1. Ecological data are collected over vast geographic areas using digital sensors such as camera traps and bioacoustic recorders. Camera traps have become the standard method for surveying many terrestrial mammals and birds, but camera trap arrays often generate millions of images that are time-consuming to label. This causes significant latency between data collection and subsequent inference, which impedes conservation at a time of ecological crisis. Machine learning algorithms have been developed to improve the speed of labelling camera trap data, but it is uncertain how the outputs of these models can be used in ecological analyses without secondary validation by a human.2. Here, we present our approach to developing, testing and applying a machine learning model to camera trap data for the purpose of achieving fully automated ecological analyses. As a case-study, we built a model to classify 26 Central African forest mammal and bird species (or groups). The model generalizes to new spatially and temporally independent data (n = 227 camera stations, n = 23,868 images), and outperforms humans in several respects (e.g. detecting 'invisible' animals). We demonstrate how ecologists can evaluate a machine learning model's precision and accuracy in an ecological context by comparing species richness, activity patternsThis is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.

Recent seizures of illegally held wildlife indicate a mounting global trade in pangolins involving all eight species. Seizures of illegally traded African pangolins are increasing as wild populations of Asian species decline. We investigated trade in pangolins and law enforcement efforts in Gabon; a country likely to have intact wild populations of three of the four species of African pangolin. We compared village sales and trade chains between 2002‐2003 and 2014. Hunters reported pangolins to be the most frequently requested species in 2014, and the value of pangolins had increased at every point along their trade chain. In Libreville, giant pangolin prices increased 211% and arboreal pangolin prices 73% whilst inflation rose only 4.6% over the same period. We documented a low rate of interception of illegally traded pangolins despite increased law enforcement. Surveys of potential export routes detected exports across forest borders, in conjunction with ivory, but not through public transport routes. We conclude that whilst there is clear potential and likelihood that a wild pangolin export trade is emerging from Gabon, traditional bushmeat trade chains may not be the primary supply route. We recommend adjusting conservation policies and actions to impede further development of illegal trade within and from Gabon.

Afrotropical forests host many of the world’s remaining megafauna, but even here they are confined to areas where direct human influences are low. We use a rare long-term dataset of tree reproduction and a photographic database of forest elephants to assess food availability and body condition of an emblematic megafauna species at Lopé National Park, Gabon. We show an 81% decline in fruiting over a 32-year period (1986-2018) and an 11% decline in body condition of fruit-dependent forest elephants from 2008-2018. Fruit famine in one of the last strongholds for African forest elephants should raise concern for the ability of this species and other fruit-dependent megafauna to persist in the long-term, with consequences for broader ecosystem and biosphere functioning.

No abstract