early spring. As this
study shows, decomposition of high-quality litter is faster and will disappear first from the
litter standing stock. The availability of low-quality litter might therefore be important to
prevent food shortage for in-stream biota. This highlights the importance of both low- and
high-quality litter for maintaining the productivity of headwater streams during the whole
year (e.g. Cummins et al. 1989, Haapala et al. 2001). Naturally, this is achieved by a diverse
tree species composition in the riparian zone, created by small-scale disturbance (i.e dying
trees), which enhance colonization of early successional species. In the managed forest that
typically consist of even age monocultures (Essen et al 1997), a more variable buffer zone can
be created by cutting down individual trees to mimic natural disturbance. My estimations
only included three of the six litter input categories. Therefore, the standing stock biomass is
likely greater, but whether an inclusion of the other categories would influence the rate of
loss is harder to predict, since it depends on the litter quality. Still, birch litter input was by
far the most important determinant for litter standing stock in the streams.
4.7 Conclusion
The results from this study show that environmental influences on litter decomposition differ
among species. This suggests that caution is needed when making general conclusions about
how processes might respond to environmental change. Further, the results suggests that
laboratory experiments on litter decomposition, where single influencing parameters often
are investigated, might not be relevant to natural conditions in which sets of interacting
parameters might influence the processes in the same or opposite direction. The observed
differences in decomposition rate between species suggests that the species contributing to
litter input can have great impact on an aquatic system’s secondary production, results that
should be considered in management of riparian zones. This study also shows the importance
of both microbial and shredder activity in the litter decomposition process. A reduced activity
of either can have huge impacts as it changes the energy base for the food web and modifies
the transport of particles and nutrients to downstream regions.
Stream food webs seem to be able to adjust to changes in the quality of litter input that
naturally occurs as the forest undergoes succession. Supporting this is the fact that streams
with riparian vegetation input dominated by early-successional species showed higher
decomposition rates of late-successional species litter. However, the non-native species
differed from this pattern, once again showing the difficulty to make general predictions for
different types of systems and litter. Further, when exotic species are intentionally
introduced to terrestrial systems, there is often little concern regarding influences on
adjacent aquatic systems. Results from this study show that a non-native tree species can
21
adversely impact in-stream litter decomposition rates, with likely negative consequences for
secondary production and energy transport to downstream aquatic ecosystem.
5 Acknowledgement
Special thanks to my supervisor Micael Jonsson and to Ryan Sponseller for making this
project happen. Also a great thanks to Ryan Burrows for field assistance and together with
Anna Sundelin and Emma Andersson have given valuable feedback on the report. Further
thanks to all the students from the masterstudent-room for help and interesting discussions
during the writing process.
22
6 References
Aerts, R. 1997. Climate, leaf litter chemistry and leaf litter decomposition in terrestrial
ecosystems: a triangular relationship. Oikos 439-449.
Allan, J.D. 2004. Landscape and riverscapes: the influence of land use on stream ecosystems.
Annual review of ecology, evolution and systematics 35:257-284.
Bärlocher, F. 1985. The role of fungi in the nutrition of stream invertebrates. Botanical
journal of the Linnean Society 91:83-94.
Benfield, E. F. 1996. Leaf breakdown in stream ecosystems. In: Methods in stream ecology,
F. R. Hauer, and G. A. Lambert (eds.), 711-720. Second edition. Burlington: Academic Press.
Bernhardt, E. S., Likens, G. E., Buso, D. C., & Driscoll, C. T. 2003. In-stream uptake dampens
effects of major forest disturbance on watershed nitrogen export. Proceedings of the
National Academy of Sciences 100:10304-10308.
Bishop, K., Buffam, I., Erlandsson, M., Fölster, J., Laudon, H., Seibert, J. and Temnerud, J.
2008. Aqua Incognita: the unknown headwaters. Hydrological processes 22:1239-1242.
Burrows, R. M., Hotchkiss, E. R., Jonsson, M., Laudon, H., McKie, B. G. and Sponseller, R. A.
2015. Nitrogen limitation of heterotrophic biofilms in boreal streams . Freshwater Biology
60:1237–1251.
