Philippa Kaur

1.Cole, M., Lindeque, P., Halsband, C. & Galloway, T. S. Microplastics as contaminants in the marine environment: A review. Mar. Pollut. Bull. 62, 2588–2597. https://doi.org/10.1016/j.marpolbul.2011.09.025,22001295 (2011).CAS 
Article 
PubMed 

Google Scholar 
2.Cauwenberghe, V. L., Vanreusel, A., Mees, J. & Janssen, C. R. Microplastic pollution in deep-sea sediments. Environ. Pollut. 182, 495–499. https://doi.org/10.1016/j.envpol.2013.08.013,24035457 (2013).Article 
PubMed 

Google Scholar 
3.Cole, M. et al. Microplastic ingestion by zooplankton. Environ. Sci. Technol. 47, 6646–6655. https://doi.org/10.1021/es400663f,23692270 (2013).ADS 
CAS 
Article 
PubMed 

Google Scholar 
4.Rodrigues, J. P., Duarte, A. C., Santos-Echeandía, J. & Rocha-Santos, T. Significance of interactions between microplastics and POPs in the marine environment: A critical overview. TrAC Trends Anal. Chem. 111, 252–260. https://doi.org/10.1016/j.trac.2018.11.038 (2019).CAS 
Article 

Google Scholar 
5.Brennecke, D., Duarte, B., Paiva, F., Caçador, I. & Canning-Clode, J. Microplastics as vector for heavy metal contamination from the marine environment. Estuar Coast Shelf Sci. 178, 189–195. https://doi.org/10.1016/j.ecss.2015.12.003 (2016).ADS 
CAS 
Article 

Google Scholar 
6.Martins, I., Rodríguez, Y. & Pham, C. K. Trace elements in microplastics stranded on beaches of remote islands in the NE Atlantic. Mar. Pollut. Bull. 156, 111270. https://doi.org/10.1016/j.marpolbul.2020.111270 (2020).CAS 
Article 
PubMed 

Google Scholar 
7.Woodall, L. C. et al. The deep sea is a major sink for microplastic debris. R. Soc. Open Sci. 1, 140317. https://doi.org/10.1098/rsos.140317 (2014).ADS 
Article 
PubMed 
PubMed Central 

Google Scholar 
8.Alomar, C., Estarellas, F. & Deudero, S. Microplastics in the Mediterranean Sea: Deposition in coastal shallow sediments, spatial variation and preferential grain size. Mar. Environ. Res. 115, 1–10. https://doi.org/10.1016/j.marenvres.2016.01.005,26803229 (2016).CAS 
Article 
PubMed 

Google Scholar 
9.Wright, S. L., Thompson, R. C. & Galloway, T. S. The physical impacts of microplastics on marine organisms: A review. Environ. Pollut. 178, 483–492. https://doi.org/10.1016/j.envpol.2013.02.031,23545014 (2013).CAS 
Article 
PubMed 

Google Scholar 
10.Fendall, L. S. & Sewell, M. A. Contributing to marine pollution by washing your face: Microplastics in facial cleansers. Mar. Pollut. Bull. 58, 1225–1228. https://doi.org/10.1016/j.marpolbul.2009.04.025,19481226 (2009).CAS 
Article 
PubMed 

Google Scholar 
11.Andrady, A. L. Microplastics in the marine environment. Mar. Pollut. Bull. 62, 1596–1605. https://doi.org/10.1016/j.marpolbul.2011.05.030 (2011). CAS 
Article 
PubMed 

Google Scholar 
12.Pakula, C. & Stamminger, R. Electricity and water consumption for laundry washing by washing machine worldwide. Energy Effic. 3, 365–382. https://doi.org/10.1007/s12053-009-9072-8 (2010).Article 

Google Scholar 
13.Belzagui, F., Crespi, M., Álvarez, A., Gutiérrez-Bouzán, C. & Vilaseca, M. Microplastics’ emissions: Microfibers’ detachment from textile garments. Environ. Pollut. 248, 1028–1035. https://doi.org/10.1016/j.envpol.2019.02.059,31091635 (2019).CAS 
Article 
PubMed 

Google Scholar 
14.Ziajahromi, S., Drapper, D., Hornbuckle, A., Rintoul, L. & Leusch, F. D. L. Microplastic pollution in a stormwater floating treatment wetland: Detection of tyre particles in sediment. Sci. Total Environ. 713, 136356. https://doi.org/10.1016/j.scitotenv.2019.136356 (2020).ADS 
CAS 
Article 
PubMed 

Google Scholar 
15.Fadare, O. O. & Okoffo, E. D. Covid-19 face masks: A potential source of microplastic fibers in the environment. Sci. Total Environ. 737, 140279. https://doi.org/10.1016/j.scitotenv.2020.140279 (2020).ADS 
CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
16.Klein, S., Worch, E. & Knepper, T. P. Occurrence and spatial distribution of microplastics in river shore sediments of the rhine-main area in Germany. Environ. Sci. Technol. 49, 6070–6076. https://doi.org/10.1021/acs.est.5b00492,25901760 (2015).ADS 
CAS 
Article 
PubMed 

Google Scholar 
17.Rummel, C. D., Jahnke, A., Gorokhova, E., Kühnel, D. & Schmitt-Jansen, M. Impacts of biofilm formation on the fate and potential effects of microplastic in the aquatic environment. Environ. Sci. Technol. Lett. 4, 258–267. https://doi.org/10.1021/acs.estlett.7b00164 (2017).CAS 
Article 

Google Scholar 
18.Claessens, M., Meester, S., Landuyt, V. L., Clerck, K. & Janssen, C. R. Occurrence and distribution of microplastics in marine sediments along the Belgian coast. Mar. Pollut. Bull. 62, 2199–2204. https://doi.org/10.1016/j.marpolbul.2011.06.030,21802098 (2011).CAS 
Article 
PubMed 

Google Scholar 
19.Song, Y. K. et al. A comparison of microscopic and spectroscopic identification methods for analysis of microplastics in environmental samples. Mar. Pollut. Bull. 93, 202–209. https://doi.org/10.1016/j.marpolbul.2015.01.015,25682567 (2015).CAS 
Article 
PubMed 

Google Scholar 
20.Lenz, R., Enders, K., Stedmon, C. A., Mackenzie, D. M. A. & Nielsen, T. G. A critical assessment of visual identification of marine microplastic using Raman spectroscopy for analysis improvement. Mar. Pollut. Bull. 100, 82–91. https://doi.org/10.1016/j.marpolbul.2015.09.026,26455785 (2015).CAS 
Article 
PubMed 

Google Scholar 
21.Majewsky, M., Bitter, H., Eiche, E. & Horn, H. Determination of microplastic polyethylene (PE) and polypropylene (PP) in environmental samples using thermal analysis (TGA-DSC). Sci. Total Environ. 568, 507–511. https://doi.org/10.1016/j.scitotenv.2016.06.017,27333470 (2016).ADS 
CAS 
Article 
PubMed 

Google Scholar 
22.Shim, W. J., Song, Y. K., Hong, S. H. & Jang, M. Identification and quantification of microplastics using Nile red staining. Mar. Pollut. Bull. 113, 469–476. https://doi.org/10.1016/j.marpolbul.2016.10.049,28340965 (2016).CAS 
Article 
PubMed 

Google Scholar 
23.Lv, L. et al. A simple method for detecting and quantifying microplastics utilizing fluorescent dyes – safranine T, fluorescein isophosphate, Nile red based on thermal expansion and contraction property. Environ. Pollut. 255, 113283. https://doi.org/10.1016/j.envpol.2019.113283 (2019).CAS 
Article 
PubMed 

Google Scholar 
24.Cauwenberghe, V. L., Devriese, L., Galgani, F., Robbens, J. & Janssen, C. R. Microplastics in sediments: A review of techniques, occurrence and effects. Mar. Environ. Res. 111, 5–17. https://doi.org/10.1016/j.marenvres.2015.06.007 (2015).CAS 
Article 
PubMed 

Google Scholar 
25.Isobe, A. et al. An interlaboratory comparison exercise for the determination of microplastics in standard sample bottles. Mar. Pollut. Bull. 146, 831–837. https://doi.org/10.1016/j.marpolbul.2019.07.033,31426225 (2019).CAS 
Article 
PubMed 

Google Scholar 
26.Uddin, S., Fowler, S. W., Saeed, T., Naji, A. & Al-Jandal, N. Standardized protocols for microplastics determinations in environmental samples from the Gulf and marginal seas. Mar. Pollut. Bull. 158, 111374. https://doi.org/10.1016/j.marpolbul.2020.111374 (2020).CAS 
Article 
PubMed 

Google Scholar 
27.Vermeiren, P., Muñoz, C. & Ikejima, K. Microplastic identification and quantification from organic rich sediments: A validated laboratory protocol. Environ. Pollut. 262, 114298. https://doi.org/10.1016/j.envpol.2020.114298 (2020).CAS 
Article 
PubMed 

Google Scholar 
28.Coppock, R. L., Cole, M., Lindeque, P. K., Queirós, A. M. & Galloway, T. S. A small-scale, portable method for extracting microplastics from marine sediments. Environ. Pollut. 230, 829–837. https://doi.org/10.1016/j.envpol.2017.07.017 (2017).CAS 
Article 
PubMed 

Google Scholar 
29.Ankley, G. T., Di Toro, D. M., Hansen, D. J. & Berry, W. J. Technical basis and proposal for deriving sediment quality criteria for metals. Environ. Toxicol. Chem. 15, 2056–2066. https://doi.org/10.1002/etc.5620151202 (1996).CAS 
Article 

Google Scholar 
30.Japanese Geotechnical Society (JGS) 0131 (JIS A1204). Test method for particle size distribution of soils (2009).31.Japanese Geotechnical Society JGS 0121(JIS A1203). Test method for water content of soils (2009).32.BS ISO 11277: Soil quality. Determination of particle size distribution in mineral soil material. Method by sieving and sedimentation (2009).33.BS ISO 1377:Part 2: Clause 3.2: Determination of the moisture content of soils (1990).34.BS EN 933-2: Tests for geometrical properties of aggregates. Determination of particle size distribution. Test sieves, nominal size of apertures (2020).35.ASTM D6913/D6913M – 17: Standard Test Methods for Particle-Size Distribution (Gradation) of Soils Using Sieve Analysis (2004).36.ASTM D2216 – 19: Standard Test Methods for Laboratory Determination of Water (Moisture) Content of Soil and Rock by Mass (2019).37.Maxwell, S. H., Melinda, K. F. & Matthew, G. Counterstaining to separate Nile red-stained microplastic particles from terrestrial invertebrate biomass. Environ. Sci. Technol. 54, 5580–5588. https://doi.org/10.1021/acs.est.0c00711 (2020).ADS 
CAS 
Article 

Google Scholar 
38.Ehlers, S. M., Maxein, J. & Koop, J. H. E. Low-cost microplastic visualization in feeding experiments using an ultraviolet light-emitting flashlight. Ecol. Res. 35, 265–273. https://doi.org/10.1111/1440-1703.12080 (2020).Article 

