Acta Botanica 2-2015.indd ACTA BOT. CROAT. 74 (2), 2015 211 Acta Bot. Croat. 74 (2), 211–232, 2015 CODEN: ABCRA 25 ISSN 0365-0588 eISSN 1847-8476 DOI: 10.1515/botcro-2015-0017 Applied use of taxonomy: lessons learned from the fi rst German intercalibration exercise for benthic diatoms MIRKO DREßLER1*, GEURT VERWEIJ2, SONJA KISTENICH1, MARIA KAHLERT3, PETRA WERNER4 1 University of Rostock, Institute of Bio-Science, Department for Botany and Botanical Garden, Wismarsche Str. 8, D-18057 Rostock, Germany 2 Koeman en Bijkerk, Oosterweg 127, 9751 PE Haren, The Netherlands 3 Swedish University of Agricultural Science, Department of Aquatic Sciences and Assessment, Box 7050, Lennart Hjelms väg 9, 750 07 Uppsala, Sweden 4 Diatoms as Bioindikators, Grainauer Str. 8, 10777 Berlin, Germany Abstract – The fi rst German intercalibration exercise for benthic diatoms was conducted to improve the application of the implementation of the European Water Framework Di- rective for running waters and lakes in Germany. The intercalibration exercise revealed several taxonomic problems. Among others, considerable problems occurred with identi- fying and differentiating species of the following four groups: (I) Amphora indistincta Levkov and A. pediculus (Kützing) Grunow, (II) Cocconeis placentula var. euglypta Eh- renberg and C. placentula var. lineata (Ehrenberg) Van Heurck, (III) Navicula cryptote- nella Lange-Bertalot and N. cryptotenelloides Lange-Bertalot and (IV) N. reichardtiana Lange-Bertalot and N. caterva Hohn & Hellermann. The taxonomic problems that emerged occurred due to both insuffi cient use of given taxonomic details (by limnologists) and ambiguous species descriptions and documentation (by taxonomists). Thus, we rec- ommend to the applied limnologist to use the mandatory identifi cation literature and to document any ambiguous valves during routine counts. Also, it would be desirable to fur- ther investigate certain species by taxonomists and, in general, to provide more basic data with species descriptions or in identifi cation manuals. These measures will improve the use of diatoms as bioindicators and consequently benefi t both applied limnologists and taxonomists. Key words: Amphora, benthic diatoms, Cocconeis, intercalibration exercise, Navicula, quality control, taxonomy * Corresponding author, e-mail: mirko.dressler@gmx.de DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 212 ACTA BOT. CROAT. 74 (2), 2015 Introduction Benthic diatoms are well-established, robust bioindicators (SMOL and STOERMER 2010) that are widely used, e.g. in paleolimnology (HÜBENER et al. 2009, DREßLER et al. 2011) or when implementing the European Water Framework Directive (EU-WFD 2000) (PRYGIEL 2002, WERNER and DRESSLER 2007, KELLY 2013). In various European countries intercali- bration exercises and taxonomic workshops are conducted: – to ensure the comparability of diatom counting results among diatomists when imple- menting the Water Framework Directive, – to facilitate a uniform approach for dealing with taxonomic problems, and – to perform the basic quality control for diatom counting results (KELLY 2001, PRYGIEL et al. 2002, KAHLERT et al. 2012). In Germany, the fi rst intercalibration exercise for benthic diatoms was conducted in 2011 and 2012 to improve the application of the German instruction protocol that assesses the water quality of lakes and rivers using diatoms according to the European Water Frame- work Directive (SCHAUMBURG et al. 2006, 2007, 2011). The 37 participants of this German intercalibration exercise came from Germany, Belgium, the Czech Republic, France, Ire- land, Italy, the Netherlands, Slovakia, Spain and Sweden and counted benthic diatom sam- ples from two rivers and two lakes according to the German method (SCHAUMBURG et al. 2006, 2007, 2011). The counting results of three auditors of these four samples were used as references. The auditor results were a statistically reliable reference for three of the four samples and provided the diatom assemblage to which the participant results were compared to and assessed with. Nine of the 37 participant results deviated signifi cantly to the results of all three auditors in at least one sample. These signifi cant differences identifi ed taxonomic problems within twelve genera. The aim of this study was a detailed description and analysis of four identifi ed taxo- nomic problems to recommend measures that may improve the applied use of diatoms as bioindicators. These four problems apparent from the fi rst German intercalibration exercise were the identifi cation and differentiation of (I) some small Amphora-taxa (mainly A. indistincta Levkov and A. pediculus (Kützing) Grunow); (II) the varieties of Cocconeis placentula Ehrenberg (mainly C. placentula var. euglypta Ehrenberg and C. placentula var. lineata (Ehrenberg) Van Heurck); (III) Navicula cryptotenella Lange-Bertalot and N. cryptotenelloides Lange-Bertalot and (IV) Navicula reichardtiana Lange-Bertalot and Navicula caterva Hohn & Hellermann. Material and methods The fi rst German intercalibration exercise was based on four samples of benthic diatoms: 1. Lake Krossinsee, Northern Germany, lowland lake, calcareous, polymictic 2. Lake Geneva, Switzerland, Alps/Alpine foothills lake, calcareous, dimictic 3. River Klepelshagener Bach, Northern Germany, lowland river, calcareous 4. River Drau, Austria, Alps/Alpine foothills river, siliceous BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 213 The taxonomic problems discussed here occurred only in sample Lake Krossinsee (Northern Germany) and in sample Lake Geneva (Switzerland). Thus, we will only address the two lake samples in the following. Lake Krossinsee (52°36'31.5"N; 13°69'46.5"E) was sampled on 12th July 2010 from various stones and various stalks (below water surface) of Phragmites australis (Cavanilles) Trinius ex Steudel and Typha latifolia L. Lake Geneva (46°22'31.3"N; 6°15'18.0"E) was sampled on 7th August 2011 from various stones. The se- lected and here presented taxonomic problems concern the differentiation and identifi cation of the following taxa: 1. Amphora indistincta and A. pediculus 2. Cocconeis placentula var. euglypta and var. lineata 3. Navicula cryptotenella and N. cryptotenelloides 4. Navicula reichardtiana and N. caterva Diatom samples were taken according to the German instructions for implementing the European Water Framework Directive (SCHAUMBURG et al. 2006, 2007, 2011). Only natu- rally prevalent substrates were sampled (stones and macrophytes in Lake Krossinsee and stones in Lake Geneva) in at least 20 cm water depth. Periphyton was scratched from at least 15 stones and plants per site into a 500 mL bottle and alcohol was added. Diatom samples were oxidised and prepared with HCl, H2O2, H2SO4, KMnO4 and C2H2O4 follow- ing modifi ed KALBE and WERNER (1974). Dried slurries containing the diatoms were then mounted with Naphrax® (refraction index 1.73) onto slides. Each slide from the two sam- ples was prepared by one person and then sent to the participants. The slides were labelled with the diatom water body type, i.e. participants knew the region and lake-type. However, participants did not know which lake exactly they were counting during the intercalibration exercise. The slides for the auditors were taken from the same batch of prepared slides as for the participants. The auditors are diatom-specialists with more than 20 years of experience of analysing diatom slides. All samples were counted three times on three different slides by the follow- ing auditors: (1) Dr. Gabriele Hofmann: three lake samples (one sample twice, i.e. two slides from one sample) and two running water samples, (2) Dr. Thomas