Canhoto, C. and Graça, M.A.S. 1996. Decomposition of Eucalyptus globules leaves and three
native leaf species (Alnus glutinosa, Castanea sativa and Quercus faginea) in a Portuguese
low order stream. Hydrobiologica 333:79-85.
Casas, J. J., Larrañaga, A., Menéndez, M., Pozo, J., Basaguren, A., Martínez, A., Pérez, J.,
González, J. M., Salvador, M., Casado, C., Descals, E., Roblas, N., López-González, J. A. and
Valenzuela, J. L. 2013. Leaf litter decomposition of native and introduced tree species of
contrasting quality in headwater streams: How does the regional setting matter? Science of
the Total Environment 458:197-208.
Cummins, K. W., Wilzbach, M. A., Gates, D. M., Perry, J. B., Taliaferro, W. W. 1989.
Shredders and riparian vegetation. BioScience 39:24-30.
Dangles, O., & Malmqvist, B. 2004. Species richness–decomposition relationships depend on
species dominance. Ecology Letters 7:395-402.
Engelmark, O., Sjöberg, K., Andersson, B., Rosvall, O., Ågren, G. I., Baker, W. L., Barklund,
P., Björkman, C., Despain, D.G., Elfving, B., Ennos, R.A., Karlman, M., Knecht, M.F., Knight,
D.G., Ledgard, N.J., Lindelöw, Å., Nilsson, C., Peterken, G.F., Sörlin, S. and Sykes, M. T.
2001. Ecological effects and management aspects of an exotic tree species: the case of
lodgepole pine in Sweden. Forest Ecology and Management 141:3-13.
Essen, P.A., Ehnström, B., Ericson, L. and Sjöberg, K. 1997. Boreal forests. Ecological
Bulletins 46:16-47.
Fierer, N., Craine, J. M., McLauchlan, K. and Schimel, J.P. 2005. Litter quality and the
temperature sensitivity of decomposition. Ecology 86:320-326.
Fisher, S. G., and Likens, G. E. 1973. Energy flow in Bear Brook, New Hampshire: an
integrative approach to stream ecosystem metabolism. Ecological monographs 43:421-439.
Freshet, G. T., Aerts, R. and Cornelissen, J. H. C. 2012. Multiple mechanisms for trait effects
on litter decomposition: moving beyond home-field advantages with a new hypothesis.
Journal of Ecology 100:619-630.
Fuller, C. L., Evans-White, M. A., and Entrekin, S. A. 2014. Growth and stoichiometry of a
common aquatic detritivore respond to changes in resource stoichiometry. Oecologia 1-12.
23
Gessner, M. O., and Chauvet, E. 1994. Importance of stream microfungi in controlling
breakdown rates of leaf litter. Ecology 75:1807-1817.
Graça, M.A.S. 2001. The Role of Invertebrates on Leaf Litter Decomposition
in Streams – a Review. International Review of Hydrobiology 86:383-393.
Gulis, V. and Suberkropp, K. 2003. Leaf litter decomposition and microbial activity in
nutrient-enriched and unaltered reaches of headwater stream. Freshwater Biology 48:123134.
Haapala, A., Muotka, T., and Markkola, A. 2001. Breakdown and macroinvertebrate and
fungal colonization of alder, birch, and willow leaves in a boreal forest stream. Journal of the
North American Benthological Society 20:395-407.
Hall, R. O. and Meyer, J. L. 1998. The trophic significance of bacteria in a detritus-based
stream food web. Ecology 79:1995-2012.
Honve, D. 1999. Production of dissolved organic carbon in forested catchments. Journal of
Hydrology 224:91-99.
Hulland, J., Ryan, M.J. and Rayner, R.K. 2010. Modeling customer satisfaction: A
comparative performance evaluation of covariance strucyure analysis versus partial least
squares. In: Handbook of partial least squares: Concepts, methods and applications. V.
Esposito Vinzi, W. W. Chin, J. Henseler, H. Wang (eds.), 307-325. Berlin-Heidelberg:
Springer.
Irons, J. G., Oswood, M. W., Stout, R., & Pringle, C. M. 1994. Latitudinal patterns in leaf litter
breakdown: is temperature really important? Freshwater Biology 32:401-411.