Google Scholar 
39.Karakolis, E. G., Nguyen, B., You, J. B., Rochman, C. M. & Sinton, D. Fluorescent dyes for visualizing microplastic particles and fibers in laboratory-based studies. Environ. Sci. Technol. Lett. 6, 334–340. https://doi.org/10.1021/acs.estlett.9b00241 (2019).CAS 
Article 

Google Scholar 
40.Penthala, R. et al. Synthesis of azo and anthraquinone dyes and dyeing of nylon-6,6 in supercritical carbon dioxide. J. CO2 Util. 38, 49–58. https://doi.org/10.1016/j.jcou.2020.01.013 (2020).CAS 
Article 

Google Scholar 
41.Prata, J. C., Costa, J. P., Duarte, A. C. & Rocha-Santos, T. Methods for sampling and detection of microplastics in water and sediment: A critical review. TrAC Trends Anal. Chem. 110, 150–159. https://doi.org/10.1016/j.trac.2018.10.029 (2019).CAS 
Article 

Google Scholar 
42.Hengstmann, E. & Fischer, E. K. Nile red staining in microplastic analysis – proposal for a reliable and fast identification approach for large microplastics. Environ. Monit. Assess. 191, 612. https://doi.org/10.1007/s10661-019-7786-4,31489505 (2019).Article 
PubMed 

Google Scholar 
43.Jung, M. R. et al. Validation of ATR FT-IR to identify polymers of plastic marine debris, including those ingested by marine organisms. Mar. Pollut. Bull. 127, 704–716. https://doi.org/10.1016/j.marpolbul.2017.12.061 (2018).CAS 
Article 
PubMed 

Google Scholar  More

Study areaThe study area (78 000 km2) is located between 68–71°N and 20–26°E, with strong climatic gradients, ranging from wet maritime to relatively dry continental, over tens of kilometers. The landscape of this climatically sensitive high-latitude region has been affected by multiple glaciations in the past. It includes the Scandes Mountains near the Arctic Ocean and low-relief areas to the south and east. The majority of the region (52%) is underlain by sporadic permafrost. Continuous and discontinuous permafrost are limited to the highest mountains of the study area (2% and 7%, respectively)17,26. This large proportion of sporadic, typically warm and shallow permafrost in the study area indicates that ground thermal response to climate warming can be rapid27. Our data do not cover low-relief plateaus of continuous permafrost (similar to northern Siberia and Alaska), where the generally high ice content of soil may lead to different and enhanced LSP responses under climate warming (e.g., ice wedge degradation and surface ponding) with altered AGB feedbacks43,44.LSP observationsThe data consist of 2917 study sites (each 25 m × 25 m) and includes previously combined observations (both in-situ [n = 581] and remote-sensing [n = 2336]) of the active surface features of three cryogenic LSP common in the area: cryoturbation, solifluction, and nivation. These LSP are mainly associated with seasonal freeze–thaw processes. Cryoturbation (i.e., frost churning) is a general term for soil movement caused by differential heave, and it creates typical surface features such as patterned ground, frost boils and hummocky terrain5. Solifluction is the slow mass wasting of surficial deposits through frost creep and permafrost flow, where gravitation causes frost-heaved soil to settle downwards during the summer thaw, creating features of lobes and terraces50. In addition, solifluction also includes gelifluction which is a mass wasting process caused by high porewater pressure in unconsolidated surface debris creating similar lobes and terraces5,50. We use the term nivation to collectively designate various weathering and fluvial processes which are intensified and depicted by the presence of snowbeds (which in general are melting in mid-July – late-August) and nivation hollows28,51. We expect the presence of such a snowbed to be an indication of active nivation processes, since in these environments the year-to-year spatial snow patters are fairly consistent31.The rationale behind LSP sampling is described in previous geomorphic studies which served as a basis for the used protocol52,53. Due to the large study domain, study objectives (focus on distribution of active surface features, not on activity itself) and modeling resolution (50 m × 50 m), we used a visual method to estimate the presence/absence of the mapped LSP. We used high-resolution aerial photography (spatial resolution of 0.25 m−2) and targeted field surveys (GPS accuracy ~5 m; Garmin eTrex personal navigator) to construct the LSP dataset. A binary variable (1 = presence, 0 = absence) was assigned to each LSP to indicate their evident activity (or absence). The activity/absence of the LSP was visually estimated based on the evidence in ground surface, indicated by e.g., frost-heaving, cracking, microtopography (e.g., erosional and depositional forms), soil displacement indicative to a process form (e.g., solifluction lobes, patterned ground), changes in vegetation cover and late-lying snow. Such indicators average the LSP activity over several years. Even small areas with slight indication of activity were considered active processes. However, such a protocol based on a visual assessment is susceptible for incorrect activity classification; solifluction may be active despite having a complete vegetation cover19 and the presence of late-lying snowbed, although being a good indication28, does not necessarily mean that active nivation processes are present.Remotely sensed vegetation indexFor obtaining remotely sensed vegetation index for the study area, we employed a maximum-value compositing approach. We downloaded all available clear sky (less than 80% land cloud cover) Landsat OLI 8 images overlapping the study area from June to September between 2013 and 2017 (total of 1086 scenes) from the United States Geological Survey (USGS) database (http:\earthexplorer.usgs.gov). Images were USGS surface reflectance products, which were preprocessed (georeferencing, projection, and atmospheric corrections) by USGS54. Landsat-8 satellite is the latest addition to the Landsat mission that has provided repeated land surface information globally since the 1970’s and is the most commonly used fine-scale satellite system for vegetation mapping. The native resolution of the Landsat OLI sensor is 30 m for the spectral bands used in the image processing steps of this study.Normalized difference vegetation index (NDVI), a widely used spectral index to estimate the amount of green vegetation, was calculated as55:$$({{{{{rm{rho }}}}}}{{{{{rm{NIR}}}}}}-{{{{{rm{rho }}}}}}{{{{{rm{red}}}}}})/({{{{{rm{rho }}}}}}{{{{{rm{NIR}}}}}}+{{{{{rm{rho }}}}}}{{{{{rm{red}}}}}})$$
(1)
where ρNIR and ρred are the surface reflectance for their respective Landsat bands, 0.851–0.879 (mu)m and 0.636–0.673 (mu)m.USGS provides pixel-based quality assessment bands for all surface reflectance products. These bands were used to mask clouds, snow, water, and other low-quality pixels from the individual NDVI scenes. Additionally, if the NDVI images still had unphysical values over 1 or under -1, these pixels and their surroundings of 100 m radius were excluded. We determined maximum values for each 30 m resolution pixel of the study area individually. After masking cloud, snow, and water from the scenes, obvious scattered erroneous NDVI values remained in some scenes. Therefore, we excluded the values outside the pixel-based 95% percentile prior to maximum composite.The CFmask cloud detection algorithm that is used to generate the quality assessment band has clear difficulties in distinguishing small snow patches from clouds. As such, a large portion of late-lying snow beds were repeatedly and incorrectly classified as clouds. Moreover, the CFmask algorithm creates buffers around the cloud pixels54, hence much information was lost around the snow patches that were incorrectly identified as clouds. After these processing steps, some pixels around the extreme late-lying snow beds had still too low number of NDVI records to provide reliable NDVI values for the maximum composite. To fill these small and scattered gaps in the initial maximum NDVI composite, we selected 74 mostly cloud-free scenes between August and September. For these 74 scenes, we manually digitized cloud masks to exclude cloud-contaminated pixels with high certainty. Moreover, every pixel must have passed the following quality checks to be included in the gap-filling composite: not classified as water in the USGS quality assessment band; normalized difference snow index (NDSI) value less than 0.4, and blue band reflectance less than 0.1 (to exclude snow); reflectance of red band between 0.03 and 0.4 (second check for water and snow, and deepest shadows); NDVI between 0 and 0.4 (lower threshold to exclude snow and water contamination; higher threshold to exclude erroneous values, as very late snowbed habitats always have very limited vegetation cover). Additionally, if the NDVI images had unphysical values over 1 or under -1, these pixels and their surroundings (200 m radius) were excluded. Pixels in the 74 selected images which passed these checks, were then used to create a secondary maximum NDVI composite that was used to fill the gaps in the initial maximum NDVI composite. The secondary composite comprised 0.4% of the pixels in the final composite. Among all 2917 LSP observation sites, 2.9% were located within the gaps in the initial maximum NDVI composite, and thus received their maximum NDVI values from the secondary NDVI composite.In the used Landsat data, the nivation sites were not covered by snow, but instead were associated with generally lower AGB values as nival processes affect the vegetation’s structure and composition (Supplementary Table 1).Above-ground biomass dataAbove-ground biomass (AGB) reference data were collected from two regions, with a total of 433 sites that represent an area of > 4000 km2 (Supplementary Fig. 9). The first dataset (hereafter BM region 1; centering to ca. 69°N, 21°E) was collected between 2008 and 2011, and the second dataset (BM region 2; centering to ca. 70°N, 26.2°E) between 2015 and 2017. Both study regions are representative of an arctic and alpine treeline ecotone and include data from mountain birch forest to barren oroarctic tundra56,57.The BM region 1 dataset consists of 309 field sites (each 10 m × 10 m), which are located around eight different massifs covering a wide range of environmental conditions (Supplementary Figs. 9–10). Sampling was performed in transects to cover various aspects of the slope (i.e., topoclimatic conditions), starting from the foothill of the mountain, and ending at the summit. A plot was systematically established at every 20 m increase in elevation and recorded with a GPS device. Four clip-harvest biomass samples (20 cm × 20 cm) were taken 5 m from the plot center in every cardinal direction. Two samples were used in bare mountaintops (north, south). The clip-harvest samples were dried for 48 h at +65 °C, and dry weight was recorded. The sample biomass values were converted to g m-2 and the average sample value was calculated for each site (Supplementary Fig. 9). The original BM region 1 dataset contains forest and treeline plots, but these were excluded from the final analyses due to an incomparable tree sampling strategy with BM region 2, which could introduce uncertainty into biomass estimates.The BM region 2 data were collected from three different massifs having an elevation range from 120 m to 1064 m (Supplementary Fig. 9). The biomasses were sampled from 102 sites (each 24 m × 24 m in size) that were chosen using a stratified sampling to cover gradients of thermal radiation (potential incoming solar radiation), soil moisture (topographic wetness index, TWI) and vegetation zone (forest, treeline, and alpine zones). Radiation and TWI were calculated from a 10 m digital elevation model (DEM, provided by the National Land Surveys of Finland and Kartverket, the Norwegian mapping authority), and assigned to one of three classes based on observation percentiles (breaks at 20% and 80%) leading to total of 27 strata. Vegetation zones were digitized based on aerial imagery. After the first field survey, 22 sites were added to account for vegetation types that were not sufficiently represented by the GIS-based stratification. Thus, the total sample size of the BM region 2 dataset is 124 AGB sites.The same clip-harvest sample protocol was used as in BM region 1; additional samples were also taken from 12 m in every cardinal direction, thus each site had eight AGB samples (Supplementary Fig. 9). Trees with diameter at breast height (DBH) greater than 20 mm were measured from a 900 m2 circular plot, which corresponds to the size of the NDVI product resolution. Large stems (DBH  > 80 mm in the forest and 40 mm at the treeline) were measured from the whole plot, whereas smaller stems were measured from five subplots. Specifically, the center subplot was 100 m2, and the four subplots located at 8 m to every cardinal direction were each 12.5 m2. For the subplot observations, we used a plot expansion factor (900/150 = 6) to generalize the observations for the whole plot assuming a homogeneous forest structure i.e., each subplot stem represents six trees within the 900 m2 plot. A total of 98% of the measured stems were mountain birch (Betula pubescens ssp. czerepanovii), making it the most abundant species in the area. For predicting stem biomass, we used the average of three allometric equations58,59,60, in order to reduce the uncertainty related to the transferability of an individual allometric model. In addition, Populus tremula (1% of the observations) were found on low-altitude south-facing slopes, and Salix caprea (1%) in moist, nutrient-rich sites. Species-specific models61,62 were used to estimate their respective stem biomasses. Individual pines (Pinus sylvestris) were scattered in the area but were not present in any of the sampled plots.The plots of above-ground tree biomass were converted to g m−2 and added to the mean clip-harvest AGB to obtain the total vascular plant AGB for each site. The BM region 1 and BM region 2 datasets were combined, and the NDVI value was extracted from the site center coordinates.Spatial autocorrelation (SAC) is a common property of any spatial dataset and means that observations are related to one another by the geographical distance63. SAC in the model residuals violates the independence assumption commonly required by statistical models and can lead to inflated hypothesis testing and biased model estimates64. To investigate whether the plot-scale AGB data are spatially autocorrelated, we calculated semivariogram which describes the spatial dependency between the observations as a function of distance between the point pairs65. Semivariogram were calculated as:$${{{{{rm{gamma }}}}}}(h)=frac{1}{2N(h)}mathop{sum }limits_{i=1}^{{N}_{h}}{left(Zleft({s}_{i}right)-Zleft({s}_{i}+hright)right)}^{2}$$
(2)
where N(h) denotes the number of data pairs within distance h, and (Zleft({s}_{i}right)) is an observation (or model residual) in location i. For the calculation, we used R package gstat66 (version 2.0-0). A visual inspection of the semivariogram indicated spatial autocorrelation at short distances (“AGB” in Supplementary Fig. 11). Therefore, for the NDVI-AGB conversion, we used a generalized least squares modeling (GLS, as implemented in R package nlme67 [version 3.1-137]) that can explicitly account for SAC in the data. For the modeling, the AGB values were log(x+0.1) transformed. The GLS, where AGB was modeled as a function of NDVI, were fitted assuming an exponential spatial correlation structure:$$gamma left(hright)={c}_{0}+cleft(1-{e}^{-h/a}right)$$
(3)
where ({c}_{0}) is the difference between the intercept and origin (i.e., the “nugget” parameter in geostatistics), (c) is the amount of variance (i.e., the “sill”) and a represents the distance of spatial dependency (i.e., the “range”). The fitted GLS was as follows:$${{log }}left({AGB}right)=-1.038629+9.725572times {NDVI}$$
(4)
The estimated spatial correlation parameters were c0 = 0.516, c = 0.484 and a = 260.605, indicating that the distance of spatial autocorrelation extends to ca. 261 m. The semivariogram for the model residuals indicated a notable reduction in the amount of spatial autocorrelation compared to the AGB data (Supplementary Fig. 11). The fitted model explained 70.6% of the deviance in the data. When the predicted values were converted back to the response scale, the model explained 60.5% of the deviance. Therefore, for the subsequent analyses we use the above-ground biomass estimated by the model.Environmental predictorsIn addition to LSP, we used climate, topography, and soil predictors to model AGB. Gridded monthly average temperatures and precipitation data (1981–2010; spatial resolution 50 m × 50 m) based on a large collection ( > 950) of Fennoscandian meteorological stations were used in a spatial interpolation scheme17. Three climate predictors—growing degree days (GDD, °C, base temperature 5 °C), mean February air temperature (Tfeb, °C) and water balance (WAB, mm)—were calculated from the gridded climate data. WAB is the difference between total annual precipitation and potential evapotranspiration (PET), which was estimated from the monthly air temperature and precipitation data68:$${PET}=58.93times {T}_{{above}0^circ C}/12$$
(5)
These climatic predictors were selected to represent different aspects of climate that are critical for tundra vegetation: heat requirements, cold tolerance and moisture availability. In addition, two local scale topographic predictors were calculated from a DEM (spatial resolution of 50 m × 50 m, provided by the National Land Survey Institutes of Finland, Norway, and Sweden): topographic wetness index69 (TWI, a proxy for soil moisture) and potential annual direct solar radiation70 (MJ cm-2 a-1). Slope angle was initially considered as a potential predictor for AGB but was later omitted due to the strong correlation with TWI (-0.93, P ≤ 0.001). We also calculated peat cover (%) from a digital land cover classification71. Here, the native resolution of 100 m was resampled at 50 m to match the resolution of the climatic and topographic predictors, using nearest-neighbor interpolation. The binary peat cover variable was transformed to a continuous scale using a spatial mean filter of 3 × 3 pixels52. Finally, the topmost soil layer of a global gridded soil database72 was used to obtain pH data. Again, the original resolution of 250 m was also resampled to 50 m resolution using bilinear interpolation.Our fine-scale data revealed strong environmental gradients over the 78,000 km2 study area (Supplementary Table 1), most of which were only moderately inter-correlated (Spearman’s correlation coefficient  More

