for Analyzing Recent Extinctions
It is difficult, if not impossible, to prove that the absence of a species (in recent times) is equivalent to its extinction. The best approach is to compile all available evidence that a species is absent, and then decide when the weight of that evidence is sufficient to assume that extinction has occurred. In order to evaluate the evidence in a rigorous and consistent manner, a system of criteria for analyzing the evidence must be developed.
To that end, prototype extinction criteria and categories were developed by MacPhee and Flemming (1999) for mammals and by Harrison and Stiassny (1999) for fishes. (See: Why Was CREO Developed? for additional information about these studies.)One of CREO's main goals is to develop extinction criteria that are rigorous, uniform, and applicable across all taxa. Yet, the prototype criteria initially developed were only designed to apply to two major taxonomic groups, mammals and fishes. In order to develop a new system of extinction criteria we proceeded as follows:
The CREO Workshop held on May 15-16 resulted in significant progress in establishing a uniform system for assessing extinction criteria. The first day of the workshop was spent discussing the issues raised on the Feasibility Surveys; and attempting to reach a consensus on these issues (see the workshop agenda) for more details about speakers and discussion topics for the focus groups).
Section 1 of this report summarizes the main discussion points raised during the first day of the workshop. This summary is based on opinions raised during individual focus groups or during the follow-up group discussions. It was not possible to reach decisions at the workshop for all of the topics under discussion. This report includes possible solutions for some of the topics that remain unresolved. These possible solutions are clearly indicated below as a "Proposal by CREO Secretariat," in order to indicate that it represents editorial comment rather than the opinions that were discussed during the workshop.
Section 2 of this report provides an explanation of a new system of criteria that is being proposed by the CREO Secretariat. This new system is based on the Panelists' comments about the extinction criteria that were submitted to the CREO Secretariat prior to the workshop, as well as discussion points raised by Panelists during the course of the workshop.Section 1. Summary of Workshop Discussions
CREO's prototype extinction criterion 1
The species name and status must be valid taxonomically.
Most of the discussion on this issue focused on species concepts and the treatment of subspecies. Rather than adopting a particular species concept, there was general agreement that CREO should advocate an approach of attempting to identify and document those entities that can be considered to be minimum diagnosable units. These entities may be considered by biologists to fulfill requirements for phylogenetic species, or they may be considered to fulfill requirements for biological species or subspecies. Regardless, as long as the entity is named and described according to the rules of nomenclature, and is acceptable to taxonomic authorities and the Advisory Panel as a minimum diagnosable unit according to documented evidence of diagnosable characters that are not based solely on geographic isolation, then the entity should be included in CREO's extinction analysis. Thus, those subspecies that represent minimum diagnosable units and are formally described and named according to the rules of nomenclature should be included in CREO's extinction analysis.
What is the rationale for adopting this approach? When compiling a survey of extinctions, we need to identify what units to count, and we want the manner in which biologists identify these units to be as uniform as possible. Yet, we know that different biologists utilize different species concepts, so if we ask biologists to adopt the approach described above, we can get as close as possible to a uniform way of counting units. Furthermore, we decided to include subspecies in the analysis because:
they are recognized by the rules of nomenclature as well as by many adherents of the biological species concept and,
if they are minimum diagnosable units, they should correspond to what adherents of the phylogenetic species concept would consider to be phylogenetic species.
There was considerable controversy about whether or not CREO's extinction analysis should include entities that have not been formally named and described according to the rules of nomenclature. No final decision was reached at the workshop.
CREO's prototype extinction criterion 2
The effective extinction date or EED (the last known date or time-interval in which the species was observed or collected, which provides a minimum date of loss for the taxon) can be positively confirmed as being later than AD 1500 either:
a) because a living specimen
has been collected or seen by a reliable source since that date (observation
of the species without collected specimens was not classified as acceptable
evidence for the study made on fishes); or
i) radiocarbon-dated, or
ii) found in close association with those of an alien species known to have been introduced post AD 1500.
