This article throws light upon the top four methods of biological sample analysis. The methods are: 1. Sampling Pattern Options for Plant Species and Communities 2. Measures of Abundance for Plant Species and Communities 3. Methods for Sampling Invertebrates 4. Bird Census Method.
Method # 1. Sampling Pattern Options for Plant Species and Communities:
i. Random Sampling (unrestricted):
ADVERTISEMENTS:
Observations are made at random points within the site or study area. It is theoretically the most statistically acceptable method, and many ecologists argue that it should be used whenever possible.
However, it assumes that species have random distributions, which is rarely, if ever, true, and it has two serious disadvantages:
(a) Unless a large sample is taken, the chance observations are unlikely to provide representative data (e.g., they could be clustered in one small area), and
(b) The results do not normally permit the detection of community gradients within the sampling area.
ii. Systematic (Regular) Sampling:
Provided that care is taken to avoid regular features that can be the most cost-effective pattern for obtaining representative samples of a site or study area, and of sampling community gradients.
ADVERTISEMENTS:
Options include:
1. Line transects which can be useful for reconnaissance or application of the line-intercept method;
2. Interrupted belt transects which can consist of:
(a) Lines with sampling points (stations) at regular intervals (and intermediate stations if required), or
ADVERTISEMENTS:
(b) Selected (e.g., alternate) squares within which sampling is conducted .They can be particularly useful for sampling community gradients;
3. A grid system which can cover the site or selected study areas, with observations taken at intersections (stations) or within grid squares. All or a proportion of (e.g., alternate) stations or grid squares may be sampled. A grid can provide the most representative sample of a study area, though precision is affected by the size of the squares which should be as small as practicable.
iii. Restricted Random Sampling:
This is sampling at random locations within areas defined by a systematic pattern, e.g., the squares of a belt transect or grid system. It is considered to be a reasonable compromise between unrestricted random sampling and systematic sampling. It does not detect community patterns within the defined areas, e.g., grid squares, which should, therefore, be small.
ADVERTISEMENTS:
iv. Stratified Sampling:
This involves the selection of study areas, and sampling each of these according to its size. It is particularly useful for large sites where adequate systematic sampling of the whole area is unnecessary or impracticable.
The study areas can be defined by features such as fairly discrete vegetation boundaries, land-use boundaries, or tracks, and can be sampled systematically or randomly. Stratified random sampling should not be confused with restricted random sampling.
v. Selective Sampling:
This involves the subjective (not random or systematic) selection of sampling points thought to be representative of a community type or to contain special feature such as a particular species under study. It may be the only practicable method, e.g., when access to sampling points is difficult or when a species only occurs in scattered locations.
However, it is also employed in the relieve method in which samples are deliberately obtained within apparently homogeneous vegetation patches that are thought to be representative of a community type.
In this case, a requirement is that the area sampled should be large enough to include all the community’s species. A popular procedure to ensure this is estimation of minimal area, though it is only reliable in truly homogeneous vegetation.
An advantage of the relieve method is that the samples should represent communities amenable to formal classification, e.g., in the NVC; disadvantages are the subjective selection of sampling areas and the deliberate disregard of community gradients.
Method # 2. Measures of Abundance for Plant Species and Communities:
With DAFOR the most abundant species is assumed to be dominant; ACFOR may be preferred because it does not require this assumption (although the estimated dominant can be noted).
Conversely, it may be difficult to distinguish between C and F in ACFOR. Either system can be supplemented by ratings such as co-dominant and subdominant, and by distributional information such as widespread or local (e.g., LA = locally abundant).
The measures permit rapid survey but are subjective, approximate, and have limited potential for analysis and presentation (although the ratings can be converted to a 5-point numerical scale). Consequently, there is little point in using them if equivalent quantitative data can be obtained.
i. Number of Individuals:
Numbers are usually counted in quadrats, and expressed as density (number per unit area) or population size (number in site or overall study area). These are appropriate for estimating abundance of selected species, provided that individuals can be readily counted, but generally not applicable in community studies because number of individuals has little meaning when comparing species of widely differing size.
ii. Cover (%):
This is the percentage of ground occupied by a perpendicular projection onto it of the aerial parts of a species.
