|
4. SCIENTIFIC IMPORTANCE OF CONTINENTAL PALEOCLIMATE RECORDS |
5.1 Planning and Advance Work
As described in the previous section, this report focuses on lacustrine records among other continental archives because of the geographic and environmental distribution of lakes and their relatively rapid and continuous sedimentation. Independent, multiproxy, climatic signatures in lacustrine sediments can be calibrated to historical and instrumental records.
Length, continuity, and time resolution are clearly each important criteria for good lacustrine records, but in many cases, trade-offs exist among the three. In particular, for a given coring or drilling capability, record length is inversely proportional to sedimentation rate, whereas time resolution is directly proportional to this rate. Fortunately, many lakes contain a variety of sedimentary environments, which in turn have a range of sedimentation rates, and thus record length and resolution.
Lakes and their sediments are sensitive to a wide range of environmental conditions. Hydrological sensitivity is perhaps the most prominent, whereby huge lakes may form or entirely evapor-ate during climatic changes. Other important types of sensitivity also result from factors such as proximity to major climatic or veget-ation boundaries that have moved repeatedly in the past, such as monsoon paths and ecotones). Paleoclimatic sensitivity also results from situations in which past climate changes produced major changes in latitudinally or altitudinally determined vegetation zones. Different types of sensitivity may or may not affect the physical and chemical properties of a given lake. The link between particular climatic proxies and climatic sensitivity of lakes is espec-ially important in designing continental paleoclimatic studies.
The quality of chronology will always be a limit on the value of paleoclimatic records. Uncertainties in dating lake sediments are significant, especially beyond the limits of radiocarbon methods (typically about 40,000 yrs). The scientific community needs to recognize the urgent need to develop or refine methods to improve dating of long continental records. Varves exist at some sites and may provide excellent age control, but most sites require additional approaches. Where tephras are present and mineralogically suitable, isotopic dating and fission-track methods may be useful, either directly on core samples, or by correlation to other dated sections. Magnetostratigraphy may also be useful for establishing age control of long records, but the presence of unconformities are difficult to establish by magnetostratigraphy alone. All reasonable approaches to dating must be considered. This is a critical problem that is a key element for interpretation and comparison of records and for developing new proxies.
The consensus of the working group was that most resources should be used initially to obtain many records of 100-300 m, spanning about two glacial-interglacial cycles, and with broad geographic coverage. This is the largest gap in our knowledge on a global scale. However, at some point serious consider-ation should be given to drilling a few key localities where there is the potential for obtaining unique, long records approaching the time spans of millions of years.
5.1.1 Site Selection Criteria
Site selection partly follows from the primary type of proxy signature being examined. Undisturbed, continuous sections are most common below about 20 m water depth (deeper in large lakes), below the depth of storm and wave reworking. Varved sediments mostly occur where bottom conditions produce anoxia. Some small lakes preserve records of local vegetation and watershed conditions. In others, such as crater lakes or lakes that occupy most of their drainage basin, a significant proportion of sediments is captured directly from the atmosphere, so that a more regional signal is recorded in their sediments. Sediments in large lakes contain signatures of aquatic systems that tend to integrate conditions over larger regions and minimize the effects of local perturbations. Larger and deeper lakes provide the opportunity for imaging with acoustic systems to guarantee continuity and representation of sedimentary environments, but require greater logistical investments. In some cases, paired nearby lakes of different size and character are advantageous for the contrast in information contained in their respective sediments.
Because of the great diversity of lakes, simple site-selection criteria are difficult to define.
However, using the general guidelines discussed above, the following criteria are suggested as the most important for setting prior-ities for lake drilling and coring:
· Potential for developing multiple proxies that can be quantified and that are sensitive to past environmental conditions such as hydrology, vegetation, and climate.
· Location that contributes to a wide geographic coverage of past continental environments, especially across climatic gradients and as part of the PEP transects.
· Potential for long, continuous records with appropriate time resolution (annual to decadal for time stream I and at least century to millennial for time stream II and beyond). Records that can be used for several time scales and resolutions are especially valuable.
· Likelihood of dating the cores by several independent methods.
5.1.2 Planning and Organization
One of the most serious deficiencies of many large coring and drilling projects is a lack of planning and pre-operational organization appropriate to the scale of the project. For large, multi-investigator projects, advance planning should include the following:
· An organizational structure, such as a chief scientist and (or) a steering committee.
· Coring location recommendations based on site surveys and reconnaissance data.
· Agreements with all principal investigators to cover the required analytical and dating work.
· Plans for collaboration with scientific institutions and scientists of host country, if applicable.
· Agreements (at least preliminary) with institutions and (or) contractors who will perform the actual drilling or coring operations.
· Arrangements for core storage and archiving.
· Plans for obtaining all necessary drilling permits, environmental impact statements, waste disposal permits, and the like.
· Preliminary protocols for sampling, analytical work, data management, publication, and authorship of reports.
An advance planning meeting of all the principal investigators in the project is highly recommended. Many of the points listed above can be discussed and implemented in such a meeting.
The size, scope, and type of project will determine exactly how these recommendations apply to individual projects. Obviously, the recommendations can be scaled back for smaller projects, but even for such projects, each of the recommendations should be considered.
5.1.3 Preliminary Site Surveys
Pre-drilling site surveys are another often-neglected part of the planning process. All available survey and core data that relates to the type of sediment, the availability of climatic proxies, and the dating of the sediments should be collected and summarized. Geophysical surveys, especially acoustic reflection profiles are particularly useful for all but the smallest lakes. These surveys require some investment and effort, but modern equipment has made such surveys easier and less expensive. For larger drilling projects, they are a necessary first step to identify promising sites that have appropriate stratigraphic sequences and depositional environments. Surveys are also necessary to show that sites have not been disturbed by slumping, erosion, or faulting, and are not unsuitable due to the presence of gas.
In most lakes, projects should plan one or more transects of cores and modern surface samples from shallow to deeper waters. Where deeper lakes are subject to turbidite underflows, acoustic profiling may often identify deep water rise morphology where uniform pelagic sections are sedimented. Often a composite section can be pieced together to gain longer age access (e.g. coring up a tilted fault block). Field sampling of watershed streams, sediments, and biota, as well as lacustrine surface sediments, are important for defining the modern geochemical systems and calib-rating various paleoclimatic proxies to modern conditions.
5.2 Drilling Technology and Operations
5.2.1 General Considerations
Discussion of operational issues and technology is focused on drilling from lake surfaces, because of the significant advantages that lake sediments offer for calibrating paleoclimate proxy signals from modern depositional environments. In some cases, active depositional sites of eolian and fluvial sediments have similar potential, but drilling at such sites is technologically much easier due to the fact that they are land-based. However, many of the recommendations in this and succeeding sections concerning operations, handling, and sampling of lake-sediment drill cores apply equally well to other types of continental sediments.
Because of the wide variety of types of lakes and character of lacustrine sediments, coring or drilling methods used must be tailored to individual situations. In the following discussion "coring" refers to single-entry, relatively short (<12-15 m) systems, whereas drilling refers to multiple entry systems that recover a series of core segments. Three variables are the primary determinants of the type of coring or drilling system that is appropriate: water depth, penet-ration of sediment required, and type of sediment.
Water depth imposes a variety of technical constraints on coring and drilling. From a technical point of view, the following water-depth classes are of interest:
1) 0 (dry lake bed), water table below surface;
2) 0 (wet lake bed), water table at surface;
3) 0-15 m;
4) 15-100 m;
5) 100-400 m;
6) > 400 m.
An additional consideration that is important for technical reasons discussed later is whether the lake is ice covered during the winter.
Sediment type also affects the type of coring or drilling system used and the depth of sediment penetration that is possible with any given system. Lacustrine sediments more than 15 m below the sediment surface are generally at a depth and degree of consolidation that requires drilling. Percussion types of coring devices or drilling are also commonly required for thick sand or gravel layers, chemically precipitated (indurated) sediments, buried soil horizons, and volcanic ashes.
Additional factors that affect the quality of cores include:
1) stability (heave and horizontal position) of the drilling platform, which is a compromise between three competing factors: stability, portability, and cost;
2) the presence of sand and gravel, tephras, or indurated layers such as paleosols or evaporite layers;
3) remote location, which affects many of the drilling logistics;
4) experience and competence of the drillers;
5) the presence of gas in organic-rich sediments-degassing commonly destroys sedimentary structures, and in extreme situations, may constitute a safety issue (ODP Technical Report 22; see Appendix 1);
6) overlying materials, such as outwash or till, that are difficult to core.
Commercial drillers tend to be oriented towards making holes rather than obtaining cores. The need for high-quality cores distinguishes commercial drilling from scientific drilling. A head drilling engineer who fully understands both the scientific aim of the project as well as the drilling techniques being used is extremely valuable. Scientific drilling operators or contract drillers with experience in scientific drilling are desirable.
In a large-scale drilling project such as those addressed here, the primary technological and operational consideration must be obtaining the drill core in as undisturbed a manner as possible, and preserving it in as close to its original condition as possible. It makes no sense to mount a large-scale effort, only to obtain cores whose quality is compromised. These considerations apply particularly to soft, near-surface sediments and to the methods used to recover deeper drill-core sections.
