Abstract
We present new evidence about the relationship between military conflict and city population growth in Europe from the fall of Charlemagne’s empire to the start of the Industrial Revolution. Military conflict was a main feature of European history. We argue that cities were safe harbors from conflict threats. To test this argument, we construct a novel database that geocodes the locations of more than 800 conflicts between 800 and 1799. We find a significant, positive, and robust relationship that runs from conflict exposure to city population growth. Our analysis suggests that military conflict played a key role in the rise of urban Europe.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
City founding years often long preceded the fall of Charlemagne’s empire. Boone (2013, p. 221) writes: “Broaching the subject of the origins of medieval urban Europe may wrongly suggest that the city was a totally new phenomenon in European history. In fact, in many cases medieval urbanity was built on a solid urban tradition, at least in those parts of Europe heavily influenced by the Roman Empire or Romanitas.” After the fall of the Roman Empire, urban centers typically saw decay, but did not typically disappear (Boone 2013, p. 222). In addition, new cities were founded during the medieval era throughout Europe including Central Europe (Livi-Bacci 2000, pp. 21–27).
Anecdotes highlight the historical role of cities as safe harbors beyond Europe. For example, Reid (2012, pp. 3–4) writes about Sub-Saharan Africa: “Elsewhere, warfare has been one of the drivers of urbanization, especially in the Western half of the continent; in some areas, the roots of urbanization lie in itinerant royal camps comprising armies and their enormous entourages, whereas in other circumstances, communities have come together in fortified urban clusters for protection, a phenomenon associated particularly with the nineteenth century—among the Yoruba, for example, or the Nyamwezi.” [Our italics].
Thus, we use all available population data, including for cities with less than 5000 inhabitants at some point in time. Table A2 of the online appendix shows the result for the most stringent specification (i.e., column 4 of Table 3) when we (i) exclude cities that never reached 10,000 inhabitants and (ii) only include city-century observations for city populations greater than or equal to 10,000. For (i), the estimate for \(C_{i,g,t-1}\) is 0.067, with p-value 0.031. For (ii), the estimate for \(C_{i,g,t-1}\) is 0.063, with p-value 0.057 (we lose nearly 40 % of observations). Table A9 of the online appendix shows the result for the most stringent specification when we normalize per-century city populations according to Black and Henderson (1999). The estimate for \(C_{i,g,t-1}\) is 0.092, with p-value 0.020.
The Bairoch et al. data do not include 1100. Thus, we interpolate (but never extrapolate) observations for 1100. de Vries (1984) is an alternative data source for European historical urban populations. However, the de Vries data do not start until 1500. Bosker et al. (2013) compare the Bairoch et al. and de Vries data for each century from 1500 to 1800. They find very similar estimates for urban populations; the correlation coefficients range between 0.986 and 0.992.
The nature of warfare changed dramatically over the nineteenth century due to improvements in transport and communications technologies and the rise of the mass army (Onorato et al. 2014).
We updated the urban population data according to Bosker et al. (2013) for Bruges, Cordoba, London, Palermo, and Paris.
We use a cylindrical equal area map projection with geometric center (longitude, latitude)=(10.00735, 46.76396), near Davos, Switzerland. Figure 1 displays data in the projection of analysis.
Reyerson (1999) claims that medieval horseback travel from Paris to the Mediterranean coast, a distance of approximately 750 km, took 12 days. At a rate of 50 km per day, this trip would take 15 days. This duration is roughly in line with Reyerson’s claim about medieval horseback travel time for Italy.
Table A8 of the online appendix shows the results for the fixed effects specification (i.e., column 1 of Table 3) and the most stringent specification (i.e., column 4 of Table 3) when we use 75 km \(\times \) 75 km grid cells. For the fixed effects specification, the estimate for \(C_{i,g,t-1}\) is 0.153, with p-value 0.002. For the most stringent specification, the estimate for \(C_{i,g,t-1}\) is 0.049, with p-value 0.095.
Table 2 shows the shares of sample cities and grid cells that saw (at least one) conflict by century. On average, 27 % of cities and 19 % of grid cells saw conflict over 800–1799. The smallest share of cities saw conflict over the tenth century (9 %), while the largest share saw conflict over the eighteenth century (54 %). Similarly, the smallest share of grid cells saw conflict over the tenth century (6 %), while the largest share saw conflict over the eighteenth century (43 %).
