Journal of Paleolimnology 18: 395���420, 1997. 395 c 1997 Kluwer Academic Publishers. Printed in Belgium. Modern diatom, cladocera, chironomid, and chrysophyte cyst assemblages as quantitative indicators for the reconstruction of past environmental conditions in the Alps. I. Climate Andre�� F. Lotter1, H. John B. Birks2, Wolfgang Hofmann3 & Aldo Marchetto4 1Geobotanical Institute, University of Bern, Altenbergrain 21, CH-3013 Bern, Switzerland, and Swiss Federal Institute of Environmental Science and Technology (EAWAG), CH-8600 Dubendorf, �� Switzerland (e-mail: lotter@sgi.unibe.ch) 2Botanical Institute, University of Bergen, Allegaten �� 41, N-5007 Bergen, Norway, and Environmental Change Research Centre, University College London, 26 Bedford Way, London, WC1H 0AP, UK (e-mail: john.birks@bot.uib.no) 3Max-Planck-Institut f�� ur Limnologie, August-Thienemann-Strasse 2, D-24302 Pl�� on, Germany (e-mail: hofmann@mpil-ploen.mpg.d400.de) 4Consiglio Nazionale delle Ricerche, Istituto Italiano di Idrobiologia, Largo Vittorio Tonolli, 50-52, I-28048 Verbania Pallanza, Italy (e-mail: marchett@iii.to.cnr.it) Received 11 October 1996 accepted 15 March 1997 Key words: transfer functions, weighted-averaging partial-least-squares, summer temperatures, surface sediments, modern training-sets, Switzerland Abstract Diatom, chrysophyte cyst, benthic cladocera, planktonic cladocera, and chironomid assemblages were studied in the surface sediments of 68 small lakes along an altitudinal gradient from 300 to 2350 m in Switzerland. In addition, 43 environmental variables relating to the physical limnology, geography, catchment characteristics, climate, and water chemistry were recorded or measured for each lake. The explanatory power of each of these predictor variables for the different biological data-sets was estimated by a series of canonical correspondence analyses (CCA) and the statistical significance of each model was assessed by Monte Carlo permutation tests. A minimal set of environmental variables was found for each biological data-set by a forward-selection procedure within CCA. The unique, independent explanatory power of each set of environmental variables was estimated by a series of CCAs and partial CCAs. Inference models or transfer functions for mean summer (June, July, August) air temperature were developed for each biological data-set using weighted-averaging partial least squares or partial least squares. The final transfer functions, after data screening, have root mean squared errors of prediction, as assessed by leave-one-out cross-validation, of 1.37 C (chironomids), 1.60 C (benthic cladocera), 1.62 C (diatoms), 1.77 C (planktonic cladocera), and 2.23 C (chrysophyte cysts). Introduction Sediments are among the few continuous proxy archives that provide information about past environ- ments over time periods of single years to millennia. Palaeoecological and palaeolimnological studies of lacustrine deposits can provide the long time-series that are needed not only to reconstruct past environmental conditions but also to assess the natural variability of biotic and abiotic systems. Moreover, they can yield information