
===== PAGE 1 =====
Page 1 of 32
Global Representation of the World’s Most-Cited 
Forestry Scientists by Number of Scientists, 
Publications and Citations
A Five-Year, Multi-Metric Geographic and Institutional Analysis (2020–2024) of the 
Stanford / Ioannidis Career-Long Citation Database
Analytical Report
Prepared 6 July 2026
Data source: Ioannidis, J. P. A. (2025), Elsevier Data Repository, doi:10.17632/btchxktzyw (v3–v8)
Abstract
This report quantifies the geographic and institutional representation of the world’s most highly cited forestry 
researchers for each annual release of the Stanford / Ioannidis “top-scientists” database from data years 
2020 to 2024, resolving every population’s output by authorship position. The cohort—termed “Forestry plus 
Ecology”—is assembled from two mutually exclusive groups, both of which require Forestry to be present: 
“Forestry” (primary Science-Metrix subfield = Forestry; 3,271 records) and “Ecology” (primary subfield = 
Ecology with secondary subfield = Forestry; 615 records), together 3,886 author-year records analysed as 
two separate groups throughout. In 2024 these 854 scientists carry 102 thousand career publications and 4.9 
million career citations. The Data section reports only raw counts—for every division, the number of 
scientists and of solo (single-authored), senior (single- or first-authored) and all publications; the Results 
derive, for each of those three authorship classes, three rates—publications per author, citations per author 
and citations per publication—a 3 × 3 scheme of nine measures per entity, shown as 3 × 3 charts. No raw 
citation total is reported. The central findings are that the United States leads the field by scale and is also 
among the most cited per paper of the large producers, whereas China and Japan show the opposite high-
volume, low-per-paper-impact profile; that per-paper impact tracks discipline and geography rather than how 
much a population publishes, and holds its solo–senior–all ordering across regions; that the ecology-rooted 
group is about twice as cited per paper as the forestry-rooted group at every authorship level; and that 
ranking scientists by undiscriminated citations measures cited collaboration more than individual authorship
—narrowing to solo- or first-authored credit reorders them substantially. A multi-root taxonomy tree organizes 
the cohort by field → primary subfield → secondary subfield, and Appendix A profiles the Ecology group 
(criterion 2) on its own.
Keywords: forestry; ecology; bibliometrics; scientometrics; citation indicators; authorship position; solo and 
senior authorship; citations per paper; Stanford/Ioannidis list; research geography; Science-Metrix taxonomy; 
choropleth mapping.

===== PAGE 2 =====
Page 2 of 32
1. Background, Concept and Related Work
Citation-based rankings of individual scientists are widely used in research evaluation. The most 
common single statistic, the h-index [1], conflates productivity with impact and scales with career length, 
field and co-authorship. To standardize comparison, Ioannidis and colleagues devised a composite 
indicator (the “c-score”) combining six field-normalized metrics [2,3] and published an openly licensed 
database ranking about the top 100,000 scientists (and the top 2% of each subfield) across 22 fields 
and 174 subfields defined by the Science-Metrix journal ontology [3,4]. Each ranked author carries a 
primary subfield (sm-subfield-1), an optional secondary subfield (sm-subfield-2), and a field-level label 
(sm-field). The database appears in career-long and single-year forms and is refreshed annually from 
Scopus. This report analyses the career-long tables.
Whereas the original database is typically used to look up individuals, this report treats forestry as a 
population and asks how that population is distributed, not only by head-count but also by the 
publications and citations those researchers carry. Forestry is a small, applied, globally distributed field 
at the centre of timber economies, biodiversity, carbon and land-use policy. Because forestry and 
ecology overlap heavily in journals, methods and personnel, a subset of the most-cited forestry 
researchers are classified primarily as ecologists but secondarily as foresters; the report keeps these as 
a distinct “Ecology” group alongside the primary-forestry “Forestry” group, and separates the two 
throughout. Both groups, by construction, contain only authors who work in forestry—there is no 
ecology-without-forestry group here.
2. Data Description
The primary data are the five career-long author tables (“Table_1_Authors_career”) from dataset 
doi:10.17632/btchxktzyw, versions 3, 5, 6, 7 and 8, for data years 2020–2024. Each ~75–85 MB 
spreadsheet holds one row per ranked scientist (about 186,000–230,000 rows) with the author name, 
one large-institution affiliation, its ISO country code, the six component citation metrics and composite 
c-score, the career publication count (np) and career citation count (nc), authorship-resolved counts—
single-authored papers and their citations (nps/ncs), single-or-first-authored (cpsf/ncsf) and single-first-
or-last-authored (npsfl/ncsfl)—and the Science-Metrix classification: primary subfield (sm-subfield-1, 
column AM), secondary subfield (sm-subfield-2, AO) and field (sm-field, AQ). All counts are on the 
Scopus basis and inherit its journal-centric coverage; Section 5 (Table 10) validates that coverage 
externally against Google Scholar for two cohort members and bounds the effect on the report’s 
conclusions.
Selection. A record enters the cohort under either of two rules, kept as separate groups: (1) “Forestry” 
— sm-subfield-1 = Forestry (3,271 records); (2) “Ecology” — sm-subfield-1 = Ecology AND sm-
subfield-2 = Forestry (615 records). Both rules require Forestry, so the two groups are mutually 
exclusive and neither admits an author who is in ecology but not forestry; their sum, 3,886 author-year 
records, is what we call “Forestry plus Ecology.” We verified against the raw releases that sm-field (AQ) 
is a field-level label taking only the 20 Science-Metrix field names and is never literally “Forestry” (the 
forestry field value is the broader “Agriculture, Fisheries & Forestry”); AQ therefore adds no records 
beyond the two subfield rules but is used as the set of roots for the taxonomy tree (Figure 3, at the end 
of this section). The cohort is delivered with this report as an Excel workbook (full per-year, per-metric 
rankings and per-group sheets) and CSV.
This section presents only the raw selected counts: for every division, the number of listed scientists 
and the number of solo (single-authored), senior (single- or first-authored) and all publications, summed 

===== PAGE 3 =====
Page 3 of 32
over the five releases (Tables 1–7). These are the data as selected; they involve no derivation beyond 
counting and classifying by authorship position. Every per-author and per-publication rate—how much 
each population publishes per scientist, and how heavily its papers are cited—is a derivation and is 
reported in the Results (Section 4), with the underlying definitions in the Methods (Section 3). Figure 1 
shows the cohort’s per-year growth; full lists and the Canadian and Australian subdivisions are in the 
companion workbook.
Table 1. “Forestry plus Ecology” cohort (Forestry 3,271 + Ecology 615 = 3,886 author-year records) by 
release year: scientist and publication counts. Here “Scientists” is the number of listed author-year 
records in each release (not a five-year average); Solo/Senior/All are the publication counts. Scientists 
= five-year average of listed author-year records; Solo = single-authored, Senior = single- or first-
authored, All = every publication — each count summed over the five releases. These are raw selected 
counts; every per-author and per-publication rate is a derivation, reported in Sections 3–4.
Release year
Scientists /yr
Solo pubs Senior pubs
All pubs
2020
709
5,777
19,985
73,804
2021
735
5,975
21,062
80,137
2022
769
6,262
22,459
87,418
2023
819
6,579
24,015
96,216
2024
854
6,850
25,125
102,096
Figure 1. Growth of the cohort’s solo, senior and all publication counts since 2020 (percent change from 
the 2020 release, with the actual paper count labelled at each point). A per-year, descriptive view of the 
selected data.
Composition by group. The cohort is 84% Forestry group (3,271 records) and 16% Ecology group 
(615). Table 2 gives the scientist and publication counts of the two groups; the field composition is 
shown by the taxonomy tree (Figure 3) at the end of this section. How the two groups differ in per-
author output and per-paper impact is a derivation and is taken up in Section 4.1.
Table 2. Scientist and publication counts by group — Forestry (criterion 1) vs Ecology (criterion 2), five-
year values. Scientists = five-year average of listed author-year records; Solo = single-authored, Senior 

