Ecology

1.
MacKenzie, D. I. et al. Occupancy Estimation and Modeling: Inferring Patterns and Dynamycs of Species Occurence (Academic Press, Cambridge, 2006).
Google Scholar 
2.
Kéry, M. & Royle, J. A. Applied Hierarchical Modeling in Ecology Vol. 1 (Academic Press, Cambridge, 2016).
Google Scholar 

3.
Lindenmayer, D. B. et al. Improving biodiversity monitoring. Austral Ecol. 37, 285–294 (2012).
Article  Google Scholar 

4.
Einoder, L. D. et al. Occupancy and detectability modelling of vertebrates in northern Australia using multiple sampling methods. PLoS ONE 13, e0206373. https://doi.org/10.1371/journal.pone.0206373 (2018).
Article  PubMed  PubMed Central  Google Scholar 

5.
Boulinier, T., Nichols, J. D., Sauer, J. R., Hines, J. E. & Pollock, K. H. Estimating species richness: The importance of heterogeneity in species detectability. Ecology 79, 1018–1028 (1998).
Article  Google Scholar 

6.
Tyre, A. J., Tenhumberg, B., Field, S. A., Niejalke, D. & Parris, K. Improving precision and reducing bias in biological surveys: Estimating false-negative error rates. Ecol. Appl. 13, 1790–1801 (2003).
Article  Google Scholar 

7.
Kellner, K. F. & Swihart, R. K. Accounting for imperfect detection in ecology: A quantitative review. PLoS ONE 9, e111436. https://doi.org/10.1371/journal.pone.0111436 (2014).
CAS  Article  PubMed  PubMed Central  ADS  Google Scholar 

8.
Iknayan, K. J., Tingley, M. W., Furnas, B. J. & Beissinger, S. R. Detecting diversity: Emerging methods to estimate species diversity. Trends Ecol. Evol. 29, 97–106 (2014).
PubMed  Article  PubMed Central  Google Scholar 

9.
Kéry, M. & Schmidt, B. Imperfect detection and its consequences for monitoring for conservation. Community Ecol. 9, 207–216 (2008).
Article  Google Scholar 

10.
Tingley, M. W. & Beissinger, S. R. Cryptic loss of montane avian richness and high community turnover over 100 years. Ecology 94, 598–609 (2013).
PubMed  Article  PubMed Central  Google Scholar 

11.
Leu, M. et al. Effects of point-count duration on estimated detection probabilities and occupancy of breeding birds. J. F. Ornithol. 88, 80–93 (2017).
Article  Google Scholar 

12.
Royle, J. A. & Dorazio, R. M. Hierarchical Modeling and Inference in Ecology. The Analysis of Data from Populations, Metapopulations and Communities (Elsevier, Amsterdam, 2008).
Google Scholar 

13.
Guillera-Arroita, G. Modelling of species distributions, range dynamics and communities under imperfect detection: Advances, challenges and opportunities. Ecography 40, 281–295 (2017).
Article  Google Scholar 

14.
Kéry, M., Royle, J. A., Plattner, M. & Dorazio, R. M. Species richness and occupancy estimation in communities subject to temporary emigration. Ecology 90, 1279–1290 (2009).
PubMed  Article  PubMed Central  Google Scholar 

15.
Sólymos, P., Matsuoka, S. M., Stralberg, D., Barker, N. K. S. & Bayne, E. M. Phylogeny and species traits predict bird detectability. Ecography 41, 1595–1603 (2018).
Article  Google Scholar 

16.
Jarzyna, M. A. & Jetz, W. Detecting the multiple facets of biodiversity. Trends Ecol. Evol. 31, 527–538 (2016).
PubMed  Article  PubMed Central  Google Scholar 

17.
Kéry, M., Royle, J. A. & Schmid, H. Modeling avian abundance from replicated counts. Ecol. Appl. 15, 1450–1461 (2005).
Article  Google Scholar 

18.
Mackenzie, D. I. & Royle, J. A. Designing occupancy studies: General advice and allocating survey effort. J. Appl. Ecol. 42, 1105–1114 (2005).
Article  Google Scholar 

19.
Jiménez-Franco, M. V. et al. Use of classical bird census transects as spatial replicates for hierarchical modeling of an avian community. Ecol. Evol. 9, 825–835 (2018).
Article  Google Scholar 

20.
Clement, M. J., Hines, J. E., Nichols, J. D., Pardieck, K. L. & Ziolkowski, D. J. Estimating indices of range shifts in birds using dynamic models when detection is imperfect. Glob. Change Biol. 22, 3273–3285 (2016).
Article  ADS  Google Scholar 

21.
Sliwinski, M., Powell, L., Koper, N., Giovanni, M. & Schacht, W. Research design considerations to ensure detection of all species in an avian community. Methods Ecol. Evol. 7, 456–462 (2016).
Article  Google Scholar 

22.
Rappole, J. H., Winker, K. & Powell, G. V. Migratory bird habitat use in Southern Mexico: Mist nets versus point counts. J. F. Ornithol. 69, 635–643 (2012).
Google Scholar 

23.
Faaborg, J., Arendt, W. J. & Dugger, K. M. Bird population studies in Puerto Rico using mist nets: General patterns and comparisons with point counts. Stud. Avian Biol. 29, 144–150 (2004).
Google Scholar 

24.
Dunn, E. H. & Ralph, C. J. Use of mist nets as a tool for bird population monitoring. Stud. Avian Biol. 29, 1–6 (2004).
Google Scholar 

25.
Lynch, J. F. Distribution of overwintering Nearctic migrants in the Yucatan Peninsula, I: General patterns of occurrence. Condor 91, 515–544 (1989).
Article  Google Scholar 

26.
Wunderle, J. M. & Waide, R. B. Distribution of overwintering Nearctic migrants in the Bahamas and Greater Antilles. Condor 95, 904–933 (1993).
Article  Google Scholar 

27.
Gram, W. K. & Faaborg, J. The distribution of neotropical migrant birds wintering in the El Cielo Biosphere Reserve, Tamaulipas, Mexico. Condor 99, 658–670 (1997).
Article  Google Scholar 

28.
Whitman, A. A., Hagan, J. M. & Brokaw, N. V. L. A comparison of two bird survey techniques used in a subtropical forest. Condor 99, 955–965 (1997).
Article  Google Scholar 

29.
Arizaga, J., Deán, J. I., Vilches, A., Alonso, D. & Mendiburu, A. Monitoring communities of small birds: A comparison between mist-netting and counting. Bird Study 58(3), 37–41 (2011).

