Ecology

RESEARCH HIGHLIGHT
03 March 2023

Human hunt at least 19% of bat species worldwide — especially flying foxes, which can have wingspans of 1.5 metres. More

Beer, C. et al. Terrestrial gross carbon dioxide uptake: global distribution and covariation with climate. Science 329, 834–838 (2010).Article 
CAS 
PubMed 

Google Scholar 
Badgley, G., Field, C. B. & Berry, J. A. Canopy near-infrared reflectance and terrestrial photosynthesis. Sci. Adv.3, e1602244 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
Chapin, F. S. III Effects of plant traits on ecosystem and regional processes: a conceptual framework for predicting the consequences of global change. Ann. Bot. 91, 455–463 (2003).Article 
PubMed 
PubMed Central 

Google Scholar 
Chu, C. et al. Does climate directly influence NPP globally? Global Chan. Biol. 22, 12–24 (2016).Article 

Google Scholar 
Yao, Y. et al. Spatiotemporal pattern of gross primary productivity and its covariation with climate in China over the last thirty years. Global Chan. Biol. 24, 184–196 (2018).Article 

Google Scholar 
Fang, J., Lutz, J. A., Wang, L., Shugart, H. H. & Yan, X. Using climate-driven leaf phenology and growth to improve predictions of gross primary productivity in North American forests. Global Chan. Biol. 26, 6974–6988 (2020).Article 

Google Scholar 
Fernández-Martínez, M. et al. The role of climate, foliar stoichiometry and plant diversity on ecosystem carbon balance. Global Chan. Biol. 26, 7067–7078 (2020).Article 

Google Scholar 
Migliavacca, M. et al. The three major axes of terrestrial ecosystem function. Nature 598, 468–472 (2021).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Reichstein, M., Bahn, M., Mahecha, M. D., Kattge, J. & Baldocchi, D. D. Linking plant and ecosystem functional biogeography. Proc. Natl. Acad. Sci. 111, 13697–13702 (2014).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Funk, J. L. et al. Revisiting the Holy Grail: using plant functional traits to understand ecological processes. Biol. Rev. 92, 1156–1173 (2017).Article 
PubMed 

Google Scholar 
Lavorel, S. & Garnier, É. Predicting changes in community composition and ecosystem functioning from plant traits: revisiting the Holy Grail. Fun. Ecol. 16, 545–556 (2002).Article 

Google Scholar 
Enquist, B. J. et al. in Advances in Ecological Research 52 (eds Samraat P, Guy W, & Anthony I. D) 249–318 (Academic Press, 2015).Garnier, E. et al. Plant functional markers capture ecosystem properties during secondary succession. Ecology 85, 2630–2637 (2004).Article 

Google Scholar 
Enquist, B. J. et al. Assessing trait-based scaling theory in tropical forests spanning a broad temperature gradient. Global Ecol. Biogeogr. 26, 1357–1373 (2017).Article 

Google Scholar 
Fyllas, N. M. et al. Solar radiation and functional traits explain the decline of forest primary productivity along a tropical elevation gradient. Ecol. Lett. 20, 730–740 (2017).Article 
PubMed 

Google Scholar 
Van der Plas, F. et al. Plant traits alone are poor predictors of ecosystem properties and long-term ecosystem functioning. Nat. Ecol. Evol. 4, 1602–1611 (2020).Article 
PubMed 

Google Scholar 
Ali, A., Yan, E.-R., Chang, S. X., Cheng, J.-Y. & Liu, X.-Y. Community-weighted mean of leaf traits and divergence of wood traits predict aboveground biomass in secondary subtropical forests. Sci. Total Environ. 574, 654–662 (2017).Article 
CAS 
PubMed 

Google Scholar 
Yang, J., Cao, M. & Swenson, N. G. Why Functional Traits Do Not Predict Tree Demographic Rates. Trend Ecol. Evol. 33, 326–336 (2018).Article 

Google Scholar 
Šímová, I. et al. The relationship of woody plant size and leaf nutrient content to large-scale productivity for forests across the Americas. J. Ecol. 107, 2278–2290 (2019).Article 

Google Scholar 
Li, Y. et al. Leaf size of woody dicots predicts ecosystem primary productivity. Ecol. Lett. 23, 1003–1013 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
He, N. et al. Ecosystem Traits Linking Functional Traits to Macroecology. Trend. Ecol. Evol. 34, 200–210 (2019).Article 

Google Scholar 
Rubio, V. E., Zambrano, J., Iida, Y., Umaña, M. N. & Swenson, N. G. Improving predictions of tropical tree survival and growth by incorporating measurements of whole leaf allocation. J. Ecol. 109, 1331–1343 (2021).Article 

Google Scholar 
Drake, J. E. et al. Increases in the flux of carbon belowground stimulate nitrogen uptake and sustain the long-term enhancement of forest productivity under elevated CO2. Ecol. Lett. 14, 349–357 (2011).Article 
PubMed 

Google Scholar 
Hilty, J., Muller, B., Pantin, F. & Leuzinger, S. Plant growth: the What, the How, and the Why. New Phytol. 232, 25–41 (2021).Article 
PubMed 

Google Scholar 
Xia, J. et al. Joint control of terrestrial gross primary productivity by plant phenology and physiology. Proc. Natl. Acad. Sci. 112, 2788–2793 (2015).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Suding, K. N. et al. Scaling environmental change through the community-level: A trait-based response-and-effect framework for plants. Global Chan. Biol. 14, 1125–1140 (2008).Article 

Google Scholar 
Liu, C., Li, Y., Yan, P. & He, N. How to Improve the Predictions of Plant Functional Traits on Ecosystem Functioning? Front. Plant Sci. 12, 622260 (2021).Reich, P. B., Walters, M. B. & Ellsworth, D. S. From tropics to tundra: global convergence in plant functioning. Proc. Natl. Acad. of Sci. 94, 13730–13734 (1997).Article 
CAS 

Google Scholar 
Reich, P. B. The world-wide ‘fast–slow’ plant economics spectrum: a traits manifesto. J. Ecol. 102, 275–301 (2014).Article 

Google Scholar 
Monteith, J. L. Climate and the efficiency of crop production in Britain. Philosophical Transactions of the Royal Society of London. B. Biol. Sci. 281, 277–294 (1977).
Google Scholar 
Garnier, E. Resource capture, biomass allocation and growth in herbaceous plants. Trend. Ecol. Evol. 6, 126–131 (1991).Article 
CAS 

Google Scholar 
Farnsworth, K. D., Albantakis, L. & Caruso, T. Unifying concepts of biological function from molecules to ecosystems. Oikos 126, 1367–1376 (2017).Article 

Google Scholar 
Zhang, R. et al. Biodiversity alleviates the decrease of grassland multifunctionality under grazing disturbance: A global meta-analysis. Global Ecol. Biogeogr. 31, 155–167 (2022).Article 

Google Scholar 
Jing, X. et al. The links between ecosystem multifunctionality and above-and belowground biodiversity are mediated by climate. Nat. Commun. 6, 1–8 (2015).Article 

Google Scholar 
Peters, M. K. et al. Climate–land-use interactions shape tropical mountain biodiversity and ecosystem functions. Nature 568, 88–92 (2019).Article 
CAS 
PubMed 

Google Scholar 
Hu, W. et al. Aridity-driven shift in biodiversity–soil multifunctionality relationships. Nat. Commun. 12, 1–15 (2021).Article 

Google Scholar 
Jing, X. et al. The influence of aboveground and belowground species composition on spatial turnover in nutrient pools in alpine grasslands. Global Ecol. Biogeogr. 31, 486–500 (2022).Article 

Google Scholar 
Jing, X. et al. Above-and belowground complementarity rather than selection drives tree diversity-productivity relationships in European forests. Funct Ecol. 35, 1756–1767 (2021).He, N. et al. Predicting ecosystem productivity based on plant community traits. Trend. Plant Sci. 28, 43–53 (2023).Maynard, D. S. et al. Global relationships in tree functional traits. Nat. Commun. 13, 1–12 (2022).Article 

Google Scholar 
Michaletz, S. T., Kerkhoff, A. J. & Enquist, B. J. Drivers of terrestrial plant production across broad geographical gradients. Global Ecol. Biogeogr. 27, 166–174 (2018).Article 

Google Scholar 
Shipley, B. Net assimilation rate, specific leaf area and leaf mass ratio: which is most closely correlated with relative growth rate? A meta-analysis. Funct. Ecol. 20, 565–574 (2006).Article 

Google Scholar 
Violle, C. et al. Let the concept of trait be functional! Oikos 116, 882–892 (2007).Article 

Google Scholar 
Jucker, T., Bouriaud, O. & Coomes, D. A. Crown plasticity enables trees to optimize canopy packing in mixed-species forests. Funct. Ecol. 29, 1078–1086 (2015).Article 

Google Scholar 
McGill, B. J. Matters of Scale. Science 328, 575 (2010).Article 
CAS 
PubMed 

Google Scholar 
Penuelas, J. et al. Increasing atmospheric CO2 concentrations correlate with declining nutritional status of European forests. Communi. Biol. 3, 1–11 (2020).
Google Scholar 
Weemstra, M. et al. Towards a multidimensional root trait framework: a tree root review. New Phytol. 211, 1159–1169 (2016).Article 
CAS 
PubMed 

Google Scholar 
Oehri, J., Schmid, B., Schaepman-Strub, G. & Niklaus, P. A. Biodiversity promotes primary productivity and growing season lengthening at the landscape scale. Proc. Natl. Acad. Sci. 114, 10160–10165 (2017).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Wright, I. J. et al. The worldwide leaf economics spectrum. Nature 428, 821–827 (2004).Article 
CAS 
PubMed 

Google Scholar 
Diaz, S. et al. The global spectrum of plant form and function. Nature 529, 167–171 (2015).Article 
PubMed 

Google Scholar 
Liu, Y. et al. The optimum temperature of soil microbial respiration: Patterns and controls. Soil Biol. Biochem. 121, 35–42 (2018).Article 
CAS 

Google Scholar 
Zhao, N. et al. Coordinated pattern of multi-element variability in leaves and roots across Chinese forest biomes. Global Ecol. Biogeogr. 25, 359–367 (2016).Article 

Google Scholar 
Zhang, J. et al. C: N: P stoichiometry in China’s forests: From organs to ecosystems. Funct. Ecol. 32, 50–60 (2018).Article 

Google Scholar 
Karger, D. N. et al. Climatologies at high resolution for the earth’s land surface areas. Sci. Data 4, 1–20 (2017).Article 

Google Scholar 
Dirk Nikolaus, K. et al. Climatologies at high resolution for the earth’s land surface areas. EnviDat. https://doi.org/10.16904/envidat.332 (2021).Kerkhoff, A. J., Enquist, B. J., Elser, J. J. & Fagan, W. F. Plant allometry, stoichiometry and the temperature-dependence of primary productivity. Global Ecol. Biogeogr. 14, 585–598 (2005).Article 

Google Scholar 
Wright, I. J. et al. Global climatic drivers of leaf size. Science 357, 917–921 (2017).Article 
CAS 
PubMed 

Google Scholar 
Zhang, Y. et al. A global moderate resolution dataset of gross primary production of vegetation for 2000-2016. Sci. Data 4, 170165 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
Jolliffe, I. T. & Cadima, J. Principal component analysis: a review and recent developments. Philos. Trans. Soc. A Math. Phys. Eng. Sci. 374, 20150202 (2016).Article 

Google Scholar 
Wieczynski, D. J. et al. Climate shapes and shifts functional biodiversity in forests worldwide. Proc. Natl. Acad. Sci. 116, 587–592 (2019).Article 
CAS 
PubMed 

Google Scholar 
Lefcheck, J. S. piecewiseSEM: Piecewise structural equation modelling in r for ecology, evolution, and systematics. Method Ecol. Evol. 7, 573–579 (2016).Article 

Google Scholar 
Bürkner, P.-C. Advanced bayesian multilevel modeling with the R Package brms. R J. 10, 395–411 (2018).Bürkner, P.-C. brms: An R package for Bayesian multilevel models using Stan. J. Stat. Software 80, 1–28 (2017).Article 

Google Scholar 
Vehtari, A., Gelman, A. & Gabry, J. Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC. Stat. Comput. 27, 1413–1432 (2017).Article 

Google Scholar 
Vehtari, A. et al. loo: Efficient leave-one-out cross-validation and WAIC for Bayesian models. R package version 2, 1003 (2019).
Google Scholar 
Gabry, J. & Mahr, T. bayesplot: Plotting for Bayesian models. R package version 1 (2017).Mac Nally, R. & Walsh, C. J. Hierarchical partitioning public-domain software. Biodivers. Conserv. 13, 659 (2004).Article 

