Plants are great to study since they can't be hurt like an animal, and they can't run away from your experiments. Even though they appear to grow slowly, there are heaps of ways in which you can see plants change within hours or days!
Plants can be used to investigate lots of important topics including the effects of temperature (and climate change!), light, gravity, as well as food issues.
We've picked out a few common species that we think would be great for a range of studies.
Sea lettuce – (Latin name Ulva lactuca) isn't really a plant, it's a seaweed (or green algae) that is very common on mudflats in the Dunedin Harbour. It forms blooms (like a growth spurt) in spring and autumn that can be very smelly when the sea lettuce starts to rot! It is very fast growing and will quickly consume all the nutrients, like nitrogen, from the seawater if it's able to photosynthesise in the sun.
Sea lettuce grows in sheets, which are a bit like leaves, but they are only one cell thick. Because it has a very simple structure, you can roughly expect an equivalent amount of sea lettuce (we call this 'biomass') to respond consistently to experiments. For example, if you repeated exactly the same experiment five times with 100g of fresh sea lettuce each time, you should see a very similar result each time. This isn't quite so easy to do in studies with land plants or even with other seaweeds, which vary a lot more, depending on factors like age, and health that you can't really see.
Some studies with sea lettuce
Nitrogen run-off from fertiliser on farms and in our gardens can cause lots of problems with pollution in rivers, lakes and our coastal environment. Sea lettuce can extract some of this nitrogen. You can buy testing kits online (we found some at watertest.co.nz) to measure nitrogen in the water. Try to pick sea lettuce that is green all over, as seaweed with white patches (bleaching) are generally not as healthy. If you are weighing the seaweed, shake the water off to weigh it but don't let it dry out. Use seawater and not freshwater (unless freshwater is part of your study).
Questions you could investigate:
- What happens to nitrogen uptake when you change the amount of light the sea lettuce receives?
- Does temperature affect nitrogen uptake by sea lettuce?
- If you add other essential nutrients, like phosphorus (buy it at a garden store) or carbon dioxide (add baking soda to mimic increased CO2), how does that affect nitrogen uptake?
Warning: large blooms of sea lettuce in Spring are so smelly they can actually be dangerous – the smell is a nasty gas called hydrogen sulfide (the same smell as rotting eggs) and it comes from the sea lettuce rotting. In Spring, perhaps wait for a windy day to collect your sea lettuce.
Horopito (pepper tree)
Native horopito (or pepper tree, called Pseudowintera colorata in Latin) is a very pretty red-and-green shrub that grows underneath the forest canopy, especially in damp areas and areas with lots of deer (since they won't eat it). You can find it in Woodhaugh Garden, Bethune's Gully and on the Pineapple Track, for example. It's leaves have a very distinct peppery taste due to a compound called polygodial (try adding a couple of leaves to a curry!). Along with polygodial, horopito leaves often have patches of red colour, caused by pigments called anthocyanins. Both polygodial and anthocyanins are made by the plant to make the leaves taste bad, as a defence against all kinds of herbivores, from deer to caterpillars. They produce more of these compounds when leaves are wounded, so we can test what happens when we mimic an animal feeding on the leaves.
Some studies with horopito
What happens to leaves when a herbivore attacks? Since both polygodial and anthocyanins are produced by the leaves to deter herbivores, we can test how well these compounds really work. You can mimic a caterpillar bite by piercing horopito leaves (still attached to the plant) with a needle and watching as red anthocyanins develop around the wound over several days. Or, you could see if there is any relationship between the red anthocyanins and the spicy polygodial taste (scientists aren't completely sure). You could come up with a way of measuring how red the leaves are, then come up with a way to test their spiciness – this might take a bit of thinking!
Questions you could investigate:
- What happens to horopito leaves when we mimic a herbivore attack? How much of the leaf changes? Do nearby leaves change too?
- Does redness (level of anthocyanin) reflect spiciness (level of polygodial)?
- Does redness or spiciness vary between forest areas that might have lots of mammalian herbivores (such as deer or possums), compared to areas with no mammalian herbivores? For example, compare horopito from Orokonui Ecosanctuary to horopito from the Silverpeaks (or another native forest without pest eradication).
Unfortunately, not all plants are desirable in certain places – we call these weeds! In New Zealand, tree lupin (Latin name Lupinus arboreus) is a major weed, particularly on sand dunes such as at Long Beach and St Kilda Beach (it's pretty much everywhere on Otago beaches). One reason weed species are so successful is that they able to produce large amounts of seeds that can easily germinate (grow). Lupin is one of these species – its large and very tough seeds are known to last for a long time in sand dunes, meaning that the weeds can come back even years after adult plants are removed by weeding or spraying. This makes lupin eradication very expensive and difficult!
