Suitability of patches and in field strips for Sky Larks Alauda arvensis in a small parcelled mixed farming area

Bird Study ISSN: 0006-3657 (Print) 1944-6705 (Online) Journal homepage: https://www.tandfonline.com/loi/tbis20 Suitability of patches and in field strips for Sky Larks Alauda arvensis in a small parcelled mixed farming area Judith Fischer, Markus Jenny & Lukas Jenni To cite this article: Judith Fischer, Markus Jenny & Lukas Jenni (2009) Suitability of patches and in field strips for Sky Larks Alaudaarvensis in a small parcelled mixed farming area, Bird Study, 56:1, 34-42, DOI: 10.1080/00063650802648127 To link to this article: https://doi.org/10.1080/00063650802648127 Published online: 24 Mar 2009. Submit your article to this journal Article views: 468 Citing articles: 11 View citing articles Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalinformation?journalcode=tbis20

Bird Study (2009) 56, 34 42 Suitability of patches and in-field strips for Sky Larks Alauda arvensis in a small-parcelled mixed farming area JUDITH FISCHER*, MARKUS JENNY and LUKAS JENNI Swiss Ornithological Institute, CH-6204 Sempach, Switzerland Capsule Sky Larks make use of agri-environment measures in winter wheat for nesting and foraging during nestling provision. Aims To determine the suitability of in-field measures (patches and strips) for breeding Sky Larks in smallparcelled Swiss lowland farmland. Methods Focal territories established over winter wheat fields with in-field measures (four patches or one strip per hectare sown with arable weed mix) and over conventionally managed wheat, were surveyed during one breeding season. Territories were digitized in a geographical information system based on the mapping of singing males. Breeding pairs were observed to determine the position of nest-sites and foraging spots during nestling provisioning. Breeding success and chick body condition were also determined. Results Winter wheat fields containing in-field measures were more likely to be part of a Sky Lark territory from June onwards than conventional winter wheat fields. Conventional winter wheat fields were used significantly less in July than in May and June. When a nest was built in winter wheat it was significantly more often in or near an in-field measure than expected. During the nestling period, the infield measures were the preferred locations for foraging followed by field borders and spring sown crops. Conclusion The tested in-field measures help to prolong the use of winter wheat for Sky Larks during the breeding season. In a mixed and small-parcelled farming system, their application is not as urgent as in larger-scale monocultures. Nevertheless, Sky Larks (as well as other wildlife species) use them actively. IP-SUISSE (the initiative to encourage integrated farming in Switzerland) encourages farmers to implement patches and in-field strips on a voluntary basis. Since the 1970s Sky Lark Alauda arvensis populations have decreased severely in western Europe (Busche 1989, Tucker & Heath 1994). This decline has mainly been attributed to the intensification of agricultural practices in both grassland and arable land. In intensively managed cultivation, vegetation grows too dense and too tall for Sky Lark nesting early in the breeding season. Moreover, the replacement of spring-sown cereals in favour of winter-sown varieties has led to a further curtailment of the Sky Lark s breeding season (Donald & Morris 2005). In landscapes dominated by winter cereal, this has caused breeding pairs to abandon their territory after one breeding attempt, which is too few to sustain populations (Jenny 1990b, Wilson et al. 1997). *Correspondence author. Email: judithfischer@vogelwarte.com Various studies have looked at how to counteract the negative effects of intensive farming practices on Sky Larks. They have investigated the potential benefits for Sky Larks of organic versus conventional farming (Wilson et al. 1997), set-aside land (Vickery & Buckingham 2001, Bracken & Bolger 2006), wildflower strips (Weibel 1999) as well as the sowing of ancient low-yield cereal varieties such as Emmer Triticum dicoccum and Einkorn Triticum monococcum (Stöckli et al. 2006). These studies all reported higher densities of breeding pairs and improved breeding conditions owing to sparser and/or lower vegetation. The latest projects included wide-spaced rows (H. Ilner pers. comm.) and undrilled patches, both with the aim of creating vegetation heterogeneity, reducing vegetation density and thereby adjusting it to Sky Lark foraging and nesting requirements. In the late 1990s, it 2009 British Trust for Ornithology

