«LANDSCAPE HETEROGENEITY AND SPECIES RICHNESS AND COMPOSITION: A MIDDLE SCALE STUDY LADA ZÁHLAVOVÁ1, MARTIN KONVIČKA2, ZDENĚK FRIC2, VLADIMÍR ...»
Ekológia (Bratislava) Vol. 28, No. 4, p. 346–362, 2009
LANDSCAPE HETEROGENEITY AND SPECIES
RICHNESS AND COMPOSITION: A MIDDLE SCALE
LADA ZÁHLAVOVÁ1, MARTIN KONVIČKA2, ZDENĚK FRIC2, VLADIMÍR
HULA3, LADISLAVA FILIPOVÁ4
Southern Bohemian University, Faculty of Agronomy, Department of Agroecology, Studentská 13, 370 01 1 České Budějovice, Czech Republic; e-mail: Zahlavova.L@seznam.cz Southern Bohemian University, Faculty of Natural Sciences, Department of Zoology, Branišovská 31, 370 01 2 České Budějovice, Czech Republic; e-mail: email@example.com, firstname.lastname@example.org ÚZRHV AF, Zemědělská 1, 613 00 Brno, Czech Republic; e-mail: email@example.com; firstname.lastname@example.org 3 University of Jan Evangelista Purkyně, Hoření 13, 400 96 Ústí n. Labem, Czech Republic; e-mail: ladislava.
4 email@example.com Abstract Záhlavová L., Konvička M., Fric Z., Hula V., Filipová L.: Landscape heterogeneity and species richness and composition: a middle scale study. Ekológia (Bratislava), Vol. 28, No. 4, p. 346–362, 2009.
The aim of this study was to confirm and prove the influence of habitat heterogeneity to species composition and richness of 4 taxa: higher plants, spiders, butterflies and true bugs. We also wanted to analyze the role of a biotope identity in species richness and composition of observed taxa on various biotopes, e.g. whether the biotope itself weakens or forcens the effect of habitat heterogeneity.
The research was taken place in the years 2003–2004 in Bochov in Doupovske vrchy, a military training zone in Western Bohemia. 4 transect lines (each 6 km long) were established here and each transect was divided into 20 rectangles, each 300 m long. True bugs, spiders and butterflies were collected here using swepping net, ground traps and transect walks. The data about higher plants were collected using classical vegetational relevés.
Habitat heterogeneity strongly affects the species richness of the observed taxons, especially butterflies and true bugs. After adding the biotope identity to the model the impact of habitat het- erogeneity dies away. But, on the contrary, species richness of higher plants and spiders were not so affected by the biotope identity. From heterogeneity predictors, the species richness of higher plants was mainly affected by number of segments and borders and butterfly species richness was strongly affected by number of segments and diversity. By true bugs the strongest impact had
346 Introduction Habitat heterogeneity is among the most discussed theme in ecology. On the macroscale the heterogenity of the landscape is predicted e.g. by the distance of large regions in one state or continent (Fischer et al., 2006; Kerr, Cihlar, 2004), global diversity of tropical rainforests (Kerr, Burkey, 2002), climatic energy (Kerr, Curie, 1999) and the heterogeneity of biomas in one continent (Kerr et al., 1998).On middle scale, the basic predictors of species richness are e.g. valley sides, tops of slopes and wood edges (Dennis, Sparks, 2005) and even the island isolation(Dennis et al., 2000). On the other hand, at local scales, heterogeneous habitats provide more resources, or niches, allowing local co-existence of more species. For instance, recent analyses from butterflies (Konvička et al., 2006) illustrate that as majority of species utilise more resource types during their lifespan, structurarly diverse environments are essential conditions for their persistence (Zimmermann et al., 2005). Especially for many invertebrate species (Krauss et al., 2003; Dennis, Sparks, 2005; Konvička et al.,
2005) the existence of mosaique landscape is very important, because they need some refugia where they live and where they lay their eggs, where to find a nectar sources, etc.
For example some butterfly species need the shrubs where they hide during hot weather and open landscape with extensively managed meadows, where they find some nectar supplies (Zimmermann et al., 2005; Dennis, 2004; Dennis, Sparks, 2005). Each invertebrate species also needs something else to survive, so the patchy landscape is much more rich in invertebrate species, than the uniform landscape. Increase of habitat heterogeneity can also cause highest migration of the species from surroundings. On the oposite side it is also known, that the lack of habitat heterogeneity causes the extinction of some in the past very common species of invertebrates, especially butterflies (Štorch et al., 2003).
