Introduction

Expansion is a consequence of the dispersal process, and the dynamics of expansion shape species distribution and community composition1,2. In light of current global changes, understanding range expansion mechanisms is an increasingly important challenge for ecology and conservation. A growing number of species are expanding their distribution range—invasive species are rapidly spreading into new areas, and native species are following their shifting climatic envelope2—but the boundary between invasive and native range expansion is not always clear.

Native species occur within their natural ranges (past or present) and in their dispersal potential3. In contrast, non-native species have been human-mediated, either accidentally or deliberately, outside of their natural ranges, regardless of their eventual impact on native ecosystems4. While “non-native” can be neutral in terms of their effect on the environment, “alien” or “invasive species” represent those non-native species that have a demonstrable impact, significantly modifying or disrupting the ecosystems that they colonise and, thus, being characterised as invaders5. Invasive species must pass through four stages to become invasive: (1) transport, (2) establishment, (3) spread, and (4) impact6. Furthermore, becoming “invasive” (both widespread and locally dominant), is always the culmination of a process that climaxes in an increase in abundance7. Due to the lack of natural enemies and appropriate conditions in the territories that they have newly occupied, invaders can introduce new pathogens, increase competition for resources (the novel weapons hypothesis6) and, ultimately, reduce the occurrence of native species to a level that threatens the indigenous species with extinction5. The responses of native species to an invasion depend critically on invasive species’ abundance and trophic position. As the abundance of an invasive species increases, native population sizes decline nonlinearly by an average of 20%, and community diversity declines linearly by 25%, with significantly larger effects on species evenness and diversity than on species richness (reviewed by9).

In the Anthropocene, alien species are no longer the only category of biological organisms establishing themselves and rapidly spreading beyond their historical boundaries10. Changes in the distribution of native plants11,12, birds13,14,15 and even small mammals16,17,18 have become increasingly common in recent decades. Nackley et al.10 pointed out that invasions by native species are a global phenomenon, and the expansion of native organisms into adjacent communities can emulate functional and structural changes and impact the biodiversity associated with invasions by alien species. However, Tong et al.19 argued against treating the expansion of native species in the same way as the arrival and expansion of alien species. Expanding native species can move together with their associated organisms (e.g., herbivores or parasites) and, thus, do not benefit from an enemy release like invasive alien species, making the effects of range expansion on local communities typically more neutral and milder than those of alien species invasion. However, Thompson and Davis20 suggested that, with continual global changes in nutrients, climate and disturbance regimes, all species can be considered to be inhabiting novel environments, and, therefore, distinctions between native and non-native species are becoming even less ecologically meaningful. Since biological invasions are an extension of normal colonisation processes, the terminology used should reflect this fact and not be based on the species’ geographic origins21,22,23. Therefore, the impact of species on the environment should be principally assessed24.

The striped field mouse (Apodemus agrarius) is an epidemiologically significant small mammal species25,26 that is widely distributed over the entire Palaearctic region, from Central Europe to the Korean Peninsula and the Russian Far East27. It spread to Europe in the first half of the Holocene, together with a specific parasite, Hystrichopsylla orientalis orientalis, a flea subspecies28. While the striped field mouse retreated from parts of Western and Central Europe, the parasite has persisted and survived thereafter on other hosts until the present day29. Over the past decade, there have been several documented expansion waves of the striped field mouse’s range into the Russian Far East30, Central and Eastern Europe31,32,33 and directly into Slovakia34. However, paleontological findings35,36 and the subrecent diet of owls37,38 indicate the occurrence of the striped field mouse in parts of Central Europe that are either supposedly not colonised or have been already re-colonised by them. But recent evidence of the species' occurrence in Western39,40 and Central Europe32,41,42 points to the considerable dynamics at the edge of their distribution, where their spread into new territory was subsequently followed by a retreat from the colonised land over a period of several decades.

