The continuing homogenization of freshwater ecosystems facilitates the establishment and spread of aquatic invasive species (Baur and Schmidlin 2007), frequently cited as a major cause of biodiversity loss and disturbance of food webs (Andersen et al. 2004; Henseler et al. 2021; Carvalho et al. 2022). The recent surge in reports of non-native species (Baur and Schmidlin 2007; Rocha et al. 2023) indicates an urgent need for quantification of the ecological impact of existing, emerging, and potentially invasive species to facilitate the focus of management efforts on the most prominent invaders (Dick et al. 2014). Invasive species are often more effective foragers than analogous native species (Dick et al. 2002; Dick et al. 2013; Dick et al. 2014). Comparative functional response (FR), i.e., predator efficiency in prey utilization (Holling 1959), represents a fundamental tool in the study of invasion ecology (Alexander et al. 2014; Xu et al. 2016; Thorp et al. 2018; Gebauer et al. 2019). Comparative FR has been used to analyse the per capita foraging efficiency among predators of different origins or characteristics (Laverty et al. 2017; Gebauer et al. 2018; Howard et al. 2018; Gebauer et al. 2019). The higher predation rate of an invader is crucial but not the only predictor of negative impact (Parker et al. 1999). In addition to the per capita effect, factors such as abundance, reproduction parameters (Parker et al. 1999; Dickey et al. 2018), and environmental variables (Dick et al. 2014) contribute to overall predator ecological impact.

Although robust comparative FR includes a range of environmental variables such as dissolved oxygen concentration (Dickey et al. 2021), temperature (Xu et al. 2016; Gebauer et al. 2018), and habitat complexity (Gebauer et al. 2018), invader abundance has been overlooked. Dick et al. (2017) recently developed an Impact Potential (IP) scale that combines abundance with the per capita impact of a predator, while Relative Impact Potential (RIP) enables comparing IP of two or more species/size classes/environmental variables. Moreover, both metrics preserve robustness to be fit for predators or plants (Dickey et al. 2020). Although abundance is a meaningful measure, the population size structure, i.e., relative abundance of size classes, might expand the informative value of the IP assessment. Fish body size influences foraging capacity (Mittelbach 1981; De Roos et al. 2003), bioenergy needs (Weitz and Levin 2006), and food preferences (Mittelbach 1981). Generally, per capita foraging efficiency increases with body size (Rudolf 2012), as large predators can ingest larger prey, for higher energy gain and can select prey from a broader diet niche (Werner 1974; Paradis et al. 1996). Smaller predators can display higher foraging efficiency towards small motile prey (Aljetlawi et al. 2004) that can be difficult to handle or provide insufficient energy for a large predator (Costa 2009). Additionally, small predators are generally more abundant (Cohen et al. 2003; Woodward and Hildrew 2005). Hence including population structure in the IP calculation might provide a more accurate prediction of novel predator-prey interaction dynamics and their consequences in freshwater communities, essential information for invasion management (Olden and Poff 2004).

The highly invasive benthic fish Neogobius melanostomus has proliferated in many European rivers and the Great Lakes of North America (Kornis et al. 2012). Its establishment is accompanied by competition for food (French and Jude 2001; Ustups et al. 2016; Herlevi et al. 2018) and space with native species (Greenberg et al. 1995; Dubs and Corkum 1996) and by a decrease in, or change of, entire macrozoobenthic assemblages (Lederer et al. 2008; Mikl et al. 2017; Pennuto et al. 2018). Neogobius melanostomus is highly fecund (Charlebois et al. 1997) and lays eggs several times a year. Hence, multiple age classes are usually present in an invaded site. Although small specimens prefer zooplankton (Števove and Kováč 2016; Olson and Janssen 2017), larger individuals are voracious and adaptable feeders, and the dietary composition usually reflects prey available in the locality (Dashinov and Uzunova 2020), with some prey overlap among size classes (Števove and Kováč 2016).

Comparing the FR of different size classes/population structure of N. melanostomus can enable accurate prediction of invasive predator impact on their potential prey in the colonized regions. Since we assume that both the parameters and type of functional response could be size-depending in fish, similarly in the African clawed frog (Thorp et al. 2018). Indeed, quantifying the impact potential of an invasive species population based on the per capita consumption rate of only one size class can result in a misestimation of impact potential concerning the prey utilization and abundance of other size classes. We aim to uncover predator body size relevance in comparative functional response and regarding that more reliable calculation of an impact potential on an example of invasive fish species N. melanostomus.

