Introduction

Densities of wild ungulates have been steadily increasing in the central Europe and North America and they have been causing extensive damages in agriculture and forests. Critical situation with forest regeneration lead also to intensive study of feeding behaviour of wild herbivores in forested areas (Homolka 1995; Cornelis et al. 1999; Gebert and Verheyden-Tixier 2001; Suter et al. 2004), their negative impact on forest regeneration (Gill 1992; Putman and Moore 1998; Sage et al. 2004) and possibilities how to maintain acceptable intensity of browsing (Reimoser et al. 1999; Augustine and deCalesta 2003). Both researchers and game managers paid attention mostly to forests and the unacceptable forest damages lead to partial reduction of deer populations at the end of the 20th century in several forested areas of the Czech Republic (Mrkva 1996). Intensity of tree browsing has therefore been generally stable during several past years and much lower than before (also due to mild winters).

Contrary to slowly but surely decreasing game damages to forests, damages to field crops are increasing in the Czech Republic (Kamler et al. 2005). This negative trend is primarily caused by increasing numbers of wild boar but also by abstracting attention of game managers from game damages to field crops resulting in insufficient reduction of ungulates in areas with lower proportion of forests (Kamler et al. 2008). Also the permanent hunting pressure in forested areas resulted in extension of game species on agricultural land; besides, landowners and farmers have been tolerating game damages for a long time due to missing experience and hunters-friendly legislation in the Czech Republic (Kamler et al. 2005).

Farmers on localities with higher densities of ungulates presently have to face extensive damages to crops. The situation forced the authorities to search for solutions that would lead to reduction of damages caused by game and this activity revealed critical lack of information about the factors affecting game impact to field crops. Especially in Europe, game damages on field crops escape the notice of the researchers. Putman and Moore (1998) evaluated the economical impacts of damage inflicted by game on agricultural plants in the UK. They divided the damage by the crop and game species. Pastures and meadows were the most commonly damaged sites but the sustained damage was classified as of low significance. In contrast, in Central Europe, damages to grassland are of the highest significance due to wild boar rooting. Trdan and Vidrih (2008) analysed even the damages caused by grazing of grassland production and found out 50% reduction of dry matter yield on average.

The potential economical impact of herbivores on agriculture production is sizeable. Conover et al. (1995) estimated the total loss caused by game on the production of field crops reaching 274 mil. USD yearly in the USA. In Europe as well as in the Czech Republic, similar data on damages is missing; however, according to partial information from farmers, it is obvious that economical impact of deer damages is not negligible (Putman and Moore1998; Kamler et al. 2005).

Wild herbivores generally damage the crops intensively in two main periods. 1. In spring, when the food available supply is limited, they browse the green parts of winter crops and growing spring crops; 2. in summer, they feed on the ripening seeds (Selting and Irby 1997; Wywialowski 1996). While the damage sustained on mature seeds means an irreversible decrease of the yield, damage caused to leaves at the beginning of the growth may have little or no impact on the yield of some crops (Yao et al. 1991; Yang and Midmore 2004). The ability to regenerate depends on the compensatory plant growth of individual crop species (Augustine and McNaughton 1998; Pilson and Decker 2002), type, intensity and vegetation stage of the damage (Obrtel and Holišová 1983; Tewolde et al. 1994), health condition of the plant and weather course in the specific year (Moriondo et al. 2003; Cerkal et al. 2007). The impact of different types and intensities of leaf damage on the yield was monitored in several studies where the damage was simulated (Table 1). Effective reduction of damages caused by game must be therefore based on detailed information about attractiveness of particular field crops in different vegetation stages and about critical periods, when the impact of browsing on subsequent yield is the greatest (Brinkman et al. 2005).

Table 1 Results of some studies determining the effect of simulated damage on field crops
Full size table

In spite of various results obtained, field crops can be generally divided into two groups: 1. crops with good ability to regenerate that can compensate for biomass loss (dense-sown crops such as winter wheat) and 2. crops in which the damage to the green parts usually leads to a decrease of the yield (e.g. soya bean, possibly sunflower and others).

The area sown with sunflower has increased over the last years in the Czech Republic. This has resulted in the necessity to grow sunflower also on less suitable sites where the contact with big herbivores cannot be avoided. Wild herbivores visit sunflower fields intensively; however, there are no data available on how their browsing affects the subsequent yield. In comparison to maize, sunflower has the advantage of losing its attractiveness for big herbivores at the end of summer; therefore, the damage it suffers is not as visible as in maize that gets completely shattered by wild boars in some parts. On the other hand, sunflower can be more significantly affected by browsing during its growth. The aim of our work was to evaluate the influence of roe and fallow deer browsing on sunflower during the vegetation. We assumed that damage inflicted on sunflower´s leading stem in any vegetation stage will result in decreased yield or in death of the whole plant.