30.
Darras, K. et al. Autonomous sound recording outperforms human observation for sampling birds: A systematic map and user guide. Ecol. Appl. 29, e01954. https://doi.org/10.1002/eap.1954 (2019).
Article  PubMed  Google Scholar 

31.
Smit, B., Woodborne, S., Wolf, B. O. & McKechnie, A. E. Differences in the use of surface water resources by desert birds are revealed using isotopic tracers. Auk 136, 1–13 (2019).
Article  Google Scholar 

32.
Lynn, J. C., Rosenstock, S. S. & Chambers, C. L. Avian use of desert wildlife water developments as determined by remote videography. West. N. Am. Nat. 68, 107–112 (2008).
Article  Google Scholar 

33.
Fisher, J. T. & Bradbury, S. A multi-method hierarchical modeling approach to quantifying bias in occupancy from noninvasive genetic tagging studies. J. Wildl. Manag. 78, 1087–1095 (2014).
Article  Google Scholar 

34.
Fisher, J. T., Heim, N., Code, S. & Paczkowski, J. Grizzly bear noninvasive genetic tagging surveys: Estimating the magnitude of missed detections. PLoS ONE 11, 1–16 (2016).
Google Scholar 

35.
Nichols, J. D. et al. Multi-scale occupancy estimation and modelling using multiple detection methods. J. Appl. Ecol. 45, 1321–1329 (2008).
Article  Google Scholar 

36.
Calvo, J. F. et al. Catálogo de las aves de la Región de Murcia (España). An. Biol. 39, 7–33 (2017).
Article  Google Scholar 

37.
Galbraith, J. A., Jones, D. N., Beggs, J. R., Stanley, M. C. & Parry, K. Urban bird feeders dominated by a few species and individuals. Front. Ecol. Evol. https://doi.org/10.3389/fevo.2017.00081 (2017).
Article  Google Scholar 

38.
McCarthy, M. A. et al. The influence of abundance on detectability. Oikos 122, 717–726 (2012).
Article  Google Scholar 

39.
Lee, A. T. K., Wright, D. & Barnard, P. Hot bird drinking patterns: Drivers of water visitation in a fynbos bird community. Afr. J. Ecol. 55, 541–553 (2017).
Article  Google Scholar 

40.
Gregory, R. D., Gibbons, D. W. & Donald, P. F. Bird census and survey techniques. In Bird Ecology and Conservation. A Handbook of Techniques (eds. Sutherland, W. J., Newton, I. & Green, R. E.) 17–55 (Oxford Scholarship, Oxford, 2004).

41.
Derlindati, E. J. & Caziani, S. M. Using canopy and understory mist nets and point counts to study bird assemblages in Chaco forests. Wilson Bull. 117, 92–99 (2005).
Article  Google Scholar 

42.
Wang, Y. & Finch, D. M. Consistency of mist netting and point counts in assessing landbird species richness and relative abundance during migration. Condor 104, 59–72 (2002).
Article  Google Scholar 

43.
Valera, F. et al. History and adaptation stories of the vertebrate fauna of southern Spain semiarid habitats. J. Arid Environ. 75, 1342–1351 (2011).
Article  ADS  Google Scholar 

44.
Rappole, J. H. Migratory bird habitat use in Southern Mexico: Mist nets versus point counts. J. F. Ornithol. 69, 635–643 (2012).
Google Scholar 

45.
Poulin, B., Lefebvre, G. & Pilard, P. Quantifying the breeding assemblage of reedbed passerines with mist-net and point-count surveys. J. F. Ornithol. 71, 443–454 (2000).
Article  Google Scholar 

46.
Armas, C., Miranda, J. D., Padilla, F. M. & Pugnaire, F. I. Special issue: The Iberian Southeast. J. Arid Environ. 75, 1241–1243 (2011).
Article  ADS  Google Scholar 

47.
Lisón, F. & Calvo, J. F. Bat activity over small ponds in dry Mediterranean forests: Implications for conservation. Acta Chiropterol. 16, 95–101 (2014).
Article  ADS  Google Scholar 

48.
Sebastián-González, E., Sánchez-Zapata, J. A. & Botella, F. Agricultural ponds as alternative habitat for waterbirds: Spatial and temporal patterns of abundance and management strategies. Eur. J. Wildl. Res. 56, 11–20 (2010).
Article  Google Scholar 

49.
Egea-Serrano, A., Oliva-Paterna, F. J. & Torralva, M. Breeding habitat selection of Salamandra salamandra (Linnaeus, 1758) in the most arid zone of its European distribution range: Application to conservation management. Hydrobiologia 560, 363–371 (2006).
Article  Google Scholar 

50.
Egea-Serrano, A., Oliva-Paterna, F. J., Tejedo, M. & Torralva, M. Breeding habitat selection of an endangered species in an arid zone: The case of Alytes dickhilleni Arntzen & García-París, 1995. Acta Herpetol. 1, 81–94 (2006).
Google Scholar 

51.
Davies, S. R., Sayer, C. D., Greaves, H., Siriwardena, G. M. & Axmacher, J. C. A new role for pond management in farmland bird conservation. Agric. Ecosyst. Environ. 233, 179–191 (2016).
Article  Google Scholar 

52.
Oertli, B. Freshwater biodiversity conservation: The role of artificial ponds in the 21st century. Aquat. Conserv. Mar. Freshw. Ecosyst. 28, 264–269 (2018).
Article  Google Scholar 

53.
MacKenzie, D. I. et al. Estimating site occupancy rates when detection probabilities are less than one. Ecology 83, 2248–2255 (2002).
Article  Google Scholar 

54.
Rich, L. N., Miller, D. A. W., Robinson, H. S., McNutt, J. W. & Kelly, M. J. Using camera trapping and hierarchical occupancy modelling to evaluate the spatial ecology of an African mammal community. J. Appl. Ecol. 53, 1225–1235 (2016).
Article  Google Scholar 

55.
Burnham, K. P. & Anderson, D. R. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach (Springer, New York, 2002).
Google Scholar 

56.
Martínez-Martí, C., Jiménez-Franco, M. V., Royle, J. A., Palazón, J. A. & Calvo, J. F. Integrating occurrence and detectability patterns based on interview data: A case study for threatened mammals in Equatorial Guinea. Sci. Rep. 6, 33838. https://doi.org/10.1038/srep33838 (2016).
CAS  Article  PubMed  PubMed Central  ADS  Google Scholar 

57.
White, G. C. & Burnham, K. P. Program MARK: survival estimation from populations of marked animals. Bird Study 46, S120–S139 (1999).
Article  Google Scholar 

58.
Laake, J. L. RMark: An R Interface for Analysis of Capture-Recapture Data with MARK. AFSC Processed Report 2013–01, 25p. Alaska Fish. Sci. Cent., NOAA, Natl. Mar. Fish. Serv., 7600 Sand Point Way NE, Seattle WA 98115 (2013).

