Conspecific and heterospecific grass litter effects on seedling emergence and growth in ragwort (Jacobaea vulgaris)

Jacobaea vulgaris Gaertn. or common ragwort is a widespread noxious grassland weed that is subject to different regulation measures worldwide. Seedling emergence and growth are the most crucial stages for most plants during their life cycle. Therefore, heterospecific grass or conspecific ragwort litter as well as soil-mediated effects may be of relevance for ragwort control. Our study examines the effects of conspecific and heterospecific litter as well as ragwort conditioned soil on seedling emergence and growth. We conducted pot experiments to estimate the influence of soil conditioning (with, without ragwort), litter type (grass, ragwort, grass-ragwort-mix) and amount (200 g/m², 400 g/m²) on J. vulgaris recruitment. As response parameters, we assessed seedling number, biomass, height and number of seedling leaves. We found that 200 g/m² grass litter led to higher seedling numbers, while litter composed of J. vulgaris reduced seedling emergence. Litter amounts of 400 g/m² had negative effects on the number of seedlings regardless of the litter type. Results for biomass, plant height and leaf number showed opposing patterns to seedling numbers. Seedlings in pots treated with high litter amounts and seedlings in ragwort litter became heavier, grew higher and had more leaves. Significant effects of the soil conditioned by ragwort on seedling emergence and growth were negligible. The study confirms that the amount and composition of litter strongly affect seedling emergence and growth of J. vulgaris. Moreover, while conspecific litter and high litter amounts negatively affected early seedling development in ragwort, those seedlings that survived accumulated more biomass and got taller than seedlings grown in heterospecific or less dense litter. Therefore, ragwort litter has negative effects in ragwort germination, but positive effects in ragwort growth. Thus, leaving ragwort litter on pastures will not reduce ragwort establishment and growth and cannot be used as management tool.


examples.
This statement is required for submission and will appear in the published article if the submission is accepted. Please make sure it is accurate.
Unfunded studies Enter: The author(s) received no specific funding for this work.   conspecific litter as well as soil-mediated effects may be of relevance for ragwort control. Our 27 study examines the effects of conspecific and heterospecific litter as well as ragwort 28 conditioned soil on seedling emergence and establishment. 29 We conducted pot experiments to estimate the influence of soil conditioning (with, without 30 ragwort), litter type (grass, ragwort, grass-ragwort-mix) and amount (2 g/dm², 4 g/dm²) on 31 J. vulgaris recruitment. As response parameters, we assessed seedling number, biomass, height 32 and number of seedling leaves. 33 We found that, 2 g/dm² grass litter led to higher seedling numbers, while litter composed of 34 J. vulgaris reduced seedling emergence. Litter amounts of 4 g/dm² had negative effects on the 35 number of seedlings, regardless of the litter type. Significant effects of the soil conditioned by 36 ragwort on seedling emergence were negligible. Results for biomass, plant height and leaf 37 number showed opposing patterns to seedling numbers. Seedlings in pots treated with high litter 38 amounts and seedlings in ragwort litter became heavier, grew higher and had more leaves. 39 The study confirms that the amount and composition of litter are strongly affecting seedling several studies aimed at trying to find effective and efficient ways how to manage this grassland 53 weed (2-5). Some of these studies come to the conclusion that specific cutting regimes are a 54 promising approach (6). However, the toxicity of ragwort biomass precludes the use as hay or 55 silage and a high number of long living seeds rules out composting the biomass. Consequently, 56 the biomass is often disposed as special waste, which is costly (7). Yet, the occurrence of 57 conspecific allelopathic effects (8, 9) suggests that leaving ragwort litter on the mown grassland 58 could be an additional component in managing the abundance of ragwort. As a first step to 59 evaluate the validity of this approach, we experimentally tested this management option.

