Distribution production and role of aqua

Freshwat Biol 1971, Volume 1, pages 3-21 Distribution, production and role of aquatic macrophytes in a southern Michigan marl lake PETER H RECH ROBERT G WETZEL on*^ NGUYEN VAN THUY W K Kellogg Biological Station, Michigan Manuscript received 19 September 1970 Summary A typical marl lake of the Upper Great Lakes region has very few quantitatively important aquatic macrophytes The macrophytes, however, dominate the total primary production of the lake Submersed vegetation is extremely sparse on the shallow (less than I m) marl bench that characterizes the littoral of these lakes, and is completely dominated by one little-known species (Scirpus subterminalis Torr.) between and m A detailed investigation of the spatial and seasonal distribution of macrophytic species and biomass showed that S subterminalis strongly dominated the lake (79% of total biomass) S suhterminalis represented an almost pure stand (to 200 g m~^ mean annual ash-free dry weight) at all times of the year at intermediate depths of macrophytic growth (1-6 m) Two species of Chara (of eight varieties and forms) were present in significant quantities (12% of total biomass; to 100 g m"^) but were severely limited to shallow depths (0-S-l m) and protected areas Several annual submersed angiosperms were present (9% of total biomass), but only two species were quantitatively important Potamogeton illinoensis Morong and P praelongus Wulfen formed brief summer peaks (less than 100 g m~2) at and 4-6 m, respectively A striking feature of the seasonal biomass distribution of Scirpus subterminalis was the higher, viable biomass (to l g m "-) throughout the winter under ice cover Cyclic fluctuations of the S subterminalis populations were discerned at different depths, each with different periodicities The population at m exhibited a fall peak; that at m had a summer maximum The lowest overall biomass of 5" subterminalis occurred in the m population in June Chara populations at 0-2 m also exhibited a relatively constant biomass throughout the year The appearance of Nitella at m in July-October and of Chara at m in September-October was interpreted as an interaction between light, thermal, and carbon stratification Estimates of macrophytic productivity of perennial ('evergreen') species populations whose biomass remains relatively constant throughout the year were made employing several different methods of calculation and turnover factors All methods Correspondence: Dr P Rich, Biology Department, Brookhaven National Laboratory, Upton, New York 11973, U.S.A 4 Peter H Rich, Robert G Wetzel and Nguyen van Thuy resulted in productivity estimates in good agreement with the conservative value of 178 g m year-i for the entire lake In comparison to the other components (phytoplanktonic, epiphytic and epipelic algae) of the primary production of Lawrence Lake, the aquatic macrophytes constituted a major portion (anuual mean 82-77 g C m-2 year-i or 48-3 %) of the total production of the lake The low diversity but relatively high quantitative importance of macrophytes in marl lakes is attributed to an adverse dissolved inorganic and organic chemical milieu which inhibits phytoplanktonic production and allows only certain adapted macrophytes to develop strongly The phenomenon of perennial biomass levels throughout the year is believed to be much more common than previously suspected and has iikely resulted from adaptations of submersed macrophytes to ameliorated conditions of water and temperatures relative to the terrestrial situation in winter Introduction Marl lakes are common in the Upper Great Lakes region and occur with moderate frequency in the southern peninsula of Michigan and northern Indiana These lakes are characterized by alkaline, hard water and marked deposits of carbonates both in the sediments and on all substrates Conspicuous in marl lakes is a generally low diversity of macrophytic vegetation; most of the production is by a few species that are adapted to this rather rigorous chemical milieu The rates of phytoplanktonic production of marl lakes are moderate lo very low as a result of a number of nutritional interactions (Wetzel, 1965, 1966a, b, 1968, 1969, 1970) The highly buffered water of the trophogenic zone particularly imposes a limited availability of free carbon dioxide, phosphorous, iron and manganese which prevents high sustained growth rates These and other factors result in low rates of bacterial metabolism and cycling of dissolved organic substrates The interactions are further complicated by strong adsorption of labile organic substrates to particulate carbonates and partly permanent losses of these compounds to the sediments (Wetzel, 1970; Wetzel & Allen, 1971) The inactivation of many organic