===== PAGE 4 =====
Page 4 of 32
= single- or first-authored, All = every publication — each count summed over the five releases. These 
are raw selected counts; every per-author and per-publication rate is a derivation, reported in Sections 
3–4.
Group
Scientists /yr
Solo pubs Senior pubs
All pubs
Forestry
654
24,343
89,539
349,175
Ecology
123
7,100
23,107
90,496
Geographic tables. Tables 3–5 give continent, world sub-region and country counts. North America and 
Europe supply most of the cohort’s output; the United States and Canada lead among countries, with 
Finland and Sweden strongly over-represented, and Central Europe (Germany, Austria, Switzerland, 
Poland, Czechia, Slovakia, Hungary, Slovenia) broken out as its own sub-region, ranking third.
Table 3. Scientist and publication counts by continent (five-year values).
Continent
Scientists /yr
Solo pubs Senior pubs
All pubs
North America
387
14,547
55,686
208,209
Europe
281
10,730
36,952
162,760
Oceania
54
2,935
10,176
30,661
Asia
37
1,516
6,359
26,347
South America
10
906
1,719
6,265
Africa
8
781
1,702
5,332
Table 4. Scientist and publication counts by world sub-region (top 8 by output). Central Europe is 
broken out of the UN Western/Eastern/Southern Europe groupings.
World sub-region
Scientists /yr
Solo pubs Senior pubs
All pubs
Northern America
377
14,032
54,150
203,553
Northern Europe
133
5,987
17,423
69,356
Central Europe
79
2,793
10,656
47,744
Australia & New Zealand
54
2,935
10,176
30,661
Southern Europe
34
974
4,885
23,634
Western Europe
35
964
3,869
21,157
Eastern Asia
25
1,136
4,121
19,592
South America
10
906
1,719
6,265
Table 5. Scientist and publication counts by country: top 25 by output (five-year values).
Country
Scientists /yr
Solo pubs Senior pubs
All pubs
United States
275
10,276
40,204
142,390
Canada
101
3,756
13,946
61,163
Finland
56
1,608
6,754
34,727
Australia
46
2,607
8,669
26,720
Germany
39
1,758
5,737
26,526
Sweden
39
1,734
5,041
18,759
France
21
552
2,208
11,627
Spain
16
425
2,357
11,482
Switzerland
15
490
1,825
9,727
United Kingdom
25
2,344
4,446
9,613
China
8
271
1,316
8,221
Japan
13
509
1,966
7,835
Netherlands
12
380
1,478
6,656

===== PAGE 5 =====
Page 5 of 32
Country
Scientists /yr
Solo pubs Senior pubs
All pubs
Portugal
5
160
653
5,684
Italy
9
307
1,499
5,017
South Africa
5
462
1,276
4,195
New Zealand
8
328
1,507
3,941
Brazil
5
655
1,018
3,925
Austria
8
219
965
3,913
Norway
6
166
647
3,164
Belgium
2
32
183
2,874
Denmark
6
97
363
2,543
Slovenia
5
63
572
2,456
India
3
117
562
2,122
Turkey
4
112
629
2,064
Subnational and institutional tables. Table 6 gives the leading U.S. states and Table 7 the most-
represented institutions (with country). The most-represented single institution is USDA Forest Service 
(United States). Full per-state tables and the complete institution list are in the workbook.
Table 6. Scientist and publication counts by U.S. state (top 15 by output).
U.S. state / category
Scientists /yr
Solo pubs Senior pubs
All pubs
Federal/Multi-site
72
2,234
9,466
29,638
North Carolina
17
1,118
3,122
12,194
Oregon
26
622
3,561
11,759
Wisconsin
15
496
2,498
9,614
California
12
452
2,182
7,168
Maine
7
213
923
5,817
Unknown
11
357
1,428
4,714
Washington
10
268
992
3,988
Florida
7
424
1,059
3,854
Missouri
3
68
646
3,610
Michigan
8
241
960
3,445
Minnesota
5
196
894
3,369
Tennessee
5
158
756
3,057
New York
6
248
796
3,029
Arizona
4
175
789
2,956

===== PAGE 6 =====
Page 6 of 32
Figure 2. Solo, senior and all publication counts by U.S. state — five-year choropleths (contiguous 
states; Alaska and Hawaii omitted from the frame). Boundaries: a public U.S. state file.
Table 7. Scientist and publication counts by institution: top 25 by output.
Institution (country)
Scientists 
/yr
Solo 
pubs
Senior 
pubs
All 
pubs
USDA Forest Service (United States)
30
840
3,819 12,262
The University of British Columbia (Canada)
18
1,086
3,066 11,476
Sveriges lantbruksuniversitet (Sweden)
23
1,039
2,851 11,289
NC State University (United States)
11
445
1,824
9,836
Natural Resources Institute Finland (Luke) (Finland)
20
474
2,264
9,821
Oregon State University (United States)
20
489
2,754
9,721
Itä-Suomen yliopisto (Finland)
10
325
1,369
7,617
Georg-August-Universität Göttingen (Germany)
9
352
1,107
7,241
University of Toronto (Canada)
10
333
1,075
5,989
Helsingin Yliopisto (Finland)
11
441
1,440
5,694
Wageningen University & Research (Netherlands)
10
278
1,090
5,443
University of Maine (United States)
6
183
818
5,186
Eidgenössische Forschungsanstalt für Wald, Schnee und Landschaft 
WSL (Switzerland)
9
306
1,003
5,117
Canadian Forest Service (Canada)
10
423
1,527
4,949
USDA Forest Products Laboratory (United States)
7
202
1,191
4,636
University of Alberta (Canada)
7
115
720
4,470
University of Wisconsin-Madison (United States)
6
255
1,083
4,330
University of New Brunswick (Canada)
5
170
636
4,133
Université Laval (Canada)
6
134
656
3,817
The Fenner School of Environment & Society (Australia)
2
166
1,191
3,706
University of Washington (United States)
8
251
893
3,601
Universität Freiburg (Germany)
5
225
577
3,522
Technische Universität München (Germany)
4
319
1,013
3,424
USDA Forest Service Rocky Mountain Research Station (United States)
9
216
1,118
3,369
Aalto University (Finland)
3
104
421
3,334
Field composition (taxonomy). The two groups occupy visibly distinct branches of the Science-Metrix 
taxonomy (Figure 3). Its multiple roots are the fields (sm-field): the Forestry group appears as gold 
“Forestry” subfield nodes (mostly under the Agriculture, Fisheries & Forestry root, some under Biology 
and other fields), while the Ecology group appears as the teal “Ecology” node under the Biology root 

===== PAGE 7 =====
Page 7 of 32
branching to a secondary Forestry leaf—showing that even a strictly forestry-defined population 
reaches across many fields.

===== PAGE 8 =====
Page 8 of 32
Figure 3 (part 1 of 2). Multi-root taxonomy of the forestry research cohort — the two principal roots 
(Agriculture, Fisheries & Forestry; Biology): field / sm-field (roots) → primary subfield / sm-subfield-1 
(gold = Forestry group, teal = Ecology group) → secondary subfield / sm-subfield-2 (blue » = 
quantitative). Right-hand numbers are author counts; minor branches aggregated. Part 2 (remaining 
roots) follows.

===== PAGE 9 =====
Page 9 of 32

===== PAGE 10 =====
Page 10 of 32
Figure 3 (part 2 of 2). Remaining cross-field roots of the taxonomy (Enabling & Strategic Technologies, 
Chemistry, Earth & Environmental Sciences, Economics & Business): field / sm-field (roots) → primary 
subfield / sm-subfield-1 → secondary subfield / sm-subfield-2. Same colour key as part 1; roots with 
fewer than 15 authors omitted (see note).
3. Methods
Population and groups. The cohort is the union of the two selection rules above; every table and figure 
distinguishes the Forestry group (primary subfield Forestry) from the Ecology group (primary Ecology, 
secondary Forestry). The overall geographic and institutional analyses pool both groups; the group split 
is in Table 2 and Section 4.1.
Raw inputs (the Data section). Three quantities are read directly from the source and simply counted, 
per division and per release year: the number of listed scientists (author-year records) and, resolving 
each author’s output by position, the number of solo (single-authored, dataset fields nps/ncs), senior 
(single- or first-authored, cpsf/ncsf—the papers an author most directly drives) and all (np/nc) 
publications. Section 2 tabulates these counts and nothing else; classifying a paper by authorship 
position is a lookup, not a derivation.
Derived measures (the nine numbers, computed in Results). Every rate is formed in pre-processing 
from the counts above and reported in Section 4. For each of the three authorship classes—solo, 
senior, all—three measures are computed over the five-year author-year sums: publications per author 
(Σclass-publications ÷ Σscientists, the output a listed scientist carries), citations per author (Σclass-
citations ÷ Σscientists, total impact per listed scientist) and citations per publication (Σclass-citations ÷ 
Σclass-publications, per-paper impact). That is a 3 × 3 scheme—(solo, senior, all) × (per author, per 

