The natural world is changing rapidly with increased human activity having dramatic consequences for biodiversity and ecosystems (IPBES, 2019). The last 50 years has seen a rapid escalation in the movement of animals and plants, locally and globally, often resulting in species establishing in habitats where they would not otherwise naturally occur (Blackburn et al., 2014; Lucy et al., 2016). The number of species being introduced to regions beyond their natural range has risen steadily over the last two centuries, with increasing trade and travel but also linked to other drivers of environmental change that facilitate biological invasions such as land and sea use change, climate change and pollution (Seebens et al., 2017; IPBES, 2019). Change in invertebrate biodiversity, including the addition of invasive alien species within ecological networks, is a major concern globally (Didham et al., 2005; Mikanowski, 2017). When introduced to a new region, generalist alien species are more likely to become invasive than specialist species; this can result in native specialist species being outcompeted and thereby leading to functional homogenisation (Clavel et al., 2011). It can be difficult to ascertain the effect that the presence of an invasive alien species (IAS) may have on ecosystem function, however in recent years, there has been evidence of the effects on the invaded ecosystem and community (Simberloff et al., 2013) due to invasive alien plants (Liao et al., 2007) and invasive alien aquatic invertebrates (Crawford et al., 2006; Constable and Birkby, 2016; Mathers et al., 2016).

Many coccinellid species provide ecosystem services in the form of pest control (Roy et al., 2012; Honěk et al., 2017). While there have been notable successes of introducing coccinellids for biological control (Dixon et al., 1997; Fowler, 2004), a number of introduced species (e.g. Harmonia axyridis, Coccinella septempunctata) have become established beyond their release sites and have subsequently had negative effects on native coccinellid species (Evans, 2000; Adriaens et al., 2008; Brown et al., 2011a; Roy et al., 2016; Sloggett, 2017); hence considered as IAS. These IAS are often studied in urban (Brown et al., 2011a; Viglášová et al., 2017) or agricultural habitats (Bianchi et al., 2007; Grez et al., 2008; Grez et al., 2014a), however less is known of the ecology of these species in rural habitats ¹ .

In any habitat, a small number of dominant coccinellid species (between two and four) are expected to comprise around 90% of the community (Honěk, 2012). Selyemová et al. (2007) reported a diverse coccinellid community in rural coniferous woodland that was dominated by four species, however, H. axyridis was not established in the region at the time. When investigating overwintering coccinellids in coniferous woodland, Holecová et al. (2018) reported that H. axyridis was not the most abundant coccinellid. In the UK, just as H. axyridis was establishing, Brown et al. (2011a) reported that H. axyridis was largely absent from coniferous woodland. Furthermore, Purse et al. (2014) predicted coniferous woodland could be a refuge for native coccinellids because climate models suggest these habitats are suboptimal for H. axyridis. Vegetation structure of a habitat can also influence coccinellid assemblages. Grassland has been shown to be a refuge for native coccinellid species with very few invasive alien coccinellids recorded in this habitat (Diepenbrock and Finke, 2013). Rural woodland generally consists of a range of tree species and areas of wild herbs/grassland. In Michigan (USA) coccinellid species richness was higher when the habitat was more complex and contained a range of vegetation structures from deciduous trees to grassland and crops (Colunga-Garcia et al., 1997). When non-crop vegetation was added to an agricultural habitat, coccinellid abundance increased (Woltz and Landis, 2014) and intraguild predation between a native coccinellid and H. axyridis decreased (Amaral et al., 2015). Additionally, trees and grassland tend to have a more diverse coccinellid community than crops (Honěk, 2012). Beginning to understand how and why certain habitats are used by particular coccinellid species would be beneficial to understanding the relationship between H. axyridis and native specialist coccinellid species (Sloggett and Majerus, 2000a). Competition for food resources from H. axyridis is one of the reasons why native coccinellids may be negatively affected (Brown et al., 2011a) and so monitoring aphid abundance adds another dimension to studies of coccinellid community dynamics.

