Slugs

The finding of larger slug numbers on MT relative to CT plots in each trial of the present study affirms the proposed hypothesis and accords with the literature. The finding of larger numbers of slugs in systems of reduced- or conservation-tillage, non-inversion tillage and direct-drilled crops when compared with that for mouldboard ploughing followed by different cultivation (CT) methods has been reported for various regions and crops (Hunter Reference Hunter1967; Grant et al. Reference Grant, Yeargan, Pass and Parr1982; Glen et al. Reference Glen, Wiltshire and Milsom1984, Reference Glen, Wiltshire and Milsom1988; Port & Port Reference Port and Port1986; Willson & Eisley Reference Willson and Eisley1992; Kendall et al. Reference Kendall, Chinn, Glenn, Wiltshire, Winstone, Tidboald, Glenn, Greaves and Anderson1995; Symondson et al. Reference Symondson, Glen, Wiltshire, Langdon and Liddell1996; Voss et al. Reference Voss, Ulber and Hoppe1998; Christian et al. Reference Christian, Bacon, Brockie, Glen, Gutteridge and Jenkyn1999; Hammond et al. Reference Hammond, Beck, Smith, Amos, Barker, Moore, Siegrist, Slates and Ward1999; Glen Reference Glen2000). Hunter (Reference Hunter1967) was among the earliest workers to show, by experimentation, that cultivation reduces slug numbers. He concluded that slugs were killed by the mechanical action of implements; even ploughing alone reduced slug numbers. Ploughing alone was also found by Glen et al. (Reference Glen, Milsom and Wiltshire1990) to reduce both slug numbers and biomass by 84 and 81%, respectively, in the upper 100 mm of soil. They also found that ploughing followed by secondary cultivations significantly reduced slug biomass relative to ploughing alone. In the UK, Christian et al. (Reference Christian, Bacon, Brockie, Glen, Gutteridge and Jenkyn1999) found slugs were much more plentiful on direct-drilled winter wheat than on ploughed wheat, and similar findings were recorded by Grant et al. (Reference Grant, Yeargan, Pass and Parr1982) for alfalfa in the USA. The MT system used in the present study involved shallow cultivation without soil inversion. This MT system, which is common with practitioners in Ireland, is broadly similar to that for non-inversion tillage in Europe (Kendall et al. Reference Kendall, Chinn, Glenn, Wiltshire, Winstone, Tidboald, Glenn, Greaves and Anderson1995; Voss et al. Reference Voss, Ulber and Hoppe1998; Rothwell et al. Reference Rothwell, Chaney, Haydock, Cole and Schmidt2005). Slug numbers in areas cultivated by non-inversion tillage are generally intermediate between ploughed plots and no-tilled plots (Kendall et al. Reference Kendall, Chinn, Glenn, Wiltshire, Winstone, Tidboald, Glenn, Greaves and Anderson1995; Symondson et al. Reference Symondson, Glen, Wiltshire, Langdon and Liddell1996) and non-inversion methods of tillage generally have less effect on slug populations than ploughing combined with subsequent cultivations (Glen et al. Reference Glen, Wiltshire and Milsom1984, Reference Glen, Wiltshire, Wilson, Kendall and Symondson1994, Reference Glen, Wiltshire, Walker, Wilson and Shewry1996; Glen & Symondson Reference Glen, Symondson and El Titi2003). Glen & Symondson (Reference Glen, Symondson and El Titi2003) concluded that the greater the number of cultivation operations and the more intensive the method of cultivation the more likely it is that slug numbers will be substantially reduced, and that non-inversion tillage is generally less effective in reducing slug numbers than CT. Tillage with tines and discs was found by Kendall et al. (Reference Kendall, Chinn, Glenn, Wiltshire, Winstone, Tidboald, Glenn, Greaves and Anderson1995) to cause greater suppression of slug populations than a single pass non-inversion tillage system (Dutzi cultivator) with slug numbers consistently higher on direct-drilled plots. Similar findings have been reported for winter oilseed rape (Voss et al. Reference Voss, Ulber and Hoppe1998) with increased slug abundance and activity in non-inversion and zero-tillage relative to conventional ploughing.