===== PAGE 11 =====
Page 11 of 32
author, per paper)—giving nine derived measures per entity; the report presents them as 3 × 3 charts 
(Section 4). No raw citation total is displayed. Because career values are cumulative and each release 
re-lists largely the same people, sums are over author-years and the scientist count used as a 
denominator is the five-year total of listed records.
Authorship position, resolved further (Section 4.6). Beyond the nested solo/senior/all classes, each 
author’s output splits into three mutually exclusive positions—single-authored, first-authored (of a multi-
author paper) and other co-authored (every paper on which the author is neither sole nor first author). 
Last-authorship is not treated separately: this study asks about the work a scientist individually or 
leadingly produces (solo and senior), so all non-lead co-authorship is pooled as “other.”
Geography. Continent, UN M49 world sub-region and country were derived from the ISO affiliation 
code. For the United States, Canada and Australia, each institution was resolved to a first-level 
subdivision by geolocating its main campus (a verified language-model pass); distributed federal 
agencies are pooled as “Federal/Multi-site” and unplaceable ones as “Unknown.” A few source records 
carry no affiliation country code or no institution name; they are included in all cohort totals but cannot 
appear in the geographic or institutional breakdowns, so those tables can sum marginally below the 
totals.
Authorship position. The database also reports, for each scientist, the number of and citations to their 
single-authored papers (nps/ncs), their single-or-first-authored papers (cpsf/ncsf) and their single-first-
or-last-authored papers (npsfl/ncsfl). Because these sets are nested, Section 4.6 differences them into 
four mutually exclusive positions—single, first, last and middle—and uses them to test how a scientist’s 
rank depends on which authorship positions are credited. All position counts use the all-citations basis, 
consistent with np and nc.
Taxonomy and tools. The taxonomy tree (Figure 3) has multiple roots given by the field values (sm-
field), each branching to the primary subfield (sm-subfield-1 = Forestry or Ecology) and then to the 
secondary subfield (sm-subfield-2); this places the two groups on visibly distinct branches. Complete 
rankings are in the companion workbook. Values were computed with Python (pandas); figures with 
matplotlib; maps use Natural Earth boundaries [6]. All totals and ratios were independently re-verified.
4. Results — the nine derived measures of output and impact
The Data section gave the raw counts. This section derives, for every population, how much its listed 
scientists publish and how heavily that work is cited, resolved by authorship position. Each entity is read 
on nine measures—the three authorship classes (solo, senior, all) crossed with three rates (publications 
per author, citations per author, citations per publication)—presented as 3 × 3 charts. Two regularities 
run through all of them and are worth stating first. (i) Per-paper impact peaks at the lead, not the sole, 
position: across the cohort solo papers draw about 37 citations each and all papers 44, but the senior 
(single- or first-authored) set sits highest at 46, because first-authored papers—an author’s own 
empirical and methodological work—are their most-cited. (ii) Output volume and per-paper impact are 
close to independent, so a population’s standing depends entirely on which of the nine numbers is 
asked.
4.1 The cohort and the two groups
Table 8 gives the cohort and the two groups on all nine measures, and Figure 4 shows the group 
contrast cell by cell. The ecology-rooted group leads the forestry-rooted group on every one of the nine: 
it publishes more per author (all 147 vs 107 papers), draws far more citations per author (all 10.7k vs 

===== PAGE 12 =====
Page 12 of 32
3.9k) and is about twice as cited per paper at every class (solo 64 vs 29, all 73 vs 37). Within each 
population the citations-per-publication row rises from solo to senior then dips for all—the lead-position 
peak noted above—while publications and citations per author climb monotonically as the class widens, 
since wider classes contain more papers.
Table 8. The cohort and the two groups on the nine derived measures (five-year values), each 
population resolved into solo, senior and all authorship classes. Each authorship class—Solo (single-
authored), Senior (single- or first-authored) and All—carries three derived measures: publications per 
author (p/au), citations per author (c/au) and citations per publication (c/pub), each formed from five-
year sums over the listed author-year records. No raw citation total is shown.
Population and authorship class Pubs / author
Cites / author
Cites / pub
Whole cohort — solo
8
301
37
Whole cohort — senior
29
1,338
46
Whole cohort — all
113
5,005
44
Forestry group — solo
7
219
29
Forestry group — senior
27
1,025
37
Forestry group — all
107
3,925
37
Ecology group — solo
12
741
64
Ecology group — senior
38
3,003
80
Ecology group — all
147
10,745
73

===== PAGE 13 =====
Page 13 of 32
Figure 4. Forestry vs Ecology group — the 3 × 3 scheme: rows are the authorship classes (solo, senior, 
all), columns the three rates (publications per author, citations per author, citations per publication), five-
year values. The Ecology group is higher in every one of the nine cells.
4.2 By continent and world region
Across continents (Figure 5) the three rates tell different stories. On citations per publication North 
America leads at every class (all 48) and Asia trails (all 27) despite carrying the most papers per author 
(all 142, against North America’s 108)—the clearest volume-versus-impact contrast in the data. A 
forensic oddity: South America and Africa post the highest solo publications per author (19 and 20, 
against ~8 elsewhere) and South America the highest solo citations per author (968)—not because the 
regions are prolific but because each lists only a handful of hyper-selected scientists whose individual 
careers are very large; the per-author means are inflated by tiny denominators and must be read as 
such. At the sub-region level (Figure 6) the same split holds: Northern America and Western Europe 
lead on per-paper impact, Eastern Asia on volume.

===== PAGE 14 =====
Page 14 of 32
Figure 5. By continent — the 3 × 3 scheme (rows: solo, senior, all; columns: publications per author, 
citations per author, citations per publication; continents ordered by output, five-year values). Asia’s 
high output but low per-paper impact, and South America’s and Africa’s tiny-base per-author spikes, 
stand out.

===== PAGE 15 =====
Page 15 of 32
Figure 6. By world sub-region (the eight most productive) — the same 3 × 3 scheme (five-year values).
4.3 By country
Country by country (Figures 7–8) the United States—by far the largest producer—is also among the 
most cited per paper at every class (solo 46, senior 56, all 52), so it leads on both output volume and 
per-paper impact. China is its mirror image: high output per author (all 196 papers) but the lowest 
citations per paper of the large cohorts (all 24), with Japan sharing the low per-paper impact. Figure 7 
isolates a single, revealing slice—each country’s solo-paper impact against its all-paper impact. Most 
large cohorts sit near the diagonal (solo and overall impact track each other), but the scatter above the 
line is itself informative: where a country’s solo papers out-cite its average paper, its most individual 
work is its strongest, whereas countries below the line lean on collaboration for their impact.

===== PAGE 16 =====
Page 16 of 32
Figure 7. A single revealing slice of the nine measures: solo-paper impact (citations per solo paper, 
vertical) against all-paper impact (citations per paper overall, horizontal), by country; bubble area = 
number of scientists; countries with at least five scientists per year. The dashed line is equality. Points 
above it (e.g. Portugal, Norway) are countries whose sole-authored work is cited more heavily than their 
average paper—their most individual output is their strongest; the United States sits high on both axes 
at large scale, China low on both. The vertical spread at a given horizontal position shows that overall 
per-paper impact does not fix solo impact—the two are related but far from identical.

===== PAGE 17 =====
Page 17 of 32
Figure 8. The full 3 × 3 scheme mapped: nine world choropleths, rows solo/senior/all × columns 
publications per author, citations per author, citations per publication (five-year values; countries with 
fewer than five listed scientists left grey). Boundaries: Natural Earth [6].
4.4 By U.S. state
Within the United States (counts in Table 6) the pooled federal-agency block and the leading forestry 
states carry most of the output; on the derived rates per-paper impact is generally high across the 
authorship classes, consistent with the national profile, though the small per-state samples make the 
per-author means noisy. Full per-state derived measures (and the Canadian and Australian 
subdivisions) are in the workbook.
4.5 By institution
Among the most-represented institutions (Figure 9) the clearest separation is again in per-paper impact 
rather than output: broad research universities reach the highest citations per publication at every 
authorship class, while several of the specialized forestry agencies and agricultural universities that top 
the output counts are markedly lower. Publications per author, by contrast, does not sort the institutions 
in any consistent way—the head-count leaders are not the most prolific per person, confirming at the 
institution level the volume–impact independence seen among countries.