Thus, the first observation of \(P_{i,g,t}\) is for 900, because the first observation of \(C_{i,g,t-1}\) measures conflict exposure over 800–99.
Table A15 of the online appendix shows the results when we adjust standard errors for spatial correlation according to Conley (1999) for the most stringent specification (i.e., column 4 of Table 3). To compute spatial correlation-adjusted standard errors, we follow the routine in Fetzer (2014), who extends Hsiang (2010). We assume that spatial correlation linearly decreases in the distance between each sample city up to cutoffs of 100, 200, 300, 400, or 500 km, respectively. The estimates for \(C_{i,g,t-1}\) are unchanged, with p-values that range from 0.003 to 0.004.
Initial log city populations refer to the first available observation for each sample city.
Table A3 of the online appendix shows the result for the most stringent specification (i.e., column 4 of Table 3) when we add city-specific time trends (rather than grid cell-specific time trends). The estimate for \(C_{i,g,t-1}\) is 0.044, with p-value 0.083 (standard errors clustered at city level) or p-value 0.147 (standard errors clustered at grid-cell level). Given that our dependent variable \(P_{i,g,t}\) is in logs, the grid cell-specific (or city-specific) time trends capture exponential trends (Wooldridge 2009, p. 361). Still, Table A3 of the online appendix shows the results for the most stringent specification when we add a quadratic time trend at the city level or grid-cell level. For the former, the estimate for \(C_{i,g,t-1}\) is 0.039, with p-value 0.139 (standard errors clustered at city level) or p-value 0.208 (standard errors clustered at grid-cell level). For the latter, the estimate for \(C_{i,g,t-1}\) is 0.038, with p-value 0.165.
Given network effects, Atlantic trade may have been important for cities near, but not along, the Atlantic coast. Table A7 of the online appendix shows the result that interacts Atlantic coast grid cells with century fixed effects for the most stringent specification (i.e, column 4 of Table 3). The estimate for \(C_{i,g,t-1}\) is 0.049, with p-value 0.076. This table also shows the result for Atlantic coast grid cells and neighboring grid cells. The estimate for \(C_{i,g,t-1}\) is very similar as before. To account for Hanseatic trade, we code Hanseatic cities according to the list in Dollinger (1964, pp. ix–x). Table A7 of the online appendix shows the results that interact Hanseatic cities or Hanseatic grid cells with century fixed effects. The estimates for \(C_{i,g,t-1}\) range from 0.068 to 0.069, with p-values that range from 0.027 to 0.029.
These data are available for grid cells of roughly 55 km \(\times \) 40 km. Bosker et al. match the soil quality data to cities based on latitudes and longitudes.
Andersen et al. (2015) provide data on soil suitability for a plow-positive crop (barley) for NUTS2 units, which we match to sample cities in our database. We test three levels of barley suitability: (i) high, where Andersen et al.’s suitability index exceeds 70, (ii) good, where this index exceeds 55, and (iii) medium, where this index exceeds 40. Table A14 of the online appendix shows the results when we interact soil suitability for barley with century fixed effects for the most stringent specification (i.e., column 4 of Table 3). The estimates for \(C_{i,g,t-1}\) range from 0.069 to 0.072, with p-values that range from 0.034 to 0.045.
To construct the potato suitability data, we follow the Nunn-Qian procedure. First, we match the GIS raster file with global coverage of soil suitability for potato cultivation to each 150 km \(\times \) 150 km grid cell in our database. There are five categories of potato suitability: (i) very suitable land, (ii) suitable land, (iii) moderately suitable land, (iv) marginally suitable land, and (v) unsuitable land. Nunn and Qian define land to be suitable for potato cultivation if it falls within categories (i) to (iii). Finally, we define potato suitability as \(Potato_{i,g}\equiv \ln (1+PotatoArea_{i,g})\).
The data for the other geographic variables are from Bosker et al. (2013).
Note that these variables are “bad” controls (Angrist and Pischke 2009) in the sense that they themselves may be outcomes of military conflict.