Airborne hyperspectral and thermal image acquisitionWe scanned over one million olive and almond trees between 2011 and 2019 with an airborne imaging spectroscopy and thermal imaging facility targeting infected and healthy trees in seven different regions located in Apulia (Italy), Majorca (Balearic Islands, Spain), Alicante, Cordoba and Seville (mainland Spain). In olive groves, over 200,000 and 372,000 trees were imaged from Xf and Vd outbreaks, respectively. In almond groves, we scanned over 132,000 trees from Xf outbreaks in Alicante and Majorca. To evaluate the effects induced by abiotic stress on spectral plant traits, we surveyed over 370,000 healthy trees (outside the outbreak areas) comprising olive and almond species subjected to a wide range of water stress conditions.We surveyed these areas with airborne hyperspectral and thermal cameras on board a manned aircraft flying at 500 m altitude above ground, yielding 40 cm and 60 cm spatial resolution, respectively. We used a hyperspectral camera (VNIR model, Headwall Photonics, Fitchburg, MA, USA) collecting 260 bands in the 400–885 nm region at 1.85 nm/pixel and 12-bit radiometric resolution with a frame rate of 50 Hz. With this spectral configuration, we captured imagery at 6.4 nm full-width at half-maximum (FWHM) bandwidth and obtained an instantaneous field of view (IFOV) of 0.93 mrad and an angular field of view (FOV) of 49.82 deg with an 8 mm focal length lens. The hyperspectral sensor was radiometrically calibrated in the laboratory using an integrating sphere (CSTM-USS-2000C Uniform Source System, LabSphere, North Sutton, NH, USA). At the time of flight, we measured aerosol optical thickness at 550 nm using a Sunphotometer (Microtops II S model 540, Solar LIGHT Co., Philadelphia, PA, USA). We then applied the resulting atmospheric correction of the calibrated radiance imagery with the SMARTS model51 to derive surface reflectance spectra. We carried out ortho-rectification of the hyperspectral imagery (PARGE, ReSe Applications Schläpfer, Wil, Switzerland) with readings acquired by the inertial measuring unit on board the airborne platform (IG500 model, SBG Systems, France). We applied spatial binning through object-based image analysis, thus increasing the signal-to-noise ratio (SNR) of the instrument. Additionally, we conducted spectral binning to reduce the number of spectral bands (260 bands at 1.85 nm resolution). SNR reached >300:1 after binning. We acquired high-resolution tree-crown temperature images with a thermal camera (FLIR SC655, FLIR Systems, USA) at 640 × 480 pixels resolution using a 24.6 mm f/1.0 lens, sensitive to the 7.5–14 μm spectral range and sensitivity below 50 mK.We identified individual trees in the high-resolution hyperspectral and thermal images using object-based crown detection and segmentation methods52. We then calculated the mean hyperspectral radiance, reflectance and temperature for each pure tree crown within every orchard under evaluation. We based our image segmentation methods on Niblack53 and Sauvola and Pietikäinen54, which allowed the isolation of tree crowns from the soil and shadow components. The segmentation of each tree crown was assessed visually to ensure a minimum number of pure vegetation pixels were selected within each tree crown and also spectrally to evaluate the purity of the reflectance extracted from the crown to avoid spectral mixture with soil, shadows and background components24,35.Collection of Xf and Vd biotic stress field dataField assessments of Xf- and Vd-infected trees were carried out from outbreaks affecting olive and almond species in the indicated regions of Italy and Spain between 2011 and 201924,35,52. During these campaigns, we performed quantitative PCR (qPCR)55 for Xf in olive and almond (Alicante), recombinase-polymerase-amplification (RPA) using the AmplifyRP XRT + test (Agdia®, Inc., Elkhart, IN)56 for Xf in almond (Majorca) or conventional PCR57 assays for Vd, as well as visual assessments in individual trees of disease incidence (DI) and disease severity (DS). A sample was considered positive if Ct values were ≤36 and amplification curves were exponential. PCR/qPCR data for model analysis were transformed to 0 and 1, for negative and positive results, respectively, and Ct values were not used in the analysis (see Supplementary Table 2 for the PCR/qPCR primer sequences for Vd and Xf). DS was scored using a 0–4 rating scale according to the percentage of the tree crown showing disease symptoms.In Apulia, the Xf-olive database comprised a total of 15 olive groves surveyed during the June 2016 and July 2017 campaigns. Visual assessments for infection were conducted on 7296 trees (3324 in 2016 and 3972 in 2017). In 2016, 1886 symptomatic (and 1438 asymptomatic) trees were surveyed (762 trees labelled as DS = 1; 802 DS = 2; 250 DS = 3 and 72 DS = 4). In 2017, 1365 were reported as symptomatic (and 2607 asymptomatic) (686 DS = 1; 542 DS = 2; 122 DS = 3 and 15 DS = 4). qPCR assays were carried out to diagnose Xf infection in 77 olive trees, whereby 39 trees tested negative (qPCR = 0) and 38 tested positive (qPCR = 1).On the island of Majorca and at the Alicante province, the field-based Xf-almond database comprised a total of 19 almond groves surveyed in 2018 and 2019, respectively. In Alicante, the field surveys covered 83 ha with 9 almond groves consisting of 943 almond trees. During the field campaigns, almond trees were visually assessed to evaluate Xf-induced DI and DS indices. From this analysis, we identified 593 symptomatic trees and 350 asymptomatic trees. Out of all symptomatic trees, 163 were rated as DS = 1, 214 DS = 2, 157 DS = 3, and 59 DS = 4. Furthermore, qPCR analysis was carried out on 226 almond trees to diagnose Xf infection, resulting in 48 non-infected (qPCR = 0) almond trees and 178 infected trees (qPCR = 1). In Majorca, field surveys in July 2019 covered a total of 2803 ha and comprised 10 almond groves. During the field campaigns, visual observations were carried out on over 4048 almond trees to assess DI and DS, yielding 1387 symptomatic and 2661 asymptomatic trees. From symptomatic trees, 537 were rated as DS = 1449 DS = 2, 359 DS = 3 and 42 DS = 4. We conducted AmplifyRP XRT + assays on 265 almond trees for diagnosing Xf infection the same day they were sampled and identified 141 negative trees (qPCR = 0) and 124 positive trees (qPCR = 1).We carried out physiological measurements of leaf chlorophyll, anthocyanins, flavonoids and nitrogen contents with a Dualex Scientific + (Force-A, Orsay, France) instrument as well as leaf reflectance (400–1000 nm spectral range) and steady-state chlorophyll fluorescence (Ft) using the PolyPen RP400 and FluorPen FP100 instruments (Photon Systems Instruments, Drasov, Czech Republic) during the field evaluations of almond and olive groves conducted in Majorca, Alicante and Apulia regions. In the Xf-olive study site in Apulia, we generated 1023 leaf measurements with Dualex, 1543 single leaf reflectance spectra, as well as 1402 Ft readings over 67 olive trees. In the Xf-almond study sites in Majorca, we measured 1242 leaves with Dualex, 1094 leaves with the PolyPen and 1218 with the Fluorpen instruments from 87 almond trees across a wide range of disease severity levels. For the Xf-almond study sites located at Alicante, we conducted 1649 leaf measurements with Dualex, 632 leaf measurements with PolyPen and 563 leaf measurements with FluorPen FP100 over 43 almond trees.We assessed Vd-infected olive trees from 11 olive groves by surveying an area of over 3000 ha in Castro del Rio and Ecija, southern Spain, in 2011 and 2013, respectively. In Castro del Rio, we conducted visual assessments in an infected area of 96 ha comprising 1878 olive trees, thus identifying 1569 asymptomatic and 283 symptomatic olive trees. Out of the 283 symptomatic trees, 218 were rated as DS = 1; 45 DS = 2; 12 DS = 3 and 8 DS = 4. We measured leaf Fs and Fm’ fluorescence parameters from 25 leaves per tree using a PAM-2100 Pulse-Amplitude Modulated Fluorometer (Heinz Walz GMBH, Effeltrich, Germany). In addition, leaf PRI570 was measured from 25 leaves per tree using a custom-made PlantPen device (Photon System Instrument, Drasov, Czech Republic). Finally, we measured leaf conductance (Gs) on five leaves per tree using a leaf porometer (model SC-1, Decagon Devices, Washington, DC, USA). In the Écija region, the surveyed area covered 3424 ha, and 5223 olive trees were evaluated. We performed visual assessment to determine DI and DS indices of Vd-infected trees, identifying 5040 asymptomatic olive trees. Of the remaining 183 olive trees that were symptomatic, 112 were trees rated as DS = 1; 41 DS = 2; 22 DS = 3 and 8 DS = 4.Trees were evaluated for disease severity and incidence by visual assessment in each outbreak region. PCR assays were carried out on a subset of these trees within each orchard to (i) validate that the pathogen (Xf or Vd) was actually present and the biotic source of symptoms; and (ii) validate that asymptomatic (DS = 0) but infected (PCR = 1) trees were detected using the hyperspectral plant traits estimated through the methodology described in this paper. In general, PCR assays are (i) time consuming and costly, and (ii) difficult to make in large infected trees due to the non-uniform distribution of the infection within each tree crown. These PCR data for each tree along with the field evaluations of DS, DI and non-destructive physiological measurements derived for each tree within every orchard were matched with the high-resolution hyperspectral images to build the biotic databases used in this study. We carried out the field work at each orchard guiding the evaluations and measurements using a high-resolution image to map the location of each tree within the orchard. Due to the planting grids typical of almond and olive species, which were not contiguous or in row-structured patterns, the identification of each individual tree in the images was straightforward.Collection of abiotic stress field dataWe monitored over 3600 ha of olive and almond groves located outside any infected area in Cordoba and Seville, Southern Spain. We performed a multitemporal analysis to study the spectral plant-trait alterations induced by abiotic stress relative to