There are several reports available for general data reporting purposes, which may be useful in compiling and evaluating your studbook. In addition, SPARKS offers several analysis reports that look at specimen relationships and evaluate population demographics and to a limited extent, genetics. The more sophisticated genetics analyses are available through export of data to the separate software GENES.
1 - Specimen Report 2 - Studbook Report 3 - Questionnaire 4 - Analysis 5 - Relationships 6 - Master Plan Report 7 - Reporting to ISIS
For most reports you may choose to have a) just a screen display, b) printed on paper immediately, or c) deferred for later printing or saving as a DOS file. All of the reports may be terminated before completion by hitting the Escape (ESC) key to stop printing.
All reports are made much more powerful by use of the retrieval criteria selections and, for some, the sort order options. To set these criteria use the left and right cursor movement keys. Be sure to read the description of these capabilities below.
This is a report, often a single page, showing all the entered data for a specimen. It lists the current status, the history of transaction events, any special data, and any UDF’s entered. If you would like this report for every animal in your studbook, enter "ALL" when prompted for the individual studbook number. If you want specimen reports for a group of specimens, define the group by setting the selection criteria to create a view of your data that includes just the ones you wish, and then request "ALL".
This is a studbook report resembling published studbooks. It is also, of course, most necessary for listing all your data in one place. This is also important when trying to organize and edit a new:set data.
Some flexibility is offered to enhance your use of the studbook report. When you request this report, additional columns may be requested to display event types, birth origins, countries, a separate death column, rearing type, and the other specimen identifiers (tag/band/ring, tattoo, ear notch). There are different choices of format for dates, too. See the section on dates for more information.
Studbook reports, in order to squeeze the most information on a page, uses your printers condensed print mode. For this reason, they do not display on your screen.
The studbook questionnaire looks similar to the studbook report mentioned above. The difference is that it is, by institution, a listing of all specimens that have ever passed through each institution. Animals that
Reports and Analysis 13
have been in several zoos are reported on each zoo’s questionnaire. Missing information is indicated and requested by filling in the blanks or sending an ARKS Taxon Report. The institutions mailing address, if known to SPARKS, is printed at the top. This report also uses condensed print mode and is not displayed on the screen. Species Management Analysis
There are five items presently available under analysis. These are:
Age Pyramid: This is a text graphic displaying the current age structure of your studbook population. Males are listed on the left, females on the right. Unknown sex animals are equally split to both sides. A "good" population should have a pyramid shape to the graph, with a wider base of younger animals which can move up and replace older animals in future years. Setting a "view" will produce an Age Pyramid of the population defined by the view.
Fecundity and Mortality: Fecundity [Mx] is a measure of reproduction. It shows, by one year age classes, the proportion of births per individual per age class in your population. Mortality [Qx] tabulates risk of death by age classes. Survivability [Sx], often seen, is one minus the mortality. To print the raw data and a description of what goes into the values, type F1 after the charts are produced. For those with a graphics board in your computer, these reports also offer a line graph of the two parameters.
Inbreeding Coefficients: This option offers rapid calculation of simple inbreeding coefficients for existing animals and for possible matings. It is based on the routine FASTINB, written by Dr. A. J. Boyce. More sophisticated genetic analysis should be accomplished by export of data to GENES, written by Dr. Robert Lacy.
Inbred animals often have reduced viability. This report shows the amount of inbreeding of individuals based on the known pedigree. Note that any specimens with unknown parentage or multiple possible parents (entered as MULT) will be treated as founders in this analysis, causing underestimation of inbreeding coefficients in their descendants. Hypothetical matings may be entered, when the report is concluded, to assist in possible mating choices. Keep in mind, however, that two animals that are highly inbreed may be crossed to produce offspring that are not inbred. This analysis is very dependent on complete pedigrees.
Census: A list of the number of living animals by year comprises a census. The census tables presented are broken down by sex. A bar chart, showing the last forty years graphically, is available if your equipment displays graphics.
