Exploring dimensions of biodiversity in Japanese ferns

class: inverse, middle, title-slide, big-margin
background-color: #000000

.pull-left-wide[
# Exploring dimensions of biodiversity in Japanese ferns
 
## Joel H. Nitta1, Brent D. Misher2, Wataru Iwasaki1, Atsushi Ebihara3

1University of Tokyo, Japan

2University of California, Berkeley, USA

3National Museum of Nature and Science, Japan

https://joelnitta.com
]

.pull-right-narrow[
![](images/title_cropped.png)
]

---
class: big-margin

## Biodiversity: more than just counting species

Species richness is only one dimension of biodiversity!

We should also consider

- Phylogenetic diversity

- Functional diversity

- How these are related to each other

---
class: big-margin

## Biodiversity: more than just counting species

Species richness is only one dimension of biodiversity!

We should also consider

- Phylogenetic diversity

- Functional diversity

- How these are related to each other

**This requires comprehensive sampling of taxa, DNA, and traits**

---
class: big-margin

## Ferns of Japan: an ideal system for studying biodiversity

.pull-left[

- Diverse flora

- ca. 675 spp.

- 98 genera

- 34 families

- One of the best-collected fern floras in the world

- Data available for DNA, distribution, and traits
]

.pull-right[
.center[
<img src="images/ja_example_ferns.png" alt="drawing" height="580"/>
]
]

.footnote[.small[Photos: A. Ebihara]]

---
class: big-margin

## Ferns of Japan: an ideal system for studying biodiversity

.pull-left[

- Diverse flora

- ca. 675 spp.

- 98 genera

- 34 families

- One of the best-collected fern floras in the world

- Data available for DNA, distribution, and traits
]

.pull-right[
![](index_files/figure-html/gbif-colls-per-country-1.svg)
]

---
## A (very brief) introduction to the biogeography of Japan

.pull-left-narrow[
- Wide variation in climate from north (subarctic) to south (subtropical)

- Wide variation in elevation

- Tertiary relict flora (never glaciated)

- Main islands continental, southern islands oceanic
]

.pull-right-wide[
<img src="images/japan_intro.png" alt="drawing" height="580"/>
]

---
class: big-margin

## Goals of this study

Using the **ferns of Japan** as a study system, seek to understand:

1. How is biodiversity distributed? 
 
2. How is biodiversity structured? 
 
3. How well is biodiversity protected?

---
class: big-margin
## Data sources

**Occurrence data**

- Specimens at herbaria in Japan

- 300,169 specimens (674 taxa)

- Convert to 0.2° (ca. 20 km x 20 km) presence-absence matrix

---
class: big-margin
## Data sources

**Phylogenetic tree**

- *rbcL* sequences from Japanese specimens (Ebihara and Nitta, 2019)

- Combine with broad sampling from GenBank

- Infer phylogeny with IQ-TREE (Nguyen et al., 2015)

- Date with 27 fossil calibration points (Testo and Sundue, 2016) using treePL (Smith and O'Meara, 2012)

- Subset to only Japanese taxa

---
class: big-margin

## Data sources

**Traits**

- 16 morphological traits mostly used for ID (Ebihara and Nitta, 2019)

- 3 binary (e.g., false indusium)

- 3 continuous (e.g., stipe length)

- 10 qualitative (e.g., indusium shape, lamina texture)
  
---
class: middle, inverse, big-margin

# How is biodiversity distributed?

---
## Species richness is humped-shaped

.footnote[Trendline: General additive model (GAM) with cubic splines]

---
class: big-margin
## Measuring diversity with a tree

.center[
<img src="images/pd.png" alt="drawing" height="250"/>
]

**P**hylogenetic **D**iversity: Total unique branch length (Faith, 1992)

**F**unctional **D**iversity: PD, but measured with a trait dendrogram (Petchey and Gaston, 2002)

---
## PD and FD are correlated with richness

---
## PD and FD are correlated with richness... as expected

.footnote[Trendline: GAM with log-transform]
---
class: big-margin
## Compare raw values to random expectation with **S**tandard **E**ffect **S**ize

.large[
.center[
`\(SES = \frac{obs - mean(null)}{sd(null)}\)`
]
]