5.2.2 Coring and Drilling Equipment
 |
 |
 |
 |
|
Sutar box corer |
Large diameter (10 cm) |
Kullenberg-type |
Alpine vibracorer (9m) |
A wide variety of coring devices, used to sample the uppermost sediment section, are available. The most common include:
· Box corers: commonly recover <1 m; used to acquire large volumes of undisturbed surface sediments.
· Freeze corers: commonly recover 0.3-2 m; often acquires undisturbed surface sediments, somewhat more reliably than box corers in water-rich sediments (Pachur et al., 1984; Renberg and Hansson, 1993).
· Push corers, such as Russian peat corer and Livingstone piston corer (Livingstone, 1965; Wright et al., 1965; Wright, 1967, 1980, 1991): recovery highly variable depending on sediment type; sediment ordinarily extruded from core barrel or sampled from exposed surfaces.
· Gravity corers: commonly recover 1-5 m, but in some cases 12 m or more; sediment disturbance varies with design, but may be minimal. Sediment-water interface may be disturbed.
· Piston corers, such as the Kullenberg type (e.g. Kelts et al., 1986). Many systems are capable of obtaining cores as much as 15 m long; new systems approach 30 m. Upper portion of the sediment section is commonly disturbed or lost; disturbance of lower sections is variable. Many systems use a small gravity corer, which may obtain a good surface sample, as a trigger or trip weight. Piston corers may or may not include a liner.
· Mackereth corer (Mackereth, 1958, 1969; Barton and Burden, 1979): commonly 6 m, but as much as 12 m recovery with minimal disturbance; requires relatively shallow water depths (but greater than length of core barrel) and pneumatic pressure. Liner not used, so extrusion required. Minicorer version (ca. 1 m) recovers undisturbed sediment-water interface.
· Percussion corers, such as the Reasoner types (Reasoner, 1993): a variety of systems that hammer the corer into the sediment; sediment disturbance variable. Capable of penetration of sand and thin hard layers.
· Vibracores (vibratory, hydrostriker): recovery as much as 10 m; corer vibrated into sediment; sediment disturbance variable depending on design and sediment type, but may be minimal.
· Selcore: combination system using gravity, hydraulic pushing, and vibration, using pressure difference between the water surface and the lake floor.
 |
 |
 |
 |
|
1 descent |
2 penetration |
3 closure |
4 ascent |
Several of these types of corers, especially Mackereth and vibracores, are capable of being configured to operate from a lake-bottom lander system. Deployment from the lake floor tends to increase reliability and minimize disturbance of the sediment core.
A wide variety of drilling systems are applicable to lacustrine sediments, including systems modeled on those of the Ocean Drilling Program (ODP), commercial drilling systems, and systems designed specifically for lake sediments. It is apparent that drilling and coring systems for continental drilling, especially in lakes, need to be much more flexible than ODP systems. Variations in lithology of continental sediments require multiple approaches, including hammering, percussion, vibration, and rotation. A list of drilling systems focused on lake sediments was compiled (see Appendix 2) and includes information on:
1) drilling and coring systems and techniques;
2) platform design and construction;
3) companies willing to work with scientists in developing specific techniques.
Commercial drilling systems may be useful for some continental drilling projects for paleoclimate research, if proper care and arrangements are made to maximize core recovery and quality. Core loss or disturbance from commercial drilling rigs usually results from failure to use a core liner (split-spoon or extrusion recovery of the cored material) or from vibrational or rotational disturbance from the coring tool itself. The latter type of disturbance is most severe for soft sediments near the top of the drill hole; other coring methods can often be used in combination with drilling to obtain a composite, minimally disturbed drill core. The use of core liners (thin, transparent plastic is most useful) is strongly recommended; a variety of wire-line, hydraulic, and hollow-stem-auger drilling systems can accommodate core liners.
 |
Figure 6b. Diagrams of Kullenberg-type piston core operations, (modified from Dixon and Karig, 1969). |
The primary technological constraints on drilling oper-ations in lakes are related to the weight of the drill string and the stability of the drilling platform. The weight of the drill string is determined by the type of material (steel, aluminum) used, the water depth, and the amount of sediment penetration. This weight affects a variety of technical and logistical factors, including the size of the drill rig, the size of the drilling platform, the type of winch required, and others. For systems that use casing or a drill string, the drill rig must be kept in place over the drill hole. The positioning and stability considerations for such systems are discussed in the next section.
5.2.3 Drilling Platforms and Positioning
A variety of drilling platforms, from large barges to canoes, are available for drilling on lakes. The larger platforms, such as barges or sophisticated rafts, are expensive, but are necessary for large-scale operations. They can hold large drilling rigs as well as recommended support facilities, such as refrigerated containers.
For many lakes, portable, lightweight rafts are desirable because of access and logistical considerations. Catamaran-style construction is the simplest and provides a central working hole, which minimizes tilting problems. Rafts need to be repairable; modular cell systems are helpful. Aluminum is excellent for strength, weight and durability, but not as easily repaired (by welding) in remote locations.
Successful raft designs include:
1) the GeoForschungsZentrum system, which uses modular heavy-duty plastic cells that are bulky but light and strong, easily assembled, and very stable and buoyant; they are of Austrian manufacture and solve many of the problems associated with raft construction, but are quite expensive;
2) an aluminum frame system with inflatable buoyancy used by a South African group; this system is light and small enough to be transported by air but is subject to problems of loss of buoyancy if holed and requires calm weather conditions;
3) the system of aluminum/polystyrene sandwich modules used by the Limnological Research Center at the University of Minnesota, which provides both construction strength and buoyancy.
In relatively small or shallow lakes, sufficient positional stability can be obtained by anchoring a raft or other platform, or by tying it to shore. However, the need to avoid significant heave may require calm weather. Heave is minimized by large platform size and by site locations with small wind fetch. In cold regions, thick winter ice can provide sufficient stability and platform capacity, either directly, or by holding a barge in position.
A major technological barrier exists for large, deep lakes that do not freeze solidly. The necessary platform stability for such lakes can be achieved by dynamic positioning, which typically keeps the platform in place over a transponder on the bottom. Robust, automatic systems such as that used by the ODP drill ship are very expensive and require deep water. To our knowledge, they have not been used in major drilling projects on lakes. However, simpler, manual, fair-weather systems are possible and can be very effective in the hands of a skilled operator.
Accurate site location can be established by using Global Positioning System (GPS) units. Simple, inexpensive receivers can achieve accuracy of less than 100 m, and differential GPS systems (using a second receiver fixed at a known location) can achieve accuracy of less than one meter.
5.2.4 Recovery Procedures
Virtually no drilling method can recover undisturbed cores of soft, near-surface lake sediments. Each lake-drilling project should thus include, for example, box cores, Mackereth minicores, and (or) freeze cores of the uppermost meter or so of sediment to provide material for calibration studies and to allow study of the most recent period of environmental change. Surface or core-top sediments are typically in high demand for such studies, and box cores provide the necessary large volumes of sediment. Depending on the drilling method used and the degree of consolidation of the sediments, the upper 3-10 meters of sediments should be sampled by multiple gravity or piston cores. As noted previously, efforts to recover undisturbed cores of the upper part of the section are relatively inexpensive and are critical to the overall success of a drilling project. Skimping on this aspect does not make sense.
Duplicate drill holes are recommended to cover gaps between core segments (segment depths should be staggered between the two drill holes) and to maximize the amount of mater-ial for analytical and dating studies. In some situations, more than two duplicate holes may be justified.
Several general procedures apply to many different types of drilling methods and serve to maximize core recovery and minimize core disturbance. They are slanted toward cased-hole, wire-line methods, but can be adapted to many other methods, such as hollow-stem augering.
· Use of casing wherever the stability of the drill-hole walls is in question (wherever sediments are less than fully consolidated).
· Coring drive should precede lowering of casing.
· If coring stroke is incomplete, casing drive should be adjusted accordingly.
· Drilling mud pressure should be carefully monitored to minimize disturbance while allowing material generated by the casing lowering to be flushed.
· Liners, preferably transparent plastic, in the coring tool are highly recommended.
· Multiple cores (multiple drill holes) are relatively inexpensive (because much of the drilling cost is associated with transportation and mobilization) and can be used to construct composite sections that solve problems such as coring gaps and gas expansion.
Complete and accurate drilling records should be kept for all aspects of the core recovery operation. These records should be a combined effort on the part of the on-duty drilling engineer and the scientist, and exact procedures should be agreed upon in advance. A variety of processes, including hole re-entry, incomplete recovery, compaction, and gas expansion commonly lead to uncertainties about the exact depths of recovered core sections. Uncertainty about which end of the core is the top is not unprecedented. Such depth uncertainties are critical to later correlation and time series analyses and need to be kept to an absolute minimum. Any other information related to the condition, treatment, or changes to core sections should be carefully recorded.
5.3 Post-drilling Operations
5.3.1 Transport and Short-term Storage
A number of procedures for storing cores at the drill site and transporting them to a permanent facility are recommended. The goal of these procedures is to keep the sediments as close to their original condition as possible:
· Cores should be stored in a dark, cool place (constant 4°C is recommended); this is strongly advised if cores are sealed in liners and essential if core materials are extruded.
· Refrigerated containers are recommended, if possible, for on-site storage and for transportation because of their availability, price, and portability.
· Strong magnetic fields and magnetic materials should be avoided during short-term storage and transportation in order to preserve the natural magnetic remanence and mineral properties of the sediment.