59.
Denis, T. et al. Biological traits, rather than environment, shape detection curves of large vertebrates in neotropical rainforests. Ecol. Appl. 27, 3218–3221 (2017).
Article  Google Scholar 

60.
Frishkoff, L. O., De Valpine, P. & M’Gonigle, L. K. Phylogenetic occupancy models integrate imperfect detection and phylogenetic signal to analyze community structure. Ecology 98, 198–210 (2017).
PubMed  Article  PubMed Central  Google Scholar 

61.
Pearman, P. B. et al. Phylogenetic patterns of climatic, habitat and trophic niches in a European avian assemblage. Glob. Ecol. Biogeogr. 23, 414–424 (2014).
PubMed  Article  PubMed Central  Google Scholar 

62.
Powell, L. A. Approximating variance of demographic parameters using the delta method: A reference for avian biologists. Condor 109, 949–954 (2007).
Article  Google Scholar  More

The results of ANOVA tests were summarised in Table 1. None of the interaction effect was significant. Therefore, the main effects of disc type and working depth were presented in the following sections.
Table 1 Summary of ANOVA test results.
Full size table

Soil cutting forces
The rippled disc required an average draft force of 675 N, which was numerically the highest among the discs (Fig. 1a). The notched disc had a minimal draft force demand of 579 N. Increasing the working depth from the shallow (63.5 mm) to deep (127 mm) resulted in the draft force increasing from 291 to 965 N. This more-than-tripled increase was significant and can be explained by the soil dynamics theory that draft force varies with the contact area between soil and tool22. The rippled edge slightly increased the surface area as compared to the smooth edge, while the notched edge slightly decreased the contact area due to the notches. As for the depth effect, a deeper operation significantly increased the portion of the disc in contact with soil regardless of the disc type.
Figure 1

Soil cutting forces of different discs at different working depths: (a) draft force, (b) vertical force, and (c) lateral force; means followed by different lower case letters or upper case letters are significantly different according to Tukey’s test at the significance level of 0.05; error bars are standard deviations.

Full size image

All the vertical forces measured were positive, which indicated that they were acting on the disc in the downward direction (Fig. 1b) that favored the disc penetration into the soil. The rippled disc had the maximal vertical force of 289 N, which will help to maintain its working depth as compared to the other two discs. As was expected, the notched disc experienced the minimal vertical force of 164 N, which was lower than that of the rippled disc. The plain disc had a medium vertical force, which was not different from the other two discs. The lower vertical force of the notched disc may not necessarily affect its superior ability of soil penetration. The deep working depth created a 65.9% higher vertical force as compared to the shallow depth. The vertical forces of similar magnitudes were also observed in previous studies, such as approximately 200 N in Nalavade et al.23.
There were no significant differences among the discs in terms of the lateral force (Fig. 1c). The notched disc had the minimal lateral force of 215 N. The lateral force increased roughly twofold from 171 to 347 N as the disc was operated from the shallow to deep depths, which was significant. The insignificant difference in the lateral force among the discs was partially attributed to their identical disc angles and similar concavity. Lower lateral forces are usually desired in terms of the frame stability of the implement. The increase of the lateral force as the depth increased indicated a great deal of attention must be paid on the frame strength when designing the disc for deep tillage application.
The soil cutting forces were resultant forces of passive cutting reaction on the concave face and the scrubbing reaction on the convex face for a concave disc24. Both the cutting force and scrubbing force acted at some angle between the horizontal and vertical directions. The projected soil cutting force was against the travel direction in the horizontal direction and downward in the vertical direction; on the other hand, the projected scrubbing force is along the travel direction and upward. The resultant draft force was against the travel direction and the resultant vertical force was downward, which was the same as that of the cutting force. This agreed with the literature that the scrubbing force on the trailing convex side of the disc tends to be minor compared to the cutting force on the leading concave side of the disc22. However, the soil cutting forces were smaller than those reported in Godwin et al.25. The combination of shallow concavity and small disc angle used in this study possibly helped in reducing the soil cutting forces in all three directions. The results agreed with that in Choi and Erback20, where the notched disc had the least forces and the forces were more dependent on the working depth than the disc shape.
Soil displacements
The soil forward displacement was maximized with the rippled disc and was minimized with the notched disc (Fig. 2a). The plain disc resulted in a medium soil forward displacement of 264 mm. During the operation of the notched disc, some soil particles might not be pushed forward, but being passed over by the notches. This could explain the small soil displacements observed for the notched disc. However, statistical analysis did not show any significant differences among the three discs with regard to soil forward displacement. The soil tracers were dislodged 184 mm on average when the discs were used at the shallow depth, which was increased by 73.4% at the deep depth.
Figure 2

Soil displacements of different discs at different working depths in three directions: (a) forward, (b) lateral, and (c) upward; means followed by different lower case letters or upper case letters are significantly different according to Tukey’s test at the significance level of 0.05; error bars are standard deviations.

Full size image

The rippled disc moved the soil tracers the furthest in the lateral direction at 197 mm (Fig. 2b). The notched disc created the minimal soil lateral displacement of 109 mm, which was less than that of the other two discs. The soil lateral displacement was increased by 42.0% as the working depth changed from the shallow to deep depth. The soil lateral displacement was the average displacement of all the tracers in the lateral direction and a positive value denoted the direction pointing toward the concave face of the disc. It was worth noting that soil tracers on the convex side tended to be pushed away in the opposite direction as compared to other tracers as observed in the experiment. This was related to the scrubbing action as described above.
No significant difference was found in the soil vertical displacement among the treatments (Fig. 2c). All the soil vertical displacements were less than 20 mm with an average of 10.6 mm. Similar to the lateral displacement, not all tracers were dislodged in the same direction. However, the majority of them were in an upward direction including the average value. The small upward displacements indicated moderate soil swelling and elevating movements and minimal soil overturning effect of the discs. This was supported by the soil failure pattern study in Nalavade et al.23, which observed that the dominating compressive shear failure pattern of the free-rolling disc discouraged soil inversion actions.
Residue mixing
The rippled disc had the highest residue mixing rate of 23.1%, which was higher than that of the notched disc, being the lowest at 14.7% (Fig. 3). The residue mixing of the plain disc was medium among the three discs. As for the working depth, the shallow depth created a residue mixing of 16.7%, which was lower than that of the deep depth.
Figure 3

Residue mixing of different discs at different working depths; means followed by different lower case letters or upper case letters are significantly different according to Tukey’s test at the significance level of 0.05; error bars are standard deviations.

Full size image

The residue mixing could be used to estimate the amount of residue being incorporated into the soil, given the surface residue before tillage was 7500 kg/ha. Therefore, the rippled disc was the most effective in terms of the residue incorporation at a rate of 2746 kg/ha. The residue incorporation increased by 606 kg/ha as the working depth increased from shallow to deep. Also, deducting the residue mixing from the original residue cover of 63.1% would be the residue cover remaining. None of the treatments resulted in a residue cover less than 30%, which suggested that all treatments would satisfy the requirement of conservation tillage.
Residue cutting
The residue cutting effectiveness of the discs varied from the highest to the lowest as the rippled, notched, and plain with no significant differences were found (Fig. 4). The total residue cutting of the notched disc consisted of one-third of partially cut while no partially cut was observed for the rippled disc. As for the plain disc, roughly a quarter of the total residue cutting was partially cut. The shallow working depth had a numerically higher residue cutting rate than the deep depth: 32.8% versus 22.2%. One in every four residue tracers being cut was partially cut when the discs were operated at the shallow depth. As a comparison, less than one residue was partially cut for every ten residue tracers being cut at the deep depth. The results suggested that the most effective treatment in cutting residues was the rippled disc at the shallow depth. On average, only 27.5% of the residue tracers were being cut, either partially or completely, by the discs. Partial cuts tended to be pushed into the soil and damaged by the discs. The majority of the remaining residue was pushed aside by the discs through disturbed soil.
Figure 4

Residue cutting including completely cut and partially cut of different discs at different working depths; means followed by different lower case letters or upper case letters are significantly different according to Tukey’s test at the significance level of 0.05; error bars are standard deviations.

