Sampling methods

Sites in Tinnerö nature reserve were sampled for carabids using pitfall traps. Traps used were polypropylene made, of 8.5cms diameter and 10cms depth. In each of the 18 sites sampled, five traps were randomly placed into the ground. The traps were spaced at not less than 10 meters intervals as carried out by Cardarelli and Bogliani (2014). The cups were partly filled with water, added with dish detergent, and set flush with the ground. Traps were left closed in the field for 3 weeks (reset on 3 weeks intervals), and after three weeks, the traps would be opened for 5 days (emptying on the 5^th day) and then closed again for 3 weeks. We repeated this for four collections between 1^st May and 9^th August 2017. Four sample collection dates were registered; A = 21st May 2017, B = 18th June 2017, C = 13th July 2017, D = 9th August 2017. This timeframe well covered the time when carabid activity was high as determined by Greenslade (1964).

During emptying, samples were placed in smaller plastic tubes filled with ethanol as a preservative and taken to the lab for cleaning and sorting. Other animals such as mice, grasshoppers, molluscs, bees, and non-carabid beetles were occasionally caught in the pitfall traps but were discarded as they were not of any relevance to this study. Some traps were found trampled or uprooted and destroyed by cattle while others were pushed out from the ground by the increasing ground water, these were replaced with new ones on every reset.

Collecting samples was done in a single day as well as setting up of all traps in all sites. The collected traps contents preserved in about 75% alcohol were pinned to Styrofoam and left to dry prior to species identification. Carabids smaller than 5mm could not be pinned but placed in small tubes with alcohol.

The identification and studying of beetles’ morphology was done under microscope as described by Lindroth (1986) in the Carabids of Fennoscandia and Denmark carabid key and other authentic online material such as Beetlebase.com (2017). The smaller carabids were sent away to a specialist (Sven Lennartsson) for identification.

Measuring canopy and sward height

Two primary metrics were measured: sward height and canopy cover. Sward growth and height is influenced by several factors including moisture, nutrients, and sunlight. In this regard, we use grass height as a relative measure of grazing. Because the current management of the sampled grasslands is exclusively grazing, sward height is also a direct measure of grazing intensity. Because carabids depend on grass for habitation, its well established that the height of grass influences their occurrence. Canopy cover was measured because sunlight exposure also affects carabids (Rainio and Niemela, 2003).

Sward height

Sward height was taken at 3 points around each trap within a 1-meter radius and averaged to give a single value for each trap. To take measurements, we cut in half a 20x30cms (190grams) writing pad. Halves were placed on the grass and the height taken from the ground to the pad using a measuring rule. Sward height was taken during each emptying occasion. For analysis, sward height will be assessed as a continuous variable.

Canopy cover

The computation of canopy cover percentage was based on aerial photographs. Tree crown photographs were taken using a Nikon digital SLR camera fitted with a fisheye lens about 30 cm above every trap perpendicular to the ground. The photographs collected were then analysed using the CIMES FISHEYE software following methods by Walter and Gonsamo (2014). Canopy cover was also assessed as a continuous variable in the statistical analysis.