Some studies with lupin
Lupin seeds germinate quite reliably in a week or less, if seeds are kept inside a folded, damp paper towel (check out some websites online for more advice). Soaking seeds in water (overnight at room temperature or for a week in the fridge) beforehand, a process called imbibing (or stratifying), seems to help. We can use this technique to test the germination rate (what percentage of seeds germinate) in different conditions. Lupin seeds can be found in the seed pods (mature seeds should rattle inside the pod) or by sieving sand or soil underneath the plants. You could get lupin seeds from different places, different soil depths (deeper seeds might be older) or treat them in some way, like exposing them to sea water, high (or low) temperature, or physical damage. Then you can test for yourself just how tough lupin seeds are, and how tricky they are to get rid of in the environment.
Questions you could investigate:
- How does seed age (indicated by seed depth in the soil) affect germination rate?
- Can you kill a lupin seed? Come up with creative ways to stress the seed and see if they can still grow (you may be surprised). Try a range of stress intensity (e.g. a gradient of temperature) so you still see a range of results, just in case your most extreme treatment does turn out to be 100% lethal…
- Can lupin seeds handle seawater? Extreme storm events frequently erode sand dunes where lupins are major weeds. Do you think the lupin seeds that get submerged in seawater (or ones that are even taken out to sea) can survive and grow after storms damage the dunes?
Warning: throw any lupin material (especially seeds) in to the rubbish when you are finished with them – you don't want to spread this weed any further into the environment!
Sycamore trees (called Acer pseudoplantanus in Latin) are another pest plant in Dunedin, particularly in places like the Town Belt, and shady valleys like Ross Creek, and Bethunes Gully. In parts of Otago, like Manuka Gorge, they are just about the only tree species around. Their large leaves falling in autumn create a thick carpet in the forest that prevents the growth of other species, and their distinctive winged seeds are able to travel effectively in the wind, allowing sycamores to spread rapidly. They are fast growing and are a particular problem when they block views from people's houses and prevent establishment of native plants. Although their flowers feed bees and insects, birds can't use sycamore flower nectar or eat their seeds. This means that compared to native forest, sycamore trees aren't as good for feeding our native bird species and don't make good habitat.
Some studies with sycamore
Several different kinds of study are possible with sycamores, but because the tree is deciduous, some studies are only possible at certain times of the year. For example, the seeds are present only in autumn, the flowers in spring, and for all of winter all you will see of sycamores is branches and stems. But, you can use sycamore at any time of year to investigate topics like seed dispersal, pest invasion and animal habitat selection.
Questions you could investigate:
- How recently have sycamores invaded certain parts of Dunedin or Otago? With permission from the landowner and definitely an adult to help, you could count the growth rings in the trunks of sycamore trees to try and figure out the age of sycamores and how quickly they are spreading. Make sure you plan carefully before cutting down any trees though!
- What animal species do sycamore trees support? 5-minute bird counts are often used to measure bird abundance in different habitats. Are the bird species different in sycamore forest compared to other forest types? What is the better habitat for native or introduced birds?
- How far can sycamore seeds travel? You could test seeds that have already fallen, or come up with a way to label seeds before they fall, then see how far away from the tree they are naturally able to spread.
- How many sycamore seedlings come up in spring? Sycamore seedlings are very distinctive and very abundant near adult trees – how many can you count in a certain area? How does that compare to seedlings from other species?
Warning: throw any sycamore material (especially seeds) in to the rubbish when you are finished with them – you don't want to spread this weed any further into the environment!
In general, the kinds of things you can test with plants involve a little bit of patience and careful planning of your experiment. Since each plant, seed or seaweed is an independent living organism, you can sometimes get different results even when you treat things the same way. It's the same with people – individuals are different. That's why we've got a few tips to help you study plants in ways that should give you reliable results. Some important words are highlighted in bold – these are the things that good scientists think of when designing their experiments.
First, choose a topic, and come up with a hypothesis – a hypothesis usually boils down to a statement like 'x affects y.' Let's take the following simple example: temperature affects germination rate of seeds.
Next, come up with a prediction about that hypothesis – 'increasing x will increase y.' In this case, you might predict that increasing temperature increases germination rate. It doesn't matter if you are right or wrong at this stage – that's the point of your test! But, it helps if you have some rationale behind your prediction – maybe you already know that higher temperatures increase many of the processes (like cell division ) that happen inside plant and animal cells (up to a point).