Hübener: three lake samples (one sample twice) and two running water samples, (3) Dr. Peter Pfi ster: two running water samples. Thus, the here discussed lake samples were counted by two audi- tors and the running water samples (which are not part of this paper) by three auditors. Ac- cordingly, each lake sample was counted twice by one auditor using two prepared slides to enable an assessment of the variability among slides per sample. Participants and auditors were instructed to base their counts on SCHAUMBURG et al. (2006, 2007) (both in English) and additionally the new and relevant changes of instruc- tions given in SCHAUMBURG et al. (2011), which were translated and send to the participants in the letter accompanying the samples. The standard identifi cation literature was HOFMANN et al. (2011). Additionally, identifi cation had to be based on the supplementary books KRAMMER and LANGE-BERTALOT (1986–91, 2004), LANGE-BERTALOT (1993, 2001), LANGE- BERTALOT and MOSER (1994), LANGE-BERTALOT and METZELTIN (1996), KRAMMER (1997a, 1997b, 2000, 2002, 2003), REICHARDT (1999), WITKOWSKI et al. (2000) and LEVKOV (2009). Participants and auditors entered the diatom counting results via the EQAT-webpage (www.planktonforum.eu) into an entry mask using the laboratory code that was given to each person with the slides. The analyses and presentation of the results are solely based on the laboratory codes. DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 214 ACTA BOT. CROAT. 74 (2), 2015 Diatom photos were taken with a ProgRes® SpeedXTcore3 (Jenoptik) camera attached to an Axioplan light microscope (Zeiss) (differential interference contrast (DIC), 100x oil- immersion objective Plan-Apochromat, aperture 1.4) at an overall magnifi cation of 1000x. Valves were measured using the software analySIS® (Soft Imaging System GmbH). Results Small Amphora-species in Lake Krossinsee The most abundant Amphora species in the sample from Lake Krossinsee were A. indis- tincta (Pl. 1: 1–4) and A. pediculus (Pl. 1: 8–12; Fig. 1). Additionally, valves of A. cf. indis- tincta (Pl. 1: 5–7), A. cf. pediculus (Pl. 1: 13–15) and Amphora spec. (Pl. 1: 16–17) oc- curred in low abundances in this sample. Rare Amphora-taxa (< 0.2%) were Amphora copulata (Kützing) Schoeman & Archibald, A. cf. copulata, Amphora eximia Carter, Am- phora inariensis Krammer, Amphora cf. subatomus Levkov, Amphora minutissima W. Smith, A. cf. minutissima and Amphora lange-bertalotii Levkov & Metzeltin, which are not further discussed here. The valves 1–4 depicted in Pl. 1 perfectly fi t the ranges given for A. indistincta and 8–12 (Pl. 1) for A. pediculus (see LEVKOV 2009 or HOFMANN et al. 2013). Thus, they are named without problems. The valves 5–7 (Pl. 1) were labelled A. cf. indistincta, because their habita morphologically corresponded to A. indistincta, but some traits differed from the species description in LEVKOV (2009). The valve widths were too small (1.9–2.9 μm in- stead of 3–4 μm) and dorsal striae density was too high (24–25 instead of 18–22 striae in 10 μm). The valves 13–15 (Pl. 1) were labelled A. cf. pediculus, because they mainly correspond to A. pediculus as even the dorsal areolae are visible using light microscopy despite their small size. However, the valves are too small (5.8–6.9 μm long instead of 7–15 μm and 1.9–2.3 μm wide instead of 2.5–4.0 μm) and the striae are too dense (e.g. dorally 25 and more striae in 10 μm instead of 18–24 in 10 μm). Pl. 1. Light microscopic pictures of Amphora indistincta (1–4), A. cf. indistincta (5–7), A. pedicu- lus (8–12), A. cf. pediculus (13–15) and Amphora spec. (16–17) from Lake Krossinsee, Ger- many. BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 215 The valves 16–17 (Pl. 1) distinctly deviate from the species descriptions of A. indis- tincta and A. pediculus. Thus, it is even more diffi cult to allocate an appropriate name to these valves, which were thus labelled Amphora spec. Overall, the sum of small Amphora-species were more or less similar among partici- pants (1.1–17.6%, average 7.2%) and auditors (6.0–8.7%, average 7.5%) in the sample of Lake Krossinsee (Fig. 1A). Just the results of laboratory 10 (1.1%) and laboratory 15 (17.6%) deviated distinctly from the rest. However, distinct differences occurred when each species is looked at by itself. All participants (0.8–17.6%, average 6.6%) and auditors (av- erage 3.3%) identifi ed A. pediculus (Fig. 1B). However, relative abundances of A. pedicu- lus of the three auditors varied distinctly with 7.8%, 1.2% and 0.9%, respectively. Comple- mentary, the auditors identifi ed none, 7.3% and 3.8% of A. indistincta (Fig. 1C). Of the 37 participants only fi ve detected A. indistincta in sample Lake Krossinsee (Fig. 1C). Despite the occurrence of some small Amphora-valves in Lake Krossinsee (see Pl. 1: 5–7 and 13–17) that do not fi t any Amphora-taxa description in HOFMANN et al. (2011) or Fig. 1. Relative abundance of small Amphora-species from Lake Krossinsee, Germany, based on the results of the fi rst German intercalibration exercise for benthic diatoms. A: Sum of A. inariensis, A. indistincta, A. cf. indistincta, A. pediculus, A. cf. pediculus and Amphora spec. B: A. pediculus, C: A. indistincta, D: Amphora spec., A. cf. indistincta and A. cf. pediculus. Dark grey bars: participants; light grey bars: auditors. DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 216 ACTA BOT. CROAT. 74 (2), 2015 Fig. 2. Relative abundance of Cocconeis placentula Ehrenberg from Lake Krossinsee, Germany, based on the results of the fi rst German intercalibration exercise for benthic diatoms. A: Sum of C. placentula var. lineata, C. placentula var. euglypta, C. placentula var. placentula and C. placentula-aggregate, B: C. placentula var. lineata, C: C. placentula var. euglypta, D: C. placentula var. placentula and E: C. placentula-aggregate, dark grey bars: participants; light grey bars: auditors. LEVKOV (2009) only four laboratories and two auditors called some valves Amphora spec., A. cf. pediculus or A. cf. indistincta (Fig. 1D), i.e. they were uncertain about the identity of some valves from this taxonomic group. Cocconeis placentula in Lake Krossinsee Cocconeis placentula was one of the dominant taxa in the Lake Krossinsee sample (Fig. 2). One main problem was the differentiation of C. placentula var. placentula (Pl. 2: 19–22), C. placentula var. lineata (Plate 2: 23–27) and C. placentula var. euglypta BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 217 Pl. 2. Light microscopic pictures of the rapheless valves of Cocconeis placentula-varieties from the intercalibration exercise sample Lake Krossinsee, Germany, labelled according to the mor- phological concept of KRAMMER and LANGE-BERTALOT (2004). C. placentula var. placentula (19–22), C. placentula var. lineata (23–27), C. placentula var. euglypta (28–31) and raphe- less valves of C. placentula with a striae density of ~24–26 in 10 μm (32–45). These latter valves are C. placentula var. placentula based on their striae density according to KRAMMER and LANGE-BERTALOT (2004). However, the arrangement and shape of the areolae better fi t with C. placentula var. euglypta (32–39) or C. placentula var. lineata (40–45) according to the concept of KRAMMER and LANGE-BERTALOT (2004). DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 218 ACTA BOT. CROAT. 