Harmonia axyridis is an intraguild predator; i.e. it preys upon other aphid natural enemies including the eggs and larvae of other coccinellid species (Roy et al., 2011; Brown et al., 2011a). It has been suggested that an ongoing increase of H. axyridis numbers may lead to the extinction of some coccinellid species locally (Harmon et al., 2007; Adriaens et al., 2010; Comont et al., 2014; Honěk et al., 2016). Harmon et al. (2007) highlighted the decline of A. bipunctata over a broad geographic range after the invasion of C. septempunctata and H. axyridis in the North America. The dramatic decline of Coccinella novemnotata in North America has also been attributed to a combination of pressures exerted by both C. septempunctata and H. axyridis (Losey et al., 2012b; Tumminello et al., 2015; Ducatti et al., 2017). In Europe there has been a decline in A. bipunctata (Belgium & UK), Adalia decempunctata and Calvia quattuordecimguttata (Czech Republic) since the arrival and establishment of H. axyridis (Brown et al., 2011a; Roy et al., 2012; Honěk et al., 2016; Brown and Roy, 2018). In the UK, distribution of H. axyridis is well known within urban and other anthropogenic habitats and Labrie et al. (2008) reported H. axyridis surviving very cold winters only where people dwell, as this species prefers to over-winter in anthropogenic structures. Additionally, Brown et al. (2011b) reported that H. axyridis tended to oviposit and feed at sites that have human structures nearby. However, much less is known on detailed habitat use of H. axyridis in the wider countryside (Brown et al., 2011a; Brown and Roy, 2018).

Considering the aforementioned studies together with the documented declines of native coccinellids in urban areas strongly correlated with the presence of H. axyridis in the UK (Roy et al., 2012), it is important to understand how native coccinellids are faring in rural areas (Viglášová et al., 2017). The aim of this study was to explore the relationship between the invasive alien H. axyridis and native coccinellid species in rural habitats.

Seventeen species of coccinellid totalling 1,330 individuals were recorded during the study period across three different vegetation gradients (tree, shrub & herb layer) from deciduous and coniferous woodland. Just eight of these coccinellid species were recorded in the herb layer in comparison to 16 species on trees & shrubs in woodland (see Supplementary Material Tables S1, S2). Five species (Myzia oblongoguttata, Myrrha octodecimguttata, Scymnus suturalis, Subcoccinella vigintiquattuorpunctata and Tytthaspis sedecimpunctata) were recorded in coniferous woodland only, while one species (Psyllobora vigintiduopunctata) was recorded in deciduous woodland only. In grassland, four coccinellid species (Exochomus quadripustulatus, S. suturalis, S. vigintiquattuorpunctata and T. sedecimpunctata) were recorded at coniferous sites only. Inclusive of all three vegetation layers, species richness was lower at deciduous (n = 12) than coniferous (n = 16) sites.

In this study, native coccinellids were more abundant than H. axyridis in both coniferous and deciduous woodland. This contrasts with studies on coccinellid assemblages in urban habitats which are dominated by H. axyridis (Brown et al., 2011a; Viglášová et al., 2017). When comparing deciduous and coniferous woodland habitats, the coniferous sites hosted a higher number of native coccinellids than did deciduous sites. Previous studies have suggested that coniferous habitat may act as a barrier to the continuing establishment of H. axyridis, thereby providing a refuge for native coccinellids from H. axyridis (Brown et al., 2011a; Purse et al., 2014), but our results do not support this. It is possible that H. axyridis has adapted to conditions found at coniferous sites in the UK and can overwinter successfully within them. Harmonia axyridis is highly phenotypically plastic and can adapt to available habitats, prey efficiently (Majerus et al., 2006) and climatic extremes (Sloggett and Majerus, 2000b).

