Weight loss in overwintering below-ground parts of Sonchus arvensis under current and temperature-elevated climate scenarios in Sweden

Summary

Weight loss in overwintering below-ground parts of perennial weeds has been attributed to respiration, but neither its temperature dependence nor its relevance for biomass dynamics under changing climate conditions have been investigated. In two experiments, we quantified weight loss of the perennial weed Sonchus arvensis, by measuring weight changes over time of sprouting roots in dark rooms at temperatures of 4, 8 and 18°C. Dry weight loss rates were 0.47, 0.64 and 1.47% day⁻¹ at 4, 8 and 18°C, respectively, giving a half-life time of 149, 110 and 47 days, respectively. A factor by which weight loss rates increase for every 10° rise in temperature (Q₁₀) was equal to about 2.3. Cumulative weight loss may comprise >40% of the below-ground biomass during overwintering periods. Applying weight loss rates and Q₁₀ to elevated soil temperature projections showed that losses during winter seasons in central Sweden will remain basically constant, the effect of increased weight loss at higher temperatures being balanced by shorter winters. This implies that need for control of S. arvensis in a changing climate will persist, but that shorter winter seasons will provide a longer time window for control of S. arvensis prior to sowing crops.

Introduction

Long term strategies for controlling perennial weeds aim to attain a greater loss of stored carbohydrates over photosynthetic gains in the plant. Optimal weed control requires an understanding of the magnitude and timing of these processes. Resource acquisition and carbohydrate sequestration in growing weeds can be hampered by means of removing and/or damaging growing tissues, while biomass loss in regenerative structures can be increased by means of stimulating activity in roots and rhizomes, thereby leading to resource depletion (Graglia et al., 2006; Vanhala et al., 2006). As climate change is anticipated to bring about shorter winters and longer vegetation seasons under Nordic conditions (Intergovernmental Panel on Climate Change, 2007), longer growth periods may provide longer time for control of growing weeds, both prior to crop establishment and after harvest.

However, little is known about biomass dynamics of perennial weeds over winter and even less about the reaction of plants to changes in winter season length and winter temperatures (Bertrand & Castonguay, 2003). For Sonchus arvensis L., which is a noxious perennial weed under Nordic conditions (Salonen et al., 2001; Andreasen & Stryhn, 2008), process-oriented studies (Eckersten et al., 2010, 2011) have been performed to assess and quantify its biomass allocation patterns, but these did not include the winter period. Fykse (1977) showed that 2-year-old roots of S. arvensis had a source and sink function, while the functional lifetime of roots of Cirsium arvense (L.) Scop. was shorter. A detailed descriptive study of the life cycle of S. arvensis, including the winter period, was made by Håkansson (1969), who found a dry matter loss of >30% over the winter period (November–April), associated with a loss of >30% in water-soluble carbohydrates. These losses were interpreted as respiratory losses, and a reference was made to Åslander (1933), who attributed a decrease of carbohydrates in C. arvense over winter to respiration. Anbari et al. (2011) found that growth performance of S. arvensis was positively correlated with root size and weight in spring. Climate induced changes in respiration over winter consequently may change the need to control S. arvensis. Håkansson (1969) also showed that below-ground losses continued during May, but at the whole plant level, this was compensated for by photosynthetic gains. Tørresen et al. (2010) also reported a reduction in below-ground biomass in S. arvensis of about 30% between late autumn and spring.

Weight losses during the overwintering period are attributed to respiration and pathogen-mediated decay. As root respiration and microbial activity are temperature dependent processes (Atkin et al., 2000), the magnitude of weight losses in regenerative structures during the period in which no photosynthesis occurs will depend on the temperature conditions, the length of this period and on the possible occurrence of a dormancy phase. This phase matches low metabolic activity (Van Der Schoot & Rinne, 2011) and low respiratory loss rates. On the basis of the occurrence of temporarily reduced shoot sprouting, S. arvensis has been claimed to be a species, which develops an innate dormancy (Håkansson & Wallgren, 1972a; Fykse, 1974), thereby preventing effective postharvest weed control. However, Pegtel (1972) did not find reduced sprouting in S. arvensis. Brandsæter et al. (2010) reported a gradually increased sprouting activity from late September and onwards, in a range of Danish and Fennoscandic ecotypes of S. arvensis.