compounds reduces the effectiveness of known complexing mechanisms by which certain ions, especially iron, are maintained in a physiologically available form for photosynthesis Hence, a number of dynamic inorganic and organic interactions are functioning simultaneously in marl lakes to suppress potential rates of phytoplanktonic photosynthesis in a cyclic causal system Most of these nutritional interactions have also been demonstrated to be effective in suppressing high sustained photosynthetic growth by Najasflexilis (Willd.) Rostk & Schmidt, a common submerged angiosperm of marl lakes (Wetzel & McGregor, 1968; Wetzel, 1969) Rates of photosynthetic carbon fixation and extracellular loss of organic compounds are intimately related to the concentrations of major cations, especially Ca^+, Mg+ ^^ and Na+ An intensive investigation of a typical marl lake, Lawrence Lake, Barry County, was initiated in 1967 The detailed treatment of the dynamics of the phytoplanktonic, epiphytic algal, benthic and bacterial metabolism will be reported elsewhere (Allen, 1969; Rich, 1970a; Miller, 1970; Wetzel et al., in prep.) The following remarks summarize the distribution and production of the aquatic macrophytes of this lake Lawrence Lake Lawrence Lake is located in south-western Michigan near the southern boundary of Macrophytes in a marl lake Barry County among undulating plains and arcs of morainic highlands The lake basin lies in an outwash apron along the southern border of the Kalamazoo morainic system of the Saginaw Lobe (Leverett & Taylor, 1915) The soil surrounding the basin is largely sandy Fox loam of medium fertility that is characterized by excessive drainage (Deeter & Trull 1928) Lawrence Lake (4-9 ha) receives drainage from two small streams and several intermittent, vernal springs along the shoreline (Fig 1) The conformation of the littoral zone is typical of marl lakes and characterized by an extensive marl bench The marl bench extends from the periphery of the lake for as much as 20 m to a depth of m Excavations of marl for commercial purposes have been extensive in Lawrence Lake and are discussed in an historical account of the lake (Rich, 1970b) Marl dredging has added 5250 m^ (10-6%) to the area and 16,450 m^ (5-6%) to volume of the lake Details of the morphometry of the basin are given in Wetzel et al, (in prep.) Where undisturbed, the broad marl bench is nearly barren of aquatic macrophytes to a depth of m as a consequence of wave action At this depth the conformation slopes precipitously downward and supports dense stands of macrophytes to a depth of 5-6 m The importance of macrophytes is relatively small because of the extreme slope to those depths where insufficient light is available to support growth The sediments below m are progressively finer, more organic and less dense than those at more shallow depths Species distribution of macrophytes The qualitative distribution of macrophytes of Lawrence Lake was determined from LAWFiENCE LAKE S E C ZT TIH.R9W COUNTY MICHIGAN Fig I Morphometric map of Lawrence Lake, Barry County Michigan, showing transects used for species distribution (Ti-Tio) and production transects (Bi, organic mat shoreline; Ba, isolated from shoreline by dredged zone; B3, typical wave-swept shoreline) Peter H Rich, Robert G Wetzel and Nguyen van Thuy minimally four samples taken at m intervals along transects Ti through Tio and Bi through B3 (Fig 1) Qualitative samples were collected with an Ekman dredge and various grapples (larger species) along Ti-Tio throughout the growing season and with the large corer discussed below in association with the quantitative estimates of production Identifications of angiosperms were determined with Muenscher (1944), Fernald (1950), Gleason (1952) and Fassett (1957); nomenclature follows Fernald (1950) and Fassett (1957) Identifications and nomenclature ofthe Characeae follow Wood &Imahori(1965) As discussed in detail below, the macrophytic biomass of Lawrence Lake is completely dominated by Scirpus subterminalis Torr and several species of Chara Numerous other species occur in the littoral; all are very poorly developed as is characteristic of marl lakes All submersed portions of the macrophytes are covered with heavy but variable quantities of precipitated monocarbonates (cf Wetzel, 1960) The floating-leaf macrophytes are sparsely represented, primarily by Nymphaea odorata Ait in small patches on the southern shore especially near the outlet (Fig 2) fieXI Us Nuphar variegatum Brosenia Schreberi Lemnp minor Nymphaea odorafa fe Species distribution of Najas flexilis and several fioating-leaf macrophytes in Lawrence Lake, 1969 Macrophytes in a tnarl lake Small colonies were also found on the western shoreline and al the southern minor inlet Nitphar