74 (2), 2015 (Pl. 2: 28–31) according to the results of this intercalibration exercise. The remaining Coc- coneis-taxa in Lake Krossinsee (C. placentula var. tenuistriata Geitler, C. neothumensis Krammer, C. pseudolineata (Geitler) Lange-Bertalot and C. pediculus Ehrenberg) were ei- ther only present in low abundances (< 1.0%) or were taxonomically not problematic ac- cording to the results of this intercalibration exercise. Thus, they are not presented or dis- cussed below. Various morphological concepts exist for differentiating the varieties of C. placentula (see e.g. HUSTEDT 1930, GEITLER 1982, KRAMMER and LANGE-BERTALOT 2004, JAHN et al. 2009, ROMERO and JAHN 2013), which are currently in practical use simultaneously. The German instruction protocol (SCHAUMBURG et al. 2006, 2007, 2011) stipulates using HOF- MANN et al. (2011) as standard identifi cation literature for counting diatoms. HOFMANN et al. (2011) refer to KRAMMER and LANGE-BERTALOT (2004) for differentiating the varieties of C. placentula. Thus, we identifi ed the varieties according to KRAMMER and LANGE-BERTALOT (2004). All 37 participants identifi ed C. placentula including the varieties euglypta, lineata or placentula or C. placentula-aggregate (12.9–44.3%, average 24.4%) (Fig. 2A). 19 partici- pants detected C. placentula var. lineata (0.5–31.8%, average 9.9%) (Fig. 2B) and 22 par- ticipants identifi ed C. placentula var. euglypta (0.2–42.8%, average 14.8%) (Fig. 2C). Eleven participants identifi ed C. placentula var. placentula (0.18–29.7%, average 7.6%) (Fig. 2D) of which two (laboratories 15 and 23) did not detect any other varieties of C. pla- centula. 14 participants detected C. placentula-aggregate (2.8–33.9%, average 21.9%) (Fig. 2E). The results of the auditors suggest that C. placentula was dominated by the varieties euglypta and/or lineata in sample Lake Krossinsee. Two auditors exclusively identifi ed C. placentula var. lineata (25.5% and 23.5%, respectively) (Fig. 2B). The third auditor de- tected 21.7% C. placentula var. euglypta (Fig. 2C) and 2.2% C. placentula var. lineata (Fig. 2B). Navicula cryptotenella and N. cryptotenelloides in Lake Krossinsee and Lake Geneva In the sample from Lake Krossinsee the genus Navicula Bory de Saint-Vincent was dominated by Navicula cryptotenella, while N. cryptotenelloides occurred with less than 0.2% (see Pl. 3: 47–68). In contrast, in the sample from Lake Geneva the auditors identifi ed N. cryptotenelloides as the dominant taxon within the genus Navicula, while N. crypto- tenella was very rare (< 1%) but present as well (Pl. 3: 47–68). In both samples other Na- vicula-taxa occurred too, such as N. antonii Lange-Bertalot, N. reichardtiana and N. tri- punctata (O.F. Müller) Bory de Saint-Vincent. However, they will not be further discussed here. In the Lake Krossinsee sample the three auditors identifi ed N. cryptotenella with 5.9– 10.0% (average 8.3%) (Fig. 3B), while 33 of the 37 participants detected N. cryptotenella with 1.4–8.9% (average: 5.4%) (Fig. 3B). N. cryptotenelloides was detected in two auditor samples with 0.4% and 0.6% (average: 0.5%) and in 24 of 37 participant samples with 0.2–3.1% (average: 0.9%) (Fig. 3C). Especially the results of the participants 15, 19, 22 and 28 expose taxonomic problems. These laboratories did not detect any N. cryptotenella in sample Lake Krossinsee (Fig. 3B). Laboratories 15 and 28 only identifi ed N. antonii and were the only participants that did not fi nd any evidence of N. cryptotenella or N. crypto- BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 219 tenelloides (see Fig. 3A). Additionally, laboratories 13, 19, 22 and 25 detected a consider- able percentage of the taxa discussed here with ambiguity (labelled with cf.). Similarly to the participants, the results of the auditors suggest taxonomic diffi culties of these Navicula- taxa. For example, auditor 40 named some counted objects as N. cf. cryptotenelloides (1.4%). In the Lake Geneva sample 35 of 37 participants detected N. cryptotenelloides (includ- ing cf.) and/or N. cryptotenella (including cf.) (1.4–11.8%, average: 7.0%) (Fig. 3D). Labo- ratories 28 and 36 did not identify any N. cryptotenelloides or N. cryptotenella (Fig. 3D). In contrast to single taxa abundances (Figs. 3E–F) the sum of N. cryptotenelloides and/or N. cryptotenella were uniform among auditors (average: 7.0%) (Fig. 3D). N. cryptotenel- loides was detected by 31 of 37 participants (1.4–11.0%, average: 6.4%) and by two of three auditors (average: 7.3%) (Fig. 3E). N. cryptotenella was not detected by any auditor in this sample (Fig. 3F). In contrast, 25 of the 37 participants identifi ed N. cryptotenella (0.2–6.1%, average: 1.6%) (Fig. 3F). One auditor (no. 40) counted solely N. cf. crypto- tenella. Similarly, four participating laboratories counted N. cf. cryptotenella (laboratories Pl. 3. Light microscopic pictures of Navicula cryptotenella (47–55), N. cryptotenelloides (56–61) and valves that can not be identifi ed with certainty with a valve width between 4.2 and 5.0 μm (62–68) from the Lake Krossinsee sample, Germany (47–55 and 61, 63–65 and 67), and Lake Geneva sample, Switzerland (56–60, 62, 66 and 68). DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 220 ACTA BOT. CROAT. 74 (2), 2015 Fig. 3. Relative abundance of Navicula cryptotenella and N. cryptotenelloides from samples Lake Krossinsee, Germany (A–C), and Lake Geneva, Switzerland (D–F), based on the results of the fi rst German intercalibration exercise for benthic diatoms. A: Sum of Navicula crypto- tenella and N. cryptotenelloides (including cf.), B: N. cryptotenella, C: N. cryptotenelloides, D: Sum of N. cryptotenella and N. cryptotenelloides (including cf.), E: N. cryptotenelloides, F: N. cryptotenella, dark grey bars: participants; light grey bars: auditors. BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 221 1 and 24) or N. cf. cryptotenelloides (laboratories 16 and 34). The results of participants 10, 16, 19, 26, 28 and 36 particularly reveal taxonomic problems. Laboratories 28 and 36 did not detect either N. cryptotenelloides or N. cryptotenella (Fig. 3D). Laboratories 10, 16, 19 and 26 did not detect any N. cryptotenelloides (Fig. 3E). Laboratories 10, 19 and 26 solely identifi ed N. cryptotenella (Figs. 3E–F). Laboratory 16 counted mainly N. cf. cryptotenel- loides. Similarly, the results of auditor no. 40, who solely identifi ed N. cf. cryptotenella, emphasize the taxonomic problems within the N. cryptotenella and N. cryptotenelloides- group. Besides general problems with identifying N. cryptotenella and N. cryptotenelloides, there is an additional problem that led to distinct diffi culties with differentiating N. crypto- tenella and N. cryptotenelloides. According to LANGE-BERTALOT (1993, 2001) the two taxa can always and certainly be distinguished by their width, as the given width for N. crypto- tenelloides is 3.7–4.2 μm and for N. cryptotenella is 5.0–7.0 μm. Overlapping valve width (4.2–5.0 μm) is not supposed to occur (LANGE-BERTALOT 1993, 2001). In contrast, valves with a valve width of 4.2–5.0 μm occurred regularly in the samples examined here. For elucidating this problem of overlapping valve width 38 valves of the N. cryptotenella and N. cryptotenelloides-group were measured and photographed in each of the Lake Krossin- see and Lake Geneva samples (Fig. 4). Of the 38 measured valves from Lake Geneva 23 valves had a width of 3.6–4.2 μm (Fig. 4), corresponding to N. cryptotenelloides. Interestingly, all 23 valves had a striae den- sity of 18.5–21.0 in 10 μm (Fig. 4), which consistently exceeds 16–18 striae in 10 μm, the given range for N. cryptotenelloides (LANGE-BERTALOT 2001, HOFMANN et al. 2011). One of the 38 Lake Geneva valves had a width of 5.5 μm and 17 striae in 10 μm. Even though the striae density is slightly too high according to LANGE-BERTALOT (2001) and HOFMANN et al. (2011), this valve is probably N. cryptotenella. The remaining 14 valves (36.8%) of Lake Geneva had a width between 4.2–5.0 μm (Fig. 4), contrasting the details given in LANGE- BERTALOT (1993, 2001), i.e. the width were in the range between N. cryptotenella and N. cryptotenelloides. These 14 valves had a striae density between 17.3 and 21.0 in 10 μm and were thus more similar to N. cryptotenelloides than to N. cryptotenella. However, it re- mains disputable which taxon these valves belong to. Thus, we suggest labelling these valves as cf. Fig. 4. Striae density and valve width from valves of the Navicula cryptotenella and N. cryptotenello- ides–complex from 38 valves from Lake Geneva, Switzerland (black diamonds), and 38 valves from Lake Krossinsee (grey circles). Vertical dashed lines denote the valve width of N. cryp- totenelloides (3.7–4.2 μm) and N. cryptotenella (5.0–7.0 μm), horizontal dashed line denotes the differentiating striae density of 16 striae in 10 μm according to LANGE-BERTALOT (2011). DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 222 ACTA BOT. CROAT. 