Use of in-field measures by Sky Larks 35 was found that Sky Larks use microhabitat structures in cereal fields when the vegetation reaches a critical height (Odderskaer et al. 1997, Schön 1999) or, when no microhabitats are available, that they are forced to nest on tramlines and field borders where predation rates are almost doubled (Donald 2004). Researchers in the UK have therefore developed in-field measures to open up the sward structure of winter wheat (SAFFIE project, Morris et al. 2003, 2004). After the first trial season of the implementation of patches, it was shown that Sky Larks can benefit from this wheat management method, especially in the second half of the breeding season in June and July (Morris et al. 2003). In Switzerland, between 1990 and 2004, Sky Lark populations decreased by 24% (Zbinden et al. 2005). In 2006, a pilot project was conducted to investigate whether or how Sky Larks use in-field measures in Switzerland. In such a small-parcelled mixed farming system, Sky Larks have a wider potential range of nesting and foraging habitats within their territories, besides winter wheat. Differences from the British results were therefore predicted as Swiss agriculture is based on mixed farming with smaller-parcelled plots (1 2 ha). Instead of implementing many small patches in the relatively small fields, it was decided to create larger and contiguous strips to keep implementation easy. The applied in-field measures were thus either patches of 3 12 m, or longer strips of 2.5 80 m, which covered about 150 200 m 2 per hectare. The aim of this study was to quantify the effects of these specially designed measures in winter wheat on the habitat use and breeding biology of Sky Larks in the small-parcelled lowland farmland typical of Switzerland. We were interested in whether Sky Larks use the in-field measures as breeding and/or foraging sites and whether territories containing in-field measures were smaller in size because of improved habitat/vegetation structure. We wanted to know whether winter wheat containing infield measures could be used throughout the entire breeding season and whether, therefore, territories with in-field measures were shifted less and whether those territories produced more chicks and/or chicks with a better body condition. METHODS Study area and design The fieldwork was conducted in the Swiss lowland midlands near Berne (47 7 N, 7 34 E, 475 m asl) in a plain of about 5.8 km 2 located in three neighbouring communities. The study area was a small-parcelled (mean field size 1.25 ha) mixed farming system typical of Switzerland and consisted for the most part of cereals (29.2%), root crops (mostly potatoes, 17.6%) and intensively used meadows (22.3%). Farmers were asked to implement four patches per hectare (each 3 12 m) or one strip (2.5 80 m) in winter wheat fields during sowing in autumn 2005. The patches and strips were managed like the rest of the field, but not sown and not sprayed with herbicide. The selected winter wheat fields with patches or strips were distributed in clusters in three neighbouring communities. Nearby conventionally managed winter wheat fields over which Sky Larks sang were selected as the control fields. Similarly, fields with patches or strips were only chosen for this study when singing Sky Larks could be observed at the beginning of the breeding season. There was a total of seven wheat fields containing patches (between four and nine, according to field size) and 14 containing strips. For the control territories, Sky Larks were observed on 16 conventional winter wheat fields. In 11 of the 14 wheat fields with strips, the strips had to be scarified with a harrow in mid-april 2006 because too few measures implemented by farmers were available in the study area. A mixture of six annual weed species (Agrostemma githago, Centaurea cyanus, Misopates orontium, Camelina sativa, Papaver rhoeas, Legousia speculum-veneris) was sown in all measures to obtain homogeneous and comparable vegetation. In April and May, the vegetation inside the patches and strips was sparse. It became suitable for nesting towards the end of June (vegetation cover then was between 35 and 50%, vegetation height ranged from 5 to 80 cm). The study was conducted in 2006 when the weather was quite exceptional in all months of the breeding period: March, April and May were unusually cold and rainy, while June and especially July were very hot and dry. Due to small sample sizes and because the areas of the two in-field measures were comparable, patches and strips were analysed as one group. Both in-field measures were implemented away from tramlines and field borders to prevent them from acting as predator routes. Land-use in the study perimeter was assessed in May, June and July and was digitized in a geographical information system. Breeding Sky Lark survey During the period of territory establishment in April, we searched for Sky Larks singing over winter wheat