Google Scholar 
Murray, K. & Conner, M. M. Methods to quantify variable importance: implications for the analysis of noisy ecological data. Ecology 90, 348–355 (2009).Article 
PubMed 

Google Scholar 
Yan, P., He, N., Yu, K., Xu, L. & Van Meerbeek, K. Integrating multiple functional traits to predict ecosystem productivity. figshare (2023). Dataset. https://doi.org/10.6084/m9.figshare.22081634.v1. More

We performed data acquisition, processing, analysis and visualization using Python23 version 3.8 with the packages Numpy24, Pandas25, Geopandas26, Matplotlib27, Selenium, Beautiful Soup28, SciPy14 and scikit-learn29. The functions used for specific tasks are explicitly mentioned to allow validation and replication studies.Data acquisition and processingHuman PUUV-incidenceHantavirus disease has been notifiable in Germany since 2001. The Robert Koch Institute collects anonymized data from the local and state public health departments and offers via the SurvStat application2 a freely available, limited version of its database for research and informative purposes. We retrieved the reported laboratory-confirmed human PUUV-infections (({text{n}}=text{11,228}) from 2006 to 2021, status: 2022-02-07). From the attributes available for each case, we retrieved the finest temporal and spatial resolution, i.e., the week and the year of notification, together with the district (named “County” in the English version of the SurvStat interface).To avoid bias through underreporting, our dataset was limited to PUUV-infections since 2006. The years 2006–2021 contain 91.9% of the total cases from 2001 to 2021. Human PUUV-incidence was calculated as the number of infections per 100,000 people, by using population data from Eurostat30. For each year, we used the population reported for the January 1 of that year. The population for 2020 was also used for 2021.In the analysis, we only included districts where the total infections were (ge {20}) and the maximum annual incidence was (ge {2}) in the period 2006–2021. The spatial information about the infections provided by the SurvStat application refers to the district where the infection was reported. Therefore, in most of the cases, the reported district corresponds to the residence of the infected person, which may differ from the district of infection. To compensate partially for differences between the reported place of residence and the place of infection, we combined most of the urban districts with their surrounding rural district. The underlying assumption was that most infections reported in urban districts occurred in the neighboring or surrounding rural district. In addition, some urban and rural districts have the same health department. Supplementary Table 1 lists the combined districts.Weather dataFrom the German Meteorological Service31 we retrieved grids of the following monthly weather parameters over Germany from 2004 to 2021: mean daily air temperature—Tmean, minimum daily air temperature—Tmin, and maximum daily air temperature—Tmax (all temperatures are the monthly averages of the corresponding daily values, in 2 m height above ground, in °C); total precipitation in mm—Pr, total sunshine duration in hours—SD, mean monthly soil temperature in 5 cm depth under uncovered typical soil of location in °C—ST, and soil moisture under grass and sandy loam in percent plant useable water—SM. The dataset version for Tmean, Tmin, Tmax, Pr, and SD was v1.0; for ST and SM the dataset version was 0. × . The spatial resolution was 1 × 1 km2.The data acquisition was performed with the Selenium package. The processing was based on the geopandas package26 using a geospatial vector layer for the district boundaries of Germany32. Each grid was processed to obtain the average value of the parameter over each district. We first used the function within to define a mask based on the grid centers contained in the district; we then applied this mask to the grid. In this method, called “central point rasterizing”33, each rectangle of the grid was assigned to a single district, the one that contained its center. The typical processing error was estimated to be about 1%, which agrees with the rasterizing error reported by Bregt et al.33; we consider that most likely this error is significantly less than the uncertainties of the grids themselves, caused by calculation, interpolation, and erroneous or missing observations.Data structureOur analysis was performed at the district level based on the annual infections, acquired by aggregating the weekly cases. From each monthly weather parameter, we created 24 records, for all months of the two previous years. Each observation in our dataset characterized one district in one year. Its target was acquired by transforming the annual incidence, as described in the following section. Each observation comprised all 168 available predictors from the weather parameters (7 parameters × 24 months), thereafter called “variables”. The notation for the naming of the variables follows the format Vx__, where “Vx” can be V1 or V2 that corresponds to one or two years before, respectively;  is the abbreviation of the weather parameter (see previous subsection: “Weather data”); and  is the numerical value of the month, i.e., from 1 to 12.The observations for combined districts retained the label of the rural district. For their infections and populations, we aggregated the individual values, and recalculated the incidence. For their weather variables, we assigned the mean values weighted by the area of each district.Target transformationTo consider the effects that drive the occurrence of high district-relative incidence, we discretized the incidence at the district level. The incidence scaled at its maximum value for each district showed extreme values for minima and maxima. About 49% of all observations were in the range [0, 0.1) and 8% in the range [0.9, 1] (Fig. 5). Therefore, we specifically selected to discretize the scaled incidence with two bins, i.e., to binarize it.Figure 5Histograms of the annual PUUV incidence from 2006 to 2021, scaled to its maximum value for each of the selected districts. Left: Raw incidence. Right: Log-transformed incidence, according to Eq. (6).Full size imageWe first applied a log-transformation to the incidence values34, described in Eq. (6).$${text{Log – incidence}} = log_{10} left( {{text{incidence}} + 1} right)$$
(6)
The addition of a positive constant ensured a noninfinite value for zero incidence, with 1 selected so that the log-incidence is nonnegative, and a zero incidence was transformed into a zero log-incidence. This transformation aimed to increase the influence of nonzero incidence values; values that are not pronounced, but still hint at a nonzero infection risk. Its effect is demonstrated in the right plot of Fig. 5, where the positive skewness of the original data is reduced, i.e., low incidence values are spread to higher values, resulting to more uniform bin heights in the range [0.05, 0.95] after the transformation. Formally, in this case the log-transformation achieves a more uniform distribution for the non-extreme incidence values.For the binarization, we performed unsupervised clustering of the log-transformed incidence, separately for each district, applying the function KBinsDiscretizer of the scikit-learn package29. Our selected strategy was the k-means clustering with two bins, because it does not require a pre-defined threshold, and it can operate with the same fixed number of bins for every district, by automatically adjusting the cluster centroids accordingly.Classification methodWe concentrated only on those variable combinations that led to a linear decision boundary for the classification of our selected target. We selected support vector machines (SVM)35 with a linear kernel, because they combine high performance with low model complexity, in that they return the decision boundary as a linear equation of the variables. In addition, SVM is geometrically motivated36 and expected to be less prone to outliers and overfitting than other machine-learning classification algorithms, such as the logistic regression. For the complete modelling process, the regularization parameter C was set to 1, that is the default value in the applied SVC method of the scikit-learn package29, and the weights for both risk classes were also set to 1.Feature selectionOur aim was to use the smallest possible number of weather parameters as variables for a classification model with sufficient performance. To identify the optimal variable combination, we first applied an SVM with a linear kernel for all 2-variable combinations of the monthly weather variables from V2 and V1, i.e., 168 variables (7 weather parameters × 2 years × 12 months). Only for this step, the variables were scaled to their minimum and maximum values, which significantly reduced the processing time. For all the following steps, the scaler was omitted, because the unscaled support vectors were required for the final model. From the total 14,028 models for each unique pair ((frac{168!}{2!cdot left(168-2right)!})), we kept the 100 models with the best F1-score, i.e., of the harmonic mean of sensitivity and precision, and counted the occurrences of each year-month combination in the variables. The best F1-score was 0.752 for the pair (V1_Tmean_9 and V2_Tmax_4); and the best sensitivity was 83% for the pair (V2_Tmax_9 and V1_ST_9).The year-month combinations with more than 10% occurrences were: V1_9 (September of the previous year, with 49% occurrences), V2_9 (September of two years before, with 12%) and V2_4 (April of two years before, with 10%). To avoid sets with highly correlated variables, we formed 3-variable combinations, with exactly one variable from each year-month combination (threefold Cartesian product). From the total 343 models (73 combinations, i.e., 7 weather parameters for 3 year-month combinations), we selected the model with the best sensitivity and at least 70% precision, i.e., the variable set (V2_ST_4, V2_SD_9, and V1_ST_9). We consider that the criteria for this selection are not particularly crucial; and we expect comparable performance for most variable sets with a high F1-score, because the variables for each dimension of the Cartesian product were highly correlated. The eight variable sets with at least 70% precision and at least 80% sensitivity are shown in Supplementary Table 2.The SVM classifier has two hyperparameters: the regularization parameter C and the class weights. By decreasing C, the decision boundary becomes softer and more misclassifications are allowed. On the other hand, increasing the high-risk class weight, the misclassifications of high-risk observations are penalized higher, which is expected to increase the sensitivity and decrease the precision. The simultaneous adjustment of both hyperparameters ensures that the resulting model has the optimal performance with respect to the preferred metric. However, in order to avoid overfitting, we considered redundant a further model optimization with these two hyperparameters. For completeness, we examined SVM models for different values of the hyperparameters and found that the global maximum for the F1-score is in the region of 0.001 for C and 1.5 for the high-risk class weight. Our selected values C = 1 and high-risk class weight equal to 1 give the second best F1-score, which is a local maximum with comparable performance, mostly insensitive to the selection of C from the range [0.2, 5.5].The addition of a fourth variable from V1_6 (June of the previous year) resulted in a model with higher sensitivity but lower precision and specificity (for V1_Pr_6). The highest F1-score was achieved for the quadruple (V2_ST_4, V2_SD_9, V1_ST_9, V1_Pr_6). Because of the increased complexity without significant improvement in the performance, we considered unnecessary a further expansion of our variable triplet. More

Purvis, A., Gittleman, J. L., Cowlishaw, G. & Mace, G. M. Predicting extinction risk in declining species. Proc. R. Soc. B 267, 1947–1952 (2000).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Crooks, K. R. Relative sensitivities of mammalian carnivores to habitat fragmentation. Conserv. Biol. 16, 488–502 (2002).Article 

Google Scholar 
Fahrig, L. Non-optimal animal movement in human-altered landscapes. Funct. Ecol. 21, 1003–1015 (2007).Article 

Google Scholar 
Fahrig, L. & Rytwinski, T. Effects of roads on animal abundance: An empirical review and synthesis. Ecol. Soc. 14, 21 (2009).Article 

Google Scholar 
Lowry, H., Lill, A. & Wong, B. B. M. Behavioural responses of wildlife to urban environments. Biol. Rev. 88, 537–549 (2013).Article 
PubMed 

Google Scholar 
Sévêque, A., Gentle, L. K., López-Bao, J. V., Yarnell, R. W. & Uzal, A. Human disturbance has contrasting effects on niche partitioning within carnivore communities. Biol. Rev. 95, 1689–1705 (2020).Article 
PubMed 

Google Scholar 
Woodroffe, R. & Ginsberg, J. R. Edge effects and the extinction of populations inside protected areas. Science 1979(280), 2126–2128 (1998).Article 
ADS 

Google Scholar 
Dressel, S., Sandström, C. & Ericsson, G. A meta-analysis of studies on attitudes toward bears and wolves across Europe 1976–2012. Conserv. Biol. 29, 565–574 (2015).Article 
CAS 
PubMed 

Google Scholar 
Owen, D. & Pemberton, D. Tasmanian Devil: A Unique and Threatened Animal (Allen & Unwin, 2005).
Google Scholar 
Yirga, G. et al. Adaptability of large carnivores to changing anthropogenic food sources: diet change of spotted hyena (Crocuta crocuta) during Christian fasting period in northern Ethiopia. J. Anim. Ecol. 81, 1052–1055 (2012).Article 
PubMed 

Google Scholar 
Knight, R. L. & Kawashima, J. Y. Responses of raven and red-tailed hawk populations to linear right-of-ways. J. Wildl. Manag. 57, 266–271 (1993).Article 

Google Scholar 
Wilmers, C. C., Stahler, D. R., Crabtree, R. L., Smith, D. W. & Getz, W. M. Resource dispersion and consumer dominance: Scavenging at wolf- and hunter-killed carcasses in Greater Yellowstone, USA. Ecol. Lett. 6, 996–1003 (2003).Article 

Google Scholar 
Lambertucci, S. A., Speziale, K. L., Rogers, T. E. & Morales, J. M. How do roads affect the habitat use of an assemblage of scavenging raptors?. Biodivers. Conserv. 18, 2063–2074 (2009).Article 

Google Scholar 
Šálek, M., Kreisinger, J., Sedláček, F. & Albrecht, T. Do prey densities determine preferences of mammalian predators for habitat edges in an agricultural landscape?. Landsc. Urban Plan. 98, 86–91 (2010).Article 

Google Scholar 
Bateman, P. W. & Fleming, P. A. Big city life: Carnivores in urban environments. J. Zool. 287, 1–23 (2012).Article 