In 2010, the occurrence of the striped field mouse was first documented in our study region8; previous studies of small mammals and the diet of owls in this area43,44,45,46,47,48,49,50,51,52 have been made for more than a century, but no evidence of the striped field mouse was recorded. The closest known populations of the species in Slovakia were approximately 120 km east of the study area from 2006 to 200753,54. Furthermore, closer new populations at another location south of our study area near the Danube river were confirmed earlier in Austria and Hungary55 (Fig. 1). Ambros et al.8 notes that no subfossil or fossil evidence of the striped field mouse has been found in the study area before. Authors hypothesised about the possible transport of striped field mice with the transit of municipal and industrial waste as a possible explanation for their presence. The fact is that in a few years, the species had expanded to a large part of south-western Slovakia56,57, and the expansion is continuing58. The enormous scale and speed of their expansion do not support the hypothesis of anthropochory, and indications and evidence supporting this supposition are lacking. However, the origin of this spreading population still remains unexplained. The change in the mean annual and summer temperatures in Central Europe has been identified as a possible driver for the current oscillation in the species’ western limit of distribution in Austria, Czechia, Hungary and Slovakia55. On the other hand, deforestation, conversion of steppe into farmland, and the construction of irrigation channels have been identified as drivers of striped field mouse range expansion in North Asia30.

Figure 1
figure 1

Study sites distributed in three study areas (green, blue, orange) and control sites (black circle), with the first striped field mouse (SFM) records in 2010 in south-western Slovakia (red triangles) and the closest known populations of striped field mice in the south of Slovakia in 2007 (red squares) and in 100 × 100 km grids in Austria and Hungary in 2013.

Full size image

Factors including the theoretical framework described above, current knowledge, the long-term presence of the striped field mouse's flea documented in the study area before the onset of the current spread59,60,61, the still unexplained origin of the spreading population and, most importantly, the fact that the fundamental and first stage of the invasion process is the transport of a non-indigenous species6, lead us to conclude that the current expansion of the striped field mouse in the study region is a range expansion of a natural species. Despite several documented changes in the range of the striped field mouse in Central and Eastern Europe, the impact of this expansion on native small mammals has never been studied. This study replicates intensive and comprehensive research conducted in the past on small mammals in the study area before the striped field mouse’s expansion, and it analyses the impact of the expansion on their native populations and communities.

This study aims to (i) analyse the impact of the striped field mouse's expansion on the native population and small mammal communities, and (ii) confront documented stages of the striped field mouse’s expansion with stages of invasion biology.

Material and methods

Study area

The research was conducted at 20 study sites divided into three study areas: the first study area was located along the original flow of the Nitra River and consisted of seven study sites; the second study area consisted of six study sites along the Paríž channel; the third area was along the original flow of the Žitava River, and it had seven study sites (Fig. 1). Both rivers in the study area originally meandered before they were straightened in the early twentieth century and channels dug to drain the floodplain, leaving behind distributaries and oxbow lakes today. All the study sites were situated adjacent to wetland habitats. The traps themselves were set at the edge of wetland habitats with agricultural habitats. The entire study area is classified as the northern Pannonian Plain and Hronská pahorkatina highlands. The mean annual temperature is 9.9 °C, with January the coldest month (mean temperature − 2.1 °C) and July the warmest (+ 20.5 °C). The study area is among the driest parts of Slovakia62, with an average annual precipitation of between 550 and 600 mm63.

Study design

Before-After Control-Impact (BACI) design was used to identify the impact of striped field mouse expansion to native communities. BACI designs are commonly used to evaluate impacts from natural perturbations64 and management actions65,66,67, as well as for a wide variety of smaller-scale field experiments68. Impact/treatment effects for BACI designs are often estimated with a significant treatment (Control–Impact) × time (Before–After) interaction, indicating that the experimental treatment has truly had an effect on the impact sites. Considering the interaction term in this manner allows treatment impacts to be distinguished from the background time effects shared by all sites, and from the background differences between control and treatment sites69. Therefore, we used three types of data:

1st trapping period: small mammal native community data

Data about the original communities of small mammals before the expansion of the striped field mouse’s range came from research conducted in 1981–1983, 1990 and 2000–2006. The data consist of 772 individuals from 10 species that were caught at 20 study sites (during 20 sessions) divided into three study areas during different seasons (Supplementary S1). The plotted rarefaction curve shows species diversity accumulating in all study areas (Supplementary S2).

2nd trapping period: Recent small mammal community data

Data were gathered in 2019–2020 after the striped field mouse’s expansion with the same methodology, same catching efforts at the same study sites and the same seasonality as data from the 1st trapping period (Supplementary S1). The data consist of 868 small mammals from 11 species which were captured at 20 locations (during 20 sessions). The communities indicated an increase and saturation of species diversity at all three study areas after the striped field mouse’s expansion (Supplementary S2).