===== PAGE 18 =====
Page 18 of 32
Figure 9. By institution (the ten most productive) — the 3 × 3 scheme (five-year values). Output per 
author varies without tracking per-paper impact; the institutions separate most clearly on citations per 
publication.
4.6 Authorship position and the attribution of credit
The nine measures still credit an author equally for a paper whether they wrote it alone, led it, or 
appeared as one of many co-authors. Resolving position one step further splits each author’s output 

===== PAGE 19 =====
Page 19 of 32
into three mutually exclusive classes, ordered from most to least individually attributable: single-
authored, first-authored (of a multi-author paper) and other co-authored (every paper on which the 
author is neither sole nor first author—last-authorship, which is not part of this study, is pooled here). So 
resolved, the cohort’s cited output is overwhelmingly collaborative (Table 9, Figure 10): sole-authored 
work is just 6% of its citations and 7% of its papers; first-authorship adds another 21%, so the two lead 
positions together account for about 27% of citations, while the remaining ~73% accrues to other co-
authored (collaborative) work. The two groups differ only slightly in this split (Figure 10), though the 
Ecology group’s citations are, as everywhere, denser per paper.
Table 9. The cohort’s five-year output resolved by authorship position (2020–2024). Categories are 
mutually exclusive and derived from the database’s nested position fields—single (nps/ncs) and single-
or-first (cpsf/ncsf)—so single + first + other = all; “other” is every non-lead co-authored paper (last- and 
middle-author positions pooled, as last-authorship is outside this study).
Authorship position
Share of papers (%)
Share of citations (%)
Citations per paper
Single-authored
7
6
37.2
First-authored (multi-author)
18
21
49.6
Other co-authored (not lead)
74
73
43.6
All positions (undiscriminated)
100
100
44.2
Figure 10. Where the cohort’s citations come from, by authorship position: single-, first- and other-co-
authored shares for the whole cohort and the two groups (five-year totals). Sole-authored work is a 
small slice; most citations accrue to non-lead co-authored work. Segment labels are rounded 
independently and may not sum to exactly 100.
Because the report’s headline counts do not discriminate authorship, they largely rank scientists by 
cited collaboration rather than by individual contribution—and the two orderings are almost unrelated. 
Ranking the 854 scientists listed in 2024 by total citations reproduces the “lead”-authorship ranking 
(single or first author) only moderately (Spearman ρ = +0.55) and bears essentially no relation to the 
ranking by single-authored citations (ρ = -0.03): of the 25 most-cited scientists overall, only 5 are among 
the 25 most-cited on their sole-authored work (Figure 11). The more one narrows toward sole individual 
contribution, the less an undiscriminated citation count reveals about it.

===== PAGE 20 =====
Page 20 of 32
Figure 11. How the undiscriminated (total-citation) ranking of the cohort’s scientists relates to rankings 
that credit only certain authorship positions: Spearman correlation and top-25 overlap fall sharply as the 
credited set narrows from lead (single + first) to single-authored citations.
The inference is direct. A citation ranking that ignores authorship position—like most of this report, and 
like the composite indicator that defines the source list itself—measures chiefly how much cited 
collaboration a scientist takes part in, not how much they individually author. That is a legitimate 
measure of participation in impactful research; but a reader who means individual contribution must 
read the single- and first-authored columns, where the ordering is very different. Section 5 makes the 
reversal concrete for two scientists at the same institution.
5. Discussion
The head-count geography (Data section) tells a familiar story—forestry’s recognized expertise is 
concentrated in the wealthy temperate North, with Canada and the Nordic countries (Finland and 
Sweden, tiny by population) ranking far above their size, and the tropical Global South, where most of 
the world’s forests are, largely absent. The gap between where forests grow and where forestry science 
is affiliated is a quantitative entry point into debates on research equity and capacity-building. But the 
more novel finding comes from resolving output by authorship position, which is why this report 
foregrounds the solo–senior–all structure.
Per-paper impact travels with discipline and geography, not with how much a population publishes. 
Within every region the ordering is the same—solo papers cited least, senior (single- or first-authored) 
papers most, all papers between—and it is stable across the five releases (Section 4.1). What moves is 
the level: North America and western Europe are the most cited per paper at every authorship tier 
(North America 48 citations per paper overall), while eastern Asia combines the highest output per 
scientist with the lowest per-paper impact (Asia 27). Publishing more does not in itself lower per-paper 
quality; the heavy variation in citations per paper is disciplinary and geographic, not a by-product of 
volume.
U.S. pre-eminence is scale. The United States leads every head-count and output ranking by fielding 
the largest number of elite scientists, and it is also among the most cited per paper of the large 
producers (solo 46, senior 56, all 52 citations per paper). China and Japan are its mirror image—high 
output per scientist, the lowest per-paper impact of the large cohorts. The lesson for benchmarking is 
that a country’s rank depends on which of the nine numbers is asked, so solo, senior and all output 
must be read together rather than collapsed into one league table.
The two groups differ in kind, not just degree. The ecology-rooted group is markedly more citation-
dense than the forestry-rooted group at every authorship level—about 2.0× the citations per paper 

===== PAGE 21 =====
Page 21 of 32
overall (73 vs 37), and a comparable multiple on solo and senior papers (Table 8, Figure 4)—so the 
roughly one-sixth of the cohort that is ecology-primary contributes a disproportionate share of its 
citations. This is the clearest single driver of quality differences in the whole dataset and reflects a real 
disciplinary fact: ecology journals and ecological questions attract heavier citation than applied 
silviculture. The taxonomy tree makes the boundary concrete, with the Ecology group sitting under the 
Biology root and the Forestry group mostly under Agriculture, Fisheries & Forestry but reaching into 
Biology, Chemistry, Earth & Environmental Sciences and Economics.
A forensic reading of the nine measures across the maps, charts and tables surfaces five regularities 
the headline counts hide. (1) Impact peaks at the lead, not the sole, position: for the cohort and in 
nearly every division, citations per paper are highest for senior work (46 for the cohort), lowest for pure-
solo work (37), with the all-paper figure between (44)—a scientist’s first-authored papers are their most-
cited output, and pooling in sole-authored work pulls the average down. (2) Volume and per-paper 
impact are near-orthogonal: Asia carries the most papers per author (142) yet is least cited per paper 
(27), North America inverts it (108 papers, 48 citations each), and neither the country nor the institution 
charts show a consistent volume-versus-quality trade-off. (3) The per-author means for the smallest 
regions are artefacts of selection, not productivity: South America and Africa post solo publications per 
author two to three times any Global-North region (19 and 20 against ~8) only because each lists a 
handful of hyper-cited veterans—large numerators over tiny denominators—so those cells must be read 
as small-sample noise, not regional strength. (4) The ecology premium is total, not partial: the Ecology 
group leads the Forestry group in all nine cells (Figure 4), the single largest structured effect in the data. 
(5) Solo impact is not predictable from overall impact at the individual scale: among the three Georgia 
reference scientists citations per solo paper run from 7 (Bettinger) to 70 (Peterson) with no tidy relation 
to their all-paper rates, so a scientist’s most individual work has to be measured directly rather than 
inferred. Together these say what the rank analysis says numerically: the axis you choose determines 
the answer.
Reference scientists. It is instructive to place individual careers against these population statistics on 
the same nine measures. Three are members of the cohort, all at the University of Georgia: the present 
author, Chris J. Cieszewski, and Pete Bettinger (both primary-Forestry authors, in roughly the top one 
percent of the 34,705 ranked Forestry-subfield authors worldwide—Bettinger 240th, Cieszewski 402nd 
in 2024), and Chris J. Peterson, a forest-disturbance ecologist (criterion 2, primary Ecology with 
secondary Forestry). Table 10 and Figure 12 give the three quantitatively against the University-of-
Georgia, Forestry-group, cohort and U.S. benchmarks. Two further scientists mark the extremes and 
are discussed in the text rather than tabulated because they sit outside the forestry cohort: the 
mathematician John Forbes Nash Jr. (not in the database at all) and the psychologist Daniel Kahneman 
(near the very top of it).
The three Georgia colleagues reproduce the cohort-level pattern within a single institution, and the nine-
number view sharpens it. Bettinger is a high-volume producer—207 career papers, 2.8 times 
Cieszewski’s 74—yet per paper the order reverses: Cieszewski’s papers average about 25 citations 
each against Bettinger’s 17. The gap widens as one narrows to individual work: on solo papers 
Cieszewski draws 40 citations each and Bettinger only 7—Bettinger’s sole-authored output is barely 
cited, his impact being overwhelmingly collaborative. Peterson, the ecologist, sits above both at every 
class (73 citations per paper overall, 70 per solo paper), a single-person instance of the ecology-over-
forestry premium (Section 4.1).
Table 10. The three University-of-Georgia reference scientists on the nine derived measures (career 
values, 2024 release), against population benchmarks (five-year values). For an individual, 
publications-per-author is the career paper count and citations-per-author the career citation count; for a 