Full size image

The effects of disc type and working depth on the residue cutting efficiency of the discs differed from the previous studies of disc openers. For example, the plain disc was found to have a much higher residue cutting efficiency than the notched and serrated discs and the efficiency increased as the working depth increased17. The primary cause of the difference was due to the difference in residue cutting mechanism between the angled tillage discs and relatively straight disc openers. The concaved discs disturbed a fair amount of soil ahead of the disc and relied on the edge to “hook” lying residues in order to cut them. Therefore, the rippled and notched discs had numerically higher residue cutting rates than the plain disc thanks to their hooking edges. The shallower the working depth, the less the soil disturbance and the higher the residue cutting efficiency is. On the other hand, a straight disc opener would ride over all possible residues on the path and penetrate the soil without causing significant disturbance to the seedbed. The difference in residue cutting effectiveness can also be accounted for in part by the difference in residue characteristics such as type, percent cover, and moisture content. For instance, wet rice residue with a moisture content of 41.4% at 2000 kg/ha15 versus dry corn stalk with a moisture content of 4.5% at 7500 kg/ha in this study. Previous studies have shown that the cutting performance of the disc openers was significantly affected by the mechanical properties of the residue26 and residue density17.
The numerically higher portion of surface residue being cut at the shallow depth was attributed to a smaller cutting angle. This cutting angle was the angle of absolute velocity vector acting on the residue with the vertical axis in Kushwaha et al.16, whose analytical model showed that the angle of absolute velocity vector for a disc is smaller at a shallower depth. The disc tended to cut or bend the residues at a smaller cutting angle, while the disc tended to push the residue ahead at a larger cutting angle. The notched disc had the numerically highest portion of partially cut among the three discs, which agreed with the results in Bianchini and Magalhaes21. Kushwaha et al.17 also observed that residue pieces were held into the notches and serrations of the discs instead of being cut, being thrown backward as the disc exited from the soil. More

We set out to disentangle the manifold and interacting drivers of fruit production of large, long-lived tropical canopy trees. We used two B. excelsa populations as models given the critical importance of this single species to ecosystem processes, Amazonian livelihoods, and tropical biodiversity conservation. Our findings uncovered that over 10 years, one site (Cachoeira) consistently generated production levels that were threefold higher than that of the other site (Filipinas). Fruit production variation at Cachoeira was also relatively constant at both individual and population levels compared to Filipinas. Yet as anticipated in the tropics (versus temperate regions) where low climate variability minimizes resource variation18, neither population exhibited masting behavior as indicated by synchrony (S).
Given that we hypothesized that fruit production would show similar patterns over time, and common driving variables, we expected weather and weather cues to play important roles in fruit production. Because our research sites are only ~ 30 km apart, we assumed that each population and individual tree experienced approximately the same weather and climatic cues. Our climate model indicated that more wet days during the narrow 3-month dry season prior to flowering resulted in increased fruit production. Furthermore, the model also indicated that when drier atmospheric conditions (represented by VAP) were present and extended beyond the dry season into the flowering period, fruit production tended to be reduced. Still, models that used the simple “year” variable to explain fruit production variation (versus multiple specific, albeit remote climate variables) had better statistical fit. This leads us to question what overall weather conditions might have caused the extremely low and highly variable production levels of 2017; in Filipinas, more than half of the trees did not produce any fruits (Fig. 1). Local Brazil nut harvesters also characterized 2017 as an exceptional nadir in production – a sentiment echoed in popular media across the Amazon basin19.
The year 2015 was a “Very Strong” El Niño year, which followed immediately on a “Weak” one (2014)20. These years relate to our 2017 production because of  > 15-month fruit maturation lag times. Such El Niño events yield sunny, dry conditions in our study region. Over the 10-year study, VAP for 2017 production was the lowest ranked (26.27 hPa), and 2016 was the second lowest (25.37 hPa) (SI Table S2), signaling back-to-back years of persistent low atmospheric moisture. While increases in solar radiation can boost forest productivity21,22, persistent dry conditions and higher accompanying temperatures induce tree stress23, and ultimately higher mortality24. As a canopy emergent, B. excelsa crowns are exposed to greater radiation levels and higher evaporative demand. Hence, they are predicted to be particularly sensitive to drought due to hydraulic stress25, potentially exacerbated by increased water column tension in such exceptionally tall trees23. Still, such large trees access stored groundwater via deep roots more than previously assumed26, and fluctuations in water supply can be moderated by internal storage in stems, roots and leaves27. It is unknown, however, the extent to which two successive El Niño years may have impacted groundwater recharge and storage, and aggravated overall tree stress. There is evidence that canopy trees are resilient to normal Amazonian dry seasons due to deep roots that access water stored from wet season precipitation3,28; yet they are more vulnerable to extended tropical droughts, as demonstrated by the higher rates of large tree, drought-related mortality29. Corlett23 suggested that this tall tree vulnerability can be attributed to the physiological challenges of transporting water from drying soil through lengthy water conduits to exposed leaves. B. excelsa demonstrates drought avoidance by losing leaves during the dry period, but only for a few days in our study region30, where deciduousness is unexceptional and average rainfall falls short of ~ 2000 mm expected for evergreen tropical forests31. Finally, drought inducement experiments have demonstrated that lower rainfall levels over time negatively affect tropical tree fruit production. Throughfall exclusion over a 4-year period had a cumulative negative effect on fruit production (− 12%) of a sub-canopy tropical Rubiaceae, but differences were only significant in 1 year32.
Delayed rainy season onset also may have influenced the extremely low 2017 fruit production. In our region, the rainy season typically begins in September, yet the key 6-month rainfall (DTF; June through November) period that influenced 2017 production was the lowest in our 10-year data set. Moreover, of the entire 117-year CRU data set, the 2017 DTF period was the 16th lowest on record (SI Table S2), indicating that rainy season onset was delayed beyond norms. Since 1979, there has been a delay in dry season end dates (or rainy season onset) and an increase in dry season length for southern Amazonia33. Grogan and Schulze34 reported that delayed rainy season onset had a negative effect on tropical canopy tree growth, but they did not track fecundity. Finally, negative correlations between fruit production and minimum temperatures during both DPF and DTF (dry season prior to, and through flowering, respectively), particularly in Cachoeira, are consistent with other tropical studies that have showed clear negative effects of high nighttime temperatures on tropical tree growth22. In sum, evidence suggests that dry, and perhaps warming, conditions may have produced cascading effects that compromised 2017 fruit production at both sites (Table S2). Still, Cachoeira responded better than Filipinas not only in 2017, but across all years, as indicated by highly significant site effects across models.
Given these results, we explored the role that site differences might play in fruit production. Previous studies have detected subtle differences in demographic structures at our sites, indicating the presence of smaller B. excelsa individuals in the Filipinas population, but without a clear attribution to ecological or socioeconomic factors9. While Cachoeira has a longer history of disturbance (i.e., low-intensity timber harvest), which could influence the dominance of B. excelsa, we lack evidence that this disturbance influences production. Despite close proximity, our sites are located in different watersheds, and are characterized by slightly different forest types and soil characteristics. Specifically, Cachoeira’s significantly higher levels of P and K (Table 1) are informative, as soil P has been positively linked to higher levels of B. excelsa production11,17. Costa35 showed that B. excelsa can be productive in acidic, less fertile soils, while suggesting that Ca is a key macronutrient for this species.
Site quality has been used extensively to explain and predict productivity across diverse forest types for decades36, and inclusion of more site variables (such as depth to water table) would likely yield improved explanations for Cachoeira’s comparatively superior production. Notwithstanding, individual tree differences, regardless of site, offer further fruit production insights. As with almost all trees, B. excelsa reproductive status and fruit production levels are explained by DBH12,16,37,38,39, with the most productive trees in the 100–150 cm DBH range11. Moreover, DBH for these trees is correlated with crown size17, which was a significant and positive explanatory variable for all our production models, although less so for large trees (≥ 100 cm DBH) in Cachoeira versus Filipinas (Table 2, Models 4a & b). Large crowns of individual trees imply greater photosynthetic capacity and sturdy physical structures that support carbohydrate and nutrient demands of the large B. excelsa fruits. Large-diameter trees with big crowns produce more fruits. Furthermore, these trees are tall; all exhibit dominant or co-dominant canopy positions, suggesting fairly unlimited access to light. Notably, while basal area growth was a significant predictor of fruit production in trees More

