TY - JOUR
KW - Bayesian occupancy models
KW - Biodiversity
KW - distributions
KW - imperfect detection
KW - long-term trends
KW - natural history collections
KW - presence-only
KW - species occurrence
KW - Vespinae
AU - Jönsson Galina M.
AU - Broad Gavin R.
AU - Sumner Seirian
AU - Isaac Nick J. B.
AB - ABSTRACT The current dearth of long-term insect population trends is a major obstacle to conservation. Occupancy models have been proposed as a solution, but it remains unclear whether they can yield long-term trends from natural history collections, since specimen records are normally very sparse. A common approach for sparse data is to coarsen its spatial and/or temporal resolution, although coarsening risks violating model assumptions. We (i) test whether occupancy trends of three social wasp (Hymenoptera: Vespidae: Vespinae) species – the common wasp (Vespula vulgaris), the German wasp (Vespula germanica) and the European hornet (Vespa crabro) – have changed in England between 1900 and 2016, and (ii) test the effect of spatiotemporal resolution on the performance of occupancy models using very sparse data. All models are based on an integrated dataset of occurrence records and natural history collection specimen records. We show that occupancy models can yield long-term species-specific trends from very sparse natural history collection specimens. We present the first quantitative trends for three Vespinae species in England over 116 years. Vespula vulgaris and V. germanica show stable trends over the time series, whilst V. crabro s occupancy decreased from 1950 to 1970 and increased since 1970. Moreover, we show that spatiotemporal resolution has little effect on model performance, although coarsening the spatial grain is an appropriate method for achieving enough records to estimate long-term changes. With the increasing availability of biological records, the model formulation used here has the potential to provide novel insights by making use of natural history collections  unique specimen assemblages.
BT - Insect Conservation and Diversity
DO - https://doi.org/10.1111/icad.12494
M1 - 5
N2 - ABSTRACT The current dearth of long-term insect population trends is a major obstacle to conservation. Occupancy models have been proposed as a solution, but it remains unclear whether they can yield long-term trends from natural history collections, since specimen records are normally very sparse. A common approach for sparse data is to coarsen its spatial and/or temporal resolution, although coarsening risks violating model assumptions. We (i) test whether occupancy trends of three social wasp (Hymenoptera: Vespidae: Vespinae) species – the common wasp (Vespula vulgaris), the German wasp (Vespula germanica) and the European hornet (Vespa crabro) – have changed in England between 1900 and 2016, and (ii) test the effect of spatiotemporal resolution on the performance of occupancy models using very sparse data. All models are based on an integrated dataset of occurrence records and natural history collection specimen records. We show that occupancy models can yield long-term species-specific trends from very sparse natural history collection specimens. We present the first quantitative trends for three Vespinae species in England over 116 years. Vespula vulgaris and V. germanica show stable trends over the time series, whilst V. crabro s occupancy decreased from 1950 to 1970 and increased since 1970. Moreover, we show that spatiotemporal resolution has little effect on model performance, although coarsening the spatial grain is an appropriate method for achieving enough records to estimate long-term changes. With the increasing availability of biological records, the model formulation used here has the potential to provide novel insights by making use of natural history collections  unique specimen assemblages.
PY - 2021
SP - 543
EP - 555
T2 - Insect Conservation and Diversity
TI - A century of social wasp occupancy trends from natural history collections: spatiotemporal resolutions have little effect on model performance
UR - https://resjournals.onlinelibrary.wiley.com/doi/abs/10.1111/icad.12494
VL - 14
ER -