60
Litter influences micro-environmental conditions, is important for nutrient cycling and can have 61 prominent effects on seedling establishment and germination via mechanical and chemical 62 effects (10, 11). Litter is affecting the physical environment by lowering light availability, 63 changing the water and temperature regime and can act as physical barrier (11). Litter may also 64 exert chemical effects by changing the availability of nutrients or releasing phytotoxins (9, 12, 65 13). Through these mechanisms litter potentially influences community organization via 66 delaying and reducing or facilitating seedling emergence or seedling establishment (11, 14).

67
Albeit a meta-study on litter effects showed an overall negative effect of litter on seedlings (15), 68 another meta-analysis demonstrated that litter can be also beneficial for seedling establishment 69 especially under dry conditions (e.g. dry grasslands or dry periods) or at low to medium litter 70 amounts (11). Positive effects were due to higher soil humidity and a lower amplitude of diurnal Cardon (19) found that more than half of the litter mixtures they reviewed behaved in a non-81 additive way concerning factors like mass loss, nutrient transfer or decomposer composition.

82
It is intuitive to hypothesize that conspecific litter is facilitating the establishment of conspecific 83 seeds. Yet, as intraspecific competition is assumed to be more intense, than interspecific 84 competition (20) also inhibiting effects may be expected. While many studies have shown 85 facilitating effects from adults on conspecific seedlings (21-23), particularly in weed species 86 (24-26), there are also studies demonstrating the opposite (27, 28). Some did not find evidence 87 of unidirectional conspecific facilitation (29). In line with the latter, studies on ragwort showed 88 autotoxic potential of ragwort litter and soil from ragwort stands (30; 9). Also, a negative plant-89 soil-feedback has been demonstrated for ragwort, which is probably driven by differences in 90 the soil community (31, 32). Under these circumstances, negative effects of ragwort litter and 91 ragwort conditioned soil may either be additive, i.e. result in stronger inhibitory effects than 92 they would do individually, or the combined effect is not additive, i.e. combined effects are 93 lower or higher than the sum of both effect sizes. However, the influence of litter during 94 seedling emergence might be weaker than during seedling establishment, since at the latter stage 95 impacts of competition are of higher importance than direct litter effects (33). Thus, if litter 96 reduces seedling density through self-thinning surviving seedlings might be in advantage later 97 on, as intraspecific competition is reduced.

98
The present study is to our knowledge the first to test the influence of the combined effects of 99 litter type, amount and conditioned soil on seedling emergence and establishment in ragwort.

100
To this end, we set up a common garden experiment, where we assessed the conspecific and 101 heterospecific effects of litter on the number of emerged ragwort seedlings, their biomass, 102 height, leaf number and specific leaf area. We posed the following questions: Does litter type influence the emergence and establishment of ragwort seedlings?

104
Assuming conspecific negative effects, we expect lower numbers of ragwort seedlings 105 to emerge from beneath conspecific litter compared to pure grass litter or a mixture of 106 grass and ragwort litter. We expect congruent patterns to occur in seedling establishment 107 after emergence.

II.
Does litter amount influence seedling emergence and establishment of ragwort and does 109 this depend on litter type? 110 We expect negative litter effects to increase with a higher litter cover and to be stronger 111 with increasing litter amount. We also expect the size of litter effects to increase with 112 the portion of ragwort litter in the litter cover. Thus, we expect an interaction of litter 113 type and amount in that seedling's emergence and establishment is supposed to be 114 lowest in high amounts of ragwort litter, higher in medium ragwort amounts and even 115 higher in low amounts of grass litter.

III.
Does ragwort-conditioned soil influence conspecific seedling emergence and 117 establishment? If so, do these effects interact with the litter effects? 118 We hypothesized an inhibitory effect of ragwort-conditioned soil on its own emergence 119 and establishment as negative plant-soil feedback for this species has been observed. 120 We expect these effects to be additive for ragwort litter. Experimental design 136 We studied the effects of litter amount (k = 2; 2 (low) vs. 4 (high) g/dm 2 ), litter type (k = 3; 137 ragwort litter, grass litter, mix of ragwort and grass litter) and soil conditioned (k = 2; soil 138 conditioned with ragwort vs. unconditioned soil) on seedling emergence and establishment.