===== PAGE 22 =====
Page 22 of 32
population both are per-scientist means. All cells are comparable across rows. Each authorship class—
Solo (single-authored), Senior (single- or first-authored) and All—carries three derived measures: 
publications per author (p/au), citations per author (c/au) and citations per publication (c/pub), each 
formed from five-year sums over the listed author-year records. No raw citation total is shown.
Reference
Solo 
p/au
Solo 
c/au
Solo 
c/pub
Sr. 
p/au
Sr. 
c/au
Sr. 
c/pub
All p/au All c/au
All 
c/pub
Cieszewski, Chris J. (this author)
8
316
40
27
802
30
74
1,878
25
Bettinger, Pete
8
56
7
71
1,338
19
207
3,495
17
Peterson, Chris J. 
(ecology→forestry)
9
633
70
27
1,715
64
71
5,194
73
University of Georgia
8
300
40
39
1,161
29
106
3,034
29
Forestry group
7
219
29
27
1,025
37
107
3,925
37
Whole cohort
8
301
37
29
1,338
46
113
5,005
44
United States
7
343
46
29
1,633
56
103
5,424
52
Figure 12. The three cohort reference scientists in the (publications per author, citations per publication) 
plane, with each scientist’s solo, senior and all placements marked and joined by a line. Peterson 
(ecologist) sits highest on per-paper impact at every class; Cieszewski’s impact falls as co-authored 
work is added (solo is his strongest); Bettinger is high-volume with low per-paper impact, and his solo 
papers (≈7 citations each) are barely cited. Two external Nobel laureates lie off the frame and are 
discussed in the text: John F. Nash (23 papers, not in the dataset) and Daniel Kahneman (196 papers, 
~571 citations each, near the top of the dataset).
John Nash is the limiting case of the selective axis, and worth stating precisely because the dataset 
omits him. His 23 papers include several of the most influential works of twentieth-century mathematics 
and economics: the 1951 article “Non-cooperative games” [7], drawn from a 28-page doctoral thesis, 

===== PAGE 23 =====
Page 23 of 32
defined the Nash equilibrium and is among the most-cited papers in the mathematical social sciences, 
cited tens of thousands of times. Its economic content directly qualified Adam Smith’s “invisible hand”: 
Nash showed that agents each pursuing rational self-interest can settle on an equilibrium that is 
collectively sub-optimal—the prisoner’s dilemma is the textbook case—so individual self-interest need 
not deliver the social optimum Smith supposed. That one short paper, and a handful of others, carry an 
impact per unit of output that no country, institution or individual anywhere in this report approaches.
Nash’s absence from the list is itself the lesson. A career of 23 papers—however foundational—yields a 
modest h-index (an h-index cannot exceed the paper count) and a composite score dominated by 
output, so a purely metric-based ranking places a field-founding laureate below thousands of prolific 
mid-tier scientists, and a coverage- and volume-sensitive database omits him entirely. The same 
structural effect is visible in miniature within the cohort: Bettinger’s h-index (30) exceeds Cieszewski’s 
(21) chiefly because he has published far more, not because his papers are individually more cited—
per paper the reverse holds. h-type and composite indicators systematically cap the low-volume/high-
impact profile, the very h-index limitation that motivated the composite indicator in the first place [1,2]. 
The practical lesson repeats the section’s theme at the scale of individuals: which scientist ‘ranks 
highest’ depends entirely on the axis chosen, so quantity and quality must be reported separately.
Kahneman marks the opposite pole from Nash, and from inside the database. His 196 indexed papers 
are almost exactly Bettinger’s count (207), but they carry 112,018 citations—about 572 per paper, some 
34 times Bettinger’s 17 and 13 times the cohort’s per-paper mean of 44—for an h-index of 88 and a 
place 4th of 36,527 in his Experimental-Psychology subfield and 218th of the roughly 200,000 scientists 
in the whole list. He is the profile forestry’s applied literature never reaches—high output and extreme 
per-paper impact together—and the mirror image of Nash: the database seats Kahneman at its summit 
yet omits Nash entirely, because it registers heavily-cited, mainstream, indexed work and loses sparse, 
foundational or non-indexed work, the same bias Table 11 quantifies. Both are here as personal 
reference points—Kahneman taught the author at UBC in 1985, as Bettinger is his present colleague—
and both show that a single citation number, stripped of authorship and of the venues a database 
happens to cover, ranks scientists on an axis that need not be the one a reader has in mind.
Authorship position makes the same point personal, and reverses it. Of Bettinger’s 3,495 citations only 
56 (from 8 sole-authored papers) are his in the strictest sense; another 1,282 come from first-authored 
papers and the remaining 2,157 from other co-authored (non-lead) work. Cieszewski’s smaller total 
(1,878) is proportionally more his own—316 single-authored and 486 first-authored citations. The effect 
on rank is decisive (Section 4.6): by total citations Bettinger (446th of the 854 scientists listed in 2024) 
sits well above Cieszewski (706th); by single-authored citations the order reverses, Cieszewski rising to 
234th and Bettinger falling to 638th. Which of the two ‘contributed more’ is not settled by the 
undiscriminated count; it depends entirely on whether one credits sole authorship or total participation.
Both Georgia foresters (Bettinger and Cieszewski) sit below the cohort’s per-paper mean of 44 citations 
per article, and below the U.S. mean of 52—as expected for applied, quantitative forestry, whose 
growth-and-yield, geomatics and forest-planning literature is cited less heavily per paper than the 
ecology-leaning work that lifts the cohort average (Section 4.1); their ecologist colleague Peterson, by 
contrast, sits above it. Their shared department, the University of Georgia, averages 29 citations per 
article across its roughly 3 listed forestry scientists per year—a figure the two of them largely set (the 
University of Georgia’s full roster of listed members meeting the two criteria over 2020–2024 is given in 
Table 12, below). They also fall on different branches of the taxonomy (Figure 3): Bettinger’s secondary 
specialty is recorded as Ecology, although his published work centres on forest management planning, 
spatial optimisation and GIS—a reminder that the Science-Metrix secondary label does not always 
capture the quantitative character of an author’s work; Cieszewski’s, by contrast, is Geomatics 

===== PAGE 24 =====
Page 24 of 32
Engineering, one of the quantitative sub-sub-fields, matching a programme in forest biometrics and 
growth-and-yield modelling. It is a distinction on which a close colleague’s first-hand knowledge of the 
work adds what the citation record alone cannot.
External validation of coverage: a case study. Because every number in this report inherits the 
coverage of the underlying database (Scopus), it is worth confronting the database record of the two 
University-of-Georgia scientists compared above—the present author and Bettinger, whose full 
publication lists are independently visible—with their public Google Scholar profiles (retrieved July 
2026; Table 11). Google Scholar’s raw lists are inflated by duplicate entries, translations of the same 
paper into other languages, and non-article items (software files, conference websites, section 
headings); de-duplicating them—done reproducibly by title—reduces Cieszewski’s 211 listed entries to 
about 170 unique works (one handbook alone is listed six times) and Bettinger’s 361 listed entries to 
about 300. Against those de-duplicated corpora the database understates both careers, but by strikingly 
different amounts. For Cieszewski the two sources roughly agree in proportion: the database holds 
1,878 of his 4,131 Google Scholar citations (45%) and about 44% of his works. For Bettinger the 
database holds 3,495 of 10,619 citations—only 33%—even though it indexes roughly two-thirds of his 
works, because his citations come heavily from books, proceedings and other documents Scopus does 
not index as citing sources. The two sources are biased in the same direction (Scopus is journal-centric 
and under-covers proceedings, books, reports and regional-language outlets) but to author-specific 
degrees. On the recent window both agree well: the database career counts grew by 10 (Cieszewski) 
and 49 (Bettinger) papers across the 2020–2024 releases, matching their Google-Scholar output at the 
resolution the analysis uses.
Table 11. External validation of database coverage: the two University-of-Georgia reference scientists’ 
records in the career database (DB, 2024 release) versus their public Google Scholar profiles (July 
2026). Google Scholar lists 211 works for Cieszewski and 361 for Bettinger; de-duplicating by title 
(removing repeated listings, translations and non-article items—one Cieszewski handbook is listed six 
times) reduces these to about 170 and 300 unique works respectively. Differences reflect source 
coverage, not errors in either source.
Measure
Cieszewski: DB
Cieszewski: Scholar
Bettinger: DB
Bettinger: Scholar
Listed works (Scholar raw)
—
211
—
361
Unique works (career)
74
≈170
207
≈300
Career citations
1,878
4,131
3,495
10,619
h-index
21
29
30
49
i10-index
—
60
—
174
Scopus share of citations
45%
(=100%)
33%
(=100%)
The implication is a disclaimer with a bounded—but not negligible—scope. Absolute magnitudes 
(papers, citations, and the ratios built from them) are Scopus-basis values and understate an author’s 
true output and impact, the more so where the work or its citing literature is proceedings-, book- or 
region-heavy; on another source they would move upward, and by author-specific amounts. The 
comparative analyses are computed within one consistently-indexed source, so they are distorted only 
insofar as coverage bias differs between the entities compared. That differential is not always small—
Bettinger’s citation coverage (≈33%) is well below Cieszewski’s (≈45%), a real difference between two 
colleagues in the same department—but in every comparison this report actually draws it preserves the 
direction of the result: Bettinger leads Cieszewski on total citations in both sources, and by a wider 
margin in Google Scholar (2.6× versus 1.9×), so a Scholar-basis analysis would sharpen, not reverse, 
the total-citation ranking. The single-authored reversal of Section 4.6 is a within-database structural fact 
about authorship credit that an all-source citation count cannot itself confirm or deny. Discrepancies 