Organic matter, such as crop residues, is believed to favour slug populations (Port & Port Reference Port and Port1986; Glen Reference Glen2000). Glen et al. (Reference Glen, Wiltshire and Milsom1984) suggested that slug populations increase in the presence of straw because it provides an additional source of food to the crop which may contribute to slug survival and that straw may also hide slugs from predators. Comparisons of slug numbers in plots where straw residues were left with those where straw was burnt have shown higher slug numbers for straw plots (Glen et al. Reference Glen, Wiltshire and Milsom1984, Reference Glen, Wiltshire and Milsom1988; Christian et al. Reference Christian, Bacon, Brockie, Glen, Gutteridge and Jenkyn1999). However, straw burning has been shown to have little direct effect on slug numbers (Edwards Reference Edwards1977); therefore, the conclusions that straw favours slug populations in the UK investigations seems reasonable. As well as the favouring of slug populations in winter wheat by straw residues, Glen et al. (Reference Glen, Wiltshire, Walker, Wilson and Shewry1996) also found that straw or stubble from the previous cereal crop resulted in higher slug populations in oilseed rape, particularly when incorporation of straw was shallow and where the tillage did not include ploughing. However, at a long-term study site in the UK, which compares more closely with the present study reported, consistently more slugs were recorded in plots where straw was chopped and incorporated compared with plots where straw was baled and removed and only the stubble was incorporated (Kendall et al. Reference Kendall, Chinn, Glenn, Wiltshire, Winstone, Tidboald, Glenn, Greaves and Anderson1995; Glen et al. Reference Glen, Wiltshire, Walker, Wilson and Shewry1996; Symondson et al. Reference Symondson, Glen, Wiltshire, Langdon and Liddell1996). Despite the evidence to-date in the literature, the incorporation of straw either in MT or CT cereal plots in the two trials of the present study, on two differing soil types, and on the same large-scale plots (recommended by Port & Port (Reference Port and Port1986) for obtaining long-term data on slugs) over a period of 9 years did not show an increase in slug numbers resulting from straw incorporation. The latter finding was unexpected and contrary to the hypothesized effect of straw on slug abundance. The barriers between plots in Trial 1 would have greatly reduced or prevented any possible intermixing of slugs between plots, thereby further supporting the finding of an absence of an effect of straw on slugs. The finding of an absence of slug proliferation resulting from straw incorporation is of considerable importance to MT practitioners in Ireland.

It is generally agreed that weather conditions influence slug abundance with damp springs and summers together with mild winters favouring populations (Martin & Kelly Reference Martin and Kelly1986). Moens (Reference Moens and Henderson1989) states that a succession of wet summers combined with mild winters favour slug populations while dry summers and severe winters reduce populations. Wet summers were also noted by Gould (Reference Gould1961) to be associated with serious slug problems in winter wheat in areas with heavy land. Similarly, Rogers-Lewis (Reference Rogers-Lewis1977) found that in wet years, slug damage is often a major problem affecting ware potatoes, particularly on the heavier silt soils. In their modelling of climate effects on slug populations, Choi et al. (Reference Choi, Bohan, Powers, Wiltshire, Glen and Semenov2004) confirmed the association between rainfall and abundance. Glen et al. (Reference Glen, Wiltshire and Milsom1984, Reference Glen, Wiltshire and Milsom1988) found that slugs were few in the upper 100 mm of soil during dry summers, however, numbers rapidly recovered in autumn when the soil became moist again; he thereby concluded that slugs had penetrated deep into the soil in dry periods. Numbers of all species of slugs found in cereal fields fluctuate widely within and between years even when agricultural practices do not change (Port & Port Reference Port and Port1986). Glen (Reference Glen and Henderson1989) states that not all fluctuations in slug numbers can be explained by weather conditions. Glen et al. (Reference Glen, Wiltshire and Milsom1988) found that slug numbers fell sharply even though conditions remained moist and there was no obvious change in husbandry practices, while at other sites slug numbers remained high. In the present study there were significant differences in slug numbers between years. However, there was no evidence that either MT or CT methods of cultivation promoted a build-up in slug numbers. There was no simple regression relationship between rainfall in any year or in the period May–September and slug abundance in November–December. Neither was there any simple relationship between rainfalls in any year with slug numbers in the subsequent year. However, examining a smoothed version of the relationship between cumulative rainfall from June–November and slug numbers (a thin plate spline fitted using SAS Proc TPSpline) and the width of confidence intervals for the fitted smooth showed strong evidence of a non-linear trend. At low levels of cumulative rainfall, there was indication of an increase in slug numbers as rainfall increased but at higher levels of cumulative rainfall there was a clear collapse in numbers. This pattern was robust to adjustments to the period involved and was consistent for each site and the mean of the two sites.