From conservation point of view, the crucial issue is the effect of heterogeneity on species richness and composition at middle scale of individual landscapes (decades of km2, i.e. farms, districts, national parks etc.). It is the scale, which is unlike the heterogeneity on the level of whole states and continents practically appliable to the conservation policy. But simultaneously, there also appear the problems, which are not important on the local scale – the diversity of owners relations, etc. Regardless, there is a strong theoretical (Hanski, 2005; Zartmann, Shaw,
2006) and empirical (Hula et al., 2004; Franzen, Ranius, 2004) evidence, that without accent to the conservation in large scales we do not stop the loss of biodiversity.
As a response, multiple studies focusing on landscape level heterogeneity recently appeared across Europe. Söderström et al. (2001) studied the diversity in 31 seminatural grassland in Sweden. They compared richness of plants, birds, butterflies, bumble bees, ground beetles and dung beetles and found, that both the landscape pattern and the way of management is very important for many species. Bengtsson et al. (2005) analysed the effects of organic farming (i.e.
without pesticides and artificial fertilizers) on species richness and abundance. They learned, that organic farming was beneficial, but the way of farming itself had only partial influence if compared to average size of field, i.e., landscape heterogeneity. Weibull et al. (2000) also compared the diversity of butterflies in organic and conventional farmlands, finding negligible effect of organic farming, but strong effects of landscape grain size. Ouinn et al. (2004) 347 studied in western France, whether butterfly use of herbaceous patches depends on the nature of those patches and their management and whether they prefer any. They came across that the butterflies need more patches, because of their different behavioral strategy during day and that is why they need a patchy landscape. Burel et al. (1998) compared the biodiversity in contrasted landscape units in selected region in western France. They measured biodiversity of small mammals, birds, insects and plants using Shannon´s diversity index, equitability and similarity indices. They wanted to study the effect of the agricultural intensification to landscape grain size and the effect of grain size to the diversity of these taxons. They learned, that the intensification of agriculture does not always lead to a decrease in species richness, but to several functional responses according to taxonomic groups, either no modification, or stability by replacement of species, or loss of species.
The major hindrance in studies of effects of habitat heterogeneity on species richness is distinguishing heterogeneity effects from effects biotope identity. Only a few studies about habitat heterogeneity did not forget, that the richness of various biotopes is different and that in heterogeneous landscape there will be also rich biotopes, which can be missing in the homogeneous landscape (Burel et al., 1998; Nikodemus et al., 2005; Herzog et al., 2001).
The second problem is a space autocorrelation (Lennon, 2000). There are only a few studies about habitat heterogeneity, which use this method. Correlation between an autocorrelated response variable and each of a set of explanatory variables is strongly biased in favour of those explanatory variables that are highly autocorrelated – the expected magnitude of the correlation coefficient increases with autocorrelation even if the spatial patterns are completely independent. Similarly, multiple regression analysis finds highly autocorrelated explanatory variables „significant“ much more frequently than it should. The chances of mistakenly identifying a „significant“ slope across an autocorrelated pattern is very high if classical regression is used. Consequently, under these circumstances strongly autocorrelated environmental factors reported in the literature as associated with ecological patterns may not actually be significant.
It is likely that these factors wrongly described as important constitute a red-shifted subset of the set of potential explanations, and that more spatially discontinuous factors are actually relatively more important than their present status suggests.
This thesis contributes to the debate by studying effects of landscape heterogeneity on species richness and composition of four taxonomic groups, plants, butterflies, true bugs and spiders and it tries to solve this problem partially. We want to bring some pieces of evidence, that the impact of landscape structure is very important for the species richness and also to bring some possible solution to this very difficult and highly discussed theme.