74 (2), 2015 Of the 38 measured valves from Lake Krossinsee two valves had a width of ~3.7 and 4.0 μm and a striae density of ~18 and 20 in 10 μm, respectively (Fig. 4). Correspondingly, these valves were probably N. cryptotenelloides. 13 valves had a width > 5.0 μm (Fig. 4), matching N. cryptotenella. Of these 13 valves 10 valves had ~14–16 striae in 10 μm (Fig. 4) and are thus probably N. cryptotenella. Three of these 13 valves had ~17–18 striae in 10 μm (Fig. 4) and are therefore only ambiguously N. cryptotenella. The remaining 23 valves (60.5%) from Lake Krossinsee had a width between 4.2–4.9 μm (Fig. 4), i.e. in a range be- tween the width of N. cryptotenella and N. cryptotenelloides. Similarly, the striae density ranged between 15.5 and 19.9 in 10 μm (Fig. 4). Thus, it remains unclear if the valves be- long to N. cryptotenella or N. cryptotenelloides and should also best be labelled as cf. Despite these distinct deviations from the description of N. cryptotenella and N. crypto- tenelloides of most valves from Lake Krossinsee and Lake Geneva (Fig. 4), only four of the 37 participants and one auditor labelled any valves with cf for Lake Krossinsee and only four other participants and the same auditor used cf. for these valves from Lake Geneva. Navicula reichardtiana and N. caterva in Lake Geneva Navicula reichardtiana was detected by 25 of the 37 participants and two auditors in the Lake Geneva sample. One additional participant and the third auditor identifi ed N. cf. reich- ardtiana. N. caterva was detected by fi ve participants and no auditor. The sum of all N. reichardtiana and N. caterva from all participants and auditors was below 2.5%. Thus, not listing these taxa does not necessarily suggest an identifi cation mistake or problem. As two auditors and most participants unambiguously identifi ed N. reichardtiana and no auditor and only fi ve participants identifi ed N. caterva, we could assume that the latter are misidentifi cations. However, with the data at hand a misidentifi cation cannot be veri- fi ed. More taxonomic examinations would be necessary. The workshop that was conducted subsequent to the intercalibration exercise identifi ed complicated taxonomic problems in- stead of simple misidentifi cation leading to some participants determining N. caterva in- stead of N. reichardtiana and to two participants identifying both taxa in the same sample. Similarly, one auditor only identifi ed N. cf. reichardtiana from this complex. In the follow- ing we will demonstrate these problems exemplary by describing the identifi cation ap- proaches of two participants (A and B) when identifying N. reichardtiana and N. caterva (Fig. 5). Both participants measured the striae density along the margin of the axial area, begin- ning at the end of the central area. Participant A counted the striae along a 5 μm scale, while participant B used a shorter scale (Fig. 5). Both extrapolated to striae density in 10 μm (Tab. 1). These different interpretations of striae counting methods and consequently vary- ing measuring methods only led to small differences in the extrapolated number of striae in 10 μm (16.2 and 15.1, respectively; Tab. 1), but have a great impact on the identifi cation process (see below). Interpretation and species identifi cation by participant A: Participant A argues that the width of the valve in the given example (Fig. 5) fi ts the range of N. caterva (4.2–5.5 μm according to LANGE-BERTALOT 2001) and is too small for N. reichardtiana (5–6 μm according to LANGE-BERTALOT 2001). The striae density (16.17 in 10 μm) only slightly exceeds the range for N. reichardtiana (14–16 in 10 μm according to LANGE-BERTALOT 2001) and fi ts the range of N. caterva ((16)18–21 in 10 μm according to BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 223 LANGE-BERTALOT 2001). Participant A denotes the central area as small, which characterises both species (LANGE-BERTALOT 2001). The striae orientation changes only once abruptly (bottom right in Fig. 5) and three times gradually, which participant A thinks fi ts better to N. caterva than to N. reichardtiana (see discussion, i.e. remarks about discrepancies be- tween the text that describes the species and the fi gures that depict the species in LANGE- BERTALOT 2001 and HOFMANN et al. 2011). In summary, participant A identifi es the valve in Fig. 5 as N. caterva. Interpretation and species identifi cation by participant B: Participant B argues that the width of the valve in the given example (Fig. 5) only so slightly exceeds the range of N. reichardtiana that this valve may still be identifi ed as N. reichardtiana. Striae density (15.14 in 10 μm) fi ts the range of N. reichardtiana and is outside the range of N. caterva. Similar to participant A, participant B denotes the central area as small, which characterises both species. The striae orientation changes abruptly once (bottom right, see Fig. 5), which distinctly characterises N. reichardtiana according to participant B. In summary, participant B argues that the valve in Fig. 5 more resembles N. reichardtiana than N. caterva. However, as the width is slightly outside of the range of N. reichardtiana and the striae orientation changes only gradually on three sides, partici- pant B identifi es the valve in Fig. 5 as N. cf. reichardtiana. Fig. 5. Example of different striae-counting approaches (vertical lines) from participant A (left) and B (right) on a valve from the Navicula reichardtiana and N. caterva-group from sample Lake Geneva. Scale bar (horizontal line) = 5 μm. Tab. 1. Results of measurements of participant A and B of the valve shown in Fig. 5 of the Navicula reichardtiana and N. caterva-group from sample Lake Geneva. Participant A measured di- rectly in μm while participant B measured in pixels and then converted into μm. In this given example 0.0549 pixels convert to 1 μm. Participant Length (μm) Width (μm) Striae/μm (pixel) Striae/ 10 μmtop left top right bottom left bottom right A 15.93 4.78 8.25/5 8/5 8/5 – 16.17 B 16.36 4.78 5/3.3 (60) 5/3.4 (62) 5/3.3 (60) 4/2.6 (47) 15.14 DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 224 ACTA BOT. CROAT. 