Ground beetles (Coleoptera: Carabidae) of rice field banks and restored habitats in an agricultural area of the Po Plain (Lombardy, Italy)

Abstract An entomological investigation was carried out in an agricultural area, mainly rice fields, of the Po river plain, located in the municipalities of Lacchiarella (MI) and Giussago (PV) (Lombardy, Italy). In 2009 and 2010, ground beetles (Coleoptera: Carabidae) were sampled along rice field banks and in restored habitats, by means of pitfall traps. The area appeared as species-rich, compared to other anthropogenic habitats in the Po river pain. Most of the collected Carabids were species with a wide distribution in the Paleartic region, eurytopic and common in European agroecosystems. The assemblages were dominated by small-medium, macropterous species, with summer larvae. No endemic species were found. Species with southern distribution, rarely found north of the Po river, were also sampled. Amara littorea is recorded for the first time in Italy.


Introduction
In the last decades, intensification and mechanization of agricultural practices, introduced in order to maximise productivity, led to a decrease in habitat quality and landscape heterogeneity throughout European agroecosystems. Diffusion of monoculture, increased use of chemicals (i.e. pesticides and fertilizers) and removal of non-cropped areas, like small woodlots and hedges, caused a wide-scale loss of biodiversity (Stoate et al. 2001).
Recently, environmentally-friendly agronomic practices and creation of non-cropped habitats have been recognized as a potential solution to this dramatic decline of biodiversity and have become key aims of European Union's Common Agricultural Policy (CAP) and, as a consequence, of national and regional ones (Stoate et al. 2009). In Lombardy lowland, environmentally-friendly measures includes reforestations, creation of hedges and buffer strips, maintenance of meadows and renaturalization of wetlands (Lombardy Region 2012, http://www.agricoltura.regione.lombardia.it).
Even if agri-environment schemes (AESs) benefit some farmland species (e.g., Peach et al. 2001), gaps in the provision of habitat quality and landscape connectivity for many others still exist (Kleijn et al. 2001, Vickery et al. 2004, Reid et al. 2007). Better understanding on effects of AESs on farmland biodiversity and exhaustive surveys on animal and plant communities in enhanced habitats are required Sutherland 2003, Stoate et al. 2009).
The aim of this research was to investigate the Carabid assemblages of an intensive agricultural area (mainly rice fields) subjected to environmental improvements since 1996, in particular the creation of buffer strips along paddy fields and the restoration of an area of 150 ha.

Study area
The study was carried out in an 4.5 km agricultural area, mainly cultivated with rice, located in north-western Italy, in the middle of the Po plain, approximately 13 km north from the city of Pavia, in the municipalities of Lacchiarella (MI) and Giussago (PV); barycentre 45°17'38.63"N, 09°08'52.08"E ( Fig. 1). 2 The study area included three adjacent rice farms, "La Darsena", "La Cadenazza" and "Necchi", and a restored area, "La Cassinazza". The Carabid fauna was sampled in: • rice field banks ( Fig. 2   Rice field with herbaceous banks. Herbaceous buffer strip along a small wetland connected to paddy field.

Sampling method and data analysis
Ground beetles were sampled using plastic pitfall traps (62 mm in diameter and 70 mm deep) buried in the soil and filled with 50 ml of wine vinegar and a drop of detergent . Pitfalls were covered with a 10 × 10 cm wooden roof to prevent flooding and emptied fortnightly.  1982, Drioli 1987; data on body size and wing development were mainly obtained from Hůrka 1996, and secondly from Jeannel 1941, Jeannel 1942. As for body size, according to Cole et al. 2002, species were divided as (a) very small (< 5 mm), (b) small (5 -9 mm), (c) medium (9 -15 mm) and (d) large (> 15 mm). Data on adult diet were not available for all species and we reported only the existing information, according to Cole et al. 2002, Purtauf et al. 2005, Melis et al. 2010and Bettacchioli et al. 2012.
A synthetic description of habitat preference, derived from Hůrka 1996 and personal observations with special reference to the Po plain, were also reported for each species. According to Fournier and Loreau (1999), we classified the species as "rare" when the total capture over the whole area was lower than 0.1% (i.e. < 35 individuals); the other species Wet meadow with reforested area on the background.
Ground beetles (Coleoptera: Carabidae) of rice field banks and restored ... were classified as "common" and the two most abundant species as "dominant". The total number of captured individuals (n) was reported in brackets.
As for rice field banks and enhanced habitats, ground beetle abundances were expressed both as absolute frequency (i.e. number of collected individuals) and as annual Activity Density (aAD; , that is the number of collected individuals during the entire sampling period (n ) divided by sampling effort (US) for each sampling station:

DAa = n / US
with US = Σ us and us = trap * (gg/10), where trap is number of traps and gg is the number of days during which the traps were active in each sampling session (Suppl. material 1).
Specimens, dried or preserved in alcohol, are stored in the author's collections (Nicola Pilon, Milano) and in the collection of the University of Pavia.
Common in the study area (n = 798). Recorded in all habitat categories.
Rare in the study area (n = 11).
Rare in the study area (n = 1); recorded in arboreal buffer strips only.
Common in the study area (n = 866). Recorded in all habitat categories.
Rare in the study area (n = 5); recorded in herbaceous buffer strips only.
Common in the study area (n = 1001). Recorded in all habitat categories.
Rare in the study area (n = 2).
Rare in the study area (n = 14).
Rare in the study area (n = 26).
Common in the study area (n = 372).
Rare in the study area (n = 2); recorded in herbaceous buffer strips only.
Rare in the study area (n = 8).
Rare in the study area (n = 8).
Rare in the study area (n = 1); recorded in arboreal restored habitats only.
Rare in the study area (n = 7).