While weight loss of S. arvensis over winter is probably caused by respiration and thereby is expected to be temperature dependent, studies thus far have not examined temperature dependence or its relevance for the life cycle of S. arvensis under climate change. Consequently, we performed two experiments which allowed for quantification of weight loss in regenerative structures of S. arvensis as a function of temperature. The hypotheses were that the relative weight loss rate (% day⁻¹) of below-ground parts of S. arvensis (i) is temperature dependent and (ii) for a given temperature is independent of the substrate in which the below-ground parts are placed. In addition, the experimental results were extrapolated to assess the effect of current seasonal temperature variability and the effects of warmer, shorter winters due to climate change on the weight loss during winter, testing the hypothesis (iii) that warmer winters will lead to an increase in cumulative weight loss of below-ground parts of S. arvensis.

Materials and methods

Experimental design and measurements

Root runners of S. arvensis were collected in the autumn of 2005 from organically grown short-term ley fields with cereals in the rotation at Sala (59°40′N, 16°40′E) and planted in buckets filled with soil. This material constitutes a plant bank, which was stored over winters in buckets with a soil volume of 10 L from October until June in a dark cold store at +2 to +4°C, and grown outdoors in the same buckets during the summers of 2006–2008 from June to October. During each year, buckets were taken from the cold store in June, roots were replanted, and an early fertilisation of about 70 kg N ha⁻¹ was provided, while the soil was kept moist during growing seasons.

On 29 June 2009, buckets were taken from the cold store, and roots were distributed into three thickness categories; fine (<2 mm), medium (2–4 mm) and thick (>4 mm) in diameter. Roots of each category were cut in 5- and 10-cm pieces, respectively, with at least two adventitious buds, but no sprouted buds, on each piece. These prepared root fragments were kept in the cold store prior to performing two experiments.

To determine the ratio between dry weight and fresh weight of the roots, 10 roots were randomly selected from each of the three thickness categories and of the 5 and 10 cm length categories, in total N = 60. After measuring fresh weight, the roots were dried at 50°C to constant weight (c. 3 days), and individual root dry weight was determined.

In the first experiment, 15 buckets with 5 cm root parts and 15 buckets with 10 cm root parts were placed in each of three dark chambers, with temperatures of 4, 8 and 18°C, respectively, on 30 June 2009. The soil in each of the 90 buckets (5 L per bucket) consisted of 85% moderately decomposed peat, 15% sand and about 1 g N per bucket and NPK proportion of 2:1:2 (Hasselfors Garden AB). For each temperature and root length, 81 roots were planted, so that three roots (one root for each thickness category) were present in each layer at a depth of 3, 10 or 17 cm below soil surface (all together 27 roots × 3 units = 81 roots), as described in Table 1. As emergence time is known to be negatively correlated with planting depth (Håkansson & Wallgren, 1972b), this design, which employs a wider depth range, was expected to give a larger time period over which soil grown (non-etiolated) shoots could be harvested.

Table 1. Experiment I and II: number of harvests, experimental units (buckets/pots), planting depths within units and number of roots per unit, for each temperature (4, 8, 18°C) and root length (5, 10 cm)

Experiment	Harvest occasion	Experimental units	Planting depth (cm)	Number of roots/unit
I	1	3	3, 10, 17	9
I	2	3	3, 10	6
I	3	3	3, 17	6
I	4	3	10	3
I	5	3	17	3
II	1–10	1	0	6

For the second experiment, referred to as the sand layer experiment, 60 pots with a diameter of 12 cm and a height of 11 cm (1 L) were prepared. A 2 cm washed, wet fine sand layer (Rådasand, Lidköping, Sweden) was placed at the bottom of each of the pots. In each of the pots, two randomly chosen roots of each of three thickness categories were placed on the sand layer, that is, six roots per pot, using 5 cm root parts in 30 of the pots and 10 cm root parts in the other 30 pots. The roots were not covered by any substrate. Ten pots with 5 cm root parts and 10 pots with 10 cm root parts were placed in each of the chambers, with temperatures of 4, 8 and 18°C, respectively, on 1 July 2009.

