02/11/15

The last big push – wrapping up the phenotyping for AdapTree

One last experiment of the AdapTree project remained in the ground: the trial comparing natural and seed orchard seed lots. The roughly 6000 plants were kept for a third season to collect periodic height measurements for the pine, and cold hardiness data for the spruce. The spruce trees announced themselves ready for freeze testing by the end of August, so September was dedicated to needle chopping and conductivity measurements. Thousands of them. Three whole weeks. Those who stuck it out were by then thinking with fond memories of the days gone by when the AdapTree team was large and fresh. But we did it, and the resulting data were clean.

Fig 1: Pine heights, diameters, and harvesting for shoot dry mass

Fig. 1: Pine heights, diameters, data logging and harvesting for shoot dry mass

Nonetheless, it was the middle of October by the time we were performing the final height and diameter measurements on the pine, while simultaneously harvesting them for shoot dry weights. (Figure 1). Harvest time was preferred for diameter measurements because it gives us easy access to the stems. We had sun, we had fog, we had beautiful autumn days and we had rain. Now, we don’t exactly melt from a little rain. But shoving wet plants in wet paper bags which are marked with sticky labels of moderate stickiness is asking for trouble. And while the recording tablets are protected, raindrops beading on a screen are not ideal for visibility. So this became a long, drawn-out affair. Of course, we only take pictures on the nicest days!

Fig. 2: Gradual progress.

Fig. 2: Gradual progress

The ideal team consisted of three people, so we organized our schedules, waited for the rain to stop, and gradually made progress (Figure 2). We finished the pine and thought we’d just continue at the same speed with the spruce. (Figure 3). Not so. While the spruce plants were much smaller than the pine, the stems were thick and asymmetric, with multiple roots spreading horizontally almost before touching the ground. Individual diameter measurements were not very repeatable, so multiple measurements were taken. Rather than taking turns at measuring, recording and bagging, for consistency’s sake all twelve blocks were measured by the same person. The repetitive bending over proved fatiguing.

03 collage

Fig. 3 : Some large pine, and the first snowfall

If you wonder what I am holding up: it’s an old toothbrush, to clean the root collar of sand and mud before taking a diameter measurement.

Then we had frost and snow in November! And the frozen ground in combination with the frozen liverworts and moss did not make a good basis for reliable height measurements – who’d have thought frozen bryophytes could bring science to a halt? As the days got shorter and shorter, we completed the heights separately.

04 collage

Fig. 4: Dry weights – so many!

In the meantime, shoot dry weights were measured in the lab, to keep the accumulating boxes of dried plants under control. (Figure 4). We didn’t finish before Christmas as planned, and Ian had to wait and wait for his data. With mild weather in the first week of January, we made one last big push, and the last spruce tree was cut on January 9 (Figure 5), with the last dry weights gathered two weeks later.

05 collage

Fig. 5 : The last spruce tree is cut …

And now all the plants are gone and no more measurements can possibly be collected from them!!!

I am grateful to all those who helped out (you can see some of them in the pictures!), and for the hot tub in the swimming pool. Did I mention the guys who (re)built the raised beds in the first place (Figure 6) ? And the summer students who helped sowing, and the technicians and students who collected weekly height measurements (Figure 7) ?

 

Fig. 6 : Re-building the beds ...

Fig. 6 : Re-building the beds …

 

Fig. 7 : Sowing, observing, measuring, ...

Fig. 7 : Sowing, observing, DNA collection, measuring, …

 

09/24/14

Science storytelling

In August, I was privileged to attend a storytelling workshop with Denise Withers, as part of the genomics entrepreneurship program. Denise is convinced that every research proposal can be boiled down to a 3 minute sales pitch that your neighbor can understand, though it may take a lot of hard work. Denise spends her time coaching scientists and entrepreneurs to do just that.

These days, it isn’t enough to have excellent science in a proposal for funding or an evaluation report. Eventually, administrators and politicians also need to be convinced that we deserve to get a slice of their scarce financial resources. Research proposals increasingly need to demonstrate impact. This is some steps removed from direct project outputs, with all the risk and uncertainty involved, so scientists are understandably hesitant to ‘promise the moon’, while at the same time feeling the pressure to do just that. The funding agency may then appear be “playing Civilization”: starting at level  3 and investing 10 units in science should automatically result in reaching level 4. But real life is not like that, is it?

Science is characterized by the unexpected, so rather than a problem, this hurdle allows the characters to shine … if we tell the story well. Seeing the story in our research work requires zooming out from the detailed focus in which we spend our daily lives. Developing a logic model and a theory of change for the project are steps which can help us develop the overview. Doing it as a team may enhance mutual understanding of the various project activities, how they are connected, and any unstated assumptions. It can be built into the funding proposal to provide ongoing evaluation. And towards the projects’ end, we scientists turn into heroes having overcome insurmountable obstacles, to reach places where we perhaps didn’t expect to get to, and having changed the world just a little bit… And this, more than the 100 page report and the 10 papers, is what evaluators (and your neighbor) will remember. Although the 100 pages and the 10 papers are still needed.

