Vulnerability of bird communities in small mid-field afforestations: searching for a hierarchy of factors

Krzysztof Kujawa, Res. Centre for Agric. & Forest Envir., Pol. Acad.of Sciences, [email protected] RCAFE, PAS

INTRODUCTION • The study (1999-2002) covered 66 mid-field afforestations (0,06-3,10 ha) located mostly within the Gen. D. Chłapowski Landscape Park METHODS • Habitats of an agricultural landscape are subject to permanent, • Main features of afforestations covered by the study: a) the area smaller then • Bird population density has been estimated with the aid of mapping negative changes, which in consequence lead to reduction of 3,5 ha, b) length bigger then breadth not more then 3-fold. method. Each plot was visited 9 times per breeding season, since 1.04 to biodiversity. 10.07. Duration of single visit in one plot (one afforestation) - 20-80 min. • It is well known that mosaic, differentiated structure of agricultural • Structure of afforestation has been characterised by 16 variables landscape (especially presence of afforestations) markedly provide for SŁONIN PUCOŁOWO (qualitative, such as species composition of tree-stand, type of tree-stand, MARIANOWO GORZYCZKI preservation of different groups of animals, including birds. GORZYCE GOŁĘBIN NW. water conditions etc. and quantitative, such as area of afforestation, percentage KRZYŻANOWO • However, the size of those afforestations is usually relatively small, WITKÓWKI cover of vegetation layers etc.) on the basis of field work and analysis of aerial WRONOWO GOŁĘBIN ST. GAJ BŁOCISZEWO photographs as well as maps. that is why natural processes are strongly influenced by adjacent areas. SPYTKÓWKI • In most studies (in as well as in other countries) the efforts DONATOWO • Neighborhood has been defined as the land adjacent to border-line of

DARNOWO were focused to the importance of structure of afforestations (their size, NOCHOWO afforestation distant up to 50 m from the border-line, defined by MapInfo on R ó w W y s k the basis of aerial photographs. Structure of neighborhood (percentage of shape, vegetation etc.), ignoring the effects of environmental factors oć WYSKOĆ RĄBINEK characterising near adjacent areas. habitats and number of habitat patches) has been analysed with the aid of a WYRZEKA software AreaMeasure. Finally, for further analyses 10 quantitative variables RO GACZEWO W. RĄBIŃ N ROGACZEWO M. have been used, such as share of grasslands, cereals, diversity index (H’) and R ó w W E W ys mean area of habitat patches. Study1 area - k oć Chlapowski2 S

0 1 2 3 km • Analysis of landscape structure around given afforestation has been done AIM OF STUDY Landscape3 Park ŁUSZKOWO WŁAWIE in respect to a circle with the radius of 1,5 km from central point of

4 J ez io and location of KOPASZEWO r o

ZBĘCHY afforestation, with the aid of aerial photographs. The structure of landscape JERKA

Zb ę sampling plots ch 5 ŚWINIEC y The aim of the study was to verify importance of three has been analysed by MapInfo and AreaMeasure and finally 20 variables (afforestations) classes of factors for avifauna of small mid-field 6 have been defined (H’, share of habitats, habitat patch number, wood isolation index (Gustafson i Parker 1994) etc.). afforestations, i.e. factors related to: Scheme of study - elements studied ¾ structure of afforestation, • In order to reduce number of independent variables a principal factors Bird communities analysis has been performed, but only if combined variance represented by ¾ structure habitat in neighborhood, principal factors amounted to more then 75%. Finally, number of variables ¾ structure of landscape. describing neighborhood has been reduced to 3, and number of variables Structure of afforestation Expected result of the research: recognition of a hierarchy characterising structure of landscape - to 11. of the importance of defined classes of factors, and better Structure of neighborhood • All statistical analyses have been done with the aid of Statistica 5.5. understanding the mechanisms, which shape avifauna of

small mid-field woodland islands. It can contribute for Structure of landscape enhancing effectiveness of planting new afforestations in terms of their importance for biodiversity preservation.

Bird community of studied afforestations: Influence of habitat structure on Influence of habitat structure on number of number of species species of visitors (non-breeders ) ¾ 61 breeding species, total bird density - 15,1 pairs/ha