· If core sediments fill their liner or container, which is securely plugged and taped, they can be transported in any orientation. Water or cap space in the liner or container can be filled with styrofoam or similar material. If cores contain water, air, or gas, they should be transported upright and with extreme care.
· Because of likely subsequent disturbance during transport, cores should not be split in the field if it can be avoided. If cores must be examined in the field, local laboratories for permanent storage should be investigated. Some examination of samples in the field can be done on core-catcher or other disturbed material.
5.3.2 Storage and Archiving of Cores
Ideally, cores should be kept sealed in a dark, moist, and cool (4°C) location that is temperature and humidity-controlled. If the cores are in liners or similar containers, 1.0-1.5 m sections are convenient and can be sealed in plastic wrap or heat-sealed polyethelene tubing. Core sections can then be packed in more rigid and stable "D-tubes," along with saturated florists foam or other water-absorbent material. A computer data base of stored and archived cores is essential.
These procedures are designed to preserve the cores as close to their original condition as possible. In long-term storage, the largest disturbances come from drying and warming. Drying commonly destroys sedimentary structures, changes oxidation state, and modifies some mineralogy. Warming can lead to a variety of geochemical changes, dissolution or precipitation of carbonate materials, and microbial or fungal growth. Light also promotes algal and other growth.
5.3.3 Down-hole, Post-drilling Measurements
These procedures are generally costly and should be subject to cost-benefit analysis. They are probably necessary for important sites with long continuous records. Special large well-logging contract companies (e.g. Schlumberger) can probably only be considered for the deepest holes. For shallower holes (<500m) there are many smaller well logging companies which can provide relatively inexpensive services, even in deep-water situations. Down-hole logging can provide precise depth references of core segments.
Gamma logs are very useful as indicators of porosity and bulk density. Magnetometer, resistivity, and other geophysical logs of the drill hole also may be useful. Vertical, down-hole seismic profiles can be valuable, but they are expensive and difficult to perform under water. Down-hole video can also be very useful. Several other measurements can provide specific types of information that may be useful for paleoclimate studies (see ODP Technical Reports, Appendix 1).
5.4 Drilling Technology for Remote Areas
The sedimentary record of many of the lakes in the world are virtually unknown because these lakes are in remote areas where logistics and transportation are difficult. Examples of these areas include Siberia, tropical West Africa, and southern South America. Because lakes in many of these areas are the main source of paleoclimate records, they are especially important for implem-enting a global grid of paleoclimate sites. These sites are needed to establish regional patterns of climate changes, to document watershed responses to those changes, and to test climatic models. Thus, there is a strong need to develop drilling technology, particularly light weight drilling rigs, that can be transported and operated in remote areas. There is also a need for funding agencies to consider the additional transportation and logistic costs of projects in remote areas the same way they consider ship operating expenses in marine projects: as required components of scientifically valuable research.
5.5 Sampling and Analytical Protocols
Sampling and analysis of core materials is one of the most visible stages of drilling projects for paleoclimatic records, yet it is one in which problems commonly develop. The Ocean Drilling Program has developed elaborate procedures for these activities, and although they may not be directly applicable in all cases, they should be seriously considered for adoption or adaptation in continental drilling projects.
5.5.1 General Considerations
Most of the issues related to sampling and analyses, such as methods, priorities, sampling density, individual responsibilities, should be decided in advance of the drilling. Sample density will be partly determined by sedimentation rate and time resolution, and to the degree possible, these should be determined during preliminary site surveys. A reference list of sample priority should be explicitly developed and agreed upon.
So called "sampling parties," where all principle investigators are present, are recommended. Based on the results of all pre-splitting logging, imaging, and analyses, a core description and sampling strategy can be agreed upon. This process can be lengthy and must be well planned with equipment backups to minimize wasted time. However, once the basic strategy is established, sampling parties have proven very useful for discussion and consistency, and they enhance productivity and insight into the record through informal discussions among principal investigators. Although not essential, the use of a single, well equipped laboratory for initial core descriptions and measurements has considerable advantages. The use of such laboratories also facilitates sampling parties, which then can be held at the same location as all of the pre-sampling analyses.
A single material or sample coordinator for each project should be appointed at the proposal stage. The coordinator is responsible for resolving sampling conflicts, recording sampling activities, maintaining a priority list of samples for different scientific questions decided by the whole team, and for organizing sampling session(s). A single depth scale for a core is critical; it may seem like an obvious goal, but a variety of problems can lead to discrepancies in sample depths. Establishing and maintaining this depth scale should be the responsibility of the sample coordinator.
Standard analytical methods are encouraged, and standard reference samples for each analysis are important for evaluating the success of each analysis. Reference samples are absolutely mandatory when more than one investigator or laboratory performs the same analysis on a subset of samples.
Initial sampling should be planned at the highest level of resolution anticipated because of difficulties in resampling soft sediment, additional disturbance related to accessing cores multiple times, and other logistical problems. However, sampling may be divided into stages, such as primary and secondary stages, or reconnaissance and detailed stages, where appropriate. This procedure, along with secondary or duplicate cores, enables sequential interrogation of the record. However, definite schedules for multistage sampling should be agreed upon in advance to prevent disruption of the overall project schedule by delays in the results for individual analyses. Samples for different proxies should be taken at the same level, and the use of the same sample material in multiple analyses should be used where practical in order to preserve core material (see discussion of archive material).
Sampling priorities should be driven mostly by the scientific questions proposed by the project members as a group. For example, if river catchment history is a primary goal of the project, then the following analyses might have the highest priority: detrital-authigenic-biogenic ratios, X-ray diffraction mineralogy, organic carbon and palynofacies, sediment magnetic properties, and elemental analyses. For other project goals, other analyses might have higher priority. At the same time, some samples or measurements need to be taken or made immediately, before they are degraded or compromised. Examples include thin sections; photography; samples affected by oxidation processes (color, organic geochemistry, sulfur geochemistry, magnetic properties, and others), pore water chemistry, radiocarbon dating, and TL/OSL dating.
5.5.2 Individual Methods and Procedures
The following summary of methods is a skeletal guide to analysis of lacustrine sediments for paleoclimatic reconstruction. A detailed discussion of these methods is beyond the scope of this report, but the next section contains notes appropriate to the context of this report for each method, along with basic references that serve as a starting points for additional information. A wealth of other information about most of these methods is contained in the sources listed in Appendix 1.
Table 1: Recommended sequence and summary of procedures and analyses :
Column 2: Type
L, Logging; D, Description; A, Analytical; Dt, Dating.
Column 3: Required condition
Ideally, the sample requirements for each technique should be defined in terms of sample condition and sample size. Neither consideration is straightforward, and each varies with the technology used, especially sample size. For these reasons, only sample condition is considered here. Required condition assumes that proper pre-cautions have been taken to avoid contamination (e.g. air-born pollen). The following categories are useful, but do not accommodate every possibility.
1: Fresh, undisturbed, unfrozen, oriented (at least up and down).
2: Frozen as soon as possible after coring or sampling.
3: Fresh, undegraded required (r) or desirable (d); not frozen.
4.: Original volume measurement required (r) or desirable (d).
5: Dry mass measurement required (r) or desirable (d).
6: Bulk sample of undetermined original volume or mass acceptable.
7: Sieved residues acceptable.
8: Exposure to light or X-rays avoided.
Column 4: Subsequent Condition
These categories are related primarily to the degree of disturbance or destruction involved in the method. They have a strong bearing on the sequence in which analyses should be performed, and on which samples can be used for multiple analyses.
Whole core
A: Non-destructive, undisturbed. Whole core, preferably in liner, remains intact and undisturbed.
B: Non-destructive; invasive. As in (A), but limited sample or pore-water is extracted, commonly by a needle or probe through the core liner.
Split core
C: Undisturbed half core.
Core samples
D: Undisturbed. Samples remain in fresh condition with sedimentary structures and grain fabrics intact.
E: Fresh. Volumetric samples remain in fresh condition, but internal structures or orientation not preserved.
F: Dried. Samples remain in uncontaminated condition, but have been dried.
G: Fractions. Extracts, digests, or other fractions remain available for limited subsequent analyses.
H: Preparations. Thin sections, smear slides, and similar preparations possibly available for other analyses.