Google Scholar 
Auman, H. J., Meathrel, C. E. & Richardson, A. Supersize me: Does anthropogenic food change the body condition of silver gulls? A comparison between urbanized and remote, non-urbanized areas. Waterbirds 31, 122–126 (2008).Article 

Google Scholar 
Coon, C. A. C., Nichols, B. C., McDonald, Z. & Stoner, D. C. Effects of land-use change and prey abundance on the body condition of an obligate carnivore at the wildland-urban interface. Landsc. Urban Plan. 192, 103648 (2019).Article 

Google Scholar 
Beckmann, J. P. & Berger, J. Using black bears to test ideal-free distribution models experimentally. J. Mammal. 84, 594–606 (2003).Article 

Google Scholar 
Fedriani, J. M., Fuller, T. K. & Sauvajot, R. M. Does availability of anthropogenic food enhance densities of omnivorous mammals? An example with coyotes in southern California. Ecography 24, 325–331 (2001).Article 

Google Scholar 
Prange, S., Gehrt, S. D. & Wiggers, E. P. Influences of anthropogenic resources on raccoon (Procyon lotor) movements and spatial distribution. J. Mammal. 85, 483–490 (2004).Article 

Google Scholar 
Tucker, M. A., Santini, L., Carbone, C. & Mueller, T. Mammal population densities at a global scale are higher in human-modified areas. Ecography 44, 1–13 (2021).Article 

Google Scholar 
Blanco, G., Lemus, J. A. & García-Montijano, M. When conservation management becomes contraindicated: Impact of food supplementation on health of endangered wildlife. Ecol. Appl. 21, 2469–2477 (2011).Article 
PubMed 

Google Scholar 
Fischer, J. R., Stallknecht, D. E., Luttrell, M. P., Dhondt, A. A. & Converse, K. A. Mycoplasmal conjunctivitis in wild songbirds: The spread of a new contagious disease in a mobile host population. Emerg. Infect. Dis. 3, 69–72 (1997).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Brittingham, M. C. & Temple, S. A. A survey of avian mortality at winter feeders. Wildl. Soc. Bull. 14, 445–450 (1986).
Google Scholar 
Hivert, L. G. et al. High blood lead concentrations in captive Tasmanian devils (Sarcophilus harrisii): A threat to the conservation of the species?. Aust. Vet. J. 96, 442–449 (2018).Article 
CAS 
PubMed 

Google Scholar 
Carrete, M., Donázar, J. A. & Margalida, A. Density-dependent productivity depression in pyrenean bearded vultures: Implications for conservation. Ecol. Appl. 16, 1674–1682 (2006).Article 
PubMed 

Google Scholar 
Bozek, C. K., Prange, S. & Gehrt, S. D. The influence of anthropogenic resources on multi-scale habitat selection by raccoons. Urban Ecosyst. 10, 413–425 (2007).Article 

Google Scholar 
Jones, J. D. et al. Supplemental feeding alters migration of a temperate ungulate. Ecol. Appl. 24, 1769–1779 (2014).Article 
PubMed 

Google Scholar 
Šálek, M., Drahníková, L. & Tkadlec, E. Changes in home range sizes and population densities of carnivore species along the natural to urban habitat gradient. Mamm. Rev. 45, 1–14 (2015).Article 

Google Scholar 
Newsome, D. & Rodger, K. To feed or not to feed: a contentious issues in wildlife tourism. In Too Close for Comfort: Contentious Issues in Human-Wildlife Encounters (ed. Lunney, D.) 255–270 (Royal Zoological Society of New South Wales, 2008).Chapter 

Google Scholar 
Tucker, M. A. et al. Moving in the anthropocene: Global reductions in terrestrial mammalian movements. Science 1979(359), 466–469 (2018).Article 
ADS 

Google Scholar 
Polis, G. A., Anderson, W. B. & Holt, R. D. Toward an integration of landscape and food web ecology: The dynamics of spatially subsidized food webs. Annu. Rev. Ecol. Syst. 28, 289–316 (1997).Article 

Google Scholar 
Prange, S. & Gehrt, S. D. Changes in mesopredator-community structure in response to urbanization. Can. J. Zool. 82, 1804–1817 (2004).Article 

Google Scholar 
Rodewald, A. D., Kearns, L. J. & Shustack, D. P. Anthropogenic resource subsidies decouple predator–prey relationships. Ecol. Appl. 21, 936–943 (2011).Article 
PubMed 

Google Scholar 
Cortés-Avizanda, A., Jovani, R., Carrete, M. & Donázar, J. A. Resource unpredictability promotes species diversity and coexistence in an avian scavenger guild: A field experiment. Ecology 93, 2570–2579 (2012).Article 
PubMed 

Google Scholar 
Arrondo, E., Cortés-Avizanda, A. & Donázar, J. A. Temporally unpredictable supplementary feeding may benefit endangered scavengers. Ibis 157, 648–651 (2015).Article 

Google Scholar 
Smith, J. A., Thomas, A. C., Levi, T., Wang, Y. & Wilmers, C. C. Human activity reduces niche partitioning among three widespread mesocarnivores. Oikos 127, 890–901 (2018).Article 

Google Scholar 
de León, L. F. et al. Urbanization erodes niche segregation in Darwin’s finches. Evol. Appl. 12, 1329–1343 (2019).Article 
PubMed 

Google Scholar 
Manlick, P. J. & Pauli, J. N. Human disturbance increases trophic niche overlap in terrestrial carnivore communities. PNAS 117, 26842–26848 (2020).Article 
ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Blair, R. B. Land use and avian species diversity along an urban gradient. Ecol. Appl. 6, 506–519 (1996).Article 

Google Scholar 
Dettori, E. E. et al. Distribution and diet of recovering Eurasian otter (Lutra lutra) along the natural-to-urban habitat gradient (river Segura, SE Spain). Urban Ecosyst. 24, 1221–1230 (2021).Article 

Google Scholar 
McKinney, M. L. Urbanization as a major cause of biotic homogenization. Biol. Conserv. 127, 247–260 (2006).Article 

Google Scholar 
Guiler, E. R. Temporal and spatial distribution of the Tasmanian Devil, Sarcophilus harrisii (Dasyuridae: Marsupialia). Pap. Proc. R. Soc. Tasman 116, 153–163 (1982).
Google Scholar 
Patton, A. H. et al. A transmissible cancer shifts from emergence to endemism in Tasmanian devils. Science (1979) 370, eabb9772 (2020).CAS 

Google Scholar 
Cunningham, C. X. et al. Quantifying 25 years of disease-caused declines in Tasmanian devil populations: Host density drives spatial pathogen spread. Ecol. Lett. 24, 958–969 (2021).Article 
PubMed 
PubMed Central 

Google Scholar 
Rose, R. K., Pemberton, D. A., Mooney, N. J. & Jones, M. E. Sarcophilus harrisii (Dasyuromorphia: Dasyuridae). Mamm. Species 49, 1–17 (2017).Article 

Google Scholar 
Guiler, E. R. Observations on the Tasmanian devil, Sarcophilus harrisii (Marsupialia: Dasyuridae) I. Numbers, home range, movements and food in two populations. Aust. J. Zool. 18, 49–62 (1970).Article 

Google Scholar 
Jones, M. E. & Barmuta, L. A. Diet overlap and relative abundance of sympatric dasyurid carnivores: A hypothesis of competition. J. Anim. Ecol. 67, 410–421 (1998).Article 

Google Scholar 
Pemberton, D. et al. The diet of the Tasmanian Devil, Sarcophilus harrisii, as determined from analysis of scat and stomach contents. Pap. Proc. R. Soc. Tasman. 142, 13–22 (2008).
Google Scholar 
Rogers, T. L., Fox, S., Pemberton, D. & Wise, P. Sympathy for the devil: Captive-management style did not influence survival, body-mass change or diet of Tasmanian devils 1 year after wild release. Wildl. Res. 43, 544–552 (2016).Article 

Google Scholar 
Andersen, G. E., Johnson, C. N., Barmuta, L. A. & Jones, M. E. Dietary partitioning of Australia’s two marsupial hypercarnivores, the Tasmanian devil and the spotted-tailed quoll, across their shared distributional range. PLoS ONE 12, e0188529 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
Department of Primary Industries Parks Water and Environment. Recovery Plan for the Tasmanian devil (Sarcophilus harrisii) (2010).Brown, O. J. F. Tasmanian devil (Sarcophilus harrisii) extinction on the Australian mainland in the mid-Holocene: multicausality and ENSO intensification. Alcheringa Aust. J. Palaeontol. 30, 49–57 (2006).Article 

Google Scholar 
Lewis, A. C., Hughes, C. & Rogers, T. L. Effects of intraspecific competition and body mass on diet specialization in a mammalian scavenger. Ecol. Evol. 12, e8338 (2022).Article 
PubMed 
PubMed Central 

Google Scholar 
Andersen, G. E., McGregor, H. W., Johnson, C. N. & Jones, M. E. Activity and social interactions in a wide-ranging specialist scavenger, the Tasmanian devil (Sarcophilus harrisii), revealed by animal-borne video collars. PLoS ONE 15, e0230216 (2020).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Jones, M. E. Road upgrade, road mortality and remedial measures: Impacts on a population of eastern quolls and Tasmanian devils. Wildl. Res. 27, 289–296 (2000).Article 

Google Scholar 
Jones, M. E. & Barmuta, L. A. Niche differentiation among sympatric australian dasyurid carnivores. J. Mammal. 81, 434–447 (2000).Article 

Google Scholar 
Andersen, G. E., Johnson, C. N., Barmuta, L. A. & Jones, M. E. Use of anthropogenic linear features by two medium-sized carnivores in reserved and agricultural landscapes. Sci. Rep. 7, 11624 (2017).Article 
ADS 
PubMed 
PubMed Central 

Google Scholar 
Hamede, R. K., McCallum, H. & Jones, M. Seasonal, demographic and density-related patterns of contact between Tasmanian devils (Sarcophilus harrisii): Implications for transmission of devil facial tumour disease. Austral. Ecol. 33, 614–622 (2008).Article 

Google Scholar 
Kitchener, A. & Harris, S. From Forest to Fjaeldmark: Descriptions of Tasmania’s Vegetation (Department of Primary Industries, Parks, Water and Environment, Tasmania, 2013).
Google Scholar 
Wiggins, N. L. & Bowman, D. M. J. S. Macropod habitat use and response to management interventions in an agricultural—Forest mosaic in north-eastern Tasmania as inferred by scat surveys. Wildl. Res. 38, 103–113 (2011).Article 

Google Scholar 
Hobday, A. J. & Minstrell, M. L. Distribution and abundance of roadkill on Tasmanian highways: Human management options. Wildl. Res. 35, 712–726 (2008).Article 

Google Scholar 
Hingston, A. B. Impacts of logging on autumn bird populations in the southern forests of Tasmania. Pap. Proc. R. Soc. Tasman. 134, 19–28 (2000).
Google Scholar 
Taylor, R. J. Notes on the diet of the carnivorous mammals of the Upper Henty River Region, western Tasmania. Pap. Proc. R. Soc. Tasman. 120, 7–10 (1986).
Google Scholar 
Hall-Aspland, S., Rogers, T., Canfield, R. & Tripovich, J. Food transit times in captive leopard seals (Hydrurga leptonyx). Polar Biol. 34, 95–99 (2011).Article 

Google Scholar 
Bell, O. et al. Age-related variation in the trophic characteristics of a marsupial carnivore, the Tasmanian devil Sarcophilus harrisii. Ecol. Evol. 10, 7861–7871 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Bell, O. et al. Isotopic niche variation in Tasmanian devils Sarcophilus harrisii with progression of devil facial tumor disease. Ecol. Evol. 11, 8038–8053 (2021).Article 
PubMed 
PubMed Central 

Google Scholar 
Bearhop, S., Adams, C. E., Waldron, S., Fuller, R. A. & MacLeod, H. Determining trophic niche width: A novel approach using stable isotope analysis. J. Anim. Ecol. 73, 1007–1012 (2004).Article 

Google Scholar 
Layman, C. A. et al. Applying stable isotopes to examine food-web structure: An overview of analytical tools. Biol. Rev. 87, 545–562 (2012).Article 
PubMed 

Google Scholar 
Crawford, K., McDonald, R. A. & Bearhop, S. Applications of stable isotope techniques to the ecology of mammals. Mamm. Rev. 38, 87–107 (2008).Article 

Google Scholar 
Bender, M. M., Rouhani, I., Vines, H. M. & Black, C. C. Jr. 13C/12C ratio changes in crassulacean acid metabolism plants. Plant Physiol. 52, 427–430 (1973).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
O’Leary, M. H. Carbon isotope fractionation in plants. Phytochemistry 20, 553–567 (1981).Article 