Control sampling

Data were gathered during two time periods, the 1st and 2nd control periods, at seven control sites. In both periods, small mammals were trapped at the time before the confirmed occurrence of striped field mouse at the control sites (Table 1).

Table 1 List of trapping at control sites with the year of first striped field mouse capture.
Full size table

Six control sites (no. 22–27) were situated in the study areas, and one (no. 14) was outside of all the study areas (Fig. 1).

This design was used due to the lack of localities (with the same or similar environmental conditions and the same native species pool) in which the striped field mice was absent at the time of this study 2019–202058. Our research thus does not have the classical concept of BACI design (after-control in a similar time as impact-control). This type of design can still help us compare changes in the composition of paired stable and disturbed communities over time.

The average time interval between trapping periods was 13.3 years ± 5.9 SD. The criteria of the control sites were the same: season, methodology of trapping, and species diversity accumulation (Supplementary S2). The data consist of 514 small mammals from 10 species, which were used in two control periods: 203 in the 1st control period and 311 in the 2nd control period (Supplementary S1).

At all control sites, more trapping sessions were carried out which showed the absence of striped field mice (none were captured). These negative sessions did not meet the methodological criteria (season, types of traps, sampling, etc.) and could not be included in further analysis (Supplementary S1). However, these data supported the later colonisation of the control sites and the absence of striped mice at the control sites during the control captures (Table 1).

Small mammal trapping

Small mammals were captured in 50 box-type single live-traps (75 × 95 × 180 mm)70 placed in a line (with a 10 m gap) and left exposed over two or three nights at each study site. The traps were baited with a mixture of cereals, apples and meal worms. Traps were checked twice a day, in the morning and evening. All the individuals caught in 1st trapping period were then euthanised with isoflurane inhalation or CO2 (lethal dose in a closed chamber). During the 2nd trapping period the following species, which together represent only 1.37% of the total captured specimens, were trapped, sexed, aged, marked and then released: pannonian root vole (Alexandromys oeconomus mehelyi), steppe mouse (Mus spicilegus), lesser white-toothed shrew (Crocidura suaveoelens), bi-coloured white-toothed shrew (Crocidura leucodon) and Miller´s water shrew (Neomys anomalus). Only their first capture (without recaptures) during the session was used in the following analyses. Other trapped small mammals were euthanised in situ with isoflurane inhalation (lethal dose in a closed chamber), in compliance with Slovakian national law (https://www.epi.sk/zz/2012-377), placed separately in a secure container and transported within two hours to a laboratory where they were identified using morphological techniques, sexed, aged, and weighed. The tissues (lung, liver, kidney, heart, stomach and part of the tail) were collected and stored at -80ºC for subsequent virological, parasitological and molecular analysis71,72,73.

Data exploration and statistical analysis

Species with minimal occurrence (< 3) in the whole dataset—namely, the steppe mouse (two occurrences), bi-coloured white-toothed shrew (two occurrences) and Miller’s water shrew (two occurrences)—were not included in the statistical analysis. Prior to analysis, data were explored for residual effects of different sampling efforts by rarefaction on basic Hill numbers; q0 = richness, q1 = exponential Shannon index, q2 = inverse Simpson index (Supplementary S2).

Community distribution and dynamics

To analyse the types of community distribution, we used rank abundance curves (RAC; also called as dominance–diversity curves) with all five models supported by the vegan R package (Broken stick, Preemption, Lognormal, Zipf and Mandelbrot). The best model for each study area and time period (before and after striped field mouse expansion) was determined by the Akaike information criterion (AIC) for the best description of the observed pattern. In general, for a smaller number of samples (sample size/number of estimated parameters < 40), it is recommended to use second-order AIC (AICc—corrected)74. Since the tool used (vegan::radfit) does not offer AICc, it was calculated manually for each model75,76. For a closer look at community changes before and after expansion, five aspects of rank abundance curves—richness, evenness, rank, species gain and species loss—were calculated.

Community metric changes

Due to the slight uncertainty regarding the q0 (species richness) of the third area shown during rarefaction analysis (Supplementary S2) and the general deficiency of q0 in diversity analyses77, we decided to use the more common species–sensitive q2 index for further analysis to meet our goal.

