James Mickley February 02, 2020
This is my exploratory analysis of the OSU herbarium’s vascular plants.
Primarily, I’m concerned with understanding sampling in space, time, and across species. But, it’s also an opportunity to explore the collection, looking at major collectors and plant families, and assessing which species are best represented.
- There’s a lot of spatial variation in collecting intensity. Lane, Benton, Harney, and Wallowa Counties stand out as the best-sampled areas.
- The 1910-1950 era was when the bulk of the specimens were collected. Collecting in the 2010s was particularly low.
- Collection through time was similar in most counties, though a few have not had much activity since the 1950s.
- Of ~4500 total species in the dataset, half are represented by 12 or fewer specimens. Nearly 1000 species are only represented by 1-2 specimens.
- 750 species have not been collected in 50 years.
- Even for the best-represented species in the herbarium, sampling has been low since 1950.
- Common trees do not have much sampling at any point in time.
- Common invasive species have been mostly ignored by collectors, with minimal sampling.
- By comparison, iNaturalist added 60,000 records in Oregon just in 2019!
The OSC herbarium has just over 4500 species of vascular plant in Oregon with good data (IDed to species, good location and date).
Here’s a summary of the number of families, species, and specimens represented, along with stats for the most common families.
| Families | Species | Specimens |
|---|---|---|
| 162 | 4539 | 111678 |
| Family | Species | Specimens |
|---|---|---|
| Asteraceae | 561 | 14531 |
| Poaceae | 390 | 8879 |
| Fabaceae | 241 | 7280 |
| Rosaceae | 186 | 4960 |
| Cyperaceae | 221 | 4611 |
| Brassicaceae | 230 | 3926 |
| Plantaginaceae | 130 | 3844 |
| Ranunculaceae | 96 | 3499 |
| Apiaceae | 132 | 3421 |
| Polygonaceae | 124 | 3294 |
| Onagraceae | 86 | 2648 |
| Ericaceae | 69 | 2520 |
| Orobanchaceae | 86 | 2460 |
| Boraginaceae | 110 | 2423 |
| Polemoniaceae | 88 | 2376 |
| Caryophyllaceae | 118 | 2365 |
| Saxifragaceae | 75 | 2118 |
| Liliaceae | 63 | 1674 |
| Hydrophyllaceae | 49 | 1668 |
| Juncaceae | 71 | 1609 |
A map of all the specimen records to look at sampling intensity across space.
It looks like there’s less sampling in the mountains, unsurprisingly.
A list of important collectors for Oregon specimens in OSC, with active dates.
Morton Peck of WILLU tops the list (Manual of the Higher Plants of OR). Louis Henderson of UofO is second, and Richard Halse is third.
Some great information on the background of these collectors is here.
| Collector | Specimens | NumSpecies | FirstAccession | LastAccession |
|---|---|---|---|---|
| M.E. Peck | 19409 | 2867 | 1900 | 1958 |
| L.F. Henderson | 6870 | 1941 | 1878 | 1987 |
| Richard R. Halse | 4496 | 1770 | 1973 | 2016 |
| Georgia Mason | 3138 | 1090 | 1956 | 1980 |
| Lilla Leach | 2919 | 1261 | 1908 | 1955 |
| W.E. Lawrence | 2492 | 913 | 1911 | 1947 |
| L.E. Detling | 2257 | 941 | 1928 | 1962 |
| Orlin L. Ireland | 1840 | 975 | 1927 | 1970 |
| K.L. Chambers | 1814 | 894 | 1946 | 2009 |
| William H. Baker | 1677 | 831 | 1927 | 1982 |
| Thomas Howell | 1418 | 913 | 1877 | 1899 |
| W.C. Cusick | 1265 | 687 | 1880 | 1918 |
| Albert N. Steward | 1218 | 734 | 1920 | 1959 |
| E.P. Sheldon | 1167 | 632 | 1882 | 1908 |
| LeRoy E. Detling | 1094 | 600 | 1923 | 1964 |
| Helen M. Gilkey | 944 | 607 | 1904 | 1960 |
| John B. Leiberg | 900 | 651 | 1886 | 1910 |
| H.C. Gilbert | 882 | 386 | 1912 | 1961 |
| S.C. Head | 852 | 422 | 1950 | 1959 |
| Charles G. Hansen | 839 | 474 | 1950 | 1957 |
Let’s look at biases or weaknesses in sampling in three axes: spatial, temporal, and across species.
Counties vary a lot in the number of specimens, from over 9000 to less than 500. This is pretty typical for an herbarium.
Lane and Benton Counties are well represented, so are Harney, Wallowa, and Josephine.
This makes clear that some counties are badly in need of more sampling. Alternatively, one could focus on those well-sampled counties and maintain those datasets.
Species diversity at the county level looks partly tied to the number of specimens. Counties with more specimens have more diversity.
This suggests that sampling is incomplete at the county level; we’d like to see diversity be more independent of sampling.
This shows a fine-grained analysis of the number of specimens collected each year. There’s a lot of variation, as collecting has ebbed and flowed, but there’s been a reduction in the last five years or so.