Google Scholar 
Farquhar, G. D., O’Leary, M. H. & Berry, J. A. On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves. Aust. J. Plant Physiol. 9, 121–137 (1982).CAS 

Google Scholar 
Cernusak, L. A. et al. Environmental and physiological determinants of carbon isotope discrimination in terrestrial plants. New Phytol. 200, 950–965 (2013).Article 
CAS 
PubMed 

Google Scholar 
NSW Parliamentary Counsel. Animal Research Act 1985 (NSW Parliamentary Counsel, 1985).
Google Scholar 
National Health and Medical Research Council (Australia). Australian Code for the Care and Use of Animals for Scientific Purposes (National Health and Medical Research Council, 2013).
Google Scholar 
du Sert, N. P. et al. Reporting animal research: Explanation and elaboration for the ARRIVE guidelines 2.0. PLoS Biol. 18, e3000411 (2020).Article 

Google Scholar 
Environmental Systems Research Institute. ArcGIS Desktop Version 10.8.1. https://www.esri.com/en-us/arcgis/products/arcgis-desktop/overview (2020).Tasmanian Vegetation Monitoring and Mapping Program. TASVEG 4.0. Natural Values Conservation Branch, Department of Primary Industries, Parks, Water and Environment thelist.tas.gov.au/app/content/data/geo-meta-data-record?detailRecordUID=b5c7a079-14bc-4b3c-af73-db7585d34cdd (2020).Land Tasmania. LIST Land Tenure. Land Tasmania thelist.tas.gov.au/app/content/data/geo-meta-data-record?detailRecordUID=9b8bf099-d668–433d-981b-a0f8f964f827 (2015).Hickey, J. E. & Wilkinson, G. R. The development and current implementation of silvicultural pratices in native forests in Tasmania. Aust. For. 62, 245–254 (1999).Article 

Google Scholar 
Whiteley, S. B. Calculating the sustainable yield of Tasmania’s State forests. Tasforests 11, 23–34 (1999).
Google Scholar 
Pemberton, D. Social Organisation and Behaviour of the Tasmanian devil, Sarcophilus harrisii (University of Tasmania, 1990).
Google Scholar 
Attard, M. R. G., Lewis, A. C., Wroe, S., Hughes, C. & Rogers, T. L. Whisker growth in Tasmanian devils (Sarcophilus harrisii) and applications for stable isotope studies. Ecosphere 12, e03846 (2021).Article 

Google Scholar 
von Bertalanffy, L. Quantitative laws in metabolism and growth. Q. Rev. Biol. 32, 217–231 (1957).Article 

Google Scholar 
Rogers, T. L., Fung, J., Slip, D., Steindler, L. & O’Connell, T. C. Calibrating the time span of longitudinal biomarkers in vertebrate tissues when fine-scale growth records are unavailable. Ecosphere 7, e01449 (2016).Article 

Google Scholar 
Qi, H., Coplen, T. B., Geilmann, H., Brand, W. A. & Böhlke, J. K. Two new organic reference materials for δ13C and δ15N measurements and a new value for the δ13C of NBS 22 oil. Rapid Commun. Mass Spectrom. 17, 2483–2487 (2003).Article 
ADS 
CAS 
PubMed 

Google Scholar 
Qi, H. et al. A new organic reference material, l-glutamic acid, USGS41a, for δ13C and δ15N measurements—A replacement for USGS41. Rapid Commun. Mass Spectrom. 30, 859–866 (2016).Article 
ADS 
CAS 
PubMed 

Google Scholar 
Bond, A. L. & Hobson, K. A. Reporting stable-isotope ratios in ecology: Recommended terminology. Guidel. Best Pract. Waterbirds 35, 324–331 (2012).
Google Scholar 
O’Connell, T. C. & Hedges, R. E. M. Investigations into the effect of diet on modern human hair isotopic values. Am. J. Phys. Anthropol. 108, 409–425 (1999).Article 
PubMed 

Google Scholar 
Jackson, A. L., Inger, R., Parnell, A. C. & Bearhop, S. Comparing isotopic niche widths among and within communities: SIBER—Stable Isotope Bayesian Ellipses in R. J. Anim. Ecol. 80, 595–602 (2011).Article 
PubMed 

Google Scholar 
R Core Team. R: A Language and Environment for Statistical Computing Version 4.2.0. https://www.r-project.org/ (2022).Bartoń, K. MuMIn: Multi-model inference. R Package Version 1.47.1. https://cran.r-project.org/package=MuMIn (2022).Burnham, K. P. & Anderson, D. R. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach (Colorado Cooperative Fish and Wildlife Research Unit, 2002).MATH 

Google Scholar 
Stock, B. C. et al. Analyzing mixing systems using a new generation of Bayesian tracer mixing models. PeerJ 6, e5096 (2018).Article 
PubMed 
PubMed Central 

Google Scholar 
Stock, B. C. & Semmens, B. X. MixSIAR: Bayesian Mixing Models in R. R Package Version 3.1.12. https://doi.org/10.5281/zenodo.1209993 (2022).Plummer, M., Stukalov, A. & Denwood, M. rjags: Bayesian graphical models using MCMC. R Package Version 4-13. https://cran.r-project.org/web/packages/rjags/rjags.pdf (2022).Newsome, S. D. et al. Variation in δ13C and δ15N diet–vibrissae trophic discrimination factors in a wild population of California sea otters. Ecol. Appl. 20, 1744–1752 (2010).Article 
PubMed 

Google Scholar 
Brooks, T. M. et al. Habitat loss and extinction in the hotspots of biodiversity. Conserv. Biol. 16, 909–923 (2002).Article 

Google Scholar 
Fahrig, L. Effects of habitat fragmentation on biodiversity. Annu. Rev. Ecol. Evol. Syst. 34, 487–515 (2003).Article 

Google Scholar 
Pardini, R., Nichols, E. & Püttker, T. Biodiversity response to habitat loss and fragmentation. Encycl. Anthr. 3, 229–239 (2018).Article 

Google Scholar 
Koch, A., Munks, S. & Driscoll, D. The use of hollow-bearing trees by vertebrate fauna in wet and dry Eucalyptus obliqua forest, Tasmania. Wildl. Res. 35, 727–746 (2008).Article 

Google Scholar 
Donázar, J. A., Cortés-Avizanda, A. & Carrete, M. Dietary shifts in two vultures after the demise of supplementary feeding stations: consequences of the EU sanitary legislation. Eur. J. Wildl. Res. 56, 613–621 (2010).Article 

Google Scholar 
Carbone, C., Teacher, A. & Rowcliffe, J. M. The costs of carnivory. PLoS Biol. 5, e22 (2007).Article 
PubMed 
PubMed Central 

Google Scholar 
Tucker, M. A., Ord, T. J. & Rogers, T. L. Revisiting the cost of carnivory in mammals. J. Evol. Biol. 29, 2181–2190 (2016).Article 
CAS 
PubMed 

Google Scholar 
Carbone, C., Mace, G. M., Roberts, S. C. & Macdonald, D. W. Energetic constraints on the diet of terrestrial carnivores. Nature 402, 286–288 (1999).Article 
ADS 
CAS 
PubMed 

Google Scholar 
Fisher, D. O. & Dickman, C. R. Body size-prey relationships in insectivorous marsupials: Tests of three hypotheses. Ecology 74, 1871–1883 (1993).Article 

Google Scholar 
Ruxton, G. D. & Houston, D. C. Obligate vertebrate scavengers must be large soaring fliers. J. Theor. Biol. 228, 431–436 (2004).Article 
ADS 
MathSciNet 
PubMed 
MATH 

Google Scholar 
Pemberton, D. & Renouf, D. A field-study of communication and social-behavior of the Tasmanian devil at feeding sites. Aust. J. Zool. 41, 507–526 (1993).Article 

Google Scholar 
Pye, R. J. et al. A second transmissible cancer in Tasmanian devils. Proc. Natl. Acad. Sci. USA 113, 374–379 (2016).Article 
ADS 
CAS 
PubMed 

Google Scholar 
James, S. et al. Tracing the rise of malignant cell lines: Distribution, epidemiology and evolutionary interactions of two transmissible cancers in Tasmanian devils. Evol. Appl. 12, 1772–1780 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
Hawkins, C. E. et al. Emerging disease and population decline of an island endemic, the Tasmanian devil Sarcophilus harrisii. Biol. Conserv. 131, 307–324 (2006).Article 

Google Scholar 
Pearse, A.-M. & Swift, K. Transmission of devil facial-tumour disease. Nature 439, 549 (2006).Article 
ADS 
CAS 
PubMed 

Google Scholar 
Wood, S. W., Hua, Q. & Bowman, D. M. J. S. Fire-patterned vegetation and the development of organic soils in the lowland vegetation mosaics of south-west Tasmania. Aust. J. Bot. 59, 126–136 (2011).Article 

Google Scholar 
Kohn, M. J. Carbon isotope compositions of terrestrial C3 plants as indicators of (paleo)ecology and (paleo)climate. PNAS 107, 19691–19695 (2010).Article 
ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Mayer, M., Ullmann, W., Sunde, P., Fischer, C. & Blaum, N. Habitat selection by the European hare in arable landscapes: The importance of small-scale habitat structure for conservation. Ecol. Evol. 8, 11619–11633 (2018).Article 
PubMed 
PubMed Central 

Google Scholar 
Barker, R. & Vestjens, W. Food of Australian Birds 1. Non-Passerines (CSIRO Publishing, 1989).Book 

Google Scholar 
Thomas, D. G. The bird community of Tasmanian temperate rainforest. Ibis 122, 298–306 (1980).Article 

Google Scholar 
DeVault, T. L., Rhodes, O. E. Jr. & Shivik, J. A. Scavenging by vertebrates: Behavioral, ecological, and evolutionary perspectives on an important energy transfer pathway in terrestrial ecosystems. Oikos 102, 225–234 (2003).Article 

Google Scholar 
DPIPWE. Annual Statewide Spotlight Surveys, Tasmania 2020/2021. Nature Conservation Report 21/2. (2021).Nguyen, H. K. D., Fielding, M. W., Buettel, J. C. & Brook, B. W. Habitat suitability, live abundance and their link to road mortality of Tasmanian wildlife. Wildl. Res. 46, 236–246 (2019).Article 

Google Scholar  More

Pianka, E. R. Niche overlap and diffuse competition. Proc. Natl. Acad. Sci. 71, 2141–2145 (1974).Article 
ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Chesson, P. Mechanisms of maintenance of species diversity. Annu. Rev. Ecol. Syst. 31, 343–366 (2000).Article 

Google Scholar 
Tokeshi, M. Species Coexistence: Ecological and Evolutionary Perspectives. (Wiley-Blackwell, 2009).Grinnell, J. Geography and evolution. Ecology 5, 225–229 (1924).Article 

Google Scholar 
Roughgarden, J. Resource partitioning among competing species—A coevolutionary approach. Theor. Popul. Biol. 9, 388–424 (1976).Article 
MathSciNet 
CAS 
PubMed 
MATH 

Google Scholar 
Syme, J., Kiszka, J. J. & Parra, G. J. Dynamics of cetacean mixed-species groups: A review and conceptual framework for assessing their functional significance. Front. Mar. Sci. 8, 1–19 (2021).Article 

Google Scholar 
Stensland, E., Angerbjörn, A. & Berggren, P. Mixed species groups in mammals. Mamm. Rev. 33, 205–223 (2003).Article 

Google Scholar 
Cords, M. & Würsig, B. A Mix of Species: Associations of Heterospecifics Among Primates and Dolphins. in Primates and Cetaceans: Field Research and Conservation of Complex Mammalian Societies (eds. Yamagiwa, J. & Karczmarski, L.) 409–431 (Springer, 2014). doi:https://doi.org/10.1007/978-4-431-54523-1_21.Goodale, E., Beauchamp, G. & Ruxton, G. D. Mixed-Species Groups of Animals: Behavior, Community Structure, and Conservation. (Academic Press, 2017).Krause, J. & Ruxton, G. D. Living in Groups. Oxford Series in Ecology and Evolution (Oxford University Press, 2002).Heymann, E. W. & Buchanan-Smith, H. M. The behavioural ecology of mixed-species troops of callitrichine primates. Biol. Rev. 75, 169–190 (2000).Article 
CAS 
PubMed 

Google Scholar 
Sridhar, H. & Guttal, V. Friendship across species borders: factors that facilitate and constrain heterospecific sociality. Philos. Trans. R. Soc. B Biol. Sci. 373, 1–9 (2018).Greenberg, R. Birds of many feathers: The formation and structure of mixed-species flocks of forest birds. in On the Move: How and Why Animals Travel in groups (eds. Boinski, S. & Gerber, P. A.) 521–558 (University of Chicago Press, 2000).Waser, P. M. ‘Chance’ and mixed-species associations. Behav. Ecol. Sociobiol. 15, 197–202 (1984).Article 

