Large-Scale Climatic Patterns Have Stronger Carry-Over Effects than Local Temperatures on Spring Phenology of Long-Distance Passerine Migrants between Europe and Africa

Abstract

Earlier springs in temperate regions since the 1980s, attributed to climate change, are thought to influence the earlier arrival of long-distance migrant passerines. However, this migration was initiated weeks earlier in Africa, where the Southern Oscillation, Indian Ocean Dipole, North Atlantic Oscillation drive climatic variability, and may additionally influence the migrants. Multiple regressions investigated whether 15 indices of climate in Africa and Europe explained the variability in timing of arrival for seven trans-Saharan migrants. Our response variable was Annual Anomaly (AA), derived from standardized mistnetting from 1982-2021 at Bukowo, Polish Baltic Sea. For each species, the best models explained a considerable part of the annual variation in the timing of spring's arrival by two to seven climate variables. For five species, the models included variables related to temperature or precipitation in the Sahel. Similarly, the models included variables related to the North Atlantic Oscillation (for four species), Indian Ocean Dipole (three), and Southern Oscillation (three). All included the Scandinavian Pattern in the previous summer. Our conclusion is that climate variables operating on long-distance migrants in the areas where they are present in the preceding year drive the phenological variation of spring migration. These results have implications for our understanding of carry-over effects.

Keywords: Africa; Europe; IOD; NAO; SOI; climate change; large-scale climate indices; long-distance migrants; passerine migration; spring phenology.

Figure A1

Location of the bird ringing station Bukowo at the Baltic coast of Poland and its two locations (green symbols) combined as one ringing site: Kopań (54°27′11″ N, 16°24′08″ E) and Bukowo (54°20′13″ N, 16°14′36″ E). Map on the right after Google Maps, 2002, modified.

Figure A2

Habitats where birds were mistnetted at the bird ringing station Bukowo (54°20′13″–54°27′11″ N, 16°14′36″–16°24′08″ E). Photos: Katarzyna Stępniewska.

Figure A3

Model diagnostics for the best model with overall spring AA as the explained variable, and 15 climate variables and the Year as explanatory variables for the Blackcap Sylvia atricapilla (Table A12).

Figure A4

Model diagnostics for the best model with overall spring AA as the explained variable, and 15 climate variables and the Year as explanatory variables for the Lesser Whitethroat Curruca curruca (Table A13).

Figure A5

Model diagnostics for the best model with overall spring AA as the explained variable, and 15 climate variables and the Year as explanatory variables for the Common Whitethroat Curruca communis (Table A14).

Figure A6

Model diagnostics for the best model with overall spring AA as the explained variable, and 15 climate variables and the Year as explanatory variables for the Common Redstart Phoenicurus phoenicurus (Table A15).

Figure A7

Model diagnostics for the best model with overall spring AA as the explained variable, and 15 climate variables and the Year as explanatory variables for the Chiffchaff Phylloscopus collybita (Table A16).

Figure A8

Model diagnostics for the best model with overall spring AA as the explained variable, and 15 climate variables and the Year as explanatory variables for the Willow Warbler Phylloscopus trochilus (Table A17).

Figure A9

Model diagnostics for the best model with overall spring AA as the explained variable, and 15 climate variables and the Year as explanatory variables for the Pied Flycatcher Ficedula hypoleuca (Table A18).

Figure 1

Geographical ranges for the seven long-distance migrants we analysed in the study. Maps by Andreas Trepte after [55], modified.

Figure 2

Many-year average cumulative curves of spring arrivals at Bukowo, Poland, over 1982–2021 for the analysed species.

Figure 3

A division of the year into periods applied in the analyses. Fields with two colours indicate possible overlapping of subsequent life stages for different populations of a species of a long-distance migrant.

Figure 4

The approximate conditions at areas visited by the long-distance Euro-African migrants related with the positive phases of the analysed atmospheric patterns: (A) in November–March at both hemispheres, (B) in March–April and June–July at the northern hemisphere. Rectangles join the opposite centres of each climate pattern. Arrows and their labels indicate approximate areas influenced by each pattern and related conditions. Symbols of the climate indices as in Table 1. The climate indices are shown in relation to the geographical range of an example of a long-distance migrant, the Willow Warbler Phylloscopus trochilus (map by Andreas Trepte after [55], modified). The negative phases have opposite effects on the environment.

Figure 5

Trends and variation for the climate indices used in the study over 1981–2021. Year is used in the equations as the explanatory variable. Statistical significance of the regression equations: * 0.05 < p < 0.1, ** p < 0.01, *** p < 0.001.

Figure 5

Trends and variation for the climate indices used in the study over 1981–2021. Year is used in the equations as the explanatory variable. Statistical significance of the regression equations: * 0.05 < p < 0.1, ** p < 0.01, *** p < 0.001.

Figure 6

Trends for the Annual Anomaly (AA) of spring migration of the seven long-distance migrants at Bukowo, Poland, over 1982–2021. Year is used in the equations as the explanatory variable. Statistical significance of the regression equations: ** p < 0.01.

Figure 7

Partial correlation coefficients for Annual Anomaly (AA) of spring passage at Bukowo, Poland, over 1982–2021, against the climate indices selected in the best multiple regression models. The details of these models are presented in Appendix B, Table A12, Table A13, Table A14, Table A15, Table A16, Table A17 and Table A18.