Cover can be measured by the following methods:
1. Visual estimation in quadrats is the most flexible and popular method, but is subjective and prone to observer error, especially in dense or tall vegetation, and tends to under-value species having small scattered individuals. For these reasons, values are usually recorded using incremental cover-abundance scale.
2. Line-intercept method, which can be accurate and quite rapid in simple vegetation, but is more difficult and time-consuming in complex vegetation.
3. Point-intercept (point quadrat) method, which can be the most objective and precise method, but is time-consuming and only readily applicable in short vegetation.
Individual observations are quantitative and should be representative; but the time needed with any of the methods can discourage the collection of an adequate overall sample of a study area.
iii. Cover-abundance:
This aims to avoid underestimating the importance of small species with scattered individuals by using cover for species with cover 4% or 5%, but abundance in the strict sense (of numbers) for species with cover 4% or 5%.
iv. Frequency (%):
This is the percentage of observation in a sample that contain the species, i.e., is derived from presence/absence observations (usually in quadrats).
Disadvantages are:
It is strictly a measure of distribution rather than abundance, and tends to over-represent small species; it fails to discriminate between high density (with many individual present in quadrats) and density that is just sufficient for at least one individual to be present in a large proportion of quadrats; and frequency values increase with increasing quadrat size, so results from surveys using different sized quadrats, and should not be calculated from sets of less than 20 observations.
In spite of its limitations, frequency can be the most cost-effective method for obtaining large representative samples in community studies because it is relatively rapid and free from observer error.
Method # 3. Methods for Sampling Invertebrates:
(i) Observer dependent methods:
1. Direct observation and identification:
General searching and recording species found, usually in areas defined as having “interesting” vegetation. Such methods are not normally quantitative, can easily lead to misidentification, only record species that are active when the observation limited to species that are either common in the study area or extremely conspicuous, e.g., dragonflies.
2. Transect walking:
It involves the observation, identification and enumeration of species on a set route, undertaken within prescribed time and weather conditions; usually restricted to butterflies and day-flying moths.
3. Sweep netting:
A handheld net is swept through vegetation up to 1 m in height. Most invertebrates are swept off the vegetation and can be collected and transferred to a preservative for later identification. The method can be quantitative if a standard number of sweeps are taken, but sweeps in different vegetation types are not directly comparable, because different vegetation type differentially resists the net.
Invertebrates occupying the basal parts of vegetation are not normally sampled, and active flying individuals often escape from a net before they are captured. The method is not suitable for thorny or wet vegetation.
4. Swish sampling:
It is like sweep netting but is restricted to the air boundary immediately above vegetation. It is especially good at collecting Diptera (flies) and Hymenoptera (bees and wasps).
5. Suction sampling:
A portable vacuum is used to collect invertebrates from the ground layer and/or basal parts of vegetation. It can be an efficient collecting method in dry conditions and where there is little vegetation litter, and can provide quantitative data if a set number of samples are obtained.
6. Soil samples:
It can be taken for identification and enumeration of soil invertebrates. A variety of physical or chemical extraction methods are used to extract the organisms from the soil samples.
7. Beating:
A stout stick is used to knock invertebrates off vegetation onto a sheet from which they are collected. It is usually used to sample the fauna of individual species of tree. With care, the method can be used to obtain quantitative data, but it is not practical in wet conditions.
8. Subsidiary methods:
Most experts have favoured methods for detecting those invertebrate groups with which they are most familiar. Methods include observing flower visitors hand- searching vegetation for plant grazers (especially molluscs), stone turning (especially for beetles, molluscs and millipedes), and investigating litter and dead wood for decomposers.