Google Scholar 
Whitesides, G. H. Interspecific associations of Diana monkeys, Cercopithecus diana, in Sierra Leone, West Africa: biological significance or chance?. Anim. Behav. 37, 760–776 (1989).Article 

Google Scholar 
Waser, P. M. Primate polyspecific associations: Do they occur by chance?. Anim. Behav. 30, 1–8 (1982).Article 

Google Scholar 
Alexander, R. D. The evolution of social behavior. Annu. Rev. Ecol. Syst. 5, 325–383 (1974).Article 

Google Scholar 
Kasozi, H. & Montgomery, R. A. Variability in the estimation of ungulate group sizes complicates ecological inference. Ecol. Evol. 10, 6881–6889 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Syme, J., Kiszka, J. J. & Parra, G. J. How to define a dolphin ‘group’? Need for consistency and justification based on objective criteria. Ecol. Evol. 12, 1–18 (2022).Article 

Google Scholar 
Hutchinson, J. M. C. & Waser, P. M. Use, misuse and extensions of ‘ideal gas’ models of animal encounter. Biol. Rev. 82, 335–359 (2007).Article 
PubMed 

Google Scholar 
Gotelli, N. J. Null model analysis of species co-occurrence patterns. Ecology 81, 2606–2621 (2000).Article 

Google Scholar 
Astaras, C., Krause, S., Mattner, L., Rehse, C. & Waltert, M. Associations between the drill (Mandrillus leucophaeus) and sympatric monkeys in Korup National Park. Cameroon. Am. J. Primatol. 73, 127–134 (2011).Article 
PubMed 

Google Scholar 
Mammides, C., Chen, J., Goodale, U. M., Kotagama, S. W. & Goodale, E. Measurement of species associations in mixed-species bird flocks across environmental and human disturbance gradients. Ecosphere 9, 1–14 (2018).Article 

Google Scholar 
Ovaskainen, O., Abrego, N., Halme, P. & Dunson, D. Using latent variable models to identify large networks of species-to-species associations at different spatial scales. Methods Ecol. Evol. 7, 549–555 (2016).Article 

Google Scholar 
Pollock, L. J. et al. Understanding co-occurrence by modelling species simultaneously with a Joint Species Distribution Model (JSDM). Methods Ecol. Evol. 5, 397–406 (2014).Article 

Google Scholar 
Warton, D. I. et al. So Many variables: Joint modeling in community ecology. Trends Ecol. Evol. 30, 766–779 (2015).Article 
PubMed 

Google Scholar 
Ovaskainen, O. et al. How to make more out of community data? A conceptual framework and its implementation as models and software. Ecol. Lett. 20, 561–576 (2017).Article 
PubMed 

Google Scholar 
Ovaskainen, O. & Abrego, N. Joint Species Distribution Modelling. (Cambridge University Press, 2020). https://doi.org/10.1017/9781108591720.Blanchet, F. G., Cazelles, K. & Gravel, D. Co-occurrence is not evidence of ecological interactions. Ecol. Lett. 23, 1050–1063 (2020).Article 
PubMed 

Google Scholar 
Haak, C. R., Hui, F. K., Cowles, G. W. & Danylchuk, A. J. Positive interspecific associations consistent with social information use shape juvenile fish assemblages. Ecology 101, 1–16 (2020).Article 

Google Scholar 
Bastianelli, G., Wintle, B. A., Martin, E. H., Seoane, J. & Laiolo, P. Species partitioning in a temperate mountain chain: Segregation by habitat vs. interspecific competition. Ecol. Evol. 7, 2685–2696 (2017).Aspin, T. & House, A. Alpha and beta diversity and species co-occurrence patterns in headwaters supporting rare intermittent-stream specialists. Freshw. Biol. n/a, (2022).Astarloa, A. et al. Identifying main interactions in marine predator-prey networks of the Bay of Biscay. ICES J. Mar. Sci. 76, 2247–2259 (2019).Article 

Google Scholar 
Parra, G. J. Resource partitioning in sympatric delphinids: space use and habitat preferences of Australian snubfin and Indo-Pacific humpback dolphins. J. Anim. Ecol. 75, 862–874 (2006).Article 
PubMed 

Google Scholar 
Parra, G. J., Wojtkowiak, Z., Peters, K. J. & Cagnazzi, D. Isotopic niche overlap between sympatric Australian snubfin and humpback dolphins. Ecol. Evol. 12, 1–11 (2022).Article 

Google Scholar 
Kiszka, J. J. et al. Ecological niche segregation within a community of sympatric dolphins around a tropical island. Mar. Ecol. Prog. Ser. 433, 273–288 (2011).Article 
ADS 

Google Scholar 
Bearzi, M. Dolphin sympatric ecology. Mar. Biol. Res. 1, 165–175 (2005).Article 

Google Scholar 
Zaeschmar, J. R. et al. Occurrence of false killer whales (Pseudorca crassidens) and their association with common bottlenose dolphins (Tursiops truncatus) off northeastern New Zealand. Mar. Mammal Sci. 30, 594–608 (2014).Article 

Google Scholar 
Elliser, C. R. & Herzing, D. L. Long-term interspecies association patterns of Atlantic bottlenose dolphins, Tursiops truncatus, and Atlantic spotted dolphins, Stenella frontalis, in the Bahamas. Mar. Mammal Sci. 32, 38–56 (2016).Article 

Google Scholar 
Kiszka, J. J., Perrin, W. F., Pusineri, C. & Ridoux, V. What drives island-associated tropical dolphins to form mixed-species associations in the southwest Indian Ocean?. J. Mammal. 92, 1105–1111 (2011).Article 

Google Scholar 
Brown, A. M., Bejder, L., Cagnazzi, D., Parra, G. J. & Allen, S. J. The north west cape, Western Australia: A potential hotspot for Indo-Pacific humpback dolphins Sousa chinensis?. Pacific Conserv. Biol. 18, 240–246 (2012).Article 

Google Scholar 
Allen, S. J., Cagnazzi, D., Hodgson, A. J., Loneragan, N. R. & Bejder, L. Tropical inshore dolphins of north-western Australia: Unknown populations in a rapidly changing region. Pacific Conserv. Biol. 18, 56–63 (2012).Article 

Google Scholar 
Palmer, C., Parra, G. J., Rogers, T. & Woinarski, J. Collation and review of sightings and distribution of three coastal dolphin species in waters of the Northern Territory. Australia. Pacific Conserv. Biol. 20, 116–125 (2014).Article 

Google Scholar 
Corkeron, P. J. Aspects of the Behavioral Ecology of Inshore Dolphins Tursiops truncatus and Sousa chinensis in Moreton Bay, Australia. in The Bottlenose Dolphin (eds. Leatherwood, S. & Reeves, R.) 285–293 (Elsevier, 1990). https://doi.org/10.1016/B978-0-12-440280-5.50018-4.Haughey, R. et al. Distribution and habitat preferences of Indo-Pacific Bottlenose Dolphins (Tursiops aduncus) inhabiting coastal waters with mixed levels of protection. Front. Mar. Sci. 8, 1–20 (2021).Article 

Google Scholar 
Hanf, D., Hodgson, A. J., Kobryn, H., Bejder, L. & Smith, J. N. Dolphin distribution and habitat suitability in North Western Australia: Applications and Implications of a Broad-Scale, Non-targeted Dataset. Front. Mar. Sci. 8, 1–18 (2022).Article 

Google Scholar 
Hunt, T. N., Allen, S. J., Bejder, L. & Parra, G. J. Identifying priority habitat for conservation and management of Australian humpback dolphins within a marine protected area. Sci. Rep. 10, 1–14 (2020).Article 

Google Scholar 
Hunt, T. N. Demography, habitat use and social structure of Australian humpback dolphins (Sousa sahulensis) around the North West Cape, Western Australia: Implications for conservation and management. PhD Thesis, Flinders University, Adelaide, Australia. (Flinders University, 2018).Cassata, L. & Collins, L. B. Coral reef communities, habitats, and substrates in and near sanctuary zones of Ningaloo marine park. J. Coast. Res. 241, 139–151 (2008).Article 

Google Scholar 
CALM MPRA. Management plan for the Ningaloo Marine Park and Muiron Islands Marine Management Area 2005–2015. (2005).Hunt, T. N. et al. Demographic characteristics of Australian humpback dolphins reveal important habitat toward the southwestern limit of their range. Endanger. Species Res. 32, 71–88 (2017).Article 

Google Scholar 
Mann, J. Behavioral sampling methods for cetaceans: A review and critique. Mar. Mammal Sci. 15, 102–122 (1999).Article 

Google Scholar 
Python Software Foundation. Python Language Reference, version 3.8.0. at https://www.python.org/ (2016).QGIS Development Team. QGIS Geographic Information System, version 3.8.3 Zanzibar. at http://qgis.osgeo.org (2019).Zanardo, N., Parra, G., Passadore, C. & Möller, L. Ensemble modelling of southern Australian bottlenose dolphin Tursiops sp. distribution reveals important habitats and their potential ecological function. Mar. Ecol. Prog. Ser. 569, 253–266 (2017).Hanberry, B. B. Finer grain size increases effects of error and changes influence of environmental predictors on species distribution models. Ecol. Inform. 15, 8–13 (2013).Article 

Google Scholar 
Gottschalk, T. K., Aue, B., Hotes, S. & Ekschmitt, K. Influence of grain size on species–habitat models. Ecol. Modell. 222, 3403–3412 (2011).Article 

Google Scholar 
Zuur, A. F., Ieno, E. N. & Elphick, C. S. A protocol for data exploration to avoid common statistical problems. Methods Ecol. Evol. 1, 3–14 (2010).Article 

Google Scholar 
Passadore, C., Möller, L. M., Diaz-Aguirre, F. & Parra, G. J. Modelling dolphin distribution to inform future spatial conservation decisions in a marine protected area. Sci. Rep. 8, 1–14 (2018).Article 
CAS 

Google Scholar 
Parra, G. J., Schick, R. & Corkeron, P. J. Spatial distribution and environmental correlates of Australian snubfin and Indo-Pacific humpback dolphins. Ecography (Cop.) 29, 396–406 (2006).Article 

Google Scholar 
Conrad, O. et al. System for Automated Geoscientific Analyses (SAGA) v. 2.1.4. Geosci. Model Dev. 8, 1991–2007 (2015).R Core Team. R version 3.6.1. at https://www.r-project.org/ (2019).RStudio Team. RStudio: Integrated Develpment for R. at http://rstudio.com/ (2019).Dormann, C. F. et al. Collinearity: A review of methods to deal with it and a simulation study evaluating their performance. Ecography (Cop.) 36, 27–46 (2013).Article 

Google Scholar 
Tikhonov, G. et al. Joint species distribution modelling with the r-package Hmsc. Methods Ecol. Evol. 11, 442–447 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Gelman, A. & Rubin, D. B. Inference from iterative simulation using multiple sequences. Stat. Sci. 7, 457–472 (1992).Article 
MATH 

Google Scholar 
Pearce, J. & Ferrier, S. Evaluating the predictive performance of habitat models developed using logistic regression. Ecol. Modell. 133, 225–245 (2000).Article 

Google Scholar 
Tjur, T. Coefficients of determination in logistic regression models—A new proposal: The coefficient of discrimination. Am. Stat. 63, 366–372 (2009).Article 
MathSciNet 
MATH 

Google Scholar 
Syme, J. The behavioural ecology of mixed-species groups of delphinids. PhD Thesis, Flinders University, Adelaide, Australia. (Flinders University, 2023).Wang, J. Y. Bottlenose Dolphin, Tursiops aduncus, Indo-Pacific Bottlenose Dolphin. in Encyclopedia of Marine Mammals (eds. Würsig, B., Thewissen, J. G. M. & Kovacs, K. M.) 125–130 (Elsevier, 2018). https://doi.org/10.1016/B978-0-12-804327-1.00073-X.Parra, G. J. & Jefferson, T. A. Humpback Dolphins. in Encyclopedia of Marine Mammals (eds. Würsig, B., Thewissen, J. G. M. & Kovacs, K. M.) 483–489 (Elsevier, 2018). https://doi.org/10.1016/B978-0-12-804327-1.00153-9.Dröge, E., Creel, S., Becker, M. S. & M’soka, J. Spatial and temporal avoidance of risk within a large carnivore guild. Ecol. Evol. 7, 189–199 (2017).Article 
PubMed 

Google Scholar 
Browning, N. E., Cockcroft, V. G. & Worthy, G. A. J. Resource partitioning among South African delphinids. J. Exp. Mar. Bio. Ecol. 457, 15–21 (2014).Article 