Jonsson, M., Malmqvist, B., Hoffsten, P.-O. 2001. Leaf litter breakdown rates in boreal
streams: does species richness matter? Freshwater Biology 46:161-171.
Kempe, G. 2014. Skog och skogsmark, Forest and forestry. In Skogsstatistisk årsbok,
Swedish statistical yearbook of forestry. L. Christiansesn (eds.), 41-70. Jönköping: Swedish
forest agency.
Kominoski, J.S., Marczak, L. B. and Richardson, J.S. 2011. Riparian forest composition
affects stream litter decomposition despite similar microbial and invertebrate communities.
Ecology 92:151-159.
Kuglerová, L., Ågren, A., Jansson, R. and Laudon, J. 2014. Towards optimizing riparian
buffer zones: ecological and biogeochemical implications for future management. Forest
ecology and management 334:74-84.
Leroy, C. J. and Marks, J. C. 2006. Litter quality, stream characteristics and litter diversity
influence decomposition rates and macroinvertebrates. Freshwater Biology 51:605-617.
Manning, D. W. P., Rosemond, A. D., Kominoski, J. S., Gulis, V., Benstead, J. P., and Maerz,
J. C. 2015. Detrital stoichiometry as a critical nexus for the effects of streamwater nutrients
on leaf litter breakdown rates. Ecology. http://dx.doi.org/10.1890/14-1582.1
McDowell, W. H. and Fisher, S. G. 1976. Autumnal processing of dissolved organic matter in
a small woodland stream ecosystem. Ecology 57:561-569.
McKie, B. G. and Malmqvist, B. 2009. Assessing ecosystem functioning in streams affected
by forest management: increased leaf decomposition occurs without changes to the
composition of benthic assemblage. Freshwater Biology 54:2086-2100.
Meyer, J.L., Edwards, R.T. and Risley, R. 1987. Bacterial growth on dissolved organic carbon
from a blackwater river. Microbial ecology 13:13-29.
Meyer, J. L., Strayer, D. L., Wallace, J. B., Eggert, S. L., Helfman, G. S., & Leonard, N. E.
2007. The contribution of headwater streams to biodiversity in river networks. Journal of the
American water resource association 43:86-103.
24
Murphy, J. F. and Giller, P. S. 2000. Seasonal dynamics of macroinvertebrate assemblage in
the benthos and associated with detritus packs in two low-order streams with different
riparian vegetation. Freshwater biology 43:617-631.
Perdue, E. M. 1998. Chemical composition, structure, and metal binding properties. In
Aquatic humic substances, D. O. Hessen and L. J. Tranvik (eds.), 41-61. Berlin Heidelberg:
Springer.
Petersen, P.C. and Cummins, K.W. 1974. Leaf processing in a woodland stream. Freshwater
Biology 4:343-368.
R Core Team. 2012. R: A language and environment for statistical computing. R Foundation
for Statistical Computing, Vienna, Austria.
Robinson, C.T., Gessner, M. O. and Ward, J. V. 1998. Leaf breakdown and associated
macroinvertebrates in alpine glacial streams. Freshwater Biology 40:215-228.
Rosemond, A. D., C. M. Swan, J. S. Kominoski, and S. E. Dye. 2010. Non-additive effects of
litter mixing are suppressed in a nutrient-enriched stream. Oikos 119:326-336.
Schmidt-Kloiber A. and Hering D. (eds.) 2012. www.freshwaterecology.info - the taxa and
autecology database for freshwater organisms, version 5.0. 2014-12-10.
Spänhoff, B., Augspurger, C., Keusel, K., 2007. Comparing field and laboratory breakdown
rates of coarse particulate organic matter: sediment dynamics mask the impacts of dissolved
nutrients on CPOM mass loss in streams. Aquatic Science 69:495-502.
Stenroth, K., Polvi, L.E., Fältström, E. and Jonsson, M. 2015. Land-use effects on terrestrial
consumers through changed size structure of aquatic insects. Freshwater Biology 60:136149.
Strickland, M. S., Osburn, E., Lauber, C., Fierer, N., & Bradford, M. A. 2009. Litter quality is
in the eye of the beholder: initial decomposition rates as a function of inoculum
characteristics. Functional Ecology 23:627-636.