Export Data Analysis File: There are several cutting-edge data analysis programs available to do further work on your animal management. These require data to be pre-processed in various ways. There are two export options to produce the needed pre-processed export files: demographic analysis, which includes Mx and Qx and age structure, and genetic analysis, which is a me,pa,ma file.
There are four reports within the relationships section. These include:
Pedigree Chart: The ancestors of a given animal are printed back four generations, in a simple text graphic form. If the number of generations goes back more than four, you will need to run several pedigrees and piece the chart together. Note that individual specimens may appear more than once on the diagram. If so, they are marked with an asterisk.
Sibling Tables: All full, as well as half, siblings are displayed. They are grouped by date, which should match up litter mates.
Reports and Analysis 14
Reproductive History: All offspring of a requested animal are listed. These are also grouped by date. The parents age at which the birth occurred is also indicated.
Descendent Lists: All descendents and their descendents. This is a potentially massive list of offspring that is the reverse of the pedigree chart. Indentations help in discriminating first and later-generation descend-ents. Sex, age, death date and/or current location are shown, along with the identity of the other parent.
When all the data is in, it is time to sit down and draw up a Master Plan for your animal management. The Master Plan Report is an attempt to bring together in one place many of the "more useful” values that are calculated in other reports. As species management concepts are refined, new "more useful" values are sure to be invented. It is hoped that later versions of SPARKS will incorporate these new ideas.
Important parts of the studbook effort are collecting quality data and making it available to others to affect management and research. The international Species Information System (ISIS) represents a central archive and distribution point for zoological specimen data - with information on hundreds of thousands of specimens from several hundred zoological facilities on 6 continents already assembled at ISIS.
For about 97% of captive species, ISIS datasets are the only broad-scope assembled specimen data available. ISIS datasets are of course limited in time and space to information contributed to ISIS. For the other 3% of species, studbooks exist - something we can be grateful for, as studbooks are typically more comprehensive in time and geography than ISIS datasets.
If you create, or materially improve, a studbook dataset for some taxon, you can add to the long-term value of your work by contributing the results to ISIS. Incorporating your data into the ISIS database will provide additional benefits to the large community which is served by ISIS.
To send your studbook data to ISIS run the "ISIS" report. This report will write a file, to be copied to a floppy disk, of any changes made to your studbool« When starting a new studbook, wait until you are comfortable with your data and it seems to be up to date. Once you have assembled your data and you are in maintenance mode, send ISIS the data after any major changes, such as your annual questionnaire.
One powerful approach to population data assembly and analysis is to look at and compare various subsets with each other. SPARKS is designed to make this very straightforward, by setting retrieval criteria - or "views" of your data. A view may consist of any number of set criteria, in any combination. Once you have established a "view" - defined a subset of interest, it stays in effect until you reset it or leave SPARKS.
You may select on any combination of the following:
| Geographic Area | Event Type | Rearing |
| Date | Cooperative Management Plan | User Defined Field |
| Age | Birth Origin | User Selected Flag |
| Sex | Founder |
Reports and Analysis 15
Geographic Area: During your data entry or data conversion through LOADEDIT, SPARKS maintained close control over location names and assured consistency with its internal location name dictionary. This standardization has its rewards here. You may restrict your view of the data to a list of geographic areas. These may be a single zoo or a dozen zoos. It may also be states, provinces, countries, continents, or any combination. SPARKS generally knows the hierarchical geographic relationships from single locations on up to continents, and will recognize many names of political and/or geographic regions. If you have not entered a specimen’s location in a way SPARKS can recognize, and SPARKS can’t determine if it should be included within your selected subset, you may be prompted as to whether or not to include this data at report time. If this happens so frequently that it becomes a nuisance, you will want to edit the location names for these specimens to become something SPARKS can recognize - or you may need to add them to the SPARKS institution list.