Google Scholar 
Kiszka, J. J., Méndez-Fernandez, P., Heithaus, M. R. & Ridoux, V. The foraging ecology of coastal bottlenose dolphins based on stable isotope mixing models and behavioural sampling. Mar. Biol. 161, 953–961 (2014).Article 
CAS 

Google Scholar 
Saayman, G. S. & Tayler, C. K. The socioecology of humpback dolphins (Sousa sp.). in Behavior of Marine Animals Current Perspectives in Research Volume 3: Cetaceans (eds. Winn, H. E. & Olla, B. L.) 165–226 (Springer, 1979).Gowans, S. & Whitehead, H. Distribution and habitat partitioning by small odontocetes in the Gully, a submarine canyon on the Scotian Shelf. Can. J. Zool. 73, 1599–1608 (1995).Article 

Google Scholar 
Clua, E. Mixed-species feeding aggregation of dolphins, large tunas and seabirds in the Azores. Aquat. Living Resour. 14, 11–18 (2001).Article 

Google Scholar 
Quérouil, S. et al. Why do dolphins form mixed-species associations in the azores?. Ethology 114, 1183–1194 (2008).Article 

Google Scholar 
Heithaus, M. R. & Dill, L. M. Food availability and tiger shark predation risk influence bottlenose dolphin habitat use. Ecology 83, 480–491 (2002).Article 

Google Scholar  More

Wajnberg, É. et al. (eds) Behavioral Ecology of Insect Parasitoids: From Theoretical Approaches to Field Applications 1st edn. (Blackwell Publishing Ltd, 2008).
Google Scholar 
Godfray, H. C. J. Parasitoids: Behavioral and Evolutionary Ecology (Princeton University Press, 1994).Book 

Google Scholar 
Morris, R. J., Lewis, O. T. & Godfray, H. C. J. Apparent competition and insect community structure: Towards a spatial perspective. Annales Zoologica Fennici 42, 1–14 (2005).
Google Scholar 
Forbes, A. A., Bagley, R. K., Beer, M. A., Hippee, A. C. & Widmayer, H. A. Quantifying the unquantifiable: Why Hymenoptera, not Coleoptera, is the most speciose animal order. BMC Ecol. 18, 1–11 (2018).Article 

Google Scholar 
Hassell, M. P. & Waage, J. K. Host–parasitoid population interactions. Annu. Rev. Entomol. 29, 89–114 (1984).Article 

Google Scholar 
Lafferty, K. D. et al. Parasites in food webs: The ultimate missing links. Ecol. Lett. 11, 533–546 (2008).Article 
PubMed 
PubMed Central 

Google Scholar 
Van Veen, F. J. F., Van Holland, P. D. & Godfray, H. C. J. Stable coexistence in insect communities due to density- and trait-mediated indirect effects. Ecology 86, 3182–3189 (2005).Article 

Google Scholar 
Schmidt, M. H. et al. Relative importance of predators and parasitoids for cereal aphid control. Proc. R. Soc. Lond. Series B Biol. Sci. 270, 1905–1909 (2003).Article 

Google Scholar 
Mills, N. J. & Wajnberg, É. Optimal foraging behavior and efficient biological control methods. In Behavioral Ecology of Insect Parasitoids: From Theoretical Approaches to Field Applications 1st edn (eds Wajnberg, É. et al.) 1–30 (Blackwell Publishing, 2008).
Google Scholar 
Vinson, S. B. Host suitability for insect parasitoids. Annu. Rev. Entomol. 25, 397–419 (1980).Article 

Google Scholar 
Benrey, B. & Denno, R. F. The slow-growth-high-mortality hypothesis: A test using the cabbage butterfly. Ecology 78, 987–999 (1997).
Google Scholar 
Chau, A. & Mackauer, M. Host-instar selection in the aphid parasitoid Monoctonus paulensis (Hymenoptera: Braconidae, Aphidiinae): Assessing costs and benefits. Can. Entomol. 133, 549–564 (2001).Article 

Google Scholar 
Strand, M. R. & Obrycki, J. J. Host specificity of insect parasitoids and predators. Bioscience 46, 422–429 (1996).Article 

Google Scholar 
Vinson, S. B. Host selection by insect parasitoids. Annu. Rev. Entomol. 21, 109–133 (1976).Article 

Google Scholar 
Wang, X. G. & Messing, R. H. Fitness consequences of body-size-dependent host species selection in a generalist ectoparasitoid. Behav. Ecol. Sociobiol. 56, 513–522 (2004).Article 

Google Scholar 
Liu, Z., Xu, B., Li, L. & Sun, J. Host-size mediated trade-off in a parasitoid Sclerodermus harmandi. PLoS ONE 6, e23260 (2011).Article 
ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Wang, X. Y., Yang, Z. Q., Wu, H. & Gould, J. R. Effects of host size on the sex ratio, clutch size, and size of adult Spathius agrili, an ectoparasitoid of emerald ash borer. Biol. Control 44, 7–12 (2008).Article 

Google Scholar 
Hardy, I. C. W., Griffiths, N. T. & Godfray, H. C. J. Clutch size in a parasitoid wasp: A manipulation experiment. J. Anim. Ecol. 61, 121–129 (1992).Article 

Google Scholar 
Scriber, J. M. & Slansky, F. The nutritional ecology of immature insects. Annu. Rev. Entomol. 26, 183–211 (1981).Article 

Google Scholar 
Moreau, J., Benrey, B. & Thiery, D. Assessing larval food quality for phytophagous insects: Are the facts as simple as they appear?. Funct. Ecol. 20, 592–600 (2006).Article 

Google Scholar 
Davidowitz, G., D’Amico, L. J. & Nijhout, H. F. The effects of environmental variation on a mechanism that controls insect body size. Evolut. Ecol. Res. 6, 49–62 (2004).
Google Scholar 
Williams, I. S. Slow-growth, high-mortality-a general hypothesis, or is it?. Ecol. Entomol. 24, 490–495 (1999).Article 

Google Scholar 
Chen, K. & Chen, Y. Slow-growth high-mortality: A meta-analysis for insects. Insect Sci. 25, 337–351 (2018).Article 
PubMed 

Google Scholar 
Waldbauer, G. P. The consumption and utilization of food by insects. Adv. Insect Physiol. 5, 229–288 (1968).Article 

Google Scholar 
Hochuli, D. F. Insect herbivory and ontogeny: How do growth and development influence feeding behaviour, morphology and host use?. Austral. Ecol. 26, 563–570 (2001).Article 

Google Scholar 
Holmes, L. A., Nelson, W. A. & Lougheed, S. C. Food quality effects on instar-specific life histories of a holometabolous insect. Ecol. Evol. 10, 626–637 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Kagata, H. & Ohgushi, T. Bottom-up trophic cascades and material transfer in terrestrial food webs. Ecol. Res. 21, 26–34 (2006).Article 

Google Scholar 
Scherber, C. et al. Bottom-up effects of plant diversity on multitrophic interactions in a biodiversity experiment. Nature 468, 553–556 (2010).Article 
ADS 
CAS 
PubMed 

Google Scholar 
Vidal, M. C. & Murphy, S. M. Bottom-up vs top-down effects on terrestrial insect herbivores: A meta-analysis. Ecol. Lett. 21, 138–150 (2018).Article 
PubMed 

Google Scholar 
Harvey, J. A. Factors affecting the evolution of development strategies in parasitoid wasps: The importance of functional constraints and incorporating complexity. Entomol. Exp. Appl. 117, 1–13 (2005).Article 

Google Scholar 
Charnov, E. L., Los-den Hartogh, R. L., Jones, W. T. & van den Assem, J. Sex ratio evolution in a variable environment. Nature 289, 27–33 (1981).Article 
ADS 
CAS 
PubMed 

Google Scholar 
Larson, A. O. The bean weevil and the southern Cowpea weevil in California. United States Department of Agriculture. Technical Bulletin No. 593, Washington, D. C. (1938).Askew, R. R. & Shaw, M. R. Parasitoid communities: their size, structure and development in Insect Parasitoids: 13th Symposium of Royal Entomological Society of London (eds. Waage, J.K. & Greathead, D.J. 225–264 (1986).Holmes, L. A., Nelson, W. A., Dyck, M. & Lougheed, S. C. Enhancing the usefulness of artificial seeds in seed beetle model systems research. Methods Ecol. Evol. 11, 1701–1706 (2020).Article 

Google Scholar 
Ellers, J., Van Alphen, J. J. M. & Sevenster, J. G. A field study of size-fitness relationships in the parasitoid Asobara tabida. J. Anim. Ecol. 67, 318–324 (1998).Article 

Google Scholar 
Wood, S. N. Stable and efficient multiple smoothing parameter estimation for generalized additive models. J. Am. Stat. 99, 673–686 (2004).Article 
MathSciNet 
MATH 

Google Scholar 
Wood, S. N. Generalized Additive Models: An Introduction with R 2nd edn. (Chapman and Hall/CRC, 2017).Book 
MATH 

Google Scholar 
Wood, S. N. Thin-plate regression splines. J. Roy. Stat. Soc. B 65, 95–114 (2003).Article 
MathSciNet 
MATH 

Google Scholar 
R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/ (2020). Accessed 3 April 2020.Burnham, K. P. & Anderson, D. R. Model Selection and Inference: A Practical Information-Theoretical Approach 2nd edn. (Springer-Verlag, 2002).MATH 

Google Scholar 
Wood, S. N., Pya, N. & Saefken, B. Smoothing parameter and model selection for general smooth models (with discussion). J. Am. Stat. Assoc. 111, 1548–1575 (2016).Article 
CAS 

Google Scholar 
Bolker, B., & R Development Core Team Tools for general maximum likelihood estimation. Version 1.0.20. (2017). Accessed 4 April 2020.Hothorn, T., Bretz, F. & Westfall, P. Simultaneous inference in general parametric models. Biometical. J. 50, 346–363 (2008).Article 
MathSciNet 
MATH 

Google Scholar 
Rose, N. L., Yang, H., Turner, S. D. & Simpson, G. L. An assessment of the mechanisms for the transfer of lead and mercury from atmospherically contaminated organic soils to lake sediments with particular reference to Scotland, UK. Geochim. Cosmochim. Acta 82, 113–135 (2012).Article 
ADS 
CAS 

Google Scholar 
Holmes, L. A., Nelson, W. A. & Lougheed, S. C. Data from: Food quality effects on instar-specific life histories of a holometabolous insect. Dryad Digital Repository. https://doi.org/10.5061/dryad.d7wm37px7 (2020).Therneau, T. A Package for Survival Analysis in R. R package version 3.2-13. https://CRAN.R-project.org/package=survival. (2021). Accessed 3 April 2020.Efron, B. The Jackknife, the Bootstrap, and Other Resampling Plans (Society for Industrial and Applied Mathematics, 1982).Book 
MATH 

Google Scholar 
Awmack, C. S. & Leather, S. R. Host plant quality and fecundity in herbivorous insects. Annu. Rev. Entomol. 47, 817–844 (2002).Article 
CAS 
PubMed 

Google Scholar 
Clancy, K. M. & Price, P. W. Rapid herbivore growth enhances enemy attack: Sublethal plant defenses remain a paradox. Ecology 68, 733–737 (1987).Article 

Google Scholar 
Loader, C. & Damman, H. Nitrogen content of food plants and vulnerability of Pieris rapae to natural enemies. Ecology 72, 1586–1590 (1991).Article 

Google Scholar 
Uesugi, A. The slow-growth high-mortality hypothesis: Direct experimental support in a leafmining fly. Ecol. Entomol. 40, 221–228 (2015).Article 

Google Scholar 
Feeny, P. Plant apparency and chemical defense. in Biochemical Interaction Between Plants and Insects. 1–40 (Springer, 1976).Teder, T. & Tammaru, T. Cascading effects of variation in plant vigor on the relative performance of insect herbivores and their parasitoids. Ecol. Entomol. 27, 94–104 (2002).Article 

Google Scholar 
Kagata, H., Nakamura, M. & Ohgushi, T. Bottom-up cascade in a tri-trophic system: Different impacts of host-plant regeneration on performance of a willow leaf beetle and its natural enemy. Ecol. Entomol. 30, 58–62 (2005).Article 

Google Scholar 
Vet, L. E. M., Lewis, W. J. & Cardé, R. T. Parasitoid foraging and learning. In Chemical Ecology of Insects 2 (eds Cardé, R. T. & Bell, W. J.) 65–101 (Springer, 1995).Chapter 

Google Scholar 
Ishii, Y. & Shimada, M. Learning predator promotes coexistence of prey species in host-parasitoid systems. Proc. Natl. Acad. Sci. 109, 5116–5120 (2012).Article 
ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Ode, P. J. & Hardy, I. C. Parasitoid sex ratios and biological control. Behavioral ecology of insect parasitoids. In Behavioral Ecology of Insect Parasitoids: From Theoretical Approaches to field applications (eds Wajnberg, E. et al.) 253–291 (Wiley, 2008).Chapter 

