The effect of hay mulching on soil temperature and the abundance and diversity of soil-dwelling arthropods in potato fi elds

The average soil temperature was signifi cantly lower in plots covered with hay. Hay mulching, however, did not increase signifi cantly the total number of soil micro-arthropods collected using a soil pin trap and a soil sampler. Signifi cant increases in the number of individuals was recorded only for certain groups. 64% of all the arthropods collected using soil pin traps were collected in hay-covered plots and 36% in control plots. This increase was statistically signifi cant for the orders Entomobryomorpha and Poduromorpha of the subclass Collembola and surface-dwelling (epigeic) Coleoptera. For the samples collected using the soil sampler, 57% of the specimens were collected from hay-covered plots and 43% from control plots. As for the pin traps this was refl ected in differences in the numbers of the orders Entomobrhyomorpha and Poduromorpha of the subclass Collembola and Pauropoda, collected in the treated and control plots. We conclude that mulching affected the different arthropod groups differently.


The signifi cance of potato production
Worldwide 19-20 million hectares of potatoes (Solanum tuberosum L.) are cultivated with an average yield of around 16 t/ha. In Hungary, 30 thousand hectares of potatoes are planted with an average yield of 16-20 t/ha while the area is 41.1 and 40.7 thousand hectares in the USA and the Netherlands, respectively (Izsáki, 2005). The yield potential of potato is mainly determined by production technology and choosing the variety most suitable for the cultivation conditions (Sárközi, 2002). There are about 3000 varieties of potato worldwide and there were 43 varieties in Hungary in 2001(Gólya, 2004. The yearly potato consumption is 50 -60 kg per capita in Hungary (Sárközi, 2002). varied between 7-8 (slightly alkaline), salt concentration (EC 2.5 mS/cm) was low under 0.02% and soil was strongly calcareous (CaCO 3 % = 23,4%). The saturation percentage (54 SP) indicated clay loam soil types at the study site (Gavlak et al., 2003). The soil sample depth was 0-8 cm.

Data collection and sorting of soil-dwelling arthropods
We installed soil pin traps, called EDAPHOLOG probes (Dombos, 2016), in each plot. The Edapholog System, including probes, was developed for the automated detection and collection of soil-dwelling arthropods and measurement of soil temperature in the upper 10-20 cm of soil (Dombos, 2016). Soil pin traps were installed randomly in the upper soil layer following Gedeon et al. (2012). Built-in thermometers provided hourly data on soil temperature (Dombos & Bánszegi, 2013). This device collects soil-dwelling micro-arthropods from the top 20 cm layer of soil into a 50 ml tube fi lled with 30 ml of 98% ethanol, and simultaneously measures soil temperature at this depth. In addition to this method of collection, two soil samples per plot (24 samples altogether) were collected at random from the top 8 cm layer of soil using a soil sampler. The height and diameter of the soil sampling cylinders were 8 cm, and each sample had a volume of 402 cm 3 . Soil pin traps were installed on 22.vi.2015 and were collected from control potato plots on 08.x.2015. The collection tubes of the instruments were changed once during the study period, on 01.viii.2015. Soil sampling using soil samplers took place on two occasions: on 30.x.2014 and 09.x.2015. The soil-dwelling arthropods were extracted from the collected samples using a Tullgren funnel. The funnels of the apparatus were equipped with sample holders of 19 × 19 × 6 cm in dimension at the upper opening; the length of the funnel was 30 cm. Due to the desiccation of the samples, the animals move from the funnel into the collecting vessel. Samples collected from the same plots were combined and placed in the same Tullgren funnel, and the arthropods were collected in 120 ml vessels. 60 ml of 70% ethanol was added to the vessels, and the liquid that evaporated was replaced regularly. The soil samples were left in the apparatus for 14 days to achieve complete desiccation, and allow suffi cient time for the arthropods to escape from the drying soil. Arthropods in each ethanol sample were sorted under a stereomicroscope and then stored in 70% ethanol for later use. We only classifi ed the arthropods to the order or family level.

Data analysis
For the comparison of abundance of specimens recorded in treated and control areas, the Mann-Whitney U test was applied (Newcombe, 2006). The software SPSS 20 and Shannon's diversity index were used (Kautz et al., 2006). For both methods, the Shannon indices calculated for the treated and control areas were compared using Paired Sample T-tests. Normality assumptions were tested using the Shapiro-Wilk test (Shapiro & Wilk, 1965). We found that normality was not violated in either method (soil pin trap: sig = 0.929; soil sampler: sig = 0.22). In the case of the samples collected using the soil sampler, as only one sample was collected each year, the data for the two years were pooled (Table 2, Fig. 2). For the comparison of the average daily soil temperatures in mulched and control plots, the related samples Wilcoxon signed rank test (Siegel, 1956) was used.