The reason for the difference in slug species composition at the two sites is most probably soil type. Based on the examination of slug damaged wheat seed in the UK, Gould (Reference Gould1961) noted Agriolimax reticulatus (Deroceras reticulatum) was usually the only species found in large numbers. This species was also regarded by Stephenson (Reference Stephenson1975) as the most important in damaging cereals by attacking the seed and grazing on the leaves. Glen (Reference Glen and Henderson1989) states that typically one or more Arion species occur with D. reticulatum. D. reticulatum is also listed as a prominent pest of agriculture and horticulture by Godan (Reference Godan1999) and attacking maize and soybean in the US (Hammond Reference Hammond and Henderson1996; Hammond et al. Reference Hammond, Beck, Smith, Amos, Barker, Moore, Siegrist, Slates and Ward1999). South (Reference South1992) considers D. reticulatum as the most economically important slug in many parts of the world.

Damage to plants

In general, there was a positive though not significant relationship between slug numbers and proportion leaf damage. For example, slug numbers were 2-fold greater in Trial 2 compared with Trial 1, with the corresponding difference in damage to leaves being 2·3-fold. The most notable and unexplainable exception to this trend was in Trial 1 in 2008 when relatively high levels of leaf damage were recorded while slug numbers in traps were relatively low. As might be expected, slug damage to leaves was significantly greater in MT when compared with CT plots, which coincides with slug abundance. In keeping with slug abundance, slug damage was greater, though not significantly so, in no-straw than in straw treated plots of each method of cultivation and at each Trial. The relationship between leaf damage and slug abundance in the present study supports that of Glen et al. (Reference Glen, Spaull, Mowat, Green and Jackson1993), who found that the proportion of seedlings grazed by slugs was correlated with the number of slugs trapped before drilling, at drilling and after crop emergence. In laboratory experiments Stephenson (Reference Stephenson1975) found that while slugs shred cereal leaves, they do not cut-off plants at ground level; in field observations he noted that, in spring, the plants grew away from such damage. Feeding by slugs on the first leaves of cereals to emerge was regarded by Moens (Reference Moens and Henderson1989) to have no impact on plant development since the terminal bud was protected at the base of the first joint.

The most serious damage to winter wheat is the feeding on the seed embryo, which prevents germination (Gould Reference Gould1961; Stephenson Reference Stephenson1975). In caged trials with various crop species and differing D. reticulatum densities, Barratt et al. (Reference Barratt, Byers and Bierlein1994) found that crop yields were inversely related to slug density, seedling damage and mortality during crop establishment. The proportion of wheat seed killed by slugs was found by Glen et al. (Reference Glen, Milsom and Wiltshire1989) to be directly related to the biomass of slugs living in the top 100 mm of soil and was inversely related to the depth of seed and the percentage of fine soil aggregates (<6 mm) in the seed bed. Earlier laboratory investigations by Stephenson (Reference Stephenson1975) showed that deep planting protected seed from damage by slugs. In the field Glen et al. (Reference Glen, Milsom and Wiltshire1989, Reference Glen, Milsom and Wiltshire1990) reported that winter wheat seed sown at 55 mm had 9% slug damage while that sown at 19 mm had 26% damage. Late-sown crops of winter cereals are usually considered to be at greater risk from slug damage than early-sown crops (Port & Port Reference Port and Port1986) because the wetter conditions later in the year promote slug activity and cereal emergence is delayed in cooler sites so that crops are at risk for longer. In addition to enhancing slug numbers, the literature also reports that straw is associated with increased seedling damage. Damage to cereal seedlings was found by Glen et al. (Reference Glen, Wiltshire and Milsom1984) and Smith et al. (Reference Smith, Jordan, Kendall, Glen and Hardcastle1985) to be greatest where straw remained on the soil surface (47%) and was reduced by incorporation (13%) or by burning (5%).