There was lengthy discussion concerning how far back in time the CREO extinction surveys should extend. It quickly became apparent that meaningful start dates vary significantly by taxa. AD 1500 had been selected as a start date in the prototype system, however, there was agreement that there is no reason to be restricted to AD 1500 as a minimum start date for those taxonomic groups in which earlier extinctions can be reliably documented. However, we did not reach any conclusions about what we might set as an earlier start date. We discussed whether it was our ultimate aim to try to document all extinctions in the period of time since the development of modern humans, or some more recent arbitrary date, but these issues were left unresolved.
Another major issue discussed was whether we could (and should) establish guidelines for evidence that allow us to establish an extinction date. The extinction date is usually set by the last date that the species is known definitely to have been extant; this represents the earliest possible extinction date because the species is not known definitively to be extant after that date. Suggested evidence for establishing this extinction date includes collected material, photographed or audio-recorded evidence of the presence of the species, observation by a reliable source, etc. It was generally agreed that it would not make sense to try to set uniform standards for evaluating this evidence. Relevant evidence can be variable across taxonomic groups, therefore, it is more appropriate to allow Panels flexibility in determining what evidence should be used for establishing the date of a species' extinction. For example, it is not useful to stipulate what constitutes a "reliable source" in the case of using observed evidence for establishing the extinction date, because the confidence that we have in the source is likely to vary in each case and, therefore, should be evaluated on a case by case basis. See the new system of criteria outlined below for the methodology that the CREO Secretariat advocates on this matter.
CREO's prototype extinction criterion 3
No individuals have been observed reliably in at least the past 50 years.
Additional criteria deemed necessary in Harrison and Stiassny's (1999) study on fishes:
A) significant but unsuccessful attempts have been made to relocate the species after its noted absence (either in a search targeted towards that species or in a sampling exercise that covers well the expected range or habitat of that species);
B) there has been a decline of that particular species, or an environmental threat to it, before the extinction date.
Much of the discussion regarding this criterion centered on retaining or rejecting the 50-year rule. Focus Group 3 raised the following objections to the use of the 50-year rule:
there are cases where species are very cryptic and can be overlooked for long periods of time
the 50-year rule discounts extinctions that are undisputed but have occurred within the last 50 years
a 50-year rule is arbitrary and meaningless; it should be taxon-specific in its definition (i.e., more closely based on a biological parameter such as generation time); even then, a waiting period can only be arbitrarily defined for some taxa which are relatively poorly known).
Furthermore, the 50-year rule is no longer used by the IUCN Species Survival Commission, so if it was utilized by CREO it would make it more difficult for SSC and other conservation monitoring organizations to utilize data gathered through the CREO effort. It was agreed, during a follow-up group discussion, that the 50-year rule should be rejected.
Thus, instead of using a waiting period to evaluate extinctions, the focus should strictly be on assessing the adequacy of survey efforts. In order to evaluate survey efforts, we must first assess what we know about the species' habitat and range. If our understanding of either the habitat or range is poor, then we cannot be confident in the thoroughness of survey efforts. On the other hand, if the species habitat and range is well understood, then surveying can be evaluated for thoroughness, and the species can be ranked as a resolved extinction in those cases where surveying is deemed adequate. Participants at the workshop stated that it is also very useful to have an understanding of the frequency and natural fluctuations that may occur in the species population size. This provides important information for assessing whether an observed period of absence should be interpreted as a within the parameters of a typical fluctuation in population size, or whether it represents a possible extinction of the species.
Participants in Focus Group 3 spent some time discussing the guidelines that should be used for evaluating whether survey efforts have been sufficiently thorough to determine whether an species is extinct. It would be very useful to quantify survey efforts (e.g. number and duration of surveys, area surveyed, number of traps set etc.) and use this as a method for evaluating surveying for thoroughness. At least some degree of quantification is essential, for example, there should be some evidence of replicate surveys. However, it was generally agreed that a very precise, quantitative assessment of surveying will not be possible in many cases.
