Extensive use of agrochemicals and habitat loss due to expansion of the agricultural area for intensification of food production have been identified as major drivers of biodiversity loss in agricultural landscapes (Newbold et al., 2015, Dudley and Alexander, 2017). Such biodiversity losses have led to the loss of key ecosystem services (ES) like pollination or natural pest regulation in agriculture (Brook et al., 2008). Sustaining the ES can be crucial in bridging the yield gap between the attainable and the actual yield (Bommarco et al., 2013). Biological control of crop pests delivered by pests’ natural predators is a key ES (Gurr et al., 2003). However, despite years of research, our understanding of the underlying ecological processes that govern biological control, particularly the complex food web interactions that regulate the process, is still incomplete (Bengtsson, 2015). For example, notwithstanding the benefits natural enemies of pests provide, their impact on other non-target arthropod populations has been highlighted (Howarth, 1991). On the other hand, the behavior and efficacy of natural predators in a multi-trophic and multi-species system are still mostly unclear and warrant scrutiny.

Amphibians have long been recognized as potential natural predators of crop pests (Hamer et al., 2004; Loman and Lardner, 2006) being both generalist and opportunist predators (Mahan and Johnson, 2007). Khatiwada et al. (2016) reported a high proportion of rice pests in frogs’ diet and recommended the introduction of frogs for biological pest control in rice fields. A sizeable abundance of amphibians is regarded as being efficient in bringing down the rice pest population (Teng et al., 2016; Fang et al., 2019). However, intensive agriculture has been found to negatively impact frog populations (Davies et al., 2008). Ghosh and Basu (2020) reported a reduced density of amphibians in areas of high agricultural intensification compared to less intensive traditional agricultural areas (for details on the basis for classifying agricultural intensification see, Ghosh and Basu, 2020). A decline in frog population in an intensive agricultural landscape can therefore be expected to negatively impact ES delivery. Apart from this, there could be other factors working in synergy with diversity loss that can disrupt the overall pest regulation by a given frog species. For example, in a multi-species and multi-trophic system within a guild, predation of the arthropod pest predators by super-predator frogs, referred to as “intraguild predation”, may also influence pest regulation (Chailleux et al., 2014).

In this study, we tested the hypothesis that population loss of ES-providing species led to ES reduction using an amphibian species as the model and pest regulation as the targeted ES. We build on the previously reported data on the density of amphibians by Ghosh and Basu (2020) as this reflects the real impact of agricultural intensification on amphibian density. We replicated the densities of a commonly found frog species in the rice fields (Fejervarya sp.) according to the density reported from areas of low and high agricultural intensification (Ghosh and Basu, 2020) in a semi-controlled field experiment to assess the impacts of two different frog densities on the rice pest population. Further, we assessed how trophic interactions between frogs (super-predator) and other natural insect pest enemies might impact the ES. Not only does this study bring out the impact of population losses of key ES providers on ES provision, it also highlights the importance of multi-trophic interaction on ES delivery which hitherto is unexplored especially in relation to amphibians as super-predators.

This study clearly brings out the importance of maintaining a threshold density of beneficial organisms in a natural multi-trophic system, below which not only does the ES provider fail to provide the desirable service it may be also counter acting. This was suggested by the fact that amphibians at lower density (in high-intensive farming) would feed on beneficial arthropods reducing the overall pest regulation. Further, it shows how population loss of key ES providing biodiversity elements can translate into the loss of ecosystem service provisioning using amphibians as model bioregulators of rice pests.

In our study, although frogs consumed pests, their existing field densities (Ghosh and Basu, 2020) were insufficient for significant pest regulation in areas of both low and high agricultural intensification. Even the highest observed frog density (10 frogs/100 m2) in the low-intensive farmlands, appears to fall short of the threshold at which they can significantly regulate pests.

There can be three possible reasons why frogs failed to regulate pests. Firstly, the existing density of frogs even in the low intensification areas is not adequate to control the pests. Teng et al. (2016) reported that an experimental density of 150 frogs/100 m2 could effectively control pests. There is a wealth of literature that shows that traditional or low-intensity farmlands harbor more biodiversity and density than conventional intensive farming areas (Gonthier et al., 2014; Stein-Bachinger et al., 2022). However, whether this higher diversity and density of an ecosystem service provider is sufficient for effective service delivery has hardly been questioned. The current density of amphibians in low-intensive agriculture is evidently below the threshold that can be expected to provide a considerable service to the ecosystem. However, reduction of intensification pressure has been reported to increase frog density (Ghosh and Basu, 2020) and as the present study shows increased frog density does impact pest regulation due to the cascading effect (of increased insect predators) in the multi-trophic system.