about the reaction of these systems to dif- ferent perturbations and, given a good time-control, it is possible to estimate phases and amplitudes of distur- bances. Despite this potential, palaeoecological stud- ies have often been descriptive and narrative (Birks, 1992, 1993). Processes driving observed patterns in the proxy records have to be inferred. Changes in past environmental conditions are often only described in Article: jopl425 GSB: Pips nr 137603 BIO2KAP *137603 jopl425.tex 13/11/1997 18:22 v.7 p.1

396 Figure 1. Map of Switzerland showing the location of the 68 sampled lakes. Numbers refer to lakes in Table 1. a qualitative way. To test hypotheses concerning past environmental changes and also to evaluate biological and climate models, it is necessary to quantify palaeo- ecological proxy data (Birks, 1995). Early attempts to quantify climatic variables such as temperature from palaeoecological data were car- ried out by Iversen (1944). However, it was only in the 1970s that mathematical methods began to be applied to the reconstruction of palaeoenvironments in a rigorous and quantitative way (e.g. Imbrie & Kipp, 1971). Diatoms have played an important role in the development of quantitative methods for environmen- tal reconstruction: early work concentrated on the rela- tionship between diatom assemblages and lake-water pH using linear regression techniques (e.g. Renberg & Hellberg, 1982). In the late 1980s much effort was put into the development of numerically robust and ecologically realistic mathematical methods for envi- ronmental reconstruction as well as for reliable error estimation (e.g. ter Braak & Barendregt, 1986 ter Braak, 1987 ter Braak & Looman, 1986, 1987 ter Braak & van Dam, 1989 Birks et al., 1990). Most of the required protocols, inferential techniques, and quality control guidelines were developed in connec- tion with lake acidification studies (e.g. Munro et al., 1990 Charles, 1990 Birks et al., 1990 Birks, 1995) and can now be directly transferred to other studies of past environmental change. In recent years, sev- eral studies have used aquatic organisms to estimate palaeo-temperatures (e.g. Walker et al., 1991a, 1997 Levesque et al., 1993, 1994 Cwynar & Levesque, 1995). High-latitude and high-altitude sites have recently become a focal point for research with respect to the expected future warming of the earth���s climate. A bet- ter understanding of the reaction of these ecosystems to environmental change in the past (e.g. Fritz, 1996) will greatly enhance the ability to predict future envi- ronmental change. In mountainous regions, however, such as the Alps of central Europe with their complex topography and climate, quantitative reconstructions, especially climatic reconstructions, need a much more refined scale of investigation than is currently available at a broad continental scale (e.g. Huntley & Prentice, 1988, 1993). jopl425.tex 13/11/1997 18:22 v.7 p.2