There was some debate concerning the importance of utilizing evidence of biological threat in extinction analyses. Participants in Focus Group 3 stated that assessments of elimination of habitat should be specific to individual species. In connection with this, some participants in Focus Group 3 stated that total destruction of habitat should be evaluated very carefully before being applied as evidence in support of an extinction; researchers should be certain that, even though the species' habitat is completely destroyed, there is no possibility that the species can withstand the elimination of its habitat.
Participants at the workshop commented that the occurrence of some biological threats (habitat disturbance, predation, hybridization) can be correlated with the disappearance of a species in some cases. Therefore, documentation of this information is very useful for assessing possible processes of extinction. However, the simple presence of a biological threat cannot be accepted as proof of an extinction; it was stated during the follow-up workshop group discussions that the only acceptable proof of extinction can come from satisfactory evidence that suitable surveying has not located the species.
The categories of "extinct in the wild" and "functionally extinct" were discussed briefly and there was divided opinion on the use of these categories. In the case of mollusks there are cases where extant species are represented only by long-lived but senescent individuals, therefore their extinction currently seems unavoidable. However, there was some opinion that these categories cannot be clearly defined and may serve to confuse the perception of extinction. These issues were not resolved at the workshop.
During the morning of the second day of the workshop, the results from the first day were distilled, and a basic flow chart (this is a PDF file) was established showing how extinction analysis should proceed.
There was also some discussion on the morning of the second day about the importance of accountability. Those who are assessing species' extinctions must be able to justify their decisions by providing references and/or an explanation that shows how a species was evaluated.Section 2. The New Criteria
We are now proposing a new, more formalized system of extinction criteria. In creating this system our objectives were to:
create a system that is rigorous, yet easy to apply
create a system that is as data rich as possible
In each step of the system described below, the information that we have about each species can simply be assigned a code. Certain code combinations are required for an extinction to be considered "resolved," other combinations mean that the extinction is "unresolved." This coding system provides a simple way of showing what evidence has been collected (so far) about the species' extinction, and what information must still be collected before the extinction can be considered resolved. The combination of codes assigned to each species can be reviewed and updated as new information comes in.
Unresolved extinctions can be subdivided into any number of categories using this system (by taxonomic problems, or by the need for additional survey data, or by missing habitat/range data, etc). Alternatively, these unresolved extinctions can all be put into one unresolved category, showing which codes they have been assigned so that we can easily see what we do and do not know about each putative extinction.
We have developed an updated flow chart (this is a PDF file) that describes the logic of the system described below. This flow chart is a slightly more detailed version of the flowchart that was developed at the workshop.
by Mary DeJong and Ian Harrison
EXTINCTION CRITERION 1
There must be some claim of a species' extinction in order for that species to be considered in this extinction analysis.
EXTINCTION CRITERION 2
The taxonomic validity of the species must be evaluated and documented. Entities allowed in this extinction analysis should be coded as one of the following:
a) species [and subspecies] that are formally named and described according to the rules of nomenclature (without any significant taxonomic dispute), and are recognized as a minimum diagnosable unit = TE = Taxonomy Established.
b) entities for which voucher material exists, and for which there is documented data to suggest that they are diagnosable units (such as a published description that includes reference to acceptable diagnosable characters), but are not yet formally named and/or described = TI = Taxonomy Incomplete.
c) entities that have been formally named and described but for which there is considerable dispute about the validity of their taxonomy = TD = Taxonomy Disputed.
All other entities are disqualified from this extinction analysis.
Once evaluated, taxonomic status should be coded and, if necessary, explained, as shown in the following examples:
2) Species that meet the other criteria for resolved extinctions (see below) but are known only from a very small hypodigm (i.e., number of collected and preserved voucher specimens) should be treated with caution. These species were evidently always very rare and may now be extinct. However, the small number of collected specimens might be a consequence of taxonomic confusion associated with the species.