I: Destruction. Samples destroyed in analysis.
| Method or Process |
Type |
Required condition |
Subsequent condition |
| 1. Magnetic susceptibility |
L |
1,5 |
A |
| 2. GRAPE |
L |
1 |
A |
| 3. P-wave velocity, resistivity, other logging |
L |
1 |
A |
| 4. Sensitive geochemistry, (e.g. pore-water Eh-pH, O2; some C and S analyses) |
A |
1 |
B |
| 5. Core splitting |
D |
1 |
C |
| 6. Photography |
D |
1 |
C |
| 7. X-ray imaging |
D |
1 |
C |
| 8. Spectral/gray-scale scan |
D |
1 |
C |
| 9. Thin sections |
D |
1 |
H |
| 10. Smear slides |
D |
6 |
H |
| 11. Visual descriptions |
D |
1 |
C |
| 12. Bulk sampling, including sample splitting and mechanical size separation |
D |
3 |
D |
| 13. Sediment magnetic properties |
A |
3, 4, 5 |
D |
| 14. Bulk density, water content |
A |
4r |
F |
| 15. Organic geochemistry |
A |
2-3 |
I |
| 16. Sulfur geochemistry |
A |
3 |
I |
| 17. Organic Carbon and Nitrogen (± LOI) |
A |
5r |
I |
| 18. Granulometry |
A |
4 |
I |
| 19. Elemental chemical analysis |
A |
4,5r |
G- I |
| 20. X-ray mineralogy |
A |
6 |
I |
| 21. Sand/silt mineralogy |
A |
7 |
H |
| 22. Carbonate conten |
A |
5r |
I |
| 23. Biogenic silica |
A |
6 |
I |
| 24. Stable isotopes (C, H, N, O, S) |
A |
6 |
I |
| 25. Pollen and spores |
A |
3d,4d,5d |
H |
| 26. Diatoms |
A |
4d,5d |
H |
| 27. Macrofossils (plant and animal) |
A |
3d,4d,7 |
G |
| 28. Ostracodes (including geochemistry and isotopes) |
A |
4d,5,7 |
G |
| 29. Radiocarbon |
Dt |
2,3,7 |
I |
| 30. 210Pb and 137Cs (and related isotopes) |
Dt |
4,5 |
F |
| 31. Magnetic remanence (paleomagnetic) measurements |
Dt |
1 |
D-F |
| 32. Varve counting |
Dt |
1 |
H |
| 33. Tephra analysis |
Dt |
4, 5, 6 |
H-I |
| 34. Other dating methods (e.g. Uranium series, TL/OSL/ESR) |
Dt |
4-8 |
F-I |
5.5.3 Notes on Specific Sampling and Analytical Methods
The following material forms a general guideline for logging, sampling, and analytical methods for paleoclimatic reconstructions from cores of lacustrine sediments. These are only general comments, so that consultation of other sources is recommended. These sources include those listed in Appendix 1, as well as review papers and books about individual methods, as cited below. Some of the material, especially that on analytical methods, is adapted from the PALE protocols document (PALE Steering Committee, 1994). Note that analytical methods discussed here do not appear in the same order as they do in Table 1, which lists them in approximately the order in which they should be performed or their samples taken. The list here is rearranged slightly to group major classes of analyses (e.g. physical, chemical, biological) together.
Logging Methods
5.5.3.1 Magnetic Susceptibility
Magnetic susceptibility is a sediment property related to the concentration of magnetite and similar magnetic minerals
(Thompson and Oldfield, 1986; Reynolds and King, 1995; see magnetic properties section below). Whole core, pass-through, volume susceptibility at resolutions of 2-3 cm has become routine and should be measured in almost all cases. It can be conveniently measured at the drilling site with a portable coil and laptop computer. Coil calibration is extremely important. Several methods are available for finer-scale measurements after the core is split, including high-sensitivity devices that pass close to the split face of the core, and devices that measure susceptibility (and other magnetic properties) on very small samples. Although susceptibility is an indication of sediment provenance and style of deposition and provides a valuable core-to-core correlation tool (e.g. Andrews and Jennings 1987; Peck et al., 1994), detailed interpretation of the susceptibility signal requires considerable supporting data from other magnetic properties.
5.5.3.2 GRAPE
GRAPE (gamma-ray attenuation porosity estimator) provides an easy, continuous measure of porosity and bulk density (Boyce, 1976). However, both porosity values and p-wave velo-cities (next section) measured by logging tools need to be calibrated against measurements made by standard methods (e.g. Bulk Density section, below). Use of gamma-ray sources may require radiation safety certification.
5.5.3.3 P-wave (sound) Velocity, Resistivity, and other Logging
P-wave velocity provides another useful measure of bulk physical properties of sediment , which is primarily related to por-osity, particle density, and degree of cementation. P-wave velocity measurements are required to produce synthetic seismograms in seismic-reflection studies (e.g. Moore et al., 1995), which are useful for correlating core data with seismic-reflection data, allowing core-site-specific information to be spatially extrapolated.
5.5.3.4 Sensitive Geochemical Samples
Pore water and dissolved gas (e.g. methane) measurements, as well as some organic and sulfur geochemical analyses, need to be made on samples that have had minimum exposure to air. Pore water can be sampled through the core liner, and Eh-pH measurements can also be measured with instruments that pene-trate the liner. In some cases, this may be the only way to obtain accurate measurements and some of the measurements are best done at the drill site immediately upon recovery of the core. However, these devices may also disturb the sediments or allow oxygen to penetrate the core liner. Decisions about the relative importance of these type of measurements compared to their potential disturbance need to be made in advance.
Description Methods
5.5.3.5 Core Splitting
Cores should be split into two halves using materials and procedures that will minimize disturbance and contamination, especially that related to organic geochemistry and sediment magnetic properties. In general, one half of the core should be designated as the working half, and the other preserved as an archive half. In some cases, preservation of one-fourth of the core or discrete samples as archives may be acceptable. Use of the archive half for non-destructive analyses or for critical samples should be done at the discretion of the project leaders or sampling coordinator.
5.5.3.6 Photography
Basic photography of split cores sediments documents changes of sediment color and texture that are often important indi-cations of changes in the rates and processes of sedimentation in lakes. These changes also aid in correlation between cores from a single lake. The cores should be carefully split and the fresh exposed surfaces photographed with a high quality color film. Each photograph should include a standard color chart so that changes in color can be calibrated along a core. It is also important to photograph the cores soon after they are split, as the colors of the sediments can often change quickly. In addition to detailed photos, overview photos of several core sections are recommended to document large-scale changes.
5.5.3.7 X-ray Images
X-ray images (commonly contact prints at a scale of 1:1) of whole or split core are commonly useful for observing bedding, laminations, and other sedimentary structures, even when the core is visually homogeneous. X-radiography also helps to locate macrofossils (e.g., calcareous fossils and wood) for dating and analysis. X-rays images may not be necessary for obviously laminated cores for which continuous thin sections will be made. Care must be taken not to X-ray samples that might be used for dating methods based on electron traps in crystalline structures (dating methods section, below).
Image analysis of X-radiograph and (or) photographic images can overcome the discontinuous nature of sediment properties derived from samples taken at fixed and discrete intervals. Image analysis can show changes in sediment color, density, and structure at very high resolutions (several pixels/mm). Studies of these "continuous" changes can be carried out by means of an image analysis system using a video camera and appropriate computer software (e.g. Schaaf and Thurow, 1994).
5.5.3.8 Gray-scale Measurements and Multi-spectral Scanning
Various kinds of light-reflection scanning have proven useful for some marine cores (Barranco et al., 1989). Gray-scale profiles can be made from photographs of the core or thin sections. Hand-held sensors that attach to portable computers are also available. Equipment for full-spectrum light scanning is expensive and not commonly available. The importance of these methods for cores from continental environments has not been extensively investigated.
5.5.3.9 Thin Sections
Thin sections are valuable for a variety of lithological studies and are required for some, such as varve counting (see dating methods, below). Thin sections commonly reveal details of sediment structure, such as varves or fine laminations (e.g. Saarnisto, 1986). Most methods include freeze-drying and impregnation of the sediment with resin (e.g. Anderson and Dean, 1988).
5.5.3.10 Smear-slides
Observations made from smear slides are simple and powerful complement to lithological descriptions, and can be performed before most other sampling. Information from smear slides is commonly very useful, and should be made available as soon as possible. Smear-slide information can be used to guide subsequent sampling, making other analyses more efficient.
5.5.3.11 Visual Descriptions
A standard protocol should be adopted for core description, including lithological and sedimentological classifications and procedures. At a minimum, all classification terms must be defined. Munsell colors and Troels-Smith symbols (Troels-Smith, 1955) are recommended. Descriptions should be done by one worker (with sedimentological training) for the entire core in order to maintain consistency. ODP guidelines (Appendix 1) and Berglund (1986) are useful for such description and sedimentological protocols.
5.5.3.12 Bulk Sampling
Several considerations are important for bulk sampling of cores for analytical work. The condition of samples required for each analysis (Table 1) must be considered foremost. In general, core material is never sufficient for all analytical needs, so conservation of material and efficiency should be considered. For example, samples taken in plastic cubes for magnetic properties and remanence analyses are often suitable for a variety of other analyses once the magnetic measurements are made. Finally, samples for some analyses should be taken at the same depths in the core as those for other analyses in order to derive the maximum information from the combined analyses. For examples, samples for sediment grain-size and mineralogical analyses should be closely associated with samples taken for magnetic property measurements.
Analytical Methods
5.5.3.13 Sediment Magnetic Properties
Measurements designed to reveal changes in the concent-ration of magnetic minerals, their grain size, and their mineralogic composition (so-called rock- or mineral-magnetic properties) have become a much more frequent and informative aspect of lake sediment studies over the last 20 years. Increasingly, these measurements are being related to climate variables and used in paleoclimatic studies (e.g. Peck et al., 1994; Creer and Morris, 1996). In addition, magnetic properties often provide a strong basis for local to regional core-to-core correlations.
Magnetic property measurements have the advantages of being relatively cheap and simple to perform, largely non-destructive, and readily combined with paleomagnetic studies. Measurements can be carried out on either a constant volume or a mass specific basis. If paleomagnetic measurements are intended, stand-ard cubes should be used and all measurement carried out on a constant volume of fresh sediment. Paleomagnetic measurements can be carried out after low field susceptibility has been measured, but must precede all other rock magnetic measurements. Both rock- and paleomagnetic properties can degrade with long storage so it is best to carry out both as early as possible after subsampling.
