The infl uence of Ambrosia trifi da on vegetative production of A. artemisiifolia

number of A. artemisiifolia plants and decrease in A. trifi da counts per m 2 caused an increase in the dry mass of A. artemisiifolia per plant. The dry mass of A. artemisiifolia ranged from 4.22 to 6.11 g/plant (July), 8.96 to 10.27 g/plant (August) and 7.04 to 19.53 g/plant (September). In the following season, these values ranged from 9.62 to 14.60 g/plant, 14.37 to 28.90 g/plant, and 23.43 to 40.47 g/plant in July, August and Sepember, respectively. Minimum values of vegetative parameters were recorded in the treatment with 2 plants, and maximum in the treatment with 10 A. artemisiifolia plants/m 2 . This means that interspeciﬁ c competition is more pronounced in this ragweed species than intraspeciﬁ c competition.


INTRODUCTION
Invasive species are considered to be one of the main factors contributing to global change nowadays, making them the prime focus of various studies in the fi elds of ecology and related sciences (Fumanal et al. 2007;Pyšek et al., 2009;Essl et al. 2009). When they grow in the same habitat, invasive weeds compete for resources and the more competitive plants suppress other species, which leads to a reduction in the number of plants of the endangered population. Invasive species are more prone to suppress other plants, resulting in changes in biodiversity and plant community structure (Shine et al., 2010). Th eir expansion has a detrimental eff ect both on agricultural production and biodiversity of a given area, and many of them also have a negative impact on human health due to allergens that they produce (Šikoparija et al., 2009(Šikoparija et al., Gerber et al., 2011. Both Ambrosia artemisiifolia and A. trifi da belong to a group of alien, harmful and invasive weed species (Kőmíves et al., 2006;Webster et al., 1994).
A. artemisiifolia (common ragweed) is native to the North American continent, whence it was introduced to Asia, Australia and Europe (Chun et al., 2010;Cunze et al., 2013). Common ragweed belongs to a group of invasive weed species harmful to the economies of many parts of Europe, including the Balkans (Kőmíves et al., 2006). Aside from being a very important weed, it is also listed as an invasive species in Serbia, where it is oft en present in rural areas (row crops, orchards and vineyards, alfalfa crops, abandoned fi elds, fi eld edges and ruderal land), as well as urban areas. At high density, it acts as a strong competitor (for light, water, nutrients and space) against other plant species, and is able to cause huge yield losses of many crops (Kőmíves et al., 2006). A. artemisiifolia can withstand strong competition, which allows it to survive in environments that are less favourable for growth, and to achieve high spread rates (Raynal & Bazzaz, 1975). According to Weber and Gut (2005), only a few species (Senecio inaequidens, Erigeron canadensis and Reynoutria japonica) in Europe have higher spread rates than common ragweed. Th e fact that it is an extremely vigorous grower is also indicated by data on its abundance per unit area, i.e. some authors state the number of over 1000 plants/m 2 and even up to 4000 plants/m 2 (Mataruga, 2004).
Unlike A. artemisiifolia, A. trifi da (giant ragweed) is currently present in Serbia only locally, along roadsides in rural areas and between settlements in the Central Bačka region of Vojvodina Province. It is also commonly found on fi eld edges, as well as in sunfl ower, maize, soybean and sugar beet fi elds. A. trifi da forms a large vegetative mass and belongs to the group of strong competitors for both natural resources (light, water and nutrients) and space. Rapid growth and development, as well as pronounced plasticity allow it to develop and occupy large surface areas, where it absorbs water and nutrients from the soil and shades out other plant species, which enables it to make better use of solar radiation (Webster et al., 1994).
Considering that A. trifi da is becoming naturalized in Serbia (Vrbničanin et al., 2004;Vrbničanin, 2015), even though it is now only locally present in the area of Central Bačka (Kucura, Savino Selo, Despotovo, Ruski Krstur, Ravno Selo, etc.) (Vrbničanin et al., 2004;Malidža & Vrbničanin, 2006;Vrbničanin et al., 2015), this research was performed to study its invasive potential and its capacity to suppress A. artemisiifolia when it is associated with it. Looking at the vegetative and generative potential of A. trifi da, it is highly likely that this ragweed species will spread in the future. Th e focus of this research was to study the inter-and intraspecifi c infl uence of A. artemisiifolia and A. trifi da when they coexist in the same habitat. Th e goal was to examine the vegetative production of A. artemisiifolia when growing in association with A. trifi da.