Prior to planting both experiments, fresh weight of the roots was measured for each experimental unit. All buckets and pots were sealed with plastic and kept moist throughout the experiment. Harvest of plants in buckets was performed well before or around the time the first shoot per bucket broke the soil surface and harvest took place during a time span ranging from about 120 days for roots at 4°C to about 20 days for roots at 18°C. Harvest of plants in pots (one pot per root length per harvest occasion) took place during a time span ranging from about 100 days for roots at 4°C to about 50 days for roots at 18°C. Dry weight of the remaining roots, fine roots and shoots per bucket or pot was determined after drying at 50°C to constant weight. In the bucket experiment, fine roots could not be retrieved from the soil and were discarded. Assuming that the allocation proportion to fine roots for a given temperature was equal in both experiments, and the fine root proportion resulting from the pot experiment was used to calculate fine root production in the buckets.

Estimation of dry weight loss rates

In analogy with weight loss due to respiration and microbial decay, we consider weight loss as a first order process, implying that dry weight loss overtime (dW/dt) is proportional to the amount of dry weight (W) present at that time:

• (1)

where k is the daily relative weight loss rate. By integration over time from t = t₀ when W = W(t₀) to t = t₁ when W = W(t₁), we obtain:

• (2)

For the relative decrease in W over time (W_Norm(t₁)  = 100 × W(t₁)/W(t₀), that is, W_Norm(t₀) = 100), the value of k equals the slope of the linear relation between ln(W_Norm(t₁)) and time with a fixed intercept equal to 4.605 (=ln(W_Norm(t₀)) = ln(100)). We then obtain:

• (3)

and estimated k by means of a linear regression procedure. Knowing the value of k, the half-lifetime (t_½) of the biomass can be calculated as ln(2)/k.

When considering how the relative weight loss rate (k) is changing as a function of temperature, we may employ the temperature coefficient (Q₁₀), which represents the factor by which k increases for every 10°C temperature (T) increment (Ryan, 1991). Q₁₀ can be calculated as:

• (4)

where k₁ and k₂ are weight loss rates at temperatures T₁ and T_2, respectively.

Application to current and elevated soil temperatures

To assess the effects of current climate variability on weight loss in below-ground overwintering structures of S. arvensis, the experimental results were extrapolated using records of daily average soil temperature, measured at 5 cm below soil surface (T₅), retrieved from the Ultuna Meteorological Station (Karlsson & Fagerberg, 1995) close to Uppsala (59.48°N, 17.39°E), for the period 2004–2010, which included six entire overwintering periods. The start of the overwintering period with root activity was defined as days with soil temperature at 5 cm depth (T₅) becoming <5.0°C, assuming a release of reduced sprouting at lower temperatures in S. arvensis roots (Brandsæter et al., 2010). The end of the overwintering season was set at T₅ ≥ 4.0°C, as the emergence time of S. arvensis in spring was conceived as a function of soil warming (Lemna & Messersmith, 1990). Weight loss was estimated on a daily basis using Eqn (1) and a k-value as function of temperature according to Eqn (4), giving:

• (5)

where k_Base, Q₁₀ and T_Base were taken from the results of the controlled experiment in pots, where all biomass could be retrieved. The accumulated loss over the season was estimated by solving W(t + 1) = W(t)−k × W(t) daily from the day of start to the day of end of the overwintering season and evaluated for the relative decrease over the whole period (W(t = t_End)/W(t = t_Start)). Equation (5) was applied with two different parameterisations, to test for sensitivity to Q₁₀ and to T_Base. First, the Q₁₀ value of 2.274, obtained over the entire experimental range of 4–18°C, a k_Base-value of 0.0147 and a T_Base of 18°C were used. Secondly, the Q₁₀ value of 2.102, obtained over the range of 4–8°C, a k_Base-value of 0.0047 and T_Base of 4°C were applied.

To evaluate possible effects of alterations in soil temperature due to climate change on average temperature and length of the active below-ground period, the daily mean T₅-values were raised constantly by 1, 2, 3 and 4°C, respectively. This gave in addition to the six actual seasons, 24 temperature-elevated seasons, for which cumulative weight loss was determined as described above.