74 (2), 2015 Both participants considered very carefully all traits when identifying th e valve in Fig. 5. Still, they arrived at different names for the very same valve. Based on the different approaches of the two participants it is not surprising that they also named many other valves differently from each other. For example, when identifying 15 different valves from the N. reichardtiana and N. caterva-complex, participants A and B agreed only in fi ve cases and differed in the naming of the remaining ten valves. Discussion The taxonomic problems that occurred in the intercalibration exercise for benthic dia- toms and that we discuss in the following sections based on a few examples represent only a fraction of the problems that occur during routine diatom-counts in practise. Interviews and discussions during the workshop of the intercalibration exercise identifi ed several rea- sons for deviating counting results. Sometimes participants differed in the interpretation of the species specifi cations in the identifi cation literature or their results differed due to insuf- fi cient use of taxonomic specifi cations. In other cases the results differed due to partly am- biguous species descriptions and to misleading recommendations about which traits to use for species identifi cation or differentiation in the identifi cation literature and also due to currently insuffi cient taxonomic and ecological knowledge of some diatom taxa. 1. Using different identifi cation literature The problem of using different identifi cation literature was especially apparent when identifying the small Amphora-species and the varieties of Cocconeis placentula. For ex- ample, Amphora pediculus and A. indistincta occurred roughly with similar abundances in Lake Krossinsee. Still, A. indistincta was detected only by fi ve of the 37 participants and by two auditors. One important reason for this discrepancy is the fact that some participants did not use the current and mandatory identifi cation literature (e.g. LEVKOV 2009, HOFMANN et al. 2011). Some participants only used KRAMMER and LANGE-BERTALOT (1986–2004), which do not incorporate A. indistincta. Thus, these participants did not fi nd any A. indis- tincta. Instead A. indistincta valves from Lake Krossinsee were counted as A. pediculus or A. inariensis. This problem can be resolved fairly easily by using the new identifi cation lit- erature (e.g. LEVKOV 2009, HOFMANN et al. 2011 or 2013). Similarly, participants used various identifi cation literatures when identifying the vari- eties of Cocconeis placentula. This is especially problematic, because several morphologi- cal concepts exist for distinguishing the varieties of C. placentula (see e.g. HUSTEDT 1930, GEITLER 1982, KRAMMER and LANGE-BERTALOT 2004, JAHN et al. 2009, ROMERO and JAHN 2013) that are concurrently in practical use. Thus, in this case counting results can only be comparable, if the same literature is used or if the used literature is detailed and the concept is documented with pictures and a comprehensive and detailed accompanying text. 2. Morphological variability Another problem apparent from the intercalibration exercise is the relatively frequent presence of valves from certain diatoms that are outside the given morphometric ranges of the species in one or more traits. Thus, identifi cation without ambiguity is not always pos- sible. This phenomenon occurred commonly in all four taxonomic groups presented here BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 225 from the Lake Krossinsee and Lake Geneva samples. For example, the valves labelled A. cf. indistincta (Pl. 1: 5–7) and A. cf. pediculus (Pl. 1: 13–15) were distinctly too small and had striae that were too dense compared to their description in LEVKOV (2009). Most rapheless valves from the Cocconeis placentula-aggregate from Lake Krossinsee had a stri- ae density of ~24–26 in 10 μm (Pl. 2: 32–45), corresponding to C. placentula var. placen- tula according to KRAMMER and LANGE-BERTALOT (2004). However, the arrangement and the shape of the areolae rather correspond to C. placentula var. euglypta or C. placentula var. lineata according to KRAMMER and LANGE-BERTALOT (2004). Also, about 37% of the valves of the Navicula cryptotenella and N. cryptotenelloides-complex from Lake Geneva had a width between 4.2–5.0 μm and ~ 61% of the Lake Krossinsee valves from this complex (Pl. 3: 62–68, Fig. 4). This width is exactly between the width of N. cryptotenella and N. cryptotenelloides and should not occur according to LANGE-BERTALOT (1993, 2001). On the contrary, the width is supposed to be a very good criterion to discriminate the two spe- cies (LANGE-BERTALOT 1993, 2001). Additionally, most valves of the N. cryptotenella and N. cryptotenelloides-complex had a higher striae density than specifi ed in LANGE-BERTALOT (2001) (see Fig. 4). At least two reasons may be responsible for the frequent occurrence of diatom-valves that are outside the given morphological description in some traits but that otherwise fi t well to the species specifi cations: (1) Some described species have a greater morphological range than in the given species specifi cations (see e.g. WOLF et al. 2002, DRESSLER and HÜBENER 2006). (2) Some species specifi cations describe species-groups instead of species, i.e. they include several similar species that have not been examined suffi ciently neither taxonomically nor ecologically so far (see e.g. EVANS et al. 2008, JÜTTNER et al. 2013). Both reasons indicate some challenges for taxonomists that will improve diatom identifi cation in practise. One problem is that some type material for species descriptions contains only a few valves and that species descriptions are often only based on one or few sites or species populations. Thus, some species descriptions include just a small part of the true morpho- logic variability and ecological range of a species. Consequently, it would be desirable if the species descriptions and details in the identifi cation literature included measures that enable the applied limnologist and other diatomists to better assess the morphological and ecological variability of a taxon. For example, it would be desirable to know on how many valves from how many populations, samples and inland waters a species description is based on. 3. Ambiguous taxa descriptions and documentations in the identifi cation literature Besides the morphological variability of taxa described above, diatom-identifi cation is further complicated by ambiguous or relatively short taxa-descriptions in some cases. For example, in the German Phylib-method for assessing water quality (SCHAUMBURG et al. 2006, 2011) HOFMANN et al. (2011) is listed as the mandatory and main identifi cation litera- ture. They (HOFMANN et al. 2011) refer to KRAMMER and LANGE-BERTALOT (2004) for the differentiation of the Cocconeis placentula varieties. Additionally, they recommend not dif- ferentiating the C. placentula varieties for now, when assessing the water quality with the German Phylib-method (except Cocconeis pseudolineata (Geitler) Lange-Bertalot). Thus, this latter comment could be responsible for just listing C. placentula in the intercalibration DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 226 ACTA BOT. CROAT. 