36 J. Fischer, M. Jenny and L. Jenni fields with in-field measures and conventional winter wheat fields for the control group (n = 16). The positions of singing males were recorded onto maps (scale 1:5000). Only fields on which Sky Larks were present during at least the first half of the breeding season were included in the analysis. Because of time constraints, not all territories present in the study area were recorded and no breeding pair density was calculated. We surveyed territories over winter wheat fields with in-field measures (patches/strips) and control wheat fields through one breeding season. The focal territories (patches/strips and control) were revisited every two or three days for 30 minutes. At the end of each month, all observations were collated on monthly maps and georeferenced. These were used to visualize the distribution of territories over control and experimental winter wheat fields during the breeding season and to calculate territory sizes and use of winter wheat fields. Nest mapping and foraging observations For each nest found, the crop type, distance to the nearest in-field measure and distance to the nearest field border, pathway or tramline were recorded. These data were used to examine whether the nests were randomly distributed or accumulated in or close to patches or strips. Because Sky Larks often walked several metres to the nest after landing, we defined nests 5 m or closer to an in-field measure as being in proximity to it. To examine foraging site preferences, nest-based foraging watches were carried out (Wilson 2001). For each foraging flight, notes were made on the duration (flying from the nest until flying back), distance and the crop/vegetation type of the foraging location. Seven crop types were distinguished as foraging locations (Table 1). Nests were visited only every two or three days in order to limit disturbance and predator attraction. Productivity per pair per season was defined as the sum of nest-leaving chicks (around day 8 after hatching) of all broods in a given territory. A nest was said to be successful if at least one young left the nest. In order to compare growth of nestlings of territories containing in-field measures to those in untreated fields, the young were weighed and tarsus length was measured using standard procedures (Morris et al. 2004). Analysis The proportions of crops used versus crops available for foraging (Table 1) were compared in a compositional analysis (Aebischer et al. 1993) in the Excel Macro version Compositional Analysis v. 6.1 plus (Smith 2006). The percentages of all available crop types around a nest area were compared to the percentages of the crops used by Sky Larks during foraging in that area. The proportions of used habitat were calculated as the sum of all foraging flights per nest (i.e. per Sky Lark pair). To render crop elements in a specific nest area independent and approximately normally distributed, log-ratios were calculated for all seven used and available crop classes (Aebischer et al. 1993). Null proportions were replaced by 0.01. The means of the resulting six pair-wise differences were tested for their significant difference from zero using the Wilks λ statistic. This was determined by randomization with 1000 iterations as recommended by Aebischer et al. (1993). Due to a considerable variance in numbers of observed flights in the nest areas, the log-ratio differences were weighted by the corresponding squareroot number of observations per nest. A significant difference indicated that crop class utilization was nonrandom and that Sky Larks preferred certain crop classes when foraging. From this analysis a ranking was established in which the most positive log-ratio Table 1. The seven crop types used in the compositional analysis. Crop classes Spring crops Autumn crops Grassland Grassy tracks Field borders Patches/strips (P/S) Others Description Spring cereals, sparse vegetation (including maize <35 cm), potatoes, sugar beets, vegetables Winter cereals, oil-seed rape, protein peas Meadows, pastures Fully or partly grassy tracks and pathways Borders were transformed into areas by buffering them with 1 m on both sides Sealed roads, pathways and fallow land without vegetation, maize >35 cm

Use of in-field measures by Sky Larks 37 differences received the highest rank and was thus the most preferred crop class. We used χ 2 tests to compare expected and observed frequencies of productivity in patches/strips and control territories, expected and observed number of nests in winter wheat and of nests in or near the patches/strips. For the latter, the Fisher s exact test was applied because the expected number of nests in or close to patches/strips was less than five. A two-way ANOVA was used to trace differences between territory sizes in patches/strips and control territories over the months. All P values reported are two-tailed with rejection levels set at 5%. All analyses except those for nestling growth were carried out in SPSS 12.0. Nestling data were analysed in an autoregressive linear mixed model in SAS 9.1.3 (PROC MIXED). Dependent variables were body mass (g) or tarsus length (mm). Along with the treatment (patches/strips versus control territory), date of hatching, brood size and age were included as fixed effects. Because body mass did not increase with a constant rate but decreased before the chicks left the nest, quadratic age was incorporated as well. To reduce multicollinearity between linear and quadratic regression coefficients such as age and age-squared, age was centred. It was thus expressed as a