Date: Viewing your data through a date "window" or "view" is an important tool - you can compare fecundity and mortality during different periods of time, for example. You may set the criteria for starting date and/or ending date to form a date span. For example: you may analyse the last five years and compare it to the five years before. To un-set a date view, either change the dates to 01/01/1800 and today or exit SPARKS and re-enter SPARKS.
Age: You may also specify an age window. All animals currently (hence living) between the starting age and ending age are included in your subset.
Sex: There are five categories of sex (female, male, sterilized female, sterilized male, and unknown) that may be used to limit the data reported and/or analyzed.
Event Type: These limit you to such occurrences as births, captures, transfers, deaths, and releases. This is perhaps most useful for printing lists of births, moves, or deaths in studbook format.
Cooperative Management Plan: If a specimen has been recorded as being in a management plan, either globally or regionally, you may select it for reporting. It is obvious that with this you can compare animals in a management plan with those not in a plan, or compare two plans, presuming you have this data in your studbook dataset.
Birth Origin: There are three choices of birth type: captive born, wild born, and unknown birth. By setting the "view" in turn, you can compare, for example, the fecundity and mortality of wild-caught versus captive-bred specimens.
Founder: You might wish to select for founders to check your studbook dataset before performing genetic analyses - to see just which specimens will be treated as founders. Founders are not just those animals that were obtained directly from the wild. For genetic calculation purposes, they are those animals whose parents are listed as wild ("WILD") or unknown ("UNK"), or one of several possibilities ("MULT"). A founder is a "potential founder" if it has not bred. See below to select on breeders.
Rearing: Choose one or more rearing types if you wish to examime the effects of different rearing situations and your data contains this information.
User Defined Fields: UDFs add a lot of power to your studbook if set up thoughtfully. See the section on UDFs for more information. By using UDFs you may select only data based on your special requirements, be it hybrid, hernia, or health.
User Selected Flag: This is the most flexible of all retrieval criteria selections. With this option you may set your "view" to a list of specific studbook numbers. You can create this list any way you choose -
Reports and Analysis 16
including saving some SPARKS report as a DOS file and editing it down to a list of specimens.
This option requires a DOS file of specimen studbook numbers, one per line,to be read by SPARKS. You must create this before entering SPARKS. Once read in, a flag is set to indicate which animals are in your subset. You may select to use or not use this flagged subset in your retrieval. Reading in another list of numbers sets the flag to the new set.
You may even run a SPARKS report, say the Reproductive History Report, and save it as a deferred print file on disk. Then, after some editing in DOS, come up with a list of breeders to be read in to set the user selected flag. Combine this with the Founders selection and you have only your reproductive founders. Editing a saved Descendant List Report could give you a list of Fl offspring, etc. The uses of this option are limited only by your needs, skills, and imagination. Sort Order
The default order of presentation in a studbook report is by studbook number. Because the data is stored in studbook number order in SPARKS, this is also quickest. However, you may request any studbook or other listing report in a number or different orders. Before each report the data must be sorted, causing a short delay. The following sort orders are available:
| Studbook Number | Owner |
| Birth Date | Sex |
| Death Date | Birth Type |
| Location | Management Plan |
Studbook Number: The normal and default production order for reports.
Birth Date: Every animal has some birth date in SPARKS, even if it is only estimated. Studbook numbers are often assigned in approximate birth date order, but not always. This gives you the option of sorting the report strictly on birth date.
Death Date: Specimens are listed in order of death. Living specimens are listed last, by studbook number, as they have nothing in the death field.
Location: Animals are presented in order of current or last location. This is a geographic location, not alphabetic name sort.
Owner: A list by current or last owner. This is different from the above location order only if an animal has been loaned out. If you wish only living animals (i.e. current owner, not last), add the criteria "living".
Sex: Females, males, and unknown.
Birth Type: Captive born, wild born, and unknown births.
Management Plan: Those in a management plan are listed first.