Based on the RAC results, we decided to include the Berger-Parker (d) dominance index to better measure differences in community evenness influenced by the presence of the striped field mouse. The Berger-Parker index has a close relationship with evenness and refers to disproportionately abundant species. While common indices—such as Shannon, inverse Simpson and evenness—increase with improving community composition, an increase in the d index indicates the instability of the community due to the most dominant species. Species evenness and dominance measures may also be related to the net primary productivity of a system, invasion susceptibility and local extinction patterns78.

Since the range of both indices did not include limit values, differences in indices between areas and time periods (before and after expansion) were tested with linear mixed models with Gaussian error distribution, where pairs of the same localities were used as random factors (paired). Due to singularity problems in restricted maximum-likelihood models (e.g., the variance of a random variable can be miscalculated), Bayesian linear mixed models were used. All models were fitted with the “Bayesian Regression Models using Stan” (brms) R package79 with four chains, where each chain had a warm-up of 2000 iterations and then 4000 samples. We confirmed algorithm convergence with visual checks and the Rhat statistic. We chose informative priors that promoted the shrinkage of effects towards zero, including N(0, 2) prior for q2 and N(0, 0.5) prior for the d index. The importance of interaction was evaluated by comparing it to models without interaction, using leave-one-out cross-validation scores (LOOCV).

Impact of striped field mouse’s dominance

The relationship between the dominance of the striped field mouse and q2 and d was described using simple linear regression. To test the unimodal rather than the linear response, we used second-degree polynomials. Maximum (for q2) and minimum (for d) values of the fitted final polynomial model were computed with a golden section search, which is an optimisation technique for finding an extremum (minimum or maximum) of a unimodal function inside a specified interval.

Community composition changes

The compositional community variation between the control and the study areas at different times was analysed using multivariate general linear models with negative-binomial error distribution and a log link function using the mvabund R package80. The significance of the individual factors was obtained by applying the PIT-trap “model-free bootstrap” sampling method81 to 999 permutations tested with likelihood-ratio test (LRT) statistics. For further multiple comparisons, post hoc analysis via a free step-down resampling procedure with Holm adjusted p-values was used.

To visualise the multivariate relationship, non-metric multidimensional scaling (NMDS) was used with the Bray Curtis dissimilarity index. Changes in community composition and the domination of the striped field mouse in the north-west direction were tested using penalised regression splines with a generalised additive model (GAM) with 10 dimensions and a thin plate regression spline smoothing basis, through the vegan package82(Oksanen et al., 2020). All analyses were performed in the R environment83.

Ethics statement

All methods were performed in accordance with the relevant guidelines and regulations of the institution or practice at which the studies were conducted. The research and methods were approved by the Ethics Committee of the Constantine the Philosopher University in Nitra. All trapping and subsequent handling of the small mammals was performed with the approval of the Ministry of Environment of the Slovak Republic in accordance with permission no. 4850/2019–6.3, dated 25th April 2019. The research was conducted by virtue of the appointment of Dr. Michal Ambros (employed by the State Nature Protection of the Slovak Republic) as mapping coordinator for important European small mammal species. This study follows the recommendations in the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines 2.084.

Results

The striped field mouse, after it had expanded its range, had a 100% frequency of occurrence at all of the 20 study sites. Prior analysis of community distribution types by RAC shows significant differences in rank change (χ2 = 9.08, df = 3, p = 0.028) when the striped field mouse became one of the most abundant species after its expansion and replaced the originally dominant bank vole (Myodes glareolus) (Fig. 2). In the first study area, the striped field mouse became the third-most abundant species. We also recorded a species re-ordering in the next position, which was previously held by the common shrew (Sorex araneus). The common shrew’s population fell after the striped field mouse’s expansion at all study areas except for the control sample, where it became the most abundant species in the control event. In contrast, the wood mouse shifted to a higher position in rank, but only became markedly more abundant in the first study area. Prior analyses showed no significant differences in other RAC aspects, such as species curve comparison, species richness comparisons, evenness comparisons, species gains and species losses (Supplementary S3).