On 14.7.2020, the field survey and N. melanostomus collection took place in the Elbe River (Czech Republic; 50.6540922N, 14.0439108E) using a backpack pulsed-DC electrofishing unit (FEG 1500, EFKO, Leutkirch, Germany) and zig-zag wading in the near-shore water (length = 100 m, width = 4 m, area = 400 m²) with stony bottom no blocked by any nets, which is heavily populated by N. melanostomus. For assessment of predator IP, the captured fish were divided into three size classes: small [wet weight (WW) = 2.25–4.24 g], medium (WW = 4.25–6.24 g), and large (WW = 6.25–10.25 g). The abundance of each predator size class per square meter of the near-shore water to a distance 4 m from the bank was calculated.

The predators were transported to the Institute of Aquaculture and Protection of Waters, Faculty of Fisheries and Protection of Waters, University of South Bohemia in České Budějovice, Czech Republic. Before the experiment, N. melanostomus were held in a recirculating aquaculture system (1600 l) for 28 days of acclimatization. Predators were fed ad-libitum with frozen Chironomus sp. larvae.

We used Asellus aquaticus as prey, a principal crustacean dietary item of N. melanostomus in the field (Vašek et al. 2014). Large crustaceans are a key component of diet for several size classes of N. melanostomus (Števove and Kováč 2016). Asellus aquaticus were collected with hand nets from submerged vegetation around banks in Kyselá voda stream (49.0195475N, 14.4640344E) and kept in a 200 l tank with aeration and organic substrate collected from the same site.

The type of predator functional response curve is one of the important predictors of the stability of a prey population utilized by a predator (Miller et al. 2006; Dick et al. 2014). As in previous research, N. melanostomus showed type II FR with respect to A. aquaticus (Laverty et al. 2017; Franta et al. 2021) and other prey types (Gebauer et al. 2018; Gebauer et al. 2019; Paton et al. 2019). In this study, all three size classes showed type II FR. Type II FR is common (Leeuwen et al. 2007) and indicates high ability to utilize prey even at low densities and thus the potential of a predator to destabilize prey populations (Miller et al. 2006; Dick et al. 2014). Nevertheless, slight differences among FR curves were evident in the present study. Small N. melanostomus showed a lower magnitude of the FR curve, hence lower interaction strength with A. aquaticus compared to medium and large individuals. Generally, larger predators can forage for larger prey over a wider area with a low risk of predation (Mittelbach 1981; Brown and Maurer 1989; Paradis et al. 1996; Costa 2009). Therefore, they can be less cautious in prey selection (Werner 1974) and show higher interaction strength. On the other hand, larger fish may be more prey-specialized (Jacob et al. 2011) or exhibit diet shifts with ontogeny. Although N. melanostomus of ~60 mm can ingest bivalves (Parker et al. 1999), they constitute a small portion of the N. melanostomus diet when prey of higher energy value is available. Neogobius melanostomus diet composition reflects the prey community abundance composition (Pennuto et al. 2010; Raby et al. 2010) rather than the highest energy yield per bite.

Mouth size is an important prey-limiting factor. Predation usually follows optimal foraging theory, i.e., a trade-off of energy gain with cost of prey capture and handling (Tytler and Calow 2012). Therefore, larger predators often avoid smaller prey, as energy cost exceeds gain (Costa 2009). On the other hand, small predators exhibit higher capture rates and lower handling time when attacking small prey (Persson 1987). Neogobius melanostomus, however, seems to be a voracious predator able to capture large prey that is only partially consumed (Roje et al. 2021). The attack rate of small fish did not significantly differ from that of large in the present study, indicating boldness against larger prey despite the longer handling time. Thorp et al. (2018) reported the highest observed attack rate in the smallest specimens of the frog Xenopus laevis, which could indicate higher efficiency of small frogs in utilizing offered prey. Larger body size can lead to a decrease in predator efficiency in catching small prey because of lower agility (Persson 1987). Also, prey that is too small (Hyatt 1979) or represents a low energy gain (Costa 2009) may be ignored by a large predator. Our assessment of attack rate did not confirm lower foraging efficacy of large N. melanostomus within the target size range. Thorp et al. (2018) observed lowest handling time in medium-sized frogs. The medium class of N. melanostomus in our study also showed the lowest handling time, but with no significant difference from that of large individuals.