Reports and Analysis 17
Specimen Report
INDIAN RHINO Studbook
===========================================================================
Taxon Name: RHINOCEROS UNICORNIS Studbook Number: 134
===========================================================================
Current Status >>>
Vital Statistics >>>
Sex: Female
Age: -18Y Est. to year
Origin >>> Identifiers >>>
Birth Type: Captive Born House Name: TUBBS
Location: METROZOO Tattoo: LL-134
Birth Date: ~ 1971 Est. to year Tag/Band:
Sire Id: 14 Transponder ID:
Dam Id: 11
Rearing: ______ Global Management Plan: No
Transaction History >>>
# Event Local ID Date In/Date Out
--------------------------------------------------------------------------
1 Birth METROZOO UNK ~ 1971 Est. to year
METRO ZOO
2 Transfer to PALM DES 134 16 Feb 1988
THE LIVING DESERT
3 Ownership to NZP-WASH 134 2 Mar 1988
NZP DEPTS. MAMMALOGY, ORNITH. AND HERP.
4 Transfer to LOSANGELZ 134 10 Mar 1988
LOS ANGELES ZOO
5 Loan to CHICAGOBR 134 10 Mar 1988
CHICAGO ZOOLOGICAL PARK (BROOKFIELD ZOO)
6 Transfer to MOSCOW Z UNK Unknown
SOJUSGOSZIRK MOSKWA Unknown
Special Data & Comments >>>
Code/Text Date
--------------------------------------------------------------------------
1 House Name while at METROZOO 1 Jan 1971
TUBBS
2 Tattoo while at METROZOO 1 Jan 1971
LL-134
PERISSODACTYLA ISIS/SPARKS
RHINOCEROTIDAE 3 May 1989
Examples of SPARKS Reports 18
Examples of SPARKS Reports 19
Examples of SPARKS Reports 20
Examples of SPARKS Reports 21
Examples of SPARKS Reports 22
This report presents life table information. Age- and sex-specific fecundity and survivorship tables are calculated and presented. Following this you have options for generating graphs, and/or smoothing the data and producing smoothed tables and/or graphs. Commonly used indicators of population growth are also presented, as is a calculated N,. Caution is advised because of the small sample sizes typically available. Effective population size is not easy to calculate accurately - be sure to check this value against the loss of heterozygosity over this many generations shown by the simulation program GENES, which is co-distributed with SPARKS. Consult the Glossary for definitions of terms and parameters. Below is an extended discussion of how missing data is handled - an important issue in the analysis of many datasets.
We calculate the vector M, for each sex as the ratio of the cells of two vectors - the numerator is the number of births credited to parents at each age class, the denominator is the number of specimens which passed through this age class while in our data set.
Missing data, such as an un-identified parent of a captive born specimen, can cause significant bias in the calculation of M,. In the case mentioned, no parental age class can be readily credited with the birth, though the unidentified individual is presumably included in the group of individuals "at risk" of birth. Hence, simply throwing out this birth leads to an underestimate of M, - the parent contributes to the de-nominator but the birth does not contribute to the numerator. Similarly, a parent of unknown age leads to the same problem as an unidentified parent - we don’t know what parental age class to credit with the birth.
For these cases where it is a good presumption that the parent is actually in our data set, but it either has an unknown age or is not identified, we are taking the approach of assuming that they have the same age-specific fecundity values as identified or known-age parents. Therefore, we provide the option of adding the births to such parents into the numerator of the M, vector, distributing the births across parental age classes using the higher-quality data to calculate the weighting factor for each age class. We generally recommend this option, as otherwise many data sets with significantly imperfect Sire and Dam data, will show erroneously low fecundity values, and therefore erroneously low growth rates. It is likely dangerous only when there is something very different about such parents. A good general common-sense procedure is to compare the calculated growth rates with the trend shown by the Census Report.
In practice, the M, calculation utilizing imperfect data requires inference of whether or not the unidentified parent was likely in the dataset as an individual at the time of the birth.