Putatively extinct species with a small hypodigm can be evaluated as resolved extinctions if they meet the necessary criteria, however, if the hypodigm if small (e.g., less than five specimens) then it should be noted whether there has been any attempt since the collection of the species to review their taxonomic validity. For example, one might describe the hypodigm in any of the following ways (providing references where appropriate):
3) Taxonomic information that must be documented for each species is as follows (showing an example):
EXTINCTION CRITERION 3:
A date for the extinction of the species must be evaluated and established and it must not pre-date (date yet to be determined). Species' extinctions that occur before this period are disqualified from this extinction analysis. (As mentioned above, we would like to set the date at AD 1500 for the purpose of the CREO-IBOY book project.)
There are several sources of evidence that can be used to establish an extinction date the source that is appropriate will depend on the particular species under scrutiny. The evidence used for establishing the extinction date should be coded as one of the following, and appropriate references and/or explanations should be provided:
a) the date when a specimen was last collected = LC = Last Collection.
b) the results of radiometric techniques, or some other appropriate scientific methodology, for dating of subfossil/fossil remains or other material found in association with those remains = FD = Fossil Dating.
c) the date when diagnosable evidence (collected material, photographs, audio-recordings, archaeological/ethnographic evidence) associated with the presence of the species was last collected = DE = Diagnosable Evidence.
d) the date when the species was last detected by a reliable source but for which there is no preserved/recorded diagnosable evidence = RS = Reliable Source.
Example: Mandibularca resinus is an endemic cyprinid fish described in 1922 from Lake Lanao in the Philippines and noted in 1982 as possibly being extinct. There is no reliable documentation on the presence of this species in Lake Lanao during the intervening 60 years. Thus, 1922 represents the earliest possible extinction date. However, the possible extinction of M. resinus has been linked to a combination of factors, culminating in the introduction of competitive/predatory species in 1963 and 1975. Under these conditions, the extinction of M. resinus can be said to have occurred between 1922 and 1982, and probably some time from 1963 onwards. Thus, the extinction date can be represented as follows:
Extinction date: 1922 LC - 1963 DT - 1982 EC
2) It is best when the evidence for an extinction date can be independently verified. In the case of preserved evidence (i.e., a to c above), independent verification can be achieved at any time through subsequent examination of the collected evidence by relevant specialists. In the case of unpreserved evidence (i.e., d above; observational or acoustic evidence that was not collected or recorded), this evidence can only be independently verified through congruent reports of two or more relevant experts who were present at the reported location and time. The acceptability of this type of evidence is expected to vary significantly for different taxonomic groups and the credibility of this reliable source will be determined case by case by consensus of the appropriate Advisory Panel.
EXTINCTION CRITERION 4
(If the species is known only from subfossil or fossil remains, [i.e. there are no accounts of live specimens] skip to Extinction Criterion 5. Species that are known from accounts of live specimens, or are known only from remains of specimens that recently died ["fresh dead" material; see Guidelines below], must be evaluated by Criterion 4, but Criterion 5 can be skipped.)
Adequacy of survey efforts (or the necessity for surveying) must be evaluated and documented. In order to evaluate the adequacy of survey efforts, it is first necessary to supply evidence (references/explanations) for our understanding of the species' habitat and range. Then, corresponding survey data should be evaluated and coded as one of the following, with appropriate references and/or explanations provided:
1. THE PARAMETERS OF THE SPECIES' ENTIRE HABITAT AND RANGE ARE WELL KNOWN OR CAN BE REASONABLY CONSTRUED
a) and surveying that would be likely to reveal the species has taken place throughout its range and any remaining habitat that has not already been completely destroyed, and this surveying has not indicated the presence of the species = HRK AS = Habitat and Range Known, and Adequately Surveyed.
--survey documentation should include those surveys targeted at the species in question as well as surveying that, while not directly targeted at the species in question, would have been likely to reveal that species; for example, survey documentation should include any relevant information from activities such as hunting, harvesting, and fishing that would be likely to reveal the species.
