Accretion of ferromanganese nodules that form pavement in Second Connecticut Lake, New Hampshire

Abstract

We discovered ferromanganese nodules that measure up to 46 cm in diameter with stromatolitic growth patterns in the western central near-shore portion of Second Connecticut Lake, Pittsburg, New Hampshire. They occur as four distinct morphotypes: irregular laminated convex plates that form concentric rings around a nucleating center (“domed-plate”), continuous pustular pavements that extend >10 m² (“pustular”), nonlaminated and nonnucleated masses with pebbles scattered throughout the matrix (“lattice”), and small nonnucleated laminated structures >2 cm (“cup”) in water depths between 5 and 12 m. An estimated growth rate of 26 mm/100 years was determined from oxide deposits on datable debris. X-ray fluorescence showed an elemental composition comparable with that of nodule structures in other lakes. These sedimentary structures, the first seen in any of the six New England states, are the most morphologically distinct and varied lacustrine ferromanganese nodules known so far.

We report morphologically distinct (“domed-plate,” “pustular,” “lattice,” and “cup”) ferromanganese nodules from Second Connecticut Lake (SCL) (lat 45°09′20″N, long 71°10′15″W) in Pittsburg, Coos County, in northern New Hampshire (Fig. 1). Described from both marine and lacustrine environments worldwide, these sedimentary structures are composed largely of iron (Fe) and manganese (Mn) oxide minerals in addition to an entire suite of cations, anions and rare earth elements (1, 2). Freshwater nodules have been documented from numerous lakes in North America, including Lake Michigan, MI (3); Oneida Lake, NY (1); Lake Vermilion, MN (4); Mosque Lake, Ontario, Canada (5); and Ship Harbour, Nova Scotia (6). Such ferromanganese structures have been referred to as nodules, concretions, pavements, oxide crusts, and precipitates. However, direct evidence of microbial activity in and on their surface leads to names such as stromatolites, microbialites, or even oncoids (4, 7). Rod and coccoid bacteria are reported to catalyze the precipitation of Fe and Mn oxides (4, 8,9–10). This microbial precipitation occurs 60 times faster than abiotic processes under comparable environmental conditions (5, 11).

Fig. 1.

Northeastern United States with inset topographic map from SCL indicating the locations of the boat ramp, Idlewilde camp, and ferromanganese nodules.

Lacustrine nodules differ in size (mm to cm) and morphology (mound, flat, and ovoid) (1, 4). Elemental composition also varies between lake systems, within the same lake, and even within an individual nodule in the same lake, particularly with respect to Fe:Mn ratios (2, 4, 6). Elemental concentration may control the nodule morphology (12). Barium (Ba), nickel (Ni), copper (Cu), zinc (Zn), and lead (Pb) elements present in freshwater nodules (4, 13) are usable as catalysts in the oxidation and reduction of both major metals (e.g., Fe or Mn) (8, 9). Moreover, ferromanganese nodules have been economically important to humans since ancient times as sources of Mn oxides and their coprecipitates (Cu, Ni, and Co) (5, 14).

SCL is one of three headwater lakes of the Connecticut River. These lakes are remnants of glacial Lake Coos that once occupied the valley created by the retreat of the Laurentide Ice sheet (15) [≈12,000 ybp (years before present)]. SCL, ≈5 km from the Quebec Canadian provincial border, is 570 m above sea level and covers an area of ≈5.2 km² (Fig. 1). This shallow lake (18.5 m maximum) has an average yearly pH of 6.6 and a Secchi clarity of 2.8 m, and it is classified as mesotrophic (16). The ferromanganese concretions were discovered by S.F.W. during a diving expedition to film an underwater video documentary on a logging camp. The distinctive sediment was noticed in water depths of 5–12 m below the dock of the abandoned Idlewilde camp that was located on SCL from 1870–1915 (17) (Fig. 2). The concretions covered at least 9,000 m² of the lake bottom sediment near the camp, indicated by the crescent on the map based on a subsequent expedition (Fig. 1). The nodule field extent was estimated by underwater videography and global positioning system (GPS) waypoint data (August 2005). Nodules measured 0.5–46 cm in diameter and 0.2–4.0 cm thick.