To test whether larger cities were more attractive targets over the short term, Table A12 of the online appendix shows the results of the target effect test when we regress the conflict exposure measure \(C_{i,g,t-1}\) on lagged city populations \(P_{i,g,t-1}\), where \(C_{i,g,t-1}\) now equals 1 if there was a military conflict in grid cell g over the first z years of century t, where \(z=10, 20, 30, 40, 50\). The coefficients for \(P_{i,g,t-1}\) are never significant.
In the online appendix, we account for the target effect in another way, by including the lagged dependent variable and re-running the specifications from Table 3. Including the lagged dependent variable induces Nickell bias (Nickell 1981), particularly if the panel’s time dimension T is small (for our sample, \(T=8\)). Still, Table A6 shows the main results when we include lagged log city population as a control. The estimates for \(C_{i,g,t-1}\) range from 0.048 to 0.080, with p-values that range from 0.005 to 0.140 (we lose 20 % of observations). For comparison, the original estimates range from 0.063 to 0.113, with p-values that range from 0.004 to 0.055. Thus, including lagged log city population does not change the main results by much. To address Nickell bias, we can use GMM estimation (Arellano and Bond 1991). However, GMM requires strong assumptions (Angrist and Pischke 2009, p. 245). Namely, GMM uses past differenced lags to instrument for the lagged dependent variable; it is unlikely that past differenced lags are not correlated with the differenced residuals. Still, Table A13 shows the main results when we use GMM. The estimates for \(C_{i,g,t-1}\) range from 0.024 to 0.062, with p-values that range from 0.013 to 0.321 (we lose more than 35 % of observations).
Figure A1 of the online appendix shows the results for the most stringent specification (i.e., column 4 of Table 3) when we exclude each country one by one (there are 26 countries). The estimates for \(C_{i,g,t-1}\) range from 0.044 to 0.077, with p-values that range from 0.014 to 0.098 (24 of 26 p-values are less than 0.050). The lowest point estimate (0.044) occurs when we exclude Spain. Still, this estimate remains significant at the 10 % level. Over the medieval period, Christian rulers expelled Muslim invaders from Spain (the Reconquista). To defend newly recaptured territory, such rulers established fortified cities. To promote repopulation, they often granted individual freedoms to new urban inhabitants (O’Callaghan 2002, pp. 699–700).
Furthermore, excluding the capital city and largest city today for each sample country does not change the main results.
The estimate for conflict exposure is positive but no longer significant when we restrict the conflict sample to the pre-early modern era (e.g., column 4 of Table A5 of the online appendix). Tilly (1992), Parker (1996), and Gennaioli and Voth (2014) highlight the importance of new urban fortifications in early modern Europe called the trace italienne. New fortifications may have reduced the probability of wartime urban destruction. For example, Bosker et al. (2013) provide data for the number of times that a sample city was plundered each century. Between the start of the ninth century and the end of the twelfth century, sample cities were plundered 105 times. However, between the start of the fifteenth century and the end of the eighteenth century, sample cities were only plundered 65 times. Furthermore, the most plundering (61 times) took place over the ninth century, while the least plundering (5 times) took place over the eighteenth century. This evidence suggests that post-fifteenth century improvements in defensive structures may have reduced the likelihood of wartime urban destruction, which may help explain why our results suggest that the safe harbor effect was more important during the early modern era than before. A related explanation is that the safe harbor effect may have become more important after the Black Death and the emergence of a post-Malthusian economy. Greater per capita income translated into greater tax revenues, which (i) increased the likelihood of military conflict (Voigtländer and Voth 2013a, b) and (ii) may have increased the ability of polities to construct new fortifications, thereby reducing the likelihood of wartime urban destruction. This explanation is consistent with the result in column 5 of Table 5, which shows that a target effect was less likely if a city had the military strength to protect its wealth.
Table A10 of the online appendix shows the result for the most stringent specification (i.e., column 4 of Table 3) when we use the number of conflicts and the squared number of conflicts as our variables of interest. The estimate for \(C_{i,g,t-1}\) is 0.023, with p-value 0.051. The estimate for the quadratic term is \(-0.001\), with p-value 0.118. The coefficient for the quadratic term is negative. However, (i) the magnitude is very small and (ii) the negative effect does not apply until 12 conflicts are reached. This number is large; nearly all sample cities were exposed to less conflict over 800–1799.