(ii) Observer-Independent Methods:
1. Pitfall traps:
A good quantitative method used chiefly for ground-dwelling beetles, which fall into the traps. Pitfalls are placed on a regular grid within selected areas. They usually contain a fluid that kills and preserves the invertebrates caught.
2. Malaise traps:
Flying insects are intercepted by a net and funnelled into a collection vessel by their flight mechanism. The method can collect large numbers of insects, especially Diptera and Hymenoptera, and is efficient for obtaining quantitative data. It does not discriminate between insects resident in, or simply flying through, the area.
3. Sticky traps:
Usually consists of a mesh screen on which a viscous oil is applied. They can be used like malaise traps, or placed within vegetation. Fragile invertebrates may become damaged in trying to escape from the trap, and sampled invertebrates have to be removed by a solvent.
4. Water traps:
Rely on the fact that a variety of flying insects (especially flower visitors) are attracted to coloured surfaces. They are simple to use but selective.
5. Light traps:
Night-flying insects are attracted to light sources, especially if these emit ultraviolet wavelengths.
They are useful, but:
(a) they require a power source and are not easily transported, and
(b) they may sample species that are not associated with the site but which are flying over it.
6. Emergence traps:
Usually consist of a closed mesh canopy placed over vegetation, and a collecting vessel. They are designed to collect most adult flying insects which were in a developmental stage on the vegetation or in the soil when the trap was erected. They can be used quantitatively, because each will have a known basal area, but must be in place for long periods.
Method # 4. Bird Census Method:
(i) General Aspects:
All methods are time-consuming, involve extensive site walking, and require expertise, e.g., in the recognition of bird calls/singing.
They are affected by:
(a) Seasonal variations, e.g., breeding seasons, in which different species sing, and seasons of migrant visitors, and
(b) Time of day, e.g., shortly after dawn being generally the best time in relation to bird song.
They require repeat sampling, e.g., early morning visits at weekly intervals, variation in routes and directions taken so that particular parts of the site are not always visited at the same time of day, and records of all contacts (sightings or call/song registrations) and the times these are made.
Factors affecting census accuracy include:
1. Birds are easier to find in some habitats than in others, e.g., they are less easily detected in dense scrub and woodland than in open habitats.
2. Some species are less conspicuous and/or noisy than others and therefore may be overlooked.
3. Recognizing different species, counting individuals and determining territories may be difficult where birds occur at high density.
4. In wet and windy weather birds may be less active and skulk out of sight.
(ii) Territory Mapping:
Can be used to determine densities, locations and territories. It is best conducted during the breeding season when most species are territorial, and territories are often marked by singing, displays and disputes with neighbors. It usually involves walking field boundaries (on agricultural land) or recording within defined plots (e.g., quadrats).
It is widely accepted in Britain but requires specialist expertise and is time-consuming, which can lead to error, e.g., if small plot sizes are used to save time. Consequently it may have to be restricted to those species which are rare or uncommon on a local, regional or national scale. The territory map should be presented in association with a habitat map, especially if impacts via habitat loss or change are to be predicted.
(iii) Line Transect Method:
Involves walking transects of fixed length (e.g., 1 km, though this can be sub-divided) and location (arranged randomly or systematically, e.g., as a grid), at a standardised speed, e.g., c. 2 km per hour.
The method can be used to estimate densities, e.g., by counting in relation to a prescribed band each side of the transect, is more rapid than mapping, and can be carried out throughout the year; but its value may be limited in small and/or heterogeneous sites.
(iv) Point Count Method:
Involves randomly located points at which observations are made. It can be useful in small sites and heterogeneous vegetation, and permits good analysis of association of bird species (including densities) with given habitat/vegetation types.
(v) Plot-less Method:
Involve the selection of observation points in relation to specific aims, e.g., at central points in the range of selected species. This can be useful in the study of scarce species that tend to be neglected by the other methods.