SES = 0: not different from random

SES > 0: more diverse than expected

SES < 0: less diverse than expected

Null distribution: 999 communities randomized using "Independent Swap" (Gotelli, 2000)

---
## SES of PD is mostly clustered throughout

.footnote[Trendline: GAM with cubic splines]
---
background-position: 40% 50%
background-image: url(images/RPD.png)
background-size: contain

## Compare branch lengths with **R**elative **P**hylogenetic **D**iversity

.pull-right[
Long branches

- Refugia

- Occurrence of a few members of clades that mainly occur outside of the study region

Short branches

- Recent speciation

.small[Mishler et al. (2014) ]
]

---
## RPD is high in the north and south

.footnote[Trendline: GAM with cubic splines]
---
## Southern islands have high RFD

(RFD = functional analog of RPD)

.footnote[Trendline: GAM with cubic splines]
---
## Southern islands are hotspots of endemism

Phylogenetic endemism = PD weighted by range size (Rosauer et al., 2009)

.footnote[Trendline: GAM with cubic splines]
---
class: middle, inverse, big-margin

# How is biodiversity structured?

---
## Classify bioregions using clustering

.center[
<img src="images/daru_2017.png" alt="drawing" height="400"/>
]

(can also use taxonomic distances)

.footnote[Daru, Elliott, Park, and Davies (2017); Daru, Farooq, Antonelli, and Faurby (2020)]

---
## Most sites fall into four bioregions

---
background-position: 50% 50%
background-image: url(images/species-background-1.png)
background-size: contain

## Most sites fall into four bioregions

.footnote[]

---
background-position: 50% 50%
background-image: url(images/species-background-2.png)
background-size: contain

## Most sites fall into four bioregions

.footnote[]

---
background-position: 50% 50%
background-image: url(images/species-background-3.png)
background-size: contain

## Most sites fall into four bioregions

.footnote[]

---
background-position: 50% 50%
background-image: url(images/species-background-4.png)
background-size: contain

## Most sites fall into four bioregions

.footnote[]

---
## Taxonomic and phylogenetic bioregions are similar

---
## Bioregions mostly match traditionally-defined forest types

.pull-left[
Taxonomic bioregions

![](index_files/figure-html/tax-bioregion-forest-1.svg)
]

.pull-right[
Forest types 
.small[Shimizu (2014) ]
<img src="images/forest_types_eng.png" alt="drawing" width="280"/>
]

---
class: middle, inverse, big-margin

# How well is biodiversity protected?

---
## Japan is well-protected overall (ca. 30% coverage)

---
## Hotspots of taxonomic richness are well-protected

---
## Hotspots of phylogenetic endemism are also well-protected

---
class: big-margin
## A growing threat: herbivory by deer (*Cervus nippon*)

.center[
<img src="images/deer_damage_1.png" alt="drawing" height="450"/>
]

---
class: big-margin
## A growing threat: herbivory by deer (*Cervus nippon*)

.center[
<img src="images/deer_damage_2.png" alt="drawing" height="500"/>
]

---
class: big-margin

# Conclusions

- How is biodiversity distributed?

- How is biodiversity structured?

- How well is biodiversity protected?

---
class: big-margin

# Conclusions

.pull-left[
How is biodiversity distributed?

- Taxonomic diversity highest in Kyushu

- Phylogenetic and functional diversity, phylogenetic endemism highest in southern islands
]

.pull-right[
<img src="index_files/figure-html/conclusions-richness-1.svg" width="450px" />
]

---
class: big-margin

# Conclusions

.pull-left[
How is biodiversity structured?

- Into four major clusters that track major forest types
]

.pull-right[
![](index_files/figure-html/conclusions-regions-1.svg)
]

---
class: big-margin

# Conclusions
.pull-left[
How well is biodiversity protected?

- Well-protected in terms of status

- May be in danger due to herbivory by deer
]

.pull-right[
![](index_files/figure-html/conclusions-protection-1.svg)
]
 
---
# Future directions

.pull-left[
 
- Understand the role that hybridization has on diversification (ca. 300 described hybrids!)