deviation from its mean (Singer 1998). Random effects were nestlings nested in the different nests. Because most nestlings were measured three times (at different ages), an autoregressive covariance structure was used. All main effects and their two-way interactions were tested. Lark territory decreased significantly from June to July (control wheat, paired t-test, t = 5.220, df = 12, P < 0.001; wheat with patches/strips, paired t-test, t = 3.068, df = 16, P = 0.007). Sky Lark males shifted their territories away from control wheat fields, so that in July a lower proportion of a control wheat field remained inside a Sky Lark territory than before (Fig. 1). However, the percentage of a winter wheat field containing patches/strips was significantly higher than the proportion of control wheat in July (t = 2.171, df = 28, P = 0.039). Thus, wheat fields with patches/strips remained to a larger proportion in Sky Lark territories than control winter wheat fields in July. Nest-site selection A total of 29 nests were located. In May, all nine nests found were in winter wheat. In June and July, however, only seven nests were attempted in winter wheat, six in potatoes and seven in other spring-sown crops. In the territories with patches/strips, the proportion of winter wheat ranged from 10.4% to 87% (mean 50.0%) during the breeding season and ten out of 19 nests (53%) were found in winter wheat. Thus, winter wheat was not a preferred crop type for nesting over the entire season, but was used according to its availability in the territories (all the broods and months pooled, n = 19, χ 2 = 0.024, P > 0.2). Among the ten nests in winter wheat with in-field measures, two were on the edge of, and four were in close proximity (<5 m) to an in-field measure (Fig. 2). RESULTS Sky Lark territory size and use of winter wheat fields The size of Sky Lark territories containing experimental wheat fields (those with patches or in-field strips) did not differ significantly from those containing control wheat fields in any month (May, t = 1.33, df = 44, P = 0.19; June, t = 1.54, df = 51, P = 0.13; July, t = 0.56, df = 46, P = 0.58). Territory size was smaller in June (1.02 ha ± 0.25) and July (1.07 ha ± 0.36) than in May (1.16 ha ± 0.29) (two-way ANOVA: treatment, F 1,141 = 1.19, P = 0.28; months, F 2,141 = 5.14, P = 0.007; interaction treatment * month, F 2,141 = 1.24, P = 0.29). Sky Lark territories shifted during the breeding season. Winter wheat lost its attractiveness for breeding Sky Larks from mid-may onwards and the percentage of a winter wheat field included in a Sky Figure 1. Mean percentage of winter wheat fields included in Sky Lark territories (control wheat (dark columns, n = 13) compared with wheat with patches/strips (grey columns, n = 17); error bars are ±1.0 se).

38 J. Fischer, M. Jenny and L. Jenni Figure 2. Distance from nests in winter wheat to either a patch or a strip (n = 10). Six nests were within 5 m from a patch or a strip. Three of the four remaining nests were early nests (grey bars) started when average height and cover were still suitable in winter wheat. Four nests were not in proximity; three of those, however, were the earliest broods started at the end of April/beginning of May when the surrounding wheat was still suitable (vegetation height 45 cm, vegetation cover 50%, Jenny 1990a, 1990c, Wilson et al. 1997). The number of observed nests near patches/strips was significantly higher than the expected number (i.e. the expected number of nests according to availability; 11.6% = 0.815 nests; Fisher s exact test, P = 0.029). Therefore, when a nest was built in winter wheat, Sky Larks preferred sites near patches/strips from mid-may onwards. Nest success and productivity Of the 29 nests found, 27 were active and 21 of those nests were discovered during the egg stage. From these 21 clutches, 13 (62%) failed and eight (38%) were successful. The outcome of a brood depended on its distance to the nearest field border, pathway or tramline. Abandoned nests (death of female or harsh weather conditions) were excluded. Failed nests had been placed significantly closer to tramlines or borders than successful broods (Fig. 3, z = 4.475, P = 0.005). All nests closer than 2 m to a field border or a tramline failed. In seven territories over fields with patches/strips and ten control territories, no nests or fledglings were found. It is very probable that several nesting activities remained unnoticed, but these attempts must have failed as no feeding Sky Larks were observed. There was not a greater number of patches/strips than control territories in which productivity could be assessed (i.e. territories with observed nesting attempts; patches/ strips versus control, χ 2 = 1.486, P > 0.05, n = 22 territories with patches/strips, 20 control territories). The Figure 3. Number of nests found and their distances (m) to the nearest tramline or field border. Failed nests (dark columns, n = 10) were very close to borders (<8 m). Successful nests (grey columns, n = 14) showed greater variability in their distribution.