Now design your experiment to see if increasing temperature increases seed germination. Obviously you need some seeds (they could be from any easily available source, or from a plant that you are interested in). Secondly, you need to decide how to make a seed germinate, as well as how and when to measure when a seed has germinated. Then, you need to choose what temperatures to vary – all of these things are called your methods.
The point of an experiment is the result, which in this case might be what percentage of seeds at each temperature germinated after a week.
You almost always need a control, that is, something to compare the result to. If you are carrying out an experiment where you change something like temperature, try to make one of the temperatures you choose as close to 'normal' in the environment as possible. For example, if you decide a seed from a certain plant probably germinates in nature at around 8oC (roughly spring soil temperature in Dunedin), then think about testing (for example) at 8oC as your control treatment. You can then compare germination in 'normal' conditions (the 8oC control) treatment to whatever other temperature treatment you chose. This makes your results easier to discuss if you are thinking about to real-life seed growing conditions out in the environment.
Figuring out what caused the result requires you to know exactly what factors were different between each experimental treatment.
This can be simple or complicated depending on your experiment. If you have a single packet of seeds that you divide evenly between two different temperatures, but keep everything else the same (the kind of germination container, humidity, light level, etc.), then you can be sure that temperature was the only factor explaining differences between germination rate at each temperature. Controlling the factors means that the only things that are different between treatments are things that you want to be different.
If you are interested in more than one factor, for example the effect of temperature (hot vs. cold) and light (shade vs. sun) on seed germination, then you need to test both light levels at both temperatures. This will mean four groups: hot+shade, hot+sun, cold+shade, cold+sun. If you only had hot/light and cold/dark seed germination measurements, how would you know which factor caused any difference in germination rate – temperature or light?
One of the most important principles in science is the idea that results of a study should be reproducible. That means that the same experiment performed by different people, or at different times, should get roughly the same result (as long as the set-up is the same). If results are reproducible, then scientists can be reasonably sure that the results are the result of real differences (like an effect caused by temperature), rather than random chance that might occur in just one study.
The best way to achieve reproducible results is to replicate your experiment.
For example, if you want to see the effect of temperature on seed germination, you wouldn't test just one seed at each temperature – you'd use lots. In this case, since it's easy to look at lots of seeds, you might have 50 or 100 seeds growing at each temperature, and compare the average germination rate at each temperature. Comparing just one seed at each temperature might mean that the differences you see are simply the result of natural variation in each seed, rather than a temperature effect. It's hard to know how many replicates you should use in each experiment but a good rule of thumb for just about any experiment is 5 or more – but preferably have as many as you think you can handle in the time you have!
Once we have a result that we think is reproducible, how do we know the meaning of our results? We call that the conclusion. This is perhaps the hardest question in science to answer. But, if we have carried out a replicated, reproducible experiment, controlled the experimental factors properly and have a good control for comparison, we can at least answer our original question. If you found the average germination rate of a type of seed was 50% at 8oC and 80% at 15oC, then you've addressed your hypothesis and prediction. In this case, we could conclude that temperature did affect germination rate overall, and, specifically, increasing temperature did increase germination rate. Knowing what your results mean depends on the question you asked in the first place – the simpler the better.
The concepts we've discussed might have been a bit unfamiliar, or even a bit difficult to understand. If you're keen to learn a few more things that will make your scientific study look seriously professional, then read on!
Once you've completed your study, you need to think about the best way to present your results. For most science, a graph will do – but what kind? It depends on the results. For most studies there is a response variable (also called the dependent variable) - this is the thing you measured, for example seed germination rate. In just about every case, the response variable would be shown on the y-axis of a graph. Then, there is the predictor variable (also called the explanatory variable) – this is the thing that you changed, like germination temperature. This is on the x-axis of the graph. In most cases your response variable is some kind of number, so it's pretty easy to know what to do with the graph's y-axis. However, it's the predictor variable that usually tells us what kind of graph we should use to best present the results.
The predictor variable could be discrete - something that is either an A, B, or C. Here, we use a bar graph – maybe one bar is seed germination temperature A and one bar is germination temperature B. To give another example, you could measure average hair length (dependent variable, y axis) in relation to hair colour (predictor variable, x axis), where each hair colour gets a different bar on the bar graph.
Alternatively, the predictor variable could be continuous (i.e. a scale of numbers where theoretically any number is possible). Perhaps you measure a person's hair length (y) in relation to their height (x). In this case, you use a scatter plot, where each person's measurement on the graph gets a point y in relation to its point x.