- Clarify the role of fern gametophytes in community assembly at broad scales
]

.pull-right[
<img src="images/vandenboschia.png" alt="drawing" width="280"/>

<img src="images/IMGP5729.jpeg" alt="drawing" width="300"/> .tiny[Photo: Joel Nitta]
]

---
class:: big-margin

# Acknowledgements

.center[
<img src="images/pags-logo.png" alt="drawing" width="600"/>

Members of the Iwasaki Lab (The University of Tokyo)

Eric Schuettpelz (Smithsonian Institution)

Alexander White (Smithsonian Institution)
]

---
## References

.small[
Daru, B. H, T. L. Elliott, D. S. Park, et al. (2017). "Understanding
the processes underpinning patterns of phylogenetic regionalization".
In: _Trends in Ecology and Evolution_ 32, pp. 845-860. DOI:
10.1016/j.tree.2017.08.013.

Daru, B. H, H. Farooq, A. Antonelli, et al. (2020). "Endemism patterns
are scale dependent". In: _Nature Communications_ 11, p. 2115. DOI:
10.1038/s41467-020-15921-6.

Ebihara, A. (2016). _The Standard of Ferns and Lycophytes in Japan,
Vol. 1_. Tokyo: Gakken Plus.

Ebihara, A. (2017). _The Standard of Ferns and Lycophytes in Japan,
Vol. 2_. Tokyo: Gakken Plus.

Ebihara, A, S. Matsumoto, and M. Ito (2009). "Taxonomy of the
reticulate Vandenboschia radicans complex (Hymenophyllaceae) in Japan".
In: _Acta Phytotaxonomica et Geobotanica_ 60, pp. 26-40. DOI:
10.18942/apg.KJ00005576220.

Ebihara, A. and J. H. Nitta (2019). "An update and reassessment of fern
and lycophyte diversity data in the Japanese Archipelago". In: _Journal
of Plant Research_ 132, pp. 723-738. DOI: 10.1007/s10265-019-01137-3.

Faith, D. P. (1992). "Conservation evaluation and phylogenetic
diversity". In: _Biological Conservation_ 61, pp. 1-10. DOI:
doi.org/10.1016/0006-3207(92)91201-3.

Gotelli, N. J. (2000). "Null model analysis of species co-occurrence
patterns". In: _Ecology_ 81.9, pp. 2606-2621. DOI: 10.2307/177478.
]

---
## References

.small[
Mishler, B. D, N. Knerr, C. E. González-Orozco, et al. (2014).
"Phylogenetic measures of biodiversity and neo- and paleo-endemism in
Australian Acacia". In: _Nature Communications_ 5, p. 4473. DOI:
10.1038/ncomms5473.

Nguyen, L. T, H. A. Schmidt, A. Von Haeseler, et al. (2015). "IQ-TREE:
A fast and effective stochastic algorithm for estimating
maximum-likelihood phylogenies". In: _Molecular Biology and Evolution_
32, pp. 268-274. DOI: 10.1093/molbev/msu300.

Petchey, O. L. and K. J. Gaston (2002). "Functional diversity (FD),
species richness and community composition". In: _Ecology Letters_ 5,
pp. 402-411. DOI: 10.1046/j.1461-0248.2002.00339.x.

Rosauer, D. F, S. W. Laffan, M. Crisp, et al. (2009). "Phylogenetic
endemism: a new approach for identifying geographical concentrations of
evolutionary history". In: _Molecular Ecology_ 18, pp. 4061-4072. DOI:
10.1111/j.1365-294X.2009.04311.x.

Shimizu, Y. (2014). "Process of the formation of Japanese forest and
typification of vegetation zone: From an East Asian viewpoint". In:
_地域学研究_, pp. 19-75.

Smith, S. A. and B. C. O'Meara (2012). "treePL: divergence time
estimation using penalized likelihood for large phylogenies". In:
_Bioinformatics_ 28, pp. 2689-2690. DOI: 10.1093/bioinformatics/bts492.

Testo, W. L. and M. A. Sundue (2016). "A 4000-species dataset provides
new insight into the evolution of ferns". In: _Molecular Phylogenetics
and Evolution_ 105, pp. 200-211. DOI: 10.1016/j.ympev.2016.09.003.
]