Swank, W. T., Vose, J. M. and Elliot, K.J. 2001. Long-term hydrologic and water quality
responses following commercial clearcutting of mixed hardwoods on a southern Appalachian
catchment. Forest Ecology and Management 143:163-178.
Suberkropp, K. and Cauvert, E. 1995. Regulation of leaf breakdown by fungi in streams:
influence of water chemistry. Ecology 76:1433-1445.
Suberkropp, K., Gulis, V., Rosemond, A. D., and Benstead, J. P. 2010. Ecosystem and
physiological scales of microbial responses to nutrients in a detritus‐based stream: Results of
a 5 year continuous enrichment. Limnology and Oceanography 55:149-160.
Temnerud J, Bishop K. 2005. Spatial variation of stream water chemistry in two Swedish
boreal catchments: implications for environmental assessment. Environmental Science and
Technology 39:1463-1469.
Wallace, J.B., Eggert, S. L., Meyer, J. L. and Webster, J. R. 1999. Effects of resource
limitation on a detrital-based ecosystem. Ecologiacal monographs 69:409-422.
Wallace, T. A., Ganf, G. G. and Brookes, J. D. 2008. A comparison of phosphorus and DOC
leachates from different types of leaf litter in an urban environment. Freshwater Biology
53:1902-1913.
Webster, J.R and Benfield, E.F. 1986. Vascular plant breakdown in freshwater ecosystems.
Annual review of ecology and systematics 17:567-94.
Webster, J. R., Golladay, S. W., Benfield, E. F., Meyer, J. L., Swank, W. T., and Wallace, J. B.
1992. Catchment disturbance and stream response: an overview of stream research at
Coweeta Hydrologic Laboratory. River conservation and management 15:232-253.
25
Weishaar, J. L., Aiken, G. R., Bergamaschi, B. A., Fram, M. S., Fujii, R., and Mopper, K.
2003. Evaluation of specific ultraviolet absorbance as an indicator of the chemical
composition and reactivity of dissolved organic carbon. Environmental Science and
Technology 37:4702-4708.
Wipfli, M.S. and Gregovich, D.P. 2002. Export of invertebrates and detritus from fish free
headwater streams in southeastern Alaska: implication for downstream salmonid
production. Freshwater Biology 47:957-969.
Whitehead, P. G., Wilby, R. L., Battarbee, R. W., Kernan, M., & Wade, A. J. 2009. A review of
the potential impacts of climate change on surface water quality. Hydrological Sciences
Journal 54:101-123.
26
Appendix 1
Table 1. Mean values for the water chemistry parameters during the study period. Values in parentheses are variance between the measurements.
Site
B1
BCC
BF
G1
G2
G3
K1
K5
K8
KL
Kr1
Kr6
Kr7
Röd
S2
S26
S3
S6
V1
V2
PO4
(µg/L)
3.20 (0.72)
6.03 (0.63)
5.04 (0.52)
3.11 (0.77)
1.06 (0.26)
1.26 (0.37)
6.12 (0.71)
1.43 (0.36)
2.51 (0.14)
2.40 (0.23)
2.64 (0.39)
3.53 (0.48)
9.87 (2.23)