Because SPARKS offers extensive ability to subset data for analysis - i.e. to set a "View" of the data to a region, period of time, etc., similar issues arise in determining whether parents were "at risk" of birth in the restricted sense of being in your currently defined view of your data. The rules SPARKS uses for M, calculation are:
In general, SPARKS uses estimated dates and accurate dates as "known" dates.
IF the specimen’s birthdate is unknown, SPARKS skips it. It will not be added to the denominator "risk" vector as a potential parent, and this birth to its parents will not be added to the numerator vector.
IF a specimen’s last removal (including death) date is unknown, it will only be treated as "at risk" (added to the denominator vector) if there is an earlier known transaction date - and then only for the period between the first known date and this last known date.
Examples of SPARKS Reports 23
IF a parent is identified, then:
IF it is not found in the dataset, SPARKS excludes the birth because the parent is not included in the group "at risk" for birth. "WILD" parents are treated this way.
IF parent is UNK (or MULT), then:
Else if the birth event is not recorded in the dataset, but the birth type is given as captive born, the missing parent is fairly likely to be in the dataset, so we include the birth. This is not a strong inference.
IF the specimen’s birth event is not recorded in the dataset and the birth type is given as wild or unknown, the missing parent is unlikely to be in the dataset, so we exclude the birth.
Data sets read in from "Omaha" format or "Houston" format do not contain a birth-type indicator. The LOADEDIT program distributed with SPARKS sets birth type to wild (2) when both parents are identified as WILD, sets birth type to captive (1) when both parents are identified and/or when the birth event is recorded in the data, and sets birth type to unknown (X) when both parents are blank or UNK and in any other cases. The consequence of these conversion rules working with the M, inference rules is that (probably rare) individuals of unidentified parentage who are captive-born but whose record chain starts somewhere after their birth will have their birth included in the pool to be distributed only if the user manually edits their birth-type indicator to captive.
Qx is calculated using the same general approach as M, - two vectors are created: the numerator vector consisting of the various possible age classes for deaths with the number recorded in each, the denominator vector consisting of the various age classes at risk to death - with the number recorded in each.
Any known or estimated aged specimen is included.
If a specimen birth date is unknown, it is skipped and does not contribute to either the numerator or denominator.
If a specimen death date is unknown, it does not contribute to the numerator, but may contribute to the denominator for the years "lived through" to the last known transaction date. This is consistent with the overall approach of assessing what part of a specimen’s history is at risk in the dataset and within any view which has been created, and only including birth or death events which also qualify.
Also see the Special Note for Death Events box in the manual section dealing with Entering Studbook Data.
Examples of SPARKS Reports 24
Examples of SPARKS Reports 25
Examples of SPARKS Reports 26
Examples of SPARKS Reports 27
Examples of SPARKS Reports 28
Examples of SPARKS Reports 29
Examples of SPARKS Reports 30
Examples of SPARKS Reports 31
Examples of SPARKS Reports 32
The Masterplan Report is organized Institution-By-Institution and then Animal-By-Animal to facilitate formulation of recommendations for each individual in the population during development of a masterplan for managment and propagation. The Report produces separate sheets for each institution. Within the sheet(s) for an institution, the Report enumerates each living individual. Only living animals are included in the Masterplan Report. The Masterplan Report may be produced for the entire (Global) studbook data set or for a defined geographic region (Restricted), e.g. North America. For each individual in the list, there are 11 columns of potential information that may be useful for masterplan formulation.
The first 2 columns each contain a single entry.
The next column may contain either or boih of 2 dates:
The next column indicates the Sire and Dam, if known, of the individual.
The next column provides information on the local identification number and the social group of the individual.
The next 3 columns present genetic metrics (measures) that can be calculated in two ways:
The 3 genetic metrics are:
32A
The next column provides another genetic characteristic of the individual:
The next column is a demographic characteristic of the individual:
The next column provides information on the number of siblings and offspring of the individual. The data are presented as either:
The final column is yet another genetic characterization of the individual.
32B
[ Table of Contents | next page ]