xariegatutn Engelm was similarly distributed but, as the white water lily, is only weakly developed Only a few specimens of the water-shield, Brasenia Schreberi Gmel were found among the Nuphar on the northern shoreline Lenina minor L was the only duckweed found in Lawrence Lake and the small colonies were always associated with the primary inlet f i! subterminalis { • • •) S acufus I V I americanus validus Fig Distribution of the four species of Scirpus in Lawrence Lake, 1969 The emergent aquatic macrophytes of Lawrence Lake were limited to three species of the bulrush Scirpus (Fig 3) The hard-stem bulrush S acutus Muhl occurred in moderate abundance in narrow bands on the undisturbed marl benches near the shoreline at a water depth usually less than 0-5 m A small colony of the three-square bulrush, S americanus Pers., was found only on the western shoreline at a depth of less than 20 cm A small patch of the soft-stem bulrush, S validus Vahl, occurred at the mouth of the southernmost inlet The submersed macrophytes of Lawrence Lake are completely dominated by the water bulrush or swaying rush, S subterminalis Torr (Fig 3) subterminalis was found to be generally sparsely distributed on the marl benches to a depth of m Peter H Rich, Robert G Wetzel and Nguyen van Thuy over the entire circumference ofthe lake At greater depths, particularly between I and m, luxuriant stands cover the precipitous gradients to a depth of m, below which no macrophytes occurred Only in a few isolated patches, on the western and northern marl benches at a depth of less than 50 cm, was S subterminalis observed to undergo a brief reproductive cycle with aerial involucres All submersed development is by extensive vegetative propagation Complete quantitative dominance of the macrophytic flora of a lake by S subterminalis has not been cited in the Great Lakes region (E G Voss, personal communication) Extensive stands have been recorded in the littoriai zone of several lakes in southern Michigan (Hanes, 1947; Hanes & Hanes, 1947) It is likely that deep-water occurrence of this species is more common than is generally suspected Najasftexilis (Willd.) Rostk & Schmidt, the slender naiad, was widely distributed in variegated fashion over the entire littoral zone (Fig 2) The growth of Najas in Lawrence Lake is moderate to poor as a result of the numerous interacting factors discussed above p illinoensis P P nafans P pect'natus P gramineus var grammfolius P foliQSus var maceHus amplifolius P praelongus Fig Distribution of the Potamogeton species of Lawrence Lake, 1969 Macrophytes in a marl lake The pondweeds of Lawrence Lake are well represented qualitatively by seven species of Potamogeton (Fig 4) Quantitatively the species oi Potamogeton contribute a minor portion to the total macrophytic production Isolated patches of F gramitieus var gratnitiifoUtis f longipedunculatus (Merat) House occurred in very shallow water along nearly the entire perimeter of the lake Small population aggregations of P illinoensis Morong were found between depths of I and m interspersed with P amplifolius Tuckerm in a narrow band between and m of depth A small development of P natans L occurred only in the vicinity ofthe mouth ofthe southern inlet P praelongus Wulfen was found only in two isolated stands in the south-east corner of the basin and in the shallow plain near the outlet at 1-2 m P pectinatus L and P.foiiosus Raf var maeellus Fern, occurred in very minor isolated patches in the littoral along the western side of the lake Utriculana cornuta U gibba H H I Ilililiiil Ceratophyllym demersum Myriophyllum heteropfiyf/um Fig Distribution of Utriciilaria, Ceratophyllum and Myriophyllum in Lawrence Lake, 1969 The two bladderworts, Utricularia corttuta Michx and V gibba L., are widely distributed in Lawrence Lake (Fig 5) in sparse populations, the former in more shallow water (0-3 m) than the latter (1-4 m) The water-milfoil, Myriophyllum heterophyllum Miehx., was found in minor quantities between and m along all of the shoreline 10 Peter H Rich, Robert G Wetzel and Nguyen van Thuy except for the wave-swept marl bench of the eastern side Stunted forms of the hornwort, Ceratophyllum demersum L., occurred only between and I m along the western shore between the two inlets The Characeae of Lawrence Lake are well developed from to m and are represented by three species: Chara globularis Thuill., C vulgaris L and Nitella fie.xilis (L) Ag (Fig 6) Several of the detailed varieties and forms of the Characeae as delineated by Wood & Imahori (1965) were differentiated, although it is recognized that many ofthe minor morphological variations may represent dynamic phenotypic responses to environmental variables Chara globularis was moderately developed on the littoral marl benches