2.00 (0.34)
11.73 (1.06)
1.53 (0.09)
3.95 (0.78)
1.90 (0.23)
2.10 (0.43)
1.76 (0.16)
NH4
(µg/L)
27.47 (8.35)
12.79 (1.47)
7.12 (0.73)
5.66 (0.76)
7.33 (1.15)
3.65 (0.30)
5.46 (0.93)
4.75 (0.69)
11.22 (1.44)
8.13 (1.83)
20.66 (3.16)
16.46 (2.45)
11.87 (1.08)
9.33 (2.04)
11.26 (1.01)
4.30 (0.52)
10.91 (1.17)
5.92 (0.19)
7.16 (1.77)
5.56 (1.04)
NO3
(µg/L)
12.10 (1.57)
60.04 (12.41)
7.00 (0.25)
5.73 (0.46)
8.07 (1.26)
7.00 (0.66)
7.52 (1.94)
11.48 (4.05)
17.69 (1.89)
8.41 (1.25)
28.14 (2.77)
16.81 (2.64)
16.49 (1.43)
18.80 (5.02)
9.20 (0.36)
8.80 (1.83)
7.88 (0.68)
16.75 (4.58)
14.25 (3.34)
9.20 (1.78)
DIN
(µg/L)
39.57 (9.88)
72.82 (12.89)
14.12 (0.95)
11.39 (1.14)
15.40 (2.36)
10.64 (0.36)
12.99 (2.51)
16.23 (3.73)
28.91 (2.68)
16.54 (3.02)
48.79 (4.39)
33.27 (5.05)
28.36 (1.36)
28.12 (7.01)
20.46 (1.32)
13.09 (2.08)
18.79 (0.87)
22.67 (4.75)
21.42 (4.78)
14.76 (2.38)
TN
(mg/L)
0.33 (0.05)
0.41 (0.05)
0.37 (0.05)
0.28 (0.05)
0.27 (0.01)
0.15 (0.02)
0.16 (0.04)
0.17 (0.01)
0.35 (0.04)
0.41 (0.10)
0.43 (0.07)
0.35 (0.02)
0.44 (0.07)
0.37 (0.08)
0.47 (0.02)
0.23 (0.04)
0.41 (0.04)
0.34 (0.01)
0.32 (0.11)
0.28 (0.09)
DOC
(mg/L)
19.71 (4.52)
18.93 (3.45)
19.55 (2.40)
17.07 (3.39)
12.97 (0.89)
6.85 (1.03)
9.23 (3.31)
7.61 (0.54)
19.24 (2.99)
23.56 (6.38)
23.15 (4.45)
18.13 (1.51)
27.32 (5.17)
17.74 (3.78)
35.73 (2.82)
10.27 (1.57)
21.26 (2.24)
15.24 (0.49)
17.22 (6.48)
15.12 (5.15)
SUVA254
(L/m-1mg-2)
5.42 (1.76)
5.50 (1.82)
4.96 (1.14)
6.39 (2.35)
4.14 (0.45)
5.47 (1.58)
8.26 (5.37)
4.26 (0.68)
5.86 (1.91)
6.11 (2.35)
5.27 (1.55)
5.24 (0.76)
5.24 (1.48)
4.97 (1.57)
4.24 (0.42)
4.44 (1.23)
4.69 (0.93)
3.95 (0.22)
6.90 (3.07)
6.32 (2.62)
Conductivit
(µS/cm3)
27.48 (1.63)
20.93 (2.63)
22.62 (2.23)
15.67 (5.39)
22.32 (7.29)
15.17 (5.14)
19.93 (3.32)
17.73 (6.02)
15.84 (5.38)
24.24 (2.41)
29.47 (1.12)
19.37 (1.02)
30.21 (2.35)
29.27 (1.28)
39.07 (3.41)
38.96 (1.37)
28.53 (3.43)
44.28 (4.11)
28.33 (1.42)
26.99 (1.15)
pH
5.57 (0.35)
5.13 (0.30)
5.43 (0.21)
5.32 (0.31)
6.40 (0.06)
6.46 (0.09)
6.24 (0.30)
6.10 (0.12)
6.00 (0.24)
5.35 (0.36)
5.45 (0.23)
5.57 (0.24)
4.92 (0.29)
6.23 (0.23)
4.95 (0.37)
6.65 (0.12)
5.42 (0.26)
5.70 (0.10)
6.20 (0.27)
6.17 (0.26)
Appendix 2
Table 1. Shredder species composition for the sites. Values are mean values for each litter-bag.
Species
PLECOPTERA
Nemurella picteti
Nemoura avicularis
Nemoura cinerea
Nemoura flexuosa
Protonemura meyeri
Amphinemura sulcicollis
Taeniopteryx nebulosa
Leuctra digitata
Leuctra nigra
Capnopsis schilleri
TRICHOPTERA
Chaetopteryx villosa
Potamophylax nigricornis
Potamophylax cingulatus
Potamophylax cingulatus latipennis
Micropterna lateralis
Micropterna sequax
Micropterna sp.