Material and methods Study area The study was carried out in Bochov, Karlovy Vary district, northwestern Czech Republic (50°10’ N, 13°01’ E, 700 m a.s.l.). The wider region, southern foothills of the volcanic Doupovské vrchy? Mts, is characterised by 348 particularly low intensity of land use. It adjoins the Hradiště military training range, but even the land outside the range is sparsely populated. The entire region is a mosaic of both traditionally managed and improved meadows and pastures, fields, small woodlots, fish ponds and alluvial wetlands. The climate is mildly warm and mildly damp, the mean air temperatures do not exceed 7 °C. The coldest month is January (3 °C), the warmest one is July (16.7 °C).
The bedrock is formed by gneiss with unique amphibolite lens with ascendent peaks of basalt eruptions. Slightly inclined bottoms of their walleys which form the protuberances going to the higher placed areas are formed with the uphill rocks. In the valleys along the rivers and in the erosion cuts are alluvial sediments.
The region is renown for high diversity of plants and animals of traditionally used submontane grasslands, hosting, e.g., strong populations of the critically endangered Marsh Fritillary butterfly (Hula et al., 2004).
Insect and plant diversity were surveyed in 2003 along four parallel transects, situated 300 metres apart and crossing the area in an approximate NW–SE direction. Each of the transects consisted of twenty sections, each 300 m long, thus giving a lattice of 20*4 sections (Fig. 1). The four taxonomic groups were recorded as follows.
Fig. 1. Residual variation attributable to the three measures of heterogeneity, as returned by the GLM multiple regression models. The analysis without the effect of biotopes. Black columns stand for significant nominally results (P 0.05) and white columns stand for non-significant results. Numbers from 0-500 are the distance from the transect line. “s” by the numbers means, that space was included in the model.
349 Higher plants A skilled botanists (L. Filipova) walked the transect in late June and again in mid July, recording all species of higher plants growing in approximate 5 metres strip along the transect path (Appendix 1). It took twenty persondays to complete the survey. Nomenclature follows Kubát (2002).
Butterflies and burnets (Lepidoptera: Papilionoidea, Hesperioidea and Zygaenidae – Recorded by M. Konvicka and Z. Fric.
Sixteen transect walks (Pollard, 1977) were carried out along the lattice, in approximately 10-days intervals, between 6th May and 20th August. All butterflies and burnets were identified to species, usually by sight, more difficult species were net-captured for identification. The walks were limited 9:30 a.m. and 4:60 p.m. (C European summer time) and to weather appropriate for counting butterflies; they were interrupted if the weather worsened.
Under fine weather, it took two person-days to complete one lattice.
Terrestrial bugs (Heteroptera) We sampled the bugs in early July by sweping herbaceous and (where appropriate) shrubby vegetation along the transects. Fifty sweps, distributed regularly along its length, were taken from each section, and care was paid to include all distinct vegetation types in proportions relatively equal to their representations along the sections. The bugs were killed en masse and subsequently sorted and identified in a laboratory. The sweping took ten persondays, the sorting/identifying ca 50 person-days. Nomenclature follows Aukema and Rieger (2004).
Spiders Spiders originated from the sweeping described above. They were identified by a specialist (V. Hula), the nomenclature follows Buchar and Růžička (2002).
Landscape variables Biotopes along the transect were distinguished in field and subsequently measured using aerial 1: 1000 photographs. We distinguished: extensively managed meadow (unfertilised, mown at most once a year), intensively managed meadow (fertilised and mown twice a year), steppe grassland, ponds, intravillan, scrub, pastures, ruderal grasslands and forests. The program ArcView3.x was used for the measurements.
Biotope heterogeneity was analysed at three distance from the transects (i) „zero“, i.e., the right at the transect;
(ii) 100 m, i.e. the heterogeity in 200*300 m rectangles dissected by individual transect sections with the width of 100 m; (iii) 500 meters, i.e. the heterogeity in 1000*300 m rectangles dissected by individual transect sections with the width of 500 m.
Out of numerous possible ways to describe biotope heterogeneity, we used three simple measures. (a) Numbers of segments of distinct biotope categories within the given rectangle (or at the transect line in case (i)), herein number. (b) Simpson’s diversity of distinct biotope categories, computed as D = 1-Σ (n/N)2, where n is the area of one biotope in the rectangle and N is the area of the whole rectangle, herein diversity. (c) Summed length of borders of the above biotope segments, herein borders.
Because lengths of individual sections, despite much care, sometimes deviated from 300 m, we used lengths of sections or (if appropriate) areas of the analysed rectangles as covariables in all analyses.