along the southern and western shores of the lake to a depth of m, although generally between and m C globularis var stachymorpha (Gant.) R.D.W was fairly distinctly iso- i S S ? : ^ CT(Wi7 globularis var stachymorpha Chara vulgar is var imperfecfa f dissolula var vulgaris f crispa var globular IS var incorynexa f, arrundensis var aspera Nitella flexilis f obtusa var vulgaris f excelsa var inconnexa Fig Species distribution of the Characeae of Lawrence Lake, 1969 Macrophytes in a marl lake 11 lated in two populations, between and m depth along a ridge between two artificial depressions in the north-eastern corner of the lake and near the outlet C globularis var globularis f connivens (Salzm ex A Br.) R.D.W was found only in an isolated strip in very shallow water along the northern shoreline and intermixed with C globularis var aspera (Deth ex Willd.) R.D.W C Yulgaris var imperfecta f dissoluta (A Br ex Leonh.) was rather widely distributed in shallow water from the shoreline to a depth of m along the northern side and especially well developed on the small delta of the major inlet Very minor populations occurred at the eastern shoreline in less than 20 cm depth C vulgaris var inconnexa f arrimdensis (Mendes) R.D.W was found in a narrow band between 0-5 and m immediately south of the major inlet Distinct patches of C vulgaris var vulgaris f excelsa (T.F.A.) R.D.W existed within the C vulgaris var inconnexa f arrundensis stand C vulgaris var vulgaris f crispa (Wallm.) R.D.W occurred abundantly along the northern and eastern sides of the lake generally from to m and only occasionally was concomitant with C globularis Nitellaflexilis var.flexilis f obtusa (T.F.A.) R.D.W was the only deep-water form of Characeae found in Lawrence Lake It occurred only in a small, narrow strip along the eastern side between and m Macrophytic production Methods Quantitative macrophyte samples were taken at 1-m depth intervals over three transects (Bi, B2 and B3; Fig I) and also taken at a depth of 0-5 m on the marl bench of each transect A sample consisted of four 40-72 cm"'^ replicates, each of which was washed and sorted separately The material was air-dried, dried at 105"C, combusted at 55O"C and re-weighed; all results are given as g m-^ ash-free dry (organic) weight Fifty-six transects were made between 1968 and 1970 with some concentration of dates in the spring and summer to follow the life-cycle of annuals more closely Each transect was usually sampled at least once in every month of the year Samples were taken by means of a specially constructed free-fall core sampler (Fig 7) The body of the sampler consisted of a 60 cm length of 7-2 cm diameter steel tubing with sharp, triangular teeth along the lower, cutting edge The upper orifice was fitted with a free-working rubber stopper which functioned as a one-way valve Table Variance estimates for the mean annual biomass (organic weight g m-2) of important depths for each transect over the entire year Transect Bi Depth (m) Mean 196-9 117-3 151-2 206-3 184-9 185-5 172-3 61-1 124-1 178-6 4 B3 B2 dredged 91[)"/o confidence interval 78-7 69-5 68-6 S40 £1 77-7 SB 43-9 34-6 27-1 51-5 441 29-5 30-2 28-6 23-6 34-5 12 Peter H Rich, Robert G Wetzel and Nguyen van Thuy Stopper guide Stopper Fig Details of closing mechanism of macrophyte sampler In use, about one-half of the length of the tube penetrated the sediments and took a very precise sample of the generally close-lying benthic vegetation In a few areas where the macrophytes {Potamogeton) were much elongated, the sampler was less efficient as evidenced by larger variance in the samples Frequently in these areas, the sampler would strike the base of a plant and exclude the stem Stems severed in this manner floated to the surface where they were collected and added to the sample Sample variance for Scirpus subterminalis was unusually low for aquatic macrophytes and was attributed to the uniform and compact growth form of the species in Lawrence Lake Sampler efficiency was also good for this species because the leaves remained matted together and close to the bottom where they were easily severed Confidence intervals (90%) and standard errors from the untransformed data for S subterminalis at 2, and m are given in Table Biotnass Biomass cycles for each species at each depth of each transect were plotted from pooled data for 1968-70 and the areas over each month of the annual plots were planimetered to provide monthly biomass values The landward end of transect B2 (B2 dredged) crosses an atypical, dredged zone which is only m deep and these data were not used in the summary annual mean presentations (Figs and 9) The monthly organic biomass values have been summarized in three ways: (1) Annual biomass pattern for the important macrophyte