healthy olive and almond trees with data we collected over a 468 ha area comprising two olive and one almond groves throughout July 2016 and August 2017 growing seasons. We analysed 2975 olive and 1964 almond trees in 2016, and 2865 olive and 2063 and almond trees in 2017. At both study sites, we monitored the midday stem water potential (SWP) using a pressure chamber (Soil Moisture Equipment Corp. model 3000, Santa Barbara, CA, USA) on 16 trees per grove. SWP values showed differences between two existing irrigation levels (well-watered and mild water stress), averaging –1.7 and –1.9 MPa across the season in the case of almonds. In olive, SWP in one of the groves reached –3.8 and –3.5 MPa. In 2017, water potential levels averaged –2.9 and –2.0 MPa. In the second grove, irrigation levels were higher, reaching an average SWP of –1.5 MPa. We used an additional study site located in Casariche (Seville province), southern Spain, to validate the results obtained from the multitemporal analysis. This study site covered 3371 ha containing 55 olive groves grown under irrigated and rainfed conditions, with 21,071 olive trees used for statistical analysis.The multitemporal dataset was used to evaluate the water-induced abiotic stress by quantifying the evolution of the importance of the most sensitive spectral traits by clustering non-stressed trees (C0) against groups of trees exposed to increasing levels of water stress (C1 to C4). The multitemporal component of this assessment enabled the evaluation of every single tree across time, therefore selecting the trees for each cluster based on a sustained water stress level, avoiding the selection of trees under short-term stress dynamics. Thus, the clusters were determined based on their CWSI levels, and only the trees with stable water stress levels across two consecutive years (2016 and 2017) were selected for the analysis. For this purpose, we did not include trees that deviated beyond 95% of the CWSI differences calculated between the first and second year in the analysis. After this trimming step, we retained 5484 olive trees (from 5566 trees) and 3652 almond trees (from 3882 almond trees). Trees were then grouped through CWSI clustering analysis using a modified three-sigma rule58. This rule describes the density of a distribution within standard deviation bands on both sides of the mean point into the 68th, 95th and 99.7th percentiles58, representing µ ± σ, µ ± 2σ and µ ± 3σ, respectively. The first interval defined by the classic three-sigma rule (µ ± σ) represented most trees, while the third interval (µ ± 3σ) consisted of very few trees, raising issues for the determination of statistical significance analysis. Based on this observation, we adjusted the breakpoints between groups as follows: we classified those trees that were in the lowest 10th percentile as C0. Trees between the 10th and 68th percentiles (µ + σ) were classified as C1, trees between the 68th and 85th percentile were classified as C2, trees between the 85th and 95th percentile were classified as C3 and finally the trees over the 95th (µ + 2σ) percentile were classified as C4. We thus selected 488 C0, 3066 C1, 1090 C2, 618 C3 and 222 C4 olives trees. Likewise, we grouped almond trees into 390 C0, 1776 C1, 1248 C2, 214 C3 and 24 C4 clusters. Moreover, the analysis of the contribution of a given trait was performed using ML modelling strategies to classify unstressed trees against the clusters defined above that were exposed to increasing levels of water stress. Furthermore, we assessed the consistency of the obtained indicators by performing the classification between stressed and non-stressed trees at an independent olive study site. For this purpose, we evaluated our predictors and compared their contribution over an additional site (Casariche).Model inversion methods for plant-trait estimationWe quantified chlorophyll content (Ca+b), carotenoid content (Cx+c), anthocyanin content (Anth.), mesophyll structure (N), leaf area index (LAI) and average leaf angle (leaf inclination distribution function or LIDF) by radiative transfer model inversion of PROSPECT-D59 and 4SAIL60, as in Zarco-Tejada et al.24. We inverted PROSPECT-D + 4SAIL using a look-up-table (LUT) generated with randomised input parameters. The LUT was generated with 100,000 simulations within fixed ranges (Supplementary Table 3). We implemented a wavelet analysis61 into six wavelets by a Gaussian kernel, estimating the parameters in the top 1% entries ranking the lowest root mean square error (RMSE) values. We then retrieved each plant trait independently by training supported vector machine (SVM) algorithms using the simulated reflectance data as input. We built SVMs in Matlab (MATLAB; Statistics and Machine Learning toolbox and Deep Learning toolbox; Mathworks Inc., Matick, MA, USA) using a Gaussian kernel (radial basis function) with hyperparameters optimised for each model. The training processes were carried out in parallel using the Matlab parallel computing toolbox. With these trained models, we then used the spectral reflectance extracted from the delineated crowns (as show in Fig. 1) to predict plant traits for each individual tree at each study site. The model inversions were carried out for each tree using the crown reflectance. The latter was calculated as an average across all the pixels belonging to the tree crown, delineated using segmentation. This method52 avoids the problem of pixels from within-crown shadows, from tree edges or from sunlit or shaded soil background affecting the spectra, as it retrieves the plant traits from pure sunlit vegetation components of the trees. We also calculated narrow-band spectral indices from reflectance spectra (Supplementary Table 1), which are sensitive to leaf traits and potentially related to disease-induced symptoms. Tree-crown radiance and temperature were used to calculate sun-induced chlorophyll fluorescence at 760 nm (SIF@760) and the crop water stress index (CWSI)37. SIF@760 was quantified using the O2-A in-filling Fraunhofer Line Depth (FLD) method63 and CWSI was calculated by incorporating the tree temperature and the weather data obtained at each study site37.Statistical analysisWe implemented random forest (RF)64 algorithms to classify healthy vs. infected (biotically stressed) trees, and non-stressed vs. water (i.e. abiotically) stressed trees for both tree species. RF algorithms have been widely used in remote sensing studies since they have shown excellent classification accuracies and high processing speeds with high-dimensional data62 and have shown to be accurate in detection of several diseases29,65,66,67. The spectral plant traits estimated by radiative transfer model inversion (Ca+b, Cx+c, Anth., LAI and LIDF), CWSI and SIF@760 were used as inputs for the models. In addition, using a recursive feature elimination approach68 the narrow-band indices that improved the classification in terms of overall accuracy (OA) and kappa coefficient (κ) were added to the models. The pool of narrow-band indices was reduced based on a variance inflation factor (VIF) analysis69 to avoid collinearity among the input features.The RF algorithms were built in Matlab and the hyperparameters were optimised using Bayesian optimisation. The importance of a feature using the RF algorithm was assessed based on the permutation of out-of-bag (OOB) predictor methodology70. To compare the relative differences of the spectral traits in classification of the biotic and abiotic stress, the importance was normalised by dividing the importance of each trait by the highest contribution obtained for each pathogen/species. For the RF models, 500 iterations were run by randomly partitioning each dataset into training (80% of samples) and testing sets (20% of samples). For the training subset, a balanced number of samples from each class was randomly selected at each iteration. The importance obtained by the OOB permutation algorithms was used to build a feature-weighted random forest algorithm (based on Liu and Zhao45), accounting for the importance of each variable on the classification process, evaluating the model against PCR data and visual observations for each biotic stress dataset in terms of OA and κ levels.Probabilities of the predictions were obtained for each sample71 and the uncertain trees were assessed. To extract the uncertainty for each individual tree on the classification, we evaluated the probability distribution for each class from each dataset independently. Then, those trees with a classification probability below the 68th percentile (µ [mean] + σ [standard deviation]) were considered as uncertain and incorporated into a second-stage classification process. The second stage consisted of an unsupervised graph theory–based spectral clustering algorithm72 and included traits selected by focusing on the divergent biotic–abiotic stress obtained from the biotic and the abiotic stress databases. Spectral clustering was performed in R using the kernlab package73.To determine the spectral traits that differed between Xf- and Vd-infected plants and those from the abiotic pathway, we first normalised the importance of the specific traits independently. Then, we compared the common traits between abiotic and biotic stress sets, selecting only biotic stress-related traits that differed in ratio by >0.5 over their homologous abiotic stress trait values. Traits that were only expressed under biotic stress conditions and that showed a normalised importance over 0.5 were included for the second-stage classification process only including those divergent-specific biotic and abiotic stress-related spectral traits as inputs. Specifically, NPQI, Anth. and SIF@760 were considered for the classification of Xf-infected olive trees. Ca+b, SIF@760 and PRIn were used for classifying Xf-infected almond trees. Furthermore, NPQI, Anth. and B spectral traits were selected for classifying uncertain Vd-infected olive trees. Finally, we validated our feature-weighted methodology coupled with the second-stage spectral clustering method against qPCR assays and visual assessment of symptom severity.Reporting summaryFurther information on research design is available in the Nature Research Reporting Summary linked to this article. More