Since the early work summarized in Thompson and Oldfield (1986), interpretative models of changes in magnetic properties in lake sediments have become increasingly complex (King and Channel, 1991; Reynolds and King, 1995). This has been in response to the realization that many factors other than simple sediment-source linkages can contribute to the properties measured under the broad heading of 'environmental magnetism'. These include particle sorting, dissolution diagenesis, authigenic sulphide formation, and biosynthesis of magnetite either as magnetosomes or as an extracellular product. In many cases, this rather complex range of environmental influences enriches the information to be gained from magnetic measurements, though it reinforces the need to avoid simplistic interpretations and to view the magnetic property record within the context of the full range of proxy measurements completed.
The range of magnetic measurements commonly carried out on sediment cores includes low field susceptibility, often at two frequencies, followed by a suite of remanence measurements that require equipment for magnetizing the sample as well as for measuring the remanence acquired. Qualitative and sometimes partially quantitative insights into the changing magnetic grain size and mineralogical components of a suite of samples can be gained in this way. An alternative or additional approach is to derive a series of parameters from direct magnetic hysteresis-loop measurements. Beyond these fairly standard properties are a range of often more time-consuming experiments designed to characterize more fully a smaller number of samples.
5.5.3.14 Bulk density, Water Content
Calculations of fluxes for any sediment constituent require measurements of sediment density. Samples should be carefully taken at appropriate intervals using standard volumes (e.g., syr-inges or paleomagnetic cubes). Wet volume density can be determined by simply weighing the sample container, subtracting the weight of the container, and dividing by the volume. Samples can then be slowly air dried (at up to about 60 °C) and re-weighed. This measurement allows the dry volume density to be calculated (g/cc) and, with comparison to the wet volume density, allows the weight percentage of water in the original sample to be calculated. An alternative method is to measure particle density. (e.g. by pycnometer) and to determine water content without using known volumes.
5.5.3.15 Granulometry
Sediment textures and grain-size distributions should be described in standard sedimentological terminology and size classes. Details of grain-size determinations are given in many texts (e.g., Lewis, 1984), and applications to lake sediments are discussed in Berglund (1986). In some cases, texture and grain size can be interpreted directly in terms of climatic variables, such as wind speed (e.g. Halfman and Johnson, 1984). The sand fraction is usually determined by shaking through a series of nested sieves. Material finer than 0.063 mm is commonly determined by use of a Sedigraph or by automatic particle counters based on light or resistivity. It can also be determined by pipette methods (e.g. Syvitski, 1991), which has the advantage of retaining size fractions for further analysis.
5.5.3.16 Elemental Chemical Analysis
Elemental composition has been used to identify distinct layers such as volcanic ash. In addition, variations in elemental composition of autochthonous and allochthonous sediment fractions have been used to document soil erosion and development, vegetation change, and limnological conditions (e.g., Mackereth 1966; Engstrom and Wright, 1984; MacDonald et al. 1993). Such changes may be a direct result of climatic changes and (or) in-direct results of changes in watershed conditions. Major-element geochemistry, especially that related to authigenic minerals, may reveal a variety of important information about the evolution of chemical processes in a lake (e.g. Kelts and Talbot, 1990; Dean, 1993).
Studies of trace metals in lake sediments mostly have focused on understanding of anthropogenic contaminants (e.g. Lottermoser et al., 1993; Renberg et al., 1994). Potential paleoclimatic signatures related to trace metals have not been established, although trace metals are of chronostratigraphic value because the timing of global dispersal of certain anthropogenic pollutants is well-documented in ice cores and lake sediments (e.g. Hong et al., 1994; Lottermoser et al., 1993)
Geochemical methods are diverse; good starting points for lake sediments include Engstrom and Wright (1984); Bengtsson and Enell (1986). Recommended analytical procedures range widely, depending greatly on the availability of hardware. Useful studies can be carried out using simple titration and petrographic microscopic analyses. Well-equipped laboratories would bring modern analytical approaches, ranging from atomic absorption (AA) spectrophotometry, neutron activation analysis (NAA), inductively-coupled plasma mass spectrometry (ICP-MS), X-ray fluor-escence (XRF), scanning electron microscopy with energy dispersive spectroscopy (SEM-EDS), and others to bear on the problem of sediment composition.
5.5.3.17 X-ray Mineralogy
Mineralogy of the clay and fine silt size fraction is useful in many paleoenvironmental studies. For example, it can be used to distinguish between sediments derived from within the basin versus those imported from extra-regional sources by wind. At somewhat lower resolution, the suite of clay minerals transported to lake basins from their watershed are responsive both to climatic variables such as temperature and precipitation, and to climatically mediated watershed processes such as soil erosion (Jones and Bowser, 1978). Mineralogy of the carbonate fractions is an important complement to information derived from carbonate-content and major-element analyses (Kelts and Talbot, 1990; Dean, 1993).
Analyses are most commonly carried out by X-ray diffract-ion methods (Lewis, 1984). Details of the quantification of X-ray diffraction curves for lake and marine sediments vary (Andrews et al., 1989; McManus 1991).
5.5.3.18 Sand/Silt Mineralogy
The use of the mineralogy of the sand and coarse silt fract-ion is similar to that for clay minerals, except that the course fract-ion may reveal information about processes that are sensitive to particle size and depositional energy, such as eolian deposition or shallow-water transport (lake-level information). Methods are simil-ar to those for clay mineralogy, except that the coarse fraction is usually ground to smaller grain size before mineralogic analysis.
5.5.3.19 Organic Carbon and Nitrogen (± LOI)
The amount of organic carbon and nitrogen in lake sediments is a fundamental property that is a function of autochthonous and allochthonous organic production, bacterial decay, and the rate of clastic sediment input. As measures of lake productivity, organic carbon and nitrogen are critical for paleoecological studies. The ratio of carbon to nitrogen is commonly an indication of the source of the organic matter (Stein, 1991).
Although the determination of the organic carbon content of the sediment is an important parameter to measure, there is no single standard method that has been applied to this measurement. Loss of weight on ignition (LOI) (Bengtsson and Enell, 1986) is the simplest method, although constituents other than carbon may contribute to weight loss. Wet-chemical methods are also used for both carbon and nitrogen (Hedges and Stern, 1984); carbon contents from LOI and wet-chemical approaches have a significant correlation when paired samples are analyzed. Organic carbon and nitrogen can also be determined by automated carbon analyzer (e.g. Weliky, et al. 1983) and by carbon-hydrogen-nitrogen (CHN) analyzer (e.g. Hedges and Stern, 1984). See Organic Geochemistry section for more information.
5.5.3.20 Organic Geochemistry
Organic matter (OM) that is preserved in lake sediments encompasses a range of compounds, from the simple such as methane, to complex biopolymeric molecules such as lignins and nucleic acids. In a lacustrine environment, OM is derived from primary production within the water column and also from terrest-rial biota by transport of leached and eroded material into the lake. The accumulation and composition of OM in the sediments is usually influenced by the environmental conditions: climate, geology of surrounding rocks, physical and chemical characteristics of the lake waters, and the nature of the OM itself (Meyers and Ishiwatari, 1993; Killops and Killops, 1993). Hence, variations in source materials and conditions can effect the composition of the organic compounds in sediments, and these can potentially be used as a chemical record of past change.
Many of the organic chemical constituents of fresh OM are labile, and may be transformed or completely remineralized within the water column and in surficial sediments. In a similar way to conventional fossil record, therefore, only resistant molecules tend to survive, and these are most commonly used to deconvolve the history of changing inputs to the lake sediments. Biological markers (or chemical fossils) are particularly useful compounds in this respect. They have a direct chemical link with a biologically occurring precursor. Certain lipids and pigments have a restricted occurrence in only a few organisms or types of organisms; if they or their derivatives can be identified, then an attempt to discern original inputs to the sediments can be made. For example, biological markers can clearly discern autochthonous and allochthonous contributions to lake sediments (Meyers and Ishiwatari, 1993), and have the potential to be used in quantitat-ive assessments of changing OM inputs (e.g. Prahl et al., 1994). These biological markers include fossil pigments (chlorophylls and their derivatives, especially carotenoids; Sanger, 1988; Leavitt, 1993) and lignin oxidation products (e.g. Orem et al., 1996). Fossil pigments are also useful for reconstructing the history of productivity, meromixes, and trophic status (Brown et al., 1984; Züllig, 1989; Guilizzoni and Lami, 1992).
The fixation of carbon by photosynthesis, by both land and aquatic plants leads to isotopic fractionation in favor of the lighter isotope of carbon 12C (more information in Stable Isotope section). This fractionation of approximately 20 parts per 1000 (20) relative to CO2 has important consequences. Again, changing environmental conditions can lead to the changing isotopic signature of OM; this can lead to isotopic shifts in the composition of sedimentary OM, for example as a result of increasing aquatic productivity (Schelske and Hodell, 1995). More interestingly, a combination of the molecular and isotopic approach can allow distinction of OM at the species level. Although this aspect of organic geochemistry is in its infancy, the potential is tremendous (Rieley et al., 1991).
The determination of molecular parameters is a rather time consuming process. Careful handling is necessary to avoid contamination and some materials need extra care (i.e. kept froz-en). Analytical methods require the separation of individual compounds by chromatography (gas chromatography, high performance liquid chromatography) and their identification, usually by mass spectrometry. Fortunately, there is a huge wealth of literature on the identification of biological markers (for reviews see Peters and Moldowan; 1993) and Engel and Macko; 1993)).