As described in Sect. 2, a city without fortified walls could still have outer rings of conjoined dwellings that functioned as barriers, or could have defensive palisades or ramparts, or be situated in a naturally protected location, thereby enabling small groups of defenders to fend off larger groups of attackers.
Tracy’s definition of “walled city” counts single or double stone walls, gun platforms placed outside of walls, and bastioned traces, but excludes defensive palisades and ramparts.
The mappings from historical provinces to modern German states are approximations. Tracy presents data for eleven historical provinces, of which we exclude two (East Prussia and Silesia) that do not fall within the borders of modern Germany. Of the nine historical provinces that we include, we map Anhalt to the modern state of Saxony-Anhalt, Brandenburg to Brandenburg, Hesse to Hesse, Meckenburg and Pomerania to Mecklenburg-Vorpommern, Rhineland to Rhineland-Palatinate, Saxony (or, Old Saxony) to Lower Saxony, Schleswig-Holstein to Schleswig-Holstein, and Thuringia to Thuringia.
References
Abramson, S., & Boix, C. (2014). The roots of the industrial revolution: Political institutions or (socially-embedded) know-how?” Working paper, Princeton University.
Acemoglu, D., Johnson, S., & Robinson, J. (2005). The rise of Europe: Atlantic trade, institutional change, and economic growth. American Economic Review, 94, 546–579.
Acemoglu, D., Cantoni, D., Johnson, S., & Robinson, J. (2011). The consequences of radical reform: The French revolution. American Economic Review, 101, 3286–3307.
Andersen, T., Jensen, P., & Skovsgaard, C. (2015). The heavy plough and the agricultural revolution in medieval Europe. Journal of Development Economics, 118, 133–149.
Angrist, J., & Pischke, J. S. (2009). Mostly harmless econometrics. Princeton: Princeton University Press.
Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. Review of Economic Studies, 58, 277–297.
Bairoch, P. (1988). Cities and Economic Development. Chicago: University of Chicago Press.
Bairoch, P., Batou, J., & Chèvre, P. (1988). La Population Des Villes Européenes. Geneva: Librarie Droz.
Bates, R. (2010). Prosperity and violence (2nd ed.). New York: Norton.
Besley, T., & Persson, T. (2009). The origins of state capacity: Property rights, taxation, and politics. American Economic Review, 99, 1218–1244.
Besley, T., & Reynal-Querol, M. (2014). The legacy of historical conflict: Evidence from Africa. American Political Science Review, 108, 319–336.
Black, D., & Henderson, V. (1999). Spatial evolution of population and industry. American Economic Review, 89, 321–327.
Blaydes, L., & Paik, C. (2015). The impact of Holy Land crusades on state formation: War mobilization, trade integration, and political development in Medieval Europe. Forthcoming, International Organization.
Bleakley, H., & Lin, J. (2015). History and the sizes of cities. American Economic Review Papers and Proceedings, 105, 558–563.
Boone, M. (2013). Medieval Europe. In P. Clark (Ed.), The oxford handbook of cities in world history. Oxford: Oxford University Press.
Bosker, M., Buringh, E., & van Zanden, J. L. (2013). From Baghdad to London: Unraveling urban development in Europe and the Arab World, 800–1800. Review of Economics and Statistics, 95, 1418–1437.
Bradbury, J. (2004). The Routledge companion to medieval warfare. London: Routledge.
Brewer, J. (1989). The Sinews of Power. London: Unwin Hyman.
Caferro, W. (2008). Warfare and economy in Renaissance Italy, 1350–1450. Journal of Interdisciplinary History, 39, 167–209.
Cantoni, D., & Yuchtman, N. (2014). Medieval universities, legal institutions, and the commercial revolution. Quarterly Journal of Economics, 129, 823–887.
Clodfelter, M. (2002). Warfare and armed conflicts: A statistical reference, 1500–2000. McFarland: Jefferson.
Conley, T. (1999). GMM estimation with cross-sectional dependence. Journal of Econometrics, 92, 1–45.
De Long, B., & Shleifer, A. (1993). Princes and merchants: European city growth before the industrial revolution. Journal of Law and Economics, 36, 671–702.
de Vries, J. (1984). European urbanization, 1500–1800. Cambridge: Harvard University Press.