Simplifying a bit, we can look at specimens by decade. The best sampling was in 1910-1950. The 2010s is the lowest decade since the 1890s.
We can also get a sense of specimens by decade by county to see if there are any spatiotemporal trends. For the most part, counties follow the broader temporal pattern. Multnomah, Klamath, Umatilla, and Marion have hardly been sampled in a long time, though.
Here, we focus on the six most-sampled counties
One way to get at data-deficient species, is to look at the number of species with very few records.
Of the 4500 species recorded, over 2000 are represented by less than 10 specimens statewide (median 12 specimens per species), and nearly 1000 species are only represented by 1-2 specimens. These would be a good target for further collecting, though it’s possible some are outdated names or something of that sort.
Here’s a breakdown of the species with 10 or fewer specimens:
| Species | Specimens |
|---|---|
| 591 | 1 |
| 344 | 2 |
| 259 | 3 |
| 170 | 4 |
| 149 | 5 |
| 150 | 6 |
| 130 | 7 |
| 130 | 8 |
| 104 | 9 |
| 98 | 10 |
Another way that we can look at sampling across species is to add a time component and look for species that have not been collected in a long time.
When we look at when each species was last collected, the median is year of last collection 1996, so half of the species have not been collected in about 25 years. There’s a significant proportion that have not been collected in over 50 years too (since 1970).
Again, these could be old names, but they could also be species that have become rare or extinct, or simply haven’t been looked for, because of inaccessibility or because of lack of taxonomic expertise.
A dataset of the species that have not been collected in 50 years, with the last year they were collected is available at oldcollections.csv
The actual number of species not collected in 50 years:
| Species_Not_Collected_In_50_Years |
|---|
| 755 |
Here’s a map of where species that have not been collected in 50 years were originally collected.
This shows temporal sampling for the 16 most common species in the herbarium.
Again, they mostly follow the broader temporal trend, with sampling falling off post-1950.
We can focus in on the three most common species. Also, I’ve included three tree species that are relatively common in the landscape for comparison.
Common trees really haven’t been sampled much. Even the really common species in the herbarium haven’t added many specimens in recent decades.
Likewise, we can look at temporal sampling for some common invasives. These invasives are not well sampled at all, regardless of decade. They’re another possible focal point of targeted collecting.
By comparison to the herbarium data, iNaturalist observations are growing exponentially.
iNaturalist has collected more observations in Oregon in each of the last four years than the best year for the herbarium, and last year it was 60,000!
R version 3.6.1 (2019-07-05)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.3 LTS
Matrix products: default
BLAS: /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.7.1
LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.7.1
locale:
[1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
[5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
[7] LC_PAPER=en_US.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] sf_0.8-0 taxize_0.9.91 knitr_1.27 cowplot_1.0.0
[5] forcats_0.4.0 stringr_1.4.0 dplyr_0.8.3 purrr_0.3.3
[9] readr_1.3.1 tidyr_1.0.0 tibble_2.1.3 ggplot2_3.2.1
[13] tidyverse_1.3.0
loaded via a namespace (and not attached):
[1] httr_1.4.1 jsonlite_1.6 foreach_1.4.7 bold_0.9.0
[5] modelr_0.1.5 assertthat_0.2.1 highr_0.8 triebeard_0.3.0
[9] urltools_1.7.3 cellranger_1.1.0 yaml_2.2.0 pillar_1.4.2
[13] backports_1.1.5 lattice_0.20-38 glue_1.3.1 digest_0.6.23
[17] rvest_0.3.5 colorspace_1.4-1 htmltools_0.4.0 plyr_1.8.4
[21] ritis_0.8.0 pkgconfig_2.0.3 httpcode_0.2.0 broom_0.5.2
[25] haven_2.2.0 scales_1.1.0 solrium_1.1.4 farver_2.0.1
[29] generics_0.0.2 withr_2.1.2 lazyeval_0.2.2 cli_1.1.0
[33] magrittr_1.5 crayon_1.3.4 readxl_1.3.1 evaluate_0.14
[37] fansi_0.4.0 fs_1.3.1 nlme_3.1-142 xml2_1.2.2
[41] class_7.3-15 tools_3.6.1 data.table_1.12.6 hms_0.5.2
[45] lifecycle_0.1.0 munsell_0.5.0 reprex_0.3.0 compiler_3.6.1
[49] e1071_1.7-3 rlang_0.4.2 classInt_0.4-2 units_0.6-5
[53] grid_3.6.1 iterators_1.0.12 rstudioapi_0.10 labeling_0.3
[57] rmarkdown_2.1 gtable_0.3.0 codetools_0.2-16 DBI_1.0.0
[61] reshape_0.8.8 curl_4.3 reshape2_1.4.3 R6_2.4.1
[65] zoo_1.8-6 lubridate_1.7.4 utf8_1.1.4 zeallot_0.1.0
[69] rprojroot_1.3-2 KernSmooth_2.23-16 ape_5.3 stringi_1.4.3
[73] parallel_3.6.1 crul_0.9.0 Rcpp_1.0.3 vctrs_0.2.0
[77] dbplyr_1.4.2 tidyselect_0.2.5 xfun_0.11