RESULTS
64% of the 7241 arthropod specimens collected using soil pin traps were collected in hay-covered plots and 36% in control plots. A total of 8321 arthropods were collected using the soil samplers. 57% of these animals were collected from mulched plots and 43% from control plots. There cultivation greatly reduces the numbers of species and abundance of Symphyla (Menta et al., 2010). Collembola are quite sensitive to soil solarization, therefore mulching may provide more favourable conditions for light-sensitive soil-dwelling arthropods due to its light shielding effect . Mulching may also have an effect on food chains. Certain soil fungi have become significantly more abundant in corn fi elds treated with white clover living mulch, which causes a signifi cant increase in the abundance of mites and springtails (Nakamoto & Tsukamoto, 2006). Living mulch provides organic matter for soil microbes and enhances moisture conservation, shelter and provides a food source for the local mesofauna, including certain mites and springtails (Kaneda et al., 2012). In soils covered by sawdust there is an increase in the abundance of decomposer organisms, such as isopods, diplopods and soil-inhabiting Coleoptera (Wardle et al., 1999). Springtails and mites play an important role in the breakdown of dead organic matter (Dányi & Traser, 2007). However, wood chips and grape marc scattered among rows of vines reduces the abundance of Collembola, most likely due to the increase in diversity of predators in the covered areas (Addison et al., 2013). The variation in the species numbers of certain arthropods in soils covered by organic plant residues may also depend on the season (Thomson & Hoffman, 2007). Among the predatory arthropods occurring in potato plots, the numbers of centipedes are not increased signifi cantly by mulching with organic materials, but their overall species and specimen numbers were originally low due to soil disturbance (Dudás et al., 2013).
The purpose of this research was to investigate the effect of hay mulching on the temperature of the top soil layer (1-20 cm) and on various groups of soil-dwelling microarthropods. We also wanted to examine and compare the effi ciency of two methods of determining abundance and estimating species diversity; traditional soil samplers and novel soil pin traps.

Study site
The study was conducted in potato plots located in highland and domestic garden environments in a town called Hidegkút (Veszprém County, Western Hungary) (47°00´20.6˝N, 17°83´05.9˝E) near a mixed community forest, Quercetum robori-cerris (Bartha et al., 1995). The variety of potato (Solanum tuberosum L.) planted was "Sárpo Mira". The size of the area studied was 168 m 2 , including tramlines, and was divided into 12 plots of 3 × 4 m. The number of treatments was 2 (hay-mulched and control plots) with 6 replicates of each. After the beginning of shoot growth, the treated potato plots were covered with a 10 cm thick layer of natural meadow hay coming from an area adjacent to the experimental site. Hay was previously dried and stored in an indoor area before it was used in the fi eld. The potatoes were planted in early May in 2014 and 2015, and were harvested at the end of autumn in both years. Potato was planted at the experimental site in previous years. The experimental area was surrounded by a Convolvulo-Agropyretum repentis (Kovács, 1995) association. Dominant plant species were perennial weeds, such as Convolvulus arvensis L., and grasses, such as Agropyron repens L. The soil in the area is brown forest soil. The soil pH were no signifi cant difference between the treated and untreated plots in terms of the total number of soil-dwelling arthropods collected by either the soil pin traps (p = 0.132) or soil samplers (p = 0.671). However, hay mulching caused a signifi cant difference in the numbers of specimen of the following arthropod groups collected by the soil pin traps: the orders Entomobryomorpha and Poduromorpha of the subclass Collembola and epigeic Coleoptera (living above the soil surface) (Table 1). There were also no significant differences in the abundances recorded in treated and untreated plots in the case of samples collected using soil samplers (p = 0.671). There were signifi cant differences in the numbers of specimens of the orders Entomobryomorpha and Poduromorpha of the subclass Collembola, and Pauropoda collected from treated and control plots using soil samplers ( Table 2).
The diversity of arthropod groups collected varied depending on the treatment: based on the results of the soil pin traps (Fig. 1), there is no signifi cant difference between the diversities recorded in the treated and untreated plots (p = 0.819), whereas based on the data obtained using the soil sampler (Fig. 2), there was a greater diversity in the treated plots (p = 0.032).
On the basis of the soil temperatures measured by soil pin traps, the average daily soil temperature in hay mulched plots was lower than that in control plots (Fig.  3). The mulching, in particular, reduces the maximum temperatures rather than evenly reducing the average temperature. The average difference between the average daily soil temperatures in treated and control plots was 1.155°C; the mean daily temperature of treated plots was signifi cantly lower (p < 0.001).