397 This study was undertaken to assess the potential of using different aquatic organisms, such as diatoms, cladocera, chironomids, and chrysophytes, as environ- mental indicators for quantitative palaeoenvironmen- tal reconstructions in Late-Glacial and Holocene time- series in the geographical region of Central Europe with particular reference to the Alps. In this paper, we focus on the different modern training-sets, and their quantitative relationship to present-day climate in the Alps. In addition, we outline the statistical bases for the numerical regression and calibration methods used. In a second, separate contribution we concentrate on the relationship of these data-sets to trophic state (Lotter et al., in press). Sites, environmental variables and microfossils studied We sampled the surficial sediments of 68 small lakes (Figure 1) of similar size between 1993 and 1994. These lakes are situated along an altitudinal gradient from 300 to 2350 m above sea level (a.s.l.), which is also a major climatic gradient (Table 1). To minimize the effects of low pH that might override effects of other important environmental variables on the aquatic organisms, only lakes in calcareous bedrock regions were chosen. At each lake several echo-soundingtracks were car- ried out to locate the deepest part of the basin and water chemistry was determined (Table 1 and Muller �� et al., in press). Four short sediment cores were taken with a modified Kajak corer (6.2 cm diameter) in the deep- est part of each lake. The top 5 cm of two of the cores were extruded in 1-cm increments in the field, the third core was cut open longitudinally in the lab, photographed, and checked for stratigraphical consis- tency of the uppermost sediment, and the remaining core has been archived in a cold room. For this study the topmost one centimetre was used for analysis. Geographical lake and catchment data are given in Table 1. Lakes and catchments were digitized from 1:25 000 topographic maps. On the basis of different GIS maps of Switzerland with a spatial resolution of 100 100 m the area covered by glaciers, loose rocks, carbonaceous, siliceous, and mixed bedrock was esti- mated for each catchment. Land-use for each catch- ment was subdivided into areas covered by agricultur- al land, urban areas, wooded areas, unwooded green areas, and bare ground. Climatic variables estimat- ed for each lake were the number of growing degree days ( 5 C), mean annual as well as mean monthly values for temperature and precipitation. Mean sea- sonal temperatures (winter: December, January, Feb- ruary spring: March, April, May summer: June, July, August autumn: September, October, November) as well as winter and summer precipitation were also cal- culated (Table 2). All organisms have been analysed by one analyst only, thus providing data sets with a consistent taxon- omy and nomenclature (Birks, 1995). Samples for diatom and chrysophyte cyst analysis comprised ca 0.5 cm3 and were treated with hot 30% H2O2 and 10% HCl before mounting on slides with Naphrax. For each slide, a minimum of 500 diatom valves was counted at a magnification of 1250 , using a Leitz DM microscope with phase contrast.For diatom identification, the floras of Krammer & Lange-Bertalot (1986���1991) were used. The taxonomy of the centric diatoms, especially the Cyclotella species, largely fol- lows