When investigating overwintering coccinellid assemblages, Holecová et al. (2018) found that even though H. axyridis was one of the most abundant species on Scots Pine, the majority of the time, either E. quadripustulatus or C. septempunctata made up a larger proportion of overwintering coccinellids, being similar to that observed in this study. Coniferous woodlands experience less extreme temperature variation than deciduous woodlands (Ferrez et al., 2011) and greater overwintering success as a result could explain the increased abundance of native coccinellids at these sites. Additionally, considering the preference H. axyridis has for more sheltered overwintering sites, it is possible that the large area of coniferous plantation in this study provided sufficient shelter to enable this species to overwinter successfully. With a shortage of knowledge on coccinellid assemblages at overwintering sites (Pendleton and Pendleton, 1997-2019; Hodek, 2012; Holecová et al., 2018) investigating the overwintering coccinellid assemblages in coniferous woodland would provide further knowledge of coccinellid behaviour and importantly how climate change may influence coccinellid assemblages in the future. Another feature of the coccinellid life cycle that may be impacted by climate change is diapause. As the onset of diapause varies among coccinellid species (Hodek, 2012) the coccinellid community can vary. Studies on coccinellid diapause tend to focus on one of a small number of species and very little is known in relation to how diapause may impact coccinellid assemblages. Exploring the influence of diapause would lead to a greater understanding of the coccinellid community.

There was a distinct difference in the coccinellid community in relation to vegetation structure (tree, shrub and herb layer) between coniferous and deciduous sites. Within these individual site types, vegetation structure affected both the abundance and distribution of different species. The tree layer in both deciduous and coniferous woodland supported the greatest coccinellid diversity and in both cases differed to the herb layer. At urban sites, Viglášová et al. (2017) found that coccinellid diversity also differed across the different vegetation types that were surveyed, with higher species diversity in trees in comparison to nettle stands. The herb layer in this current study also hosted a different coccinellid community to that of the tree and shrub layer, likely due to the very different food sources available in the herb layer (mildew, plant material, different aphids). Similar findings were reported by Viglášová et al. (2017) for the coccinellid species observed on nettle stands. The differences reported in this current research relate to both site types with a unique coccinellid assemblage at both coniferous and deciduous sites. Grassland specialists dominated the herb layer at coniferous sites (e.g. T. sedecimpunctata & S. vigintiquattuorpunctata), while generalist coccinellids, such C. septempunctata dominated at deciduous sites. Interestingly, Viglášová et al. (2017) reported seasonal differences in how C. septempunctata used different vegetation structures, with greater numbers in nettles in the summer, and higher abundance on trees later in the year. No such seasonal effect was evident within our study, however, C. septempunctata did make use of the different vegetation layers as previously illustrated.

The abundance of H. axyridis, when present, was not influenced by native coccinellid diversity, however, when native coccinellid diversity was higher, the probability of occurrence of H. axyridis was lower. In an agricultural habitat, Grez et al. (2021) reported native coccinellid diversity to be negatively associated with non-native coccinellid abundance. It is widely documented that H. axyridis is often the most abundant coccinellid at urban sites, however this was not the case for rural woodland and grassland habitats in this study. Adalia bipunctata is reported to have high niche overlap with H. axyridis (Sloggett, 2008), however, the extent of co-occurrence between H. axyridis and A. bipunctata is likely to vary between habitats (urban/rural, tree/grass). Recently, Gardiner et al. (2021) reported native coccinellid diversity to be positively associated with forested habitat, while being negatively associated with urbanised habitats.