Generally, handling time increases with the size of prey (Hoyle and Keast 1987) until size exceeds predator gape size (Kislalioglu and Gibson 1976; Aljetlawi et al. 2004), but handling small prey can be difficult for a large predator (Costa 2009). Our results show that the size range of offered A. aquaticus was suitable for both medium and large N. melanostomus size classes, while small fish showed longest handling time. A possible explanation is the high ratio of A. aquaticus size to small N. melanostomus gape size, as described in Micropterus salmoides (Kislalioglu and Gibson 1976). Factors such as digestion capacity (Brown and Maurer 1989; Aljetlawi et al. 2004; Li et al. 2017) and satiation level (Kislalioglu and Gibson 1976; Li et al. 2017) can also influence handling time. Digestion is a component of handling time (Woodward and Hildrew 2002), and high handling time can reflect the limited digestion capacity of smaller predators (Li et al. 2017). Generally, gastric evacuation rate, i.e., the quantity of food evacuated per body weight of predator per time unit, is similar among predator sizes. However, a large predator can ingest more food (Brown and Maurer 1989), reaching satiation more rapidly, and handling time declines with increasing satiation (Kislalioglu and Gibson 1976). Although N. melanostomus were starved to ensure sufficient time for evacuating the gut, we could not measure the speed of satiation of a particular size class. Specifying the above mentioned factors influencing handling time in N. melanostomus would require techniques exceeding the scope of this study.

Field abundance provides a numerical estimate of predator response (Dickey et al. 2021) and, when combined with FR data, substantially extends predictive accuracy of ecological impact (Dick et al. 2017). Although invasive species commonly reach high abundance, this alone cannot be sufficient to assess predation impact (Laverty et al. 2017). Laverty et al. (2017) presented a much lower FR (per-capita effect) of the non-native topmouth gudgeon Pseudorasbora parva than the native analogous European bitterling Rhodeus amarus. However, P. parva reaches several times the field abundance of R. amarus, which explains its high ecological impact. Moreover, IP can include various proxies for abundance, including fecundity, lifespan, or propagule pressure (Dickey et al. 2018) that are closely related to population size and fluctuation. Neogobius melanostomus shows different IP and RIP in response to prey type, oxygen conditions, and the commonly used field abundance data (Laverty et al. 2017; Dickey et al. 2021). In our study, the most abundant size class showed the lowest maximum feeding rate. Although IP of individual size classes differed significantly, the final comparison of RIP showed that body size plays only a minor role in assessing N. melanostomus total impact. However, depending on field abundance and population structure of established or recently invaded sites (Taraborelli et al. 2010; Brandner et al. 2018), predator size can be important factor in total impact.

Neogobius melanostomus can be considered a voracious invasive predator of A. aquaticus across various size classes. There are significant body-size differences in the magnitude of N. melanostomus interaction with prey with respect to per capita foraging efficiency, which can affect its IP. Abundance as well as the size structure of a N. melanostomus population may fluctuate with time post-colonization (Rakauskas et al. 2013; Denys et al. 2015; Brandner et al. 2018), season (Blair et al. 2018), and habitat (Uspenskiy et al. 2021). Although basing analysis on a limited size range can overestimate the IP of N. melanostomus because of high per capita consumption rate in small individuals combined with their high abundance in the population and vice versa (Dick et al. 2017), the size-group comparison of IP in our study does not show a major difference. Hence, for simplicity and rapid calculation of N. melanostomusIP, there may be no need to consider the maximum feeding rates of individual size classes. Nevertheless, our results show size to play an important role in the per capita effect and trophic interactions in the food web.

We encourage considering population structure for future quantification of invasive predator consumption pressure. Although only a minor effect of body size was determined in N. melanostomus on precise calculation of impact potential. We assume that increasing numerical differences among size classes can fundamentally increase the importance of body size in IP calculation because of the size dependency of the per capita consumption rate. Additionally, the role of body size can be different in other invasive species or even higher in a predator-prey system with multiple prey species carried on in the field where not a subset, but an entire population is included in the calculation. Unfortunately, the FRs from the field are only occasionally published (Zimmermann et al. 2015; Sorial-Diaz et al. 2018). However, FR for estimating a predator consumption rate on the population level in a lab or the field brings new challenges, including the importance of many other species-specific features or behaviour which incorporation should be solved in the future.

This study was financially supported by the Ministry of Education, Youth and Sports of the Czech Republic – projects CENAKVA (LM2018099) and by the Grant Agency of the University of South Bohemia, project no. 065/2022/Z.

P.F., R.G. and B.D. research conceptualization, P.F., R.G. sample design and methodology, P.F., R.G., N.Z.S. investigation and data collection, L.V., P.F. data analysis and interpretation, B.D. funding provision and P.F. and R.G. writing – original draft; B.D. and R.G. writing – review.