Wunsam et al. (1995). The chrysophyte cyst to diatom ratio (C/D) was determined by counting the number of cysts occurring together with the first 1000 diatom valves. Samples having a C/D ratio higher than 0.02 were further exam- ined. A minimum of 125 cysts or at least 1000 fields of view, whichever came first, were analysed. Cyst taxonomy follows Duff et al. (1995). Sample sizes for zoological microfossil analysis ranged from 0.6 to 27.2 g dry weight (mean = 6.8 g). The sediment was treated with 10% KOH on a mag- netic stirrer. For cladoceran analysis, the remains of Bosminidae and Chydoridae were examined in the sep- arate fractions 100 m and 55���100 m. Subsamples equivalent to 0.009���3.4 g dry weight (mean = 0.3 g) were counted under a microscope at 100 magnifi- cation. As different fragments of the animals such as shells, head shields, and post-abdomens are general- ly well-preserved, only the most abundant component, which in most cases is the shell, is taken into account (Frey, 1986). Densities were calculated as numbers per gram dry weight. If chydorid numbers per sample were 100, additional non-quantitative samples were counted. For identification, the keys by Frey (1958, 1959), Flossner �� (1972), and Lieder (1983) were used. Nomenclature follows Flossner �� (1972). Identification of Bosmina (Eubosmina) remained unclear in several cases. Bosmina (Eubosmina) morphs with a very long mucro and a very short mucro are called Bosmina sp. A and Bosmina sp. B, respectively. In the chydorids, a quantitative separation of the small Alona species is difficult. As identification is mainly based on the pores jopl425.tex 13/11/1997 18:22 v.7 p.3

398 Table 1 . Major variables characterizing the climate, limnology, and catchment of the investigated lakes. The water chemistry refers to spring circulation concentrations Lakes Geography Catchment Climate Land-use Geology Biology Number Name Abbre- Eleva- Lati- Longi- MaximumOpen InflowingCatchment Annual Winter Spring Summer Autumn Degree Annual Bare Agri- UrbanWoodedGreen GlaciersLooseCarbonateSiliceousMixed Dia-Chryso-Clado-Chiro- in Fig. 1 viation tion tude tude depth water area streams area tempera- tempera- tempera- tempera- tempera- days precipita- land culture[%] [%] unwooded [%] rock bedrock bedrock bedrock tom phyte cera nomid [m asl] [N] [E] [m] [km 2 ] [km 2 ] ture [ C]ture [ C]ture [ C]ture [ C]ture [ C] tion[mm] [%] [%] [%] [%] [%] [%] [%] taxa taxa taxa taxa 1 Lago di Muzzano MUZ 337 45 59 0 53 00 8 55 0 42 00 3.2 0.22 0 2.20 11.3 2.9 10.7 20.0 11.8 22948 1785 0.0 29.1 50.5 16.5 3.9 0 24 5 66 5 48 15 12 16 2 Lago d���Origlio ORI 416 46 03 0 07 00 8 56 0 38 00 6.0 0.07 1 1.19 11.1 2.8 10.4 19.6 11.4 22041 1817 1.0 5.1 14.3 78.6 1.0 0 17 0 83 0 48 18 10 8 3 Burg�� aschisee BUR 465 47 10 0 10 00 7 40 0 09 00 31.0 0.19 2 4.29 8.9 0.7 8.3 17.1 9.3 16721 1049 1.0 61.3 9.6 25.6 2.5 0 96 4 0 0 40 18 20 25 4 Moossee MOO 521 47 01 0 25 00 7 28 0 42 00 21.0 0.31 1 10.41 8.7 0.4 8.2 17.0 9.1 16410 1048 1.5 66.3 9.0 21.9 1.3 0 60 40 0 0 45 11 18 12 5 Le Loclat LOC 432 47 01 0 13 00 6 59 0 52 00 9.2 0.05 0 0.88 9.1 1.0 8.5 17.3 9.5 17226 892 0.0 60.8 2.9 33.3 2.9 0 29 