DISCUSSION
The characteristics of soil affect the occurrence and frequency of soil insects. Decomposing organisms, such as Podorumorpha springtails, are abundant in crops growing in soils with a high carbon and nitrogen content. In crop cultures where ryegrass is used as green weight the abundance of herbivorous Symphypleona increases (Crotty et    . As a result of mulching with compost, decomposing and predatory soil-dwelling arthropods proliferate but herbivores do not (Mathews et al., 2002). Among the arthropods collected using soil pin traps, Collembola were the most abundant (4032 individuals), especially the Entomobryomorpha (3775 individuals); followed by Acari (1762 individuals). However, in the case of samples collected using soil samplers this ratio was reversed, with the Acari the most abundant (6072 individuals), followed by Collembola (1373 individuals); but within this subclass, Entomobryomorpha were again the most abundant (919 individuals). Different sampling methods resulted in different ratios of arthropod groups. The advantage of soil pin traps was that they continuously collect soil arthropods. Among soil-dwelling arthropods in English fi elds, Collembola are the most abundant, with the next most abundant group, mites, not far behind; the abundance of both these arthropods is approximately 10,000 individuals/m 2 (Dunger, 1983). Despite the relatively high numbers collected, we could not detect a signifi cant difference between the mite populations in treated and untreated plots (Tables 1, 2). Mulching the soil surface with plant cuttings has no signifi cant effect on the numbers of armoured mites, however, it results in an increase in the numbers of predatory mites and springtails (Wang et al., 2011). In the case of the application some organic mulches (wheat straw, bark and wood chips, dried municipal sludge) there is an increase in the abundance soil microfauna, and as a result of more fungi, there are more fungal-grazing pygmephorid and tarsonemid mites and Collembolans (Whitford et al., 1989). As a result mulching our potato plots with hay, the abundance of the orders Entomobryomorpha and Poduromorpha of the subclass Collembola increased signifi cantly but not that of Acari (Tables 1, 2). Collembola also feed on plant-pathogenic fungi (Hopkin, 1997), therefore, they benefi t agriculture. Various organic mulches may affect the abundance of different soil-dwelling arthropod groups in diverse ways; pine bark mulch has a more positive effect on fl y larvae inhabiting the soil surface presumably due to its slower decomposition (Gill et al., 2011). There was no signifi cant increase in the numbers of fl ies recorded following mulching with hay in this study (Tables 1, 2). Abundance of Coleoptera increases as a result of mulching with cowpea cuttings (Gill & McSorley, 2012a). Based on the results of the soil pin traps, the abundance of epigeic Coleoptera (living above the soil surface) was also increased by hay mulching in this study. Other sampling methods, such as Berlese funnels, differ in their effi ciency for sorting soil Arthropods. For example, Berlese funnels can only process small amounts of litter  and the   number of taxa collected by this method is lower than by other sampling methods (Gill & McSorley, 2012b). There were no differences in taxa collected by the two sampling methods used in this study (Tables 1, 2). The variations in the temperature at the soil surface can be substantial both spatially and temporally. The majority of the soildwelling animals can tolerate occasional major changes in soil temperature, but generally are very sensitive to higher temperatures (Dunger, 1983). Therefore, we assume that lower soil temperatures in mulched plots may also have contributed to the increase in the numbers of individuals of some groups of soil-dwelling micro-arthropods in these plots. Soil solarisation using plastic mulch can be used to control weeds (Gill et al., 2009).

CONCLUSIONS
With the exception of certain groups, hay mulching did not result in a widespread and signifi cant increase in the numbers of soil-dwelling micro-arthropods. There were marked differences in the numbers of certain groups collected associated with both the treatment and method of sampling. The effi ciency the sampling techniques used differed, therefore depending on the method of collection, statistically detectable increases in the numbers was only recorded for: Entomobryomorpha and Poduromorpha of the subclass Collembola, Pauropoda and epigeic Coleoptera (living above the soil surface). In terms of diversity of arthropod groups, differences between treated and untreated plots were only detected in the samples collected using a soil sampler, which revealed a greater diversity in the treated plots. Consequently, further studies are needed to verify this fi nding using other methods. On the other hand, the effect of mulching on epigeic groups was only detected using soil pin traps. A detectable reduction in the average soil temperature was recorded in the potato plots that were mulched with hay. Based on the analysis of the samples collected using the two methods described above, it is assumed that soil pin traps are not suitable for sampling certain groups of soil-inhabiting organisms such as Symphyla, Pauropoda and Psocoptera. Our results indicate that hay mulching in potato fi elds results in an increase in the activity of soil-dwelling micro-arthropods, consequently it is recommended as a preferable management practice.