Given that elasmobranchs are well known for the rate at which they replace their teeth, it is perhaps surprising that anterior teeth are retained long enough for dietarily informative microwear textures to develop. Yet our results demonstrate that tooth microwear textures vary with diet in C. taurus, and show that DMTA can provide an additional, potentially powerful tool for dietary discrimination in elasmobranchs. Furthermore, recent analysis indicates that C. taurus mostly consume prey in one piece30, implying less interaction of teeth with prey than would the case in animals that process their food before swallowing. We predict that for elasmobranchs that bite their prey the relationship between diet and microwear texture will be even stronger than that reported here.
Sampling individuals with different diets reveals increases in PC 1 values that in turn correspond to changes in a number of different ISO texture parameters. In general terms, as noted above, there is a trend towards ‘rougher’ surfaces with increases in the proportion of elasmobranchs in C. taurus diets, and with increasing consumption of benthic elasmobranchs30,31,32 (which may be associated with an increase in the amount of sediment consumed with prey). The increase in variance of PC1 values may also reflect increased diversity of prey types30,31,32 in larger individuals. To a degree, the greater variance might reflect the greater difference between maximum development of ‘rough’ microwear texture in a tooth near the end of its functional life compared to a smooth, recently erupted tooth. Either way, our results indicate that microwear texture tracks diet, but more work will be required to tease apart these additional factors.
Our analyses indicate that the tooth microwear textures of Specimen 5, from a different geographic area to other specimens, and for which we have no dietary data, are closely comparable to those of samples 1, 2 and 3, in terms of both values and variances. On this basis we interpret specimen 5 to have had a diet dominated by fish. The larger size of this specimen (at ca. 335 cm, larger than any other specimens analysed) lends further support to the hypothesis that microwear texture is tracking diet, and not size. Our dietary predictions regarding C. taurus from this area could be tested using traditional stomach contents, or stable isotope analyses, but this is outside the scope of the present study.
Our results also suggest that application of DMTA to analysis of the diet of individual sharks will produce more reliable results if multiple teeth are sampled rather than a single tooth. Comparing the six teeth of the aquarium individuals (fed only fish) with six teeth sampled randomly from the wild individuals (which had more varied diets) revealed significant differences in every sub-sampling (Supplementary Table S5). However the number of parameters displaying a significant difference between wild and aquarium teeth varied, and fewer significant differences than were found than analyses comparing the aquarium teeth to multiple teeth from each wild individual. This suggests that analyses based on single isolated teeth rather than those from jaws, a situation that would commonly arise in analyses of fossil teeth, have the potential to detect differences between populations and species with different diets, but will be less sensitive than analyses based on multiple teeth per individual. To a certain extent, this will be offset in collections of isolated fossil teeth because the vast majority are teeth that were shed at the end of the functional cycle, so there will be much less sampling of recently erupted teeth with less well-developed microwear textures. (Due to the rate of tooth replacement in elasmobranchs, the number of teeth shed by an individual in its lifetime outnumber the number of teeth in the individuals jaw at time of death by several orders of magnitude).
Drawing wider comparisons with microwear texture analyses in other groups of vertebrates, of the relationship between diet and 3D microwear texture based on ISO parameters, the number of parameters that differ between samples of C. taurus is larger than most previous studies, probably due to greater differences in material properties of food between the samples compared. Wild C. taurus consume a wider variety of prey than aquarium fed C. taurus. Wild individuals consume ‘harder’ prey items, whilst interacting with the natural environment. A wild individual consuming a benthic elasmobranch will have to bite through dermal denticles, a larger cartilage skeleton and inevitably will ingest some sediment during the process. In contrast aquarium individuals are largely fed whole and partial fish within the water column, a much ‘softer’ diet. Comparison of this study to others analysing vertebrate diet, repeatedly display significant differences in certain parameters when comparing groups with harder/softer diets. Purnell and Darras23 found that Sdq, Sdr, Vmc, Vvv, Sk and Sa discriminated best between the specialist durophagous and more opportunist durophagous fish in their study (based on ANOVA and PCA), with these parameters also differing between populations of the opportunist durophage Archosargus probatocephalus with different proportions of hard prey in their diets. Of these parameters, Sk, Sa, Vmc, and Vvv produce pairwise differences between C. taurus samples (between 1 and 4). These parameters capture aspects of surface heights and the volumes of material within the core and voids in valleys, respectively (Supplementary Table S1 online). All increase in value as the proportion of elasmobranchs in the diet increases, the same as the pattern of increase with durophagy seen in Archosargus probatocephalus and Anarhichas lupus23. Vmc, Vvv, and Sk were also found to increase with the amount of hard-shelled prey in the diet of cichlids24. This means that ‘harder’ diets produce tooth surface textures with greater core depth and an increase in the volumes of core material and valleys. In short ‘harder’ diets produce rougher tooth surfaces.
This conclusion is also supported by a recent DMTA study on reptiles29, which exhibit significant overlap with sharks in the parameter trends correlating with ‘harder’ diets. Of the parameters correlating with increasing PC 1 values in sharks, parameters correlated with increasing dietary ‘hardness’ in reptiles include those capturing aspects of texture height (Sa, Sq, S5z), the number of peaks (Spk), and the depth, void volume and material volume of the core (Sk, Vvc, Vmc). Once again ‘harder’ diets produce rougher tooth surfaces.
Other studies, although focussed on terrestrial rather than aquatic vertebrates, have found similar patterns. Vmc, Vvc, Vvv, and Sa increase with more abrasive diets in grazing ungulate mammals34; Vmc, Vvv and Sk increase with increasingly ‘hard’ prey in insectivorous bats21. Unlike other studies, the latter found Sa (the average surface height) to decrease with harder diets26. A recent study of bats and moles35 found that, like sharks, increasing the ‘hardness’ of the prey creates rougher tooth surfaces that can be defined by increases in Sa, Vmc, VVc values (amongst others) and a decrease in Sds values (amongst others). More