groups as a mean transect summed with respect to depth (Fig 8); (2) the annual biomass pattern of a mean transect including depth distribution (Fig 9); and (3) the distribution of mean annual biomass over depth for each transect, including the atypical area of transect B2 (B2 dredged) separately (Fig 10) Macrophytes in a marl lake 13 Table Mean annual biomass as ash-free dry weight (g"-) of each macrophyte group by transect Transect Scirpus subterminalis Characeae 533-1 (67-7%) 179-3 (22-8%) 75-5 (9-6%) 787-9 (100-0%) 704-0 (86-5%) 39-5 (4-9%) 70-7 (8-7%) 814-1 (100-0%) 594-4 (84-2%) 46-4 (66%) 65-4 (9-3%) 706-3 (100-0%) 277-5 (58-6%) 188-5 (39-8%) (1-5%) 2109 (75-8%) 453-7 (16-3%) 218-9 (7-9%) B2 dredged* Annuals Total 7-3 473-3 (iOO-0%) 2781-6 * Transect B2 dredged is only m deep Water bulrush (S subterminalis) was the dominant component (76 % of all sampled material) of the macrophytic flora at all times of the year (Table 2) Distinct fall and late-winter biomass maxima were evident (Fig 8); however, the species is best described as a perennial ('evergreen') population Two annual maxima were evident and resulted from dissonant peaks at different depths (Fig 9a) The larger fall maximum consisted of a major October peak at m and the terminal stages of a June-August peak at m The late winter maximum was compounded from minor peaks at and m the beginning ofthe summer m peak, and a subsequent minimum created by a marked decrease in the m population in June A fall maximum in the Characeae (16% of total lake-biomass) was evident within a relatively constant seasonal biomass distribution Again, this group is best described as a perennial population (Figs and 9b) The appearance of Nitella at m (transect 600 400 200- M M N Fig Annual biomass (g m-^ ash-free dry (organic) weight) of the important macrophytic groups as a mean transect summed with respect to depth —, Scirpus subterminalis; , Characeae; , annuals 14 Peter H Rieh, Robert G Wetzel and Nguyen van Thuy (a) Fig Annual biomass (g m"^ ash-free dry (organic) weight) of a mean transect with depth distribution, (a) Scirpus sublermirtalis; (b) Characeae () and annual macrophytes ( ) B3 only) in July-October and Chara at m in September-October was interpreted as an interaction between light, thermal and carbon stratification Nitella is known to be limited by dissolved CO2 availability above pH 7-3 (Smith, 1967) During the month of July, the hydrogen ion concentration at m in the deep water column shifted from pH 8-3 to 8-1 as summer stagnation progressed The shift may have been even greater at the m contour where the lake sediments were in contact with the m water stratum The disappearance ofthe Nitella population at the end of October coincided with fall overturn The other species of macrophytes, collectively termed annuals (8% of total macrophytic biomass) were not sufficiently numerous in the samples to provide very reliable estimates of biomass As a whole, these species died in the fall, and their maximum biomass occurred during the commonly accepted growing season, June-^September Two maxima are apparent in Fig and may also be discerned in Fig 9(b) The early Macrophytes in a marl lake 15 deeper peak (4-6 m) was largely the result o^ Potamogeton praelongus Wulfen, and the later, shallower maximum (3 m) largely P illinoensis Morong Differences among transects, excluding the dredged zone of transect B2, were not great (Fig 10) Qualitatively, transect B3 difTered from the others in the presence of Nitella at m Only traces of Nitella were found at B2 and no Characeae were ever found below m at transect Bi Chara was more important in the shallow littoral of transects Bi and the dredged portion of B2, which are protected from wave action, than along the exposed transects B2 and B3 Chara thrived in the presence of Nuphar variegatum at 0-5 m along transect Bi {Nuphar was a minor component and was not included in the biomass or productivity figures and appeared in no other transect.) In all cases, the biomass of Scirpus subterminalis at m was inversely correlated with the amount of Chara present With the exception of transect B2 dredged, the mean annual biomass of Scirpus subterminalis was highest and consistently close to 200 g"^ at m Biomass for this species was also consistently greater at m than at m Field Or* 100 200 Biomass {g m"^) Fig 10 Distribution of mean annual biomass (g m-^ ash-free dry (organic) weight) with depth for each transect, including the atypical area of transect B2 (B2 dredged) separately , Scirpus subtertninalis; , Characeae; , annuals 16 Peter H Rich, Robert G Wetzel and Nguyen van Thuy observations suggest that a spongy, fibrous peat stratum at m may have caused the apparent biomass low by either reducing actual colonization by macrophytes