Figure 2
figure 2

Rank abundance curves (RAC) of species in each study area before and after the striped field mouse’s expansion and in control sites during 1st and 2nd control periods. The lines represent the best-fit models of the abundance–rank relationship; mgl = Myodes glareolus, sar = Sorex araneus, asy = Apodemus sylvaticus, afl = Apodemus flavicollis, aur = Apodemus uralensis, mar = Microtus arvalis, moe = Microtus oeconomus mehelyi, smi = Sorex minutus, msu = Microtus subterraneus, mmi = Microtus minutus, csu = Crocidura suaveolens. Equations in the upper left-hand corner represent best-fit models of RACs.

Full size image

The results also show significant interaction between the study areas and community metrics caused by the striped field mouse’s expansion (for q2, Δelpd = − 4.4 with Δse = 2.2, and for the Berger–Parker index (d), Δelpd = 4 with Δse = 2.4). In the case of diversity (Fig. 3a) and evenness disturbance (Fig. 3b), contradictory effects were noticed between the first and third study areas. While the third study area showed a decrease after expansion, both metrics increased in the first study area. The second study area and control sample showed no significant differences (Supplementary S4).

Figure 3
figure 3

Expected values of posterior predictive distribution for the a) q2 and b) Berger-Parker index, before and after the striped field mouse’s expansion into each study area. A lower Berger-Parker index indicates an increase in evenness. Whiskers represent 95% CI of mean estimate.

Full size image

A comprehensive view of the relationship between the striped field mouse's dominance and community metrics is evident in a polynomial regression that shows the significant polynomial, unimodal relationship of q2 (F2,17 = 10.05; p = 0.001) and the Berger-Parker index (F2,17 = 22.83; p < 0.001). Both models show an initial increase in community diversity (Fig. 4a) and evenness (Fig. 4b), with these metrics subsequently deteriorating as the striped field mouse’s dominance rises. Further exploration set a negative impact break-even point for striped field mouse’s dominance at 22% and 28%, respectively.

Figure 4
figure 4

Second-degree polynomial relationship of the striped field mouse’s dominance for: (a) q2 (inverse Simpson index), and (b) Berger–Parker index. Grey areas represent 95% CI. The dashed line shows the results from an optimised search of the minimum and maximum values of the functions. The dashed vertical line illustrates the break-even point for the negative impact of the striped field mouse’s dominance on diversity metrics of small mammal communities.

Full size image

The similarity of small mammal communities was significantly affected by the striped field mouse’s expansion (MGLM: LRT = 57.45, p < 0.001), by the study area (MGLM: LRT = 102.59, p < 0.001) and by their interaction (MGLM: LRT = 62.60, p = 0.015) (Fig. 5a). While all three study areas indicate a high similarity before the expansion (MGLM: LRT = 52.72, p = 0.091), the post-expansion communities show similarity depending on the area (MGLM: LRT = 102.38, p < 0.001). Post hoc testing of data after expansionthen shows the differences between the control and the first area (MGLM: LRT = 45.27 padj = 0.015); control and second area (MGLM: LRT = 62.58, p = 0.003) and control and third area (MGLM: LTR = 44.07, p = 0.018).

Figure 5
figure 5

Non-metric multidimensional scaling (NMDS; stress = 0.194) of (a) small mammal community differences before and after the striped field mouse’s expansion (central figure = all study areas and control sites together, figures in corners = each study area and control separately) and (b) the effect of latitude and dominance on study sites and study area distribution; colours = study areas and control sites; solid line = study areas before striped field mouse expansion, dashed line = study areas after striped field mouse expansion; circle size = dominance of striped field mouse, colour of surface = changing latitude.

Full size image

The generalised additive model indicates a significant gradient in community composition changes (F1.5, 9 = 0.65, p = 0.026) and in the domination of the striped field mouse (F2.9, 3.6 = 7.06, p = 0.002) in the area towards the north-west (Fig. 5b).