Analysis
Overall, we collected 34,108 individuals belonging to 98 carabid species. We recorded 65 species in rice field banks, 73 species in buffer strips and 78 in restored habitats. Eight species were found only in rice field banks ( Number and percentage of carabid species for each ecological categories (chorological complexes, body size, larval and wing development) in rice field banks, buffer strips and restored habitats.
Also rice field banks, buffer strips and restored habitats, analyzed separately, were dominated by Holoartic, medium-small, winged species, with summer larvae (Table 1); species number and percentages for chorotype, body size, larval and wing development were similar in the different habitat categories (Figs 6,7,8,9).  Table 1).

Figure 6.
Percentage of carabid species for each chorological complexes (Subcosmopolitan, Holoartic, European, Mediterranean) in rice field banks, buffer strips and restored habitats (plotted after data in Table 1).

Figure 8.
Percentage of carabid species for each larval development (summer, winter, poliennal) in rice field banks, buffer strips and restored habitats (plotted after data in Table 1).

Figure 9.
Percentage of carabid species for each wing development (macropterous and pteridimorphic) in rice field banks, buffer strips and restored habitats (plotted after data in Table 1).

Discussion
On the whole, 98 carabid species were collected in rice field banks, buffer strips adjacent to paddy fields, and restored habitats (herbaceous and arboreal). Species number could be slightly underestimated because of the sampling method which is not very well suited for some taxa as Lebiinae and Bembidinae. Nevertheless, the area resulted species-rich, especially when you consider that it is not placed inside a riverine corridor and when you compare the species number with that recorded in other anthropogenic habitats of the Po plain: 60-70 species in rye, oat and fallow fields (Pescarolo 1990, Pescarolo 1993; 48 species in a complex of habitats composed by one poplar grove, one artificial wetland, banks of irrigation canals and cropped areas (Casale et al. 1993); 55 species in poplar groves of different ages (Casale et al. 1993); 60 species in meadows of different ages (Gobbi et al. 2005); 60 species in meadows, crops and reforested areas of two urban parks in Milan (Pilon et al. 2010).
Most of the collected carabids, both in the whole area and in each habitat categories, were species with a wide distribution in the Paleartic region, eurytopic and common in European agroecosystems. The assemblages were dominated by small-medium, macropterous species, with summer larvae; we didn't find any endemism.
No brachypterous and strictly forest-dwelling species were sampled, despite the presence of some recent woodlots (i.e about 10 years old). In fact, species unable to disperse by flight were prevented to colonize these stands (including Abax continuus Ganglbauer 1891, very common in woods of the Lombardy plain), because of the absence of ecological corridors connecting woodlots with forest remnants (Macarthur and Wilson 1967). As a consequence, the Carabid fauna was mainly composed by species of open habitats. Most of the species were also hygrophilous, due to a dense network of artificial irrigation canals and a superficial water-table.
The most interesting aspect of this Carabid coenosis is the presence of several species with southern distribution, quite common in clay soil on the right bank of the Po river, and known only in few stations north of the Po river. Among these species, we list Acinopus picipes, Amblystomus niger, Dinodes decipiens, Harpalus cupreus, Harpalus oblitus, Parophonus mendax, Parophonus planicollis, Pterostichus macer. Although a comparison with the past coenosis is not possible for the lack of similar surveys in the area, it could be hypothesized that these species are recent colonizers (7-10 years). They are not reported in the historical catalogue of Magistretti (1965), and are also not listed in several recent faunistic investigations carried out in the Lombardy lowland, particularly along the Ticino river (Pasquetto 1992, Bogliani et al. 2003, Adda river (Conti 1991), Po river (Pilon et al. 1991, Rancati andSciaky 1994) and in Milan (Pilon et al. 2010), where potentially suitable habitats were sampled. Even in an intensive survey along the Po river included in Piedmont region, only some of these species have been collected (Allegro and Sciaky 2001). If so, we could assume a tendency to a northward shift in the distribution of these species, according to what has been observed for other zoological groups well studied and that have great mobility, such as birds (Chen et al. 2011) and dragonflies (Ott 2010).
We underline also the presence of Brachinus plagiatus, an uncommon halophilous species. Moreover Amara littorea, an Asiatic-European distribution species, has been recorded with certainty for the first time in Italy (Cardarelli and Pilon 2012).