Fig. 2.

S.F.W. holds nodules collected from a depth of 7 m that were discovered while he was diving in SCL in June 2002.

Results and Discussion

The sedimentary structures rest loosely at the sediment–water interface and form a reef-like arrangement. The most conspicuous and abundant of the morphotypes are the domed-plates that are supplemented by pavements, lattice, and small cup structures (Fig. 3). The full extent of the pavement, at least 10 m², is yet to be measured. The domed-plate nodules form irregular but concentric rings around a pebble, cobble, or shell “nucleus,” range from 10 to 43 cm in diameter and 0.3 to 0.6 cm thick, and are found in situ in both concave up and down positions. (Fig. 3 A and B). The domed-plates often have several satellite plates attached to the central nodule that, in many samples, extend the size to >46 cm in diameter (Fig. 3B).

Fig. 3.

Distinct nodule morphologies. (A) Nucleated domed-plate nodules that show concentric ring patterns at the sediment–water interface on clay- to silt-sized substrate. (B) The same nodule type as that shown in A with satellite nodules attached to the larger nodule. (C) A flat, porous crust (with a pustular texture that is 1–2 cm thick) creates an underwater pavement that covers >10 m² and is at least 15 cm deep. (D) Undersurface of the flat porous crust pavement. Concentric rings surround a central nucleus stone as in the domed-plate nodules. (E) A lattice without a nucleus or concentric ring pattern. Stones are scattered throughout the dense matrix. (F) Cup-shape nodules (<2 cm in diameter) are found only on sand- to medium-pebble-sized loose sediments.

The sedimentary structures that create the underwater pavement have a distinct porous, pustular texture that forms the upper layer of these nodules (Fig. 3C). This crust varies in thickness from 1 to 2 cm and lacks growth rings. From above, the pustular pavement appears as a flat sidewalk, but when an individual section is removed one sees a series of joined nucleated nodules with concentric ring similar to the domed-plate structures but without the dome (Fig. 3D). The fact that nucleated nodules are joined suggests the pavement accreted by fusion of synoptic contemporary nodule deposits growing out from many centers.

Lattice structures that range in length from 10 to 45 cm and in thickness from 0.5 to 6.0 cm lack both the central nucleus and concentric rings. However, when cut perpendicular to the long axis, the cross-section reveals variegated internal laminae (Fig. 3E). Lattice nodules contain a sporadic distribution of particles, granule to coarse pebble in size, scattered throughout their matrix. The name “lattice structure” derives from the voids that occur throughout its matrix.

The small cup morphotypes with concentric ring patterns (0.2 to 2.0 cm in diameter and 0.2 to 0.4 cm thick) lie on a substrate composed of sand- to medium-size pebbles (Fig. 3F). They, like lattice nodules, lack an obvious central nucleus. In common to all morphologies (5–6 m depth) is the periphyton coating on the top of the nodules comprised of algae and aquatic bryophytes, described also from Lake Vermilion (4).

The growth rate of SCL nodules is estimated from a 20-mm section of concentric rings nodule material found growing on a glass bottle discarded after 1930 to be ≈26 mm/100 years. These deposits could have developed in fewer than 1,000 years if a constant growth rate is assumed. Growth rates of freshwater nodules average from 3.0 × 10⁻¹ mm/100 years_{(radiometric isotopic data)} (4) to 150 mm/100 years_(estimated) (6). However, isotopic data (²³⁶Ra) showed that nodule growth is nonlinear with extensive periods of stasis or erosion that last from 500 to 1,000 years (13). But even if we consider irregular nodule growth rates during this 75-year period, the growth rate of ferromanganese structures in SCL (except Mosque Lake, Ontario, Canada) is greater than that previously reported in freshwater lakes in North America (6).