===== PAGE 25 =====
Page 25 of 32
relative to a Google-Scholar-basis analysis are therefore expected—material for the absolute levels, but 
direction-preserving for the comparisons.
The University of Georgia roster makes the coverage and data-quality issues concrete at the smallest 
scale. Over the five releases the database attaches five distinct names to UGA under the two criteria, 
but two are spurious and are set aside here: the criterion-1 records for I. D. Cave and J.-P. Schütz carry 
a “University of Georgia” affiliation that is evidently a Scopus assignment error—both are eminent non-
UGA wood scientists (Cave a New Zealand wood-physics researcher, Schütz an ETH-Zürich 
silviculturist), and both are among the blank-country records the Methods flag. That leaves the three 
genuine UGA cohort members of Table 12: under criterion 1 (primary subfield Forestry) the present 
author and Pete Bettinger, both of the Warnell School of Forestry and Natural Resources; under 
criterion 2 (primary Ecology, secondary Forestry) the forest-disturbance ecologist Chris J. Peterson of 
the Department of Plant Biology. The database records only the institution, so the departments here are 
added from public university sources; the two spurious affiliations are a reminder that even the affiliation 
field—on which all of this report’s geography rests—carries occasional error.
Table 12. The three genuine University-of-Georgia members of the cohort, by selection criterion (1 = 
primary subfield Forestry; 2 = primary Ecology with secondary Forestry), 2020–2024, with name 
variants across releases consolidated. The source carries only the institution, not the department; 
departments are added from public UGA sources. Two further criterion-1 records the source attaches to 
UGA—I. D. Cave (secondary Microscopy) and J.-P. Schütz (secondary Evolutionary Biology)—are 
internationally-known non-UGA wood scientists whose “University of Georgia” affiliation appears to be 
an indexing artifact, and are excluded above.
Name
Crit. Primary → secondary subfield
Department (added from public sources)
Cieszewski, Chris J. (this 
author)
1
Forestry → Geomatics 
Engineering
Warnell School of Forestry and Natural 
Resources
Bettinger, Pete
1
Forestry → Ecology
Warnell School of Forestry and Natural 
Resources
Peterson, Chris J.
2
Ecology → Forestry
Department of Plant Biology
A paradox at the source. The University of Georgia’s presence on this list is the smallest the analysis 
can register while still being non-zero: two primary-Forestry scientists (criterion 1) and a single ecologist 
working in forestry (criterion 2)—and, once the two spurious affiliations of Table 12 are set aside, that is 
the whole of the University of Georgia’s forestry cohort as the ‘Forestry’ classification counts it (a fuller 
Warnell tally, across other subfield labels, follows below). Against its context the figure is jarring. The 
University of Georgia is the state’s flagship land-grant institution and the nation’s first state-chartered 
public university (1785, the birthplace of public higher education in America); its Warnell School of 
Forestry and Natural Resources, endowed in 1906, is the oldest forestry school in the American South. 
And Georgia is, by a wide margin, the number-one forestry state in the United States: it holds more 
commercial timberland than any other state—about 22 million acres—leads the nation in timber 
harvested and in forest-product exports, and supports a forest industry of roughly forty billion dollars a 
year. That the flagship university of the country’s foremost forestry state should field only two or three of 
the world’s most-cited forestry scientists is, on its face, hard to reconcile.
Part of the shortfall is measurement, and the report’s own findings say so. The list ranks only the most-
cited two percent or so of scientists, on a Scopus basis that—as Table 11 documents—systematically 
under-covers the conference proceedings, books, technical reports and extension literature in which 
applied, quantitative and management-oriented forestry, the working core of a school like Warnell, 
largely appears. The same coverage shortfall that roughly halved the present author’s indexed citations 

===== PAGE 26 =====
Page 26 of 32
depresses the visibility of exactly this kind of forestry. The count of two therefore understates the 
science the university actually does; the real forestry research effort is larger than the metric can see.
Yet measurement does not dissolve the paradox, and the present author contends that the residue is 
symptomatic of a deeper malaise in the university system. Across American higher education the past 
half-century has seen administrative ranks expand far faster than faculty or research capacity: from 
1975 to 2005 administrators grew by roughly 85 percent and administrative staff by some 240 percent, 
and between 1993 and 2009 administrative positions grew about 60 percent—on the order of ten times 
the growth rate of tenured faculty—with professional-staff growth outpacing faculty hiring across every 
type of institution [8]. The argument advanced here is that this proliferation of administration, with the 
incentives and the diversion of resources it imposes, has come at the expense of the sustained, 
discipline-deep research programmes that produce highly-cited science—that a flagship’s vanishing 
presence in global forestry science, in the very state where forestry matters most, is a visible symptom 
of a system that has grown its bureaucracy while starving the scholarly enterprise it exists to serve. This 
is an interpretation, not a demonstration: the data assembled here cannot isolate cause, and other 
factors—retirements and hiring cycles, the field’s proceedings-heavy applied character, the metric’s 
own biases—contribute. But the juxtaposition is itself a datum. The country’s premier forestry state; its 
oldest southern forestry school; its flagship, land-grant, first-chartered university—and, on the world’s 
most-used ranking of scientific impact, almost no one. Any honest account of the health of the 
academic research enterprise has to explain that.
Full disclosure, however, requires a correction to the headline count—and it cuts the other way. The 
Science-Metrix taxonomy files each scientist under a single primary subfield, and for a natural-
resources faculty that scatters people across labels. Checking the affiliations of all 280 University-of-
Georgia scientists on the 2024 list against the Warnell School’s own roster shows that several Warnell 
faculty appear not as “Forestry” but under Plant Biology, Fisheries, Agronomy, veterinary parasitology, 
and even Recreation and Tourism (Table 13). Counted properly, the school’s presence on the list is 
about eight, not two: R. O. Teskey (tree physiology, filed under Plant Biology), D. Markewitz (forest 
soils, filed under Agronomy), G. D. Grossman and G. S. Helfman (fisheries, both emeritus), M. J. 
Yabsley (wildlife disease, filed under parasitology) and K. M. Woosnam (parks and tourism), alongside 
the two Forestry-classified foresters. The dispersion is the report’s own taxonomy point (Figure 3) in 
miniature: a single-subfield label cannot hold a discipline as broad as forestry and natural resources, so 
any tally keyed to “Forestry” alone undercounts the field, at UGA and everywhere. Eight highly-cited 
scientists is still a modest showing for the flagship of the nation’s first forestry state—the paradox is 
softened, not dissolved—but the honest figure is eight, and the two-scientist headline is in part an 
artifact of classification.
Table 13. Warnell School of Forestry and Natural Resources faculty on the 2024 list who are classified 
under a non-Forestry primary subfield, by publication class (career values, 2024 release; Solo = single-
authored, Senior = single- or first-authored, All = every paper; c/pub = citations per publication, no 
citation totals shown). Identified by matching all 280 University-of-Georgia listees against the Warnell 
roster and public sources; the database carries no department, so this is a best-effort identification and 
may omit joint USDA-Forest-Service or Savannah-River-Ecology-Laboratory appointments affiliated 
elsewhere in the source. Grossman and Helfman are emeritus. With the two Forestry-classified 
foresters (Bettinger, Cieszewski) the school’s presence on the list is about eight.
Name
Listed subfield (primary → 
secondary)
Solo 
pubs
Solo 
c/pub
Senior 
pubs
Sr. 
c/pub
All 
pubs
All 
c/pub
Yabsley, Michael J.
Mycology & Parasitology → 
Veterinary Sci.
5
26
46
31
335
25