Analyses

Linear regression analysis procedures were used to establish relations between weight values and time. In case of known intercepts, the general model Y = a + b × x was reduced into a fixed intercept model. A general factorial anova (SAS Institute, 2002–2004) was used to assess the effects of root size categories on initial dry matter content, after having ensured an approximate normal distribution of the latter (SAS Institute, 2002–2004). A mixed model (Proc Mixed) was used to assess the effects of temperature (α), substrate (β) and root length (γ) on weight loss overtime (b). After testing for significance of interactions, the final model used was as follows:

• (6)

In Eqn (6), r_ijk is normally distributed (N(0, )), where s_ijk is the standard error of b_ijk, and e_ijk is N(0, σ²).

Results

Quantification of weight loss

Neither thickness, length nor their interaction had a significant effect on the dry matter content of the root parts at planting (Table 2). Consequently, the average value of the root dry weight to fresh weight ratio (0.178) was used to estimate root dry weight at planting from root fresh weight at planting for all roots, regardless of size category, in both experiments. Fine root dry weight as a percentage of total plant dry weight for plants grown on sand layers was larger (P < 0.001) for plants grown at higher temperatures (Fig. 1). Biomass loss rates of plants grown on sand layers in the dark increased with temperature, being −0.466, −0.628 and −1.472% per day for temperatures of 4, 8 and 18°C, respectively (Fig. 2). By means of regression analysis, it was shown that biomass loss rates, for given substrates at given temperatures, did not differ significantly between root parts of 5 and 10 cm length (Table 3). We obtained half-life times of 149, 110 and 47 days for dry weight of the sand layer grown plants of S. arvensis at 4, 8 and 18°C, respectively. For the plants which were grown in soil, loss rates of the retrieved biomass component (planted root and shoots, but no fine roots) were significantly higher than those for plants on sand layers at the same temperatures (Table 3). However, when adjusting for the non-retrieved fine root biomass using the fine root percentages from the sand layer grown plants (Fig. 1), there were no significant differences in biomass loss rates for a given temperature between soil grown and sand layer grown plants (Table 3). An anova, which included the standard errors around the weight loss rates, showed that neither root length nor substrate had an effect on biomass loss rate. The effect of temperature was highly significant (Table 4).

Table 2. anova results, specifying the effects of Sonchus arvensis root thickness, root length and their interaction, on initial dry weight over fresh weight ratio (dry matter content, DMC) at planting

Dependent variable	Experimental factor	DF	F-value	P > F
Type 3 tests of fixed effects. DF, degrees of freedom; F-value, treatment mean square/error mean square; P > F, significance probability value.
DMC	Root thickness	2	0.198	0.821
	Root length	1	0.038	0.846
	Root thickness × root length	2	0.201	0.819

Figure 1.  Fine root dry weight expressed as a proportion of total plant dry weight (FRP) for plants grown on sand layers at temperatures of 4, 8, and 18°C in the dark.

Figure 2.  Relative weight loss, ln(100 × W(t₁)/W(t₀)), over time (day) at 4, 8 and 18°C, where 100 × W(t₁)/W(t₀) represents the percentage dry weight remaining over time (in days) of plants grown on sand layers, k is the weight loss proportion per day, W(t₁) is the dry weight of old root, fine roots and shoots at harvest and W(t₀) is the dry weight of the root at planting.

Table 3. Estimates of weight loss rates and their standard errors at temperatures of 18, 8 and 4°C, by means of a fixed intercept model: ln(100 × W(t₁)/W(t₀))=4.605 + k × time, where 100 × W(t₁)/W(t₀) represents the percentage dry weight remaining over time (days), k the relative weight loss rate (% day⁻¹), W(t₁) the dry weight at harvest and W(t₀) the dry weight at planting. Roots were either grown in soil (substrate = 1) or on a sand layer (substrate = 2). Substrate 3 denotes roots grown in soil, but with a weight correction for non-retrieved fine roots