71 0 0 34 14 13 27 6 Lac de Seedorf SEE 609 46 47 0 47 00 7 02 0 29 00 7.5 0.10 3 7.38 8.2 0.0 7.5 16.4 8.6 15160 1031 0.3 77.1 7.0 14.6 1.0 0 95 5 0 0 24 n.a. 6 17 7 Gerzensee GER 603 46 49 0 55 00 7 32 0 53 00 10.0 0.27 2 2.70 8.3 0.1 7.8 16.6 8.8 15578 1044 0.0 78.2 14.9 5.0 2.0 0 60 15 25 0 32 n.a. 17 26 8 Uebeschisee UEB 641 46 44 0 07 00 7 34 0 00 00 14.5 0.15 2 2.16 8.2 0.0 7.7 16.3 8.6 15139 1187 6.3 80.0 5.3 6.1 2.3 0 97 3 0 0 37 18 21 22 9 L�� utzelsee L �� UT 500 47 15 0 38 00 8 46 0 22 00 6.0 0.13 4 6.02 8.4 0.2 7.9 16.6 8.9 15721 1352 7.1 67.0 7.1 10.2 8.6 0 25 40 35 0 41 n.a. 19 8 10 Wilersee WIL 730 47 10 0 20 00 8 37 0 13 00 21.0 0.03 1 0.50 7.3 0.7 6.6 15.4 7.9 13274 1612 5.1 89.8 1.0 4.1 0.0 0 100 0 0 0 35 n.a. 16 24 11 H�� uttwilersee H �� UT 434 47 36 0 39 00 8 50 0 42 00 15.0 0.35 2 3.71 8.6 0.3 8.3 17.0 8.9 16351 922 3.1 68.2 3.1 25.4 0.3 0 96 4 0 0 35 14 20 18 12 Husemersee HUS 409 47 37 0 22 00 8 42 0 20 00 14.0 0.08 1 1.33 8.7 0.4 8.4 17.0 8.9 16537 903 5.1 32.7 4.1 54.1 4.1 0 100 0 0 0 36 16 19 16 13 Mettmenhasler See MET 418 47 28 0 31 00 8 29 0 35 00 12.5 0.03 0 0.50 8.8 0.6 8.5 17.1 9.1 16684 1043 4.0 44.6 25.7 19.8 5.9 0 72 28 0 0 46 n.a. 17 26 14 Unterer ChatzenseeCHA 439 47 25 0 58 00 8 29 0 26 00 7.8 0.19 1 1.29 8.8 0.6 8.4 17.0 9.1 16495 1037 12.0 25.0 10.0 45.0 8.0 0 100 0 0 0 51 n.a. 22 31 15 Soppensee SOP 596 47 05 0 31 00 8 04 0 56 00 26.5 0.23 0 1.59 8.2 0.1 7.5 16.3 8.7 15172 1266 2.0 80.8 5.1 12.1 0.0 0 94 6 0 0 40 11 22 9 16 Burgseeli BUG 613 46 41 0 55 00 7 53 0 11 00 19.0 0.09 1 1.18 8.3 0.2 7.9 16.3 8.9 15431 1193 4.1 31.6 8.2 51.0 5.1 0 25 75 0 0 28 30 15 22 17 Blausee BLA 887 46 32 0 01 00 7 39 0 55 00 10.0 0.01 0 0.09 6.9 0.9 6.2 14.9 7.6 12446 1213 0.0 0.0 1.0 89.0 10.0 0 100 0 0 0 36 n.a. 8 19 18 Schwarzsee SCH 1046 46 40 0 10 00 7 17 0 06 00 9.5 0.46 10 19.70 6.0 1.8 5.1 14.0 6.7 10671 1620 7.1 53.2 2.2 27.0 10.5 0 15 66 0 19 58 29 14 8 19 Rotsee ROT 419 47 04 0 18 00 8 19 0 01 00 16.0 0.50 2 4.60 9.1 0.9 8.6 17.3 9.5 17190 1179 0.5 36.1 35.1 18.5 9.8 0 10 90 0 0 49 6 19 22 20 Mauensee MAU 504 47 10 0 20 00 8 04 0 35 00 9.0 0.60 2 4.30 8.7 0.5 8.1 16.9 9.1 16258 1110 2.0 76.5 8.2 12.2 1.0 0 85 15 0 0 38 16 21 20 21 Seelisberg Seeli SEL 738 46 57 0 32 00 8 34 0 16 00 37.5 0.18 1 2.70 7.3 0.6 6.6 15.3 8.0 13301 1591 4.1 46.3 3.6 46.0 0.0 0 0 100 0 0 18 15 20 20 22 Lac Brenet BRE 1002 46 40 0 29 00 6 19 0 31 00 17.0 0.63 0 2.85 6.3 1.1 5.0 14.0 7.1 10710 1515 2.2 31.9 7.2 55.9 2.7 0 30 70 0 0 66 n.a. 20 24 23 Lac des Taill` eres TAI 1036 46 57 0 57 00 6 34 0 07 00 8.5 0.44 0 33.16 5.5 2.1 4.4 13.4 6.4 9691 1603 0.2 44.6 2.0 52.7 0.5 0 4 96 0 0 51 27 11 19 24 Schwendisee SCW 1159 47 11 0 19 00 9 19 0 55 00 9.5 0.04 2 5.06 5.1 2.6 4.2 12.7 6.2 8865 1793 7.3 48.5 0.5 37.4 6.3 0 13 84 0 3 42 16 13 27 25 Voralpsee VOR 1123 47 09 0 30 00 9 22 0 43 00 3.3 0.15 5 13.52 5.3 2.5 4.4 12.9 6.3 9184 1681 12.2 49.8 1.0 28.1 9.0 0 16 84 0 0 41 n.a. 4 7 26 Tschingelsee TSC 1150 46 33 0 10 00 7 44 0 37 00 1.6 0.11 2 36.65 5.6 2.0 4.6 13.4 6.4 9798 1304 57.0 25.6 1.2 10.1 6.1 8 13 55 0 24 55 n.a. 1 6 27 Lac des Chavonnes CHV 1690 46 20 0 03 00 7 05 0 09 00 29.5 0.05 1 0.74 3.7 3.4 2.1 