5.5.3.21 Sulfur Geochemistry
The amount, kind, and isotopic composition of sulfur species in lake sediments can potentially reveal much information about productivity, sulfate availability, surface runoff, and post-depositional alteration (Berner, 1984, Tuttle et al., 1990; Bates et al., 1993, 1995; Vairavamurthy and Schoonen, 1994). Sulfur exists in a variety of oxidation states and its geochemistry is linked with that of carbon. In addition, sulfur diagenesis affects magnetic mineralogy. Although sulfur geochemistry is complex, it has been successfully applied to studies of ancient and modern lakes to help understand depositional and diagenetic processes related to changes in climate, productivity, and water chemistry. Sulfur geochemical methods, however, have not been widely used for paleoclimate studies, and their potential has not been fully tapped. For this reason, it is relatively undeveloped as a paleoenvironmental proxy, despite its great potential. Nevertheless, large systematic changes in the abundance and isotopic composition of pyritic sulfur are generally the result of variations in the supply of sulfate to a lake from rainfall runoff. A shift to warmer, wetter conditions, such as that experienced by some areas during the early Holocene, may be marked by an increase in the abundance of pyritic sulfur and a negative shift in sulfur isotope composition (Spiker and Bates, 1993).
Sulfur species commonly analyzed include disulfide "pyrite" sulfur (S2-), "acid-volatile" sulfur (S- in monosulfide minerals, e.g. greigite and pyhhrotite), organic sulfur (S in kerogen and bitumen), and sulfate sulfur (Tuttle et al., 1986, Bates et al., 1993). Sulfur isotopes (34S/32S ratios) are generally measured on the pyrite and organic-sulfur fractions, although they can be measured in all major fractions depending on abundance. Analytical methods are discussed in Bates et al. (1993).
5.5.3.22 Carbonate content
The carbonate content of lake sediments, composed of detrital, authigenic, and biogenic components, reflects the chemical limnology of the basin as well as biological activity (Kelts and Hsü, 1978; Talbot, 1990; Eugster and Kelts, 1983). The relative amounts of detrital and authigenic carbonate are particularly sens-itive to drainage basin conditions of many lakes; both transport of detrital sediment to the lake and chemical precipitation of carbonates respond in different ways to climatic changes. The chemistry and mineralogy of authigenic carbonate phases are also sensitive to climatic and drainage basin conditions (e.g. Dean, 1973). Authigenic and biogenic carbonates are the primary materials for many isotopic studies (Stable Isotopes section).
Total carbonate is usually measured by volumetric methods on CO2 evolved during reaction with acids, or by difference between total carbon and organic carbon measured by other means (Organic Carbon section). Carbonate mineralogy and chemistry are ordinarily determined by X-ray diffraction (X-ray Mineralogy section) and elemental methods (Elemental Chemistry section), respectively.
5.5.3.23 Biogenic Silica
Biogenic silica, produced mostly by diatoms, is a major component of lake sediments in many environments. Other components that contribute to biogenic silica include phytoliths, chrysophyte cysts, and sponge spicules. Down-core variations in biogenic silica reflect primarily changes in diatom productivity (e.g. Colman et al., 1995), which is a complex response to a variety of physical and chemical properties of the water column, as well as such variables as nutrient and light availability. Water chemistry and sedimentation rates also affects the amount of biogenic silica that escapes dissolution to be preserved in the sediments.
Biogenic silica is equated with opaline silica, which can be measured by wet-chemical methods. A variety of methods are in use (Battarbee, 1986; DeMaster, 1981; Mortlock and Froelich, 1989). In some methods, a correction can be made for the silica contributed by fine-grained clays that inevitably dissolve along with the opaline silica (Mortlock and Froelich, 1989; Colman et al., 1995).
5.5.3.24 Stable Isotopes (C, H, N, O, S)
A wide variety of chemical and biological processes cause fractionation of stable isotopes in lake sediments, so that analysis of the isotopic composition of various components of the sediment allows inferences about the history of climatic and other environmental conditions in the lake. Oxygen and carbon isotopes in carbonates precipitated from the water column, either organically or inorganically, have been most commonly analyzed (e.g. Lister, 1988; 1989; Schwalb et al., 1994). Oxygen and carbon isotopes in temperate lakes reflect the isotopic composition of met-eoric water, the composition and source of which has important paleoclimatic implications (Stuiver, 1970). These isotopic compositions are subsequently modified by limnological (e.g. evaporation) and biological processes (Talbot, 1990). Studies that monit-or and model modern isotopic composition of lake and meteoric water (e.g. Hostetler and Benson, 1994) are important for reconstructing these processes. Although carbonates are the most common and straight forward materials for isotopic studies, siliceous materials (such as diatoms, chrysophyte cysts, and sponge spicules) have some potential.
Carbon isotopes of organic matter are another important paleolimnological tool, since photosynthesis causes significant fract-ionation of carbon isotopes (Organic Geochemistry section). Although the processes are complex. carbon isotopes in organic matter are important reflections of productivity, aquatic vegetation type, and other biological factors (McKenzie, 1985). Also promising are isotopic measurements on specific organic matter components (e.g. hydrogen isotopes in lignin; Krishnamurthy et al., 1995).
Valuable paleolimnological and paleoclimatic information is also potentially available from nitrogen and sulfur (Sulfur Geochemistry section) isotopes. However, these methods still need much additional basic research.
A variety of methods are used to separate materials for isotopic analyses. The actual isotopic measurements are usually made by standard mass spectrometer methods. The references cited above and Fritz and Fontes (1980) are useful sources of additional information on methods.
5.5.3.25 Pollen and Spores
Pollen and spores in lake sediments have always been a mainstay of continental paleoclimatic reconstructions. Pollen and spores provide information about changes in the composition and spatial patterning of late Quaternary vegetation that can be used to infer regional paleoclimatic changes. Pollen analyses provide information on changes occurring both on land and in the water body itself. Pollen spectra also show how different vegetation types, whose response times can vary dramatically, react to climate change. Pollen can also be combined with plant microfossils to reconstruct lake-level changes (e.g. Schneider and Tobolski, 1985).
Different types of lakes preserve different vegetation signals, ranging from aquatic vegetation in the lake itself, to vegetat-ion in the drainage basin, to composites of regional vegetation. In small lakes or in shallow sections of large lakes, plant macrofossils and pollen grains and spores of aquatic vegetation are more abundant. However, the risk of hiatuses is higher in these environments. In deeper and larger lakes, pollen is more likely to provide information about regional upland vegetation. Large deep lakes typically have higher sedimentation rates and fewer hiatuses. Ideal-ly, both shallow and deep lakes in the same area should be cored for their complementary information. In varved lakes, pollen analys-es can yield annual to seasonal resolution of reconstructed vegetat-ion (e.g. Peglar, 1993). High-resolution studies in which pollen is combined with other paleoecological indicators are especially useful for documenting rapid changes (e.g. Van Geel et al., 1989), including human disturbance of natural vegetation.
Theoretically, pollen in lake sediments can be used to reconstruct precise vegetation assemblages, and vegetation can be used to directly reconstruct climate variables (e.g. Elk Lake: Whitlock et al., 1993; Bartlein and Whitlock, 1993). These linkages lead to close relationships between vegetational and climatic modeling (e.g. Harrison et al., 1995). Statistical methods for these reconstructions include closest-analog methods (Overpeck, 1989) and transfer functions (e.g. Guiot, 1987). Two major problems, lack of modern analogs for past pollen spectra and human disturbance of modern vegetation, introduce uncertainty into these reconstructions, so they should be used with caution. In any case, modern samples from comparable depositional settings need to be collected for the study region, in order to provide the best calibrations possible for the relationships of pollen to vegetation and veget-ation to climate. In large lakes, a network of surface-sediment samples in different depositional environments is needed.
Major regional reconstructions of paleovegetation from pollen studies include Bartlein et al., 1986; COHMAP members, 1988; Webb et al., 1993; and Guiot et al., 1993). Pollen spect-ra of long sections can be compared to the marine oxygen-isotope record, either directly in the same (marine) cores (e.g. Dupont, 1993), or indirectly, through spectral and correlation statistical techniques (e.g. Hooghiemstra et al., 1993).
Pollen and spores must be extracted from the sediments and analyzed in a consistent manner to ensure comparability of results among laboratories. Recommended guides to laboratory procedures, especially physical and chemical extraction of microfossils, include Faegri and Iversen (1989) and Moore et al. (1991). Exotic markers should be added to all samples, so that pollen concentrations can be calculated in addition to percentages. Where time control is sufficient, pollen fluxes can be calculated from concentrations.
5.5.3.26 Diatoms and other Siliceous Microfossils
Diatoms and other limnic microfossils are especially important because their abundance and distribution is controlled directly by physical and chemical limnology. Direct reconstruction of paleolimnological variables is possible (e.g. Bradbury, 1988; Bradbury and Dieterich-Rurup, 1993; Gasse and Fontes, 1989), although taxonomic uncertainties and incomplete understanding of environmental factors that control species distributions limit paleoenvironmental interpretations. Nevertheless, diatoms are the primary source of paleolimnological information for many lakes.