Temperature (°C)	Substrate	Root length (cm)	k	SE	R^2
4	1	5	−0.68	0.080	0.178
4	1	10	−0.69	0.079	0.000
4	1	5 + 10	−0.69	0.091	0.091
4	2	5	−0.47	0.070	0.086
4	2	10	−0.48	0.058	0.140
4	2	5 + 10	−0.47	0.046	0.046
4	3	5	−0.46	0.055	0.255
4	3	10	−0.42	0.060	0.049
4	3	5 + 10	−0.44	0.040	0.150
8	1	5	−1.52	0.158	0.000
8	1	10	−1.85	0.129	0.456
8	1	5 + 10	−1.69	0.105	0.209
8	2	5	−0.53	0.150	0.260
8	2	10	−0.82	0.129	0.248
8	2	5 + 10	−0.64	0.103	0.202
8	3	5	−0.70	0.122	0.000
8	3	10	−0.88	0.103	0.178
8	3	5 + 10	−0.79	0.081	0.000
18	1	5	−4.33	0.408	0.000
18	1	10	−5.45	0.209	0.713
18	1	5 + 10	−4.90	0.248	0.256
18	2	5	−1.58	0.120	0.759
18	2	10	−1.35	0.219	0.000
18	2	5 + 10	−1.47	0.121	0.403
18	3	5	−1.46	0.269	0.000
18	3	10	−1.71	0.250	0.218
18	3	5 + 10	−1.58	0.182	0.077

Table 4. anova results, specifying the effects of substrate (grown on sand layers or in soil and corrected for non-retrieved fine roots), temperature (4, 8, or 18°C) and root length (5, 10 cm) on the daily biomass loss rate of Sonchus arvensis plants

Dependent variable	Experimental factor	DF	F-value	P > F
Type 3 tests of fixed effects. DF, degrees of freedom; F-value, treatment mean square/error mean square; P > F, significance probability value.
Biomass loss rate (k)	Substrate	1	0.17	0.695
	Temperature	2	115.20	<0.001
	Root Length	1	1.66	0.239

Using the data from the sand layer experiment and calculating over the entire temperature range (4–18°C), a value of 2.274 was obtained for Q₁₀. For the range from 8°C to 18°C, the Q₁₀ value was 2.346, while the Q₁₀ value was 2.102 for the range from 4°C to 8°C.

Application to elevated soil temperatures

The average daily soil temperature at 5 cm depth (T₅) during the six actual overwintering seasons was 1.34°C, and at a 10 cm depth, the average daily soil temperature (T₁₀) was only 0.09°C higher than at 5 cm. Temperatures below 0°C were only occasionally measured in the soil. While the increased temperature, when applying the elevated soil temperature projections, caused higher daily biomass loss rates, the length of the active below-ground period of S. arvensis prior to shoot emergence decreased. This is exemplified in Fig. 3 for a temperature rise of 3°C, during the seasons 2004–2007. Further quantification of the length of the overwintering seasons showed that the length of the active below-ground period of S. arvensis prior to shoot emergence was decreased by about 10 days for a rise of 1°C (Fig. 4), under moderately (0–3°C) elevated temperature projections.

Figure 3.  Determination of winter season start by a soil temperature <5°C (- - - -), at actual soil temperatures during the seasons 2004–2007 (…….) and under a projected soil temperature increment of 3°C (^_______).

Figure 4.  Average length of the active below-ground period (ABG) (days) of Sonchus arvensis under elevated temperature scenarios (0–4°C). The 95% confidence limits are shown.

Using a Q₁₀ value of 2.274, k_Base-value of 0.0147 and a T_Base of 18°C to calculate daily loss rates, we found that cumulative weight loss in below-ground parts of overwintering S. arvensis during six actual seasons at Ultuna on average was 44% (Table 5), with a small variation between seasons. Average estimated losses under elevated soil temperatures were around 44%, and did not change significantly, except for the highest soil temperature rise, which would lead to a decrease in weight loss overwinter. However, the variation around the mean seasonal losses increased constantly with elevated temperatures. When using the alternative values for Q₁₀, k_Base and T_Base, all calculated seasonal mean values differed by <1% and displayed the same variation in mean seasonal losses, compared with the first parameterisation.

