Consistent diel activity patterns of forest mammals among tropical regions

Abstract

An animal's daily use of time (their "diel activity") reflects their adaptations, requirements, and interactions, yet we know little about the underlying processes governing diel activity within and among communities. Here we examine whether community-level activity patterns differ among biogeographic regions, and explore the roles of top-down versus bottom-up processes and thermoregulatory constraints. Using data from systematic camera-trap networks in 16 protected forests across the tropics, we examine the relationships of mammals' diel activity to body mass and trophic guild. Also, we assess the activity relationships within and among guilds. Apart from Neotropical insectivores, guilds exhibited consistent cross-regional activity in relation to body mass. Results indicate that thermoregulation constrains herbivore and insectivore activity (e.g., larger Afrotropical herbivores are ~7 times more likely to be nocturnal than smaller herbivores), while bottom-up processes constrain the activity of carnivores in relation to herbivores, and top-down processes constrain the activity of small omnivores and insectivores in relation to large carnivores' activity. Overall, diel activity of tropical mammal communities appears shaped by similar processes and constraints among regions reflecting body mass and trophic guilds.

Fig. 1. Hypotheses (H1–H3) to determine processes that shape diel activity patterns in tropical forest mammal communities, with associated predictions (P1–3).

If the energetic cost of thermoregulation dominates (H1), we expect a positive relationship between body mass and nocturnality (1), regardless of trophic guild. If bottom-up regulation dominates (H2), predators will follow the diel activity of their prey (2). If top-down regulation dominates (H3), then we predict that small predators and potential prey species (herbivores and insectivores) will avoid top-predators (3). “+” represents a positive relationship between the activity of species groups (bottom-up process), and “−” represent a negative relationship between the activity of species groups (top-down process). Silhouettes from phylopic.org: jaguar, ocelot, and agouti by Gabriela Palomo-Munoz; tapir no license; browsing ruminant by Nobu Tamura (vectorized by T. Michael Keesey) http://creativecommons.org/licenses/by/3.0/.

Fig. 2. Map of the study areas and activity density examples.

Mammal activity data were collected using the standardized TEAM camera-trapping protocol in 16 protected areas (black dots in background) situated in 14 countries and tropical forests (areas shaded green on the map in the background) in three biogeographic regions. Activity density plots in each column show examples of species in each region (from left to right: Neotropics, Afrotropics, and Indo-Malayan tropics). Illustrations by John Meaghan.

Fig. 3. Diel activity in relation to body size and trophic guilds of tropical ground-dwelling and scansorial mammals in three regions.

Estimates correspond to the probability of activity during the day, night, and twilight extracted from the multinomial logit models fitted to TEAM camera-trap data (n = 126,382). Tick marks above the x-axis indicate the body mass of species included in the analysis. Lighter colours indicate model predictions for body masses that are below or above the range for species included in the analysis in each region. “n” represents the number of independent events. n_{carnivores_Neotropics} = 2182, n_{carnivores_Afrotropics} = 1474, n_{carnivores_Indo-Malayan_tropics} = 152, n_{omnivores_Neotropics} = 4656, n_{omnivores_Afrotropics} = 4656, n_{omnivores_Indo-Malayan_tropics} = 435, n_{herbivores_Neotropics} = 45,839, n_{herbivores_Afrotropics} = 47,458, n_{herbivores _Indo-Malayan_tropics} = 7803, n_{insectivores_Neotropics} = 4399, n_{insectivores_Afrotropics} = 3886, n_{insectivores_Indo-Malayan_tropics} = 212.

Fig. 4. Bottom-up and top-down processes as determinants of the diel activity of tropical mammals.

Centre of bars represent the mean coefficient estimates and bars show the 95% confidence intervals of the (GLMM) fitted to assess the relationship between the activity of species groups. The first column includes the relationship between the activity of large carnivores (n = 747) and prey (a large herbivores n = 191,294, b small herbivores n = 58392, d small omnivores n = 8098, and f small insectivores n = 7120) and h the relationship between the activity of large carnivores and small carnivores (n = 2280). The second column includes the relationship between small carnivores and potential prey (c small herbivores, e small omnivores, and g small insectivores). Note that n represents the total number of independent events for each species group and size. Green symbols illustrate a positive effect (bottom-up) and brown symbols illustrate a negative (top-down) relationship. Effects were considered significant when the 95% CI did not overlap zero (dashed horizontal lines). Neotropical sites “Neo” are denoted with squares, Afrotropical sites “Afro” with triangles, and Indo Malayan “Indo” with circles.