Dincecco, M., & Prado, M. (2012). Warfare, fiscal capacity, and performance. Journal of Economic Growth, 17, 171–203.
Dittmar, J. (2011). Information technology and economic change: The impact of the printing press. Quarterly Journal of Economics, 126, 1133–1172.
Dollinger, P. (1964). The German Hansa. Palo Alto: Stanford University Press.
Downing, B. (1992). The military revolution and political change. Princeton: Princeton University Press.
Fetzer, T. (2014). Can workfare programs moderate violence? Evidence from India. Working paper, London School of Economics.
Galor, O., & Weil, D. (2000). Population, technology, and growth: From the malthusian regime to the demographic transition and beyond. American Economic Review, 90, 806–828.
Gennaioli, N., & Voth, H. J. (2014). State capacity and military conflict. Review of Economic Studies (forthcoming).
Glaeser, E. (2011). Triumph of the city. New York: Penguin.
Glaeser, E., & Shapiro, J. (2002). Cities and warfare: The impact of terrorism on urban form. Journal of Urban Economics, 51, 205–224.
Greif, A., & Tabellini, G. (2010). Cultural and institutional bifurcation: China and Europe compared. American Economic Review Papers and Proceedings, 100, 135–140.
Guiso, L., Sapienza, P., & Zingales, L. (2008). Long-Term Persistence. NBER Working Paper 14278.
Gutmann, M. (1980). War and Rural Life in the Early Modern Low Countries. Princeton: Princeton University Press.
Hale, J. (1985). War and society in renaissance Europe, 1450–1620. Baltimore: Johns Hopkins University Press.
Henderson, V., Storeygard, A., & Deichmann, U. (2015). Has climate change driven urbanization in Africa? Working paper. London: London School of Economics.
Hohenberg, P., & Lees, L. (1995). The making of urban Europe, 1000–1994. Cambridge: Harvard University Press.
Hsiang, S. (2010). Temperatures and cyclones strongly associated with economic production in the Caribbean and Central America. Proceedings of the National Academy of Sciences, 107, 15367–15372.
Iyigun, M., Nunn, N., & Qian, N. (2010). Resources and Conflict in the Run-up to Modern Europe. Working paper, University of Colorado.
Jaques, T. (2007). Dictionary of battles and sieges. Westport: Greenwood Press.
Jia, R. (2014). Weather shocks, sweet potatoes, and peasant revolts in Historical China. Economic Journal, 124, 92–118.
Livi-Bacci, M. (2000). The population of Europe. Oxford: Blackwell.
Mann, M. (1986). The autonomous power of the state: Its origins, mechanisms, and results. In J. Hall (Ed.), States in history. Oxford: Oxford University Press.
Motamed, M., Florax, R., & Masters, W. (2014). Agriculture, transportation, and the timing of urbanization: Global analysis at the grid cell level. Journal of Economic Growth, 19, 339–368.
Mumford, L. (1961). The city in history. New York: Harcourt Brace.
Nickell, S. (1981). Biases in dynamic models with fixed effects. Econometrica, 49, 1417–26.
Nunn, N., & Qian, N. (2011). The potato’s contribution to population and urbanization: Evidence from a historical experiment. Quarterly Journal of Economics, 126, 593–650.
O’Brien, P. (2011). The nature and historical evolution of an exceptional fiscal state and its possible significant for the precocious commercialization and industrialization of the British economy from Cromwell to Nelson. Economic History Review, 64, 408–446.
O’Callaghan, J. (2002). Reconquest and repopulation. In M. Gerli (Ed.), Medieval Iberia: An encyclopedia. London: Routledge.
Onorato, M., Scheve, K., & Stasavage, D. (2014). Technology and the era of the mass army. Journal of Economic History, 74, 449–481.
Parker, G. (1996). The military revolution. Cambridge: Cambridge University Press.
Pirenne, H. (1925). Medieval cities. This version published in 1969. Princeton: Princeton University Press.
Reid, R. (2012). Warfare in African history. Cambridge: Cambridge University Press.
Reyerson, K. (1999). Commerce and communications. In D. Abulafia (Ed.), New Cambridge medieval history (Vol. 5). Cambridge: Cambridge University Press.
Reyerson, K. (2000). Medieval walled space: Urban development vs. defense. In J. Tracy (Ed.), City walls. Cambridge: Cambridge University Press.