Discussion

Establishment

Less than a decade after the first record of the striped field mouse’s expansion into south-western Slovakia8, the species had colonised all study sites. This finding corresponds with the strong ability of striped field mouse individuals to disperse and spread over a distance of 400–600 m, occasionally surpassing one kilometre85,86. Dudich29 estimated the speed of the species’ spread in southern Slovakia at between 25 and 30 kms over a 10–12 year period in the 1970s and 1980s, or three kilometres per year. The high degree of dispersion and success in colonisation are both characteristic of all invasive animal species, ranging from birds87,88 to larger mammals89,90, as well as invasive small mammals, such as the house mouse (Mus m. domesticus)91 and rat species (Rattus rattus, Rattus axulans and Rattus norvegicus)92,93,94. Ireland’s bank vole invasion began in the first decades of the nineteenth century and is still happening today, with the expansion rate estimated at 2.5 km per year95. The greater white-toothed shrew (Crocidura russula) has also expanded its range in Ireland since 2007 at an estimated rate of 0.5–14.1 km per year depending on the landscape96. However, detailed literature describing native small mammal range expansions is weak. Only Jareño et al.17 have recorded a rapidly expanded range (approx. 50,000 km2) for the common vole (Microtus arvalis) in Spain’s agricultural land over the past 40 years. Borrowing from invasion ecology theory, one stage in the invasion process is the individuals who are spreading having to establish a self-sustaining population within their new non-native range6. An established or naturalised non-native population may then grow in abundance and expand its geographic range. Both the occurrence of striped field mice in the diet of owls recorded from the Late and Early Holocene throughout the sixteenth century36,38,97,98, and changes in the range of occurrence during the last decades (see the introduction), describe a dynamic expansion and subsequent contraction of the striped field mouse's range or re-expansion of its distribution in Western and Central Europe in the past. However, no signs of this species’ range contracting in south-western Slovakia were observed until 2022. On the contrary, the species is still slowly penetrating the foothills and mountain valleys of the Carpathian Mountains, north of the study sites58.

Impact on native population

Both Ambros et al.8 and Dudich29 expressed concerns about interspecific competition as the striped field mouse penetrated into the native populations of small mammals. In the newly colonised environment, the striped field mouse became either the most abundant or third-most abundant species. Our RAC results show the change in community level since the striped field mouse expanded its range and re-ordered the species, with a negative impact mainly on the bank vole and common shrew. The impact on bank voles was visible mainly in study areas with a higher striped field mouse dominance. Conversely, the abundance of bank voles rose where the striped field mouse was less dominant. While the abundance of the common shrew decreased in all study areas after the striped field mouse’s expansion, an abundance of wood mice increased only in the study area in which the striped field mouse was less dominant. The first indications of the negative impact of striped field mouse on the bank vole and common shrew had already been drawn five years after it expanded its range into south-western Slovakia57. Density-dependent competition is common in animal species99 and an important factor regulating population growth100,101. The growth rate of bank vole populations was negatively related to increasing densities of field voles (Microtus agrestis) in the increase phase of the vole cycle102, and, contrarily, its density grew twice in the absence of competitors103. The presence of the bank vole’s dominant competitors decreased females’ survival, their number, and the size of their territories104,105. Aggressiveness as a consequence of the competition between striped field mice and bank voles for available burrows has been directly observed106. Similarly, increasing total rodent numbers in the community negatively affected common shrew population growth rates102. Shrews had different habitat selection in the absence of competitors107, and Zub et al.108 indicated a possible competitive relationship between striped field mice and common shrews. Interspecific competition between shrews and rodents occurs primarily for space rather than for food, as herbivorous rodents and insectivorous shrews mostly subsist on different diets109. However, the high level of food plasticity among striped field mice110,111,112—an animal diet (mainly invertebrates) can constitute up to 30% of its diet113,114—can lead to higher competition as a result of food niche overlapping. Observations of a negative correlation between population sizes of sympatric small mammal species provided early indications that interspecific competition could have consequences on both population size and habitat use115,116,117. Our results on the community level again suggest concordance with invasion ecology, where becoming the dominant species in a colonised environment represents the next stage of invasion7 and competition and other interactions between non-native and native individuals are a part of a host of extrinsic forces determining whether new species are going to persist in the colonised environment6.