Google Scholar 
Xiaoyi, W. & Zhongqi, Y. Behavioral mechanisms of parasitic wasps for searching concealed insect hosts. Acta Ecol. Sin. 28, 1257–1269 (2008).Article 

Google Scholar 
Otten, H., Wäckers, F., Battini, M. & Dorn, S. Efficiency of vibrational sounding in the parasitoid Pimpla turionellae is affected by female size. Anim. Behav. 61, 671–677 (2001).Article 

Google Scholar 
Kaplan, I., Carrillo, J., Garvey, M. & Ode, P. J. Indirect plant-parasitoid interactions mediated by changes in herbivore physiology. Curr. Opin. Insect Sci. 14, 112–119 (2016).Article 
PubMed 

Google Scholar 
Ode, P. J. Plant toxins and parasitoid trophic ecology. Curr. Opin. Insect Sci. 32, 118–123 (2019).Article 
PubMed 

Google Scholar 
Barbosa, P., Gross, P. & Kemper, J. Influence of plant allelochemicals on the tobacco hornworm and its parasitoid, Cotesia congregate. Ecology 72, 1567–1575 (1991).Article 
CAS 

Google Scholar 
Barbosa, P. Natural enemies and herbivore–plant interactions: Influence of plant allelochemicals and host specificity. In Novel Aspects of Insect–Plant Interactions (eds Barbosa, P. & Letourneau, L. D. K.) 201–230 (Wiley, 1988).
Google Scholar 
Ode, P. J. Plant chemistry and natural enemy fitness: Effects on herbivore and natural enemy interactions. Annu. Rev. Entomol. 51, 163–185 (2006).Article 
CAS 
PubMed 

Google Scholar  More

Terraces are a land type that is defined by its shape. They have a distinct morphological structure and edge features that distinguish them from other land types. In this study, we define terraces as agricultural land with strip or wavy sections built on slopes greater than 2° along the contour direction. Figure 1 depicts Google Maps satellite images of terraces in the Loess Plateau region. Terraces can be distinguished from other features in remote sensing images based on their colour, morphology, texture, and structure. Terraces can be distinguished from construction land, water, glaciers, and deserts by their colours. Figure 1b–d shows terraces that are primarily green and yellow. Furthermore, terraces are generally distributed along the contour direction, and can therefore be identified based on their morphology. Terraced field ridges curve downward and resemble strips in Fig. 1b,d or circles or ovals in Fig. 1c rather than a neat grid-like distribution. These features differ in morphology from the flat land shown in Fig. 1h. Based on texture and structure, the field area of terraces can be identified based on their strong edge features, as shown in Fig. 1b–d. The edges of terraces have dark stripes caused by oblique illumination received from the sun, and the field ridge of terraces often intercepts part of the sunlight due to their height. Sloping cultivated land, as shown in Fig. 1g, has no evident terraced wall. The outline of sloping cultivated land in the high-resolution image is curved, with no prominent edge features. These findings are critical differences distinguishing terraces and sloping land in high-resolution images.Fig. 1The spatial location of the Loess Plateau and images of various types of cultivated land. (a) The spatial location of the Loess Plateau and Spatial distribution of various cultivated land types images, (b) wide strip-mounted terraces in Longxi, (c) circular wide terraces in central Yulin, (d) high resolution image of Zhuanglang County in July 2019, (e) Zhuanglang County in February 2020, (f) narrow terraces in Shangbao, Chongyi, Jiangxi Province, (g) sloping cropland in Zhenjiang Town, Laibin, Guangxi, and (h) horizontal cropland in the North China Plain.Full size imageDeep learning-based terrace extraction modelThe DLTEM is a terrace extraction model that uses deep learning algorithms and other supplementary information. Initially, a preliminary terrace distribution map was obtained using a deep learning algorithm. It was then combined with the spectral and digital elevation model (DEM) elevation information to fine-tune the results. The final spatial distribution of the terraces was produced by manual correction (Fig. 2). Traditional land classification models or methods typically superimpose spectral, elevation, and morphological texture information from remote sensing images together for training, such as random forest, which is easily ignored in training since morphological texture information accounts for a relatively small amount of the total information. This leads to significant errors while identifying land classes with textural characteristics. In contrast, the DLTEM focuses on morphological texture information from remote sensing images and classifies it into land classes, followed by auxiliary correction through additional information. Thus, this method is more suitable to extract terraces enriched with texture structure information.Fig. 2Flow chart of the deep learning-based terrace extraction model.Full size imageThe UNet++ network is a classic deep learning algorithm that is uniquely unrivaled in extracting colour, morphology, texture, and structure features from images and applying them for classification. In comparison with other Convolutional Neural Network (CNN) classification models (e.g., Fully Convolutional Networks (FCN)), it has high classification accuracy, fast computation speed, strong robustness, and provides variable importance metrics. Therefore, in this study, the UNet++ network was adopted as the network framework for deep learning; the primary data source used was high-resolution satellite imagery from 2019. DEM (SRTM v4.1) data were used to obtain the elevation information and GlobeLand30 data were used to obtain the spectral information. The results were corrected to construct the final map of the distribution of terraces in the Loess Plateau.Study areaThe Loess Plateau, one of China’s four major plateaus, is located in northern central China (34°–40° N and 103°–114° E) (Fig. 1). It is covered by a thick loess layer that ranges in thickness from 50 to 80 m, and is the world’s largest loess deposition area, covering 648,700 km2. The altitude of the Loess Plateau ranges from 800 to 3,000 m, its average annual temperature is 6–14 °C, and its average annual precipitation is 200–700 mm. Since ancient times, the Loess Plateau has been used for agriculture because of its fine grains, fluffy soil texture, and rich soluble mineral nutrients, all of which are conducive to crop cultivation. However, long-term unsustainable land use caused the degradation of the vegetation cover in the Loess Plateau. Moreover, the land is degrading due to considerable nutrient loss caused by long-term water erosion in conjunction with natural conditions, such as arid climate, loose soil, concentrated and heavy rainfall. The fragmented ground in the region has made it susceptible to soil erosion. It has also become the primary source of Yellow River sediment as a result of the massive flow of eroded sediment into the Yellow River, posing a serious threat to the economic and social development of the lower Yellow River basin.Terracing is one of the main measures used to enhance crop yield and conserve soil and water in the region. Since the 1980s, the Chinese government has implemented many large-scale slope-to-terrace projects in the Loess Plateau. Especially in recent years, the outline of the comprehensive management plan for the Loess Plateau area (2010–2030) has been promulgated with a planned area of 2.608 million hectares for slope to terrace conversion, making it the core area of slope to terrace conversion projects in the country.Data preparationAlthough high-resolution satellite images can be an important data source for the spatial distribution of terraces on the Loess Plateau, they are not ideal for terraces classification. On the one hand, a higher resolution image requires more storage space. On the other hand, it reduces the efficiency, prolongs the interpretation time, and increases the noise in the image, affecting the interpretation accuracy. Most of the terraces on the Loess Plateau are wider than 7 m (Fig. 1b–d). These are wide terraces in comparison with the narrow terraces of southern China (Fig. 1f), which are less than 2 m wide. Furthermore, it is also easy to mistake the fish-scale pits constructed for soil and water conservation for terraces because of their similarity in form. However, as the width of their field surface is less than 1.5 m, remote sensing images with a 2 m resolution can effectively prevent the false extraction of such features. Based on the actual situation of this study area, we chose a high-resolution image with a spatial resolution of 1.89 m from Google Maps 16 level as the data source. The colour, texture, and morphological features of terraces in the images show seasonal variations. In autumn and winter, the weather is dry, and the vegetation is less shaded in the Loess Plateau. During this time, even the edge features become more visible and easier to identify. As a result, we selected images from October 2018 to February 2019 whenever possible (Fig. 1c,d).Deep learning network selectionLand classification is the extraction of land types from remote sensing images using image segmentation techniques. As the key technology of image segmentation, the Fully Convolutional Network (FCN) classifies images at the pixel level. FCN follows the network structure pattern of encoding and decoding, which adopts AlexNet as the encoder of the network and then employs transposed convolution to up-sample the feature map output from the final convolutional layer of the encoder to the resolution of the input image to achieve pixel-level image segmentation. However, due to the large error in image pixel boundary localization, Ronneberger et al.29 improved the FCN structure in 2015 by expanding the capacity of the network decoder by adding a contracting path to the encoding and decoding modules to achieve more accurate pixel boundary localisation29. The U-Net network is commonly used in medical image processing because it requires a small number of training samples and is effective in classifying objects with a fixed structure and limited semantic information. This network is comparable to natural image semantic segmentation such as Deeplab v3+, which has a smaller number of model parameters and the same effect.Since the texture and morphological features of terraces and human organs have certain similarities, they are primarily manifested by simple semantic information contained within the terrace images themselves. Thus, high-level semantic information and low-level features of such images become more important. However, high-resolution images are more complicated and variable than medical image patterns, and errors in terrace extraction edge identification using the U-Net network, such as boundary segmentation of terraces and flatlands, still occur. To fully utilize the semantic information of the network, we adopted a nested U-Net architecture, namely the UNet++ network proposed by Zhou et al.28. The network integrates long-connected and short-connected architectures to capture features at different levels by adding a shallower U-Net structure and integrates them via feature superposition to make the scale difference of feature maps smaller when fused to enhance the correct rate of image segmentation edges. However, because the U-Net++ network increases the number of model parameters, this study adopted the sparse matrix approach to accelerate model training and decrease the number of parameters.Data pre-processingData pre-processing is a prerequisite for UNet++ network training, that is, valid input according to the standard format annotation before training can be performed. Since the UNet++ network proposed by Zhou et al.28. is primarily used for medical images, which have characteristics such as fixed image structure, no spatial information, and less pattern variation, labelling medical images is comparatively easier using this method. In contrast, high-resolution remote sensing images have a large number of rasters, many pattern changes, irregular image structure, and spatial information. Therefore, determining how to better annotate high-resolution remote sensing images and reduce the annotation workload becomes critical. First, we vectorized the training sample area and generated the terrace vector dataset using ArcGIS with a high-resolution remote sensing image as the primitive map. Second, we converted the terrace vector dataset into raster data. The information of the raster had to be identical to that of the primitive map, including the size of the raster, its processing range, and its coordinate system. The output was converted to TIFF format to complete the image annotation. Since the raster size input to UNet++ network training is a fixed size, it is much smaller than the original image. To simplify the process of inputting the original image and its annotation information, we added an image import module to DLTEM, which was a sliding window of 400*400, and read the image automatically by setting the corresponding judgement conditions. Finally, the entire high-resolution image was processed automatically into the model in accordance with the established rules for training.The goal of the data enhancement was to improve the universality and robustness of the UNet++ network training results. As mentioned above, the high-resolution images taken simultaneously often included clouds or other anomalies in some areas, as the images were stitched together using multiple sources of data fusion. This can easily form evident stitching traces (Fig. 1c,d) due to the different shooting times and image quality of various data sources, i.e., brightness, saturation, and colour contrast of the images. Thus, the model trained on the original image data has strong limitations, and in many scenes, there are notable matrix-type misclassification regions due to image differences, making extraction work challenging. Therefore, in this study, we first adjusted the brightness, grayscale, and contrast of the training data after input to enhance its colour feature recognition ability. We then altered the scaling of the image, and rotated and transformed the training image from 0° to 360° to enhance morphological feature recognition and the accuracy of the training network in terrace extraction.Parameter settingThe network parameter setting is the most critical hyperparameter for UNet++ network training. They are mainly divided into input image size, batch size, learning rate, number of iterations, objective function, gradient descent strategy, momentum, decay rate, and activation function. Among them, we set the image size to 400*400 pixels based on the actual situation of the terraced area, where the UNet++ network has four scaling times, and the image size must be a multiple of 16. The batch size primarily affects the convergence of the model. If the batch limit is set to one, the model is easily affected by the random perturbation phenomenon and cannot converge to find the optimal solution. Since the batch size is determined by the size of the video memory, the value of the batch is limited by equipment constraints. The model in this study used a 2080Ti video card with 11 GB of video memory, and the batch was set to 8. The learning rate, gradient descent strategy, and objective function play a role in whether the network can find the best classification model better and faster. The learning rate was set to 0.001 for the first 500 generations, with the goal of achieving fast convergence to the target region. The learning rate was then set to 0.0001 for 500–1,000 generations, and the model was fine-tuned by choosing a smaller learning rate to find the model with the highest classification accuracy. Adam was chosen for the gradient descent strategy. The momentum and adaptive learning rate were used to increase the convergence rate. The cross-entropy classification loss function was chosen as the objective function to improve the differentiation between terraced and non-terraced areas. Momentum, decay rate, and activation function were all adopted from the previous default settings of the UNet++ network.Data correctionIn this study, we primarily used high-resolution images from Google Earth as the data source to extract the distribution of terraces on the Loess Plateau. Because this image source only contains a large amount of texture structure information and no vegetation information, it is easy to misjudge and misclassify features with the same morphological structure and edge features, such as permanent snow and ice, water bodies, bare land, and artificial surfaces. Vegetation information was generally processed based on waveband data from multispectral/hyperspectral images. It requires topographic correction, atmospheric correction, radiometric calibration, de-clouding, and other operational processes, which are extremely sophisticated30.GlobeLand30 is a 30 m spatial resolution global surface coverage dataset developed by the National Geomatics Center of China. The most recent GlobeLand30 dataset (v2020) has been updated with data sources from 2017 to the present. Its extensive data sources enable effective reduction of the impacts of cloud cover, with an overall accuracy of 85.72%. The classification accuracy of permanent snow and ice, water bodies, bare land, and artificial surfaces of this dataset is as high as 75.79%, 84.70%, 81.76%, and 86.70%, respectively. Since the update time of v2020 data is similar to that of high-resolution images, it can be used as correction data for vegetation information31.Since the training image data are two-dimensional planar data with no elevation or slope information (Fig. 1g), certain flat fields with visible field bumps are easily misclassified as terraces. The Space Shuttle Radar Topography Mission (SRTM v4.1) DEM has a spatial resolution of 30 m and ranges from 60° N to 56° S, completely covering the Loess Plateau32,33. In this study, these data were treated as terrain correction data. The amendment standard corrects the areas that have been extracted as terraces below 2° to non-terraced areas according to the requirements of the Ministry of Natural Resources of China.The spatial resolution of our extracted terraces is 1.89 m, whereas the spatial resolution of GlobeLand30 and DEM as correction data sources is 30 m, which is difficult to meet the requirements of data processing. Hence, we up-sample the two correction data sources, and then used multi-source data fusion. First, we extracted and up-sampled the terraced areas of glaciers, rivers, and deserts from GlobeLand30 to a spatial resolution of 1.89 m. Secondly, we up-sampled the DEM to 1.89 m using spatial interpolation for its raster centre as the true value of the region and performed a slope calculation for the up-sampled DEM. Further, the spatial distribution maps of glaciers, rivers, deserts, and slope maps of the Loess Plateau with the same resolution as the spatial distribution maps of terraces were available. Finally, we superimposed these images, used the terrace range in the TDMLP as a mask, and assessed the pixels in the mask area one by one. If a pixel belonged to permanent snow and ice, a water body, bare land, or an artificial surface, or had a slope less than 2°, it was modified to the background value. Otherwise, the original value was retained.We made artificial corrections to the data based on the extracted results for the arid areas of the Loess Plateau as well as for the flatter basins, given that these areas do not feature terraces.Training and validation dataFor supervised classification, the selection of sample areas and sample features is crucial. The focus and core of any land classification work is representative and effective training sample selection. To obtain a better sample area selection, we considered the selection of sample areas from three perspectives, i.e., colour texture features, topographic features, and spatial distance of the training samples. First, the terraces in this study are in agricultural land, including cultivated land, woodland, grassland, and other types of land; thus, different types of land will present different texture details. At the same time, high-resolution images from Google Earth are mosaicked. Because of the different acquisition times, the same region and land type will have visible colour differences and stitching traces, which is more common in the Loess Plateau region. Therefore, these factors should be considered in the selection of training samples as much as possible to improve the generability of the model and the correct rate of its extraction. Second, the state of the terraces varies according to topographic features. Among them, gradient, direction, altitude, and climate are the most significant factors. Terraces can be categorised as shallow-slope or steep-slope terraces. Based on slope aspect, altitude, and climate characteristics, they can also be categorised as either easy to identify or hard to identify. Thus, the sample should be inclusive of these types of terraces. According to the first law of geography, terraces in different spatial locations have different morphologies. Therefore, the spatial location of the samples should also be at a certain distance.In summary, we selected one county in each region based on the geomorphic zoning characteristics of the Loess Plateau. In addition, we added one more in the area where the density of terraces may be higher. Finally, we selected the whole area of seven counties (Fig. 3) as the training sample area distribution, covering 2.18% of the overall Loess Plateau area. The colour morphological features, topographic features, spatial location, and imaging quality of terrace images in these regions are highly representative. This method was unique from other classification methods. Most of the traditional methods are based on the single-pixel information of feature layers such as random forests, which tend to ignore the neighbouring information around the point, and thus are subject to misclassification and under classification for land types with outstanding texture information. In our study, we adopted the visual interpretation of the whole domain, which can cover the neighbourhood information of each pixel point more comprehensively. To ensure the uniformity and correctness of visual interpretation, the terraces in the training area were visually interpreted by seven interpreters after uniform professional training. For the disputed and uncertain areas, the seven interpreters carried out interactive interpretation and scoring according to the interpretation results. Finally, two other interpretation experts made the final review and corrections. The interpretation results of the training area were re-examined and revised based on the results of the later interpretations.Fig. 3Distribution of training sample areas and validation sites in terraces on the Loess Plateau.Full size imageTo better assess and compare the validity and correctness of the terraced agricultural area datasets on the Loess Plateau in quantitatively, the validation dataset was divided into two parts: a per-pixel point-based validation set and a field validation dataset of terraces with location information. The extracted datasets were comprehensively evaluated in terms of both pixel scale and field validation.We constructed a single-pixel validation point that evaluates the TDMLP. We applied the Icosahedral Snyder Equal Area Discrete Global Grid created by ArcGIS. Based on this strategy, the study area was partitioned into 972 regions (Fig. 3). To better validate the terrace classification results (excluding non-terrace classes), we placed more validation points within the grid where the terrace distribution is more concentrated. First, we calculated the proportion of terraces in each hexagonal grid to the total area of the hexagonal grid. Second, we separated the terraces into four levels according to the proportion of terraces to the whole grid area as 0–20%, 20–50%, 50–80%, and 80–100% and the number of validation points was 10, 20, 40, and 50, respectively.Since the proportion of the extracted terraced area to the total area was only 14%, direct random point deployment would have led to fewer terraced validation sets and thus would have affected the final data evaluation. Therefore, in the deployment strategy, we ensured that the validation points distributed in the extracted terraces in each grid account for at least one-fifth of the total number of validation points, but for the grid with a smaller proportion of terraces or even 0, this practice was meaningless. Hence, we stipulated that in the grid with a proportion of terraces ≤1%, direct random scattering was to be performed. The final scattered verification points in the terraced and non-terraced areas were 5,194 and 6,226, respectively, with a ratio close to 1:1 for easy verification. The spatial distribution is shown in Fig. 3.We validated the spatial distribution map of terraces on the Loess Plateau from 14 April 2021 to 1 May 2021 and constructed a field validation dataset of terraces with location information. Considering the longitudinal, latitudinal, and vertical heterogeneities of the Loess Plateau, the verification route was divided into two sections, north to south and east to west, to more comprehensively cover all regions of the Loess Plateau. The verification route started at Hohhot in the northeast of the Loess Plateau. It passed through the Datong Basin, followed the Yellow River to the south and the Weihe Plain, and then travelled westward through Mount Liupan to the westernmost part of the Loess Plateau. The route was through 54 counties/districts in 16 cities and six provinces on the Loess Plateau, with a total distance of 3,680 km, covering 15.8% of the counties on the Loess Plateau (total of 341 counties). We also surveyed and sampled the verification points approximately every 5 km along the route and collected data from a total of 815 sample points, covering various types of terraces on the Loess Plateau. The results are shown in Fig. 3. More