Use of in-field measures by Sky Larks 39 number of young fledging was unusually low with 1.4 chicks per territory for both groups (21 chicks in 15 territories with patches/strips, 14 young in ten control territories). There were not more territories with patches/strips than control territories with fledging chicks (χ 2 = 0, P = 1, n = 15 territories with patches/strips, ten control territories). Hence, it could not be concluded that territories with patches/strips produced more nests or fledging chicks than control territories. Foraging sites Foraging flight destinations could be observed for ten nests in nine territories with patches/strips during foraging watches. Ninety-six per cent of the observed flights were within a distance of 150 m around the nestsites (n = 166). Therefore, a radius of 150 m was chosen as the appropriate range in which the available foraging habitats were determined. On average, seven flights per hour took place (median = 7, range 2 14). Seven out of nine Sky Lark pairs with patches/strips present in their territories made use of them during chick feeding. Patches and strips covered only from 0.17% to 0.63% of the available foraging area (150 m radius). However, they were accessed on 12.6% of the observed flights. Thus, the in-field measures received the highest rank (Table 2) and were the most preferred of all crop type classes. Autumn crops and the class others (sealed roads, bare fallow land and maize >35 cm) were avoided (compositional analysis, Fig. 4, Table 2). Grassy tracks were neither preferred nor avoided, but Sky Larks could often be seen foraging on grassy pathways when they did not have a nest and foraged for themselves. Grassland was avoided, but cut grass within the nest range was used significantly more often during foraging than expected. Only 18.8% of the available grassland was freshly mown but 69.1% of the grassland flights were made into freshly cut grass (χ 2 = 21.5, P < 0.001). Nestling growth Growth of nestlings was measured in nine nests from territories with patches/strips and six nests from control territories. All the models investigating the treatment effect on growth parameters found no significant difference in body mass and tarsus growth between nestlings from territories with patches/strips and control territories. Tarsi were non-significantly longer in nestlings from patches/strips territories, but body mass was slightly lower (Table 3). Temperature and date of hatching were highly correlated (Pearson correlation, r = 0.821, n = 122, P = 0.0001). Later hatching dates (and therefore increasing mean temperature) had a positive influence on both body mass and tarsus development. With increasing brood size, nestlings were on average 0.7 ± 0.27 g lighter. DISCUSSION Sky Lark territory size was not affected by the presence of in-field measures (patches or strips) in winter wheat fields. Both control territories and territories with patches/strips decreased in size from May to July, in agreement with findings in NE Switzerland (Stöckli et al. 2006). From June to July the percentage of control fields included in Sky Lark territories decreased significantly (from 60% to 38%). This shift away from progressively less suitable winter cereals was also noticed by Stöckli et al. (2006). In contrast, the percentage of experimental wheat fields used decreased less than that of controls and remained at about 55% (Fig. 1). From late June onwards, there was significantly more wheat containing in-field measures Figure 4. Comparison of crop type availability and use (each adding up to 100%) by foraging Sky Larks provisioning nestlings at ten nests in nine territories with in-field measures (n = 166 flights). P/S, Patches/strips.