Table 5. Cumulative weight losses (%) in below-ground parts of overwintering Sonchus arvensis during six seasons and under simulated climate conditions with daily soil temperatures 1–4°C above those recorded during the seasons indicated. Daily k was calculated as , using a Q₁₀ value of 2.274, a k_Base-value of 0.0147 and a T_Base of 18°C. Mean values and their standard deviations for six seasons at each of the temperature rise projections are given

Season	Actual temp. (T)	T + 1°C	T + 2°C	T + 3°C	T + 4°C
2004–2005	42.7	44.2	46.7	41.0	35.5
2005–2006	43.9	45.6	41.3	41.5	27.4
2006–2007	45.7	46.3	46.6	47.1	31.7
2007–2008	43.7	45.2	39.9	36.4	26.8
2008–2009	43.4	45.5	48.0	49.0	39.6
2009–2010	44.4	40.2	41.9	43.2	28.9
Mean	44.0	44.5	44.1	43.0	31.7
SD	1.02	2.21	3.42	4.53	5.04

Discussion

Dependency of weight loss rates to temperature and substrate

When testing the hypothesis that the relative weight loss rate of below-ground parts of S. arvensis is temperature dependent, we found an increase in weight loss rates with temperature in both experiments. Furthermore, the obtained value Q₁₀-value falls well within the ranges found in the literature on respiration; the Q₁₀-value often is modelled to be 2.0 (Lambers et al., 2008). Landsberg (1986) found values for Q₁₀ of 2.0–2.3 to be reasonable for modelling purposes. Biomass loss in overwintering below-ground structures prior to emergence also may be attributed to the activity of pathogens (microbes and fungi). Apart from some discolorations in shoot tips, we did not find any signs of pathogen-mediated decay or biomass loss in our experiments. Consequently, we suggest that the detected weight losses might mainly be attributed to maintenance respiration and respiration associated with allocation of assimilates to fine roots and shoots.

When testing the second hypothesis that biomass loss rates for given temperatures are independent of the substrate, we found larger losses for plants grown in soil (bucket experiment), compared with plants grown on sand layers (pot experiment). However, realising that biomass loss in the bucket experiment also was caused by the non-retrieved fine roots, we recalculated respiratory losses in the bucket experiment, using the assumption that the allocation proportion to fine roots for a given temperature was equal in both experiments. This assumption was supported by the visually encountered amount of fine roots in the buckets in comparison with the amount of roots on the sand layers. We found that the losses attributed to respiration did not differ between the two substrates. This suggests that the respiration, concurring with allocation of biomass to shoots merely is temperature driven and not determined by the degree of etiolation of structures developing in the dark.

Cumulative respiratory losses over the winter season

We hypothesised that warmer winters will lead to an increase in cumulative weight loss of below-ground parts of S. arvensis, as higher temperatures are known to increase respiration. However, when applying the encountered weight loss rates and Q₁₀ to elevated soil temperature projections, it was shown that cumulative losses during winter seasons in central Sweden will remain basically constant, the effect of increased weight loss at higher temperatures being balanced by shorter winter seasons.

To anticipate weight loss in below-ground parts of S. arvensis under future climate change, we did not perform an impact assessment of possible climate scenarios on soil temperatures and length of concurrent vegetation periods, nor did we make a statement on the likelihood of the occurrence of detailed specific climate scenarios, which is outside the scope of this paper. Our calculations only predict the amount of biomass loss of overwintering S. arvensis for a combination of soil temperatures and lengths of the overwintering period.