11.1 4.8 6179 1665 1.0 26.5 1.0 67.3 4.1 0 0 22 0 78 53 14 9 6 28 Lac Retaud RET 1685 46 21 0 41 00 7 12 0 00 00 4.5 0.01 0 0.22 3.5 3.7 2.1 11.0 4.7 6057 1762 8.9 44.6 5.0 9.9 31.7 0 59 0 0 41 46 14 18 11 29 F�� alensee F �� AL 1452 47 15 0 08 00 9 24 0 59 00 31.0 0.15 0 4.25 3.5 3.9 2.2 10.9 4.7 5887 2018 28.9 44.8 0.0 5.5 20.9 0 15 85 0 0 13 12 7 10 30 Seealpsee SAL 1141 47 16 0 10 00 9 23 0 56 00 15.0 0.14 1 11.33 5.3 2.4 4.3 12.9 6.3 9044 1874 29.3 51.9 0.5 3.5 14.9 0 18 82 0 0 40 n.a. 8 7 31 Grosssee GRO 1618 47 04 0 48 00 9 14 0 53 00 11.5 0.05 2 2.20 2.9 4.5 1.6 10.3 4.2 5142 1824 3.0 79.4 1.0 14.5 2.0 0 0 43 57 0 61 n.a. 8 2 32 Engstlensee ENG 1850 46 46 0 27 00 8 21 0 32 00 49.0 0.45 2 7.40 1.9 5.1 0.2 9.0 3.3 3470 1793 58.5 29.3 0.3 4.0 8.0 11 10 54 0 25 31 n.a. 9 21 33 Seebergsee SEB 1831 46 34 0 41 00 7 26 0 38 00 15.5 0.06 0 0.23 2.2 4.9 0.6 9.7 3.5 4163 1683 15.2 42.3 0.0 15.2 27.3 0 8 46 0 46 48 13 7 13 34 Tannensee TAN 1976 46 46 0 30 00 8 18 0 22 00 16.0 0.34 10 1.12 1.3 5.6 0.5 8.4 2.7 26971861 17.8 70.3 2.0 0.0 9.9 0 0 100 0 0 92 20 9 18 35 Melchsee MEL 1891 46 46 0 19 00 8 16 0 07 00 15.5 0.49 7 5.92 1.7 5.2 0.1 8.9 3.2 3279 1843 19.2 71.6 1.2 2.0 6.0 0 26 74 0 0 45 n.a. 7 11 36 Sewenseeli SEW 1689 46 53 0 21 00 8 05 0 15 00 4.5 0.03 2 0.21 3.3 4.0 1.8 10.7 4.5 5671 1744 0.0 23.8 0.0 70.3 5.9 0 65 0 0 35 47 15 9 11 37 Lac de Bret BRT 674 46 30 0 53 00 6 46 0 26 00 18.0 0.50 1 2.97 8.2 0.3 7.5 16.3 8.8 15131 1278 0.3 83.3 5.1 10.3 1.0 0 89 11 0 0 43 n.a. 16 11 38 Lac Tanay TAY 1408 46 20 0 47 00 6 50 0 33 00 31.0 0.18 1 7.93 5.0 2.2 3.5 12.5 6.0 8265 1500 21.5 35.4 0.8 25.3 17.2 0 1 99 0 0 36 n.a. 9 14 39 Lac de Nervaux NER 1493 46 22 0 44 00 6 59 0 07 00 10.0 0.01 1 0.92 4.3 2.9 2.9 11.9 5.4 7287 1609 8.0 44.0 0.0 30.0 18.0 0 0 95 0 5 35 16 10 17 40 S�� agistalsee S �� AG 1935 46 40 0 51 00 7 58 0 39 00 9.7 0.07 3 3.85 1.7 5.3 0.1 8.9 3.0 3290 1663 39.5 41.5 0.0 2.2 16.8 0 0 100 0 0 48 18 5 13 41 Wannisbordsee WAN 2103 46 40 0 58 00 8 17 0 58 00 14.0 0.02 1 1.64 0.5 6.2 1.4 7.6 2.1 1876 1970 84.2 3.0 0.0 0.0 12.9 0 1 0 99 0 59 n.a. 5 16 42 Bannalpsee BAN 1587 46 52 0 12 00 8 25 0 42 00 17.0 0.16 2 8.23 3.3 3.9 1.9 10.7 4.5 5659 1674 58.4 32.2 0.5 2.0 7.0 2 29 69 0 0 49 15 6 14 43 Iffigsee IFF 2065 46 23 0 16 00 7 24 0 33 00 30.0 0.10 1 4.61 1.5 5.5 0.3 8.8 2.9 3152 1666 73.1 18.0 0.0 0.0 9.0 13 18 43 0 26 35 21 6 11 44 Flueseeli FLU 2045 46 24 0 38 00 7 29 0 59 00 8.5 0.04 1 0.79 1.5 5.5 0.3 8.9 2.9 3181 1487 92.0 8.0 0.0 0.0 0.0 0 0 100 0 0 47 n.a. 2 9 45 L�� ammerensee L �� AM 2296 46 24 0 08 00 7 35 0 13 00 2.5 0.07 1 1.55 0.4 6.4 1.6 7.6 1.9 1878 1530 53.0 30.0 0.0 0.0 17.0 0 12 88 0 0 35 n.a. 4 4 46 Tr�� uebsee TR �� U 1764 46 47 0 32 00 8 23 0 38 00 7.0 0.26 1 7.07 2.3 4.7 0.8 9.6 3.7 4111 1749 60.8 31.2 0.3 4.8 3.0 19 7 48 0 26 44 n.a. 6 14 47 Bachsee BAC 2265 46 40 0 12 00 8 01 0 24 00 16.0 0.07 4 1.87 0.0 6.7 1.9 7.0 1.6 1394 1782 45.5 45.5 0.0 0.0 8.9 0 0 99 0 1 55 n.a. 3 9 48 Lutersee LUT 1702 46 50 0 12 00 8 21 0 05 00 4.5 0.02 0 0.59 2.6 4.5 1.0 9.9 3.9 4490 1735 25.0 56.0 0.0 5.0 14.0 0 0 100 0 0 53 n.a. 6 11 49 Hagelseewli HAG 2339 46 40 0 26 00 8 02 0 11 00 18.5 0.03 0 0.36 0.4 7.0 2.4 6.6 1.3 1046 1821 55.4 41.6 0.0 0.0 3.0 0 0 100 0 0 40 11 2 6 jopl425.tex 13/11/1997 18:22 v.7 p.4