A hippo swims in Colombia’s Magdalena River, near where Pablo Escobar’s compound was located.Credit: Fernando Vergara/AP/Shutterstock

Colombian environment minister Susana Muhamad has triggered fear among researchers that she will protect, rather than reduce, a growing population of invasive hippos that threaten the country’s natural ecosystems and biodiversity. Although she did not directly mention the hippos — a contentious issue in Colombia — Muhamad said during a speech in late January that her ministry would create policies that prioritize animal well-being, including the creation of a new division of animal protection.
Landmark Colombian bird study repeated to right colonial-era wrongs
The hippos escaped from drug-cartel leader Pablo Escobar’s estate after he died in 1993. Left alone, the male and three females that Escobar had illegally imported from a US zoo established themselves in Colombia’s Magdalena River and some small lakes nearby — part of the country’s main watershed. After years of breeding, the ‘cocaine hippos’ have multiplied to about 150 individuals, scientists estimate.Given that the hippos (Hippopotamus amphibius) — considered the largest invasive animal in the world — have no natural predators in Colombia and have been mating at a steady rate, their population could reach 1,500 in 16 years, according to a modelling study published in 20211. “I do not understand what the government is waiting for to act,” says Nataly Castelblanco Martínez, a Colombian conservation biologist at the Autonomous University of Quintana Roo in Chetumal, Mexico, and co-author of the study. “If we don’t do anything, 20 years from now the problem will have no solution.”Researchers have called for a strict management plan that would eventually reduce the wild population to zero, through a combination of culling some animals and capturing others, then relocating them to facilities such as zoos. But the subject of what to do with the hippos has polarized the country, with some enamoured by the animals’ charisma and value as a tourist attraction and others concerned about the threat they pose to the environment and local fishing communities.‘A bit surreal’Several studies and observations suggest how destructive it could be to allow the Colombian hippo population to explode. A 2019 paper2, for example, showed that, compared with lakes without hippos, those where the animals have taken up residence contain more nutrients and organic matter that favour the growth of cyanobacteria — aquatic microbes associated with toxic algal blooms. These blooms can reduce water quality and cause mass fish deaths, affecting local fishing communities.

A sign near Doradal, Colombia, warns passersby of the danger of invasive hippos.Credit: Juancho Torres/Anadolu Agency via Getty

Other scientists have predicted that the hippos could displace endangered species that are native to the Magdalena River, such as the Antillean manatee (Trichechus manatus manatus), by outcompeting them for food and space. They caution that traffic accidents and attacks on people caused by the hippos will become more common. And they warn that wildlife traffickers are already taking advantage of the situation by illegally selling baby hippos — a trend that could intensify.“It’s a bit surreal,” says Jorge Moreno Bernal, a vertebrate palaeontologist at the University of the North in Barranquilla, Colombia. “This is just a taste of what may come.”When Colombian authorities first recognized the speed at which the hippo population was growing, during the 2000s, they acted to reduce their numbers. But in 2009, when photos appeared online after soldiers gunned down Pepe, Escobar’s fugitive male hippo, the outcry from animal-rights activists and others plunged the environment ministry into an “institutional paralysis”, says Sebastián Restrepo Calle, an ecologist at Javeriana University in Bogotá.Researchers say that the hippos don’t belong in Colombia — they are native to sub-Saharan Africa. Simulations run by Castelblanco Martínez and her colleagues suggest that to reduce the population to zero by 2033, about 30 hippos would need to be removed from the wild population per year1. No other course of action, including sterilization or castration, would eradicate them, according to the modelling of various management scenarios, says Castelblanco Martínez.The cost of inactionThe worry now is that, instead of basing decisions on evidence and expertise in conservation, the government is listening to popular opinion, says Restrepo Calle. Neither Muhamad nor representatives of the environment ministry replied to Nature’s requests for comment.
Ancient stone tools suggest early humans dined on hippo
“Why prioritize one species over our own ecosystems?” — especially a species that isn’t native, asks Alejandra Echeverri, a Colombian conservation scientist at Stanford University in California. Along with her colleagues, Echeverri published a study last month showing that Colombia has few policies governing invasive species compared with its overall number of biodiversity policies3.Animals-rights advocates, meanwhile, argue that they aren’t ignoring environmental concerns. Luis Domingo Gómez Maldonado, an animal-rights activist and specialist in animal law at Saint Thomas University in Bogotá, says “It’s not about saving the hippos on a whim,” but rather about solving the issue while also giving the hippos justice. “My indisputable position is: let’s save as many individuals as possible, let’s do it ethically.”Researchers, too, say they have the animals’ best interests at heart. “Even if [advocates] don’t see it, we care about the hippos,” Castelblanco Martínez says. “The more time that passes, the more hippos will either have to be culled, castrated or captured.”The question is whether environmental authorities will act swiftly to draft and enforce a management plan that is both ethical and effective. Should they sit on the issue for too long, Castelblanco Martínez warns, rural communities that are most affected by the hippos might take matters into their own hands.If the government doesn’t cull them, she says, people will use shotguns to do it. More