MATERIALS AND METHODS
Th e experiment was performed in April 2016 and 2017 on a piece of land in the village of Dobrić (Šabac) (44 ° 41'N 19 ° 34'E) where nothing had been sown in the previous two years. Since the presence of A. trifi da was not detected in the area of Western Serbia, seeds of this plant were obtained from a location in Centralna Bačka region (45°30'N19°31'E). Aft er collection in the previous season, the seeds were stored under controlled conditions (in a refrigerator at t = 4 °C). Considering that A. artemisiifolia was already present in the test area in high numbers (>100 plants/m2), it was not necessary to sow it. Up to 100 seeds of A. trifi da were sown depending on treatment.
Th e experiment was set up on the principle of completely randomized block design with four replications, using the model of replacement series ("Replacement design") (Kropff & van Laar, 1993). A total of fi ve treatments were conducted, i.e. fi ve diff erent density ratios of A. trifi da/A. artemisiifolia: 2/8, 4/6, 6/4, 8/2 and 10/0 (total number of plants was 10/m2). Th e given plant density was maintained by thinning every 7 to 10 days over the season. When establishing any given number of A. trifi da and A. artemisiifolia, care was taken to ensure that their distribution per unit area was uniform (that the plants of both ragweeds were at an equal distance from each other) and that samples of both species were in the same stage of development.
Plot size was 6 m 2 , and each plot was divided into six sub-plots of 1 m 2 . Vegetative parameters, i.e. plant height (cm) and dry mass (g) of А. Artemisiifolia, were measured on four sub-plots during the vegetative season three times (in July, August and September). Plant height was measured on experimental plots with a wooden pull-out meter, while dry mass was measured on precise scales in the laboratory.
Th e results of texture and chemical analysis of powdered clay soil regarding the basic agrochemical properties [pH in KCl and H 2 O, P 2 O 5 (mg/100 g), K 2 O (mg/100 g) and humus (%)] are presented in Table 1.

Statistical analysis
To calculate the eff ects of treatments (diff erent plant ratios) on the measured vegetative parameters of A. artemisiifolia, the analysis of variance (ANOVA) was used in the statistical program SPSS 23, and diff erences between means were tested using the LSD test at the levels of signifi cance of: * p<0.05, ** p<0.01 and ns p>0.05.

RESULTS
In general, the values of vegetative parameters were higher in 2017 due to more favourable meteorological conditions, primarily the amount and distribution of precipitation ( Figure 1). Vegetative production of A. artemisiifolia plants, growing under diff erent ratios of two Ambrosia species during July, August and September, had a similar trend in both years. With increasing number of A. artemisiifolia and decrease in A. trifi da/m 2 , the height and dry weight of A. artemisiifolia increased.    (Table 3).  Figure  3). Statistically signifi cant diff erences (P≤0.05) were confi rmed between treatments with 10 A. artemiisifolia plants/m 2 compared to 2, 4, 6 and 8 A. artemisiifolia plants/m 2 ; treatments with 2 compared to 4 and 6 plants, and also between treatments with 8 and 6 A. artemisiifolia plants/m 2 ( Table 2).