In the present study, slug damage to cereal leaves was widespread and in some years severe, while damage to seed and seedlings was very slight. Slug damage was not related to either ear density prior to harvest or grain yields and is not considered to have impacted on these parameters. These findings in regard to leaf damage support similar findings by Stephenson (Reference Stephenson1975) and Moens (Reference Moens and Henderson1989). The very low level of slug damage to seed and seedlings is due in part to the depth of sowing (40 mm) and to the relatively early date of sowing which would have favoured the reasonably quick establishment of plants thus avoiding the risk of slug damage.

Earthworms

Earthworms were significantly greater in MT compared with CT plots in Trials 1 and 2. This finding agrees with the proposed hypothesis on the effects of cultivation on earthworm abundance and concurs with the literature. Cultivation affects earthworm habitats by destroying established burrow systems and by changing the soil temperature, moisture and food availability (Edwards Reference Edwards and Satchell1983). Ploughing was estimated by Cuendet (Reference Cuendet and Satchell1983) to damage 27–41% and kill 18–28% of all earthworms. Ploughing also exposes earthworms to predation and desiccation (House & Parmelee Reference House and Parmelee1985) and is especially damaging to deep burrowing anecic species, e.g. L. terrestris and A. longa (Kladivko Reference Kladivko2001). The more drastic the type of cultivation and the more operations involved the greater the detrimental effect on the soil fauna (Edwards Reference Edwards1975, Reference Edwards1977). Where drastic forms of soil cultivation are involved, such as that currently used for potato production, together with the mechanical harvesting of the crop, earthworm populations can be virtually eliminated within a single year (Curry et al. Reference Curry, Byrne and Schmidt2002). Many studies, involving various crops and locations, have shown greater earthworm abundance in direct-drilled crops when compared with CT systems (Barnes & Ellis Reference Barnes and Ellis1979; Gerard & Hay Reference Gerard and Hay1979; Edwards & Lofty Reference Edwards and Lofty1982; House & Parmelee Reference House and Parmelee1985; Jordan et al. Reference Jordan, Stecker, Cacnio-Hubbard, Li, Gantzer and Brown1997; Kladivko et al. Reference Kladivko, Akhouri and Weesies1997; Wuest Reference Wuest2001; Peigné et al. Reference Peigné, Cannavaciuolo, Gautronneau, Aveline, Giteau and Cluzeau2009; Umiker et al. Reference Umiker, Johnson-Maynard, Hatten, Eigenbrode and Bosque-Perez2009). However, conservation or reduced cultivations, comparable to MT in the present study, have also been reported to support grater earthworm populations than conventional plough-based cultivation systems (Haukka Reference Haukka1988; Jordan et al. Reference Jordan, Leake and Ogilvy2000; Emmerling Reference Emmerling2001; Hutcheon et al. Reference Hutcheon, Iles and Kendall2001; Wuest Reference Wuest2001; Rothwell et al. Reference Rothwell, Chaney, Haydock, Cole and Schmidt2005). In comparing conservation and plough based tillage systems over a 10-year period in Britain, Jordan et al. (Reference Jordan, Leake and Ogilvy2000) and Hutcheon et al. (Reference Hutcheon, Iles and Kendall2001) found earthworm biomass increased by 36% for the former method. Over the same time period El Titi & Ipach (Reference El Titi and Ipach1989) found that both numbers and biomass of earthworms, in Germany, were up to 6-fold higher in conservation tillage relative to CT. A 6-fold increase in earthworm numbers in conservation tillage relative to CT, over 30 years, was recorded in NW USA by Wuest (Reference Wuest2001). In Australia, Chan & Heenan (Reference Chan and Heenan2006) recorded 6·6 times more earthworms in wheat grown in no-till plots with crop stubble retained relative to wheat in tillage of 3-passes of a ‘scarifier’ and stubble burnt. Intensive rotary cultivation followed by ploughing was found by Bostrom (Reference Bostrom1995) to kill 73–77% of earthworms; however, one year later, he found that the biomass in tilled plots returned to levels seen in undisturbed lucerne leys. He concluded that the large amount of organic matter ploughed under, which served as a food for earthworms together with the supply of cocoons in the soil, allowed endogeic earthworms to make a fast recovery. The decline in earthworm populations under continuous arable cropping was regarded by Edwards & Lofty (Reference Edwards and Lofty1982) as being due to a combination of mechanical damage to earthworms during cultivation and to decreased availability of food resulting from accelerated disappearance of organic matter.