74 (2), 2015 exercise in Lake Krossinsee by 14 laboratories, next to the morphological problems. How- ever, when using the German Phylib-method (SCHAUMBURG et al. 2006, 2011) a separation of the varieties is essential, as the indicator values of the C. placentula-varieties differ from one another (German water quality assessment software Phylib, version 4.1; October 2012). Thus, using the mandatory German water quality assessment software, the lumping or split- ting of C. placentula (and varieties) will affect the results of the ecological water quality assessment. According to KRAMMER and LANGE-BERTALOT (2004) the differentiation of the varieties should be based on the fi ne structures of the rapheless valves that are visible in the light microscope. Accordingly, the varieties placentula and tenuistriata can be easily distin- guished based on the striae density (KRAMMER and LANGE-BERTALOT 2004). Problems occur with the most common varieties, i.e. lineata and euglypta, as the striae density and number of areolae per stria of euglypta (19–22 striae in 10 μm and 3–5 areolae per stria) are entirely within the range of lineata (16–23 striae in 10 μm and 3–10 areolae per stria) (KRAMMER and LANGE-BERTALOT 2004). Additionally, the areolae of euglypta are supposed to be more robust compared to lineata, whereas the striae of lineata are supposed to appear distinctly dotted (KRAMMER and LANGE-BERTALOT 2004). However, the differences between robust and distinct areolae remain unclear. The variety lineata is supposed to often have slit-like areo- lae which appear somewhat isolated from one another and an irregular or zigzag pattern of the longitudinal lines of areolae (areolae along the apical axis) (KRAMMER and LANGE-BER- TALOT 2004). Overall, it remains diffi cult to clearly distinguish the varieties euglypta and lineata based on these rather vague traits. Thus, these vague descriptions probably explain some of the inconsistent results of the intercalibration exercise with respect to the euglypta and lineata varieties. Another problem is the mismatch between pictures and species description in the identi- fi cation literature for the varieties euglypta and lineata (e.g. KRAMMER and LANGE-BERTALOT 2004). For example, the valves on Plate 53, Figures 17–18 (page 354) depict C. placentula var. euglypta in KRAMMER and LANGE-BERTALOT (2004), but show distinctly more than fi ve areolae per stria and a pattern of the longitudinal lines of areolae that can be called irregular or an arrangement in a zigzag pattern (which should both be typical for lineata not euglypta according to KRAMMER and LANGE-BERTALOT 2004). Based on the currently inadequate state of information about the taxonomy and ecology of the varieties of C. placentula, it is diffi cult to make recommendations for the practical use (except: document your choices with pictures and text). Further taxonomic and ecologi- cal work is necessary for a distinct differentiation of the C. placentula varieties. During the identifi cation of valves from the N. reichardtiana and N. caterva-complex two participants (A and B) carefully considered the whole combination of characters (see Tab. 1) to identify the valve in Fig. 5. However, they differ in their identifi cation of the valve in Fig. 5 due to different approaches and interpretation, especially of the traits »striae density" and »changes of striae orientation towards the poles". It is diffi cult or even impos- sible to decide, which participant is correct, as there are some ambiguities that cannot be resolved. It is not documented in suffi cient detail in which way the describing authors counted the striae density. Thus, no decision can be made about which measuring approach is more appropriate. Ultimately, striae density should be measured in the same manner as the describing authors. However, this is often not known or documented. BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 227 The text of LANGE-BERTALOT (2001) and HOFMANN et al. (2011) describe that the striae orientation towards the poles of N. reichardtiana changes abruptly and of N. caterva gradu- ally. However, the pictures in LANGE-BERTALOT (2001) and HOFMANN et al. (2011) do not al- ways correspond to this description. For example, HOFMANN et al. (2011) depict valves of N. caterva with abruptly changing striae orientation towards the poles (Pl. 31, Fig. 38, stri- ae top left and bottom left). The same valve is also depicted in LANGE-BERTALOT (2001) (Fig. 7, Pl. 33) and also named N. caterva. Similarly, valves of N. reichardtiana are depict- ed with gradually changing striae orientation towards the poles in HOFMANN et al (2011) (Pl. 31, Figure 29 and 30, for both: top striae) and in LANGE-BERTALOT (2001) (Pl. 13 Fig. 26) (gradual 3-times except bottom right). Thus, for identifying N. reichardtiana and N. cater- va the question arises what to do with the mismatch between describing text and describing pictures in the identifi cation literature. Currently, there is a rather wide range of possibilities to interpret the trait »striae orientation towards the poles«, which partly explains the differ- ent approaches of participants A and B (see Tab. 1). Again, it would be key to know how the describing authors interpreted this trait. There are more mismatches of pictures and species descriptions similar to the C. pla- centula-varieties and to N. reichardtiana / caterva in the identifi cation literature. For exam- ple, the areolae on the dorsal striae of A. pediculus should typically be visible in a light mi- croscope according to LEVKOV (2009) and HOFMANN et al. (2011), but are not discernable on most valves labelled A. pediculus in HOFMANN et al. (2011; page 785). Overall, we recommend always using all three: the pictures, the measurable dimensions (e.g. width, length, striae density) and also the verbal species descriptions when identifying diatoms. In case of mismatching pictures and descriptions the applied limnologist should document with text and pictures the approach and reasoning taken to identify the diatom. The taxonomist may correct the current mismatches and avoid them in future by choosing the pictures more carefully or by providing appropriate explanations. 4. Marking identifi cation uncertainties Another substantial problem became apparent in the groups of Navicula cryptotenella / N. cryptotenelloides, N. reichardtiana / N. caterva and the small Amphora-species. For ex- ample, 37% (Lake Geneva) and 61% (Lake Krossinsee) of the measured 38 valves had a width between the width of N. cryptotenella and N. cryptotenelloides. Still, only eight of the 37 participants and one auditor labelled the Navicula cryptotenella / N. cryptotenelloi- des results with uncertainties by using cf. in either lake. Similarly, part of the Amphora- valves in the Lake Krossinsee sample deviated distinctly in their morphology from the de- scriptions in HOFMANN et al. (2011) and LEVKOV (2009) (see Pl. 1: 5–7, 13–15, 16–17). However, only four laboratories and two auditors indicated any uncertainty about the iden- tifi ed small Amphora-valves by using »spec.« and »cf.«. Overall, it seems very likely that A. indistincta and A. pediculus have wider morpho- logical ranges than given in the identifi cation literature (LEVKOV 2009, HOFMANN et al. 2011). Thus, for example the valves 5–7 (Pl. 1) (A. cf. indistincta) and 13–15 (Pl. 1) (A. cf. pediculus) may actually belong to their respective species. However, as long as fur- ther taxonomic and morphological examinations do not clarify the dimensional ranges, these and other valves that are outside their species specifi cations, should be counted with a »cf.« and be documented with a picture. DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 228 ACTA BOT. CROAT. 74 (2), 2015 One reason for submitting results with more certainty than actually present was identi- fi ed during the workshop of the intercalibration exercise. Several participants reported that they noted deviations of some valves from the descriptions without labelling them with a cf. They reasoned that the German method for implementing the Water Framework Direc- tive (SCHAUMBURG et al. 2006, 2011) only allows a maximum of 5.0% diatom objects that could not be determined (sp., spp.) and/or could not unambiguously be determined (cf.). Otherwise the diatom-indices are assumed to be unreliable and the sample cannot contrib- ute to the water quality assessment. Thus, the participants wanted to avoid the exclusion by indicating ambiguity (using cf.) as little as possible. As a consequence of this German method it is common practise to pool certain ambigu- ous taxa with other taxa to avoid labelling objects with »spec.« or »cf.