--surveys must take place during times (daily and seasonal cycles), environmental conditions, and throughout places that are suitable for the detection of the species
--the intensity, extent, and duration of surveying should be appropriate for the species and its particular habitat and range; for example, in most cases surveys should be repeated over a reasonable time period for species that are cryptic, naturally rare, and/or known to go undetected for long periods.
b) but surveying is not necessary because the habitat has been completely destroyed = HRK HD = Habitat and Range Known, and Habitat Destroyed.
c) but while some surveying has taken place, this surveying is not yet sufficient according to the stipulations of "a" above = HRK IS = Habitat and Range Known, but Insufficiently Surveyed.
d) and it is not apparent that any surveying has taken place = HRK NS = Habitat and Range Known, and apparently Not Surveyed.
2. THE PARAMETERS OF THE SPECIES' ENTIRE HABITAT AND/OR RANGE ARE NOT WELL KNOWN (for example, the species may be known only from a single collection of one or two specimens)
e) and the parameters of the habitat and/or range that are currently understood have undergone adequate surveying without revealing the species, but the species' extinction is inconclusive until the parameters of its entire habitat and range (and the associated need for additional surveying) are better understood = H/RQ ASP = Habitat and/or Range in Question, so Adequate Surveying is Provisional.
f) and the parameters of the habitat and/or range that are currently understood have been completely destroyed, but the species' extinction is inconclusive until the parameters of its entire habitat and/or range (and the associated need for surveying) are better understood = H/RQ HDP = Habitat and/or Range in Question, so Habitat Destruction is Provisional.
g) and the parameters of the habitat and/or range that are currently understood have undergone some surveying without revealing the species, but the species' extinction is inconclusive until the parameters of its habitat and/or range are better understood and additional surveying has occurred = H/RQ IS = Habitat and/or Range in Question, as well as being Insufficiently Surveyed.
h) it is not apparent that any surveying has taken place = H/RQ NS = Habitat and/or Range in Question, and apparently Not Surveyed.
It is also useful, when possible, to document surveying efforts before the extinction date. These data can then be compared to the surveying after the extinction date, and may show a decline in the population of the species, or they might show natural fluctuations in the overall population size of the species. This comparative analysis can be useful in assessing the evidence for extinction, particularly for any species that undergoes periodic absences as a normal part of its life cycle, and appears to be absent under surveying regimes, only to turn up later. Documentation of previous, temporary periods of absence of the species from surveys can ensure that any contemporary periods of absence are not treated prematurely as evidence of extinction.
For many species it will not be possible to document detailed information about each survey effort, especially when surveying is largely the result of activities such as hunting, harvesting, and fishing that are not specifically attempts to scientifically survey the species (but relevant survey attempts, nonetheless). These survey efforts should be still be considered in the extinction analysis even if detailed information is not available.
2) In order to assist in the development of database entries, once information about survey efforts has been collected and assessed, brief explanations should be assigned to each species regarding what we know (or don't know) about the habitat, range, and the adequacy of survey efforts. For example, when habitat and range are known, they should be documented as follows:
If either the habitat or range is poorly understood, it should be coded according to which of the two (or both) are in question:
Survey data should also be briefly explained after the assigned code, as shown in the two examples below:
3) In some cases, an extinct
species might only be known from one or more organisms that died only
relatively recently. Some researchers refer to this material as "fresh
dead" or a "death assemblage," rather than subfossil remains (which are
usually assumed to have been dead for a longer period of time). Remains
can usually be identified as "fresh dead" rather than subfossil based
"Fresh dead" material does not represent species that have been absent for a long period of time and, therefore, unlike species known only from subfossil or fossil material, these extinctions cannot be verified simply by the lack of evidence over a long period of time. Therefore, species known only from fresh dead material should be evaluated according to criterion 4 in the same way as putatively extinct species that are otherwise known from accounts of live specimens.
EXTINCTION CRITERION 5
For species known only from subfossil or fossil specimens, life history, habitat, and range cannot be documented. However, these extinctions can be verified simply by the lack of any other recorded evidence within the time period since the remains have been dated, and should be coded as S/FEO SNA = Subfossil/Fossil Evidence Only, Surveying Not Applicable.
