Sign in

Jonas Trepel

@jonastrepel.bsky.social
2K followers 521 following 37 posts

Postdoctoral researcher at Harvard University. Broadly interested in conservation biology, rewilding, megafauna ecology, global change & biodiversity.

PostsRepliesMedia
Jonas Trepel @jonastrepel.bsky.social · 02/08/2026
We also compared strictly protected areas with matched unprotected sites. Greening was slightly slower inside PAs, but protection did not alter the relationship between nitrogen deposition and vegetation greening, suggesting these global pressures extend well beyond reserve boundaries. (3/5)
Figure showing that protection status does not moderate the relationship between changes in green plant biomass and global change drivers. (A) Spatial distribution of strictly protected areas (IUCN category Ia, Ib and II) and paired control areas in similar environmental conditions and of the same size and shape; (B) interaction of global change drivers and protection status. The x-axis is restricted to the 2.5–97.5th percentile of the respective variable. The units for the predictors are as follows: fire frequency is given in number of fires between 2001 and 2024, human modification index (HMI) change is in percent of land modified, mean annual temperature (MAT) trend and max. annual temperature trend is °C year−1, N deposition is in kg km−2 year−1 and precipitation trend is mm year−1. Shading represents the 95% confidence intervals. All variables have been scaled and centered for modeling but are plotted on their original scale to ease interpretation. The models further contained MAT and mean annual precipitation (MAP) to account for the environmental context (MAT: ß = 0.33 [95% CI: −0.24, 0.91]; MAP: ß = −0.01 [−0.37, 0.36]. Variables were considered significant when the 95% confidence interval of the coefficient estimate did not overlap zero.
100
Jonas Trepel @jonastrepel.bsky.social · 02/08/2026
Using satellite observations from 2001–2024, we assessed how vegetation greening relates to N deposition alongside climate change, fire and human land-use change. Across most major climate zones and ecosystem types, nitrogen deposition showed the strongest association with greening. (2/5)
Figure showing positive association between N deposition and increases in green plant biomass across biomes. Models for changes in EVI as a function of N deposition were fitted for a variety of subsets separately, including Köppen–Geiger climatic regions (A)–(B) and aggregated biomes based on the functional biomes of Higgins et al. (2016) (C)–(D). (A) Coefficient estimates for each Köppen–Geiger climatic region separately. Error bars indicate the 95% confidence interval; (B) spatial distribution of climatic regions; (C) coefficient estimates for aggregated biomes separately; (D) spatial distribution of aggregated biomes. All variables have been scaled and centered for modeling. Variables were considered significant when the 95% confidence interval of the coefficient estimate did not overlap zero. A positive coefficient estimate indicates a positive relationship between the respective predictor variable and the response.
111
Jonas Trepel @jonastrepel.bsky.social · 02/08/2026
Our new paper is out in @ecography.bsky.social! We show that atmospheric nitrogen deposition is strongly associated with global vegetation greening, and may be an important - but often underestimated - driver of global vegetation change nsojournals.onlinelibrary.wiley.com/doi/10.1002/... (1/5)
Fig. 1 of the manuscript: Global increases in green plant biomass are strongly linked to atmospheric N deposition. (A) Spatial distribution of trends calculated with an autoregressive model that accounts for temporal autocorrelation; EVI was scaled by 10 000. (B) Model estimates from a generalized linear mixed effects model which accounts for spatial autocorrelation using spatial random fields. The x-axis is restricted to the 2.5–97.5th percentile of the respective variable; see the Supporting information for the full range and raw data points. A solid line indicates a statistically significant relationship, a dashed line a non-significant relationship. The units for the predictors are as follows: fire frequency is given in number of fires between 2001 and 2024; human modification index (HMI) change is the difference in percent of land modified between 2020 and 2000; mean annual temperature (MAT) is in °C and mean annual precipitation in mm, MAT trend and max. annual temperature trend is °C year−1 (from 1951 to 2024), N deposition is in kg km−2 year−1 and precipitation trend is mm year−1 (from 1951 to 2024). All variables have been scaled and centered for modeling but are plotted on their original scale to ease interpretation.