Rokkan, S. (1975). Dimensions of state formation and nation-building: A possible paradigm for research on variations within Europe. In C. Tilly (Ed.), The formation of national states in Western Europe. Princeton: Princeton University Press.
Rosenthal, J. L., & Wong, R. B. (2011). Before and beyond divergence. Cambridge: Harvard University Press.
Shultz, K., & Weingast, B. (1998). Limited governments, powerful states. In R. Siverson (Ed.), Strategic politicians, institutions, and foreign policy. Ann Arbor: University of Michigan Press.
Smith, A. (1776). The wealth of nations. This version published in 2008. Oxford: Oxford University Press.
Stasavage, D. (2011). States of credit. Princeton: Princeton University Press.
Stasavage, D. (2014). Was Weber right? City autonomy, political oligarchy, and the rise of Europe. American Political Science Review, 108, 337–354.
Tilly, C. (1975). Reflections on the history of European state-making. In C. Tilly (Ed.), The formation of states in Western Europe. Princeton: Princeton University Press.
Tilly, C. (1992). Coercion, capital, and European states, 990–1992. Cambridge: Blackwell.
Tracy, J. (2000a). Introduction. In J. Tracy (Ed.), City walls. Cambridge: Cambridge University Press.
Tracy, J. (2000b). To wall or not to wall: Evidence from medieval Germany. In J. Tracy (Ed.), City walls. Cambridge: Cambridge University Press.
van Bavel, B., Bosker, M., Buringh, E., & van Zanden, J. L. (2013). Economy. In P. Clark (Ed.), Oxford handbook of cities in world history. Oxford: Oxford University Press.
van Zanden, J. L. (2009). The long road to the industrial revolution. Leiden: Brill.
van Zanden, J. L., Bosker, M., & Buringh, E. (2012). The rise and decline of European parliaments, 1188–1789. Economic History Review, 65, 835–861.
Voigtländer, N., & Voth, H. J. (2013a). The three horsemen of riches: Plague, war, and urbanization in early modern Europe. Review of Economic Studies, 80, 774–811.
Voigtländer, N., & Voth, H. J. (2013b). Gifts of Mars: Warfare and Europe’s early rise to riches. Journal of Economic Perspectives, 27, 165–186.
Weber, M. (1922). Economy and Society. This version published in 1978. Berekeley: University of California Press.
White, L. (1962). Medieval technology and social change. Oxford: Oxford University Press.
Wolfe, M. (2000). Walled towns during the French wars of religion (1560–1630). In J. Tracy (Ed.), City walls. Cambridge: Cambridge University Press.
Wolfe, M. (2009). Walled towns and the shaping of France. New York: Palgrave Macmillan.
Wooldridge, J. (2009). Introductory econometrics (4th ed.). Mason, OH: Thomson South-Western.
Acknowledgments
We thank editor Oded Galor and three anonymous referees for valuable comments. Similarly, we thank Pablo Beramendi, Timothy Besley, Carles Boix, Roberto Bonfatti, Eltjo Buringh, Eric Chaney, James Fearon, Jeffry Frieden, Edward Glaeser, Philip Hoffman, Horacio Larreguy, James Morrow, Tommaso Nannicini, Nathan Nunn, Margaret Peters, Hugh Rockoff, Jean-Laurent Rosenthal, Kenneth Shepsle, Ugo Troiano, Julian Wucherpfennig, Jan Luiten van Zanden, Daniel Ziblatt, Fabrizio Zilibotti, and seminar participants at Birmingham, Bristol, Harvard, Harvard PIEP, LSE, Michigan, Modena, NES, Nottingham, PSE, UCL, and numerous conferences. We thank Maarten Bosker, Eltjo Buringh, and Jan Luiten van Zanden for generous data-sharing, and Nicola Fontana, Giovanni Marin, Michael Rochlitz, Nicole Scholtz, and Kerby Shedden for excellent data help. Finally, we thank the National Science Foundation for financial support through Grant SES-1227237.
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Dincecco, M., Onorato, M.G. Military conflict and the rise of urban Europe. J Econ Growth 21, 259–282 (2016). https://doi.org/10.1007/s10887-016-9129-4
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10887-016-9129-4