Several assumptions were made when making the calculations that lead to a resulting cumulative weight loss of >40% of the below-ground biomass during overwintering periods: (i) dormancy, which may cause a decreased respiration (Van Der Schoot & Rinne, 2011), did not occur in our experiments and (ii) does not occur during the periods over which we calculated cumulative biomass loss. Brandsæter et al. (2010) showed that the reduction in sprouting capacity is a quantitative phenomenon, which gradually disappears after exposure to lower temperatures. When harvesting the roots from the cold store, prior to planting, sprouting activity had commenced in our material. Therefore, we inferred that the calculated weight loss rates were not affected by a reduced sprouting capacity. For the cumulative weight loss calculations, we assumed that a temperature dependent release of reduced sprouting in S. arvensis was completed when T₅ < 5.0°C. For the six actual seasons, this occurred at the end of October or during early November. This also coincides with the findings of Brandsæter et al. (2010). Consequently, our calculations of the cumulative weight loss during the overwintering periods were not expected to be affected by dormancy. A further assumption was made when (iii) determining the end of the overwintering period of S. arvensis by setting shoot emergence time in spring at T₅ ≥ 4.0°C. For the six actual seasons, this occurred on average in the second week of April, which coincides with phenological observations on emergence made around Uppsala (Håkansson, 1982) and elsewhere (Lemna & Messersmith, 1990) under comparable conditions. A final assumption (iv) was made by choosing a soil temperature at 5 cm below ground as the basis for our calculations, because S. arvensis has the majority of its roots close to the soil surface (Lemna & Messersmith, 1990). Using the temperature at 10 cm depth, instead of T₅, would have influenced the results very little, as the average difference between T₅ and T₁₀ only was 0.09°C. The two different parameterisations of Eqn (5) to estimate daily k-values on basis of our experimental results showed that, in relation to the annual variations, the model neither is sensitive for the encountered variation in estimates of Q₁₀, nor for the choice of T_Base. While root respiration in plants, as a percentage of acquired carbon, is known to be substantial, ranging from 10 to 50% according to Lambers et al. (2008) and from 8 to 52% according to Jing et al. (2007), our cumulative value for weight loss in S. arvensis during current winters (>40%) is higher than the one found by Tørresen et al. (2010), who reported a reduction in below-ground biomass in S. arvensis (of about 30%) between late autumn and spring. The lower value found in their investigation might have been caused by an exposure of their roots to lower temperatures, as their experimental units were placed on the soil surface and could have been subjected to low air temperatures. Graphs provided by Håkansson (1969) display a dry matter loss of >30% from November to April, but also that a photosynthetic gain at the plant level is counterbalancing continuing weight losses in underground parts during late April and May. This may explain why we attained larger cumulative losses, because our experiments were performed entirely in the dark.

Implications of life cycle dynamics for future weed control

Weight losses of the encountered order of magnitude in below-ground S. arvensis during the overwintering season need to be accounted for in life cycle models of this perennial weed. Given an average weight loss of 30 and 45%, respectively, during winter, weight increments of 43 and 82%, respectively, would be required during each vegetation season, for a population of S. arvensis to maintain its weight over the seasons. If future winter seasons develop according to the assumptions that soil temperatures will rise, and concurrent winter seasons will become shorter, weight loss rates in below-ground part of S. arvensis will increase, while cumulative weight loss is counterbalanced by shortening of the winter seasons. A significant climate warming already is occurring in Sweden (Swedish Meteorological and Hydrological Institute, 2006), and evaluation of existing climate scenarios (Intergovernmental Panel on Climate Change, 2007) showed that a rise in air temperature by 4°C might be attained by the end of this century for the southern and middle part of Sweden. An increase in cropping season length will be due to temperature rise in spring, rather than during the autumn. This implies that S. arvensis would emerge earlier in the season and that release of decreased sprouting capacity by low temperatures is expected to occur even later in the autumn, thereby counteracting the effects of cultivation methods which promote depletion of roots during autumn. Spring cultivation at the developmental stage of 5–7 leaves (the stage where S. arvensis is considered to be most susceptible for soil cultivation) currently is the most commonly recommended weed control method (Håkansson, 1982). Any future temperature rise will provide an extended time period to combat this perennial weed prior to sowing crops. While average future winter conditions probably will not impact on the magnitude of biomass loss in overwintering S. arvensis, our calculations suggest that the variance in biomass loss during future warmer and shorter winters might increase. To combat S. arvensis efficiently, this would require an adaptive control strategy that can respond to such increased variability between years.

Acknowledgements

The project was funded by SLU (Swedish University of Agricultural Sciences), Uppsala, Sweden, and by SLU EkoForsk, Uppsala, Sweden. Two reviewers are acknowledged for their constructive comments.