1122
Jonas Trepel @jonastrepel.bsky.social · 16/05/2026
Huge thanks to the fantastic co-author team: Liza le Roux, @cmlmagneville.bsky.social, Anika Oosthuizen, Debbie du Preez, Savannah Reichert, @jcsvenning.bsky.social and Rob Buitenwerf - and to everyone else who helped on the ground and maintained these remarkable exclosures over decades. 5/5
The field team crawling through the thicket inside an exclosureField team setting up a plot in the thicketA giraffe browsingTwo more white rhinos
162
Jonas Trepel @jonastrepel.bsky.social · 16/05/2026
Herbivores also made vegetation more open by reducing vegetation height and density - but these structural changes did not explain impacts on diversity. Instead, reducing dominance seems to be the key mechanism linking herbivores to more diverse and potentially more resilient plant communities. 4/5
A buffalo, disturbed while grazing and looking mildly displeasedAn elephant, eating a shrub.
161
Jonas Trepel @jonastrepel.bsky.social · 16/05/2026
We found that large herbivores increased plant species richness and functional redundancy, mainly by reducing the dominance of a few plant species and allowing more herbaceous plants to coexist. Interestingly, these effects were strong in savannas but largely absent in the thicket. 3/5
Effects of large herbivores on plant taxonomic diversity, dominance, functional diversity and vegetation structure. Estimates from intercept only GLMMS for plant community responses on plot (n = 108 pairwise comparisons) and site scale (n = 6 pairwise comparisons). Plant functional distance serves as a proxy for functional redundancy, with a lower distance representing a higher redundancy. Orange points right of the dashed line represent a significant increase of the variable in the presence of herbivores and orange points left of the dashed line represent a significant decrease in the presence of herbivores. Points in the background are the raw effect sizes, coloured by their value. Error bars represent the 95% confidence interval.Two white rhinos
141
Jonas Trepel @jonastrepel.bsky.social · 16/05/2026
We used six long-term (21-71 years!) exclosure experiments across South African savannas and thicket ecosystems to test how (nearly) intact herbivore communities affect plant communities. 2/5
Study area, exclosure locations and plot design. (a) Study area. The shaded polygons represent the protected areas we worked in (Addo Elephant National Park [AENP], Phongolo Nature Reserve [PNR], Kruger National Park [KNP]). The inset maps show the exclosures and the black dots correspond to the true plot locations. The Nyathi exclosures extend approximately 200 m at their longest site, while Jack's exclosure reaches about 2200 m. The PNR border fence is approximately 4400 m long and 20 m wide, and therefore appears as a line on the map. The black dots indicate clusters of three plots inside and three paired plots outside the fence. The Nkuhlu exclosure spans approximately 1350 m, the Satara exclosure about 550 m and the N'waswitshumbe exclosure approximately 2000 m; (b) plot layout. We identified all herbaceous plants on small plots (0.5 × 2 m), nested in bigger plots (2.5 × 10) on which all woody plants were identified. The LiDAR scanner was placed in the centre of the larger plot.Conceptual diagram of the tested hypotheses.
151
Jonas Trepel @jonastrepel.bsky.social · 16/05/2026
We found that large herbivores increased plant species richness and functional redundancy, mainly by reducing the dominance of a few plant species and allowing more herbaceous plants to coexist. Interestingly, these effects were strong in savannas but largely absent in the thicket. 3/5
Effects of large herbivores on plant taxonomic diversity, dominance, functional diversity and vegetation structure. Estimates from intercept only GLMMS for plant community responses on plot (n = 108 pairwise comparisons) and site scale (n = 6 pairwise comparisons). Plant functional distance serves as a proxy for functional redundancy, with a lower distance representing a higher redundancy. Orange points right of the dashed line represent a significant increase of the variable in the presence of herbivores and orange points left of the dashed line represent a significant decrease in the presence of herbivores. Points in the background are the raw effect sizes, coloured by their value. Error bars represent the 95% confidence interval.Two white rhinos
000
Jonas Trepel @jonastrepel.bsky.social · 16/05/2026
We used six long-term (21-71 years!) exclosure experiments across South African savannas and thicket ecosystems to test how (nearly) intact herbivore communities affect plant communities. 2/5
Study area, exclosure locations and plot design. (a) Study area. The shaded polygons represent the protected areas we worked in (Addo Elephant National Park [AENP], Phongolo Nature Reserve [PNR], Kruger National Park [KNP]). The inset maps show the exclosures and the black dots correspond to the true plot locations. The Nyathi exclosures extend approximately 200 m at their longest site, while Jack's exclosure reaches about 2200 m. The PNR border fence is approximately 4400 m long and 20 m wide, and therefore appears as a line on the map. The black dots indicate clusters of three plots inside and three paired plots outside the fence. The Nkuhlu exclosure spans approximately 1350 m, the Satara exclosure about 550 m and the N'waswitshumbe exclosure approximately 2000 m; (b) plot layout. We identified all herbaceous plants on small plots (0.5 × 2 m), nested in bigger plots (2.5 × 10) on which all woody plants were identified. The LiDAR scanner was placed in the centre of the larger plot.Conceptual diagram of the tested hypotheses.
