Do horse chestnut trees grow better in a forest or in a mid-field habitat?

Although the horse chestnut is not a native element of the Polish dendroflora, it is frequently planted in cities (parks, avenues, yards), and forms many road−side avenues outside of cities. At present, most horse chestnut trees display the effects of infection by the horse chestnut leaf miner (HCLM) or by the fungus Guignardia aesculi. The present study is based on a group of horse chestnut trees forming an avenue in NW Poland, in the vicinity of Krzymów. The avenue was planted in the 2nd decade of the 20th century. The trees forming the avenue are growing in a mid−field habitat (FIELD) and in a forest habitat (FOREST), which enables a comparison among trees growing in these habitats. Dendrometric analyses were performed on 100 trees, and dendrochronological analyses were performed for 32 trees, with equal number of trees sampled in each habitat. The FIELD chronology spans 103 years (1916−2018), the FOREST chronology spans 101 years (1918−2018). Our study aimed to compare the height and crown diameter of the horse chestnut trees growing in a forest and in a mid−field habitat, to examine the rate of radial growth, and to compare the growth−climate relationships in trees from each habitat. The horse chestnut trees growing in the forest habitat are higher than those from the mid−field habitat. The crown diameter is comparable in both habitats, and the breast−height diameter (1.3 m above ground, DBH) is higher in the trees growing in the mid−field habitat. The horse chestnut trees from the mid−field habitat are healthier, their leaves are infected to a lesser extent and fall off later. Also their annual increments are wider. Also the minimum, mean and maximum tree−ring widths, and the cumulative radial growth are higher for the mid−field habitat. We also found differences in the growth−climate relationships established for each site, especially for the year preceding growth (the analyzes were performed for temperature (T) over the period 1948−2018, for precipitation (P) over the period 1951−2018 and for sunshine duration (SD) over the period 1965−2018). For the trees growing in the mid−field habitat, air temperature and sunshine duration for July and August of the previous year are highly significant (negative correlation). Precipitation (posi− tive correlation) and sunshine duration (negative correlation) in September of the year preceding growth are significant for the forest trees. These differences may result from the disparate features of the studied habitats, and the contrasting intensity of tree infestation by HCLM and by the fungus G. aesculi.