399 Table 1 . Continued Lakes Geography Catchment Climate Land-use Geology Biology Number Name Abbre- Eleva- Lati- Longi- MaximumOpen InflowingCatchment Annual Winter Spring Summer Autumn DegreeAnnual Bare Agri- UrbanWoodedGreen GlaciersLooseCarbonateSiliceousMixed Dia-Chryso-Clado-Chiro- in Fig. 1 viation tion tude tude depth water area streams area tempera- tempera- tempera- tempera- tempera- days precipita- land culture[%] [%] unwooded [%] rock bedrock bedrock bedrock tom phyte cera nomid [m asl][N] [E] [m] [km 2 ] [km 2 ] ture [ C]ture [ C]ture [ C]ture [ C]ture [ C] [mm] [%] [%] [%] [%] [%] [%] [%] taxa taxa taxa taxa 50 Lag Grond GRD 1016 46 48 0 32 00 9 15 0 29 00 5.0 0.02 1 1.89 6.3 1.4 5.5 13.9 7.2 10999 1200 1.0 49.0 22.4 25.5 2.0 0 92 0 8 0 52 n.a. 15 18 51 Schwellisee SCE 1933 46 45 0 56 00 9 38 0 55 00 12.0 0.03 4 9.58 1.8 5.0 0.1 8.9 3.3 3267 1162 58.3 25.7 0.0 0.2 15.8 0 57 29 14 0 67 n.a. 2 15 52 Obersee OBE 1734 46 47 0 07 00 9 40 0 53 00 14.5 0.08 1 2.71 2.7 4.4 1.2 9.9 4.0 4595 1140 2.3 62.8 6.3 25.6 3.0 0 75 18 7 0 38 11 10 18 53 T�� urlersee T �� UR 643 47 16 0 17 00 8 30 0 02 00 21.0 0.50 5 5.19 7.9 0.2 7.3 16.0 8.4 14495 1339 0.2 58.0 15.0 25.0 1.8 0 57 43 0 0 35 10 10 20 54 Seewli See SWL 2028 46 48 0 45 00 8 43 0 02 00 16.0 0.08 1 2.70 0.8 6.0 1.0 7.9 2.4 2224 1729 57.0 27.0 0.0 1.0 15.0 0 61 33 0 6 42 n.a. 3 11 55 Bichelsee BIC 590 47 27 0 32 00 8 54 0 04 00 6.5 0.09 5 2.70 6.8 1.1 6.1 14.9 7.3 12317 1437 0.0 28.2 3.2 68.5 0.0 0 1 99 0 0 28 12 14 20 56 Egelsee EGE 667 47 24 0 08 00 8 21 0 39 00 10.0 0.02 0 0.29 7.6 0.4 7.0 15.7 8.1 13923 1247 3.0 0.0 0.0 80.0 17.0 0 100 0 0 0 30 18 16 18 57 Dittligsee DIT 652 46 45 0 25 00 7 32 0 09 00 16.5 0.07 0 3.13 7.9 0.2 7.3 16.0 8.4 14582 1169 3.1 69.4 5.9 21.4 0.3 0 100 0 0 0 17 8 14 21 58 Inkwilersee INK 461 47 11 0 58 00 7 39 0 50 00 5.0 0.12 1 2.13 8.8 0.7 8.2 17.1 9.3 16663 1073 1.0 67.9 9.3 20.6 1.3 0 100 0 0 0 30 n.a. 15 17 59 Hasensee HAS 434 47 36 0 29 00 8 49 0 58 00 5.5 0.11 0 2.52 8.6 0.3 8.3 17.0 8.9 16336 926 0.3 75.9 5.4 17.4 1.0 0 100 0 0 0 39 21 15 20 60 Nussbaumsee NUS 434 47 36 0 29 00 8 49 0 58 00 8.2 0.25 1 5.87 8.6 0.2 8.3 16.9 8.8 16263 927 0.5 68.7 7.2 23.6 0.0 0 94 6 0 0 40 n.a. 20 19 61 Gattiker WaldweiherGAW 545 47 16 0 48 00 8 33 0 35 00 5.5 0.03 1 1.88 8.3 0.2 7.8 16.5 8.8 15463 1265 0.0 27.3 2.0 70.7 0.0 0 96 4 0 0 53 16 14 20 62 H�� uttnersee H �� UN 658 47 11 0 05 00 8 40 0 36 00 12.0 0.17 1 2.33 7.7 0.4 7.0 15.8 8.2 14087 1574 1.0 81.6 9.2 6.1 2.0 0 100 0 0 0 39 n.a. 18 12 63 Lac des Rousses ROU 1058 46 30 0 12 00 6 05 0 11 00 11.5 0.89 4 15.68 5.9 1.6 4.6 13.4 6.8 10314 1976 0.2 39.8 10.0 49.8 0.2 0 25 75 0 0 60 n.a. 20 19 64 Lac de l���Abbaye ABB 871 46 31 0 48 00 5 54 0 42 00 18.0 0.80 1 25.93 6.6 0.9 5.3 14.1 7.5 10800 1884 0.0 50.0 10.0 40.0 0.0 0 25 75 0 0 41 19 16 18 65 Lago di Montorfano MON 394 45 46 0 59 00 9 08 0 23 00 6.5 0.52 0 1.57 13.2 4.2 12.0 21.8 14.1 24395 1592 1.0 49.0 14.7 34.3 1.0 0 55 35 0 10 48 24 15 19 66 Lago del Segrino SEG 374 45 49 0 51 00 9 16 0 02 00 8.5 0.34 1 2.66 11.7 2.9 