EXTINCTION CRITERION 6
Evidence of biological threats to the species prior to the extinction date is not proof of extinction. However, if such evidence exists, it should be evaluated and documented, as this evidence may help us understand the factors involved with the demise of the species.
If any reliable documentation on threats can be found, then these documented threat(s) should be coded as follows and references and/or explanations should be provided:
If there is not any reliable information about threats then the species should be coded as TU = Threats Unknown.
(This is an example of how one might evaluate an extinction according to this new system of extinction criteria.)
Taxonomic Validity --Evaluate and code taxonomic status, providing an explanation and/or appropriate references if necessary.
Taxonomic Status: TE Lozano-Vilano, M.L. and S.C. Contreras-Balderas (1993). Four new species of Cyprinodon from southern Nuevo León, Mexico, with a key to the C. eximius complex (Teleostei: Cyprinodontidae). Ichthyological Exploration of Freshwaters 4: 295-308.
--Evaluate and code the hypodigm, providing an explanation and/or appropriate references if necessary.
Hypodigm: at least 26 (evaluated in 1993).
--Provide taxonomic information and relevant references for the following:
Species: Cyprinodon longidorsalis Lozano-Vilano and Contreras-Balderas, 1993
Common name: Charco Palma pupfish
(for species known from other than sub-fossil or fossil remains)
Cyprinodon longidorsalis is endemic to a single spring, Charco La Palma, described by Lozano-Vilano and Contreras-Balderas (1993) as having clear water, a muddy bottom, predominant algae and aquatic plants; the population size for the pupfish was estimated at 50 to 70 individuals (Lozano-Vilano and Contreras-Balderas, 1993; Contreras-Balderas and Lozana-Vilano, 1996). The spring and surrounding area have been surveyed on several occasions over the last 10-15 years by S. Contreras-Balderas and colleagues. The spring water was used for human consumption and for irrigation and during the period 1984-1994 the outflow from the pool dried up, the water level in the pool dropped consistently, and it appeared as though the banks of the pool had collapsed and the deeper parts were filled in and leveled. In 1994 no more fish were found in the pool and almost no vegetation was present. Information provided by Lozano-Vilano and Contreras-Balderas (1993) and Contreras-Balderas and Lozana-Vilano (1996) indicates that the habitat and range of Cyprinodon longidorsalis are well known (HK and RK) and have been adequately surveyed (AS).
--To assist in the development of database entries, provide a shortened explanation about what we know or do not know about the habitat, range and survey efforts.
Range: RK endemic to the single spring of Charco La Palma, Aramberri, southern Nuevo León, Mexico
Surveying notes: AS The spring of Charco La Palma and area around appear to have been surveyed on several occasions over the last 10-15 years by S. Contreras-Balderas and colleagues. Information provided by Lozano-Vilano and Contreras-Balderas (1993) and Contreras-Balderas and Lozana-Vilano (1996), indicates that the region has been adequately surveyed.
(for species known only from sub-fossil or fossil remains)
Documented Threats: HM Threats to Cyprinodon longidorsalis included habitat modification as a consequence of falling water levels in the pool to which the species was endemic.
in the CREO/IBOY Book?
Species: Cyprinodon longidorsalis Lozano-Vilano and Contreras-Balderas, 1993
Common name: Charco Palma pupfish
Taxonomic Status: TE
Hypodigm: at least 26
Effective extinction date: 1988 LC - 1994 EC
Habitat: HK clear water spring, with muddy bottom, predominant algae and aquatic plants
Range: RK endemic to the single spring of Charco La Palma, Aramberri, southern Nuevo León, Mexico
Surveying notes: HK AS spring of Charco La Palma and surrounding area around have been surveyed on several occasions over the last 10-15 years.
Documented Threats: HM
References: Lozano-Vilano and Contreras-Balderas (1993) and Contreras-Balderas and Lozana-Vilano (1996).
|Home | References Cited | Contact Us|