Materials and methods

Study area

We studied the impact of roe and fallow deer on sunflower between May and September 2006 in a field located in the south-eastern part of the Czech Republic, approximately 30 km south of Brno near the village of Morkůvky. This area is characterised by intensive agriculture and small forests surrounded by fields. Large herbivores occupy mostly the forests and often graze on attractive field crops in near surroundings. Our field (48°58′N,16°50′E) was of approximately rectangular shape with dimensions of 1 km × 0.5 km with a forest adjacent to its short side. The size of the forest was about 5 km2, deciduous trees prevailed. The forest management focused on the maintenance of high deer densities. Stable population of roe and fallow deer occupying this forest was estimated to 20 fallow and 15 roe deer per km2.

Data collection and analyses

We collected data about deer impact in permanent plots and transects. We located nine permanent paired plots in three lines in the distances of 55, 250 and 520metres from the forest edge (Fig. 1). Plots were 4 rows in width and 40 m in length and were divided into two parts – a control part treated with repellent and an untreated part. Deer impact was monitored regularly from the emergence of plants until the harvest. Treatments by repellent and data sampling were performed 10, 22, 36, 54 and 77 days after emergence (DAE). We monitored the number of plants on each plot, signs of browsing (terminal clipping or leaves browsing), and further development of the previously damaged plants. During each visit, we sprayed the plants on the control plots with repellent Lanol N (Eko cheming Strakonice, Czech Republic) to prevent deer browsing. This repellent is intended for protection of trees during vegetation period and is recommended for time limited protection up to 20-30 days. We collected additional data on transect along the plots where we registered the density of the browsed plants, presence of deer footmarks, and density of broken plants. The plots were harvested manually just before the commercial harvest 144 DAE and the number of all plants on the plots, number of damaged plants and yield were recorded.

Fig. 1
figure 1

Location of nine paired plots in three lines in the distances from forest edge

Full size image

Differences in intensity of browsing and yield between protected and control plots were tested by ANOVA using the statistical programme SPSS 11.0.

Results

Usage of the monitored sunflower by herbivores during the growth

Large herbivores used the monitored area regularly; we found fresh track and faeces during every visit to the studied area. We recorded intensive browsing at first visit of study area (10 DAE), when was damaged 9.7 and 9.9% of plants on control and treated plots, respectively (treated plots had not been protected jet) in the line closest to the forest (Table 2); in the next two lines, 5% of plants were damaged on average. Deer browsed the plants at the height of 1 to 2 cm above the ground. During further growth, the plants continued to be extensively affected by browsing. When the plants had reached approximately 12 cm in height, we identified 4.3, 2.3 and 0.6% of the plants to be damaged in the first, second and third line, respectively. At this stage of growth, deer browsed the plants at the height of 5 to 10 cm above the ground. As the plants had reached the height of approximately 40 cm, 13, 7.9 and 6.7% of the plants were damaged on the first, second and third line, respectively. At this stage of growth, deer browsed the plants at about 30 cm. Later on, the plants developed disc florets and herbivore browsing was confined to leaves or substitute stems of the previously browsed plants. At the flowering stage, the sunflower ceased to be attractive as a diet component for big herbivores; they did not damage the developed seeds either.

Table 2 Mean height and proportion of damaged plants of studied sunflower at the time of data sampling (days after emergency – DAE) on the line 1 (nearest to forest edge)
Full size table

Treating the areas with game repellents did not fully protect the control plots; however, the intensity of the damage inflicted upon the plants was less than half compared to that suffered by the non treated areas. Intensity of deer browsing was highest near the forest edge in line 1 (Table 2).

Effect of herbivore browsing

Herbivore browsing affected density of plants on plots and size of disc. Effect of browsing at early stage of development of sunflower was serious. Approximately one half of the plants that had been damaged at the time of our first check died immediately after browsing. 12.5% of the plants that had been damaged (by browsing at 5 to 10 cm above the ground) when they reached the height of 10 to 15 cm died, and the rest produced substitute stems. Number of plants on control plots is in (Table 3).

Table 3 Number of sunflower plants on studied plots at 10, 36 and 144 days after emergency (DAE)
Full size table

Approximately one third of substitute stems developed no or only minimal disc florets. One third of the damaged plants that had formed substitute stems developed disc florets with a diameter of 6 to 8 cm, and one third grew disc florets with a diameter of 8 to 10 cm. The yield of seeds established on the browsed areas from the line closest to the forest was significantly lower than from the control plots (N = 12 = rows; t = -3.301; p = 0.003); in the second and the third line, the yield from the browsed plots did not differ from the yield of the control plots (second line: N = 12; t = -0.199; p = 0.844; third line: N = 12; t = -2.059; p = 0.0522; Table 4).