DISCUSSION
In general, a large number of factors (biotic and abiotic) can aff ect the behaviour of weeds, their appearance, abundance and distribution, and thus the manifestation of inter-and intraspecifi c competition (Gibson et al., 2017;Adeux et al., 2019). Also, competition largely depends on meteorological conditions (temperature, precipitation and insolation). Th e results of this research showed that meteorological conditions greatly infl uenced the vegetative production of ragweed plants. Diff erences between the two years were more pronounced in terms of precipitation amounts and distribution, i.e. 2016 was drier and 2017 richer in precipitation (Figure 1). Th is statement is in accordance with results of previous studies. Nelson et al. (2006) emphasized the importance of optimal conditions for germination, growth and development of plants, and their fruiting, for their competitive potential. Consequently, a noticeable precipitation defi cit during 2016 caused lower vegetative production of common ragweed plants at all A. trifi da/A. artemisiifolia density ratios. Most oft en, weed species in competitive relationships undergo various changes under the conditions in which they grow, depending, among other factors, on the density of plants per unit area (Adeux et al., 2019). According to many authors, vegetative production is a very important indicator of competition between plants (Tollenaar et al., 1994;Patterson, 1995). Also, the ability of species to compete for resources is infl uenced by their very morphological characteristics. Among the most indicative morphological features in that sense are plant height and weight (Bertholdsson, 2005).
According to some authors, plant height is a parameter that largely contributes to suppressing other plants (Cosser et al., 1997;Piliksere et al., 2013). Mason et al. (2007) asserted that the interaction between plant height and other species characteristics could contribute to more effi cient use of natural resources, so it oft en happens that plants growing at higher density can have better growth and higher yield. Also, the height of plants greatly aff ects the extent to which they can use solar energy. For example, under conditions of moderate soil moisture and nutrient content, competition for light between Abuthilon theoprhrasti and soybean is one of the most important factors that aff ect plant height (Lindquist & Mortensen, 1999). Accordingly, A. theoprhrasti makes better use of light energy, which provides it with a competitive advantage over soybean crop (Akey et al., 1990). According to Irwin and Aarssen (1996), the height and branching of a shoot in common ragweed can oft en vary depending on habitat conditions in which it grows, and especially on humidity (Bollinger et al., 1991). In Europe, the average height of A. artemisiifolia plants ranges from 1.5 to 2 m (Šilc, 2002), while in Serbia the height ranges from 20 to 150 cm, and sometimes up to 2 m (Vrbničanin et al., 2007). In general, the results of this study indicate that variable ratios of A. trifi da/A. artemisiifolia aff ected the height of A. artemisiifolia plants.
Th e height of A. artemisiifolia plants ranged from 35.00 to 99.80 cm (from July to September) in 2016, and the growth trend remained the same in the following season, except that plants generally increased growth due to more favourable weather conditions (from 56.19 up to 148.50 cm) (Figure 2). However, even though the raised number of A. artemisiifolia plants in association with A. trifi da increased the height of A. artemisiifolia plants, this cannot be attributed to competition for light because the total number of plants per unit area was constant (A. trifi da + A. artemisiifolia = 10 plants/m 2 ). Th us, the reason for the increased height of A. artemisiifolia plants that goes with the increase in its ratio in association with A. trifi da, can be connected to absent intraspecifi c competition, i.e. when the number of A. trifi da increases, the number of A. artemisiifolia/m 2 (with the same total number) drops, and the interspecies competition comes to the fore. Similar to these results, Oljača et al. (2000) found that with an increase in the number of Datura stramonium from 1 to 10 plants per meter length (1, 3, 6, and 10) in the inter-row space of corn, the height of D. stramonium had the following upward trend: 116 cm, 117.7 cm, 119.7 cm and 141.7 cm.
Although competition is a process that can manifest in several ways, it is most common for a more competitive species to act on a less competitive one by eff ecting its lower mass production (Bussan et al., 1997). Based on height and dry weight, stronger interspecies competition was confi rmed in A. artemisiifolia in our experiment (Figure 3). In 2016, maximum dry weight/plant of A. artemisiifolia was 19.53 g (10 A. artemisiifolia/m 2 ). Dry mass/plant in 2017 in all treatments was higher compared to the previous season, which was characterized as drier of the two years ( Figure 1). Th e highest dry mass of A. artemisiifolia was recorded in monoculture (40.47 g/plant), and the lowest in the treatment with 2 A. artemisiifolia/m 2 (23.40 g/plant) (Figure 3). Th is fact can partly explain the success of common ragweed's invasiveness, as it survives even when there is a great physical pressure within the population, leaving abundant off spring, which is a prerequisite for its further spreading and colonization of new habitats. A. artemisiifolia is a species that has increased plasticity and adapts better to stressful conditions such as nutrient, water and physical space defi cits (Leskovšek et al., 2012). However, there is also an opposite view to be found in literature. Vidotto et al. (2007) found that with the number of A. artemisiifolia increasing from 4 to 25 plants/m 2 , the dry weight of these plants fell from 687 to 140 g/plant (4.9 x), which was associated with nutrient defi ciency. In contrast, and similar to our results, Brassica napus produced a higher mass/plant in monoculture compared to the situation when it grew alongside Malva paravifl ora at diff erent ratios, namely: 25/75%, 50/50% and 75/25%. In a treatment with B. napus/M. paravifl ora 25/75% abundance ratio, the dry mass of B. napus was reduced by 68%, compared to monoculture (Bakhtiari & Saeedipoor, 2014).

CONCLUSION
Th e results of this experiment showed that the highest average height of A. artemisiifolia plants was in the treatment with 10 A. artemisiifolia plants/m 2 , and the lowest with 2 A. artemisiifolia plants /m 2 (8/2 A. trifi da/A. artemisiifolia). Also, an increase in the number of A. artemisiifolia plants/m 2 , versus A. trifi da, caused an increase in the dry weight of A. artemisiifolia/plant.
In general, the lowest values of vegetative parameters were found in treatments where the number of A. artemisiifolia plants was the lowest, compared to A. trifi da. With increasing number of plants per area unit, vegetative parameters increased, and the maximum value was recorded in treatments with the highest number of A. artemisiifolia (10 plants/m 2 ). It infers that interspecifi c competition is more pronounced than intraspecifi c competition in A. artemisiifolia. Increasing numbers of A. artemisiifolia plants per area unit did not suppress their growth, and this facilitates its successful invasion of new habitats, where it acts as a strong competitor and a great potential danger to crops. Considering all of the above, maximum responsibility of experts is needed, along with education of agricultural producers about the danger of this invasive weed species, the damage it can do to crops, as well as all available control methods, emphasising prevention.