1.
Pantoja-Hernández, L. & Martínez-García, J. C. Retroactivity in the context of modularly structured biomolecular systems. Front. Bioeng. Biotechnol. 3, 85 (2015).
PubMed  PubMed Central  Article  Google Scholar 
2.
Jayanthi, S. & Del Vecchio, D. Retroactivity attenuation in bio-molecular systems based on timescale separation. IEEE Trans. Autom. Control 56, 748–761 (2011).
MathSciNet  Article  Google Scholar 

3.
Gyorgy, A. et al. Isocost lines describe the cellular economy of genetic circuits. Biophys. J. 109, 639–646 (2015).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

4.
Summers, D. The kinetics of plasmid loss. Trends Biotechnol 9, 273–278 (1991).
CAS  PubMed  Article  PubMed Central  Google Scholar 

5.
Mishra, D., Rivera, P. M., Lin, A., Del Vecchio, D. & Weiss, R. A load driver device for engineering modularity in biological networks. Nat. Biotechnol. 32, 1268–1275 (2014).
CAS  PubMed  PubMed Central  Article  Google Scholar 

6.
Weiße, A. Y., Oyarzún, D. A., Danos, V. & Swain, P. S. Mechanistic links between cellular trade-offs, gene expression, and growth. Proc. Natl. Acad. Sci. USA 112, E1038–E1047 (2015).
ADS  PubMed  Article  CAS  PubMed Central  Google Scholar 

7.
Brenner, K., You, L. & Arnold, F. H. Engineering microbial consortia: a new frontier in synthetic biology. Trends Biotechnol 26, 483–489 (2008).
CAS  PubMed  Article  PubMed Central  Google Scholar 

8.
Kennedy, T. A. et al. Biodiversity as a barrier to ecological invasion. Nature 417, 636–638 (2002).
ADS  CAS  PubMed  Article  PubMed Central  Google Scholar 

9.
Beyter, D. et al. Diversity, productivity, and stability of an industrial microbial ecosystem. Appl. Environ. Microbiol. 82, 2494–2505 (2016).
CAS  PubMed  PubMed Central  Article  Google Scholar 

10.
Butler, G. J. & Wolkowicz, G. S. K. A mathematical model of the chemostat with a general class of functions describing nutrient uptake. SIAM J. Appl. Math. 45, 138–151 (1985).
MathSciNet  Article  Google Scholar 

11.
Foster, K. R. & Bell, T. Competition, not cooperation, dominates interactions among culturable microbial species. Curr. Biol. 22, 1845–1850 (2012).
CAS  PubMed  Article  PubMed Central  Google Scholar 

12.
Hibbing, M. E., Fuqua, C., Parsek, M. R. & Peterson, S. B. Bacterial competition: surviving and thriving in the microbial jungle. Nat. Rev. Microb. 8, 15–25 (2010).
CAS  Article  Google Scholar 

13.
Freilich, S. et al. Competitive and cooperative metabolic interactions in bacterial communities. Nat. Commun. 2, 589 (2011).
ADS  PubMed  Article  CAS  PubMed Central  Google Scholar 

14.
Zelezniak, A. et al. Metabolic dependencies drive species co-occurrence in diverse microbial communities. Proc. Natl. Acad. Sci. USA 112, 6449–6454 (2015).
ADS  CAS  PubMed  Article  PubMed Central  Google Scholar 

15.
May, A. et al. Kombucha: a novel model system for cooperation and conflict in a complex multi-species microbial ecosystem. PeerJ 7, e7565 (2019).
PubMed  PubMed Central  Article  Google Scholar 

16.
Czaran, T. L., Hoekstra, R. F. & Pagie, L. Chemical warfare between microbes promotes biodiversity. Proc. Natl. Acad. Sci. USA 99, 786–790 (2002).
ADS  CAS  PubMed  Article  PubMed Central  Google Scholar 

17.
Dinh, C. V., Chen, X. & Prather, K. L. J. Development of a quorum-sensing based circuit for control of coculture population composition in a naringenin production system. ACS Synth. Biol. 9, 590–597 (2020).
CAS  PubMed  Article  PubMed Central  Google Scholar 

18.
Stephens, K., Pozo, M., Tsao, C.-Y., Hauk, P. & Bentley, W. E. Bacterial coculture with cell signaling translator and growth controller modules for autonomously regulated culture composition. Nat. Commun. 10, 4129 (2019).
ADS  PubMed  PubMed Central  Article  CAS  Google Scholar 

19.
Liu, F., Mao, J., Lu, T. & Hua, Q. Synthetic, context-dependent microbial consortium of predator and prey. ACS Synth. Biol. 8, 1713–1722 (2019).
CAS  PubMed  Article  PubMed Central  Google Scholar 

20.
Gupta, A., Reizman, I. M. B., Reisch, C. R. & Prather, K. L. J. Dynamic regulation of metabolic flux in engineered bacteria using a pathwayindependent quorum-sensing circuit. Nat. Biotechnol. 35, 273–279 (2017).
CAS  PubMed  PubMed Central  Article  Google Scholar 

21.
Scott, S. R. & Hasty, J. Quorum sensing communication modules for microbial consortia. ACS Synth. Biol. 5, 969–977 (2016).
CAS  PubMed  PubMed Central  Article  Google Scholar 

22.
Balagaddé, F. K. et al. A synthetic Escherichia coli predator–prey ecosystem. Mol. Syst. Biol. 4, 187 (2008).
PubMed  PubMed Central  Article  Google Scholar 

23.
Kong, W., Meldgin, D. R., Collins, J. J. & Lu, T. Designing microbial consortia with defined social interactions. Nat. Chem. Biol. 14, 821–829 (2018).
CAS  PubMed  Article  Google Scholar 

24.
Rebuffat S. M. (ed. Kastin, A. J.) In Handbook of Biologically Active Peptides 129–137 (Elsevier, 2013).

25.
Geldart, K., Forkus, B., McChesney, E., McCue, M. & Kaznessis, Y. pMPES: a modular peptide expression system for the delivery of antimicrobial peptides to the site of gastrointestinal infections using probiotics. Pharmaceuticals 9, 60 (2016).
PubMed Central  Article  CAS  PubMed  Google Scholar 

26.
Fedorec, A. J. H. et al. Two new plasmid post-segregational killing mechanisms for the implementation of synthetic gene networks in Escherichia coli. iScience 14, 323–334 (2019).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

27.
MacDonald, J. T., Barnes, C., Kitney, R. I., Freemont, P. S. & Stan, G.-B. V. Computational design approaches and tools for synthetic biology. Integr. Biol. 3, 97 (2011).
Article  Google Scholar 

