Ecology is not a dirty word

There are very few (if any) true ‘wilderness’ areas left, those completely untouched by human influence. This isn’t a tragedy – it’s an opportunity to grow, learn and discover more about the amazing planet we live on. Many ‘natural’ ecosystems have become social-ecological systems, where humans and nature can co-exist, not out-compete each other.

Agricultural systems are a perfect example. It’s hard to keep wild animals out of agroecosystems. They affect crop yields directly and indirectly across the growing season through positive (e.g. insects pollinating flowers) or negative (e.g. birds damaging fruit) interactions with crop plants. Because humans tend to label and categorise things (labels are easier to manage, justify or remove) we generally label these animals as either ‘bad’ or ‘good’ – aphids are annoying pests, bees are little angels. That’s all there is to it.

In reality, no organism is completely ‘bad’ or ‘good’ to the extreme; the effect it has on other organisms around it, including us, varies with context. All the individual plant-animal interactions happening in a single crop system are influenced by seasons, landscapes, management practices, and the social, cultural and economic values of the local farming community.

Sounds complicated…because it is. But this is what ecological science is all about. Understanding how changing contexts influence the way an organism interacts with their environment will help ensure that the landscapes we live in protect biodiversity, while also providing us with food and other resources.

As contexts change, the impacts wild animals have on crop plants can vary between ‘good’ and ‘bad’. Indeed, the very same species can be ‘good’ in one system and ‘bad’ in another. A single mirid bug species can provide biological control services by preying on crop pests in Mediterranean vegetable crop systems, but can also damage fruit when insect prey are scarce. One Brazilian bee species provides beneficial pollination services in cashew orchards, and causes minor crop losses in broccoli fields by collecting parts of broccoli heads to make nests. These examples, and many others, show that creating a positive or negative ‘label’ for a species doesn’t always represent the overall effect it has on the agroecosystem it interacts with.

So designing research and farm management programs based on simplistic assumptions of these cost/benefit labels is not the best way to balance sustainable food production and biodiversity conservation goals.

In a new study (published with co-authors Rebecca Peisley, Gary Luck and Romina Rader), we synthesised knowledge gained from a couple of hundred papers that have documented how wild bird and insect activity affects crop yields. We found that this type of plant-animal interaction has usually been studied from two separate perspectives: one group of studies (mostly agricultural sciences) focused on identifying and managing the costs that animals create for farmers through direct damage to crops; while the second group of studies (mostly ecology and conservation focused) aimed to document the benefits that animals provide to farmers, like pollination and other ecosystem services. These studies also often ignored other contextual factors that influence final yields, like seasonal changes in animal activity, or interactions between different species groups. This reductionist approach makes a system easier to study and it provides valuable knowledge on individual components of the system. But it also means the overall body of knowledge on how wild animals affect crops becomes disjointed and sometimes contradictory.

In our study, we suggest a new approach that can help researchers, conservationists and managers working in agricultural systems to address these complex issues. The approach relies on acknowledging that an animal’s role in an ecosystem can change across time and space. Attempting to associate changes in yields with an individual animal species, a single ecosystem service, or an isolated crop stage is not the most useful way to understand how net crop yields are influenced by animal activity. Instead, we suggest that it is more useful to try and identify the net outcomes of multiple positive and negative interactions that happen across an entire growing season, while also considering the social and environmental contexts that influence those outcomes.

Here’s an example. In Australian almond orchards, native birds have often been considered pests because they can cause crop losses by pecking at developing fruit. But after harvest has finished, the same birds also remove the decaying ‘mummy’ nuts left on trees. Growers often use paid manual labour to remove these nuts, because they harbour pathogens and pests. Gary Luck’s recent cost-benefit analysis of this social-ecological system showed that the positive economic value that growers gained from the birds’ ‘cleaning’ behaviour outweighed the cost of crop losses from damaged fruit. This resulted in an overall positive net return for farmers, when averaged across the entire plantation. So, in this situation, allowing the birds to ‘do their thing’ could be more cost effective for growers than deterring birds from the orchard and using manual labour to remove the mummy nuts by hand.

Very few studies have considered how wild animal activity creates cost-benefit trade-offs in agroecosystems. However, studying similar trade-offs in natural ecosystems is central to ecological science, and recent research in crop systems has identified synergies between different ecosystem services, like pollination and pest control.

Farms are ecosystems too. Most scientists, farmers and agronomists understand that we need to find a way to maintain sustainable crop production, while also protecting biodiversity and ecosystem function. To do this, agriculture needs to move beyond biological simplification and intensifying production. Cycles of interactions between crops, wild animals and people are inherent to productive agroecosystems and need to be sustained, not isolated from the system. Identifying net outcomes of all these interactions, and understanding how these outcomes vary under different environmental contexts, is a step toward achieving this.

© Manu Saunders 2015