40 J. Fischer, M. Jenny and L. Jenni Table 2. Compositional analysis matrix of the crop type classes. Spring Autumn Field Patches/ Grassy Crop type class Others crops crops borders strips Grassland tracks Rank Others 2.997** 1.179 4.500** 6.272** 3.546** 2.733 0 Spring crops 2.997** 1.834 0.833 3.229** 0.084 0.826 4 Autumn crops 1.179 1.834 2.667 5.064** 1.233 0.679 2 Field borders 4.500** 0.833 2.667 2.397 1.382 2.030 5 Patches/strips 6.272** 3.229** 5.064** 2.397 3.118 4.823** 6 Grassland 3.546** 0.084 1.233 1.382 3.118 0.300 2 Grassy tracks 2.733 0.826 0.679 2.030 4.823** 0.300 2 Matrix of t-values comparing all classes against each other with indication of their significance (**P < 0.05). Weighted mean λ = 0.0827. P randomized = 0.034. Log-ratio differences derived from each of the ten nest areas were weighted by the corresponding square-root number of observations. A significantly positive t-value indicates that the habitat type in the first column is preferred over the habitat type indicated in the top row. In the last column the ranks for all crop type classes are given according to their preference by Sky Larks; the highest rank indicates the most preferred habitat type (patches/strips, rank 6). Table 3. GLMM for nestling body mass and tarsus development. Body mass (g) Tarsus (mm) Fixed effects Estimate se df P Estimate se df P Intercept 1141.25 198.54 34 <0.0001 907.55 104.89 35 <0.0001 Centred age 3.253 0.061 83 <0.0001 2.465 0.074 82 <0.0001 Centred age squared 0.155 0.020 82 <0.0001 Treatment: control 0.194 0.399 34 0.63 0.385 0.324 35 0.24 Treatment: P/S 0 0 Brood size 0.699 0.267 34 0.01 Date of hatching 0.068 0.012 34 <0.0001 0.054 0.006 35 <0.0001 Data of nestlings from patches/strips territories were set to 0 and then compared to control nestling data. Compound symmetry covariance structure was assumed for the tarsus model. included in territories than control wheat. This suggests that Sky Larks shift away less from winter wheat with in-field measures than from conventional fields and that it can still serve as breeding and foraging habitat later in the breeding season. In the SAFFIE project (Morris et al. 2004), the number of singing males and nesting birds on plots with patches decreased only by half of that on the control plots. Thus, the establishment of patches and in-field strips can prevent Sky Larks from shifting or abandoning winter wheat, which is especially important in large-scale wintercereal-dominated areas. It became evident that nests can be placed several metres from an in-field measure and it will still be used as an access route to the nest (Morris et al. 2004, A. Pille pers. comm.). Hence, by leaving only small patches unsown, a large benefit could be gained for breeding Sky Larks. In many territories, with or without patches or strips in winter wheat fields, Sky Larks chose a more suitable crop for breeding after winter wheat had grown too tall and dense. But when a nest was built in winter wheat, it was more often in or near a patch or strip than expected by availability. This demonstrates a preference for breeding sites near infield measures from mid-may onwards. The most important factor determining breeding success was the distance to the nearest tramline or field border. Failed nests were significantly closer to tramlines or borders than successful broods (Fig. 3). All eight nests within 2 m from a tramline or a field border failed. As the availability of sparse and low vegetation decreases drastically around the middle of the breeding season, Sky Larks are forced to build their nests very close to borders or tramlines where the risk of predation is almost doubled (Donald et al. 2002). Therefore, infield measures can serve as good quality breeding habitats provided that they are implemented as far away from tramlines and field borders as possible. Productivity was identically low in territories with in-field measures and control territories (1.4 chicks as opposed to 1.6 1.8 nest-leaving chicks per pair per season reported by Jenny (1990c) and S. Stöckli (pers.