28.
Kirk, P., Thorne, T. & Stumpf, M. P. H. Model selection in systems and synthetic biology. Curr. Opin. Biotechnol. 24, 767–774 (2013).
CAS  PubMed  Article  Google Scholar 

29.
Barnes, C. P., Silk, D., Sheng, X. & Stumpf, M. P. H. Bayesian design of synthetic biological systems. Proc. Natl. Acad. Sci. USA 108, 15190–15195 (2011).
ADS  CAS  PubMed  Article  Google Scholar 

30.
Woods, M. L., Leon, M., Perez-Carrasco, R. & Barnes, C. P. A Statistical approach reveals designs for the most robust stochastic gene oscillators. ACS Synth. Biol. 5, 459–470 (2016).
CAS  PubMed  PubMed Central  Article  Google Scholar 

31.
Leon, M., Woods, M. L., Fedorec, A. J. H. & Barnes, C. P. A computational method for the investigation of multistable systems and its application to genetic switches. BMC Syst. Biol. 10, 130 (2016).
PubMed  PubMed Central  Article  Google Scholar 

32.
Yeoh, J. W. et al. An automated biomodel selection system (BMSS) for gene circuit designs. ACS Synth. Biol. 8, 1484–1497 (2019).
CAS  PubMed  Article  PubMed Central  Google Scholar 

33.
Beal, J. et al. An end-to-end workflow for engineering of biological networks from high-level specifications. ACS Synth. Biol. 1, 317–331 (2012).
CAS  PubMed  Article  PubMed Central  Google Scholar 

34.
Rodrigo, G. & Jaramillo, A. AutoBioCAD: full biodesign automation of genetic circuits. ACS Synth. Biol. 2, 230–236 (2013).
CAS  PubMed  Article  PubMed Central  Google Scholar 

35.
Friedman, J. & Gore, J. Ecological systems biology: the dynamics of interacting populations. Current Opinion in Systems Biology 1, 114–121 (2017).
Article  Google Scholar 

36.
Toni, T., Welch, D., Strelkowa, N., Ipsen, A. & Stumpf, M. P. H. Approximate Bayesian computation scheme for parameter inference and model selection in dynamical systems. J. R. Soc. Interface 6, 187–202 (2009).
PubMed  Article  PubMed Central  Google Scholar 

37.
Kass, R. E. & Raftery, A. E. Bayes factors. J. Am. Stat. Assoc. 90, 773–795 (1995).
MathSciNet  Article  Google Scholar 

38.
Salis, H. M., Mirsky, E. A. & Christopher, C. Automated design of synthetic ribosome binding sites to control protein expression. Nat. Biotechnol. 27, 946–950 (2009).
CAS  PubMed  PubMed Central  Article  Google Scholar 

39.
Marisch, K. et al. A Comparative analysis of industrial Escherichia coli K-12 and B strains in high-glucose batch cultivations on process-, transcriptomeand proteome level. PLoS ONE 8, e70516 (2013).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

40.
Treloar, N. J., Fedorec, A. J. H., Ingalls, B. & Barnes, C. P. Deep reinforcement learning for the control of microbial co-cultures in bioreactors. PLOS Comput. Biol. 16, e1007783 (2020).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

41.
Lee, D. D. & Seung, H. S. Learning the parts of objects by non-negative matrix factorization. Nature 401, 788–791 (1999).
ADS  CAS  PubMed  Article  PubMed Central  Google Scholar 

42.
Kerner, A., Park, J., Williams, A. & Lin, X. N. A programmable Escherichia coli consortium via tunable symbiosis. PLoS ONE 7, e34032 (2012).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

43.
Zhou, K., Qiao, K., Edgar, S. & Stephanopoulos, G. Distributing a metabolic pathway among a microbial consortium enhances production of natural products. Nat. Biotechnol. 33, 377–383 (2015).
CAS  PubMed  PubMed Central  Article  Google Scholar 

44.
Shou, W., Ram, S. & Vilar, J. M. G. Synthetic cooperation in engineered yeast populations. Proc. Natl. Acad. Sci. USA 104, 1877–1882 (2007).
ADS  CAS  PubMed  Article  PubMed Central  Google Scholar 

45.
Pande, S. et al. Fitness and stability of obligate cross-feeding interactions that emerge upon gene loss in bacteria. ISME J 8, 953–962 (2014).
CAS  PubMed  Article  PubMed Central  Google Scholar 

46.
Yurtsev, E. A., Conwill, A. & Gore, J. Oscillatory dynamics in a bacterial crossprotection mutualism. Proc. Natl. Acad. Sci. USA 113, 6236–6241 (2016).
CAS  PubMed  Article  PubMed Central  Google Scholar 

47.
Hosoda, K. et al. Cooperative adaptation to establishment of a synthetic bacterial mutualism. PLoS ONE 6, e17105 (2011).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

48.
Zhang, X. & Reed, J. L. Adaptive evolution of synthetic cooperating communities improves growth performance. PLoS ONE 9, e108297 (2014).
ADS  PubMed  PubMed Central  Article  CAS  Google Scholar 

49.
Chen, Y., Kim, J. K., Hirning, A. J., Josi, K. & Bennett, M. R. Emergent genetic oscillations in a synthetic microbial consortium. Science 349, 986–989 (2015).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

50.
Bernstein, H. C., Paulson, S. D. & Carlson, R. P. Synthetic Escherichia coli consortia engineered for syntrophy demonstrate enhanced biomass productivity. J. Biotechnol. 157, 159–166 (2012).
CAS  PubMed  Article  Google Scholar 

51.
Scott, S. R. et al. A stabilized microbial ecosystem of self-limiting bacteria using synthetic quorum-regulated lysis. Nat. Microbiol. 2, 17083 (2017).
CAS  PubMed  PubMed Central  Article  Google Scholar 

52.
Ziesack, M. et al. Engineered Interspecies amino acid cross-feeding increases population evenness in a synthetic bacterial consortium. mSystems 4, e00352–19 (2019).
PubMed  PubMed Central  Article  Google Scholar 

53.
Liao, M. J., Din, M. O., Tsimring, L. & Hasty, J. Rock-paper-scissors: engineered population dynamics increase genetic stability. Science 365, 1045–1049 (2019).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

54.
Ahn, J. et al. Human gut microbiome and risk for colorectal cancer. J. Natl Cancer Inst 105, 1907–1911 (2013).
CAS  PubMed  PubMed Central  Article  Google Scholar 

55.
Stokell, J. R. et al. Analysis of changes in diversity and abundance of the microbial community in a cystic fibrosis patient over a multiyear period. J. Clin. Microbiol. 53, 237–247 (2015).
CAS  PubMed  Article  PubMed Central  Google Scholar 

56.
Louca, S. et al. Function and functional redundancy in microbial systems. Nat. Ecol. Evol. 2, 936–943 (2018).
PubMed  Article  PubMed Central  Google Scholar 

57.
Tyson, G. W. et al. Community structure and metabolism through reconstruction of microbial genomes from the environment. Nature 428, 37–43 (2004).
ADS  CAS  PubMed  Article  PubMed Central  Google Scholar 