1.Lande, R. & Shannon, S. The role of genetic variation in adaptation and population persistence in a changing environment. Evolution 50, 434–437 (1996).PubMed 
Article 
PubMed Central 

Google Scholar 
2.Barrett, R. D. & Schluter, D. Adaptation from standing genetic variation. Trends Ecol. Evol. 23, 38–44 (2008).PubMed 
Article 
PubMed Central 

Google Scholar 
3.Young, A., Boyle, T. & Brown, T. The population genetic consequences of habitat fragmentation for plants. Trends Ecol. Evol. 11, 413–418 (1996).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
4.Cushman, S. A. Effects of habitat loss and fragmentation on amphibians: a review and prospectus. Biol. Conserv. 128, 231–240 (2006).Article 

Google Scholar 
5.Opdam, P. & Wascher, D. Climate change meets habitat fragmentation: linking landscape and biogeographical scale levels in research and conservation. Biol. Conserv. 117, 285–297 (2004).Article 

Google Scholar 
6.Broadhurst, L. M. et al. Seed supply for broadscale restoration: maximizing evolutionary potential. Evol. Appl. 1, 587–597 (2008).PubMed 
PubMed Central 
Article 

Google Scholar 
7.Vitt, P., Havens, K., Kramer, A. T., Sollenberger, D. & Yates, E. Assisted migration of plants: changes in latitudes, changes in attitudes. Biol. Conserv. 143, 18–27 (2010).Article 

Google Scholar 
8.Aitken, S. N. & Bemmels, J. B. Time to get moving: assisted gene flow of forest trees. Evol. Appl. 9, 271–290 (2016).PubMed 
Article 
PubMed Central 

Google Scholar 
9.Evans, B. J. et al. Speciation over the edge: gene flow among non-human primate species across a formidable biogeographic barrier. R. Soc. Open Sci. 4, 170351 (2017).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
10.Weeks, A. R. et al. Assessing the benefits and risks of translocations in changing environments: a genetic perspective. Evol. Appl. 4, 709–725 (2011).PubMed 
PubMed Central 
Article 

Google Scholar 
11.Pavlova, A. et al. Severe consequences of habitat fragmentation on genetic diversity of an endangered Australian freshwater fish: a call for assisted gene flow. Evol. Appl. 10, 531–550 (2017).PubMed 
PubMed Central 
Article 

Google Scholar 
12.Aitken, S. N. & Whitlock, M. C. Assisted gene flow to facilitate local adaptation to climate change. Annu. Rev. Ecol. Evol. Syst. 44, 367–388 (2013).Article 

Google Scholar 
13.Rajpurohit, S. & Nedved, O. Clinal variation in fitness related traits in tropical drosophilids of the Indian subcontinent. J. Therm. Biol. 38, 345–354 (2013).Article 

Google Scholar 
14.Kawecki, T. J. & Ebert, D. Conceptual issues in local adaptation. Ecol. Lett. 7, 1225–1241 (2004).Article 

Google Scholar 
15.Kottler, E. J., Dickman, E. E., Sexton, J. P., Emery, N. C. & Franks, S. J. Draining the swamp hypothesis: little evidence that gene flow reduces fitness at range edges. Trends Ecol. Evol. https://doi.org/10.1016/j.tree.2021.02.004 (2021).16.Kelly, E. & Phillips, B. L. Targeted gene flow for conservation. Conserv. Biol. 30, 259–267 (2016).PubMed 
Article 
PubMed Central 

Google Scholar 
17.Macdonald, S. L., Llewelyn, J., Moritz, C. & Phillips, B. L. Peripheral isolates as sources of adaptive diversity under climate change. Front. Ecol. Evol. 5, 88 (2017).Article 

Google Scholar 
18.Edmands, S. Between a rock and a hard place: evaluating the relative risks of inbreeding and outbreeding for conservation and management. Mol. Ecol. 16, 463–475 (2007).PubMed 
Article 
PubMed Central 

Google Scholar 
19.Edmands, S. Heterosis and outbreeding depression in interpopulation crosses spanning a wide range of divergence. Evolution 53, 1757–1768 (1999).PubMed 
Article 
PubMed Central 

Google Scholar 
20.Frankham, R. et al. Predicting the probability of outbreeding depression. Conserv. Biol. 25, 465–475 (2011).PubMed 
Article 
PubMed Central 

Google Scholar 
21.Whiteley, A. R., Fitzpatrick, S. W., Funk, W. C. & Tallmon, D. A. Genetic rescue to the rescue. Trends Ecol. Evol. 30, 42–49 (2015).PubMed 
Article 
PubMed Central 

Google Scholar 
22.Schierup, M. H. & Christiansen, F. B. Inbreeding depression and outbreeding depression in plants. Heredity 77, 461–468 (1996).Article 

Google Scholar 
23.Bjorkman, A. D., Vellend, M., Frei, E. R. & Henry, G. H. Climate adaptation is not enough: warming does not facilitate success of southern tundra plant populations in the high Arctic. Glob. Change Biol. 23, 1540–1551 (2017).Article 

Google Scholar 
24.Frankham, R. Where are we in conservation genetics and where do we need to go? Conserv. Genet. 11, 661–663 (2010).Article 

Google Scholar 
25.Tallmon, D. A., Luikart, G. & Waples, R. S. The alluring simplicity and complex reality of genetic rescue. Trends Ecol. Evol. 19, 489–496 (2004).PubMed 
Article 
PubMed Central 

Google Scholar 
26.Weeks, A. R. et al. Genetic rescue increases fitness and aids rapid recovery of an endangered marsupial population. Nat. Commun. 8, 1071 (2017).PubMed 
PubMed Central 
Article 
CAS 

Google Scholar 
27.Le Cam, S., Perrier, C., Besnard, A.-L., Bernatchez, L. & Evanno, G. Genetic and phenotypic changes in an Atlantic salmon population supplemented with non-local individuals: a longitudinal study over 21 years. Proc. Roy. Soc. B-Biol. Sci. 282, 20142765 (2015).Article 
CAS 

Google Scholar 
28.Fitzpatrick, S. W. et al. Gene flow from an adaptively divergent source causes rescue through genetic and demographic factors in two wild populations of Trinidadian guppies. Evol. Appl. 9, 879–891 (2016).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
29.Robinson, Z. L. et al. Experimental test of genetic rescue in isolated populations of brook trout. Mol. Ecol. 26, 4418–4433 (2017).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
30.Byrne, P. G. & Silla, A. J. An experimental test of the genetic consequences of population augmentation in an amphibian. Conserv. Sci. Pract. 2, e194 (2020).31.Stuart, S. N. et al. Status and trends of amphibian declines and extinctions worldwide. Science 306, 1783–1786 (2004).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
32.Urban, M. C., Richardson, J. L. & Freidenfelds, N. A. Plasticity and genetic adaptation mediate amphibian and reptile responses to climate change. Evol. Appl. 7, 88–103 (2014).PubMed 
Article 
PubMed Central 

Google Scholar 
33.Carey, C. & Alexander, M. A. Climate change and amphibian declines: is there a link? Divers. Distrib. 9, 111–121 (2003).Article 

Google Scholar 
34.Parmesan, C. & Yohe, G. A globally coherent fingerprint of climate change impacts across natural systems. Nature 421, 37–42 (2003).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
35.Pounds, J. A. et al. Widespread amphibian extinctions from epidemic disease driven by global warming. Nature 439, 161–167 (2006).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
36.Thomas, C. D. et al. Extinction risk from climate change. Nature 427, 145 (2004).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
37.Rudin-Bitterli, T. S., Evans, J. P. & Mitchell, N. J. Geographic variation in adult and embryonic desiccation tolerance in a terrestrial-breeding frog. Evolution 74, 1186–1199 (2020).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
38.Eads, A., Mitchell, N. J. & Evans, J. Patterns of genetic variation in desiccation tolerance in embryos of the terrestrial-breeding frog, Pseudophryne guentheri. Evolution 66, 2865–2877 (2012).PubMed 
Article 
PubMed Central 

Google Scholar 
39.Cummins, D., Kennington, W. J., Rudin‐Bitterli, T. & Mitchell, N. J. A genome‐wide search for local adaptation in a terrestrial‐breeding frog reveals vulnerability to climate change. Glob. Change Biol. 25, 3151–3162 (2019).Article 

Google Scholar 
40.Bureau of Meteorology. Climate Data Online, http://www.bom.gov.au/climate/data/ (2020).41.Turelli, M. & Moyle, L. C. Asymmetric postmating isolation: Darwin’s corollary to Haldane’s rule. Genetics 176, 1059–1088 (2007).PubMed 
PubMed Central 
Article 

Google Scholar 
42.Dobzhansky, T. Studies on hybrid sterility. II. Localization of sterility factors in Drosophila pseudoobscura hybrids. Genetics 21, 113 (1936).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
43.Muller, H. J. Isolating mechanisms, evolution and temperature. Biol. Symp. 6, 71–125 (1942).
Google Scholar 
44.Orr, H. A. The population genetics of speciation: the evolution of hybrid incompatibilities. Genetics 139, 1805–1813 (1995).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
45.Arntzen, J. W., Jehle, R., Bardakci, F., Burke, T. & Wallis, G. P. Asymmetric viability of reciprocal-cross hybrids between crested and marbled newts (Trituris cristatus and Trituris marmoratus). Evolution 63, 1191–1202 (2009).PubMed 
Article 
PubMed Central 