===== PAGE 27 =====
Page 27 of 32
Name
Listed subfield (primary → 
secondary)
Solo 
pubs
Solo 
c/pub
Senior 
pubs
Sr. 
c/pub
All 
pubs
All 
c/pub
Teskey, Robert O.
Plant Biology & Botany → Forestry
2
72
17
110
120
61
Woosnam, Kyle M.
Sport, Leisure & Tourism → Reg. 
Planning
5
99
41
54
179
39
Grossman, Gary D. 
(emeritus)
Fisheries → Ecology
11
38
43
49
112
43
Markewitz, Daniel
Agronomy & Agriculture → Forestry
2
42
13
66
110
43
Helfman, Gene S. 
(emeritus)
Fisheries → Behavioural Science
11
70
17
73
30
122
Nonetheless, the core figure stays glaring. Whatever the exact tally—two scientists in the top “Forestry” 
recognition category, about eight across the whole school—it is conspicuously thin for the flagship 
university of a state where forestry is no marginal pursuit but one of the largest agricultural sectors of 
all: by the industry’s economic footprint, on the order of forty billion dollars a year and roughly 140,000 
jobs, it stands second only to poultry. A discipline of that economic weight, housed at the institution 
charged above all others with serving it, ought to field far more than two of its scientists among the 
world’s most-cited. That it does not is the datum with which this section began, and enlarging the count 
from two to eight does not enlarge it enough to dissolve the point.
Robustness and integrity of the cohort. The database carries several diagnostics—beyond output and 
impact—that test how trustworthy and how concentrated the cohort’s citation record is; the section 
closes by reading them (Table 14, Figure 13). Three point the same way: the cohort is clean. Self-
citation is moderate (median 15% of an author’s citations, none above 49%, only 83 of 854 authors 
above 25%); citation concentration shows no anomaly (the ratio of citations to distinct citing papers is 
1.3 at the median and never exceeds 2.4, well below the range that flags citation-farming); and 
retraction exposure is negligible—just 7 retracted papers across all 854 authors (5 authors), with 4 
authors citing any retracted work. Visibility is high: the median author has had 94% of their papers cited 
at least once, so impact is broad, not confined to a few blockbusters.
Table 14. Robustness, integrity and distribution diagnostics for the cohort (2024 release, 854 authors). 
Self-citation, citation concentration and retraction exposure are screening variables, not misconduct 
labels; the cohort is unremarkable on all three. The hm-index fractionally counts multi-authored papers 
[10], so hm ≈ ½·h quantifies the cohort’s heavy collaboration. Retraction columns are populated only in 
the 2023–2024 releases.
Diagnostic (2024 release, n = 854)
Value
Self-citation rate — median / 90th pct / max
15% / 25% / 49%
Authors with self-citation above 25%
83 (10%)
Citations per distinct citing paper (cprat) — median / max
1.33 / 2.37
Retracted papers in cohort / authors affected
7 / 5
Papers ever cited — median share per author
94%
h-index / hm-index (coauthorship-adjusted) — median
31 / 16  (hm ≈ 0.50·h)
Citation inequality — Gini / top-10% / top-1% share
0.45 / 34% / 7%
Career age (2024 − first publication year) — median
35 yr

===== PAGE 28 =====
Page 28 of 32
Figure 13. Left: Lorenz curve of career citations across the 854 listed scientists—even within this 
citation elite the top tenth holds about a third of all citations (Gini 0.45), so means are pulled by a heavy 
right tail. Right: the coauthorship-adjusted hm-index against the h-index; the cloud sits along the hm = 
½·h line, so accounting for multi-authorship roughly halves the index. The three University-of-Georgia 
reference scientists are marked.
Two readings follow. Methodologically, the concentration and the h-to-hm gap reinforce the report’s 
recurring caution—population means are pulled by a heavy right tail and by collaboration, so the 
medians, per-paper rates and authorship-resolved figures used throughout are the honest summaries, 
not raw totals or a single index, in line with responsible-metrics guidance that indicators be read with 
context and never collapsed into one score [11]. Substantively, the integrity screen matters for a cohort 
of elite authors precisely because citation-based selection can reward gaming: that this forestry cohort 
shows low self-citation, no citation-concentration anomaly and almost no retraction exposure means its 
standing rests on ordinary, broadly-cited work. The reference scientists are likewise unremarkable (self-
citation 9–23%), and the h-to-hm halving holds for each: Bettinger’s h of 30 falls to an hm of 18 once his 
heavy co-authorship is discounted—the same collaboration signal Section 4.6 read from his authorship 
positions.
Caveats on the comparisons. All of these authors already sit at the top of the citation distribution, so 
every comparison is among an elite; citations accumulate with career length, so nations or institutions 
with older cohorts look more impactful; per-paper impact is shaped by field mix (the ecology effect 
above), publication venue and language; and the small-cohort leaders (Netherlands, Brazil, Portugal) 
rest on a handful of authors and should not be over-read. Affiliation geography credits the current host 
institution, not origin or training.
Uses. Reporting solo, senior and all output with its per-paper impact lets funders and forestry ministries 
benchmark national standing on volume and on per-paper impact independently, and at the level of 
individually-authored work rather than total collaboration; forestry schools benchmark peers and target 
recruitment; editors widen reviewer and speaker pools using the geography; and science-of-science 
researchers can study the authorship-credit structure and the forestry–ecology boundary directly.
Limitations. Affiliation country is current institution, not nationality or origin. Scopus coverage biases 
toward English-language journals and under-covers proceedings, books and reports, likely understating 
tropical, non-English and applied-technical forestry—the case study above (Table 10) quantifies the 

===== PAGE 29 =====
Page 29 of 32
effect for two authors at roughly two- to three-fold on citations, and shows the coverage shortfall is 
author-specific yet direction-preserving for the report’s comparisons. Publication and citation sums 
inherit the skew of citation distributions—a few superstar authors dominate—so ratios for small divisions 
are volatile (hence the minimum-size filters). The counting unit is the author-year, so annual changes 
mix genuine entry with re-computation. The two-group selection is deliberately conservative on the 
ecology side; a looser rule would enlarge the Ecology group. Subnational placement leaves some U.S. 
institutions “Unknown” and pools federal agencies.
6. Conclusion
By head-count the forestry research cohort—3,271 primary-forestry authors plus 615 ecology-rooted 
authors who also work in forestry, 3,886 in all—is small, growing, and strongly concentrated in the 
Global North: North America and Europe dominate, the United States and Canada lead among 
countries, Finland and Sweden are conspicuously over-represented, and USDA Forest Service is the 
most represented institution (Data section). The analytical contribution, however, is in resolving output 
by authorship position. Reading solo, senior and all publications with their citations per paper together 
shows that per-paper impact travels with discipline and geography, not with how much a population 
publishes; that the United States leads by scale and is also among the most cited per paper of the large 
producers, while China and Japan show the opposite high-volume, low-per-paper-impact profile; that 
the ecology-rooted group is about twice as cited per paper as the forestry-rooted group at every 
authorship level; and that an undiscriminated citation ranking measures cited collaboration more than 
individual authorship, so narrowing to solo- or first-authored credit reorders scientists substantially. 
Which country, institution or individual looks strongest depends on which of the nine measures one 
asks for—so they must be reported together.
7. Project Summary
Five career-long releases of the Stanford / Ioannidis database (2020–2024) were filtered to a “Forestry 
plus Ecology” cohort of 3,886 records—Forestry (3,271: primary subfield Forestry) plus Ecology (615: 
primary Ecology with secondary Forestry)—both groups requiring forestry. The Data section (Tables 1–
7) gives only raw counts—the number of scientists and of solo (single-authored), senior (single- or first-
authored) and all publications for every division (group, continent, world sub-region, country, U.S. state 
and institution)—with the two groups distinguished and a multi-root field→subfield→subfield taxonomy. 
The Results (Section 4) derive, for each authorship class, three rates—publications per author, citations 
per author, citations per publication—a 3 × 3 scheme of nine measures shown as 3 × 3 charts, and read 
the patterns: per-paper impact tracks discipline and geography rather than publication volume, and 
holds its solo–senior–all ordering across regions. A further analysis (Section 4.6) resolves each author’s 
output into single, first and other (non-lead) co-authored positions and shows that an undiscriminated 
citation ranking measures cited collaboration rather than individual contribution (ranking by total 
citations is essentially uncorrelated with ranking by single-authored citations). The Discussion adds 
individual scientists as concrete reference points—the author and two UGA colleagues (all in the 
cohort); the mathematician John Nash as an external low-volume/high-impact case the database omits 
entirely; and Daniel Kahneman as an in-database high-impact extreme—showing the same authorship-
credit reversal at the scale of a single career; it also lists the University of Georgia’s cohort members by 
criterion (Table 12) and bounds the database’s Scopus coverage against Google Scholar (Table 11); 
and a closing robustness screen (Table 14, Figure 13) confirms low self-citation, no citation-
concentration anomaly and negligible retraction exposure, while quantifying the cohort’s heavy 