A negative relationship was observed between native coccinellid and aphid abundance at deciduous sites but at coniferous sites H. axyridis and aphid abundance were positively correlated. The majority of coccinellids recorded at deciduous sites were C. septempunctata, which is a species that is known to tolerate areas with low aphid density (Honěk, 1985). The third and fourth most frequently observed coccinellids at deciduous sites were P. quattuordecimpunctata and A. decempunctata, both of which are also tolerant of low aphid abundance (Honěk, 1985). The relationship between coccinellid abundance and that of aphids, however, is not an easy one to tease apart, in part due to the dynamic nature of both aphid and coccinellid populations. Vandereycken et al. (2013) reported a positive relationship between aphids and coccinellids in a range of crop habitats. Conversely, when investigating coccinellids in urban areas, Viglášová et al. (2017) found the relationship between common coccinellid species and aphid abundance to be non-linear with coccinellid abundance increasing with that of aphids, however when aphid abundance became very high, coccinellid abundance decreased. Furthermore, Brown et al. (2011a) and Brown and Roy (2018) did not find any correlation between H. axyridis or aphidophagous coccinellids and aphid abundance. A network ecology approach may reveal more about the complex relationship between coccinellids and aphids.

The decline of native coccinellids is not solely a consequence of the arrival of H. axyridis and indeed some studies show that native species were in decline prior to its arrival (e.g. A. bipunctata & C. quinquepunctata in Czech Republic, Honěk et al., 2016). Climate change, land use change, intensification of agricultural practices (Honěk et al., 2016) and increased anthropogenic disturbance (Brown and Roy, 2018) may all have contributed to the decline of these species alongside the arrival of H. axyridis. There are suggestions that the initial decline of native species will reverse and that the invasive alien and native populations may stabilise and co-exist (Hentley et al., 2016). Research by Honěk et al. (2016) illustrates just how important long-term population studies are in having baseline data prior to the establishment of an IAS but also in determining how native coccinellid abundance can fluctuate over several years. Long-term studies in a range of habitats are needed to reveal a more complete picture on native coccinellid communities and how they change in the presence of IAS and other drivers of environmental change. Other process may be influencing the differences we observed between coniferous and deciduous sites, perhaps connectivity or adjacent land use or a combination of the two. Including these aspects in future work would enhance our knowledge of coccinellid communities and contribute to informing conservation action.

Coccinellid communities are not often the sole focus of studies and information on their structure tends to come as an add-on to other works (Honěk, 2012). More research needs to be initiated to investigate the coccinellid community as a whole and not just focus on individual species. In this study H. axyridis was less abundant than native coccinellids as a group within both rural woodland and rural grassland. A distinct native coccinellid assemblage was present within all three vegetation layers (tree, shrub and herb layer) sampled. Moving this research further forward, it would be interesting to explore how the coccinellid community varies with seasonality.

With increasing pressures from multiple drivers acting at varying temporal and spatial scales (Bonebrake et al., 2019), it is important to continue research into the dynamics of native coccinellid communities to inform appropriate conservation action. There is increasing evidence that the composition of landscapes can determine community composition (Gardiner et al., 2009), indeed Grez et al., 2014b demonstrated the importance of heterogeneous landscapes for increasing coccinellid diversity and abundance within agricultural systems. Indeed, conserving insect diversity, particularly as the pressure from global environmental change increases, will depend on improving spatial and temporal heterogeneity, including maintaining unique habitats and ensuring functional connectivity, to create landscape mosaics (Samways et al., 2020) that benefits insects including coccinellids. We have demonstrated the importance of different woodland habitats in supporting diverse coccinellid communities. Considering the complex relationship between aphids and generalist coccinellids it is important to further understand the importance of coccinellid diversity, how the communities exist in different landscape contexts and their role in ecosystem functioning.

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

RAF, HER, PMJB contributed to conception and design of the study. RAF carried out field work and performed statistical analysis. RAF wrote the first draft of the manuscript. All authors contributed to revision of the manuscript, read and approved the submitted version.

RAF received a studentship from Anglia Ruskin University to complete this research. HER was supported by the Natural Environment Research Council award number NE/R016429/1 as part of the UK-SCAPE programme Delivering National Capability.

Thank you to the following who granted land access for RAF to undertake surveys: Bedfordshire Cambridgeshire Northamptonshire Wildlife Trust; Forestry Commission; Natural England. Thank you to A. Pickwell, I. Bradley & M. Farrow for assistance during survey work.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.