Impact of the striped field mouse’s dominance at the community level

The variability of the striped field mouse’s dominance between the study areas led to a different impact on the researched community metrics. While low dominance increased diversity and evenness (higher q2 values and lower Berger-Parker index), the high dominance of the striped field mouse showed a contradictory effect. No significant differences in the control sites indicated that these changes in community metrics are consequences of the striped field mouse’s expansion. Polynomial models confirmed an initial positive impact on community diversity (higher q2 values) and evenness (lower Berger-Parker index), with low striped field mouse dominance alongside a subsequently rapid decrease in diversity (lower q2 values) and evenness (higher Berger-Parker index) as dominance rose. On average, the break-even point of the negative impact on native small mammal communities was at 25% of the striped field mouse’s dominance. Having a negative impact on the native community and its diversity as a whole is one of the main criteria of becoming an invasive species4,5, which has been well documented for a wide spectrum of invaders, ranging from plants118,119, non-mammalian vertebrates120,121,122 and mammals123, to small mammals on islands124,125—though it is less documented for small mammals on continents91,126. Most of these studies have mainly measured the impact on species richness. Bradley et al.9 documented that the negative effect of an invader´s abundance is significantly stronger for evenness and diversity than for richness, which as a conservative measure of community-level changes requires species extinctions or additions to register change. Bradley et al.9 again reviewed that as invaders become more abundant community-level impacts are clearly negative; these statements fully correspond with our results.

In invasion ecology, non-native species can reduce variability between communities over time in a process of homogenisation where a widespread species becomes dominant and replaces other native, often rare species127,128. In our results, communities of small mammals were already showing high levels of similarity before the striped field mouse’s expansion, which was expected due to the similarities between the study sites. Our analysis, thus, suggests that the differences between recent (post-expansion) communities are a consequence of expansion and different levels of the striped field mouse’s dominance across the study areas.

Dominance as a dispersal direction indicator?

Population spread is a consequence of dispersal and population growth. Dispersal moves individuals through space (some of these individuals move into uncolonised territory), and population growth causes the low-density edge populations to increase in density over time129. The invasion front is characterised by lower densities relative to those behind it130. This density gradient moves through space and is persistent over time, so long as the population is spreading129. In a study from 2016, we suggested a north to north-east dispersal direction of the striped field mouse’s expansion57. Based on the decrease in the striped field mouse’s dominance, however, our current results indicate a north-west expansion direction. Therefore, we assume that the first studied area was colonised after the other two. This hypothesis is supported by Stanko131, who in many cases documented the gradual increase in the dominance of the striped field mouse in colonised communities, and by new occurrences of the striped field mouse found farther north of the first study area58. However, determining the spread’s direction was not the primary goal of this study, and confirmation of this hypothesis would require a different, extended type of study design.

“Invasion” of a native species?

Transport—whether biotic, abiotic or human-mediated—to novel areas is considered the first fundamental stage of the invasion process6. The origin of the striped field mouse population in the study region has not yet been clarified and based on known indications of its life history (see Introduction) we recently considered it a native species expanding its range. However, expansions of native organisms into adjacent communities can emulate the functional and structural changes, as well as the impact on biodiversity, associated with invasions by alien species10. Tong et al.19 argued that expanding native species can move together with their associated organisms and, thus, do not benefit from enemy release like invasive alien species, so the effects of native species’ range expansion on local communities are typically more neutral and milder than alien invasions. Colautti and MacIsaac7 proposed that transport and other invasion stages as an introduction or establishment could be used to model the local spread of more types of potential colonisers, and might be available from a regional species pool including native species. Several studies have suggested that processes affecting local spread and establishment in novel areas may be independent of species origin21,22,23,132. Indeed, nonindigenous species are actually only nonindigenous populations of species; the same “species” that are nonindigenous, naturalised, or invasive in one area are native somewhere else7. Our results suggest several identical stages of invasions and invasion success for the striped field mouse’s expansion in south-western Slovakia. The species was able to establish a self-sustaining population, expand its range, become the dominant unit and, overall, affect the native population and communities. The results are, thus, in accordance with both Thompson et al.132, who argued that the attributes of invasive aliens are shared by native species and not unique, and Nackley et al.10, who argued that, in many respects, expanding native species can be functionally indistinguishable from invading alien species.

In conclusion, over the several years since the striped field mouse expanded into south-western Slovakia it has successfully colonised, and continues to persist in, all study sites. Although neither the causes nor the origins of its current expansion have yet been elucidated, the dominance of the species has nevertheless had a determinedly significant impact on the native community, specifically: (i) the rise of its domination led to a re-ordering of other species, and (ii) the initial positive effect of the expansion on the community metrics, when the striped field mouse exercised low dominance, turned markedly negative after crossing the 25% dominance threshold. The experience gained from a live history of striped field mice, and both the consequences and risks from similar invasions, oblige us to focus future studies on the current range of occurrence and possible retreat from colonised territory, to identify the origin of expanding populations, and to analyse potential threats of this epidemiologically major species.