Use of in-field measures by Sky Larks 41 comm.); Schläpfer (1988) obtained similar values in years with suboptimal weather). No significant differences in body mass were found between nestlings in territories with in-field measures and control territories. From the findings of this study, no clear-cut conclusions can be drawn on whether or not in-field measures enhance productivity and nestling condition. Productivity may be higher in years with less extreme weather conditions. In Switzerland, Sky Larks can attempt nesting in alternative crops and do not abandon their territory as quickly as in large-scale monocultures, so that an increase in productivity like the one found in the UK (Donald & Morris 2005) cannot be expected. Especially in small-parcelled mixed farming systems, the effect of in-field measures could be enhanced by implementing them in other crops. The structure of generally unsuitable oil-seed rape and winter barley could be opened up. In maize and sunflowers, lack of ground cover prevents Sky Larks from nesting. In these crops, the wildflower mix sown into the in-field measures could improve the nesting situation. Field borders, spring-sown crops and freshly cut grass were attractive foraging habitats for Sky Larks, confirming earlier studies (Jenny 1990a, Wilson 2001). The in-field measures covered on average only 0.3% of the area, but were used for 12.6% of the foraging flights (Fig. 4) and were the most preferred crop class during foraging observations. Clearly, the patches and strips are covered with a very suitable vegetation structure which might improve food accessibility for Sky Larks. Increased activity of wolf spiders (Lycosidae) was discovered during this study (pers. obs.). This could be a reason for the preference of in-field measures by Sky Larks. Holland (2005) emphasized that more weeds can enhance chick food abundance. Hence, the sown wildflower mix could further increase the value of in-field measures as foraging spots because a larger variety of arthropod species could develop there. Adult Sky Larks frequently used the in-field measures when foraging for themselves. Unfortunately, they could not be tracked during foraging. Hence, by observing foraging flights during nestling provision, only a fraction of the total use of the in-field measures by foraging Sky Larks was captured. Conclusions Several studies have highlighted the importance of altered crop vegetation structure, e.g. wide-spaced rows and patches, for farmland birds like the Sky Lark. This study has shown that in-field measures are used by Sky Larks as landing, foraging and breeding spots, confirming other work. Owing to the patches/strips, Sky Larks were able to use winter wheat for a longer time period. For Sky Larks, patchiness in otherwise uniform and dense vegetation is crucial for successful foraging and breeding. The in-field measures enhance food accessibility which is important in foraging site selection. Even though the in-field measures are rather small in size (compared to other measures), their benefit exceeds the area on which they are implemented. Nests were more successful when they were built further than 2 m from field borders and tramlines. After winter wheat reaches a certain height and density, Sky Larks build nests in alternative crops or are forced to shift to sparse spots, field borders or tramlines where nest failure rates are very high. Thus, in-field measures are important additional breeding habitats provided that they are implemented as far from borders and tramlines as possible. Contrary to findings from the UK, productivity and nestling growth was not improved in territories with in-field measures. However, this might also have been a result of suboptimal weather conditions during the breeding season. Nonetheless, these in-field measures provide valuable additional habitat during the breeding season, and they become all the more important in intensively managed farmland. ACKNOWLEDGEMENTS We thank IP-SUISSE for collaboration, and the farmers in the study area for allowing us access to their land. Christian Marfurt helped with GPS mappings and GIS analyses. Samuel Klinkert assisted in collecting the data. We are grateful to Eva Knop and Simon Birrer who commented on an earlier draft of this paper. This study was part of an MSc thesis by Judith Fischer at the Zoological Museum, University of Zurich. We thank Lukas Keller for supervision and comments, as well as Simon Gillings and Rob Field for valuable remarks on an earlier draft. REFERENCES Aebischer, N.J., Kenward, R.E. & Robertson, P.A. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology 74: 1313 1325. Bracken, F. & Bolger, T. 2006. Effects of set-aside management on birds breeding in lowland Ireland. Agric. Ecosyst. Environ. 117: 178 184. Busche, G. 1989. Drastische Bestandeseinbussen der Feldlerche auf Grünflächen in Schleswig-Holstein. Vogelwelt 110: 51 59. Donald, P.F. 2004. The Skylark. T. & A.D. Poyser, London.