or possibly by reducing the efficiency of the sampler, although the latter is less probable Some quantitative differences existed between transects with respect to S subterminalis and Chara (Fig 10) Transect Bi had more Chara than the other two complete transects The total mean annual biomass for transect B2, which had the smallest amount of Chara and the greatest biomass of Scirpus subterminalis was similar to that of transect Bj Transect B3 had low biomass of Chara and Scirpus subterminalis, and the lowest total mean annual biomass Transect B3 also had the most pronounced m low in S subterminalis, as well as a wave-swept shallow littoral The biomass proportions of transect B2 dredged were very different from the undisturbed transects Chara replaced Scirpus subterminalis at m and represented a much higher proportion of the mean annual biomass Annuals were poorly represented in the transect, but observations at other dredged sites suggested that annual biomass can be very important in disturbed areas Productivity Estimating the productivity of the perennial populations of macrophytes whose biomass levels remained relatively constant throughout the year was problematical Generally, productivity of aquatic macrophytes in the temperate zone has been based upon measurements of biomass increments of emergent and submergent annuals between seed germination and maximum biomass The simple-increment method may be modified to account for the low mortality and losses during the growth period by a turnover rate of 2-20% The method is confounded in the case of perennial populations by lack of clearly defined increments uncomplicated by losses of material persisting from prior growth and by an indeterminate growth period Borutskii (1950) made a detailed investigation of biomass of Elodea canadensis, a perennial population in Lake Beloie A similar observation was made for this species in a marsh of western Lake Erie (Rich, 1966) Based upon estimated losses throughout the year, Borutskii concluded that productivity could be as high as five times maximum biomass for this species Much of the turnover in this case was attributed to damage caused by human interference; a condition not true of Lawrence Lake Summarizing several other instances where biomass persisted from one growing season to the next, Westlake (1965) suggests that annual net production is only 50-80% of maximum biomass in such cases Thus, productivity values in the literature for submerged perennials and perennial populations in the temperature zone range from 0-5 to 5-0 times maximum biomass The high portion of the range is certainly unusual While grazing, damage, and mortality may be negligible or low in the brief period prior to the biomass peak of fast-growing annuals, the assumption is not realistic for perennials maintaining populations throughout a year Further, a turnover factor established for the observed growth of the perennial populations would not account for the annual maintenance of the significant annual minimum biomass In the absence of large storage structures as perennating organs for either Scirpus subterminalis or Chara, maintenance metabolism must be significant at all times of year Scirpus subterminalis, particularly, is a very fragile plant consisting of a rosette of long (35 dm), narrow (1-2 mm), delicate leaves Chara replaced Scirpus subterminalis in the shallow littoral probably because it is more resistant to wave action However, Chara is subject to rapid marl incrustation which makes the plant very brittle Ice move- Macrophytes in a marl lake 17 ments have been observed which must inflict much damage to the Chara at 0-5 m in winter as well On the basis of the above observations, the biomass increment method was discarded in favour of two newer concepts of turnover estimation The better documented method, and probably the one applicable to the Lawrence Lake flora, is the Allen curve technique (Allen, 1951) as modified for Glyceria maxima and other plants by Mathews & Westlake (1969) The method follows from the observation that Glyceria continuously produces cohorts of leaves and stems which go through annual life cycles Those cohorts whose growth periods coincide with spring and summer dominate the annual biomass and productivity, but growth and mortality are experienced by all cohorts at all times The most conservative turnover factor of 1-5 for overall annual net production as a function of maximum biomass was selected and applied to the biomass maximum (summer or winter) within each depth population of Scirpus subterminalis and Chara of each transect Both summer and winter maxima were selected and entered into the calculations as distinct in a second case The calculation is not warranted in that winter biomass peaks occurred at