Google Scholar 
46.Lee-Yaw, J. A., Jacobs, C. G. C. & Irwin, D. E. Individual performance in relation to cytonuclear discordance in a northern contact zone between long-toed salamander (Ambystoma macrodactylum) lineages. Mol. Ecol. 23, 4590–4602 (2014).PubMed 
Article 
PubMed Central 

Google Scholar 
47.Sanchez, S. et al. Within-colony spatial segregation leads to foraging behaviour variation in a seabird. Mar. Ecol. Prog. Ser. 606, 215–230 (2018).Article 

Google Scholar 
48.Sasa, M. M., Chippindale, P. T. & Johnson, N. A. Patterns of postzygotic isolation in frogs. Evolution 52, 1811–1820 (1998).PubMed 
Article 
PubMed Central 

Google Scholar 
49.Sánchez‐Guillén, R., Córdoba‐Aguilar, A., Cordero‐Rivera, A. & Wellenreuther, M. Genetic divergence predicts reproductive isolation in damselflies. J. Evol. Biol. 27, 76–87 (2014).PubMed 
Article 
CAS 
PubMed Central 

Google Scholar 
50.Coyne, J. A. & Orr, H. A. Patterns of speciation in Drosophila. Evolution 43, 362–381 (1989).PubMed 
Article 
PubMed Central 

Google Scholar 
51.Kelemen, L. & Moritz, C. Comparative phylogeography of a sibling pair of rainforest Drosophila species (Drosophila serrata and D. birchii). Evolution 53, 1306–1311 (1999).PubMed 
PubMed Central 

Google Scholar 
52.Hercus, M. J. & Hoffmann, A. A. Desiccation resistance in interspecific Drosophila crosses: genetic interactions and trait correlations. Genetics 151, 1493–1502 (1999).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
53.Rudin-Bitterli, T. S., Mitchell, N. J. & Evans, J. P. Extensive geographical variation in testes size and ejaculate traits in a terrestrial-breeding frog. Biol. Lett. 16, 20200411 (2020).PubMed 
PubMed Central 
Article 

Google Scholar 
54.Shaver, J., Barch, S. & Shivers, C. Tissue-specificity of frog egg-jelly antigens. J. Exp. Zool. 151, 95–103 (1962).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
55.Bradford, D. F. & Seymour, R. S. Influence of environmental PO2 on embryonic oxygen consumption, rate of development, and hatching in the frog, Pseudophryne bibroni. Physiol. Zool. 61, 475–482 (1988).Article 

Google Scholar 
56.Seymour, R. S., Geiser, F. & Bradford, D. F. Metabolic cost of development in terrestrial frog eggs (Pseudophryne bibronii). Physiol. Zool. 64, 688–696 (1991).Article 

Google Scholar 
57.Warkentin, K. M. Adaptive plasticity in hatching age: a response to predation risk trade-offs. Proc. Natl Acad. Sci. USA 92, 3507–3510 (1995).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
58.Webb, P. Effect of body form and response threshold on the vulnerability of four species of teleost prey attacked by largemouth bass (Micropterus salmoides). Can. J. Fish. Aquat. Sci. 43, 763–771 (1986).Article 

Google Scholar 
59.Watkins, T. B. Predator-mediated selection on burst swimming performance in tadpoles of the Pacific tree frog, Pseudacris regilla. Physiol. Zool. 69, 154–167 (1996).Article 

Google Scholar 
60.Wilson, R. & Franklin, C. Thermal acclimation of locomotor performance in tadpoles of the frog Limnodynastes peronii. J. Comp. Physiol. B 169, 445–451 (1999).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
61.Teplitsky, C. et al. Escape behaviour and ultimate causes of specific induced defences in an anuran tadpole. J. Evol. Biol. 18, 180–190 (2005).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
62.Walker, J., Ghalambor, C., Griset, O., McKenney, D. & Reznick, D. Do faster starts increase the probability of evading predators? Funct. Ecol. 19, 808–815 (2005).Article 

Google Scholar 
63.Langerhans, R. B. Morphology, performance, fitness: functional insight into a post-Pleistocene radiation of mosquitofish. Biol. Lett. 5, 488–491 (2009).PubMed 
PubMed Central 
Article 

Google Scholar 
64.Plowman, M. C., Grbac-lvankovic, S., Martin, J., Hopfer, S. M. & Sunderman, F. W. Jr Malformations persist after metamorphosis of Xenopus laevis tadpoles exposed to Ni2+, Co2+, or Cd2+ in FETAX assays. Teratog. Carcinog. Mutagen. 14, 135–144 (1994).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
65.Lynch, M. & Walsh, B. Genetics and Analysis of Quantitative Traits. Vol. 1 (Sinauer Sunderland, MA, 1998).66.Remington, D. L. & O’Malley, D. M. Whole-genome characterization of embryonic stage inbreeding depression in a selfed loblolly pine family. Genetics 155, 337–348 (2000).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
67.Lynch, M. The genetic interpretation of inbreeding depression and outbreeding depression. Evolution 45, 622–629 (1991).PubMed 
Article 
PubMed Central 

Google Scholar 
68.Armbruster, P., Bradshaw, W. E., Steiner, A. L. & Holzapfel, C. M. Evolutionary responses to environmental stress by the pitcher-plant mosquito, Wyeomyia smithii. Heredity 83, 509–519 (1999).PubMed 
Article 
PubMed Central 

Google Scholar 
69.Marr, A. B., Keller, L. F. & Arcese, P. Heterosis and outbreeding depression in descendants of natural immigrants to an inbred population of song sparrows (Melospiza melodia). Evolution 56, 131–142 (2002).PubMed 
Article 
PubMed Central 

Google Scholar 
70.Marshall, T. & Spalton, J. Simultaneous inbreeding and outbreeding depression in reintroduced Arabian oryx. Anim. Conserv. 3, 241–248 (2000).Article 

Google Scholar 
71.Rudin-Bitterli, T. S., Mitchell, N. J. & Evans, J. P. Environmental stress increases the magnitude of nonadditive genetic variation in offspring fitness in the frog Crinia georgiana. Am. Nat. 192, 461–478 (2018).PubMed 
Article 
PubMed Central 

Google Scholar 
72.Drummond, E., Short, E. & Clancy, D. Mitonuclear gene X environment effects on lifespan and health: How common, how big? Mitochondrion 49, 12–18 (2019).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
73.Morales, H. E. et al. Concordant divergence of mitogenomes and a mitonuclear gene cluster in bird lineages inhabiting different climates. Nat. Ecol. Evol. 2, 1258–1267 (2018).PubMed 
Article 
PubMed Central 

Google Scholar 
74.Schmid, M., Evans, B. J. & Bogart, J. P. Polyploidy in amphibia. Cytogenet. Genome Res. 145, 315–330 (2015).PubMed 
Article 
PubMed Central 

Google Scholar 
75.Silla, A. J. Artificial fertilisation in a terrestrial toadlet (Pseudophryne guentheri): effect of medium osmolality, sperm concentration and gamete storage. Reprod. Fertil. Dev. 25, 1134–1141 (2013).PubMed 
Article 
PubMed Central 

Google Scholar 
76.Phillip, G. B. & Keogh, J. S. Extreme sequential polyandry insures against nest failure in a frog. Proc. Roy. Soc. B-Biol. Sci. 276, 115–120 (2009).Article 

Google Scholar 
77.Brandies, P., Peel, E., Hogg, C. J. & Belov, K. The value of reference genomes in the conservation of threatened species. Genes 10, 846 (2019).CAS 
PubMed Central 
Article 

Google Scholar 
78.Scheele, B. C. et al. Interventions for reducing extinction risk in chytridiomycosis‐threatened amphibians. Conserv. Biol. 28, 1195–1205 (2014).PubMed 
Article 
PubMed Central 

Google Scholar 
79.Osborne, W. S. & Norman, J. A. Conservation genetics of Corroboree frogs, Psuedophryne corroboree (Anura: Myobatrachidae): population subdivision and genetic divergence. Aust. J. Zool. 39, 285–297 (1991).Article 

Google Scholar 
80.Browne, R. K. et al. Sperm collection and storage for the sustainable management of amphibian biodiversity. Theriogenology 133, 187–200 (2019).PubMed 
Article 
PubMed Central 

Google Scholar 
81.Silla, A. J. & Byrne, P. G. Hormone-induced ovulation and artificial fertilisation in four terrestrial-breeding anurans. Reprod. Fertil. Dev. https://doi.org/10.1071/RD20243 (2021).82.O’Brien, D. M., Keogh, J. S., Silla, A. J. & Byrne, P. G. Female choice for related males in wild red-backed toadlets (Pseudophryne coriacea). Behav. Ecol. 30, 928–937 (2019).Article 

Google Scholar 
83.Gosner, K. L. A simplified table for staging anuran embryos and larvae with notes on identification. Herpetologica 16, 183–190 (1960).
Google Scholar 
84.Anstis, M. Tadpoles and Frogs of Australia. (New Holland Publishers, 2013).85.CSIRO, and Bureau of Meteorology. State of the Climate 2018 (CSIRO Publishing, 2018).86.Andrich, M. A. & Imberger, J. The effect of land clearing on rainfall and fresh water resources in Western Australia: a multi-functional sustainability analysis. Int. J. Sustain. Dev. World Ecol. 20, 549–563 (2013).Article 

Google Scholar 
87.Raut, B. A., Jakob, C. & Reeder, M. J. Rainfall changes over southwestern Australia and their relationship to the Southern Annular Mode and ENSO. J. Clim. 27, 5801–5814 (2014).Article 

Google Scholar 
88.Arnold, G. in Greenhouse: Planning for Climate Change (ed. Pearman, G. I.) 375–386 (CSIRO Publishing, 1988).89.Hobbs, R. J. Effects of landscape fragmentation on ecosystem processes in the Western Australian wheatbelt. Biol. Conserv. 64, 193–201 (1993).Article 

Google Scholar 
90.Silla, A. J. Effect of priming injections of luteinizing hormone-releasing hormone on spermiation and ovulation in Gϋnther’s toadlet, Pseudophryne guentheri. Reprod. Biol. Endocrinol. 9, 68 (2011).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
91.Lymbery, R. A., Kennington, W. J. & Evans, J. P. Multivariate sexual selection on ejaculate traits under sperm competition. Am. Nat. 192, 94–104 (2018).PubMed 
Article 
PubMed Central 

Google Scholar 
92.Browne, R. K., Clulow, J. & Mahony, M. Short-term storage of cane toad (Bufo marinus) gametes. Reproduction 121, 167–173 (2001).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
93.Kouba, A. J., Vance, C. K., Frommeyer, M. A. & Roth, T. L. Structural and functional aspects of Bufo americanus spermatozoa: effects of inactivation and reactivation. J. Exp. Zool. A. Comp. Exp. Biol. 295, 172–182 (2003).PubMed 
Article 
PubMed Central 