The second possible reason for the unaltered pest population at both high and low frog densities (in low and high-intensive farming respectively) could be due to trophic interactions interfering with pest regulation. As the results show arthropod predators were regulated by the frogs at a low density due to intraguild predation, which could have reduced the predation pressure by arthropod predators on pests. In a multi-species and multi-trophic environment with a super-predator like frog, natural pest regulation may remain unrealized. It is well established that generalist predators may have an antagonistic effect on biological control (Perez-Alvarez et al., 2019) especially because of diffused predation pressure (Wells, 2007). However, if frogs feed on crop pests only, the growth rate of beneficial arthropods would decrease as they are robbed off their resource, and eventually, this could further lead to an increase in crop pest abundance. Again, if the harvesting is shifted completely towards arthropod predators there would be an increase in crop pest abundance due to the release of interspecific interaction from the pest predator population. Therefore, the presence of frogs would cause negative feedback on the crop pest predator population, consequently increasing crop pest abundance. Our study shows frogs exhibit a preference for arthropod predators over pests, especially at a lesser density (Table 2, Fig. 4). The diet of amphibians in croplands has been found to be dominated by non-pests (94% of diet) (Brown, 1974; Khatiwada et al. 2016). Lisiecki (2019) also highlights the absence of any significant effect of lizards on crop damage due to intraguild predation on the beneficial arthropods. Though generalist species have a wide range of prey choices, they often weaken the biocontrol efficiency as they kill and eat natural enemies as well (Snyder et al., 2022). Therefore, intraguild predation can be a cause for disrupting pest regulation in a multitrophic system.

Fig. 4.

Impact of prey preference shift in super-predator frog resulting in disrupted pest regulation. The preference of frogs for crop pests (black arrow) reduces the arthropod pest predator density due to resource depletion (broken arrow) lowering the predation pressure on crop pests by the pest predators (grey arrow) (A). The preference of frogs for the arthropod pest predators (black arrow) causes an increase in crop pests due to lack of predation pressure from the predators (arrow in grey) (B). Width of arrows signify feeding intensity.

(0.3MB).

The third possible factor that may affect pest regulation by frogs could be their feeding behavior. The model frog species seems to follow a Type II functional response pattern, however, this conclusion is based on a small number of replicates. This means that the feeding rate remains constant beyond a certain prey density and is dependent upon the prey handling time. Handling time (Th) relates to every prey item consumed and is indifferent to the prey density. Even with a higher density of grasshoppers, the maximum number of prey consumed was determined by the time spent in prey handling (T/Th). This means that the rate of prey consumption would decline at higher prey densities (that includes both crop pests as well as arthropod pest predators) due to handling constraints (Thorp et al., 2018).

The high prey availability per arthropod pest predator especially for high frog density compared to control can be explained by the following:

• i

  Availability of pest refuge — at sampling phase 3 (30–40 days after transplanting) the crops have maximum vegetative mass (leaves and tillers) (http://www.agritech.tnau.ac.in/expert_system/paddy/Botany.html). The pest build-up was also the maximum at this phase. The pests found refuge in increased vegetative mass helping them to evade the frogs. Prey activity, crypsis, and refuge are important components that determine the outcome of prey-predator interactions (Sih and Moore, 1990; Chakraborty et al., 2017) and can destabilize the prey–predator cycle. Increase in search time also influences the profitability of the prey causing the predators to attack the more conspicuous prey, in our case the arthropod predators.

• ii

  High predation pressure of frogs on the arthropod predators - our model species is an ambush forager (field observation) that forages using visual cues of prey movement (Cooper, 1999). Behaviourally predatory arthropods are more agile than the herbivore pests (Ahmed et al., 2016) who remain increasingly well hidden in the higher available vegetative mass and this makes the natural predator easier prey for the super-predator frog.

What still remains to be explained is why higher frog density did not have any significant impact on the arthropod pest predators compared to lower frog density. A possible reason could be increased resource competition among frogs at higher frog density that compels them to seek out other food options i.e., crop pests. However, more conclusive results would require further experimental scrutiny.

The vulnerability of amphibians to agricultural intensification is globally acknowledged (Ghosh and Basu, 2020). Agricultural intensification is only going to increase in the coming years more so in developing countries to feed the growing population. Our study brings out the need to conserve frogs in the intensive agricultural landscape for effective natural pest regulation service. Scientific management of appropriate landscape structures at the patch, and landscape levels, like semi-natural habitats, hedgerows, ephemeral water bodies, degree of connectivity with adjoining remnant natural vegetation, and habitat heterogeneity (Ghosh and Basu, 2020), will help in maintaining a healthy frog community. Vegetated buffer strips that include vegetation along ditches, and levees connecting various crop areas are important for maintaining ecosystem health (Marshall and Moonen, 2002), at the same time, they also provide habitat refugia for the amphibians. Further, the presence of vegetated strips provides continuous prey availability for amphibians (Martori et al., 2005). Such linear features and their degree of connectivity to adjoining natural or semi-natural vegetation would eventually increase the abundance of herpetofauna in general. The consequent environmental heterogeneity between natural and croplands would increase species co-existence among natural arthropods as well as the amphibian assemblage (Da Silva et al., 2012). Alongside bolstering the frog density and diversity, complex landscapes can weaken the strength of intraguild predatory effects by reducing niche overlap through spatial separation or even alternate resource availability (Perez-Alvarez et al., 2019).

As our study shows, despite the limitations, we discuss in this study, amphibians can still make good pest regulators provided they are maintained in a threshold abundance. Conservation of species or groups of species and maintaining their population density above specific threshold levels in an agricultural landscape at which they can be effective is the only desirable pathway towards developing an ecologically healthy sustainable agricultural landscape.

DG and PB conceived the idea, DG worked on the methodology, and collected and analysed data; DG prepared the first draft and DG and PB wrote the final manuscript. Both authors approve the publication of the manuscript.

The authors report no declarations of interest.