10.4 20.1 13.0 23900 1586 5.1 25.3 8.1 60.6 1.0 0 25 70 0 5 41 n.a. 21 27 67 Lago di Endine END 334 45 46 0 55 00 9 56 0 50 00 8.0 0.49 3 8.34 12.2 4.0 10.2 20.6 13.8 24000 1425 0.0 39.6 9.9 49.5 1.0 0 20 20 50 10 37 15 16 17 68 Lac Lioson LIO 1848 46 23 0 14 00 7 07 0 46 00 26.0 0.07 0 1.50 2.6 4.4 1.0 10.0 3.9 4659 1777 28.0 30.0 1.0 14.0 27.0 0 13 0 0 87 52 n.a. 5 18 Maximum 2339 49.0 0.89 36.65 13.2 4.2 12.0 21.8 14.1 24395 2018 92.0 89.8 50.5 89.0 31.7 19 100 100 99 87 92 30 22 31 Mean 1094 13.8 0.20 5.07 5.9 1.8 4.9 13.7 6.8 11039 1467 16.3 46.2 5.8 25.1 6.7 1 43 43 7 7 43 16 12 16 Median 945 11.8 0.12 2.70 6.4 1.1 5.4 14.1 7.3 10900 1552 4.0 44.6 3.0 19.2 3.5 0 25 40 0 0 41 15 13 17 Minimum 334 1.6 0.01 0.09 0.4 7.0 2.4 6.6 1.3 1046 892 0.0 0.0 0.0 0.0 0.0 0 0 0 0 0 13 6 1 2 of the head shield, species such as Alona rustica are possibly underestimated. For chironomid analysis, the fractions 200 m and 100���200 m were examined separately under a stereo microscope at 20���25 magnification. The results of both analyses were then combined for the numerical treatment of the data. The larval head- capsules were picked out, dehydrated in 96% alcohol, and mounted in Euparal (Hofmann, 1986). The head capsules were identified at 200���400 magnification. Densities were calculated as numbers per gram dry weight. In 46 of the 68 lakes, numbers of specimens per sample were 50 and examination of the topmost 2 cm of sediment was therefore necessary. For iden- tification, the keys by Hofmann (1971), Wiederholm (1983), and Moller Pillot (1984) were used. Nomen- clature follows Wiederholm (1983). The Pentaneuri- ni were separated by the positions of the pores SSm, S9, S10, and VP on the submental surface (Kowa- lyk, 1985). With respect to pore arrangement, Penta- neurini sp. A, B, and C correspond to Paramerina-, Zavrelimyia-, and Telopelopia-types, respectively. The distinction of Orthocladius with six lateral mentum teeth from Cricotopus is uncertain: brownish head cap- sules in which the lateral mentum teeth were equal in size are associated with Orthocladius. Orthocladius sp. A has more than six pairs of lateral mentum teeth and refers to the subgenus Euorthocladius. Tanytarsus sp. A is characterized by a short, distally rounded spur at the antennal pedestal. Numerical analyses To determine whether to use linear- or unimodal- based numerical techniques (ter Braak & Prentice, 1988), each biological data-set was initially analysed by detrended correspondence analysis (DCA Hill & Gauch, 1980) with detrending-by-segments, non- linear rescaling, and downweighting of rare taxa to estimate the lengths of the compositional gradients in each data-set. As all the data-sets have composi- tional gradient lengths of 2 or more standard devia- tion units, all subsequent numerical analyses involved techniques that are based on an underlying uni- modal species-response model (ter Braak & Prentice, 1988 Birks, 1995), namely canonical correspondence analysis (CCA), detrended CCA, weighted-averaging (WA) regression and calibration, and WA-partial least squares regression (WA-PLS ter Braak & Juggins, 1993). jopl425.tex 13/11/1997 18:22 v.7 p.5