Google Scholar 
94.Abràmoff, M. D., Magalhães, P. J. & Ram, S. J. Image processing with Image. J. Biophotonics Int. 11, 36–42 (2004).
Google Scholar 
95.Noldus, L. P., Spink, A. J. & Tegelenbosch, R. A. EthoVision: a versatile video tracking system for automation of behavioral experiments. Behav. Res. Methods Instrum. Comput. 33, 398–414 (2001).CAS 
PubMed 
Article 
PubMed Central 

Google Scholar 
96.Bates, D., Mächler, M., Bolker, B. & Walker, S. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67, 1–48. https://doi.org/10.18637/jss.v067.i01 (2014).97.Bolker, B. M. et al. Generalized linear mixed models: a practical guide for ecology and evolution. Trends Ecol. Evol. 24, 127–135 (2009).PubMed 
Article 
PubMed Central 

Google Scholar 
98.Harrison, X. A. Using observation-level random effects to model overdispersion in count data in ecology and evolution. PeerJ 2, e616 (2014).PubMed 
PubMed Central 
Article 

Google Scholar 
99.Rudin-Bitterli, T. S., Evans, J. P. & Mitchell, N. J. Fitness consequences of targeted gene flow to counter impacts of drying climates on terrestrial-breeding frogs. Data sets. https://doi.org/10.5061/dryad.6m905qg09 (2021). More

1.Steenweg, R. et al. Scaling-up camera traps: monitoring the planet’s biodiversity with networks of remote sensors. Front. Ecol. Environ. 15, 26–34 (2017).Article 

Google Scholar 
2.Rich, L. N. et al. Assessing global patterns in mammalian carnivore occupancy and richness by integrating local camera trap surveys. Global Ecol. Biogeogr. 26, 918–929 (2017).Article 

Google Scholar 
3.Barnosky, A. D. et al. Has the Earth’s sixth mass extinction already arrived? Nature 471, 51–57 (2011).Article 

Google Scholar 
4.Ahumada, J. A. et al. Wildlife insights: a platform to maximize the potential of camera trap and other passive sensor wildlife data for the planet. Environ. Conserv. 47, 1–6 (2020).MathSciNet 
Article 

Google Scholar 
5.Norouzzadeh, M. S. et al. Automatically identifying, counting, and describing wild animals in camera-trap images with deep learning. Proc. Natl Acad. Sci. 115, E5716–E5725 (2018).Article 

Google Scholar 
6.Miao, Z. et al. Insights and approaches using deep learning to classify wildlife. Sci. Rep. 9, 8137 (2019).Article 

Google Scholar 
7.Liu, Z. et al. Large-scale long-tailed recognition in an open world. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 2537–2546 (IEEE, 2019).8.Liu, Z. et al. Open compound domain adaptation. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition 12406–12415 (IEEE, 2020).9.Hautier, Y. et al. Anthropogenic environmental changes affect ecosystem stability via biodiversity. Science 348, 336–340 (2015).Article 

Google Scholar 
10.Barlow, J. et al. Anthropogenic disturbance in tropical forests can double biodiversity loss from deforestation. Nature 535, 144–147 (2016).Article 

Google Scholar 
11.Ripple, W. J. et al. Conserving the world’s megafauna and biodiversity: the fierce urgency of now. Bioscience 67, 197–200 (2017).Article 

Google Scholar 
12.Dirzo, R. et al. Defaunation in the Anthropocene. Science 345, 401–406 (2014).Article 

Google Scholar 
13.O’Connell, A. F., Nichols, J. D. & Karanth, K. U. Camera Traps in Animal Ecology: Methods and Analyses (Springer Science & Business Media, 2010).14.Burton, A. C. et al. Wildlife camera trapping: a review and recommendations for linking surveys to ecological processes. J. Appl. Ecol. 52, 675–685 (2015).Article 

Google Scholar 
15.Kays, R., McShea, W. J. & Wikelski, M. Born-digital biodiversity data: millions and billions. Divers. Distrib. 26, 644–648 (2020).Article 

Google Scholar 
16.Swanson, A. et al. Snapshot Serengeti, high-frequency annotated camera trap images of 40 mammalian species in an African savanna. Sci. Data 2, 1–14 (2015).Article 

Google Scholar 
17.Ahumada, J. A. et al. Community structure and diversity of tropical forest mammals: data from a global camera trap network. Philos. Trans. R. Soc. B Biol. Sci. 366, 2703–2711 (2011).Article 

Google Scholar 
18.Pardo, L. E. et al. Snapshot Safari: a large-scale collaborative to monitor Africa’s remarkable biodiversity. South Africa J. Sci. https://doi.org/10.17159/sajs.2021/8134 (2021).19.Anderson, T. M. et al. The spatial distribution of African savannah herbivores: species associations and habitat occupancy in a landscape context. Philos. Trans. R. Soc. B Biol. Sci. 371, 20150314 (2016).Article 

Google Scholar 
20.Palmer, M., Fieberg, J., Swanson, A., Kosmala, M. & Packer, C. A ‘dynamic’ landscape of fear: prey responses to spatiotemporal variations in predation risk across the lunar cycle. Ecol. Lett. 20, 1364–1373 (2017).Article 

Google Scholar 
21.Tabak, M. A. et al. Machine learning to classify animal species in camera trap images: applications in ecology. Methods Ecol. Evol. 10, 585–590 (2019).Article 

Google Scholar 
22.Whytock, R. C. et al. Robust ecological analysis of camera trap data labelled by a machine learning model. Methods Ecol. Evol 12, 1080–1092 (2021).Article 

Google Scholar 
23.Beery, S., Van Horn, G. & Perona, P. Recognition in terra incognita. In Proc. European Conference on Computer Vision (ECCV) 456–473 (IEEE, 2018).24.Tabak, M. A. et al. Improving the accessibility and transferability of machine learning algorithms for identification of animals in camera trap images: MLWIC2. Ecol. Evol. 10, 10374–10383 (2020).Article 

Google Scholar 
25.Shahinfar, S., Meek, P. & Falzon, G. How many images do I need? Understanding how sample size per class affects deep learning model performance metrics for balanced designs in autonomous wildlife monitoring. Ecol. Inform. 57, 101085 (2020).Article 

Google Scholar 
26.Norouzzadeh, M. S. et al. A deep active learning system for species identification and counting in camera trap images. Methods Ecol. Evol. 12, 150–161 (2020).Article 

Google Scholar 
27.Willi, M. et al. Identifying animal species in camera trap images using deep learning and citizen science. Methods Ecol. Evol. 10, 80–91 (2019).Article 

Google Scholar 
28.Schneider, S., Greenberg, S., Taylor, G. W. & Kremer, S. C. Three critical factors affecting automated image species recognition performance for camera traps. Ecol. Evol. 10, 3503–3517 (2020).Article 

Google Scholar 
29.Kays, R. et al. An empirical evaluation of camera trap study design: how many, how long and when? Methods Ecol. Evol. 11, 700–713 (2020).Article 

Google Scholar 
30.Prach, K. & Walker, L. R. Four opportunities for studies of ecological succession. Trends Ecol. Evol. 26, 119–123 (2011).Article 

Google Scholar 
31.Mech, L. D., Isbell, F., Krueger, J. & Hart, J. Gray wolf (Canis lupus) recolonization failure: a Minnesota case study. Can. Field-Nat. 133, 60–65 (2019).Article 

Google Scholar 
32.Taylor, G. et al. Is reintroduction biology an effective applied science? Trends Ecol. Evol. 32, 873–880 (2017).Article 

Google Scholar 
33.Clavero, M. & Garcia-Berthou, E. Invasive species are a leading cause of animal extinctions. Trends Ecol. Evol. 20, 110 (2005).Article 

Google Scholar 
34.Caravaggi, A. et al. An invasive-native mammalian species replacement process captured by camera trap survey random encounter models. Remote Sens. Ecol. Conserv. 2, 45–58 (2016).Article 

Google Scholar 
35.Arjovsky, M., Bottou, L., Gulrajani, I. & Lopez-Paz, D. Invariant risk minimization. Preprint at https://arxiv.org/abs/1907.02893 (2019).36.Yosinski, J., Clune, J., Bengio, Y. & Lipson, H. How transferable are features in deep neural networks? In Advances in Neural Information Processing Systems 3320–3328 (IEEE, 2014).37.Deng, J. et al. ImageNet: a large-scale hierarchical image database. In Proc. 2009 IEEE Conference on Computer Vision and Pattern Recognition 248–255 (IEEE, 2009).38.Pimm, S. L. et al. The biodiversity of species and their rates of extinction, distribution and protection. Science https://doi.org/10.1126/science.1246752 (2014).39.Liu, W., Wang, X., Owens, J. & Li, Y. Energy-based out-of-distribution detection. In Advances in Neural Information Processing Systems (eds Larochelle, H. et al.) 21464–21475 (Curran Associates, 2020).40.Lee, D.-H. Pseudo-label: the simple and efficient semi-supervised learning method for deep neural networks. In Workshop on Challenges in Representation Learning, ICML, Vol. 3 (2013).41.He, K., Zhang, X., Ren, S. & Sun, J. Deep residual learning for image recognition. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 770–778 (IEEE, 2016).42.Hinton, G., Vinyals, O. & Dean, J. Distilling the knowledge in a neural network. Preprint at https://arxiv.org/abs/1503.02531 (2015).43.Gaynor, K. M., Daskin, J. H., Rich, L. N. & Brashares, J. S. Postwar wildlife recovery in an African savanna: evaluating patterns and drivers of species occupancy and richness. Anim. Conserv. 24, 510–522 (2020).Article 

Google Scholar 
44.Paszke, A. et al. in Advances in Neural Information Processing Systems Vol. 32 (eds Wallach, H. et al.) 8024–8035 http://papers.neurips.cc/paper/9015-pytorch-an-imperative-style-high-performance-deep-learning-library.pdf (Curran Associates, 2019)45.Chen, T., Kornblith, S., Norouzi, M. & Hinton, G. A simple framework for contrastive learning of visual representations. Preprint at https://arxiv.org/abs/2002.05709 (2020).46.He, K., Fan, H., Wu, Y., Xie, S. & Girshick, R. Momentum contrast for unsupervised visual representation learning. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition 9729–9738 (IEEE, 2020).47.Xiao, T., Wang, X., Efros, A. A. & Darrell, T. What should not be contrastive in contrastive learning. Preprint at https://arxiv.org/abs/2008.05659 (2020). More

The net uptake of CO2 by the biosphere offsets roughly a quarter of current fossil fuel emissions. However, climate change is expected to impact photosynthesis and ecosystem respiration differently. Quantification of these individual processes is required to better understand and predict the consequences for carbon cycling. Variations in oxygen isotope signatures (δ18O and Δ17O) in atmospheric CO2 can be used as tracers for photosynthesis. Δ17O is much less dependent on variations in the hydrological cycle, which often obscure photosynthesis signals in the more widely measured δ18O. Although, measurement techniques for Δ17O in tropospheric CO2 only became sufficiently accurate to interpret variations since the ~2010s, providing new insights into the carbon cycle. More