139
High litter densities refer to average regional productivity of grasslands in Northern Germany

140
(37) and thus represent average litter quantities that ragwort seeds are exposed to at heavily 141 ragwort-infested sites. While litter type and litter amount both affected seedling emergence significantly the amount 210 of explained variance was more than twice as high for litter type than litter amount (37.88% vs. 211 17.01%; Table 1). Seedling number differed significantly between all three litter types but only 212 grass and ragwort litter effects differed significantly from the control (95% CI does not include 213 zero; Fig 1). In grass litter seedling emergence was one third higher (28 ± 1; mean ± SE) than 214 in ragwort (18 ± 1). Interestingly, ragwort litter had a negative effect on seedling number across 215 both litter amounts, while the opposite was true in case of pure grass litter. Across different 216 litter types an increased litter cover, i.e. 4 g/dm², led to a significantly reduced seedling number, 217 both in comparison to the control and the low litter cover of 2 g/dm² (Fig 1). Effects of litter type and amount on biomass per plant were also significant but diametric to the 231 results for seedling emergence (Fig 2). More than half of the variance within biomass per plant 232 was explained by litter type, whereas litter amount only explained around one fifth of the 233 variance ( Table 1). The interaction between litter type and amount was also significant, but  were considered significant (i.e. different from the controls) when 95% CI did not overlap with 249 zero.

250
Seedling height was significantly influenced by litter type and amount (Fig 3). The pattern is 251 the same as for biomass as such that all litter types differed and seedlings in mixed and ragwort 252 litter grew highest. More than half of the variance was explained by litter type (  Mean leaf number was significantly affected by litter type and amount (Fig 4). The explained 266 variance for litter amount was twice as high as in litter type. Thus, leaf number is the only 267 measured variable that seems to be stronger influenced from litter amount than litter type. Leaf Our experimental results suggest that litter type as well as litter amount do influence seedling 281 emergence and establishment. It is interesting to note that although ragwort litter reduced the 282 number of conspecific emerging seeds, seedlings were actually performing better in ragwort 283 litter, i.e. successful ragwort seedlings developed higher biomass, height and leaf numbers.

284
Litter type 285 While both, litter type and litter amount, influenced seedling emergence significantly, litter type 286 accounted for a much higher portion of the variance than litter amount. In contrast, other studies 287 reported that litter amount was more relevant than litter type (43, 44). This highlights the species 288 specific nature of litter effects (13) and is a hint that ragwort litter is indeed autotoxic and 289 negatively influences emergence of conspecific seedlings. litter (high amount: grass 90 ± 3% (mean ± SE), mixed 85 ± 4% and ragwort 67 ± 8%, low 295 amount: grass: 85 ± 3%, mixed 70 ± 9%, ragwort 37 ± 5%). In accordance to the study of 296 Eckstein and Donath (45), which revealed a rise in soil moisture when litter cover increased, 297 we assume that seedlings experienced more favourable water availability in the grass litter. As ragwort produces many seeds with high germination potential and low dispersal ability (36, 317 1, 35) autotoxic effects may avoid intraspecific competition in ragwort and thus may lead to 318 lower seedling numbers but increased seedling vitality.

319
Mixed litter resulted in fewer emerged seedlings than in grass litter, but more emerged seedlings 320 compared to ragwort litter, which suggests that effects of grass and ragwort litter simply added 321 up to a neutral effect. This is supporting the mass-ratio hypothesis by Grime (18) Another explanation for superior performance of emerged seedlings in ragwort litter might be 348 an improved availability of nutrients (10), as ragwort litter decomposed quite fast. This is 349 supported by our observation that even though the number of seedlings in the control pot and 350 in the ragwort litter pot with 2 g/dm² was about the same and thus competition seedlings 351 experienced was comparable, biomass of seedlings growing in ragwort litter pots was still 352 higher.