Overall, earthworm numbers increased in both MT and CT plots and in both straw and no-straw plots of each tillage system over the course of the present study. Total earthworm numbers and biomass as well as numbers of A. chlorotica, Lumbricus species and juvenile epilobous and tanylobous worms for the 9 years, in each trial, showed significant differences between years. This trend was consistent and significant in both trials over the first 7 years of the study. The reversal in this trend in Trial 2 in the last 2 years was most probably due to unusually severe and prolonged periods of frost prior to sampling of earthworms, which had to be delayed until late February/early March rather than the targeted sampling period of November–December. In Trial 1, earthworm sampling occurred on 3 November in the second last year but not until 20 January 2010 for the 2009/10 season. The unexpected finding of the present study was the increase in earthworm numbers in CT plots with succeeding years, particularly for the no-straw plots, albeit from somewhat low initial populations. This finding supports the trend observed by Edwards (Reference Edwards and Appelhof1981) that continuous winter wheat may lead to increased populations of earthworms based on comparisons with wheat grown in rotation with other crops. The maximum earthworm densities for the no-straw CT plots in Trials 1 and 2 were 174/m² and 452/m², respectively. Earlier, at the site of Trial 1 but separated from it by 620 m, Schmidt et al. (Reference Schmidt, Curry, Hackett, Purvis and Clements2001) recorded 55 earthworms/m² after 4 years of CT wheat on a loam to sandy loam soil while in direct drilled wheat having a clover under-storey they recorded 483 worms/m². The latter cropping system was found by Schmidt et al. (Reference Schmidt, Clements and Donaldson2003) to support an earthworm density of 1097/m² at a UK site. The highest earthworm densities recorded in the present study may well have been the maximum each trial could sustain under the cultivation and husbandry employed. However, the reduced earthworm densities in the last 2 years of the study are attributed to unfavourable weather conditions which may have resulted in earthworms retreating to lower soil depths, entering enforced quiescence or possible mortality due to extreme frosts (down to –15 °C).

Over the 9 years there were significantly more earthworms in straw-treated plots, for the combined cultivation systems, than in those without straw in Trials 1 and 2. In CT plots, there were significantly more worms over the 9 years in straw-treated plots than in no-straw plots in each trial. Earthworm biomass showed similar trends to that for earthworm numbers. Earthworm abundance was generally increased where straw and crop residues were retained on the soil surface. In comparing the effects of chopping and spreading straw with straw burning in direct-drilled crops, Barnes & Ellis (Reference Barnes and Ellis1979) found no difference in the overall earthworm population. However, they found significantly larger numbers of deep burrowing (anecic) species on the chopped and spread areas whereas on burnt areas numbers of the horizontal burrower (endogeic) A. chlorotica were significantly greater. Crop residues on the soil surface in no-till systems are available as a food supply to earthworms for a longer period of time than if incorporated with a tillage implement (Kladivko Reference Kladivko2001). Additionally, Kladivko (Reference Kladivko2001) considers that surface residues serve as a mulch giving protection from inclement weather, thus greatly favouring earthworms. These surface residues, which contribute to higher soil organic carbon, are particularly important to deep burrowing species since these feed primarily on residues at the soil surface, pulling them into their permanent burrows (Edwards & Lofty Reference Edwards and Lofty1980; Kladivko Reference Kladivko2001; Umiker et al. Reference Umiker, Johnson-Maynard, Hatten, Eigenbrode and Bosque-Perez2009). Direct seeding of crops into living mulch or bicrop (e.g. clover) can greatly increase numbers of some earthworm species (Schmidt et al. Reference Schmidt, Clements and Donaldson2003; Pelosi et al. Reference Pelosi, Bertrand and Roger-Estrade2009). Retaining stubble as surface mulch for successive years was found by Mele & Carter (Reference Mele and Carter1999) to support consistently higher earthworm densities while the incorporation of stubble reduced densities by 53% compared with stubble mulch across texturally different soils. The absence of a positive response to stubble incorporation reported by Mele & Carter (Reference Mele and Carter1999) may possibly have been due to the mechanical damage to earthworms by the rotary hoe used to incorporate the stubble and working at a depth of 100 mm. Soil incorporation of straw was found by Fraser & Piercy (Reference Fraser and Piercy1998) to result in greater earthworm abundance in the 4th year of their investigations, while the removal of straw residues was reported by Haukka (Reference Haukka1988) to reduce earthworm populations by 50%.

