Modeling the influence of temporal and spatial factors on the assessment of impacts of pesticides on skylarks

Abstract

Spatio-temporal factors strongly influence the population dynamics of animals; thus there have been calls to integrate these factors in environmental impact assessment of toxic compounds. To date, methodological difficulties have probably prevented this union. However, new modeling techniques that could help are available. This paper presents the construction and application of an agent-based simulation model of skylarks in agricultural landscapes and its use to assess the impact of pesticides relative to changes in landscape structure and mortality assumptions. Simulations indicated that pesticides had a negative impact on skylark population size. The annual reduction in numbers was variable and depended primarily upon migration mortality and an interaction between weather and pesticides. Altering landscape structure, crop diversity, or migration mortality assumptions resulted in a population change of approximately 37%, compared to a mean of 4% for pesticides. It was concluded that factors other than pesticides are likely to be limiting skylark numbers in most landscapes. This study demonstrates the importance of modeling the interactions between spatio-temporal environmental factors and the study organisms. Agent-based models (ABMs) are able to extract these relationships as emergent properties of their mechanistic nature. Therefore, we recommend the use of ABM models in future regulatory assessment of pesticides.

Interpreting outputs of agent-based models using abundance–occupancy relationships

“…The underlying cause of the increase in skylark numbers with landscape fragmentation was the temporal variation in food availability. With small fields, there was a higher diversity of vegetation types within the home range of the skylarks, and hence better foraging opportunities for birds while incubating and raising chicks; but this is only the case, when a variety of crops are grown in the fields (Gevers et al, 2011;Topping and Odderskaer, 2004), hence the reversal of the effect of landscape structure with the CROP scenario. Such effects can also be seen in real world studies.…”

“…The ALMaSS model framework is implemented for six model animal species resembling six real-world species: a carabid beetle (Bembidion lampros) (Bilde and Topping, 2004), a sheet web spider (Erigone atra) (Thorbek and Topping, 2005), the grey partridge (Perdix perdix) (Topping et al, 2010a), the skylark (Alauda arvensis) (Topping and Odderskaer, 2004), the European brown hare (Lepus europaeus) (Topping et al, 2010b) and the field vole (Microtus agrestis) (Dalkvist et al, 2009). These species were chosen because they are common in agricultural ecosystems and their ecology is well understood which allowed the construction of detailed behavioural rules in the species models.…”

The Potential for the Use of Agent-Based Models in Ecotoxicology

“…The first 11 years were discarded as a burn-in period, and the results were expressed as mean population size over the last 44 years. Weather data were as used by Topping and Odderskaer [30] and were a continuous loop of 11 years of weather data from a weather station near to the landscape simulated.…”

“…Crop husbandry is highly detailed (see [30]) and simulates all farming activities that would be carried out on that crop (e.g., plowing, harrowing, sowing, fertilizer applications, pesticide applications, harvest, and postharvest operations). Application of pesticides and fertilizers can be allocated specific characteristics (e.g., amount and type) and may result in changes in the vegetation growth, arthropod biomass, and provide field-specific information for animal models such as the type and amount of toxicant present.…”

“…Those used as examples here are Alauda arvensis (skylark) [30,31], Microtus agrestis (field vole) [29], Bembidion lampros (beetle) [32], Erigone atra/Oedothorax fuscus (spider) [33], and Capreolus capreolus (roe deer) [34]. These range from species with highly detailed behavior but low numbers (roe deer) to spiders with simple behavior but the necessity to handle over 1 million agents concurrently.…”