Table 2.

Best fitting regression models for the impact of weather conditions on pollen concentration and fruit production of European beech stands.

Generative variable	Model	R2	Difference of summer temperatures 1 and 2 years before target year	2 years before target year (summer)		1 year before target year (summer)		target year (spring)
				Temperature	Precipitation	Temperature	Precipitation	Temperature	Precipitation
Pollen	Basic	0.47	xxx	↘	0	↘	0	0	↘
Fruits	Basic	0.30	xxx	↘	0	↘	0	0	0
Pollen	ΔT	0.47	Δ	xxx	0	xxx	0	0	↘
Fruits	ΔT	0.30	Δ	xxx	0	xxx	0	0	0

Meteorological measurements from stations inside the stands, deviations from long-term mean. Basic model: includes summer (June and July) mean temperatures and precipitation sums of the two years before the target year, and spring (April and May) mean temperatures and precipitation sums of the target year. ΔT model: includes difference between summer (June and July) temperatures of the two years before the target year, summer precipitation sums of the two years before the target year, and spring (April and May) mean temperatures and precipitation sums of the target year. xxx: parameter not part of the model. ↘ = lower than average, ↘ = higher than average, Δ = summer temperature difference relevant, 0 = not included in the best fitting model. R² from linear regression models.