Example: A story of the AdapTree Project in three minutes, based on logic models and theory of change.

Due to climate change, the seed sources being planted in our forest are becoming increasingly mismatched to the climates they will experience. This mismatch can be addressed by moving seed over the landscape where needed. This requires adjusting the rules of what can be planted where. However, we must first know which seed sources are best adapted to given sites, both now and in the future.

We reduced the scope of this problem to the two provinces of BC and AB and two important commercial species, interior spruce and lodgepole pine. However, we still can’t plant every seed source on every planting site and wait thirty years for the results. Instead, we looked at the underlying patterns of adaptation in the genome of these species.

Because of the enormous size of conifer genomes, we zoomed in on >25,000 candidate genes. These genes were then sequenced in 500 natural seedlots from all over to find genotypic variation at the base-pair level. The adaptive value was revealed by growing plants of the same seedlots in controlled environments simulating present and future climates.

We then developed a SNP chip to evaluate adaptive genetic potential in orchard seed lots, whose top quality seed is used by preference in tree planting. Using that, we can evaluate the ecological risks of moving seed sources over the landscape or ‘assisted migration’, including the status quo. This allows policy makers to update seed transfer guidelines as climate changes. A socio-economic research component ensures that stakeholders can make informed decisions regarding new strategies.

The main outcome of AdapTree is information about the adaptive potential of seed sources regarding present and near-future climates.  The resulting impact, long-term, would be reduced plantation failures, increased forest productivity, and resilient forest-dependent communities. A short term outcome is the development of sampling and research methods that are transferable to other seed sources, species and climates.

 

06/8/14

Simulating climates in growth chambers – The AdapTree project

This post is part of the series Simulating Climates in Growth Chambers.

The AdapTree project evaluates the genetic and phenoptyic variation of two commercially important conifers in western Canada, interior spruce (Picea glauca, P. engelmannii, and their natural hybrids) and lodgepole pine (Pinus contorta). More than 580 seed sources were grown under controlled climate conditions to quantify genetic diversity and geographic structure for adaptive traits such as phenology, frost hardiness, seedling growth, and response to drought and heat. Concurrently, sequence capture and resequencing of much of the exome for ~600 individuals of each species reveals genetic variation, some of which is associated with this adaptation. Around 5,000 additional individuals per species growing in various other controlled climate regimes and outdoor common gardens will then be genotyped using a cheaper SNP array. All of these markers will be tested for a potential role in local adaptation to climate through 1) association with climate-relevant phenotypes; and 2) gene-environment correlations. The SNPs with evidence of local adaptation will then be used to evaluate the suitability of populations to future climates. Field-based validation studies have been established to confirm the genomic results.

Location of the target simulated climates on a map.

Location of the target simulated climates on a map.

To bring out the differences in adaptive characteristics of populations from all over British Columbia and Alberta, three temperature regimes were developed to represent four different climates with mean annual temperatures (MAT) of 1, 6 and 11 °C (all well watered), as well as an MAT 11 dry climate.
Realistic climates were needed to yield realistic bud break and bud set data in growth chambers. The photoperiod regime was identical for all plants and day length corresponds to that at 54.5°N at the relevant time of the season. Time constraints on the project necessitated germination modification and growing season compaction to reach the desired plant sizes quickly for all experiments. Not all experiments could be grown under fully controlled conditions, and some of the plants grown in the greenhouse were out of sync with nature yet needed to be planted outdoors in the following season. Their blackout regime gave us trouble, since covering them up to keep light out caused conditions ideal for fungal growth. During the second growing season, simulated drought was applied in cycles to the MAT11 dry treatment. Half of the plants in MAT11 wet and half of the plants in MAT11 dry were also subjected to a heat wave in the middle of summer. Plants were well-watered and chlorophyll fluorescence was measured to evaluate plant stress as a consequence of the heat wave. Carbon isotope composition was used to evaluate plant response to drought integrated over the growing season. At the end of the second growing season, cold hardiness measurements required us to simulate real winter, and not just a chilling period. The night frost pre-treatments were successful and good cold hardiness data were obtained. After this, the plants could be destructively sampled for dry weights. This required washing the soil off the roots. Thanks to our foresight in using plant cones during the first season (see root washing) we were able to separate the roots even for the largest (MAT 6 and 11) plants and after two seasons of growth.

Further information about the AdapTree project can be found on the website.