For diatoms, general separation and counting methods are fairly well established (Battarbee, 1986). Surficial lake sediments should be collected near the center of the basin and in other depositional environments (e.g., littoral or euphotic zone, fluvial communities) for modern ecological calibration. Measurements of water conductivity, pH, surface temperature, and secche disk depth should be made in the field. In addition, water samples must be collected for the calibration of water chemistry with modern and fossil assemblages.
For other siliceous microfossils (such as sponge spicules, phytoliths, and chrysophyte stomatocysts and scales), the diatom processing procedure is suitable in most cases. These microfossils will be observed in the course of diatom enumeration, and notes should be taken to determine if further exploration is warranted.
5.5.3.27 Macrofossils (Plant and Animal)
Studies of macrofossils, "any part (of an organism) preserved after death which does not require a high-power microscope to see it, and which can be manipulated by hand (Birks, 1980)," in lakes contribute most to paleoecological reconstructions when used in combination with pollen analysis. Macrofossils include seeds, leaves, wood, conifer needles, fish bones, mollusks, insects, and other remains. The advantage of macrofossils is that they may be identified to species level much more easily and frequently than pollen. Thus they can provide critical taxonomic clarification. Macrofossil presence in sediments indicates local presence on the paleolandscape (Birks 1980; Glaser 1981), which allows spatial refinement of paleo-distributions not usually possible with pollen. Macrofossil distribution and abundance in lakes varies considerably, depending on specific characteristics of the lake and vegetation. In general, however, lakes selected as regional pollen sites will not be the most suitable sites for macrofossil studies: a compromise usually arises in which multiple cores are taken and macro-fossils retrieved from lake-margin cores, which must be correlated to central pollen cores. Guidelines for methods and interpretation of macrofossils include Berglund (1986); Birks (1980); and Wasylikowa (1986).
5.5.3.28 Ostracodes (including Geochemistry and Isotopes)
Continental ostracodes are small bivalved crustaceans that live in many permanent and ephemeral aquatic environments. The physical and chemical properties of the environment determine which species may live there; environmental tolerances vary widely with species. Ostracode valves, which are made of calcite, are common fossils. Ostracode species assemblages as well as the isotope values and trace-metal ratios of their valves provide a wide variety of paleolimnological information (DeDeckker and Forester, 1988). For many species, the modern distribution and related environmental variables have been documented (e.g. Delorme, 1989). From such information, hydrochemistry and temperature response surfaces have been constructed for various species assemblages (Forester, 1991; Smith, 1993).
Trace metal ratios (Mg/Ca, Sr/Ca) from ostracode valves are related to water temperature, salinity, and perhaps other fact-ors during calcification (DeDeckker, 1988; Engstrom and Nelson, 1991; Holmes, 1996). Ostracodes also are a source of biogenic calcite that is at or near isotopic equilibrium with lake water, for stable isotope analyses (e.g. Forester et al., 1994; Schwalb et al., 1994; Heaton et al., 1995; see also stable isotope section). They are also useful material for AMS radiocarbon analyses (Colman et al., 1990).
Continental ostracodes are small and their valves are oft-en fragile. The sand-size valves must be separated from the bulk sediment with care, using techniques such as freezing the sediment to aid dispersion, followed by gentle washing and sieving, and hand picking.
Dating Methods
5.5.3.29 Radiocarbon
Radiocarbon analyses are, of course, the mainstay of chronology for lake sediments less than about 40,000 years old. Many books have been written about sampling and analytical methods for radiocarbon dating as well as interpretation of the results. An entire journal (Radiocarbon) is devoted to the subject. Accelerator-Mass Spectrometer (AMS) methods have allowed small samples of individual components of lakes sediments, such as macrofossils, to be analyzed separately (e.g. Hajdas et al., 1995). An important issue related to lake sediments is the calibration of the radiocarbon time scale with the sidereal (calendar) one (Stuiver et al., 1991).
Three problems are especially prevalent in analyses of lake sediments:
· reservoir effects, in which dissolved CO2 in the lake is not in equilibrium with the atmosphere (e.g. the "hard-water effect" produced by groundwater containing CO2 dissolved from bedrock);
· the contribution to lakes of detrital carbon, which is older than the sediments in which it is deposited;
· recycling, resuspension, and redeposition of previously deposited material.
The carbon fraction that yields the most accurate age depends on the lake system studied. In some cases, biogenic carbonate may give the best results if reservoir effects can be evaluated by dating shell collected live before atmospheric nuclear testing (e.g. Rea and Colman, 1995). Where detrital organic carbon input is high, the humic acid fraction may concentrate less refractory, autochthonous organic carbon; in other cases, particularly old samples not kept frozen, humic acids may concentrate modern microbial contamination. Pollen extracts may also yield accurate results (Brown et al., 1989). In general, analyses of total organic carbon (TOC) are difficult to evaluate, although, where other data show that most of the carbon is autochthonous (e.g. Colman et al., 1996), such ages may be acceptable.
5.5.3.30 210Pb and 137Cs (and related Isotopes)
Lead-210 profiles offer one of the best ways of obtaining high-resolution chronologies for the recent past (100-150 years). However, chronologies and dry-mass sedimentation rates derived from 210Pb measurements may be highly model-dependent (Appleby and Oldfield, 1992). This is especially true where sedimentation regimes have changed through time. Measurements of dry mass and wet volume coupled with the use of additional dating constraints and (or) multiple profiles often allow the assumptions underlying the alternative models to be tested.
Measurements of 210Pb are made either by direct gamma assay (Appleby et al., 1988) or through the decay of its alpha-emitting grand-daughter 210Po (Robbins, 1978). The advantages of the former approach are that the procedures are non-destruct-ive and simultaneous measurements of 137Cs, 134Cs and, in a sufficiently low-background detector, 241Am (Appleby and Oldfield, 1992) can be made. Moreover, estimates of changing 226Ra-supported 210Pb levels can be obtained by measuring the activity of the short-lived parent 214Pb, provided procedures are adopted either to preclude or to allow for the effects of radon escape. Direct gamma measurements are thus very useful where samples are required for subsequent analysis, where the additional constraints on age-depth calculation provided by fall-out radioisotope profiles are needed, or where changes in sediment type and source make variations in the level of 226Ra-supported activity likely. The sample mass required varies with the type of detector used, but may be less than a gram in relatively high-activity samples.
The advantages of the second approach, using alpha spectroscopy, include speed and economy. In most cases, 226Ra-supported activity is estimated from the 'tail' of the 210Pb versus depth curve and assumed to be relatively constant. Alternatively, separate measurements of 226Ra activity can be made.
5.5.3.31 Magnetic Remanence (Paleomagnetic) measurements
Depositional remanent magnetism (DRM), one form of natural remanent magnetism (NRM), is a critical chronological component of lake sediment studies, to the degree to which those sediments faithfully record the earth's magnetic field at the time of deposition. Magnetic reversal stratigraphy is somewhat peripheral, because the youngest magnetic reversal, the Brunhes-Matuyama (ca. 780 ka; Baksi et al., 1992), is beyond the time period emphasized here. However, the Brunhes-Matuyama boundary commonly is useful for constraining younger age estimates based on sedimentation rates or on the number of 100,000-yr climate cycles.
Within the Brunhes chron, about a dozen so-called excursions have been identified and dated (e.g. Champion et al., 1988). These large-scale transient changes in apparent field position are potentially valuable chronostratigraphic markers. Nothing uniquely identifies a given excursion, however, and not all of them appear to be recorded globally. Apparent excursions can be caused by core disturbance, so great care must be used in identifying them.
Changes in the non-dipole component of the earth's magnetic field lead to small-amplitude, regional magnetic-field changes (secular variation). Secular variation curves for different regions dated by radiocarbon or varve-counting methods can be used for time control by correlation (e.g. Lund, 1993), but, as for excursions, additional time constraint is required for unique matching of curves. Core disturbance and post-depositional alteration must also be eliminated.
Finally, changes in the relative intensity (measured intensity normalized to another magnetic property) of the earth's magnetic field have the potential to provide correlations to dated sequences (e.g. Peck et al., 1996). Special care must be taken to test whether the sediments are suitable for recording relative intensity (King et al., 1983) and that variations in relative intensity are not caused by one of many depositional and post-depositional processes, rather than the field intensity at the time of deposition.
Samples for paleomagnetic analyses (Collinson, 1983) are generally taken in small plastic boxes pushed carefully into the split face of the core. Samples need to be "cleaned" to appropriate levels determined from step-wise alternating field and (or) thermal demagnetization. A variety of other tests are needed to ensure that the measured remanance is primarily depositional.
5.5.3.32 Varve Counting
Varved lake sediments are extremely valuable in limnological and paleoclimatic studies because of their annual, and in some circumstances, seasonal, time resolution (Anderson and Dean, 1988). In most cases, the annual nature of the laminations needs to be demonstrated using independent time control, and even then, varves may occasionally be missing. Varve formation can also be documented by sediment-trap studies (e.g. Thunell et al., 1993).
Varve thickness itself is an important indication of limnological and watershed processes, and when the annual nature of the varves is combined with other climatic and limnological proxies, extremely valuable inferences can be drawn (e.g. Lotter, 1991; Anderson, 1993; Goslar et al., 1995; Zolitschka and Negendank, 1996).
Varve thickness is the most common parameter measured in paleolimnological studies, but other variables can be measured as well. Varves are usually measured and counted in thin section (see Thin Sections, above). Changes in thickness of varves or laminae can be determined by adapting methods derived from dendrochronology (Thetford et al., 1991). X-radiography of uniform-thickness core slabs also provides useful images for varve and lamination analysis.