353
Regarding leaf number, only seedlings in ragwort litter had significantly more leaves than plants 354 in grass or mixed litter. This again might be linked to lower intraspecific competition as there 355 were lower seedling numbers in the ragwort pots or by less physical hampering compared to 356 grass litter (51).

357
Litter amount 358 We can affirm that ragwort emergence is influenced by litter amount, albeit less strongly than 359 by litter type. Seedling emergence was lowest from beneath high litter amounts and higher from 360 beneath low litter amounts, although not significantly different from the control. Positive effects 361 could be due to physical effects such as constant moisture and temperature regime discussed 362 above. Still, when litter cover is too high negative effects prevail over positive effects.

363
Accordingly, compared to the control treatment, seedling emergence was lowest from beneath 364 4 g of litter and not reduced from beneath 2 g of litter cover. In case of ragwort, the reduced 365 seedling emergence from beneath 4 g of litter is very likely linked to the relatively small seed 366 mass of ragwort (0.22 mg; (52)). The smaller the seeds are, the less resources a seed can spent 367 to successfully penetrate through a litter cover (53). In addition, small seeded species, like 368 ragwort, tend to depend on higher light levels for germination than large seeded species (54, 369 55). Since photosynthetic active radiation (PAR) decreases with increasing litter cover (45) not 370 only germination will be lower, but seedlings germinated will also lack light resources for 371 successful penetration of the litter cover. This is also in accordance with van der Meijden and 372 van der Waals-Kooi (46), who found that emergence in ragwort significantly decreased when 373 seeds were placed beneath a sand cover of more than 4 mm. Furthermore, Facelli et al. (56) 374 showed that ruderal species, such as ragwort, are potentially negatively impacted by high 375 amounts of litter. Therefore, ragwort seedlings might be reduced by a mulching regime on 376 productive sites were litter amounts are medium to high.

378
The effects of litter amount on the establishment of seedlings was quite the opposite of the 379 effects on seedling emergence. Thus, seedlings in dense litter got heavier (at least in mixed and 380 ragwort litter), grew higher, and tended to have more leaves than seedlings in lower amounts stage as seedlings in our experiment and found that plants tend to grow higher, when there is 386 more shade. Thus, seedlings may grow higher in higher litter amounts to avoid its shade and in 387 order to reach light. Another explanation for heavier and higher seedlings with more leaves 388 could be higher nutrient content related to higher amounts of litter (33).

390
Conditioned vs. unconditioned soil 391 We expected potentially most inhibitory effects of ragwort litter on itself on sites that are 392 infested with ragwort for a longer time. However, effects on soil on seedling emergence and 393 establishment were very weak (Table 1). In fact, the only significant effect of conditioned soil 394 occurred on seedling height, i.e. seedlings in unconditioned soil grew bigger than in conditioned 395 soil. In general, this finding is in line with findings of van de Voorde (30), who also reported 396 autotoxic effects of ragwort on its seedling development. However, on average the observed 397 difference in plant height between ragwort seedlings growing in conditioned versus those 398 growing in unconditioned soil was only 2 mm (less than 4% of the average height of about 399 54.5 mm). Therefore, we expect that this difference is too small to be of ecological relevance.

400
In other studies, looking at several seedling cohorts, those cohorts usually differed several 401 centimeters in height (58-60). physical ragwort litter effects in the stage of seedling emergence, those seedlings that emerged 407 successfully may have profited from a lower intraspecific competition and may had more 408 resources to invest than seedlings that emerged from beneath grass or mixed litter or they 409 profited from higher nutrient amounts in ragwort litter. Due to the ambivalent nature of our 410 results, these findings cannot easily be translated into management recommendations for 411 ragwort-infested sites. However, the effects of mixed litter on ragwort recruitment did not differ 412 from the control treatment. Therefore, we expect that leaving litter containing ragwort biomass 413 on the grassland will have no regulating influence on recruitment of ragwort seedlings.