Limnephilidae
COLEOPTERA
Helophorus
B1
Bcc
BF
G1
G2
G3
K1
K5
K8
KL
Site
Kr1
Kr6
Kr7
Rod
S2
S26
S3
S6
V1
V2
7
306
1
3
-
38
522
-
2
1
64
3
14
-
8
108
1
-
2
37
15
23
12
2
2
-
9
6
1
-
32
339
3
6
-
4
51
2
-
3
1
138
5
-
8
220
11
1
21
1
-
4
112
10
-
6
306
1
-
4
128
1
-
1
133
79
6
6
19
11
23
693
2
-
1
13
1
-
284
-
225
-
5
6
32
6
8
17
-
15
9
27
8
11
-
-
2
1
1
1
-
2
23
2
2
-
1
8
-
1
6
-
7
5
-
1
1
2
23
1
1
5
-
2
1
1
-
2
-
2
2
-
5
24
-
-
5
-
3
-
7
1
-
1
1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Appendix 3
Table 1. Mean values of mass input (g) of different litter categories to litter trays in summer and fall
Site
B1
BCC
BF
G1
G2
G3
K1
K5
K8
KL
Kr1
Kr6
Kr7
Rod
S2
S26
S3
S6
V1
V2
Alder
0.22
0.00
0.00
0.00
0.00
0.14
0.08
0.12
0.01
1.10
0.00
0.00
0.01
0.61
0.00
0.00
0.02
1.76
0.00
0.00
Birch
0.98
0.19
0.65
0.62
0.28
0.22
0.89
2.70
1.66
1.67
0.60
0.00
0.72
0.16
0.63
0.22
1.29
0.29
1.10
1.74
Other
1.58
1.01
1.03
1.04
1.51
0.78
1.42
1.13
2.38
0.73
1.19
1.24
1.26
0.81
2.74
1.90
1.04
2.68
0.94
1.05
Summer
Pine Spruce
0.12 4.04
0.03 0.03
0.05 1.15
0.50 0.03
0.94 1.22
0.01 1.62
0.30 0.10
0.17 0.08
0.00 2.18
0.00 0.01
0.23 1.75
0.04 2.61
0.01 2.82
0.00 1.68
3.45 4.08
0.47 2.05
1.76 0.45
0.00 0.36
0.01 0.70
0.01 0.12
Twigs
0.32
0.06
0.76
0.76
0.49
1.30
0.71
0.01
1.96
0.57
1.38
1.24
0.70
0.17
0.87
0.41
1.05
3.26
1.31
0.39
Tot
7.26
1.32
3.64
2.96
4.44
4.08
3.49
4.21
8.19
4.08
5.16
5.13
5.52
3.43
11.77
5.07
5.62
8.35
4.06
3.31
Alder
1.57
0.00
0.00
0.00
0.57
1.57
0.04
3.55
1.87
7.90
0.00
0.00
0.00
1.94
0.33
0.00
0.89
23.01
0.61
0.45
Birch
9.55
6.42
8.73
23.34
4.50
4.66
35.72
9.78
28.99
6.91
8.89
1.14
9.69
32.12
12.72
7.76
24.79
13.67
11.28
23.64
Other
1.01
1.30
2.39
4.70
1.93
1.20
1.72
1.61
3.93
2.06
0.54
1.42
0.35
6.20
1.58
2.35
1.23
5.49
3.11
5.71
Fall
Pine
0.27
0.50
0.48
0.18
1.68
0.05
0.90
1.80
0.00
0.00
1.34
0.11
0.04
0.04
7.74
2.46
2.44
0.06
0.13
0.06
Spruce
2.78
0.16
3.60
0.21
5.62
4.37
0.39
0.63
5.78
4.83
2.94
6.60
8.82
0.02
9.53
8.97
2.19
1.43
2.87
0.41
Twig
5.83
0.00
0.44
0.63
0.32
0.25
0.49
0.64
0.39
0.31
0.19
0.45
0.19
0.50
1.42
0.66
0.58
1.15
0.08
0.42
Tot
21.01
8.37
15.64
29.06
14.63
12.10
39.26
18.00
40.95
22.01
13.89
9.71
19.09
40.82
33.32
22.19
32.12
44.80
18.09
30.69
Dept. of Ecology and Environmental Science (EMG)
S-901 87 Umeå, Sweden
Telephone +46 90 786 50 00
Text telephone +46 90 786 59 00
www.umu.se

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project