Download PDF

It was turtle-nesting season when this photograph was taken one night in June. I am on Needham’s Point beach measuring a critically endangered female hawksbill turtle (Eretmochelys imbricata). As field director of the Barbados sea turtle project, I run the day-to-day conservation activities and train and manage volunteers.We also run research projects that inform our conservation activities. We collect data such as shell length, which can tell us the age at which females become sexually mature and can indicate growth rates. These data help us to keep track of turtle health and survival. For example, if we start seeing smaller turtles, this could indicate that they are maturing faster, or that food is scarce and the turtles are growing more slowly.In August, the baby turtles hatch. I was on call 7 days a week for around 8 hours a day, responding to emergencies. These included hatchlings wandering off in the wrong direction, putting them at risk of being hit by a car or eaten by predators such as rats and cats. We took the hatchlings to a safe spot on the beach and released them. I also had to prepare for the expected swells as Hurricane Ida passed us by: when beaches flood, nests can wash away. We took rescued eggs and premature hatchlings to a makeshift intensive-care unit until they were ready for release. We aim to leave no turtle behind.I have worked at the project for 15 years. I recently finished a master’s degree on the coloration of the Barbados bullfinch (Loxigilla barbadensis) at the University of the West Indies, which hosts the turtle project. Next year I hope to start a PhD, part of which will look at the conflict between tourism and sea-turtle survival in Barbados. Here, interactions between sea turtles and humans occur at every stage of the turtles’ lives and can affect their survival. After my doctorate, I will continue to focus on helping sea turtles in the Caribbean. There is something addictive about making a real-time, tangible difference to their lives.

Nature 598, 532 (2021)
doi: https://doi.org/10.1038/d41586-021-02851-6

Related Articles

230-million-year-old turtle fossil deepens mystery of reptile’s origins

Breeding the sweetest biofuels in the business

Dexterous sea turtles use flippers as grippers

Subjects

Careers

Conservation biology

Latest on:

Careers

How I wrote a pop-science book
Career Column 15 OCT 21

How AI is helping the natural sciences
Career Guide 13 OCT 21

Eight career tips from Nobel Laureates
Career Column 08 OCT 21

Jobs

W 3-Professorship for Biochemistry

University of Kiel (CAU)
Kiel, Germany

Laboratory Associate

Center for Molecular Fingerprinting (CMF) Research Nonprofit LLC
Multiple locations

Scientific Computing Officer – Structural Studies – Scientific Computing – LMB 1697

MRC Laboratory of Molecular Biology
Cambridge, United Kingdom

Head of Laser Lab Operations

Center for Molecular Fingerprinting (CMF) Research Nonprofit LLC
Multiple locations More

1.Hughes, W. O. H. & Goulson, D. The use of alarm pheromones to enhance bait harvest by grass-cutting ants. Bull. Entomol. Res. 92, 213–218 (2002).CAS 
Article 

Google Scholar 
2.Staab, M. & Kleineidam, C. J. Initiation of swarming behavior and synchronization of mating flights in the leaf-cutting ant Atta vollenweideri Forel, 1893 (Hymenoptera: Formicidae). Myrmecol. News 19, 93–102 (2014).
Google Scholar 
3.Sales, T. A., Toledo, A. M. O. & Lopes, J. F. S. The best of heavy queens: Influence of post-flight weight on queens’ survival and productivity in Acromyrmex subterraneus (Forel, 1893) (Hymenoptera: Formicidae). Insectes Soc. 67, 383–390 (2020).Article 

Google Scholar 
4.Camargo, R. S., Forti, L. C., Fujihara, R. T. & Roces, F. Digging effort in leaf-cutting ant queens (Atta sexdens rubropilosa) and its effects on survival and colony growth during the claustral phase. Insectes Soc. 58, 17–22 (2011).Article 

Google Scholar 
5.Autuori, M. Contribuição para o conhecimento da saúva (Atta spp.) (Hymenoptera: Formicidae). I. Evolução do sauveiro (Atta sexdens rubropilosa Forel, 1908). Arq. Inst. Biol. 12, 197–228 (1941).
Google Scholar 
6.Aylward, F. O. et al. Leucoagaricus gongylophorus produces diverse enzymes for the degradation of recalcitrant plant polymers in leaf-cutter ant fungus gardens. Appl. Environ. Microbiol. 79, 3770–3778 (2013).ADS 
CAS 
Article 

Google Scholar 
7.Costa, A. N., Vasconcelos, H. L., Vieira-Neto, E. H. M. & Bruna, E. M. Do herbivores exert top-down effects in Neotropical savannas? Estimates of biomass consumption by leaf-cutter ants. J. Veg. Sci. 19, 849–854 (2008).Article 

Google Scholar 
8.Bollazzi, M., Forti, L. C. & Roces, F. Ventilation of the giant nests of Atta leaf-cutting ants: Does underground circulating air enter the fungus chambers?. Insectes Soc. 59, 487–498 (2012).Article 

Google Scholar 
9.Sousa-Souto, L. et al. Increased CO2 emission and organic matter decomposition by leaf-cutting ant nests in a coastal environment. Soil Biol. Biochem. 44, 21–25 (2012).CAS 
Article 

Google Scholar 
10.Hasin, S. et al. CO2 efflux from subterranean nests of ant communities in a seasonal tropical forest, Thailand. Ecol. Evol. 4, 3929–3939 (2014).Article 

Google Scholar 
11.Tschinkel, W. R. The nest architecture of the Florida harvester ant, Pogonomyrmex badius. J. Insect Sci. 4, 21 (2004).Article 

Google Scholar 
12.Kleineidam, C. & Roces, F. Carbon dioxide concentrations and nest ventilation in nests of the leaf-cutting ant Atta vollenweideri. Insectes Soc. 47, 241–248 (2000).Article 

Google Scholar 
13.Currie, J. A. Gas diffusion through soil crumbs: The effects of compaction and wetting. J. Soil Sci. 35, 1–10 (1984).CAS 
Article 

Google Scholar 
14.Kleineidam, C., Ernst, R. & Roces, F. Wind-induced ventilation of the giant nests of the leaf-cutting ant Atta vollenweideri. Naturwissenschaften 88, 301–305 (2001).ADS 
CAS 
Article 

Google Scholar 
15.Vogel, S., Ellington, C. P. & Kilgore, D. L. Wind-induced ventilation of the burrow of the prairie-dog, Cynomys ludovicianus. J. Comp. Physiol. 85, 1–14 (1973).Article 

Google Scholar 
16.Jonkman, J. C. M. The external and internal structure and growth of nests of the leaf-cutting ant Atta vollenweideri Forel, 1893 (Hym: Formicidae) Part II. Zeitschrift für Angew. Entomol. 89, 158–173 (1980).Article 

Google Scholar 
17.Gutiérrez, J. L. & Jones, C. G. Physical ecosystem engineers as agents of biogeochemical heterogeneity. Bioscience 56, 227–236 (2006).Article 

Google Scholar 
18.Fernandez-Bou, A. S. et al. The role of the ecosystem engineer, the leaf-cutter ant Atta cephalotes, on soil CO2 dynamics in a wet tropical rainforest. J. Geophys. Res. Biogeosciences 124, 260–273 (2019).ADS 
CAS 
Article 

Google Scholar 
19.Moitinho, M. R. et al. Does fresh farmyard manure introduce surviving microbes into soil or activate soil-borne microbiota?. J. Environ. Manag. 11, 1–15 (2021).
Google Scholar 
20.Roces, F. Variable thermal sensitivity as output of a circadian clock controlling the bimodal rhythm of temperature choice in the ant Camponotus mus. J. Comp. Physiol. A 177, 637–643 (1995).Article 

Google Scholar 
21.Römer, D., Bollazzi, M. & Roces, F. Carbon dioxide sensing in an obligate insect-fungus symbiosis: CO2 preferences of leaf-cutting ants to rear their mutualistic fungus. PLoS ONE 12, e0174597 (2017).Article 

Google Scholar 
22.Halboth, F. & Roces, F. The construction of ventilation turrets in Atta vollenweideri leaf-cutting ants: Carbon dioxide levels in the nest tunnels, but not airflow or air humidity, influence turret structure. PLoS ONE 12, e0188162 (2017).Article 

Google Scholar 
23.Kleineidam, C. & Tautz, J. Perception of carbon dioxide and other “air-condition” parameters in the leaf cutting ant Atta cephalotes. Naturwissenschaften 83, 566–568 (1996).ADS 
CAS 

Google Scholar 
24.Kleineidam, C., Romani, R., Tautz, J. & Isidoro, N. Ultrastructure and physiology of the CO2 sensitive sensillum ampullaceum in the leaf-cutting ant Atta sexdens. Arthropod Struct. Dev. 29, 43–55 (2000).CAS 
Article 

Google Scholar 
25.Camargo, R. S. & Forti, L. C. Queen lipid content and nest growth in the leaf cutting ant (Atta sexdens rubropilosa) (Hymenoptera: Formicidae). J. Nat. Hist. 47, 65–73 (2013).Article 

Google Scholar 
26.Seal, J. N. Scaling of body weight and fat content in fungus-gardening ant queens: Does this explain why leaf-cutting ants found claustrally?. Insectes Soc. 56, 135–141 (2009).Article 

Google Scholar 
27.Camargo, R. D. S., Fonseca, J. A., Lopes, J. F. S. & Forti, L. C. Influência do ambiente no desenvolvimento de colônias iniciais de formigas cortadeiras (Atta sexdens rubropilosa). Ciência Rural 43, 1375–1380 (2013).Article 

Google Scholar 
28.Silva, E. J., da Silva Camargo, R. & Forti, L. C. Flight and digging effort in leaf-cutting ant males and gynes. Sociobiology 62, 334–339 (2015).Article 

Google Scholar 
29.Kuzyakov, Y. Sources of CO2 efflux from soil and review of partitioning methods. Soil Biol. Biochem. 38, 425–448 (2006).CAS 
Article 

Google Scholar 
30.Camargo, R. S., Silva, E. J., Forti, L. C. & Matos, C. A. O. Initial development and production of CO2 in colonies of the leaf-cutting ant Atta sexdens during the claustral foundation. Sociobiology 63, 720–723 (2016).Article 

Google Scholar 
31.Cribari-Neto, F. & Zeileis, A. Beta regression in R. J. Stat. Softw. 34, 1–24 (2010).Article 

Google Scholar 
32.Ferrari, S. & Cribari-Neto, F. Beta regression for modelling rates and proportions. J. Appl. Stat. 31, 799–815 (2004).MathSciNet 
Article 

Google Scholar 
33.Smithson, M. & Verkuilen, J. A better lemon squeezer? Maximum-likelihood regression with beta-distributed dependent variables. Psychol. Methods 11, 54 (2006).Article 

Google Scholar  More