A reduced beneficial effect of straw on earthworm abundance might have been anticipated within the CT plots since the straw and indeed the stubble from the pervious crop was widely dispersed within the plough-cultivation soil layer and as such was likely to be less available to earthworms than that in MT plots in which a considerable portion of straw and stubble remained on the soil surface. The positive impact of straw on earthworm abundance within each system of cultivation and at each trial site was rather similar thus increasing the robustness of the finding.

Species

Over the 9 years a total of 11 earthworm species were recorded of which 7 species were common to both trials. Endogeic species comprised almost 84% of the earthworm fauna at Trial 1 and 92% at Trial 2. The percent anecic species at Trial 1 was almost 10% and at Trial 2 was 1·2%. A. chlorotica (endogeic), the numerically dominant species in the present trials, was more plentiful in MT than CT and in straw than no-straw plots. These differences were significant for cultivation treatment in Trial 2 and for straw treatment in Trial 1. Over the 9 years and across the two cultivation systems, Lumbricus species (anecic and epigeic) in each trial were significantly more abundant in MT than in CT and in straw relative to no-straw. However, within cultivation systems numbers of Lumbricus worms did not differ significantly between straw and no-straw treatments. The species composition and relative abundance of species recorded in the present study were broadly similar to those recorded in other Irish studies (Curry Reference Curry1976; Schmidt & Curry Reference Schmidt and Curry2001; Schmidt et al. Reference Schmidt, Curry, Hackett, Purvis and Clements2001). The finding of fewer A. chlorotica in CT relative to MT concurs with the findings of Edwards & Lofty (Reference Edwards and Lofty1982) who found 3·4 times more of this species in direct drilled than ploughed plots but also found these differences were much less than that recorded for the anecic species L. terrestris and A. longa. This differential effect of cultivation on different earthworm species has also been recorded by other workers. Gerard & Hay (Reference Gerard and Hay1979) found that the impact of cultivations was less detrimental to A. chlorotica than all other species. Chan (Reference Chan2001) considered endogeic species can increase under tillage especially if there is an increased food supply. This view is also supported by Rothwell et al. (Reference Rothwell, Chaney, Haydock, Cole and Schmidt2005) who concluded endogeic species could withstand CT practices. In contrast with the above, Bostrom (Reference Bostrom1995) found that while A. caliginosa was initially killed by cultivation, one year later the biomass and abundance of this species was greater in ploughed treatments than in undisturbed leys and attributed these findings to initially low earthworm abundance and increased food supply provided by ploughing down crop residues. Additionally, Pelosi et al. (Reference Pelosi, Bertrand and Roger-Estrade2009) found 2·8 times more endogeic earthworms in conventional tilled and organic wheat than in wheat direct drilled into a living mulch.

There is extensive evidence of the detrimental impact of tillage on anecic earthworms which is mostly reported for comparisons between direct-drilled crops and those sown by CT (Gerard & Hay Reference Gerard and Hay1979; Kladivko et al. Reference Kladivko, Akhouri and Weesies1997; Kladivko Reference Kladivko2001; Chan Reference Chan2001; Rothwell et al. Reference Rothwell, Chaney, Haydock, Cole and Schmidt2005; Peigné et al. Reference Peigné, Cannavaciuolo, Gautronneau, Aveline, Giteau and Cluzeau2009). Edwards & Lofty (Reference Edwards and Lofty1982) reported that after 8 years L. terrestris and A. longa were 17·5 times more numerous in direct drilled relative to ploughed plots while at a second site the difference was 37·3-fold. Edwards & Lofty (Reference Edwards and Lofty1982) also found that populations of these species were intermediate between those for direct drilling and ploughing for less intensive cultivations such as chisel-plough.