« for the ecological water quality assessment to allow for a seemingly reliable assessment according to some participants. Ultimately, this issue is also a psychological problem. A contractor may in- crease (or think to increase) the chances of future assignments without further quality con- trol (such as participating in an intercalibration exercise), if he or she always delivers »good, clean and reliable results« and who seemingly identifi es all or almost all diatoms in a sample with certainty. Thus, there is the danger that a low number of ambiguous taxa be- comes an alleged quality attribute, especially for the German method. Nonetheless, ambig- uous taxa should be labelled with »cf.« or »spec.« also in practise (i.e. by the applied lim- nologist) and be documented with pictures to facilitate or allow a later comparison and verifi cation of counting results. In principle, quality assessments of counting results are very useful. However, the < 5%-cf cut-off is problematic, as there are samples in practise with less than 5% ambiguous taxa that contain only very few indicative taxa (WERNER and DRESSLER 2007). Thus, they would still be considered reliable according to the German method (SCHAUMBURG et al. 2006, 2007). In this respect a new quality assessment was intro- duced for lakes in the German method, whereby samples with < 60% indicative taxa are deemed unreliable (SCHAUMBURG et al. 2011). However, it is problematic that they retained the < 5%-cf cut-off in this context, because samples are still deemed unreliable that may have a high abundance (>> 60%) of indicative taxa. 5. Ecological references and species identifi cation Another problem for the applied limnologist is the recommendation in the identifi cation literature to use the ecological preferences of a taxon as an additional criterion for taxa identifi cation for some taxa. For example, in contrast to A. pediculus, Amphora indistincta is supposed to live in nutrient-poor waters (HOFMANN et al. 2011). This may lead to the pre- clusion or disregard of certain taxa during the identifi cation process. However, in some cases the ecological preferences are not verifi ed suffi ciently. For example, A. indistincta and A. pediculus lived concurrently in Lake Krossinsee, despite their different ecological amplitude described in HOFMANN et al. (2011). Diatom assemblages are used to infer the ecological class status of a water body in the routine counts for the European Water Framework Directive. It would thus be a circular ar- gument to use ecological preferences for identifying the diatom taxa, which in turn, are used to determine the water quality. Thus, taxa that are not supposed to occur in a certain water body or lake/river-type should not be disregarded during the identifi cation process. BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 229 Conclusions The fi rst German intercalibration exercise for benthic diatoms identifi ed the following fi ve major issues and thus generated the subsequent recommendations. 1) Use of different identifi cation literature led to different species names of the same taxa. 2) Some diatom valves of the intercalibration exercise had morphological dimensions outside the given species ranges to a certain degree in one or more traits. 3) Participants and auditors also allocated different names to the same taxa due to am- biguous species descriptions and certain recommendation of traits that should be used to differentiate species in the mandatory identifi cation literature. Also, the pic- tures do not always correspond to the species description in the literature, which led to further deviation among counting results. 4) Some taxa were reported with more certainty than actually present, i.e. despite mor- phologic deviations from the defi ned ranges in the identifi cation literature the valves were not labelled with a »cf.«. 5) Insuffi cient knowledge about the ecology of some species with misleading recom- mendations based on this ecology for some diatoms led to different naming. Recommendation for diatom counts by applied limnologists: During routine diatom counts in practise we recommend to carefully consider all traits defi ning a species and to use the mandatory identifi cation literature. Also, for problematic taxa or ambiguous taxa the actually used traits for identifi cation should be specifi ed and the used identifi cation books should be specifi ed to the contracting authority, as it may make a difference if you, for example, only use HOFMANN et al. (2011) or if you can additionally use KRAMMER (1997a, b) when identifying Encyonopsis ssp. or Encyonema ssp.. Additionally, we recommend a photographic documentation of ambiguous taxa (spec., aff., cf.) for a bet- ter comparability of counting results among diatomists and for the possibility of a later ad- justment of the results according to new research results. Ambiguous diatom valves should be labelled with »spec.«, »cf.« or »aff.« in the counting results. In the German method for water quality assessment using diatoms (SCHAUMBURG et al. 2011) the relative abundance of 5% ambiguous taxa (or more) should not make the results of a count unreliable. Instead, this measure should be replaced by other measures, such as the percentage of indicative taxa in a sample that contributed to the water quality assessment that should be above a certain threshold. Recommendations for taxonomists: When describing new diatom-taxa and when compiling identifi cation books the taxono- mist should keep in mind that his or her work will also be used by the applied limnologist and not just by fellow scientists. Accordingly, ambiguous documentations should be avoid- ed, i.e. all traits necessary for identifi cation should be depicted in detail and without ambi- guity and according to unambiguous text descriptions. For an assessment of the possible morphological variability of a taxon the taxonomist needs to provide basic data, i.e. on how many valves are the given ranges in the identifi cation literature based on, from how many samples and/or populations and from how many different inland waters (sites). Similarly, data about the ecology of each taxon should be part of the basic data (if available). Also important is the method description that was used to generate the measured ranges, espe- DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 230 ACTA BOT. CROAT. 74 (2), 2015 cially of striae density. Surely, these recommendations have often already been followed, but this intercalibration exercise demonstrates that there is room for improvement. Overall, the recommendations for the applied limnologist may help to reduce the error when assessing the water quality with diatoms. An improved method that already works well will remain an important tool for water managers and may thus be used even more of- ten. Also, the intercalibration exercise revealed the need for funding fundamental diatom research, as the taxonomist can help to further improve the use of diatoms as bioindicators by investigating species specifi cations. Acknowledgements Our thanks go to all 37 participants of the fi rst German intercalibration exercise, the auditors (Gabriele Hofmann, Thomas Hübener and Peter Pfi ster), Sven Adler for the statis- tical analyses and Andreas Meybohm for the administrative operation of the intercalibra- tion exercise and for providing the internet platform (http://www.planktonforum.eu/). References DRESSLER, M., HÜBENER, T., 2006: Morphology and ecology of Cyclostephanos delicatus (Genkal) Casper & Scheffl er (Bacillariophyceae) in comparison with C. tholiformis Stoermer, Håkansson & Theriot. Nova Hedwigia 82, 409–434. DRESSLER, M., SCHWARZ, A., HÜBENER, T., ADLER, S., SCHARF, B. W., 