100
Jonas Trepel @jonastrepel.bsky.social · 13/11/2025
Thanks to the great team, including @joe-atkinson.bsky.social , @andrewabraham.bsky.social, @jessekalwij.bsky.social, @jcsvenning.bsky.social, @econovoau.bsky.social, Rob Buitenwerf, Liza le Roux, Londiwe Mokeona, Michelle Greve, Steven Khosa, Caroline Makofane, Anika Oosthuizen & more! (3/3)
Screenshot of the paper title page, showing the full authorlist: Jonas Trepel, Joe Atkinson, Elizabeth le Roux, Andrew Abraham, Marge Aucamp, Michelle Greve, Marilize Greyling, Jesse Kalwij, Steven Khosa, Lukas Lindenthal, Caroline Makofance, Londiwe Mokoena, Anika Oosthuizen, Bent Rech, Erick Lundgren, Jens-Christian Svenning, Robert Buitenwerf
1232
Jonas Trepel @jonastrepel.bsky.social · 13/11/2025
higher herbivory intensity in South African savannas increases herbaceous plant richness and functional redundancy. Effects were largely consistent across scales, but strongest at the landscape level, highlighting large herbivores’ role in promoting diversity and ecosystem resilience (2/3)
Graphical abstract for our new paper, showing sketches of two reserves, one with and one without large herbivores. The one with large herbivores has higher herbaceous species richness and functional redundancy.
1220
Jonas Trepel @jonastrepel.bsky.social · 13/11/2025
Excited to share that my first PhD chapter just got published in @animalecology.bsky.social! You can check it out here: doi.org/10.1111/1365.... Based on fieldwork in the beautiful Waterberg Biosphere Reserve in South Africa, we show that.. (1/3)
Beautiful Sunset in Kaingo Private Game Reserve, one of our fieldsites A white rhino in Dabchick Wildlife Reserve, one of our fieldsites Woody savanna landscape in Swebeswebe Nature Reserve, on of our fieldsites Overview of the reserves in our study sites. Shows the location of 10 included reserves with pictograms indicating the largest herbivore in each.
25513
Jonas Trepel @jonastrepel.bsky.social · 17/08/2025
Setup & findings: In the Bavarian Forest NP (and many other places), managers use winter feeding to manage deer populations. We found this raises plant nutrient levels not just at feeding sites, but hundreds of meters beyond. Read the full paper for more detail! (2/5)
The figure shows the predicted plant nitrogen concentration in a radius of 2500m around the enclosures/feeding sites, clearly highlighting higher plant nutrient concentrations closer to the enclosures.
130
Jonas Trepel @jonastrepel.bsky.social · 29/04/2025
Check out our new paper, led by Bent Rech: we investigate space use of semi-feral water buffalo, cattle, and horses in a Danish rewilding area. We found that the three species differ markedly in their seasonal habitat selection and responses to resources and infrastructure. doi.org/10.1007/s002...
Figure 2 from Rech et al. (2025) shows habitat selection estimates for cattle, horses, and buffalo, separated by feeding status and season.
12813
Jonas Trepel @jonastrepel.bsky.social · 30/01/2025
It was a huge pleasure to talk about our @econovoau.bsky.social fieldwork & research at the Waterberg Research Convention today. Wonderful to meet many scientists, conservation practitioners & interested reserve owners from the Waterberg. Thanks to the WRSC for their invaluable work & organization!
Name tag and brochure of the Waterberg Research Convention 
Abstract of a talk with the title „Megafauna, mega-impact: how wild large herbivores shape ecosystems across the globe (and in the waterberg) Conference venue (still empty on the picture but completely full later) Panorama view from the research camp in our field site Kaingo game reserve
071
Jonas Trepel @jonastrepel.bsky.social · 23/11/2024
Great to have everyone here :). A quick intro: I’m a broadly interested ecologist, doing my PhD at @econovoau.bsky.social . I’m investigating how large mammals and global change impact biodiversity and ecosystem functioning, using both field-based approaches and large-scale remote sensing analyses.
A white rhino in the savannaAn elephant breaking branches from a big treeThree people collecting data in the savannaTwo giraffes walking
0729