Table 4 Yield of seeds from browsed areas compared to control plots in%. * indicate significant differences between control and treated plots
Full size table

Discussion

It is understood that agricultural plants form a common part of wild herbivores diet and their presence in the landscape will always result in some damage to the crops. Therefore, the aim of game management is to curb the damage to an acceptable level by reducing the numbers of herbivores, and by additional measures (e.g. supplemental feeding, chemical repellents, electrical fencing).

In spite of the fact that the vegetation in forests is fully developed at the time when the sunflower plants emerge through the soil surface, the monitored field still represented a very attractive food resource for big herbivores. This is probably due to the high attractiveness of young sunflower plants. Both roe and fallow deer used our study area intensively and damaged the plants by browsing on leaves and apical growing points. Damages to sunflower caused by wild herbivores are therefore absolutely different from damages to maize: 1.Maize suffers the biggest damage during the summer and autumn during ripening the corn, sunflower sustains the most significant damage at the very beginning of vegetation; later on, the intensity of the browsing drops (Wywialowski 1996). 2. Maize is damaged mainly by wild boar; sunflower by large herbivores and also brown hare (Herrero et al. 2006; Kamler et al. 2008; Schley et al. 2008). Therefore, protection of sunflower must respect its high, but temporary attractiveness in the early stages of growth. Effect of damage is highest just after emergency, when majority of sunflower plants damaged at the beginning of vegetation died.

Based on the recorded browsing intensity and its impact on further growth of the sunflower, it is possible to identify six main periods (Table 5): 1st period – the plants emerge through the soil and grow until they reach the height of approximately 10 cm. Plants at this stage were extremely attractive for herbivores, and browsing usually resulted in death of the attacked plant. However, the number of the damaged plants was very difficult to determine as there was almost nothing left of them. They withered up and disappeared within a few days. After three weeks all that was seen were gaps in the rows.

Table 5 Periods of sunflower growth according to its attractiveness for herbivores and to effect of browsing
Full size table

2nd period – the plants reach the height of 10 to 20 cm, they have already developed some true leaves, a stronger stem and a root system. Herbivores were browsing the plants from the top to the height of approximately 2 - 4 cm above the ground. Plants damaged during this period usually survived and formed substitute stems with lower yield.

3rd period - the plants reach the height of 20 to 50 cm. At such height, herbivores browsed leaves on the upper part of the stem, or on the whole plant top. Such damage did not result in the death of the plants, and substitute stems were formed.

4th period – the plants are higher than 50 cm and they are just about to start flowering. In case of our monitored area, the plants reached 120 to 200 cm and only leaves of these plants were damaged.

5th period – from the beginning to the end of the flowering. Attractiveness of the sunflower plants was dropping fast and the game eventually stopped browsing on leaves. Browsing was recorded on plants damaged during the previous vegetation stages because these plants had formed substitute stems offering younger biomass.

6th period - the end of flowering till the harvest. No significant damage was inflicted upon the monitored area during this period. Mature sunflower plants did not provide attractive green biomass for the herbivores to feed on, and they were not interested in the developed seeds either. The damage recorded was caused by small mammals and birds. Big herbivores damaged sunflowers only by trampling.

The recorded yield loss was significant on the monitored area, although the control plots were not completely protected by the repellents. Based on the collected data, the calculated loss was approximately 30% of the yield on the half of the monitored area closer to the forest; the other half of the area was considered as not damaged.

Conclusion

Significant influence of fallow and roe deer browsing on the sunflower was found. Young plants were very attractive for the herbivores, and the game preferred them in comparison to natural food supply. Browsing on small plants had significant impact on their further development, and intensive damage often resulted in the death of the whole plant. Therefore, the sunflower proved to be very sensitive to damage inflicted by big herbivores and browsing on the plants cause serious loss in the yield. Determining the level of the damage is very difficult, because the most significant damage is suffered in the course of the intensive growth; during harvest, the damage is no more apparent. An option is to fence the control plots to exclude the influence of the game as application of repellents did not totally prevent browsing. Sunflower should be grown on sites that are less accessible for big herbivores, including brown hare; otherwise, some form of protection is necessary at least at the beginning of the season. The critical period in which the most significant browsing occurs starts right after the beginning of the growth and lasts till the start of flowering (20 -30 days). Damages by hares are often of high significance at early stages of vegetation.

With regard to the increasing area of sunflower fields, further research is necessary in order to determine the exact reaction of sunflower to browsing and to examine the possibilities of the sunflower protection on the threatened areas. Both farmers and game managers would profit from protection of sunflower fields during the first days of growth.