Straw incorporation in each trial of the present study resulted in greater numbers of A. chlorotica, with the effect being significant in Trial 1. In contrast Barnes & Ellis (Reference Barnes and Ellis1979) found A. chlorotica were more abundant in areas where straw was burnt than where it was chopped and spread. Also, Edwards & Lofty (Reference Edwards, Lofty and Gossbard1979) found A. chlorotica increased significantly in numbers as a result of all straw removal treatments and was most in plots where straw was spread and burnt. The incorporation of straw, in the present study, had a significant beneficial effect on Lumbricus (anecic and epigeic) numbers for combined but not within cultivation systems. The chopping and spreading of straw on the soil surface was also found by Barnes & Ellis (Reference Barnes and Ellis1979) to significantly increase the number of deep burrowing species, relative to straw removal. It is possible some conditions favouring L. terrestris and A. longa may be unfavourable to A. chlorotica and A. caliginosa; Edwards & Lofty (Reference Edwards and Lofty1982) reported a significant negative correlation between populations of L. terrestris and A. longa and between those of A. chlorotica and A. caliginosa.

Beetles

The aggregate numbers of 45 carabid species captured, over a 12-week period, for each of 5 years did not show a significant effect for either method of cultivation or straw treatment but did show a significant difference between years. Trapping of beetles commenced in the 5th year of the present study and this should have contributed to a stabilization of the effects of treatments. Further enhancement of this stabilization was also likely to result from the physical separation of treatments by the use of plastic barriers which were erected four years before the commencement of beetle trapping. P. melanarius, the most common and largest beetle encountered, was significantly more numerous in MT relative to CT plots in 2 of the 5 years as well as for the aggregate data for the 5 years. Small species (5 Bembidion + T. quadristriatus) were significantly more abundant in CT relative to MT plots in 3 of the 5 years as well as for aggregate numbers over the 5 years. Straw incorporation had no effect on carabid numbers within either method of cultivation or for combined cultivations. The results support the proposed hypothesis regarding effects of cultivation on carabid species but not that for the effect of straw on beetles.