42 J. Fischer, M. Jenny and L. Jenni Donald, P.F. & Morris, A.J. 2005. Saving the Sky Lark: new solutions for a declining farmland bird. Br. Birds 98: 570 578. Donald, P.F., Evans, J., Muirhead, L.B., Buckingham, D.L., Kirby, W.B. & Schmitt, S.I.A. 2002. Survival rates, causes of failure and productivity of Skylark nests on lowland farmland. Ibis 144: 652 664. Holland, J. 2005. Arable farming for an improved environment. In Review of 2005, The Game Conservancy Trust. Available at: http://www.gct.org.uk/text03.asp?pageid=325 Jenny, M. 1990a. Nahrungsökologie der Feldlerche Alauda arvensis in einer intensiv genutzten Agrarlandschaft des schweizerischen Mittellandes. Ornithol. Beob. 87: 31 53. Jenny, M. 1990b. Populationsdynamik der Feldlerche Alauda arvensis in einer intensiv genutzten Agrarlandschaft des schweizerischen Mittelland. Ornithol. Beob. 87: 153 163. Jenny, M. 1990c. Territorialität und Brutbiologie der Feldlerche Alauda arvensis in einer intensiv genutzten Agrarlandschaft des schweizerischen Mittelland. J. Ornithol. 131: 241 265. Morris, A.J., Bradbury, R.B. & Evans, A.D. 2003. Sustainable arable farming for an improved environment: the effects of novel winter wheat sward management on Skylarks (Alauda arvensis). In BCPC International Congress Crop Science & Technology 2003, Congress Proceedings: 227 232. BCPC, Alton. Morris, A.J., Holland, J.M., Smith, B. & Jones, N.E. 2004. Sustainable arable farming for an improved environment: managing winter wheat sward structure for Skylarks Alauda arvensis. Ibis 146 (Suppl. 2): 155 162. Odderskaer, P., Prang, A., Poulsen J.G., Andersen, P.N. & Elmegaard, N. 1997. Skylark (Alauda arvensis) utilisation of micro-habitats in spring barley fields. Agric. Ecosyst. Environ. 62: 21 29. Schläpfer, A. 1988. Populationsökologie der Feldlerche in einer intensiv genutzten Agrarlandschaft. Ornithol. Beob. 85: 309 371. Schön, M. 1999. Zur Bedeutung von Kleinstrukturen: Bevorzugt die Feldlerche Störstellen mit Kümmerwuchs? J. Ornithol. 140: 87 91. Singer, J.D. 1998. Using SAS PROC MIXED to fit multilevel models, hierarchical models, and individual growth models. Educ. Behav. Sci. 24: 323 355. Smith, P.G. 2006. Compos Analysis Version 6.2 User s Guide. Version 6.2.3. Smith Ecology Ltd, Abergavenny. Website: http://www.smithecology.com/software.htm Stöckli, S., Jenny, M. & Spaar, R. 2006. Eignung von landwirtschaftlichen Kulturen und Mikrohabitat-Strukturen für brütende Feldlerchen Alauda arvensis in einem intensiv bewirtschafteten Ackerbaugebiet. Ornithol. Beob. 103: 145 158. Tucker, G.M. & Heath, M.F. 1994. Birds in Europe: Their Conservation Status. BirdLife Conservation Series No. 3. BirdLife International, Cambirdge. Vickery, J.A. & Buckingham, D.L. 2001. The value of set-aside for Skylarks Alauda arvensis in Britain. In Donald, P.F. & Vickery, J.A. (eds) Ecology and Conservation of Skylarks Alauda arvensis: 161 175. RSPB, Sandy. Weibel, U.M. 1999. Effects of wildflower strips in an intensively used arable area on Skylarks (Alauda arvensis). PhD thesis. Swiss Federal Institute of Technology, Zurich. Wilson, J.D. 2001. Foraging habitat selection by Skylarks Alauda arvensis on lowland farmland during the nestling period. In Donald, P.F. & Vickery, J.A. (eds) Ecology and Conservation of Skylarks Alauda arvensis: 91 101. RSPB, Sandy. Wilson, J.D., Evans, J., Browne, S.J. & King, J.R. 1997. Territory distribution and breeding success of Skylarks Alauda arvensis on organic and intensive farmland in southern England. J. Appl. Ecol. 34: 1462 1478. Zbinden, N., Keller, V. & Schmid, H. 2005. Bestandsentwicklung von regelmässig brütenden Vogelarten der Schweiz 1990 2004. Ornithol. Beob. 102: 271 282. (MS received 11 September 2007; revised MS accepted 4 April 2008)