5.5.3.33 Tephra Analysis
Volcanic tephra, in addition to its direct and indirect affects on climate and vegetation, is an important part of the chrono-logy of lake sediments in many areas. This is because tephra layers form essentially isochronous horizons that are relatively easy to correlated on the basis of chemistry and mineraology, from independently dated sites to other sites. Recent research into tephra distribution has shown that the geographical extent of volcanic ash distribution previously has been consistently and seriously underestimated. Extensive studies in European peats and lake sediments show that these deposits contain microscopic layers of fine Icelandic volcanic ash (e.g. Pilcher and Hall, 1996). Advanced analytical techniques, especially for single shards, allows geochemically confirmed site linkage over extensive areas, which, combined with high-precision multi-sample 14C ages of sediments containing tephra, has resulted in chronologies that, under optimum conditions, are comparable to the dendrochronological time scale.
Tephra studies are especially useful when combined with regional stratigraphy and geological records to form a chronological framework for large areas (e.g. Sarna-Wojcicki and Davis, 1991). Chemical and mineralogical methods for identification and correlation of tephras are diverse (e.g. Westgate and Gorton, 1981) but relatively standard except for the application to small samples or single shards. Because tephras are preserved in many types of paleoclimate records (ice cores, tree rings, marine sediments) in addition to lake sediments, their usefulness in lacustrine paleoclimate studies is unlimited. PAGES currently has a major project involving explosive volcanism (PAGES Workshop report, Series 96-3: Climatic Impact of Explosive Volcanism; 1996).
5.5.3.34 Other Dating Methods
A variety of experimental or developing dating methods are applicable to lacustrine sediments. Uranium-series methods have provided reliable ages for several kinds of authigenic minerals (e.g. Bischoff et al., 1985; Lao and Benson, 1988). Experimental schemes based on Uranium-series isotopes in bulk sediment (e.g. Edgington et al., 1996) also have shown promise.
Various luminescence measurements-thermal (TL), optic-ally stimulated (OSL), and infra-red stimulated (IRSL)-also have potential for use with lake sediments (e.g. Berger, 1988; Godfrey-Smith et al., 1988). However, because of uncertainties about initi-al doses and degree of bleaching at the time of deposition of lake sediments, these methods have more commonly been used with other sediments such as loess and eolian sand. Electron-spin resonance (ESR) methods (Grün, 1989) have been used for chemically precipitated lake sediments, but still entail large uncertainties. Care must be taken to avoid exposure of samples for luminescence dating to light or X-rays.
These and other methods are particularly important for time control beyond the range of radiocarbon dating.
5.5.4 Other General Logging, Sampling, and Analysis Notes
· Whole-core, multi-sensor, logging instruments are available that measure magnetic susceptibility, GRAPE, and P-wave velocity in a single, automated pass-through of the core.
· Description, photography, scanning, and sampling should be performed as soon as possible after the cores are split.
· Samples for many methods that are of secondary importance and that would disturb the cores can be taken from material in the core catcher or from the very top and bottom ends of the core sections.
· The natural breaks in activities come: (1) after the cores have been logged, but before they are split, and (2) after time-critical description, measurement, and sampling activities.
5.5.5 Sample and Archive Material Availability
Information on the condition, storage, and availability of sample and (or) archive material is important (see Data Archive, below). Exclusive access to core and sample material by active project participants should be limited to a specific period of time, commonly two years. Unused core material should be made available to the general research community after this period.
5.6 Data Management and Dissemination
Data management is of vital importance for developing a global paleoclimate history from continental settings. Developing the continental record requires the synthesis of paleoclimate reconstructions from a mosaic of sites. In order to achieve this goal, both primary and secondary data from terrestrial sites must be archived in coherent file formats and made accessible to the research community. For this reason, it is a primary responsibility of any scientific investigator to contribute data to the PAGES Data Center, the World Data Center-A for Paleoclimatology in Boulder, Colorado, USA.
5.6.1 Data Coordination
Effective data coordination is essential for maximizing the success of any project. A data coordination program implemented at the point of project inception can expedite data processing and synthesis thereby making the data and results available to the general research community as soon as possible. Good data coordination is important for developing syntheses of project data in local, regional, and global contexts.
Facilitating data flow among project members is particul-arly important for large projects. An internal project database or data server may be necessary to provide access to all data associ-ated with the project. Access to the server could be through the internet, by mail using diskettes, and (or) through non electronic methods. The project data coordinator should be responsible for maintenance of the project database including:
· Development of effective and complete means of archiving group data and disseminating those data among group scientists;
· Construction and maintenance of a complete core depth scale, including drilling depths, drive depths, and core section depths;
· Keeping a sample inventory (depth, sample number, interval, thickness, amount of sediment, proxy(ies), investigator;
· Core archive material inventory, which is part of core storage and is needed to keep track of who gets archive material and information;
· Obtain and facilitate access to data necessary to complete the research outlined by the science questions for the overall project (e.g. historical meteorological data, solar forcing time series);
· Maintenance of the data for calibration and modern system characterization.
Recommendations for Data Coordination are already discussed at length in the PALE Research Protocols (PALE Steering Committee, 1994), and those recommendations should be consulted when building a data coordination program. However, some aspects of effective data coordination are important enough to be repeated here.
Data Coordinator
Every project should have a data coordinator. This person may be one of the principal investigators on the project. If the project is large, it might be appropriate for a non-researcher to be appointed data coordinator. The data coordinator should be responsible for maintaining a working database for the project, and making that database available to all members of the research group. The coordinator might also obtain other datasets of use to the project, or provide information on how to obtain useful data. Finally, the data coordinat-or should be responsible for collecting and formatting data for easy submission to the PAGES/WDC-A for Paleoclimatology database.
Data Synthesis
It is important to interpret proxy records obtained from a single site or group of sites in a global and regional context, and to assess the impacts of climate changes indicated by these proxies on ecosystems. These syntheses should be made with the idea that the data produced from the analyses and interpretations can be used for larger scale syntheses. Of particular importance is the usefulness of the data for GCM simulations and sensitivity experiments. For many projects, it may also be important to write summary papers in common language to be read by the general public.
Data Availability
Data should be put into the public domain as soon as possible, and no later than two years after the data set is complete, or immediately after its publication. Rapid availability of data is a responsibility of publicly funded research, and it is important to modern quantitative research activities related to global change science, among other things.
5.6.2 PAGES Data Archive
A primary responsibility of an investigator is to contribute data to the World Data Center-A for Paleoclimatology in Boulder, Colorado, USA. Data should be submitted to the data center in a timely manner, and a plan for submitting data should be made at the beginning of the project. Data should be submitted to the data center immediately after publication, and(or) two years after sample collection, whichever comes first.
Data submitted to the archive should be prepared in a logical file format that is consistent with the guidelines for submission of data to the World Data Center-A for Paleoclimatology. Data files may be ASCII, Spreadsheet, or other common computer formats. Data files can be sent by E-mail, FTP transfer, or via the postal system on diskettes or other media. Please direct inquires to: Mr. Bruce Bauer, data manager, bab@mail.ngdc.noaa.gov.
Submitted files should include the following information:
· Point or Time Series Data:data set name, latitude, longitude, bathymetry/elevation, contribut-ors, references, variable names, variable unit, data precision, data format, contact information (mail, phone, electronic mail).
· Mapped or Gridded Data Sets:data set name, latitudes of grid points, longitudes of grid points, time period represented by data, contributors, references, variable names, variable unit, data precision, data format, contact informat-ion (mail, phone, electronic mail).
· Optional Information: additional site information, brief description of data, methods including references, chronostratigraphic information, date of data generation.
Contributors are urged to examine existing WDC-A data sets for examples of important archival information. These data can be viewed via anonymous FTP (ftp.ngdc.noaa.gov; look in the paleo directory) or on the WDC-A's home page on the World Wide Web (http://www.ncdc.noaa.gov/paleo/paleo.html).
Addresses for World Data Center-A for Paleoclimatology:
E-mail: paleo@mail.ngdc.noaa.gov; FTP: ftp.ngdc.noaa.gov.
Files can be placed in the /pub directory.
Postal, WDC-A for Paleoclimatology, 325 Broadway, code E/GC, Boulder, CO 80303, USA
5.7 Publication Protocols
Research results for any project must be published in a timely manner. To ensure timely publication of results and interpreta-tions, and to avoid conflicts, a publication procedure and schedule should be developed at the beginning of the project. The nat-ure of such a schedule will vary according to the goals and size of each project, but some plan for the publication of results should be agreed upon in advance.
One procedure that has worked well in a variety of circumstances is publication of a comprehensive "Preliminary Results" report, authored by the generic team project members, which, in addition to the first set of results, contains an overview of the project, its organization, and its members. Project members generally agree to cite this initial publication in all subsequent topical and synthesis reports.
Publication of topical investigations should be completed in a timely manner so that syntheses of the proxy data in a regional and global context can be made. Syntheses and summaries need to be carefully balanced among topical studies and principle investigators; project Workshops or Symposia facilitate this proced-ure. Summary papers for the general public commonly prove helpful to the project as a whole, especially if they are focused on the impacts of climate change on landscapes and ecosystems.
| 0. TABLE OF CONTENTS |