Elemental composition data from the x-ray fluorescence (XRF) (Table 1), confirms that our newly discovered ferromanganese nodules in SCL resemble those in other North American freshwater lakes (6, 13). The Fe:Mn ratio in nodules of similar morphotypes from Lake Vermilion, Ship Harbour Lake, Mosque Lake, and Lake Oneida was compared with those in SCL (Table 2). In the pustular sequences (SCL and Lake Vermilion), a relative decrease was seen in the amount of Fe compared with the reef-type or domed-plate nodules. Canadian nodules from Ship Harbour and Mosque Lake with nearly identical morphologies to those in SCL also show higher Fe:Mn concentration, but an order of magnitude higher. Comparable morphotypes to SCL in Oneida Lake, however, are enriched in Fe relative to Mn. Morphotypes do not linerarly correlate with elemental metal concentrations (Fe:Mn) in any of these nodules.

Table 1.

Concentration of selected elements from SCL nodules

SCL nodule type	Elements
	Fe, wt %	Mn, wt %	Ba, ppm	Cr, ppm	Ni, ppm	Zn, ppm	Pb, ppm
Domed-plate	62.57	14.39	79	48	153	774	20
Pavement (pustular layer only)	11.42	6.52	1,357	189	80	463	16
Pavement (full thickness)	35.59	14.33	4,748	101	118	696	17
Lattice	16.76	4.42	1,705	119	93	309	70

Table 2.

Comparison of Fe and Mn ratios in sedimentary structures from SCL, Lake Vermilion, Ship Harbour Lake, Mosque Lake, and Lake Oneida

Nodule locality (ref.)	Fe:Mn ratio	Nodule description
SCL, NH	4.4	Domed-plate structures
SCL, NH	1.8	Pustular portion of pavement nodules
Lake Vermilion, MN (4)	6.3	Granular segment of reef-type nodules
Lake Vermilion, MN (4)	0.8	Massive (pustular) segments of nodules
Ship Harbour Lake, NS (6)	34.2	Concretions image and description are homologous to SCL's domed-plates
Mosque Lake, ON (6)	15.7	Concretions image and description are homologous to SCL's domed-plates
Lake Oneida, NY (13)	0.5	Average Fe:Mn ratio of 60 nodules

Ratios were measured in SCL samples with XRF, in Lake Vermilion samples with energy dispersive x-ray spectroscopy (EDS), in Ship Harbour Lake and Mosque Lake samples with XRF and microprobe, and in Lake Oneida samples with atomic-absorption spectroscopy (AAS).

Diagenetic models that claim to predict nodule growth rates and morphologies from elemental concentrations (12) are not verified by our data. Why SCL and comparable nodules are morphologically varied with accelerated growth rates relative to other freshwater nodules is unknown. To what extent do the physicochemical properties of the limnological and geological settings, the aqueous details of the environment, and the relative abundance and composition of the microbial community determine size, form, and growth rate (8, 18)? Our discovery makes possible quantitative answers to these questions. Because of the extent, variety, and accessibility of these structures, problems that require integrated microbial-ecology, geological, and geochemical methods of study are amenable to direct investigation. We welcome any information about similar structures.

Materials and Methods

Nodules (n = 20) representative of the four distinct morphotypes were collected by divers in September 2002 and immersed in lake water for 2 days before their delivery to the University of Massachusetts, Amherst, for study. They were photographed, catalogued, and allowed to air dry for several weeks. Samples (≈15 g) from each morphotype were powdered in a standard ball-mill grinder before pellet preparation for analysis by Siemens XRF, SRS 303 Sequential X-Ray Spectrometer (Department of Geosciences, University of Massachusetts, Amherst). Nine major and 17 minor elements were detected, and their relative proportions were measured, some of which are reported in Table 1. Although XRF analysis is limited as a tool for accurate identification in sediments composed of many different mineral elements and the results sensitive to the quality of sample digestion during preparation, the technique sufficed for initial identification. Powdered nodule material of each morphotype was treated with a 10% hydrochloric acid to detect calcium carbonate (CaCO₃). The absence of any reaction shows CaCO₃ to be absent or below the level of detection.

Abbreviations

SCL

Second Connecticut Lake

XRF

x-ray fluorescence.