In general, conservation/non-inversion tillage supports greater arthropod communities including carabid beetles than CT (Edwards Reference Edwards1977; House & All Reference House and All1981; Kromp Reference Kromp1999; Witmer et al. Reference Witmer, Hough–Goldstein and Pesek2003; Thorbek & Bilde Reference Thorbek and Bilde2004; Menalled et al. Reference Menalled, Smith, Dauer and Fox2007). The effects of cultivations on carabid beetles have been outlined by Kromp (Reference Kromp1999), Holland & Luff (Reference Holland and Luff2000), Hance (Reference Hance and Holland2002) and summarized by Holland (Reference Holland2004). Using emergence traps, which avoid the known limitations of pitfall traps (Luff Reference Luff1975; Adis Reference Adis1979; Sunderland et al. Reference Sunderland, De Snoo, Dinter, Hance, Helenius, Jepson, Kromp, Lys, Samu, Sotherton, Toft and Ulber1995) Purvis & Fadl (Reference Purvis and Fadl1996, Reference Purvis and Fadl2002) and Holland & Reynolds (Reference Holland and Reynolds2003) showed ploughing affected survival in many species. Holland & Reynolds (Reference Holland and Reynolds2003) quantified these reductions for some species but also found that almost 90% more Notiophilus biguttatus emerged from ploughed than from unploughed ground. Despite the evidence of the general impact of cultivations on carabids some species are found in greater abundance on ploughed than on unploughed land. Holland & Luff (Reference Holland and Luff2000) list 47 taxa of which 20 favoured ploughed, 21 favoured unploughed and six were neutral to either form of tillage. Adults of small carabid species (body length <6 mm), such as Bembidion species and T. quadristriatus were more prevalent in ploughed plots whereas six of the eight taxa associated with non-inversion tillage, including P. melanarius, were larger (mostly >10 mm) while yet other species were equally distributed between these cultivation systems. Holland & Luff (Reference Holland and Luff2000) also found this species-specific response to cultivation method was reasonably consistent in most years. In contrast, Baguette & Hance (Reference Baguette and Hance1997) considered smaller species, e.g. Bembidion species, were more vulnerable to ploughing and were favoured by conservation tillage. The findings of Hatten et al. (Reference Hatten, Bosque–Perez, Labonte, Guy and Eigenbrode2007) of a relationship between beetle size/species and tillage regime in the USA is similar to that reported for the UK by Kendall et al. (Reference Kendall, Chinn, Glenn, Wiltshire, Winstone, Tidboald, Glenn, Greaves and Anderson1995). Hatten et al. (Reference Hatten, Bosque–Perez, Labonte, Guy and Eigenbrode2007) found larger species (>14 mm) more common in no-tillage, small species (<7 mm) in CT and intermediate sizes (7–14 mm) occurring equally between tillage systems. There is evidence to indicate the association of carabids with different systems of cultivation is not size-specific. Based on the catches of beetles in emergence traps, Holland & Reynolds (Reference Holland and Reynolds2003) concluded size was not an important determinant for carabids to survive cultivation. The results of the present study, which shows the most abundant and large species P. melanarius was more numerous on MT than CT while the smaller Bembidion species and T. quadristriatus were more numerous on CT than on MT plots, clearly supports the findings of Kendall et al. (Reference Kendall, Chinn, Glenn, Wiltshire, Winstone, Tidboald, Glenn, Greaves and Anderson1995) and Hatten et al. (Reference Hatten, Bosque–Perez, Labonte, Guy and Eigenbrode2007). The carabid species assemblage present in fields is determined by multiple factors (House & Stinner Reference House and Stinner1983) the more important being soil type and moisture status and crop type (Holland & Luff Reference Holland and Luff2000) as well as field size, soil pH, organic matter and sand content of soil (Irmler Reference Irmler2003). Since most of these factors were common to all treatments and as plots were separated by plastic barriers and sampled over a 12-week period coinciding with the peak emergence period for many carabids (Holland & Reynolds Reference Holland and Reynolds2003), the results of the present study are considered to provide a reasonably accurate reflection of the impact of treatments on carabid beetles.

The incorporation of straw did not have any effect on abundance of either total beetles (45 species), the large beetle P. melanarius or the smaller five Bembidion species+T. quadristriatus between MT and CT tillage either for individual years or for the aggregate beetle numbers for the 5 years. While the pitfall traps used have known limitations (Luff Reference Luff1975; Adis Reference Adis1979; Sunderland et al. Reference Sunderland, De Snoo, Dinter, Hance, Helenius, Jepson, Kromp, Lys, Samu, Sotherton, Toft and Ulber1995) this method is commonly used for carabid investigations and in the present study their limitations were likely to have been mitigated by the use of physical barriers between plots. This finding differs from that of Symondson et al. (Reference Symondson, Glen, Wiltshire, Langdon and Liddell1996) who, also using pitfall traps, reported greater numbers of P. melanarius where straw was incorporated by non-inversion tillage. Symondson et al. (Reference Symondson, Glen, Wiltshire, Langdon and Liddell1996) showed P. melanarius to be an important predator of slugs, finding that 84% of beetles contained slug remains; they also found this beetle to be more plentiful in areas of high slug biomass while Bohan et al. (Reference Bohan, Bohan, Glen, Symondson, Wiltshire and Hughes2000) confirmed a significant spatial aggregation between P. melanarius and slugs. The carabid species encountered in the present study include predators of the aphid pests of cereals, including those vectoring virus diseases of cereals (Sunderland Reference Sunderland1975; Sunderland et al. Reference Sunderland, Crook, Stacey and Fuller1987; Kennedy Reference Kennedy1994; Kennedy & Connery Reference Kennedy and Connery2001).

It is concluded that MT increases slug numbers, slug damage and earthworm numbers, particularly of deep-burrowing species, relative to CT cultivations. Straw incorporation does increase earthworm numbers but not slugs, slug damage or carabid beetles. The large variation in slug populations between years was affected by factors other than husbandry including method of cultivation and straw incorporation. One factor identified was cumulative rainfall in the period June to November. Slug damage to cereal leaves does not affect ear density or grain yield in either MT or CT crops when sown at 40 mm and before 18 September.