¾ Average bird density per plot - from 4,4 to 47,9 p/ha Structure of Positive effect Negative effect Structure of afforestations ??? –unexplained variance ¾ Average number of species per plot per year - from 1 to 18 ??? afforestations 25% AREA - area of afforestation AREA These „ring” figures show what part of AREA ¾ Total number of species per plot per year - from 3 to 27 variability in species number (or other Number 40% 33% BORDER - relative length of border- bird attributes) may be explained by line of afforestation ¾ Most common (recorded in more then 50% of plots) species : F. SHR_ Number of species step-wise multiple regression (coloured coelebs, E. citrinella, T. merula, S. atricapilla , P. major, H. icterina, P. COV ??? of species caeruleus, C. carduelis and M. striata. 10% part of a ring) and illustrate relative 56% TREE_ importance (%) of individual variable BOR BORD DIVER ¾ DER for analysed bird characters. E.g., (fig. ER Most rare (recorded in single plots) species: T. pilaris, S. rubetra, P. 12% modularis, P. pica, P. perdix, L. naevia, L. fluviatilis, J. torquilla, D. 13% on left) step-wise regression explains... 11% martius, C. oenas, A. atthis, A. scirpaceus. ¾ Most abundant species (dominance >5%): F. coelebs, S. atricapilla, ... 75% of variability of species number E. citrinella, H. icterina and T. merula. and shows that most important variable is AREA. One may see which class of No variables in the model - Number variables is most significant for given Number no influence of neighborhood of species attribute of bird community and which of species variable from given class is most Structure of Frequency distribution of Frequency distribution of important. ??? mean density in plots (p/ha) mean sp. number in plots neighborhood 100% 25 15 ??? Structure of 100% 20 12 neighborhood LSOR 15 9 LP BUILT WOO D 10 6 13% LPWOOD - principal factor, correlated LPWAT AREA - area of afforestation 17% 5 3 positively with number of wood patches and No. of observations No. of observations ER BORDER - relative length of border-line of negatively - with number of large crop fields 0 0 9% 0 5 10 15 20 25 30 35 40 45 50 55 -2 0 2 4 6 8 10 12 14 16 18 20 Number afforestation Number KPWA ??? TER LPWATER - principal factor, correlated of species SHR_COV - percentage cover of shrub layer 54% of species 16% positively with number of water bodies and Birds of interior index of isolation of afforestation Share of habitat-guilds in total density TREE_DIVER - diversity index (H’) of species of forest ??? composition of tree-stand LSORB LSORBUILT - principal factor, correlated 78% Structure of UILT Birds of forest-crops Birds of Forest birds without any Others Structure of positively with share of woodland, and 13% ecotone margin of preferences to margin or landscape LSORBUILT, LPWATER – see Fig. on right landscape negatively - with share of built-up areas and forest interior side orchards

Influence of habitat structure on species Influence of habitat structure on bird density Influence of habitat structure on the rate of richness and abundance in „habitat-guilds” local extinction of breeding species Determination coefficients of step-wise regression model in respect to „habitat- guilds”: EDGE - Birds of forest-crops ecotone, W_EDGE - Birds of margin of AREA - area of afforestation forest, W_ALL - Forest birds without any preferences to margin or interior, W_IN AREA - Birds of interior of forest, S - number of species, D - bird density BORDER - relative length of border-line of AREA 17% 14% afforestation Habitat-guild Structure of Structure of Structure of TREE_ ??? BOR TREE_TYPE - type of tree-stand; its value is TYPE AREA - area of afforestation afforestation neighborhood landscape 45% Bird DER higher when tree-stand is more leafy Extinc- 11% EDGE_S 32 9 density 11% TREE_TYPE - type of tree-stand; its value EDGE_D 35 26 HERB_COV - cover of herb layer tion rate is higher when tree-stand is more leafy TREE_ W_EDGE_S 49 19 TYPE TREE_DIVER - diversity index (H’) of HERB 10% species composition of tree-stand ??? W_EDGE_D 33 – Structure of TREE_ _COV 75% Structure of afforestation DIVER 9% W_ALL_S 68 33 8% afforestation W_ALL_D 45 9 42 N2_ CROPS W_IN_S 32 11 12% W_IN_D 23 5 10 N2_CROPS - principal factor, correlated positively with share of crop fields and Bird Extinc- negatively - with share of woodland and density No variables in the model - grassland no influence of neighborhood tion rate

SUMMARY AND CONCLUSIONS ??? Structure of ??? 88% Variability of species number and bird density was most strongly Structure of neigborhood 100% related to structure of afforestations. Landscape structure influenced neigborhood LSWO avifauna usually less (several-fold) then afforestation structure. LSFRA ODDIV GM Only for W_ALL and EDGE the effect of landscape structure was 11% LSWOODDIV -principal factor, correlated LPOR 17% almost the same as the effect of afforestation structure. Effect of BUILT positively with share of woodland and structure of habitat in neighborhood was very weak. The influence 11% diversity index and negatively - with share Extinc- LSFRAGM - principal factor, correlated of habitat structure on community dynamics was relatively small, Bird of large crop fields density tion rate positively with share of small crop fields, and what suggests that changes in bird community in such small LPORBUILT -principal factor, correlated negatively - with share of large crop fields afforestations is random process or they are dependent on other positively with number of patches with and shelterbelts. ??? factors, which have not been identified in this study. Structure of built-up areas and number of orchards, and ??? 78% negatively - with index of isolation of 83% Structure of landscape afforestations landscape