2011: Use of sedimen- tary diatoms from multiple lakes to distinguish between past changes in trophic state and climate: evidence for climate change in northern Germany during the past 5,000 years. Journal of Paleolimnology 45, 223–241. EU-WFD 2000: European Water Framework Directive 2000, Directive 2000/60/EC of the European Parliament and of The Council of 23 October 2000 establishing a framework for the Community action in the fi eld of water policy. Offi cial Journal L 327. EVANS, K. M., WORTLEY, A. H., SIMPSON, G. E., CHEPURNOV, V. A., MANN, D. G., 2008: A molecular systematic approach to explore diversity within the Sellaphora pupula spe- cies complex (Bacillariophyta). Journal of Phycology 44, 215–231. GEITLER, L., 1982: Die infraspezifi schen Sippen von Cocconeis placentula des Lunzer See- bachs. Algological Studies 30, 1–11. HOFMANN, G., WERUM, M., LANGE-BERTALOT, H., 2011: Diatomeen im Süßwasser-Benthos von Mitteleuropa. Bestimmungsfl ora Kieselalgen für die ökologische Praxis. Über 700 der häufi gsten Arten und ihre Ökologie. A.R.G. Gantner Verlag K.G., Rugell. HOFMANN, G., WERUM, M., LANGE-BERTALOT, H., 2013: Diatomeen im Süßwasser-Benthos von Mitteleuropa. Bestimmungsfl ora Kieselalgen für die ökologische Praxis. Über 700 der häufi gsten Arten und ihre Ökologie. 2. korrigierte Aufl age. Koeltz Scinetifi c Books, Königstein. HUSTEDT, F., 1930: Bacillariophyta (Diatomeae). In: PASCHER, A. (ed.) Die Süsswasser-Flora Mitteleuropas. Heft. 10, 2. Aufl age, Jena. HÜBENER, T., ADLER, S., WERNER, P., SCHULT, M., ERLENKEUSER, H., MEYER, H., BAHNWART, M., 2009: A multi-proxy paleolimnological reconstruction of trophic state reference BENTHIC DIATOMS IN GERMANY: APPLICABLE APPROACH ACTA BOT. CROAT. 74 (2), 2015 231 conditions for stratifi ed carbonate-rich lakes in northern Germany. Hydrobiologia 631, 303–327. JAHN, R., KUSBER, W. H., ROMERO, O. E., 2009: Cocconeis pediculus Ehrenberg and C. pla- centula Ehrenberg var. placentula (Bacillariophyta): Typifi cation and taxonomy. Fottea 9, 275–288. JÜTTNER, I., ECTOR, L., REICHARDT, E., VAN DE VIJVER, B., JARLMAN, A., KROKOWSKI, J., COX, E. J., 2013: Gomphonema varioreduncum sp. nov., a new species from northern and western Europe and a re-examination of Gomphonema exilissimum. Diatom Research 28, 303–316. KAHLERT, M., KELLY, M., ALBERT, R.-L., ALMEIDA, S. F. P., BESTA, T., BLANCO, S., COSTE, M., DENYS, L., ECTOR, L., FRANKOVA, M., HLUBIKOVA, D., IVANOV, P., KENNEDY, B., MARVAN, P., MERTENS, A., MIETTINEN, J., PICINSKA-FAŁTYNOWICZ, J., ROSEBERY, J., TORNES, E., VIL- BASTE, S., VOGEL, A. 2012: Identifi cation versus counting protocols as sources of uncer- tainty in diatom-based ecological status assessments. Hydrobiologia 695, 109–124. KALBE, L., WERNER, H., 1974: Das Sediment des Kummerower Sees. Untersuchungen des Chemismus und der Diatomeenfl ora. Internationale Revue der Gesamte Hydrobiologie 59, 755–782. KELLY, M. G., 2001: Use of similarity measures for quality control of benthic diatom sam- ples. Water Research 35, 2784–2788. KELLY, M. G., 2013: Building capacity for ecological assessment using diatoms in UK riv- ers. Journal of Ecology and Rural Environment 36, 89–94. KRAMMER, K., 1997a: Die cymbelloiden Diatomeen, Eine Monographie der weltweit bekannten Taxa, Teil 1 Allgemeines und Encyonema Part. Bibliotheca Diatomologica Band 36. J. Cramer, Berlin Stuttgart. KRAMMER, K., 1997b: Die cymbelloiden Diatomeen, Eine Monographie der weltweit bekannten Taxa, Teil 2 Encyonema Part., Encyonopsis and Cymbellopsis. Bibliotheca Diatomologica Band 37. Cramer, Berlin Stuttgart. KRAMMER, K., 2000: The genus Pinnularia. Diatoms of Europe Vol. 1. A.R.G. Gantner Ver- lag K.G., Ruggel. KRAMMER, K., 2002: Cymbella. Diatoms of Europe Vol. 3. A.R.G. Gantner Verlag K.G., Ruggel. KRAMMER, K., 2003: Cymbopleura, Delicata, Navicymbula, Gophocymbelloides, Afrocym- bella. Diatoms of Europe Vol. 4. A.R.G. Gantner Verlag K.G., Ruggel. KRAMMER, K., LANGE-BERTALOT, H., 1986–1991, 2004: Süßwasserfl ora von Mitteleuropa, Bacillariophyceae. 2/1: Naviculaceae; 2/2: Bacillariaceae, Epithemiaceae, Surirellace- ae; 2/3: Centrales, Fragilariaceae, Eunotiaceae; 2/4: Achnanthaceae (ergänzter Nach- druck). Fischer Verlag, Stuttgart. LANGE-BERTALOT, H., 1993: 85 Neue Taxa und über 100 weitere neu defi nierte Taxa ergän- zend zur Süßwasserfl ora von Mitteleuropa Vol. 2/1–4. Bibliotheca Diatomologica 27. J. Cramer, Berlin, Stuttgart. LANGE-BERTALOT, H., 2001: Navicula sensu stricto. 10 Genera separated from Navicula sen- su lato. Frustulia. Diatoms of Europe 2. A.R.G. Gantner Verlag K. G., Ruggel. LANGE-BERTALOT, H., METZELTIN, D., 1996: Oligotrophie-Indikatoren. 800 Taxa repräsenta- tiv für drei diverse Seen-Typen. Iconographia Diatomologica 2. Koeltz Scientifi c Books, Königstein. DREßLER M., VERWEIJ G., KISTENICH S., KAHLERT M., WERNER P. 232 ACTA BOT. CROAT. 74 (2), 2015 LANGE-BERTALOT, H., MOSER, G., 1994: Brachysira. Monographie der Gattung. Bibliotheca Diatomologica 29. J. Cramer, Berlin, Stuttgart. LEVKOV, Z., 2009: Amphora sensu lato. Diatoms of Europe 5. A.R.G. Gantner Verlag K. G., Ruggel. PRYGIEL, J., 2002: Management of the diatoms monitoring networks in France. Journal of Applied Phycology 14, 19–26. PRYGIEL, J., CARPENTIER, P., ALMEIDA, S., COSTE, M., DRUART, J.-C., ECTOR, L., GUILLARD, D., HONORÉ, M.-A., ISERENTANT, R., LEDEGANCK, P., LALANNE-CASSOU, C., LESNIAK, C., MER- CIER, I., MONCAUT, P., NAZART, M., NOUCHET, N., PERES, F., PEETERS, V., RIMET, F., RUMEAU, A., SABATER, S., STRAUB, F., TORRISI, M., TUDESQUE, L., VAN DE VIJVER, B., VI- DAL, H., VIZINET, J., ZYDEK, N., 2002: Determination of the biological diatom index (IBD NF T 90–354): results of an intercomparison exercise. Journal of Applied Phycol- ogy 14, 27–39. REICHARDT, E., 1999: Zur Revision der Gattung Gomphonema. Iconographia Diatomologi- ca 8. A.R.G. Gantner Verlag K.G., Ruggel. ROMERO, O., JAHN, R., 2013: Typifi cation of Cocconeis lineata and Cocconeis euglypta (Bacillariophyta). Diatom Research 28, 175–184. SCHAUMBURG, J., SCHRANZ, C., MEILINGER, P., STELZER, D., VOGEL, A., 2011: Bewertung von Seen mit Makrophyten & Phytobenthos gemäß EG-WRRL – Anpassung des Verfahrens aufgrund erster Ergebnisse und Erfahrungen aus den Bundesländern Endbericht April 2011. Bayerisches Landesamt für Umwelt. SCHAUMBURG, J., SCHRANZ, C., STELZER, D., HOFMANN, G., 2007: Action instructions for the ecological evaluation of lakes for implementation of the EUWater Framework Direc- tive: Makrophytes and Phytobenthos. Bavarian Environment Agency. SCHAUMBURG, J., SCHRANZ, C., STELZER, D., HOFMANN, G., GUTOWSKI, A., FOERSTER, J., 2006: Instruction protocol for the ecological assessment of running waters for implementation of the EC Water Framework Directive: Macrophytes and Phytobenthos. Bavarian Envi- ronment Agency. SMOL, J.P., STOERMER, E. F., 2010: The diatoms: applications for the environemental and earth sciences. Cambridge University Press, Cambridge. WERNER, P., DRESSLER, M., 2007: Assessment of the ecological status of eight lakes from northern Germany according to the Water Framework Directive (WFD) using benthic diatoms: problems and achievements of the newest German WFD guideline. In: KUS- BER, W.-H., JAHN, R., (eds.): Proceedings of the 1st Central European Diatom Meeting 2007. Botanic Garden and Botanical Museum Berlin-Dahlem, Freie Universität Berlin, Berlin. WITKOWSKI, A., LANGE-BERTALOT, H., METZELTIN, D., 2000: Diatoma fl ora of marine coasts. Iconographia Diatomologica 7. Koeltz Scientifi c Books, Königstein. WOLF, M., SCHEFFLER, W., NICKLISCH, A. 2002: Stephanodiscus neoastraea and Stephano- discus heterostylus (Bacillariophyta) are one and the same species. Diatom Research 17, 445–451.