Perennial energy crops (PECs) can reduce negative impacts of intensive silage maize cultivation on the environment. Further, remaining vegetation of PECs after harvest may provide suitable habitat and more beneficial overwintering conditions for arthropods than maize. We hypothesized that after harvest and in winter, arthropod abundance and biomass are higher in PECs than in silage maize.

In a field experiment, we compared the two PECs cup plant and field grass with silage maize regarding their suitability as autumn (post-harvest) and overwintering habitats for arthropods. We measured soil temperature and moisture and analyzed biomass as well as abundance of autumn-active and overwintering arthropods of these three crops. During autumn we assessed arthropods by suction sampling in 24 plots of the experimental field. In spring we assessed soil-emerging arthropods for four times in 18 plots.

In PEC plots, soils were moister and less exposed to cold temperatures than in maize. Compared to maize, total arthropod abundance and biomass were higher in PEC plots for both sampling periods. Results were similar for most examined arthropod taxa.

Our results demonstrate that, compared to silage maize, the PECs provide suitable post-harvest habitats and constitute more suitable overwintering habitats for arthropods. We assume that differences are based on lack of disturbance and the provision of vegetation structures after harvest that function as overwintering habitat for arthropods. We conclude that positive effects of PECs on ground arthropods are not limited to their growing time but continue after harvest and during winter.

Maize, cup plant and field grass were grown on an experimental field in Braunschweig, Germany. Details for crop management can be found in Grunwald et al. 2020 GCB Bioenergy. The habitat suitability of the three crops for arthropods was assessed with two methods. Arthropods were collected with suction sampling after harvest and emerging arthropods were assessed using ground photo-eclectors in spring. Further, soil temperature and moisture were assessed in spring and winter.

Temperature was measured with Data loggers ‘Tinytag Plus 2-TGP-4500’ (Gemini Data Loggers, Chichester, UK) (range: –25 °C to +85 °C and an accuracy of 0.01 °C). Data loggers were placed in waterproof plastic boxes for protection, positioned 15 cm below soil level and covered with soil.

Moisture was measured with the device ‘HD2’ and the probe ‘TRIME-PICO 64’ (IM-KO Micromodultechnik GmbH, Ettlingen, Germany) relative soil moisture was measured at three points around the emergence trap set of each plot on every first and last day of emergence sampling in 2019. The device measures volumetric water content of soils based on conductivity and is suited for different types of soil and works efficiently at a temperature range from –15 °C to +70 °C.

For descriptive analysis and plotting winter moisture was summarized to subplot level. The respective two measurements of soil moisture in winter per plot were averaged and used as a response variable in models. For measurements in spring mean temperatures per day and subplot were calculated and moisture data was averaged across sampling positions and subplot.

A suction sampling device (‘ecoVac’, ecoTech GmbH, Bonn, Germany) was used on 2016-09-26 in all 24 plots (eight replicates per crop) between 10:30 and 16:30. Per plot, 18 consecutive sampling points with a distance of 2 m along a central line, were assessed, starting 2 m in from the edges of the plots (sum:  0.28 m² per plot). Diameter: 14 cm, suction strength: 14 m s^-1, duration: 20 seconds.

Ground photo-eclectors (‘Modell 250’ by ecoTech GmbH Bonn, Germany) were used in 18 plots. They had a height of 95 cm and consisted of a plastic cylinder with a cone-shaped tent-structure on top of it. The cylinder was dug partly into the ground, encircling an area of 0.25 m². Let-in flush with the inner wall of the cylinder was one pitfall trap (plastic cups with a volume of 0.5 l and a diameter of 9.5 cm) dug into the ground. Pitfall traps were partly filled with monoethylene glycol (MEG).Transparent eclector head boxes attached to the top of the tents were also filled with MEG and conserved arthropods moving toward light at the top of the tent. One emergence trap set was placed on the longitudinal side of each plot, one meter from the edge.

From suction samples and ground photo-eclectors all arthropods (per plot) were poured onto a sieve of 0.5 mm mesh width and rinsed with tap water. The sieve was placed over a vessel until the time between two drops reached more than 20 seconds after which the sample was weighted (‘Sartorius handy h51’ by Sartorius AG, Göttingen, Germany; accuracy ± 0.3 mg).

Afterwards, arthropods were sorted into coarse taxonomic groups (Araneae, Opiliones, Isopoda, Chilopoda, Diplopoda, Diptera, Hymenoptera, Carabidae, Staphylinidae, other Coleoptera, Hemiptera, Dermaptera and ‘others’). For both methods data on abundance was analyzed separately for individual taxa and for all taxa combined.