Nutrient addition, but not predator exclusion, shapes arthropod communities and herbivory in a temperate forest

Plants support diverse arthropod communities, and arthropod herbivores respond differently to plant traits, nutritional content, and defences, which influence their host plants selection, survival and performance (bottom-up control). At the same time, arthropod herbivores are affected by their interactions with predators (top-down control). Investigating how these forces interact, and how they are affected by nutrient availability, is crucial for understanding the mechanisms driving herbivore populations and their impact on ecosystems.

We investigated how predator exclusion and fertilisation affect arthropod densities, sizes, and herbivory on two temperate forest tree species. Using a factorial design, we compared fertilised and unfertilised trees with and without the exclusion of flying vertebrate predators in the forest understory during September 2020 and 2021. We collected and identified arthropods into feeding guilds.

Fertilisation, but not predator exclusion, increased herbivory damage on trees as well as the size of predatory arthropods on the fertilised trees. Nutrient addition and predator exclusion had no significant effect on arthropod density. These patterns may indicate that the additional nutrients could have attracted herbivores, which in turn attracted their predators and may have enhanced their activity, thus potentially offsetting detectable changes in herbivore density.

These results suggest that nutrient enrichment and predator exclusion interact, with nutrient addition affecting plant growth and herbivory damage primarily, but also increasing the size of predatory arthropods, but not the size of all arthropods. Effects of predator exclusion were less pronounced, potentially due to larger predatory arthropods compensating for the absence of flying vertebrate predators.

Our study provide fully factorial field tests of top-down and bottom-up forces in a temperate forest understory, underscores the critical need to evaluate how diverse ecological interactions mediate the synergistic effects of nutrient pollution and the ongoing decline of insectivorous vertebrate predators on arthropod communities and herbivory damage, particularly as these preassures intensify in our rapidly changing environment.

Methods
Arthropod survey

We used beating sheets to collect all arthropods from each of the trees. We measured all arthropods (to the nearest mm) and identified them into morphospecies, assigned them to an order or family if necessary for their assignment to one of the following four feeding guilds: predator, leaf chewer, leaf sucker, and as having “no relationship” (i.e., arthropod with no consumptive relationship with other arthropods or plants). Considering that the adult arthropod may belong to a different feeding guild than the juvenile, the developmental stage was considered in the guild assignment. 

Herbivory damage survey

Individual leaves with a piece of cm-scale were scanned (EPSON, 600 dpi, colour tiff format) within 12 hours after collection. To analyse herbivory, we first outlined the missing part of the leaves based on the expected shape in Photoshop® (following Sam et al., 2020). Then we calculated the remaining area (a) and the full expected area of each leaf (b, both in cm2) using ImageJ version 1.47 (National Institute of Health, USA), and we calculated the area eaten by the herbivores (c = b - a). We then calculated the proportion of leaf area lost as c/b for each leaf, which we used in our analyses as a mean across the 10 leaves per individual branch. For visualisation, we used percentages of herbivory. Next, we estimated the total number of leaves for each tree by taking the average of three independent estimates. We then determined the total leaf area per tree by multiplying the estimated number of leaves by the mean leaf area calculated from the leaves we used for the herbivory measurement.

Leaf toughness survey

We collected 10 random fresh leaves from each sapling. We measured leaf toughness using a FL50 penetrometer with a Stepper Motor Powered Test Stand TVO 500N55S (Sauter GmbH) and a 1.5-mm-wide tip. We measured toughness once for each leaf, avoiding leaf veins (Onoda et al., 2011)