Another common predictor variable might be something like time, where the response variable is measured repeatedly, perhaps over a series of days, weeks or months (again, let's use hair length as our y). Time becomes the x axis, and we use a line graph, where each data point (a measurement of hair length y at time x) is connected to the next measurement of y and x – this is useful for following things like growth rate, because you can see if the growth rate is regular (linear) or increasing at a faster rate as time progresses (exponential), for example. If you are testing multiple groups (like hair length of different individual people, or classrooms, for example), you give each group its own line on the line graph.
Other types of graph can be used too. You've probably seen pie charts, which are usually used when there is only a response variable measured. For example, you might count how many of each species of bird visit a bird feeder. Each bird species you saw would get a share on a pie chart based on what percentage of visits to the feeder were made by that species. If we didn't use multiple bird feeders, make multiple measurements, or change the food (for example), there isn't a predictor variable, and so we use a simple pie chart.
All scientific results, even when we replicate the conditions in the experiment well, have natural variation due to random factors we can't control. We call this error. If you take the average seed germination rate in three identical experiments, how different was one experiment from another? We can calculate standard deviation or standard error, and include error bars on the graphs we use to present the data. Written, this is the average, plus or minus the standard deviation (or error). This helps us see if the averages are really different, or just so variable that it's possible for them to look different on a given day. For example, if the germination rate of seeds in three separate experiments was 50% +/- 20% at 8oC and 80% +/- 25% at 15oC, we probably shouldn't say that temperature had an effect, since there is a lot of overlap between the two averages when we take error into account.
As described above, let's carry out a simple, pretend experiment about seed germination and temperature. We'll use all the same concepts discussed above. Let's say we want to know about the possible effect of climate change (increasing temperature) on a plant whose seeds germinate in spring (let's use sycamore trees, which are an annoying pest tree in Dunedin's Town Belt that produces lots of seeds that come up annually each September).
We hypothesise that temperature affects germination rate. We predict that increasing temperature will increase germination rate, since increasing temperatures normally increases processes that occur within plant and animal cells (the science behind this is complicated, but this is normally true).
We already know sycamore seeds germinate in Spring, and soil temperature in Spring in Dunedin is about 8oC (you can find this sort of information online). So for our methods, we will germinate some seeds at 8oC, and that will be our control treatment. Then we decide on a higher temperature – this is where you get to choose, but try to be realistic. Let's say 15oC – high enough above normal to (probably) see a result, but probably not high enough to cook the seeds. Of course, you could choose as many temperatures as you want, but it depends how much work you want.
So, we take 200 sycamore seeds, and divide 100 each into the 8oC and 15oC group. We read on the internet or in a book how to get the seeds to germinate (let's say we're putting them on a damp paper towel in an ice cream container in a dark cupboard). We use the same kind of paper towel, container, and even cupboard, if we can – we're trying to make as much as we possibly can exactly the same, except for temperature. How you control the temperature depends on what materials you have – maybe you've got a spare fridge or a cellar that you can control. These are also part of our methods.
Let's say it normally takes a week for sycamore seeds to germinate. Decide before the experiment what counts as germination – is it the first root (called the radicle) coming out? Is it the first leaf? It's up to you – as long as you're consistent. Let's say you count the number of radicles visible in each treatment after one week – then – simple – you've got a result! Since we used 100 seeds, the number of seeds at each temperature with a radicle is the same as the percentage of seeds that germinated. We present a bar graph, where germination rate is the y axis, and the rate at each temperature is a bar on the x axis.
In our pretend experiment, 50 seeds germinated after a week at 8oC and 80 seeds at 15oC. In this case, we can conclude that there is indeed an effect of temperature, and that increasing temperature increased germination – addressing our hypothesis and prediction. We could then speculate, based on this result, that sycamore seeds might germinate more in Dunedin if temperatures increase. But, be careful not to draw too many conclusions from one study – think more about what your data actually shows rather than what it might show.
What do judges look for in a good science fair project?
Science, really, is the process of answering questions. By far the most important thing to do in a science fair project then, is to ask a good, simple question, come up with a way to answer it, and then discuss how (or if) your results answered the question. That means showing a clear hypothesis and prediction, having clearly presented methods and results, and then a conclusion that makes sense based on your data.
Having a great idea for a project in the first place definitely helps. Having good pictures, good layout or a catchy title will also add points, but don't forget it's the science that's the most important thing! Don't be worried if your experimental results didn't show anything new or exciting – even if you carried out a good experiment then no results (or results you think are boring) are perfectly possible. It's only a few studies out of thousands that ever make huge breakthroughs or get the Nobel Prize, after all.