===== PAGE 30 =====
Page 30 of 32
collaboration (hm ≈ ½·h) and citation concentration (Gini 0.45). Appendix A profiles the Ecology group 
on its own terms; Appendix B notes that single-authored (solo) output is carried in the “Solo” columns of 
every table throughout. Deliverables: this report; a comprehensive workbook with full per-year, per-
metric rankings and per-group sheets; and the source career database (Excel and CSV).
8. References and Data Sources
[1] Hirsch, J. E. (2005). An index to quantify an individual’s scientific research output. PNAS, 102(46), 16569–
16572. https://doi.org/10.1073/pnas.0507655102
[2] Ioannidis, J. P. A., Klavans, R., & Boyack, K. W. (2016). Multiple citation indicators and their composite across 
scientific disciplines. PLoS Biology, 14(7), e1002501. https://doi.org/10.1371/journal.pbio.1002501
[3] Ioannidis, J. P. A., Baas, J., Klavans, R., & Boyack, K. W. (2019). A standardized citation metrics author 
database annotated for scientific field. PLoS Biology, 17(8), e3000384. 
https://doi.org/10.1371/journal.pbio.3000384
[4] Ioannidis, J. P. A., Boyack, K. W., & Baas, J. (2020). Updated science-wide author databases of standardized 
citation indicators. PLoS Biology, 18(10), e3000918. https://doi.org/10.1371/journal.pbio.3000918
[5] Archambault, É., Beauchesne, O. H., & Caruso, J. (2011). Towards a multilingual, comprehensive and open 
scientific journal ontology. Proc. 13th ISSI Conference, 66–77. (Science-Metrix classification.)
[6] Natural Earth (2024). Free vector and raster map data, 1:110m Admin 0 – Countries. 
https://www.naturalearthdata.com (public domain).
[7] Nash, J. F. (1951). Non-cooperative games. Annals of Mathematics, 54(2), 286–295. 
https://doi.org/10.2307/1969529 (Nash equilibrium; Nobel Memorial Prize in Economic Sciences 1994).
[8] Ginsberg, B. (2011). The Fall of the Faculty: The Rise of the All-Administrative University and Why It Matters. 
Oxford University Press. See also Desrochers, D. M., & Kirshstein, R. (2014), Labor Intensive or Labor 
Expensive? Changing Staffing and Compensation Patterns in Higher Education, Delta Cost Project / 
American Institutes for Research.
[9] Ioannidis, J. P. A. (2025). August 2025 data-update for “Updated science-wide author databases of 
standardized citation indicators” (Version 8) [Data set]. Elsevier Data Repository. 
https://doi.org/10.17632/btchxktzyw.8
[10] Schreiber, M. (2008). To share the fame in a fair way, hm modifies h for multi-authored manuscripts. New 
Journal of Physics, 10, 040201. https://doi.org/10.1088/1367-2630/10/4/040201 (coauthorship-adjusted h-
index).
[11] Hicks, D., Wouters, P., Waltman, L., de Rijcke, S., & Rafols, I. (2015). The Leiden Manifesto for research 
metrics. Nature, 520, 429–431. https://doi.org/10.1038/520429a. See also DORA (2013), San Francisco 
Declaration on Research Assessment, https://sfdora.org.
Data sources.
Primary: career-long author tables (Table_1_Authors_career), dataset doi:10.17632/btchxktzyw, versions 
3/5/6/7/8, via the Elsevier Digital Commons / Mendeley Data public API. Field/subfield taxonomy: Table_2 
threshold tables in the same dataset. Base maps: Natural Earth (world) and a public US-state boundary file. 
External coverage validation (Table 11): the public Google Scholar profiles of the two reference scientists, 
https://scholar.google.com/citations?user=npyjSdoAAAAJ (Cieszewski) and https://scholar.google.com/citations?
user=rvCOS60AAAAJ (Bettinger), retrieved July 2026 and de-duplicated by title. Analysis and figures: Python 
(pandas, openpyxl, matplotlib).

===== PAGE 31 =====
Page 31 of 32
Appendix A. The Ecology Group (Criterion 2) in Detail
This appendix profiles only the Ecology group—selection criterion 2: authors whose primary Science-
Metrix subfield is Ecology and whose secondary subfield is Forestry (615 author-year records, 16% of 
the 3,886-record cohort). These are ecologists who also work in forestry; like the Forestry group, every 
one of them has Forestry as a subfield, so this is still a strictly forestry-defined population—just the 
ecology-primary half of it. Following the main report, Section A.1 gives its raw counts and Section A.2 
its derived nine-measure premium; full per-country and per-institution breakdowns are in the companion 
workbook.
A.1 Counts by release year and continent
The Ecology group is even more Global-North-concentrated than the cohort as a whole—North America 
and Europe supply almost all of it. Tables A1 and A2 give its raw scientist and publication counts by 
release year and by continent; the derived rates follow in A.2.
Table A1. Ecology group (criterion 2) scientist and publication counts by release year (scientists = listed 
records that year). Scientists = five-year average of listed author-year records; Solo = single-authored, 
Senior = single- or first-authored, All = every publication — each count summed over the five releases. 
These are raw selected counts; every per-author and per-publication rate is a derivation, reported in 
Sections 3–4.
Release year
Scientists /yr
Solo pubs Senior pubs
All pubs
2020
117
1,349
4,272
15,385
2021
127
1,458
4,627
17,797
2022
123
1,422
4,665
18,130
2023
127
1,469
4,886
19,582
2024
121
1,402
4,657
19,602
Table A2. Ecology group — scientist and publication counts by continent (five-year values, ordered by 
output).
Continent
Scientists /yr
Solo pubs Senior pubs
All pubs
North America
69
3,369
11,710
44,726
Europe
32
1,602
5,331
24,753
Oceania
10
721
3,156
10,964
South America
5
750
1,295
4,316
Africa
5
542
1,307
4,092
Asia
2
116
308
1,645
A.2 The ecology impact premium (derived)
On the derived rates the group’s defining feature is per-paper impact. Its solo papers draw about 64 
citations each against the Forestry group’s 29, its senior papers 80 against 37, and across all papers 73 
against 37—roughly a two-fold premium at every authorship class—while it also publishes more per 
author (all 147 vs 107 papers). The full nine-measure comparison against the Forestry group is Table 8 
and Figure 4 of the main report. Country- and institution-level rates are not tabulated here: with only 
about 123 scientists per year the group has too few members per country to give stable per-author 
means, so the reliable comparison for this group is at the group and continent level.

===== PAGE 32 =====
Page 32 of 32
Appendix B. Note on single-author (solo) publications
An earlier version of this report carried two further appendices replicating the whole analysis for single-
authored publications only. In the presentation adopted throughout, single-authored output is no longer 
a separate analysis: solo is the first of the three authorship classes carried everywhere—its publication 
counts in the Data section (the “Solo pubs” column of Tables 1–7 and A1–A2) and its derived rates 
(solo publications per author, citations per author and citations per publication) as the top row of every 3 
× 3 Results chart and of the reference-scientist table (Table 10). The solo figures can be read directly 
against the senior and all-publication figures alongside them, which is more informative than the 
standalone tabulation those appendices provided; the separate appendices are therefore no longer 
needed.
Two features of solo output stand out across those tables. First, single-authored work is a small slice of 
the cohort—about 7% of its papers are solo—so absolute solo counts run far below the all-publication 
counts. Second, across the cohort solo papers are cited somewhat less per paper (37) than senior (46) 
or all papers (44) (Section 4.1), consistent with sole-authored work often being reviews or conceptual 
pieces rather than large empirical collaborations. But the between-group and geographic orderings that 
hold for all papers hold for solo papers too: the Ecology group out-cites the Forestry group, and North 
America out-cites Asia, on solo work as on the rest. The complete per-scientist single-authored 
rankings by division are in the companion workbook.