58.
Wang, X., Policarpio, L., Prajapati, D., Li, Z. & Zhang, H. Developing E. coli– E. coli co-cultures to overcome barriers of heterologous tryptamine biosynthesis. Metab. Eng. Commun. 10, e00110 (2020).
PubMed  Article  PubMed Central  Google Scholar 

59.
Yuan, S. F., Yi, X., Johnston, T. G. & Alper, H. S. De novo resveratrol production through modular engineering of an Escherichia coli–Saccharomyces cerevisiae co-culture. Microb. Cell Factor 19, 143 (2020).
CAS  Article  Google Scholar 

60.
Friedman, J., Higgins, L. M. & Gore, J. Community structure follows simple assembly rules in microbial microcosms. Nat. Ecol. Evol 1, 109 (2017).
PubMed  Article  Google Scholar 

61.
Carmona-Fontaine, C. & Xavier, J. B. Altruistic cell death and collective drug resistance. Molecular Systems Biology 8, 627 (2012).
PubMed  PubMed Central  Article  Google Scholar 

62.
Tanouchi, Y., Pai, A., Buchler, N. E. & You, L. Programming stress-induced altruistic death in engineered bacteria. Mol. Syst. Biol. 8, 626 (2012).
PubMed  PubMed Central  Article  CAS  Google Scholar 

63.
Ackermann, M. et al. Self-destructive cooperation mediated by phenotypic noise. Nature 454, 987–990 (2008).
ADS  CAS  PubMed  Article  Google Scholar 

64.
Williams, G. T. Programmed cell death: a fundamental protective response to pathogens. Trends Microbiol 2, 463–464 (1994).
CAS  PubMed  Article  Google Scholar 

65.
Calles, B., Goñi-Moreno, Á. & Lorenzo, V. Digitalizing heterologous gene expression in Gram-negative bacteria with a portable ON/OFF module. Mol. Syst. Biol. 15, e8777 (2019).
CAS  PubMed  PubMed Central  Article  Google Scholar 

66.
Fedorec, A., Karkaria, B., Sulu, M. & Barnes, C. Single strain control of microbial consortia. bioRxiv, https://doi.org/10.1101/2019.12.23.887331 (2019).

67.
Bell, T., Newman, J. A., Silverman, B. W., Turner, S. L. & Lilley, A. K. The contribution of species richness and composition to bacterial services. Nature 436, 1157–1160 (2005).
ADS  CAS  PubMed  Article  Google Scholar 

68.
Hsu, R. H. et al. Venturelli. Microbial interaction network inference in microfluidic droplets. Cell Syst 9, 229–242.e4 (2019).
CAS  PubMed  PubMed Central  Article  Google Scholar 

69.
Doekes, H. M., De Boer, R. J. & Hermsen, R. Toxin production spontaneously becomes regulated by local cell density in evolving bacterial populations. PLoS Comput. Biol. 15, e1007333 (2019).
CAS  PubMed  PubMed Central  Article  Google Scholar 

70.
McNaughton, S. J. Stability and diversity of ecological communities. Nature 274, 251–253 (1978).
ADS  Article  Google Scholar 

71.
Sterner, R. W., Bajpai, A. & Adams, T. The enigma of food chain length: absence of theoretical evidence for dynamic constraints. Ecology 78, 2258–2262 (1997).
Article  Google Scholar 

72.
Barabás, G., Michalska-Smith, M. J. & Allesina, S. Self-regulation and the stability of large ecological networks. Nat. Ecol. Evol. 1, 1870–1875 (2017).
PubMed  Article  PubMed Central  Google Scholar 

73.
Thébault, E. & Fontaine, C. Stability of ecological communities and the architecture of mutualistic and trophic networks. Science 329, 853–856 (2010).
ADS  PubMed  Article  CAS  PubMed Central  Google Scholar 

74.
Tang, S., Pawar, S. & Allesina, S. Correlation between interaction strengths drives stability in large ecological networks. Ecol. Lett. 17, 1094–1100 (2014).
PubMed  Article  PubMed Central  Google Scholar 

75.
Harris, C. R. et al. Array programming with NumPy. Nature 585, 357–362 (2020).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

76.
Virtanen, P. et al. SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat. Methods 17, 261–272 (2020).
CAS  PubMed  PubMed Central  Article  Google Scholar 

77.
Siek, J. G., Lee, L.-Q., Lumsdaine, A. The Boost Graph Library, 243 (Addison-Wesley, 2002).

78.
Pedregosa, F. et al. Scikit-learn: machine learning in Python. J. Mach. Learn. Res. 12, 2825–2830 (2011).
MathSciNet  Google Scholar 

79.
Harper, M., et al. python-ternary: ternary plots in python. Zenodo https://doi.org/10.5281/zenodo.594435 (2019).

80.
Wickham, H. ggplot2-Positioning Elegant Graphics for Data Analysis (Springer-Verlag New York, 2016).

81.
Kylilis, N., Tuza, Z. A., Stan, G. B. & Polizzi, K. M. Tools for engineering coordinated system behaviour in synthetic microbial consortia. Nat. Commun. 9, 2677 (2018).
ADS  PubMed  PubMed Central  Article  CAS  Google Scholar 

82.
Senn, H., Lendenmann, U., Snozzi, M., Hamer, G. & Egli, T. The growth of Escherichia coli in glucose-limited chemostat cultures: a re-examination of the kinetics. BBA—Gen. Subj. 1201, 424–436 (1994).
Article  Google Scholar 

83.
Destoumieux-Garzón, D. The iron-siderophore transporter FhuA is the receptor for the antimicrobial peptide microcin J25: role of the microcin Val11-Pro16 β-hairpin region in the recognition mechanism. Biochem. J. 389, 869–876 (2005).
PubMed  PubMed Central  Article  Google Scholar 

84.
Kaur, K. et al. Characterization of a highly potent antimicrobial peptide microcin N from uropathogenic Escherichia coli. FEMS Microbiology Letters 363, fnw095 (2016).
PubMed  Article  CAS  PubMed Central  Google Scholar 

85.
Andersen, K. B. & Meyenburg, K. V. Are growth rates of Escherichia coli in batch cultures limited by respiration? J. Bacteriol. 144, 114–123 (1980).
CAS  PubMed  PubMed Central  Article  Google Scholar 

86.
Marenda, M., Zanardo, M., Trovato, A., Seno, F. & Squartini, A. Modeling quorum sensing trade-offs between bacterial cell density and system extension from open boundaries. Sci. Rep. 6, 39142 (2016).
ADS  CAS  PubMed  PubMed Central  Article  Google Scholar 

87.
Destoumieux-Garzón, D. et al. Microcin E492 antibacterial activity: evidence for a TonB-dependent inner membrane permeabilization on Escherichia coli. Mol. Microbiol. 49, 1031–1041 (2003).
PubMed  Article  CAS  PubMed Central  Google Scholar 

88.
Karkaria, B. D., Fedorec, A. J. H. & Barnes, C. P. Automated design of synthetic microbial communities. Zenodo https://doi.org/10.5281/zenodo.4266261 (2020). More